Quartz: Deposits, Mineralogy and Analytics
Quartz: Deposits, Mineralogy and Analytics
Quartz: Deposits, Mineralogy and Analytics
Editors
Quartz: Deposits,
Mineralogy and Analytics
123
Editors
Jens Gtze
TU Bergakademie Freiberg
Institute of Mineralogy
Brennhausgasse 14
09596 Freiberg
Germany
ISBN 978-3-642-22160-6
DOI 10.1007/978-3-642-22161-3
Robert Mckel
TU Bergakademie Freiberg
Institute of Mineralogy
Brennhausgasse 14
09596 Freiberg
Germany
ISBN 978-3-642-22161-3
(electronic)
Preface
vi
Preface
The editors of this book are highly appreciated because this book represents a
fruitful international collaboration between scientists from Australia, Brazil,
Canada, Germany, Norway, Switzerland and the USA.
Jens Gtze
Robert Mckel
Contents
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
1
2
2
2
3
12
12
13
16
17
19
19
20
21
22
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
29
29
30
31
33
33
34
34
35
35
37
39
39
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
vii
viii
Contents
2.6.3
Analysis of Fluid (Liquid and Gaseous) Inclusions .
High Purity Quartz Processing . . . . . . . . . . . . . . . . . . . . .
2.7.1
Pre-processing . . . . . . . . . . . . . . . . . . . . . . . . . .
2.7.2
Physical Processing. . . . . . . . . . . . . . . . . . . . . . .
2.7.3
Chemical Treatment . . . . . . . . . . . . . . . . . . . . . .
2.7.4
Thermal Treatment . . . . . . . . . . . . . . . . . . . . . . .
2.8 Conclusion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
42
45
45
47
48
49
49
50
.
.
.
.
.
.
.
.
53
53
53
54
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
55
55
56
59
60
62
63
64
65
66
67
69
69
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
71
72
73
73
77
81
82
84
...
84
...
85
2.7
.
.
.
.
.
.
.
.
Contents
4.5
ix
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
85
85
96
100
103
106
108
111
113
119
120
122
123
123
123
123
124
125
126
126
127
128
129
134
134
135
.
.
.
.
.
.
.
.
139
140
141
141
145
147
149
149
.
.
150
152
Contents
156
156
.
.
.
.
.
.
.
.
.
.
.
.
.
.
161
161
165
167
167
168
172
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
177
178
181
182
187
188
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
191
192
193
196
196
197
206
210
214
215
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
219
220
221
221
223
225
228
229
231
231
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Contents
xi
10.8 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
10.9 Conclusion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
11 Cathodoluminescence Microanalysis of the Defect
Microstructures of Bulk and Nanoscale Ultrapure
Silicon Dioxide Polymorphs for Device Applications . .
11.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . .
11.2 Materials and Methods . . . . . . . . . . . . . . . . . . . .
11.3 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
11.4 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
11.5 Bulk Single Crystal a-SiO2 (Quartz) . . . . . . . . . . .
11.6 Bulk Amorphous a-SiO2 . . . . . . . . . . . . . . . . . . .
11.7 Dry Thermal Oxide (Amorphous SiO2) on Si (001)
11.8 Buried Amorphous SiO2 in Si (001) . . . . . . . . . . .
11.9 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . .
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
232
234
234
.
.
.
.
.
.
.
.
.
.
.
237
238
239
242
246
247
248
254
256
260
262
.
.
.
.
.
.
.
.
265
266
269
269
269
270
270
272
277
277
.
.
279
280
.
.
.
280
282
283
.
.
.
.
287
288
289
291
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
xii
Contents
13.3.1
13.3.2
13.3.3
13.3.4
Mineralogy . . . . . . . . . . . . . . .
Cathodoluminescence . . . . . . . .
Geochemistry . . . . . . . . . . . . . .
Sources of Silica and Conditions
of Quartz Formation . . . . . . . . .
13.4 Conclusions . . . . . . . . . . . . . . . . . . . . .
References . . . . . . . . . . . . . . . . . . . . . . . . . .
...............
...............
...............
291
293
296
...............
...............
...............
299
302
303
.....
.....
.....
.....
.....
.....
.....
.....
Quartz
307
308
309
313
313
317
317
317
318
.....
320
.....
.....
.....
323
325
325
345
347
350
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
357
331
332
333
337
341
Contributors
Sandra B. Barreto Departamento de Geologia, Universidade Federal do Pernambuco (UFPE), Recife, PE 50740-530, Brazil
Mario L. S. C. Chaves Departamento de Geologia, IGC, Universidade Federal de
Minas Gerais (UFMG), Belo Horizonte, MG 30123-970, Brazil
Florian Eichinger Hydroisotop GmbH, Woelkestrasse 9, 85301 Schweitenkirchen, Germany, e-mail: fe@hydroisotop.de
Enchbat Dombon Institute of Mineralogy, TU Bergakademie Freiberg, Brennhausgasse 14, 09596 Freiberg, Germany; Department of Sciences, Technology,
and Innovation, Mongolian University of Sciences and Technology, P.O. Box 520,
Ulaanbatar 210646, Mongolia
Belinda Flem Geological Survey of Norway, Leiv Eirikssons vei 39, 7491
Trondheim, Norway, e-mail: Belinda.Flem@NGU.NO
Thomas Gtte Institute of Geosciences, Goethe-University Frankfurt, Altenhferallee 1, 60438 Frankfurt, Germany, e-mail: goette@em.uni-frankfurt.de
Jens Gtze Institute of Mineralogy, TU Bergakademie Freiberg, Brennhausgasse
14, 09596 Freiberg, Germany, e-mail: goetze@mineral.tu-freiberg.de
Reiner Haus Dorfner Analysenzentrum und Anlagenplanungsgesellschaft mbH
(ANZAPLAN), Scharhof 1, 92242 Hirschau Germany, e-mail: reiner.haus@
dorfner.com
Peter M. Ihlen Geological Survey of Norway, Postboks 6315 Sluppen, 7491
Trondheim, Norway
Ulf Kempe Institute of Mineralogy, TU Bergakademie Freiberg, Brennhausgasse
14, 09596 Freiberg, Germany, e-mail: kempe@mineral.tu.freiberg.de
Klaus Krambrock Departamento de Fsica, ICEx, Universidade Federal de Minas
Gerais (UFMG), CP 702, Belo Horizonte, MG 30123-970, Brazil
xiii
xiv
Contributors
Matthias R. Krbetschek Institute for Applied Physics, TU Bergakademie Freiberg, Leipziger Str. 23, 09596 Freiberg, Germany, e-mail: quatmi@mailserver.tufreiberg.de
Andreas Kronz Geowissenschaftliches Zentrum, Universitt Gttingen, Gttingen,
Germany, e-mail: akronz@gwdg.de
Zucheng Li Department of Geological Sciences, University of Saskatchewan,
Saskatoon, SK S7N5E2, Canada
Messias G. de Menezes Departamento de Geologia, Escola de Minas, Universidade Federal de Ouro Preto (UFOP), Ouro Preto, MG 31400-000, Brazil
Thomas Monecke Institute of Mineralogy, TU Bergakademie Freiberg, Brennhausgasse 14, 09596 Freiberg, Germany; Department of Geology and Geological
Engineering, Colorado School of Mines, 1516 Illinois Street, Golden, CO 80401,
USA
Giulio Morteani Hydroisotop GmbH, Woelkestrasse 9, 85301 Schweitenkirchen,
Germany; Departamento de Geologia, IGC, Universidade Federal de Minas Gerais
(UFMG), Belo Horizonte, MG 30123-970, Brazil, e-mail: gmorteani@gmx.de
Axel Mller Geological Survey of Norway, Leiv Eirikssons vei 39, Sluppen,
6315, 7491 Trondheim, Norway, e-mail: axel.muller@ngu.no
Yuanming Pan Department of Geological Sciences, University of Saskatchewan,
Saskatoon, SK S7N5E2, Canada, e-mail: yuanming.pan@usask.ca
Maurcio V. B. Pinheiro Departamento de Fsica, ICEx, Universidade Federal de
Minas Gerais (UFMG), CP 702, Belo Horizonte, MG 30123-970, Brazil
Michael Pltze ClayLab, Institute for Geotechnical Engineering, ETH Zurich,
8093 Zurich, Switzerland, e-mail: michael.ploetze@igt.baug.ethz.ch
Mikhail Poutivtsev Institute of Mineralogy, TU Bergakademie Freiberg, Brennhausgasse 14, 09596 Freiberg, Germany; Faculty of Physics, TU Munich, MaierLeibnitz-Laboratory, James-Franck-Strasse, 85748 Garching, Germany
Christoph Priess Dorfner Analysenzentrum und Anlagenplanungsgesellschaft
mbH (ANZAPLAN), Scharhof 1, 92242 Hirschau, Germany
Sebastian Prinz Dorfner Analysenzentrum und Anlagenplanungsgesellschaft
mbH (ANZAPLAN), Scharhof 1, 92242 Hirschau, Germany
Karl Ramseyer Institute of Geological Sciences, University Bern, Baltzerstr.
1+3, 3012 Bern, Switzerland
Brian Rusk Economic Geology Research Unit, School of Earth and Environmental Sciences, James Cook University, Townsville, QLD 4811, Australia,
e-mail: brain.rusk@jcu.edu.au
Contributors
xv
Chapter 1
Abstract The various modifications of silica, especially quartz, play a central role
in the composition of geological materials. Owing to their abundance and properties, SiO2 minerals and rocks have been used since the beginning of human being
in different applications such as tools, weaponries, jewelleries or building materials. In modern industries, silica minerals are widely used as raw materials in
high-tech applications (e.g. piezo quartz, optical devices, solar industry) or as mass
products (e.g. glass sands, refractory industry, foundry industry, etc.). The
occurrence of different silica minerals and the similarity in mineral composition of
SiO2 rocks require a clear terminology and nomenclature of silica polymorphs and
varieties as well as the different types of SiO2 rocks and their genesis. The
knowledge about the development of specific properties (typomorphic properties)
in dependence on the specific conditions of formation can be used both for the
reconstruction of geological processes and for specific technical applications. The
present work gives an overview about the state of the art of the mineralogical and
petrographical nomenclature of SiO2 minerals and rocks, the analytical approach
for the identification and classification of these materials, and their industrial
applications.
J. Gtze (&)
Institute of Mineralogy, TU Bergakademie Freiberg,
Brennhausgasse 14, 09596 Freiberg, Germany
e-mail: goetze@mineral.tu-freiberg.de
J. Gtze
1.1 Introduction
SiO2 minerals and rocks play an important role in geological processes and industrial
applications. Quartz is not only the most important silica mineral, it is abundant in the
Earth crust in igneous, metamorphic and sedimentary rocks. SiO2 minerals and rocks
have been formed by primary and secondary magmatic, hydrothermal or sedimentary processes or during diagenesis and metamorphosis (Heaney et al. 1994). Results
of these processes are pegmatite or hydrothermal quartz, quartz sands and sandstones, chert, flint or quartzite. The diagenesis of organic silica skeletons, e.g., diatoms, radiolaria, siliceous sponges, may result in the formation of siliceous rocks
such as porcellanites, diatomites or radiolarites (Fchtbauer 1988).
Owing to their abundance and properties, SiO2 minerals and rocks have been
used in different applications (e.g. tools, weaponries, jewellery, etc.) since early
human being and as traditional building materials (e.g. sandstones) worldwide for
centuries. Natural silica raw materials represent a wide group of industrial rocks
and minerals with interesting chemical and physical properties. Both single
crystals of quartz and polycrystalline material or silica rocks are being used in the
industry. The application of SiO2 materials is widespread including the use of
idiomorphic natural quartz crystals or mass products of SiO2 rocks, e.g. as highpurity quartz sands, refractory materials or silicon ore. Because of the increasing
requirements concerning the quality of silica raw materials, synthetic quartz
crystals or SiO2 materials are necessary for certain highly advanced applications
(Blankenburg et al. 1994).
Silica materials, in particular quartz, are characterized by specific properties
(e.g., crystal shape, colour, trace element and isotopic composition, luminescence
properties)ranging from point defects to macroscopic appearancewhich are
dependent on the geological history and specific conditions of formation. The
knowledge of the interrelation between genetic conditions and such properties can
be used both for the reconstruction of geological processes and for specific
technical applications (Gtze 2009a).
The aim of the present work is to provide an overview of the mineralogical and
petrographical nomenclature of SiO2 minerals and rocks, the analytical techniques
for the characterization of these materials, and their geological importance and
industrial applications.
1982)
Trigonal
Hexagonal
Monoclinic
Hexagonal
Tetragonal
Cubic
Cubic
Orthorhombic
Monoclinic
Tetragonal
Monoclinic
Tetragonal
Orthorhombic
Natural silica glass
H2O-bearing, solid SiO2 gel
J. Gtze
Table 1.2 Classification of non- and microcrystalline silica phases (modified after Graetsch
1994)
Crystal
structure or
phase
Variety
Subvariety/
synonymous
name
Quartz
Microquartz
Disordered
Chalcedony
quartz
Quartzine
(often with
moganite)
Moganite
Disordered
Opal-C
cristobalite
Cristobalite/
Opal-CT
tridymite
Noncrystalline
Lussatine
Microstructure
Optical
character
Water
content
(weight-%)
Granular
Fibrous [1120]
Fibrous [0001]
Positive
Length-fast
Length-slow
\0.4
0.52
0.51
Platy (110),
Length-slow
lepidospheric
Platy (111)*
Length-fast
1.53
Lussatite
Fibrous [110]*
Platy,
Common
lepidospheric
structureless
opal
Opal-AG
Precious opal
Close packing of
Potch opal
homometric
spheres
Close packing of
heterometric
spheres
Opal-AN
Hyalite
Botryoidal crusts
and lumps
Lechatelierite Fulgurites
Vitrified tubes
Impact glass
Meteoritic silica
glass
13
Length-slow
Nearly isotropic
38
310
Play of colour,
anomalous
birefringence
Isotropic
48
48
Strain
birefringence
Isotropic
Isotropic
37
\0.3
\0.3
J. Gtze
Fig. 1.2 Schematic structure of a-quartz projected along [001] showing piezoelectric axes XX
(a modified after Beall 1994) and scheme of most common point defects (b modified after Gtze
2009a)
radical (Friebele et al. 1979; Baker and Robinson 1983; Pan et al. 2009), whereas
hydrogen excess from the H2O crystallization medium can result in the formation
of OH- centres (silanol groups) in quartz (Weil 1984).
The analysis of point defects is essential for the use of quartz and SiO2
materials in several highly-advanced technical applications. All kinds of point
defects can strongly alter the structural, electrical and optical properties resulting
in electrical instabilities in SiO2 insulator layers, absorption effects in Piezo quartz
or lowering of the quality of optical materials due to loss by dispersion and
absorption effects (e.g. formation of smoky quartz) (Blankenburg et al. 1994).
Furthermore, condensation of point defects during crystal growth or by mechanical
and thermal treatment can result in the formation of dislocations (line defects).
Such line defects can be visualized using etching techniques or X-ray topography.
J. Gtze
brownish (e.g., Ramseyer et al. 1988; Ramseyer and Mullis 1992; Gtze et al.
2001; Gtze 2009a). Radiation halos in quartz due to lattice damage by alphaparticles are also characterized by a strong 650 nm CL emission band (e.g., Owen
1988; Meunier et al. 1990; Ramseyer et al. 1988; Gtze et al. 2001; Komuro et al.
2002; Botis et al. 2005; Krickl et al. 2008).
b Fig. 1.3 Micrographs in polarized light and cathodoluminescence (CL) of different quartz
samples demonstrating the potential of CL studies in natural and synthetic quartz; a, b Polarized
lightCL pair of a hydrothermal quartz crystal from Chemnitz, Germany showing distinct
growth zoning, which is not visible in transmitted light; c CL micrograph of quartz in a rhyolite
from Euba, Germany; the CL image reveals growth zoning and syneusis of two quartz
phenocrysts, which show strong features of resorption; d CL micrograph of a sandstone sample
from Dohna, Germany; different CL colours of the detrital quartz grains point to varying source
rocks; authigenic quartz overgrowths are clearly visible under CL (see arrows); e CL micrograph
of a U-bearing quartz conglomerate from Witwatersrand, RSA; radioactive fluids have caused
radiation damage along the grain boundaries and corroded the quartz grains (see arrows); f CL
micrograph of silica glass showing heterogeneities (bright stripes in CL) caused by Al impurities;
g, h Polarized lightCL pair of a hydrothermal synthetic quartz crystal; the CL image reveals
complex internal structures not visible in transmitted light
Applications of the characteristic luminescence properties of quartz in geosciences are numerous. One of the first applications was the evaluation of the
provenance of detrital quartz grains in sands and sandstones (Sippel 1968;
Zinkernagel 1978; Richter et al. 2003). Furthermore, processes of compaction,
brittle deformation, quartz cementation, and porosity evolution in the diagenetic
history of reservoir sandstones can also be evaluated using CL (e.g., Housknecht
1991; Millikan and Laubach 2000). CL in combination with trace elements was
used to reconstruct the geological evolution of granitic systems, ore-forming
processes or the metamorphic history of rocks (e.g., Mller et al. 2000, 2002,
2003a; Graupner et al. 2000; Rusk et al. 2006, 2008; Van den Kerkhof et al. 2004).
Furthermore, effects of shock damage in quartz due to impact events could be
revealed by CL in terrestrial and extraterrestrial samples (e.g., Sippel 1971;
Ramseyer et al. 1992; Gtze 2009b).
CL studies can also be helpful in technical applications such as for the evaluation of the quality of SiO2 raw materials and technical SiO2 products (Gtze
2000). Impurities and heterogeneities in high-purity quartz and SiO2 materials are
often only detectable by CL (see Fig. 1.3 fh). Moreover, SiO2 plays an important
role in many technologically important applications such as silicon semiconductor
device technology, optics, or SiO2 insulator layers in electronic devices. The
optical, electrical and mechanical properties of these high-tech materials are
dependent on the presence and/or generation of defects (imperfections and
impurities), which can be detected and characterized by CL spectroscopy (e.g.,
Barfels 2001; Stevens-Kalceff 2009). For instance, the generation of defects and
traps and the related electronic transitions can significantly influence the usability
of those materials. Characterization of the defect microstructure of silicon dioxide
(SiO2) allows the performance of these applications to be optimized (Fig. 1.4).
10
J. Gtze
Fig. 1.4 Cathodoluminescence spectra of undoped SiO2 layers (SiO2) and those implanted by
oxygen (SiO2:O) and silicon (SiO2:Si), respectively (black lines = initial spectra, red
lines = spectra after electron irradiation); the CL spectra show a significant increase of the red
emission band (650 nm) after doping with oxygen and an increased blue emission (450 nm) after
doping with silicon (data from Barfels 2001)
SiO2
Fe2O*)
3
TiO2
Al2O3
Cr2O3
CoO
CuO
MnO
NiO
V2O5
11
Fibre optics
Special optical
glass
[99.9 wt.-%
\2.0 ppm
\1.5 ppm
[99.8 wt.-%
\20 ppm
\25 ppm
\500 ppm
\0.1 ppm
\0.05 ppm
\0.1 ppm
\1.0 ppm
\0.15 ppm
\15 ppm
\0.01 ppm
\0.05 ppm
\0.05 ppm
\0.1 ppm
\0.1 ppm
\0.5 ppm
*)
Fe2O3 contents for crystal glass are \150 ppm, window glass
\0.1 mass-%, bottle glass 0.54 mass-%
Blankenburg et al. 1994; Gtze 1997; Mller et al. 2005). Correlations of impurity
concentrations and the temperature of formation of the host quartz indicate that
lowest trace-element contents can be expected for the temperature range between
ca. 480 and 530C (Wnsch 1987; Blankenburg et al. 1994). In conclusion, pegmatite quartz, metamorphogenically mobilised quartz veins and lenses as well as
some hydrothermal quartz veins have the greatest potential for high-purity quartz
raw materials.
The concentration limits of certain trace elements in SiO2 raw materials for
high-tech applications are very low (Tables 1.3, 1.5) and often require synthetic
instead of natural SiO2 material. For instance, the production of lamp tubing and
optics requires high-purity materials with Al concentrations below 20 ppm
(compare Table 1.3). Semiconductor base materials and crucibles use material
with Al \ 10 ppm, in micro-electronic devices U and Th concentrations should be
below 2 ppb, and the production of solar silicon needs raw materials with very low
concentrations of P and B (Table 1.5).
12
J. Gtze
1985; Leeder et al. 1987; Van den Kerkhof and Hein 2001). Recently, a couple of
modern analytical methods such as microchemical analysis, capillary electrophoresis, infrared spectroscopy, Raman spectroscopy, isotope measurements,
INAA or LA-ICP-MS is being used to provide data concerning the chemical
composition of fluid inclusions (e.g., Klemm 1986; Gerler and Schnier 1989;
Gtzinger 1990; Ghazi et al. 1993; Hanson et al. 1996; Hallbauer 1997; Channer
et al. 1999; Flem et al. 2002; Mller et al. 2003b; Gtze et al. 2004).
The information about type and number of inclusions provides important data
about impurities in SiO2 raw materials. Mineral and fluid inclusions can significantly influence the trace-element composition of quartz. On the other hand, the
type and amount of fluid inclusions can influence the melting behaviour of SiO2
raw materials (e.g. Gemeinert et al. 1992). Traces of refractory minerals (e.g.
zircon, aluminium silicates) have also to be considered if quartz raw materials are
used in melting processes (e.g. glass industry).
13
impurities during the mobilisation and crystallization processes resulting in a highpurity SiO2 material. In contrast, quarzite is a bedded massive metamorphic rock
with at least 90% quartz, formed from pre-existing quartz-rich sandstones by
compaction and metamorphosis. A less focused nomenclature for quartzite is used
in the industry. Here, all very hard, resistant rocks with [96% SiO2 are called
quartzite, independent on the formation history.
14
J. Gtze
Fig. 1.5 Micrographs in polarized light (except c = transmitted light) of different siliceous
rocks showing typical microstructures. a Massive chert from Texas, USA with fine-grained
granular quartz and chalcedony; b Lydite (Kieselschiefer) from Lommatzsch, Germany with
extremely fine-grained silica matrix and cross-cutting microcrystalline quartz veins of different
generations; c Siliceous sinter from Yellowstone, USA showing partially a layered microstructure; d Tertiary quartzite (silcrete) from Glossen, Germany with fine-grained siliceous matrix and
large detrital quartz grains
Another type of siliceous rocks is siliceous sinter. This is a porous, layered and
fine-grained siliceous rock, which originates from the evaporation of silica in hot
springs (Fig. 1.5c). Tripoli (or polishing slate) is a fine-grained, highly porous
siliceous rock with inorganic silica source. These fine-grained limnic sediments
are preferentially formed during Tertiary.
A special group of siliceous rocks is represented by silcretes. This group of
terrestrial siliceous rocks is especially reported from Australia (silcrete) and
Europe, where the rocks are called Tertiary quartzite (because of the preferred
occurrence in Tertiary sediments) or cement quartzite (in contrast to the metamorphic quartzite). These rocks originate from the silicification of pre-existing
rocks (mostly sediments) by silica from weathering solutions and may cover large
areas (Walther 1993). Probably the term limnoquartzite (or opalite) also describes
this type of rocks considering that a significant amount of silica may originate
from volcanic and/or hydrothermal activities.
Because of the intergrowth of both detrital and newly precipitated quartz and
silica material, these rocks represent more or less the transition between siliceous
15
Fig. 1.6 Formation of different types of chert by the diagenesis of opal-A precursors (siliceous
oozes) during burial and solution-reprecipitation steps via opal-A ?opal-CT ?microquartz; the
geological ages relate to the first occurrence; some of the images are modified from Fchtbauer
(1988)
and siliciclastic rocks. The texture and appearance of the silcretes drastically vary
due to the amount of detrital material and the kind of cementing material
(Figs. 1.5d, 1.7).
16
J. Gtze
minerals. This results in complex mineral composition and reflects at least partially
the primary composition of the host rock(s). In contrast, relatively slow erosion
together with warm and wet climatic conditions cause decomposition of unstable
minerals and result in a sedimentary rock with high amounts of detrital quartz.
Sandstones (Fig. 1.8) have in average 65 vol% quartz (Blatt et al. 1980), but it
may be C99% in some mature sands (Gtze 1997).
In addition, the granumlometric properties of the sedimentary rocks, e.g. grainsize distribution, roundness and surface properties of quartz grains are especially
influenced by the distance and intensity of the transport conditions and provide
information concerning the geological environment during rock formation. During
diagenesis, compaction/cementation as well as alteration and neoformation of
minerals can significantly change the properties of the siliciclastic rocks. Diagenetic and metamorphic processes can cause a compaction of sands/sandstones and
a transformation into quartzite (Fig. 1.8d).
17
Fig. 1.8 Micrographs in polarized light of different siliciclastic rocks showing typical mineral
composition and microstructures, respectively; a Arcosic sandstone from Altendorf, Germany
consisting of mainly quartz, feldspar and some sheet silicates; b Glauconitic sandstone from
Goslar, Germany with rounded glauconite grains and infiltrated illite on grain surfaces; c Mature
sandstone from Lohmen, Germany consisting almost exclusively of detrital quartz and diagenetic
silica cement; d Metamorphic quartzite (Dubrau quartzite, Germany) formed during high pressure
(and temperature) from a quartz rich sediment; note the typical sutured grain contacts
sands and gravels in the foundry and glass industry as well as for the production of
semiconductor silicon and silicon alloys, or quartzites for refractory materials
(Table 1.4). Knowledge of the interrelation between genesisspecific propertiesparameters for technical application of the raw material is necessary for a
successful use in many industrial applications. Therefore, for the evaluation of the
usability of SiO2 raw materials for a specific application, the limiting properties,
e.g. chemical purity, are essential (e.g., Blankenburg et al. 1994; Gtze 1997;
Mller et al. 2005).
18
J. Gtze
Table 1.4 Interrelation between genesis and specific properties of different types of SiO2 raw
materials and their preferred application in the industry
Quartz type
Properties
Preferred application
Magmatic/postmagmatic
Quartz of alaskite
(Iota quartz)
Chemical purity
Pegmatite and
Hydrothermal quartz
Chemical purity,
perfect crystal order
Metamorphic
Quartzite
Metamorphogenic
quartz mobilisates
Sedimentary
Quartz sands
Quartz gravel
Sedimentary quartzite
SiO2 up to [98%,
lumpy
Chemical purity
Refractory materials,
silicon and Si alloys (e.g. FeSi)
Quartz synthesis (lascas)
Chemical purity,
granulometric properties
Chemical purity,
grain size
Chemical purity,
cryptocrystalline silica
Table 1.5 Chemical composition (in ppm) of ultra-high purity Iota quartz standards (IOTA
2005) and high purity quartz HPQ (Norwegian Crystallites AS 2006)
Iota Std
Iota CG
Iota 4
Iota 6
HPQ*)
Al
Ti
Fe
Na
Zr
P
B
*)
16.2
1.3
0.2
0.9
1.3
0.1
0.08
14.7
1.1
0.2
1.0
1.3
0.1
0.08
8.6
1.4
0.3
0.3
\0.01
0.05
0.04
8.0
1.4
0.2
0.1
\0.01
0.05
0.04
\30
\10
\3
\8
\2
\1
silicates. Therefore, quartz from these rocks does not play an important role as raw
material. The only exception is the extraction of high-purity silica material by
chemical leaching from alaskites (so called Iota quartzJung 1992).
In contrast, pegmatite bodies and hydrothermal veins may provide large
amounts of high-purity quartz (Table 1.5). Such deposits can reach dimensions of
several tenths or hundreds of meters and with extremely low concentrations of
impurities (Blankenburg et al. 1994). This material is preferentially used as raw
material for the hydrothermal quartz synthesis (Table 1.4). High-purity natural
19
pegmatite and hydrothermal quartz crystals have formerly been used directly as
piezo quartz or optical quartz.
20
J. Gtze
Fig. 1.9 Scheme of an autoclave for hydrothermal growth of quartz single crystals (left) and
synthetic quartz crystal (rightwidth ca. 15 cm) showing typical morphology. Note the
development of large basal faces, which are uncommon in natural quartz crystals
21
1.5 Conclusions
SiO2 minerals and rocks are important constituents of the Earths crust and they
play an important role as usable materials since the beginning of human being.
Owing to their abundance and properties, SiO2 minerals and rocks have been used
in different applications such as tools, weaponries, jewelleries or building materials. Recently, quartz and other silica minerals and rocks cover a broad field of
geological and technical applications including high-tech materials. The identification and characterization of these SiO2 materials require both a clear nomenclature and a valuable analytical approach.
The classification of silica minerals and rocks is based on the mineralogical and
petrographical nomenclatures. Accordingly, we have to differentiate between SiO2
minerals (including different varieties) and SiO2 rocks. Quartz is the most
important silica mineral in respect to its appearance in the lithosphere and the
industrial use. Therefore, investigations concerning characteristic properties in
dependence on the specific conditions of formation can help to evaluate the
potential of raw material for industrial applications. Among the SiO2 rocks, in
particular sedimentary rocks have to be considered as potential raw material. The
group of sedimentary SiO2 rocks can be subdivided into the two sub-groups of
siliceous rocks and siliciclastic rocks. The properties of these rocks (chemistry,
mineral composition, texture, etc.) mainly depend on the geological history.
In recent years, the development and application of advanced analytical
methods have provided a large amount of new data concerning the structure and
properties of natural and synthetic SiO2 material, which can help to reconstruct
geological processes and to determine the viability of specific industrial
applications.
Acknowledgments I gratefully acknowledge the comments of K. Ramseyer (Bern) and an
anonymous reviewer, which helped to improve the quality of the paper.
22
J. Gtze
References
Baker JM, Robinson PT (1983) EPR of a new defect in natural quartz: possibly O2-. Solid State
Commun 48:551554
Bambauer HU (1961) Spurenelemente und c-Farbzentren in Quarzen aus Zerr-klf-ten der
Schwei-zer Alpen. Schweizerische Mineralogische Petrographische Mitteilungen 41:335367
Barfels T (2001) Kathodolumineszenz amorpher und kristalliner Modifikationen von SiO2 und
GeO2. Ph.D. thesis, University Rostock, p 168
Beall GH (1994) Industrial applications of silica. Rev Mineral 29:469506
Blankenburg H-J, Gtze J, Schulz H (1994) Quarzrohstoffe. Deutscher Verlag fr Grundstoffindustrie, Leipzig-Stuttgart, p 296
Blatt H, Middleton GV, Murray RC (1980) Origin of sedimentary rocks, 2nd edn. Prentice-Hall,
Inc., Englewood Cliffs, p 782
Botis S, Nokhrin SM, Pan Y, Xu Y, Bonli T (2005) Natural radiation-induced damage in quartz.
I. Correlations between cathodoluminescence colors and paramagnetic defects. Can Mineral
43:15651580
Channer DM DeR, Bray CJ, Spooner ETC (1999) Integrated cation-anion/volatile fluid inclusion
analysis by gas and ion chromatography; methodology and examples. Chem Geol 154:5982
Cohen AJ (1956) Color centers in alpha-quartz. Part I. Smoky quartz. J Chem Phys 25:908914
Demars C, Pagel M, Deloule E, Blanc P (1996) Cathodoluminescence of quartz from sandstones:
interpretation of the UV range by determination of trace element distributions and fluidinclusion P-T-X properties in authogenic quartz. Am Mineral 81:891901
Dennen WH (1964) Impurities in quartz. Geol Soc Am Bull 75:241246
Dennen WH (1966) Stochiometric substitution in natural quartz. Geochim Cosmochim Acta
30:12351241
Dennen WH (1967) Trace elements in quartz as indicators of provenance. Geol Soc Am Bull
78:125130
Ermakov NP (1950) Research on the nature of mineral-forming solutions (in Russ.). University of
Kharkov Press, Kharkov, p 460
Flem B, Larsen RB, Gromstvedt A, Mansfeld J (2002) In situ analysis of trace elements in quartz
by using laser ablation inductively coupled plasma mass spectrometry. Chem Geol
182:237247
Friebele EJ, Griscom DL, Stapelbroek M, Weeks RA (1979) Fundamental defect centers in glass:
the peroxy radical in irradiated, high-purity, fused silica. Phys Rev Lett 42:13461348
Fchtbauer H (1988) Sediments and sedimentary rocks. Schweizerbart, Stuttgart
Gemeinert M, Gaber M, Hager I, Willfahrt M, Bortschuloun D (1992) On correlation of gasliquid-inclusions properties and melting behaviour of different genetic quartzes for
production of transparent fused silica. Neues Jahrbuch Mineralogie, Abhandlungen 165:1927
Gerler J (1990) Geochemische untersuchungen an hydrothermalen, metamorphen, graniti-schen
und pegmatitischen Quarzen und deren Flssigkeitseinschlssen. Ph.D. thesis, University
Gttingen, p 169
Gerler J, Schnier C (1989) Neutron activation analysis of liquid inclusions exemplified by a
quartz sample from the Ramsbeck Mine, F.R.G. Nucl Geophys 3:4148
Ghazi AM, Vanko DA, Roedder E, Seeley RC (1993) Determination of rare earth elements in
fluid inclusions by inductively coupled plasma-mass spectrometry (ICP-MS). Geochim
Cosmochim Acta 57:45134516
Gtze J (1997) Mineralogy and geochemistry of German high-purity quartz sands. In: Papunen H
(ed) Mineral deposits: research and exploration. Balkema, Rotterdam, pp 721724
Gtze J (2000) Cathodoluminescence microscopy and spectroscopy in applied mineralogy.
Freiberger Forschungsheft, C 485 Geowissenschaften, p 128
Gtze J (2009a) Chemistry, textures and physical properties of quartzgeological interpretation
and technical application. Mineral Mag 73:645671
23
Gtze J (2009b) Cathodoluminescence microscopy and spectroscopy of lunar rocks and minerals.
In: Gucsik A (ed) Cathodoluminescence and its application in the planetary sciences.
Springer, Berlin, pp 87110
Gtze J, Blankenburg H -J (1990) Zur Methodik der komplexen mineralogischgeochemischen
Charakterisierung von Glassanden. Sprechsaal International Ceramics & Glass Magazine 123,
8:795803, 9:928940, 12:11841189
Gtze J, Lewis R (1994) Distribution of REE and trace elements in size and mineral fractions of
high purity quartz sands. Chem Geol 114:4357
Gtze J, Pltze M (1997) Investigation of trace-element distribution in detrital quartz by electron
paramagnetic resonance (EPR). Eur J Mineral 9:529537
Gtze J, Zimmerle W (2000) Quartz and silica as guide to provenance in sediments and
sedimentary rocks. Contributions to Sedimentary Petrology, vol 21. Schweizerbartsche
Verlagsbuchhandlung, Ngele & Obermiller, Stuttgart, 91 S
Gtze J, Siedel H (2007) A complex investigation of building sandstones from Saxony
(Germany). Mater Charact 58:10821094
Gtze J, Pltze M, Fuchs H, Habermann D (1999) Defect structure and luminescence behaviour
of agateresults of electron paramagnetic resonance (EPR) and cathodoluminescence (CL)
studies. Mineral Mag 63:149163
Gtze J, Pltze M, Habermann D (2001) Cathodoluminescence (CL) of quartz: origin, spectral
characteristics and practical applications. Mineral Petrol 71:225250
Gtze J, Pltze M, Graupner T, Hallbauer DK, Bray C (2004) Trace element incorporation into
quartz: a combined study by ICP-MS, electron spin resonance, cathodoluminescence,
capillary ion analysis and gas chromatography. Geochim Cosmochim Acta 68:37413759
Gtze J, Pltze M, Trautmann T (2005) Structure and luminescence characteristics of quartz from
pegmatites. Am Mineral 90:1321
Gtzinger MA (1990) Determination of aqueous salt solutions in fluid inclusions by infrared
investigations. Neues Jahrbuch Mineralogie, Monatshefte H.1:112
Graetsch H (1994) Structural characteristics of opaline and microcrystalline silica minerals. Rev
Mineral 29:209232
Graupner T, Gtze J, Kempe U, Wolf D (2000) Cathodoluminescence imaging as a tool for
characterization of quartz and trapped fluid inclusions in multistage deformed mesothermal
Au-quartz vein deposits: a case study from the giant Muruntau Au-ore deposit (Uzbekistan).
Mineral Mag 64:10071016
Griffiths JHE, Owen J, Ward IM (1954) Paramagnetic resonance in neutron-irradiated diamond
and smoky quartz. Nature 173:439442
Griscom DL (1985) Defect structure of glasses. J Non-Cryst Solids 73:5177
Hallbauer DK (1992) The use of selected trace elements in vein quartz and quartz pebbles in
identifying processes of formation and source rocks. Geologcal Society of South Africa 24th
Congress, Bloemfontein, Abstracts, 157159
Hallbauer DK (1997) The application of capillary ion analysis to the geochemistry of natural
aqueous fluids and in particular to the analysis of fluid inclusions in minerals. In: Proceedings
of the 30th International Geological Congress, vol 9, pp 409424
Hanson B, Delano JW, Lindstrom DJ (1996) High-precision analysis of hydrous rhyolitic glass
inclusions in quartz phenocrysts using the electron microprobe and INAA. Am Mineral
81:12491262
Heany PJ, Prewitt CT, Gibbs GV (1994) Silicaphysical behaviour, geochemistry and materials
application. Reviews in Mineralogy, vol 29. Mineralogical Society of America, Washington,
p 606
Heynke U, Leeder O, Schulz H (1992) On distinguishing quartz of hydro-thermal or metamorphogenic origin in different monomineralic veins in the eastern part of Germany. Mineral
Petrol 46:315329
Houseknecht DW (1991) Use of cathodoluminescence petrography for understanding compaction, quartz cementation, and porosity in sandstones. In: Baker CE, Kopp OC (eds)
Luminescence microscopy: quantitative and qualitative aspects. SEPM, Dallas, pp 5966
24
J. Gtze
Ioannou SE, Gtze J, Weiershuser L, Zubowski SM, Spooner ETC (2003) Cathodoluminescence
characteristics of Archean VMS-related quartz: Noranda, Ben Nevis, and Matagami districs,
Abitibi Subprovince, Canada. G3 Online Publication, 5(2), doi:10.1029/2003GC000613
IOTA (2005) IOTA high-purity quartz. http://www.iotaquartz.com/welcome.html. Accessed
20 May 2005
Jourdan A-L (2008) Elemental and isotopic zoning in natural alpine quartz. Ph.D. thesis,
University Lausanne
Jung L (1992) High purity natural quartz. Part I: high purity natural quartz for industrial use. Part
II: High purity natural quartz markets for suppliers and users. Quartz Technology. Liberty
Corner, New Jersey, p 657
Klemm W (1986) Beitrge zur analytischen Geochemie von Gas-Flssigkeits-Einschlssen in
hydrothermalen Mineralen. Habilitation thesis, TU Bergakademie Freiberg
Knauth LP (1994) Petrogenesis of chert. In: Heaney PJ, Prewitt CT, Gibbs GV (eds) Silica
physical behaviour, geochemistry and materials applications. Reviews in Mineralogy, vol 29.
Mineralogical Society of America, Washington, pp 233258
Komuro K, Horikawa Y, Toyoda S (2002) Development of radiation-damage halos in low-quartz:
cathodoluminescence measurement after He+ ion implantation. Mineral Petrol 76:261266
Konhauser K (2007) Introduction to geomicrobiology. Blackwell Publishing, Malden, p 425
Kostov RI, Bershov LV (1987) Systematics of paramagnetic electron-hole centres in natural
quartz (in Russian). Izvestiya Akademii nauk USSR, Seria Geologia 7:8087
Kostova B, Pettke T, Driesner T, Petrov P, Heinrich CA (2004) LAICP-MS study of fluid
inclusions in quartz from the Yuzhna Petrovitsa deposit, Madan ore field, Bulgaria. Swiss Bull
Mineral Petrol 84:2536
Krickl R, Nasdala L, Gtze J, Grambole D (2008) Alteration of SiO2 caused by natural and
artificial alpha-irradiation. Eur J Mineral 20:517522
Larsen RB, Polv M, Juve G (2000) Granite pegmatite quartz from Evje-Iveland: trace element
chemistry and implications for high-purity quartz formation. Norges Geologiske Underskelse Bulletin 436:5765
Larsen RB, Henderson I, Ihlen PM, Jacamon F (2004) Distribution and petrogenetic behaviour of
trace elements in granitic pegmatite quartz from South Norway. Contrib Mineral Petrol
147:615628
Larsen RB, Jacomon F, Kronz A (2009) Trace element chemistry and textures of quartz during
the magmatic hydrothermal transition of Oslo Rift granites. Mineral Mag 73:691705
Leeder O, Thomas R, Klemm W (1987) Einschlsse in Mineralen. VEB Deutscher Grundstoffverlag, Leipzig
Lehmann K, Berger A, Gtte T, Ramseyer K, Wiedenbeck M (2009) Growth related zonations in
authigenic and hydrothermal quartz characterized by SIMS-, EPMA-, SEM-CL- and SEMCC-imaging. Mineral Mag 73:633644
Lyakhovich VV (1972) Trace elements in rock-forming minerals of granitoides (in Russian).
Nedra, Moscow, p 200
Mackey JH (1963) EPR study of impurity-related color centers in germanium-doped quartz.
J Chem Phys 39:7483
Meunier JD, Sellier E, Pagel M (1990) Radiation-damage rims in quartz from uranium-bearing
sandstones. J Sediment Petrol 60:5358
Milliken KL, Laubach SE (2000) Brittle deformation in sandstone diagenesis as revealed by
scanned cathodoluminescence imaging with application to characterization of fractured
reservoirs. In: Pagel M, Barbin V, Blanc P, Ohnenstetter D (eds) Cathodoluminescence in
geosciences. Springer, Berlin, pp 225243
Mineeva RM, Bershov LV, Petrov I (1991) EPR of surface-bound Fe3+ ions in polycrystalline
quartz (in Russian). Dokladi Akademii Nauk SSSR 321:368372
Miyoshi N, Yamaguchi Y, Makino K (2005) Successive zoning of Al and H in hydrothermal vein
quartz. Am Mineral 90:310315
25
26
J. Gtze
27
Weil JA (1984) A review of electron spin spectroscopy and its application to the study of
paramagnetic defects in crystalline quartz. Phys Chem Mineral 10:149165
Weil JA (1993) A review of the EPR spectroscopy of the point defects in a-quartz: the decade
19821992. In: Helms CR, Deal BE (eds) Physics and Chemistry of SiO2 and the Si-SiO
interface 2. Plenum Press, New York, pp 131144
Wopfner H (1978) Silcretes of northern South Australia and adjacent regions. In: Langford-Smith
T (ed) Silcrete in Australia, Armidale, pp 93141
Wopfner H (1983) Environment of silcrete formation: a comparison of examples from Australia
and the Cologne Embayment, West Germany. In: Wilson RCL (ed) Residual deposits: surface
related weathering processes and materials. Geological Society of London Special
Publications 11, pp 151158
Wright PM, Weil JA, Buch T, Anderson JH (1963) Titanium colour centers in rose quartz. Nature
197:246248
Wnsch K (1987) Mineralogische, geochemische und strukturelle Untersuchungen an
metamorphogenen Quarzmobilisaten. Ph.D. thesis, TU Bergakademie Freiberg, Germany
Zinkernagel U (1978) Cathodoluminescence of quartz and its application to sandstone petrology.
Contributions to Sedimentology, 8. Schweizerbartsche Verlagsbuchhandlung, Ngele and
Obermiller, Stuttgart, p 69
Chapter 2
Abstract Very high purity quartz for advanced high-tech applications is currently
sourced from just a few locations around the world. Such is the expected growth in
demand that more sources are required to be found. For successful high purity raw
quartz resource identification detailed analysis and appropriate process technology
selection is essential. This article reviews general aspects of high purity quartz
deposits, exploration requirements, quality evaluation of raw quartz, and provides
basic insight into the different specifications and market developments of the hightech end-user industries reliant on very high purity refined quartz products.
2.1 Introduction
Quartz is one of the most abundant minerals. It occurs in many different settings
throughout the geological record (Gtze 2009). As the prime source of silica the
mineral has wide large volume application in the manufacture of glass, ceramics,
refractory materials and other traditional uses (Blankenburg et al. 1994). However,
only very few deposits are suitable in volume, quality and amenability to tailored
refining methods for speciality high purity applications. As such high purity quartz
has become one of todays key strategic minerals with applications in high-tech
industries that include semiconductors, high temperature lamp tubing, telecommunications and optics, microelectronics, and solar silicon applications (Blankenburg
et al. 1994; Haus 2005, 2010; Moore 2005; Dal Martello et al. 2011a, b).
29
30
R. Haus et al.
Whilst most processing plants for standard quartz applications deploy off-the-shelf
technology solutions, quartz for high-tech applications requires tailored processing
techniques and specially designed equipment to achieve essential high purity specifications. Beneficiation of raw quartz into refined high-purity products involves several
refinement steps which need to be adapted to effectively minimise the specific
impurities of the individual raw quartz feed to comply with stringent end-use specifications (Haus 2005). As a result, high purity quartz with total impurity levels less than
20 ppm may be achieved so creating a highly valuable raw material which commands
up to 5 EUR/kg.
31
and further development of Brazilian quartz for high quality products. Potential
deposits in Madagascar and Angola also suffer from poor infrastructure and lack of
interest on the part of their governments.
In Asia, Japans influence is large especially in Southeast India and in Sri
Lanka. Japan once imported quartz lumps from these regions. However, government-led efforts on the part of the quartz supplier countries were launched to stop
the export of unrefined quartz lumps and to support the development of quartz
processing within the country. Today these countries specialise in the production
of high purity filler materials for epoxy moulded compounds (EMC) low in uranium and thorium used in the manufacture of computer chips.
Given its strategic relevance to the semiconductor and photovoltaic industries
many more high purity quartz deposits are under development. However, the
exploration and exploitation of new suitable quartz deposits is hampered by quality
assurance regulations, which are globally applied. Whilst traditional sources of
quartz have been questioned in quality terms, it is in their interest to demonstrate
that their material is tested and meets the standards of silica glass production.
Suppliers from new quartz deposits have yet to achieve this status. This is achieved
normally in several consecutive cycles of tests with increasing quantities of test
material. These tests must be carried out for many of the product groups in the
various application areas.
These pre-business services are expensive but made necessary by increasingly
strict quality regulations. They require certain personnel expertise from the
potential new suppliers (e.g. for the provision of relevant raw material) and highly
specialised cooperation partners well recognised within the industry.
32
R. Haus et al.
Fig. 2.1 High temperature Xenon lamp made from high purity quartz
Fig. 2.2 High purity quartz is applied as filler material in epoxy moulding compounds (EMC)
used as cover in the microelectronics chip production
thermal shock resistance and thermal stability. It is used in the high performance,
high temperature lamp manufacturing sectors for UV lamps, mercury, xenon and
halogen bulbs, and high intensity discharge lamps (Fig. 2.1).
Silica glass is widely used as basic material for optical fibers and additional
optoelectronic devices in the telecommunications industry. It is used in the optical
industry in microlithographic applications, excimer laser optics, beamers and other
specialised applications. In the microelectronics industry, a major application is as
filler material in epoxy moulding compounds (EMC) for electronic components
(Fig. 2.2).
Silicon is the most common material for the production of solar cells in the
photovoltaic industry either in mono- or polycrystalline form.
Specific requirements as to tolerable limiting values differ from industry to
industry. In the lamp tubing and optics industries aluminium content in the refined
33
quartz concentrate should not exceed 20 ppm, other metals should be less than
1 ppm, and total impurities less than 30 ppm. For semiconductor base materials
and crucibles aluminium content should be even lower, specified to less than
10 ppm, other metals less than 0.1 ppm, and total impurities not to exceed 15 ppm.
Feedstock for solar silicon used in the photovoltaic industry should generally
have boron and phosphorus contents in the sub-ppm range since these elements are
most difficult to remove and negatively affect the performance of the solar cells
(Geerligs et al. 2002).
For microelectronics applications, e.g. in epoxy moulding compounds, uranium
and thorium, both responsible for soft errors by alpha radiation, should be less than
2 ppb, and in low alpha applications, even less than 0.5 ppb.
34
R. Haus et al.
well as exceptional shock resistance (Chap. 3), the demand in high purity quartz will
benefit from the increasing demand in the lamp tubing and automotive industries.
35
used for PV applications. More recently, however, the same sizes for crucibles as
are used in the semiconductor industry are being requested. Therefore, the increase
in high-purity quartz granule consumption will not equal the upcoming growth in
c-Si monocrystal production, but it is still estimated at above 5% p.a.
Manufacturers of solar silicon are looking to remain competitive by pushing
production costs down. The result might be a process of consolidation: in 2012, the
four largest producers aloneOCI, Hemlock, GCL and Wackercould probably
cover worldwide demand. Prices, which are currently still dominated by demand,
should then fall and relate more to costs.
This could attract new technologies which are based on high purity materials
and less investment and operating costs to enter the solar silicon market. On the
other hand it will raise the demand in high quality quartz, specifically low in boron
and phosphorus.
36
R. Haus et al.
Fig. 2.3 Typical vein quartz deposit (Africa), width of the picture approx. 15 m
37
Fig. 2.4 Aluminium distribution in the vertical drill core of a massive pegmatite quartz deposit,
elevation in m
In order to reduce exploration costs information from initial drilling is combined with geophysical data derived from methods such as seismic, gravimetric,
geoelectric or geomagnetic field surveys.
Geophysical resistivity profiles apparent lateral variations in resistivity of rocks
using a specific electrode array over constant distances in specific locations.
Vertical Electrical Sounding (VES) provides information on vertical variations in
resistivity within a geological formation. Apparent resistivity values are correlated
with the geological formations present to provide an interpretation of the extent of
a deposit (Fig. 2.5). This assists efficient drill hole siting, the results of which
ideally confirm and complement, in greater detail, resistivity data.
Once a quartz deposit has been identified the precise quality of its mineral
content and potential for quality improvement are key factors in determining its
economic value. Representative samples are taken for detailed investigation to
evaluate the potential of the raw material to be processed into a high value refined
product. Determinative mineralogical techniques characterise any fluid and/or
mineral inclusions that need to be removed by applying tailor-made processes. The
combination of mineralogical characterisation across the whole deposit by means
of representative sample analysis with the identification of appropriate specific
processes to remove impurities is crucial to the thorough evaluation of any raw
quartz deposit for high purity and high value applications.
2.6 Analytics
Naturally occuring high purity quartz always has inclusions which are present
either in the form of finely dispersed solids (mineral phases such as e.g. muscovite,
rutile, calcite) or fluid (liquid and gaseous) inclusions which can provide valuable
38
R. Haus et al.
Fig. 2.5 Interpretation of lateral resistivity profiling and Vertical Electrical Sounding (VES) data
presenting the thickness and subsurface distribution of a quartz body
insight into the conditions prevalent at the time of formation of the quartz. By the
analysis of trace elements and impurities the conditions of formation of the quartz
can be unraveled. Typically the bulk sample will be chemically characterized first.
Then the size and distribution as well as chemical composition of the mineral and
fluid inclusions are investigated in detail. The different steps in the analysis of raw
quartz samples are discussed in the following.
39
40
R. Haus et al.
Fig. 2.6 Left: photomicrograph (reflected light) of a large xenomorphic calcite crystal with wavy
deformation twins. Note also the presence of smaller irregularly shaped crystals nearby. Width of
the picture is 600 lm; Right: photomicrograph (reflected light) of quartz with a cluster of apatite
crystals on the left and tremolite fibres on the right side. Picture width is 2.3 mm
Optical microscopic analysis of raw quartz is the starting point. It allows rapid
detection and a first overview of the sample texture and structure as well as impurities
such as fluid and solid phase inclusions. An example of large xenomorphic calcite
crystals with characteristic deformation twinning and apatite and tremolite inclusions
determined by optical microscopy within the quartz matrix are shown in Fig. 2.6.
Impurities which are directly accessible to the analytics can most conveniently
be detected by Micro X-ray fluorescence. An example of K-feldspar in quartz
identified by Micro XRF is shown in Fig. 2.7. EPMA is another technique which is
suitable for the precise identification of the chemical composition of minor
inclusions such as K-feldspar in quartz shown in Fig. 2.8. Trace element contents
ranging in the 100 ppms can be detected by conventional EPMA. When improved
background modeling, new quantitative blank correction routines and multiple
spectrometers to improve the geometric efficiency are used, detection limits for Ti
and Al as low as 23 ppm and 67 ppm can be achieved (Donovan et al. 2011).
However, critical questions such as the localization of aluminium (isomorphic
substitution for silicon in the quartz crystal lattice and/or mineral inclusions) can only
be determined by combining optical microscopy with advanced spectroscopic
methods (Tlili et al. 1989), e.g. Raman spectroscopy (Fig. 2.9), or electron paramagnetic resonance (EPR) spectroscopy. In Raman spectroscopy the sample is
irradiated by laser light. A small part of the scattered light shows a shifted frequency
with respect to the primary light resulting from inelastic scattering processes. From
the resulting spectrum the oscillation frequency of the molecules can be determined.
To unambiguously identify if substitutional elements are incorporated into the
quartz lattice structure electron paramagnetic resonance is used.
EPR is a method which is able to detect the presence of unpaired electrons in a
material. The sample is placed in an external uniform magnetic field and irradiated
by microwaves which interact with the unpaired electrons. In quartz unpaired
electrons are present in the paramagnetic [AlO4]0 centre or as electron- and hole
centres which are caused by radiation (Gtze and Pltze 1997; Nuttall and Weil
41
Fig. 2.7 Photograph of K-feldspar in quartz (whitish spots in the area framed by black ink dots)
and corresponding analysis by Micro XRF
Fig. 2.8 Microphotograph (crossed polarizers) of a small grain of K-feldspar in quartz and
corresponding results of electron probe microanalysis
1981; Okada et al. 1971). Figure 2.10 shows an example of EPR spectra of two
quartz samples (labeled 1 and 2) from different deposits. In the upper part of the
figure the simulated spectra of [AlO4]0 and [TiO4-/Li+]0 centers are shown.
Measured spectra match those of the simulations and clearly confirm the presence
of both types of centers in sample 1 (1a, 1b) and the presence of the former in
sample 2. The aluminium contents in the samples, as determined by ICP-OES
analyses, are 64 ppm and 10 ppm respectively.
42
R. Haus et al.
Fig. 2.9 Laser micro-Raman spectra and photomicrograph (inset upper right) of a muscovite
inclusion in host quartz (grey). Note the presence of the OH stretching peak at *3630 cm-1
(inset upper left) that is characteristic for aluminous micas (muscovite)
43
Fig. 2.10 Determination of substitutional elements Ti and Al in the quartz lattice by EPR;
simulated EPR spectra of [AlO4]0 and [TiO4-/Li+]0 centers are shown in the top part; measured
spectra of two different quartz samples (labeled 1 and 2) are shown below: sample 1 shows both
the typical signature of [TiO4-/Li+]0 and of [AlO4]0 centers (spectra 1a, 1b), sample 2 indicates
only the presence of a minor amount of [AlO4]0 centers
44
R. Haus et al.
Fig. 2.11 Secondary aqueous two-phase liquidvapor inclusions with low degree of fill in
quartz. Plane Polarized light
Fig. 2.12 Microphotograph of a large re-equilibrated two liquidvapor inclusion with a crossshaped black twinned crystal. Plane polarized light
45
impurity level when element concentrations enriched in the fluid inclusions are
forced in the silica glass during the melting process. Fluid inclusions may form
bubbles making the silica glass less valuable. Since silica glass melt is highly
viscous, smaller bubbles are not able to rise to the surface of the melt and so
survive to unacceptably impair the quality of the silica glass.
In order to define the properties of the fluid inclusions, optical microscopy in
combination with micro thermometry and Raman spectroscopy is employed.
Specific options for process development, e.g. thermal treatment or opening of
fluid inclusions by specific comminution technology, are explored according to the
results of these analyses. Some quartz samples consist of clear and milky parts, the
latter being richer in fluid inclusions than the former. The size of these parts is
determined during the characterisation procedure in order to evaluate options for
separating clear from milky quartz, by e.g. optical sorting. All these analyses are a
full quality assessment that is a prerequisite for the design of optimum bench, pilot
or technical scale processing tests.
Pre-processing (mechanical)
Physical processing
Chemical leaching
Thermal treatment
2.7.1 Pre-processing
Based on the specific characteristics of the quartz deposit, one or more processing
stages are required in order to liberate mineral impurities and fluid inclusions for
further physical treatment:
Initial crushing
Optical sorting
Comminution
Classification to product particle size
46
R. Haus et al.
47
Fig. 2.13 Fraction 38 mm, clear quartz (left) and milky quartz (right) after optical sorting
0.3
0.8
6.2
10
0.6
1.6
2.1
2.4
1.3
1.4
\0.1
\0.1
Table 2.2 Chemical analyses of quartz sample 0.10.3 mm after conventional comminution and
electrodynamic fragmentation
Al (ppm) Fe (ppm) Na (ppm) K (ppm) Li (ppm) Ti (ppm) Zr (ppm)
Quartz raw
41
material
Conventional
23
comminution
Electrodynamic
28
fragmentation
4.9
12
15
0.5
1.3
\0.1
464
10
3.5
0.6
1.9
\0.1
1.3
13
4.7
0.6
0.5
\0.1
new technology is ideal for selective liberation of gas and liquid inclusion trails
within the quartz crystals (Table 2.2).
Attrition
Magnetic separation
High tension separation
Flotation
48
R. Haus et al.
Attrition is applied to clean the surfaces of the quartz particles. Thereby fine
particles attached to quartz surfaces, e.g. clay minerals or iron oxide coatings, are
either washed away or liberated for their subsequent physical separation. Magnetic
separation removes heavy minerals from quartz as they are mostly paramagnetic or
ferromagnetic. These minerals are attracted by a magnetic field. Quartz, being
diamagnetic, is repelled. Since magnetic susceptibility is strong in the case of
ferromagnetic minerals only moderate magnetic field strengths are necessary for
their separation, whereas higher field strengths are required to separate paramagnetic minerals.
High tension techniques separate minerals owing to differences in their surface
conductance. For this processing step particles are uniformly passed through an
electrostatic field. The electrostatic separator consists of a heated chamber where
the electrodes are situated. The generated electrostatic field is up to 120 kV. Feed
material is activated by heating the sample or by the addition of diluted acids to the
feed material prior to heating. Typically, feldspar impurities may be separated
from quartz via high tension as a dry alternative to froth flotation.
Froth flotation selectively separates minerals according to differences in their
ability to be wetted, enhanced or suppressed by conditioning reagents. Separation takes place in a water-filled medium into which the ore is fed to form a
suspension which is agitated to avoid sedimentation processes. A frothing agent
is added and air introduced to form rising air bubbles. Hydrophobic/Hydrophobized mineral particles (such as heavy minerals, feldspar or mica) attach to
the air bubbles and rise to the surface forming froth whereas hydrophilic
(wetted) particles remain below the froth layer in the suspension. The mineralcarrying froth is removed.
Flotation process designs vary in complexity depending primarily on the type of
mineral, degree of liberation, and the desired purity of the product.
49
Table 2.3 Chemical analyses of quartz sample 0.10.3 mm after chemical treatment
Al (ppm) Fe (ppm) Na (ppm) K (ppm) Li (ppm) Ti (ppm) Zr (ppm)
Quartz after
magnetic
separation
Acid washing
Leaching
Hot chlorination
21
0.2
3.1
1.0
2.2
1.2
\0.1
21
20
21
\0.1
\0.1
\0.1
2.8
0.7
0.2
0.9
0.3
\0.1
2.2
2.2
1.6
1.2
1.2
1.1
\0.1
\0.1
\0.1
Fig. 2.14 Comparison of quartz grains of raw quartz, fraction 0.10.3 mm (left) and quartz after
thermal treatment (right)
(Table 2.3) which are highly restricted in lamp tubing and semiconductor
applications.
2.8 Conclusion
Whilst many industrial minerals deploy off-the-shelf technology solutions most speciality minerals require tailored processes and specially designed equipment that
cannot be conceived prior to thorough analytical evaluation. That is why detailed
investigations of the specific impurities ubiquitously found in quartz need to be
performed before tailor-made processing concepts may be aligned with realistic
50
R. Haus et al.
Fig. 2.15 Comparison of melting results of raw quartz, fraction 0.10.3 mm (left) and quartz
after thermal treatment (right)
investment costs and quality requirements. Engineering services need to cover all the
technology and economic details such as estimation of investment costs, details of
main plant and equipment, calculation of mass balances, energy requirements, and
specific production costs before the final investment decision can be made. A precise
and sustainable definition of the product split, in terms of relevant chemistry and
appropriate physical characteristics, and end-user acceptance are two prerequisites for
the success of any minerals-based project. End-user approval of processed samples at
potential qualities is the ultimate, and necessary, risk management process.
References
Andres U, Jirestig J, Timoshkin I (1999) Liberation of minerals by high voltage electrical pulses.
Powder Technol 104:3749
Blankenburg H-J, Gtze J, Schulz H (1994) Quarzrohstoffe. Deutscher Verlag fr Grundstoffindustrie, Leipzig, p 296
Dal Martello E, Bernardis S, Larsen RB, Tranell G, Di Sabatino M, Arnberg L (2011a) Electrical
fragmentation as a novel refining route for hydrothermal quartz for SoG-Si production. Miner
Eng. doi: 10.1016/j.powtec.2012.02.055
Dal Martello E, Tranell G, Gaal S, Raaness OS, Tang SK, Arnberg L (2011b) Study of pellets and
lumps as raw materials in silicon production from quartz and silicon carbide. Metall Mater
Trans B, p 12
Donovan JJ, Lowers HA, Rusk BG (2011) Improved electron probe microanalysis of trace
elements in quartz. Am Mineral 96:274282
Flem B, Larsen RB, Grimstvedt A, Mansfeld J (2002) In situ analysis of trace elements in quartz
by using laser ablation inductively coupled plasma mass spectrometry. Chem Geol
182:237247
Geerligs LJ, Wyers GP, Jensen R, Raaness O, Waernes AN, Santen S, Reinink A, Wiersma B
(2002) Solar grade silicon by a direct route based on carbothermic reduction of silica:
requirements and production technology. Energy Research Centre of the Netherlands (ECN)
publication ECN-RX02-042
51
Chapter 3
Abstract Quartz sand is a valuable raw material for the building and construction
industry. Individual branches demand very different sand qualities. The requirements of the following products are discussed: autoclaved aerated concrete, calcium silicate units, cement, clay bricks and roof tiles, concrete, mortar and render.
Not only very pure sand qualities with [98 wt% SiO2 are required, some applications need only 95, 88 or 70 wt% SiO2. The requirements for the individual uses
are, however, strict. New market developments and changes in legislature force the
construction industry to improve its products. These developments cause modifications in the production process as well as new requirements to the raw materials
used. Consequently, traditional branches may have to change their raw material
deposits. There is a strong need for data concerning the chemical and granulometrical composition of near surface sand deposits.
3.1 Introduction
3.1.1 The Idea Behind
The aim of this chapter is to increase the understanding of little researched but
important commercial raw materials that suffer from some kind of ignorance.
Although sand deposits form the basis of many important industrial activities, it is
very difficult to learn something about their existence and it is almost impossible to
find mineralogical or chemical data. If data are available they mainly cover special
H. B. Walther (&)
Xella Technologie und Forschungsgesellschaft mbH, Kloster Lehnin, Germany
e-mail: Hartmut.Walther@xella.com
53
54
H. B. Walther
sand of high puritybut 95% of sand and gravel are used in civil engineering and
only 5% for special uses (Table 3.1).
New market requirements and changes in legislature force the construction
industry to improve its productsbuilding materials for the safe use in sustainable
buildings. This development causes changes in the production process as well as
new requirements to the raw materials used. Unfortunately, most geologists and
even raw material specialists have little concern about the needs of the construction industry and its requirements for raw materials. This chapter intends to
reduce this gap of knowledge. It will inform about the needs of building material
producers and will encourage people to have a closer look to standard materials.
Quartz sand according to the definition of the German law contains minimum 80
wt% quartz (Anonymus 1985). The needs from the industry differ from that value
(Blankenburg et al. 1994; Lorenz and Gwosdz 1999). There is a strong need for
data about sand and gravel everywhere in the world.
55
On the other hand, the demand for high sound insulation requires massive walls
of high density produced from calcium silicate blocks. Wall blocks of high density
are needed on streets with heavy traffic or between semi-detached houses to ensure
privacy.
In addition, it is the aim of product research to allow a short construction time
of the buildings and to avoid possible sources of mistakes by craftsmen. As a
consequence, individual elements increased in size. Traditional bricks of normal
format NF have a size of 7.1 9 11.5 9 24 cm but modern blocks have sizes like
25 9 30 9 36.5 cm or 50 9 100 9 36.5 cm. Even storey-high panels of different
materials are available today.
The whole production process was optimized and developed to produce construction elements of sizes as described above and with reduced or increased raw
densities. This process resulted in new requirements to the raw materials including
quartz sand.
56
kg/m
H. B. Walther
600
400
200
0
1920 1940 1960 1980 2000 2020
year
moisture and improvements in the phase composition. Figure 3.1 illustrates the
development of the raw density. Beside an ongoing reduction of the raw density of
traditional products new products of lower raw densities were developed. Lightweight mineral insulation board of a raw density of about 100 kg/m3 is state of the
art. Its brand name is YTONG-Multipor and it is used for insulation purposes.
The compressive strength of a material with a given recipe/composition is of
course directly related to the density of this material. Thus, the compressive
strength of YTONG-Multipor is lower than the compressive strength of normal
wall materials. However, an identical compressive strength of 2 N/mm2 can be
achieved today with a raw density of 300 kg/m3 which is about half the value of
that from the 1950s.
This development caused a remarkable increase in the required quality of the
raw sand (Table 3.2). Products of high raw density and low strength can be produced with sand of lower quality. However, a high content of alkaline elements
can lead to efflorescing salt crystals forming a white covering of the wall.
57
Table 3.2 Requirements of siliceous raw materials for the use in modern AAC
Property
Requirement
Comments
Chemical composition
SiO2
K2O ? Na2O
Content of humic acids
Maximum grain size
Fraction \63 lm
Critical components
[80 wt%
\1 wt%
\0.5 wt%
No performance required
\4 wt%
Montmorillonite
Coal and wood
Table 3.3 Grain size distribution of sand for the use in calcium silicate blocks according to
Quincke and Eden (2002)
Unfavourable grain size distributions
Uniformly grained
Favourable grain size distributions
Gap grading
grain is the smaller is its geometrical resistance against pressure (Table 3.4).
Optimal raw sand allows a reduction of the lime without any influence on the
mechanical strength. Single grained raw sands restrict the possible variety of
products and require higher amounts of binder.
58
H. B. Walther
Table 3.4 Grain shape groups and its importance for calcium silicate units according to
Gundlach (1973)
Grain shapes
Group / suitability
sharp edged
very well suitable
edged
well suitable
moderately rounded
suitable
rounded
sufficient
well rounded
restricted use
Fig. 3.2 Production program of a calcium silicate unit producer in 2010, 50 years ago only block
sizes of the lower corner were produced
At the beginning, more than 100 years ago, calcium silicate units were produced in the traditional brick formats like 240 9 115 9 71 mm (Normalformat,
NF, in Germany) or 250 9 125 9 65 mm (Reichsformat, RF, in Germany) or
240 9 115 9 52 mm (Dnnformat, DF, in Germany). Later the height of the
blocks increased and in the 1950s a large portion of the production was delivered
in 2DF i.e. 240 9 115 9 113 mm. Later improvements in the moulding press
technology allowed the production of larger and even higher blocks of 6 DF or 12
DF. A new technological era started in 1973 (Kendel 1973). Calcium silicate units
reached a height of 50 cm (Fig. 3.2) and at this point the compressibility of the raw
materials and the internal structure of the compressed and hardened product
obtained a new importance. Single grained raw sands could not be used any longer
59
Table 3.5 Requirements of siliceous raw materials for the use in calcium silicate units
Property
Requirement
Comment
Chemical composition
SiO2
K2O ? Na2O
Content of humic acids
Maximum grain size
Fraction \125 lm
Critical components
[50 wt%
\1 wt%
\0.8 wt%
About 20 mm
[5 wt%
Mica
Coal and wood
Pyrite/
marcasite
Clay minerals
for these large products. Their technological disadvantages could not be compensated even by using larger amounts of lime.
A second development influenced the requirements of the mineralogical composition of the raw sand. Increasing legal demands concerning sound insulation
require walls of high weight and consequently higher raw density.
Increasing block heights require raw material with sufficient fine fraction which
will be well compacted (Table 3.5). The fine fraction has not necessarily to consist
of quartz only, but may contain also clay minerals. For example: In one case a
minimum content of about 5 wt% of quartz \125 lm is needed to produce 50 cm
high elements of the strength class 20 N/mm2. The sand for the use in the production is sometimes mixed from 3 up to 5 different qualities of sand. In collaboration with the University of Kassel the German Research Center for the Calcium
Silicate Industry developed a computer program to optimize the grain size distribution from up to 5 individual sand qualities (Eden et al. 2008).
High sound insulation of a wall requires a high density of the building material.
This can be achieved by adding coarse particles into the production mixture,
particles up to 20 mm are possible. To increase the density, rocks of higher density
like basalt can be added. High end products contain heavy mineral fractions
(Bertran 1995, Table 3.8 heavy aggregates) and the remaining quartz grains
assure the stability of the building material. Its structure complies with that of a
gap grading (Table 3.3).
60
H. B. Walther
Table 3.6 Requirements of siliceous raw materials for the use in the cement industry
Property
Requirement
Comment
Chemical composition
SiO2
K2O ? Na2O
Maximum grain size
Critical components
[95 wt%
\1 wt%
No performance required
Alkaline elements
61
Table 3.7 Requirements of siliceous raw materials for the use in the clay brick/roof tile industry
Property
Requirement
Comment
Chemical composition
SiO2
Al2O3
K2O ? Na2O
Content of humic acids
Maximum grain size
Fraction \63 lm
Critical components
[85 wt%
No performance required
No performance required
No performance required
\4 mm
No performance required
Coarse carbonate grains
drying sheds were typical for old brick factories. The drying is accompanied by
shrinkage of the clay material. During burning the final material properties are
generated and a second volume reduction occurs. Modern factories include a drying
plant for well defined drying with minimal stress to the structure of the material.
The stringent requirements for building materials introduced in recent decades
also affected the traditional brick industry. The height of the bricks increased
remarkably. The Romans used formats only a few centimetres high, subsequently
the height was increased to 5.2 cm and today, most blocks are about 24 cm high.
New thermal requirements demanded a reduction of raw density of the bricks. This
goal is mainly reached by a honeycomb or cellular structure of the bricks. In some
cases the cell walls are also porous.
The final dimensions of the block/tile have to be reached after two shrinkage
steps of several millimetres each. Consequently, the producers take special care to
limit shrinkage. One approach is to reduce the moisture of the ductile material, a
second one is to reduce the content of clay minerals in the initial raw material by
adding fine grained quartz sand. However, this approach introduces a new source
of defectsat burning temperature quartz occurs in its hexagonal high temperature
modification and its transformation into the trigonal low temperature modification
during cooling results in a volume decrease of about 0.8% at 573 C. This volume
change causes cracks in the material and is one of the most important scrap
sources, especially in roof tile production. Consequently, a small grain size is
needed to reduce quartz stress in the ceramic material. Because of mixing of sand
and clay before processing the ceramic raw mass, the grain size of quartz is of
minor interest. The material is passing two or three rolling mills and the final one
has an opening of 0.65 mm and all larger grains are crushed.
The quartz grains in the ceramic mass for the production of clay bricks and roof
tiles should be very small because of the volume change from the high temperature
modification to the low temperature modification. Coarse grains are broken in
rolling mills during processing of the raw materials. Coarse carbonate grains must
not be contained within the sand because of their thermal decomposition during
the burning process (Table 3.7).
62
H. B. Walther
C3.0
Light weight
Aggregates
0.4 2.0
Barite
Magnetite
Ilmenite
Hematite
Sand
Gravel
Crushed stone
Pumice
Perlite
Steel sand
Heavy metal slag
Ferrophosphorous
Blast furnace slag
Corundum
Crushed concrete
Expanded shale
Expanded clay
Expanded perlite
Crushed bricks
63
suspension. The latter is the joker of the concrete specialists: dispersed silica
tremendously increases the strength of concrete. Using these ingredients and
natural aggregates with a high content of quartz grains and pebbles, a concrete can
be produced having a compressive strength corresponding to quartzitic sandstone
or dense limestone. Such building projects can be a challenge in terms of homogeneity of material and continuity of operation. A construction site of a cooling
tower in Niederauem/Germany required 17,650 m3 UHPC permanently delivered
in a constant quality of aggregates and mixture during several weeks without any
interruption (Weber and Riechers 2003).
UHPC is a material providing a new level of stability and strength compared to
regular concrete. This offers the possibility to produce products using UHPC that
were formerly produced from metal. An example is given by Niemann et al. (2008) in
the German patent application DE 10 2007 016 719 A1 concerning calendar rolls in
the production of paper and cardboard. Even compressive strength values between
150 and 200 N/mm2 can be reached by using Reactive Powder Concrete (RPC).
The requirements for the use of sand and gravel in concrete are mandatory
regulated by the European standard DIN EN 12620. This standard as well as the
German standard DIN 1045-2 must be considered in Germany together with some
other regulations. Amongst them, the Alkali-Richtlinie is of outstanding
importance. Many damages on German highways (Autobahnen) were caused by
the long term reaction of alkaline elements released from the cement together with
reactive silica from sand and gravel (ASRalkali-silica reaction). It is the aim of
this regulation to prevent further damages based on this reaction.
These standards and regulations restrict the use of sand and gravel in concrete.
They are generally available and can not be reduced down to a few paragraphs and
a single table. Therefore, this chapter will not describe the quality requirements of
aggregates for the use in concrete.
64
H. B. Walther
support of the fresh mortar structure, thus the use of natural quartz sand is limited.
Nevertheless, quartz sand and broken limestone are equal in this application.
Coarse quartz grains cause structural effects in the render and special products
contain grains up to 6 mm size.
Floor pavements on the basis of cement also belong to the mortar group.
The requirements of this industry in context of grain size distribution are
manifold but for the single product quite detailed. The relevant European standard
is DIN EN 13139. The preparation of mortar and render on site by the craftsmen
mixing sand and cement by volume in a mortar mixer is decreasing year by year.
Plants producing premixed dry mortar and render have been established also in
Eastern Europe.
Sand fractions used in render and mortar are characterized by narrow grain size
distributions like 0.30.5 mm, especially in thin bed mortar. Mortar layers on
exposed masonry have a normal thickness of 12 mm and require grains of a size up
to 4 mm. Sand fractions used for render are rather fine grained, render with
structural effects contains several grains of 23 mm diameter, sometimes even
grains of 6 mm diameter may be used. Floor screeds require sand with a main
fraction of 08 mm.
However, all of these materials must be dried and should have moisture contents clearly below 1 wt%. The material should be free of clay minerals because
they will cause high rates of shrinkage resulting in cracks.
A special requirement of this industry is the increasing demand on very light
coloured white sands. These very pure sands are not only used for render, but are
requested increasingly for white mortar. Market leader of white masonry blocks
demand white mortar for combined supply with the blocks.
No tabulated values were given, because of the wide variety of requirements
and the existence of a standard for the raw materials.
65
66
H. B. Walther
Total amounts in wt-%
70
60
90
calcite + dolomite
quartz
80
70
50
30
60
wt-%
40
50
40
30
20
20
10
10
losses
quartz relative
carbonate rel.
Fig. 3.4 Left: total amounts of carbonates and quartz in the single fractions. Right: slot 1 until
the named slot were discarded, contents of the processed material and theoretical losses
Fig. 3.5 Feeding material for autogenous milling; coarse fraction left and fine fraction (wet) right
in the dry processed material was increased from 64 to 76%, accepting losses of .
This was achieved without the use of any chemicals.
67
fraction within the fine material. The main disadvantage of this method is that it is
not possible to achieve a constantly fine grained product, which is the aim of
milling. Step by step the milling systems are now changed into ball mills.
Some applications, for example in the chemical industry, require low iron
contents in the milled sand. Consequently, steel balls must not be used in grinding.
Alternatively, balls from corundum or even ZrO2 are used. Very high requirements
concerning the chemical purity of quartz sand require grinding with silica balls,
normally flint balls. These grinding balls are commonly no spheres but rather
ellipsoids. This use of flint in the mill is also called autogenous milling, because of
the chemical and mineralogical identity of quartz sand and flinthowever, true
autogenous milling requires the same quartzite for grinding-boulders and fine
material as well.
Two main principles are applied in grinding: dry and wet milling. For dry
milling, the material has to be dried before the grinding process. After grinding,
the fine material can be stored in silos and can easily be shipped. Wet milling saves
the drying energy but instead of powder a slurry is produced which has to be
stirred continuously. Wet mills are normally plated with rubber and are not as
noisy as steel plated dry mills. Furthermore, there is no quartz dust problem. Main
disadvantage is the necessity of stirring the slurry and the impossibility of shipping. The ground slurry has to be used on site.
The finest material can be achieved by wet milling. But use of economically
prized steel balls for fine grinding introduces a new problem. If ground to long, the
metallic iron abrasion oxidizes and takes the oxygen out of the water releasing
hydrogen. Consequently, only balls from corundum or ZrO2 can be used.
68
H. B. Walther
Southern Europe was affected by the Alpine orogenesis causing the formation
of numerous mountains and massive plains of debris. Neither valleys in mountain
regions do exhibit quartz sand deposits nor river plains in their surroundings. If
sand deposits exist, their phase composition is controlled by the weak source rocks
in the mountains and by weathering products in the river plains. Only deposits of
Tertiary and older age contain suitable sand. The efforts to produce high silica
products are immense; for instance the company Sibelco mines massive quartzite
in the northwest of Italy, close to the French border (Fig. 3.6). The material has to
be crushed, ground and purified by grain size fractionation and magnetic sorting
before selling. Even glass sand quality can be produced in that way.
Istria is dominated by massive lime sediments but containing one sand layer of up
to 5 m thickness, which is sometimes solidified by calcareous cement. This sand
material was mined, processed by flotation and exported from Yugoslavia to Italy.
The raw material was mined in the last underground mine of Croatia until 2000
(Fig. 3.7). After the political change, the material was not flotated any longer but only
broken and solely used for the production of AAC in the YTONG factory of Pula.
69
Physical use
Physical and
chemical use
Chemical use
Clay bricks
Calcium silicate
units
Ceramic products
(pottery)
Autoclaved aerated
concrete
Cement
Concrete products
Foundry sand
Grit
Mortar and render
Chemicals
Glass
3.5 Conclusions
The different scenarios of usage within individual branches call for different raw
material characteristics. As long as the grain is used in its original state, its size and
shape is a characteristic feature. The grain size is of minor importance, if the sand
is ground and used chemically. An overview about the characteristic uses is given
in the following Table 3.9.
Quartz sands are valuable raw materials for the building and construction
industry. The individual branches demand very different sand qualities and the
requirements grow. Consequently, there is a strong need for information concerning potential raw materials. The acquisition of this data exceeds the capability
of individual producers. Any data about mineralogical and chemical composition
as well as granulometric data about sand deposits and occurrences are worthy of
publication.
Acknowledgments I would like to thank numerous colleagues for their help, information and
stimulating discussions. Special support was given by W. Eden, U. Jakobs, M. Kanig, W. Krcmar,
H.-J. Riechers, T. Schoch, C. Wertel and, last but not least, H. Wopfner. Comments by T. Gtte
and an anonymous reviewer led to a considerable improvement of the original manuscript.
References
Anonymous (1985) Gemeinsamem Runderlass vom 23.09.1985 des Umwelt- und des
Wirtschaftsministeriums NRW
Anonymous (2011) Verwendung von Kies und Sand im berblick. http://www.bks-info.de/
images/ks-baust/verwendung_ueberblick.gif. Accessed 30 May 2011
Bertran A (1995) DE 43 39 916 A1, published 24.05.1995, Zusammensetzung zur Herstellung
von Baustoffen mit schalldmmender Wirkung
Blankenburg H-J, Gtze J, Schulz H (1994) Quarzrohstoffe. Leipzig and Stuttgart
DAfStb Alkali-Richtlinie 2007-02 Vorbeugende Manahmen gegen schdigende Alkalireaktion
im Beton (Alkali-Richtlinie)Teil 1: AllgemeinesTeil 2: Gesteinskrnungen mit Opalsandstein und FlintTeil 3: Andere alkaliempfindliche Gesteinskrnungen
DIN 1045-2 2008-08 Tragwerke aus Beton, Stahlbeton und Spannbeton- Teil 2: Beton
Festlegung, Eigenschaften, Herstellung und KonformittAnwendungsregeln zu DIN EN
206-1
70
H. B. Walther
Chapter 4
71
72
A. Mller et al.
4.1 Introduction
High-purity quartz (HPQ), which is generally defined as quartz containing less
than 50 lg g-1 of contaminating elements (Harben 2002; Fig. 4.1), is a valuable
commodity used in a wide range of high-technology products. Demand for HPQ is
increasing strongly due to the rapid development and expansion of the HPQconsuming industry. Security of supply necessitates the identification and characterization of new HPQ deposits in more countries, particularly in Europe, possibly including deposits of a different kind compared to those currently in
production. Prerequisites for developing exploration tools for such deposits are the
application of state-of-the-art microanalytical methods for appropriate petrological
and chemical characterisation of potential deposits in order to achieve a better
understanding of the environment and conditions of HPQ formation.
Industrial HPQ products are commonly delivered as processed fine-grained
sand. Therefore, all types of medium- to coarse-grained quartz-rich ([20 wt.%)
rocks possess, theoretically, the potential for being HPQ deposits, given that the
deposit is large enough ([80,000 t; Fig. 4.1) and the other minerals of the rock can
be separated by conventional processing. Impurities within quartz crystals (intracrystalline impurities) are the principal control on the quality of quartz because
they can be removed by processing to some extent only. The impurities which are
discussed in the first part of this study include: (i) lattice-bound trace elements,
(ii) submicron inclusions \1 lm, and (iii) mineral and fluid micro-inclusions
([1 lm). In addition, intercrystalline impurities may occur along grain boundaries
in the form of mineral coatings or micro-crystals. This type of impurities are not
discussed further because they can be eliminated if the raw material is crushed
down to its average crystal size, allowing the removal of these impurities by
processing.
High-purity quartz (HPQ) is often thought of as a relatively easily available
commodity. In fact, HPQ is very rare in economic and near-economic quantities
([80,000 t), but at microscopic scale (\1 mm) it is rather common and occurs as
recrystallised and neocrystallised micro-domains (secondary quartz) within quartz
crystals (e.g. Mller et al. 2000; Van den Kerkhof and Hein 2001; Van den
Kerkhof et al. 2004; Mller et al. 2008a). Another reason for this misconception is
uncertainty about the definition of HPQ. A revised definition of HPQ based on
previous classifications by Harben (2002) and Mller et al. (2007) is therefore
suggested in the second part of this study.
In the third part of the paper examples of potential economic and economic
Norwegian HPQ deposits are introduced. Some of the deposits are in operation,
others are under investigation. The deposits described represent a wide range of
genetic environments, including hydrothermal, igneous and metamorphic settings.
Attention is paid to the micro-inclusion inventories of the quartz and how they can
be described, since micro-inclusions are the major contaminants of HPQ products.
Finally, aspects of HPQ formation in the Norwegian deposits are discussed.
73
Fig. 4.1 Classification of chemical quartz qualities (hyper to low) and their approximate price
range according to Harben (2002). High-purity quartz is defined containing 850 lg g-1
impuritites (grey shaded field). The lower limit of 8 lg g-1 is represented by the quartz product
IOTA 8 which is the puriest quartz product on the market produced from natural quartz (IOTA
2011). The approximate economic limit of the deposit size bases on data of active quartz mines
74
A. Mller et al.
Fig. 4.2 Schematic quartz structure showing the configuration of trace elements in the quartz
lattice (modified from Gtze 2009). McLaren et al. (1983) proposed the substitution of Si4+ by
four H+ is possible (silanol groups). Because of the two-dimensional illustration the fourth H+ is
not shown on the figure
have shown that electron defects have only a minor or no contribution to the charge
balance (Mller and Koch-Mller 2009). The equation is important for the determination of the chemical quality of quartz. Aluminium is the most common trace
element in natural quartz, in concentrations of up to several thousand lg g-1 and
can thus be more easily determined by analytical methods than the other elements.
If high Al concentrations are detected then concentrations of Li, K, Na and H will
be high as well and possibly also the concentrations of B and P. The Al concentration in quartz is thus an important quality indicator.
Aluminium, Fe, and alkali metals tend to form minute atomic clusters which are
incorporated along specific growth axes: this is especially true for low-temperature
quartz (Pfenninger 1961; Flicstein and Schieber 1974; Siebers 1986; Ramseyer
and Mullis 1990; Brouard et al. 1995). Investigations of hydrogen isotopes by
Simon (2001) documented two H-reservoirs in quartz: (i) fluid inclusions and
(ii) structurally-bound water in small, homogeneously distributed micro-clusters
and bubbles. According to Simons (2001) model, such micro-clusters should be
accompanied by high Al3+ and K+, Na+, Li+ abundances. The possible configuration of such a cluster is illustrated in Fig. 4.3. The relatively high abundance of
H-compensated Al defects and structurally bound water molecules observed by the
Fourier transform infrared spectroscopy (FTIR) may indicate the existence of such
micro-clusters (e.g. Mller et al. 2003a; Mller and Koch-Mller 2009).
75
76
A. Mller et al.
Fig. 4.4 Average abundance and variations of trace elements in natural quartz. Data shown in
grey are from Gerler (1990), Blankenburg et al. (1994) and Gtze (2009). Data shown in black
represent 2117 LA-ICP-MS analyses of quartz carried out at NGU over the last 6 years.
Concentrations of Cl are uncertain due to the very high detection limit of about 100 lg g-1. Cl is
presented because Gerler (1990), Blankenburg et al. (1994), and Gtze (2009) suggested that it
occurs in considerable amounts in quartz
77
78
A. Mller et al.
Fig. 4.5 a Trace element evolution of quartz in granitic rocks. Quartz found in dacitic/dioritic
rocks is characterised by high Ti and moderate Al. During further magmatic evolution Ti in
quartz decreases and Al increases. Aluminium and Ti contents of igneous quartz plot far from the
HPQ field (modified from Mller et al. 2010a). Data were acquired using EPMA and SIMS.
b Compared to igneous quartz shown in a, pegmatite quartz has low Ti due to the generally low
crystallization temperatures. Some of the Niobium-Yttrium-Fluorine (NYF)-type pegmatites plot
in the HPQ field. Lithium-Cesium-Tantalum(LCT)-type pegmatites are characterized by high Al.
Data were acquired using LA-ICP-MS
79
Fig. 4.6 Iron concentration profiles of quartz crystals at the contact of Fe-rich minerals. The
profiles illustrate that high Fe concentrations in quartz are predominantly diffusion-controlled.
a Quartz crystal enclosed in tourmaline (schorl) (modified from Mller et al. 2005c).
b Hydrothermal quartz in the contact with pyrite (modified from Mller et al. 2010b). c Igneous
quartz in contact with biotite (modified from Mller et al. 2002b)
80
A. Mller et al.
Fig. 4.7 BSE images of submicron inclusions in quartz (with permission from Seifert et al.
2011). a Submicron rutile needles (whiskers) in blue quartz from Broken Hill, Australia.
b Distribution of submicron inclusions mainly comprising mica and minor rutile in blue quartz
from the Faja de Eruptiva Oriental rapakivi granite, NW Argentina. The density of
submicrometre inclusions corresponds to 8,910 particles per mm2
81
zoned blue-quartz grains resulted in 1.11.7 particles per lm3 in the cores and
0.40.6 in the rims.
Mller et al. (2002b) interpreted AlK spikes observed in EPMA profiles across
quartz grains in deformed granites from the Lachlan Fold Belt, Australia, as
submicron inclusions of mica-like composition. They concluded that multiple
deformation of quartz caused the redistribution of Al and K in the quartz lattice,
which results in the accumulation of these elements in submicron inclusions
(\0.5 lm) of muscovite-like composition.
In summary, submicron inclusions may occur in high densities in quartz
crystallised in specific environments: where these inclusions are present, they are
the major source of contamination. Further systematic studies of submicron and
nano-inclusions in quartz are needed.
82
A. Mller et al.
concentrations of F, Cl, B, P, Li, Cs, and Rb, up to several weight percent (e.g.,
Thomas et al. 2006) and, thus can be the major source of contamination, e.g. of B
and P in quartz raw materials produced from pegmatitic quartz.
In theory, all mineral phases which occur in the host rock may also occur as
micro-inclusions in quartz. Mineral species included in igneous quartz are mainly
feldspar, mica, rutile, zircon, apatite, Fe oxides, etc. (Roedder 1984; Leeder et al.
1987). The spectrum of mineral inclusions in metamorphic quartz depends on the
conditions of metamorphism. Whereas mineral inclusions of chlorite, muscovite or
amphibole are more characteristic of low-grade metamorphic rocks, kyanite,
staurolite or garnet occur especially in high-grade metamorphic rocks. Inclusions
of anhydrite, gypsum, polyhalite, calcite, several salt minerals, organic matter, etc.
have been described in sedimentary authigenic quartz (e.g. Richter 1971; Fruth and
Blankenburg 1992; Hyrsl and Niedermayr 2003; Gtze 2012). The formation of
mineral inclusions in quartz is manifold. They were enclosed during growth from
melt and fluids or by solid-state grain boundary migration during metamorphism
and subsequent crystal lattice recovery. The latter process seems to be very
common in metamorphic rocks since mineral inclusions are most common in
quartz of quartzites. Exsolution of rutile needles as a result of cooling or
decompression of Ti-rich quartz is the third process leading to the formation of
mineral inclusions (e.g., Adachi et al. 2010).
83
and Ni give a colouration of the silica glass that reduces its transmission properties. Both P and B are unwanted in photovoltaic or semiconductor products
manufactured from quartz. For these reasons upper concentration limits for the
most critical trace elements should be included in a revised HPQ definition. The
classification should be applicable not only to processed quartz raw material but
also to unprocessed natural quartz in order to allow useful classification of
potential quartz deposits. For the exploration of HPQ deposits it is important to
characterise unprocessed quartz by in situ analysis of quartz crystals.
Mller et al. (2007) made a first attempt at HPQ classification based on the
concentration limits of certain elements. They suggested upper concentration
limits of 25 lg g-1 for Al and 10 lg g-1 for Ti. With the background knowledge
mentioned above, together with trace element concentrations of HPQ products
available on the market, a more comprehensive definition of HPQ is suggested.
The sum of the nine elements Na, K, Li, Al, Ca, Fe, Ti, B and P analysed on quartz
crystals (single grain analysis) or processed quartz sand (bulk product analyses)
should be \50 lg g-1. The maximum content of each element is suggested as: Al
\30 lg g-1, Ti \10 lg g-1, Na \8 lg g-1, K \8 lg g-1, Li \5 lg g-1, Ca \5
lg g-1, Fe \3 lg g-1, P \2 lg g-1 and B \1 lg g-1 whereby the sum of all
elements should not exceed 50 lg g-1. Figure 4.8 illustrates the suggested upper
concentration limits of HPQ compared with average abundances in natural quartz
and the concentrations of the IOTA standard of these nine elements. The proposed
upper concentrations limits for Na, K and Ca are relatively high compared to their
average abundance in natural quartz, because in processed quartz the concentrations are superimposed by contributions from fluid inclusions (containing NaCl
and KCl), and mica and feldspar micro-inclusions (containing K, Na and Al)
which are the most common intracrystalline impurities. Other elements that also
occur in the HPQ products in quantities of up to a few lg g-1 (e.g. Mg, Zr and Ge)
are rarely given attention, because they have only minor effects on silica glass
properties. The content of lattice-bond H might be several tens of lg g-1 but is not
84
A. Mller et al.
included in the definition because there are no published requirements and because
of the analytical challenges in quantifying the H concentration.
It should be mentioned that trace element concentrations in HPQ products do
not necessarily represent hundred percent lattice-bound elements if processed HPQ
sand is analysed, for example by solution ICP-MS. Portions of the element concentrations might originate from inclusions or other minerals which were not
completely removed during processing. The analysis of inclusions or foreign
minerals can be avoided by the application of in situ micro-beam techniques on
single quartz crystals such as EPMA, LA-ICP-MS or SIMS.
4.4 Methods
4.4.1 Laser Ablation Inductively Coupled Plasma Mass
Spectrometry (LA-ICP-MS)
Concentrations of Li, Be, B, Al, P, K, Ti, Mn, Fe, and Ge were analysed in situ by
laser-ablation inductively-coupled plasma mass spectrometry (LA-ICP-MS). The
analyses were performed on the double-focusing sector field mass spectrometer
model ELEMENT 1 from Finnigan MAT which is combined with the NewWave
193-nm laser probe. The laser had a repetition rate of 15 Hz, a speed of
15 lm s-1, a spot size of 75 lm, and energy fluence of about 14 mJ cm-2 on the
sample surface. Raster ablation was applied in the centre of quartz crystals on an
area of approximately 225 9 300 lm. The approximate depth of ablation was
between 40 and 100 lm depending on the crystallographic orientation and
absorption behaviour of the individual quartz crystals. The carrier gas for transport
of the ablated material to the ICP-MS was He mixed with Ar. External calibration
was performed using four silicate glass reference materials (NIST SRM 610, 612,
614, 616), the NIST SRM 1,830 soda-lime float glass, the BAM No.1 amorphous
SiO2 glass and the synthetic pure quartz monocrystal Qz-Tu. Certified, recommended and proposed values for these reference materials were taken from the
certificates of analysis where available, or otherwise from the web site Geological
and Environmental Reference Materials (GeoReM 2011). The isotope 29Si was
used as the internal standard. A linear regression model, including several measurements of the different reference materials, was used to define the calibration
curve for each element. For the calculation of P concentrations, the procedure of
Mller et al. (2008b) was applied. Ten sequential measurements on the SiO2
blank crystal were used to estimate the limits of detection (LOD) which were
based on 3 9 standard deviation (3r) of the ten measurements. LODs vary for each
analysis sequence (measurement day). Examples of LODs are given in Table 4.2.
The analytical error ranges within 10% of the absolute concentration of the
element. More details of the measurement procedure are provided by Flem et al.
(2002).
85
86
A. Mller et al.
Fig. 4.9 Simplified geological map of the Melkfjell area (modified from Gustavson and Gjelle
1991). The outline of the Melkfjell quartzite is according to Wanvik (2009)
the quartzite belt turns eastwards and widens to 600 m, forming the most interesting section from an economic point of view (Fig. 4.11). Several quartzite bodies
of similar lithology and tectonostratigraphic position occur in the area (Wanvik
2001). Wanvik (2009) distinguished white-weathering and brownish-weathering
quartzites. The latter is characterized by a higher Fe content. The average bulk
composition of the Melkfjell quartzite is given in Table 4.1. The quartzite contains
minor muscovite and accessory biotite, feldspar and graphite. The quartzite is
commonly cross-cut by pegmatite and amphibolite veins. The amphibolite veins
are between 0.1 and 10 m wide end extend up to lengths of 100 m. Irregular lenses
and veins of pegmatite are 0.11 m in size and have quartz-dioritic to trondhjemitic compositions. The crystal sizes of quartz and feldspar are 12 cm. The
pegmatites become dominant in the quartzite at the eastern end of the quartzite
body, east of the Melkfjelltjnna (Fig. 4.11). The high pegmatite proportion makes
this part of the Melkfjell quartzite uneconomic. However, several million tons of
quartzite are present in the more pure and homogeneous parts of the deposit
(Wanvik 2009).
87
Fig. 4.10 a Outcrops of the Melkfjell quartzite in the foreground with Melkfjell to the left. View
towards W. b Outcrop of kyanite quartzite at Juovvacorr in Skjomen. The bluish shade of the
rock is caused by fine-grained kyanite. c Contact of the massive quartz core and the feldspardominated wall zone of the Nedre yvollen pegmatite exposed in the underground mine. The
short edge of the photograph corresponds to 2 m. d Exposure of the Nesodden quartz vein with
the Hardangerfjord in the background. The yellow dashed line marks the outline of the Nesodden
vein. View towards NE. e Outcrop of the Kvalvik quartzite with irregular, cross-cutting quartz
veins of high-purity quality. f Temporary outcrop of the Svanvik quartz vein. The quartz block in
the foreground is about 1 m high
Quartz petrography. The quartz crystals range in size from 10 to 2,000 lm with
an average of about 500800 lm. The quartz is strongly recrystallised and
large grains show undulatory extinction. Inclusions in quartz are very common.
88
A. Mller et al.
Fig. 4.11 Detailed geological map of the Melkfjell quartzite according to Wanvik (2009)
Table 4.1 Average whole rock composition of the Melkfjell quartzite according to Wanvik
(2009)
Al2O3
Fe2O3
CaO
MgO
Na2O
K2O
TiO2
P2O5
SiO2
98.05
0.67
\0.41
0.02
0.11
0.09
0.11
0.02
0.02
The inclusions observed are (in order of frequency) biotite and muscovite
(202,000 lm), albite (1002,000 lm), plagioclase (202,000 lm), apatite
(1050 lm), rutile (\110 lm), monazite (10100 lm), zircon (1040 lm),
pyrite (1050 lm), Fe-oxides (520 lm) and arsenopyrite (10 lm; Figs. 4.12a,
4.13a). CL imaging revealed that grain-boundary migration is a widespread phenomenon, whereby grains with low CL intensity replacing large grains with higher
CL intensity (Fig. 4.14).
Quartz crystal chemistry. The quartz of the Melkfjell quartzite is characterized
by low concentrations of Li (mean 1.1 lg g-1), B (mean 0.8 lg g-1) and Al (mean
7.7 lg g-1; Table 4.2; Fig. 4.15). The Ti concentration is moderately high due to
high-temperature Caledonian metamorphism. Application of the Ti-in-quartz
geothermometer by Wark and Watson (2006) reveals the peak metamorphic
temperature of about 520 8C (Ti = 8.76 lg g-1) which is somewhat higher
than the maximum temperature of 506C determined by Mller and Koch-Mller
(2009). Ti saturation is assumed due to the common exsolution of rutile needles
(Fig. 4.12). The analysis with the low Ti content of 2.09 lg g-1 was from neocrystallized quartz with low CL intensity, reflecting retrograde crystallization
conditions. This low-Ti quartz shows also the lowest Al and Li contents.
89
Fig. 4.12 Optical microscope images of micro inclusions in quartz. a Rutile needles in quartz
from Melkfjell. b Rutile and zircon inclusions in quartz from the Gullsteinberg kyanite quartzite.
c Silicate melt inclusion in pegmatite quartz from Nedre yvollen. d Fluid inclusions with gas
bubble and solids (presumably halite) in quartz from the Nesodden vein. e Fluid inclusion with
gas bubble and solids (presumably halite and sylvite) in quartz from the Kvalvik deposit.
f Liquid-rich fluid inclusions in quartz from the Svanvik vein
Economic assessment and remarks on HPQ formation. The common occurrence of rutile needles in the Melkfjell quartz will make it very hard to achieve a
low Ti content in the final, processed product and, thus, only applications with
low-quality requirements for Ti content are feasible. The common occurrence of
apatite and monazite as interstitial accessories and inclusions in quartz will cause
elevated P concentrations in the product, limiting possible photovoltaic applications. However, the quality of the quartzite is adequate for ferrosilicon applications
if contaminating rocks of amphibolite and the pegmatite dykes are properly handled (Wanvik 2009). Apart from its Ti and P contents the quartz might be of HPQ
90
A. Mller et al.
Fig. 4.13 Backscattered electron images of micro inclusions in quartz. a Polyphase inclusion in
quartz of the Melkfjell quartzite comprising weathered biotite, monazite and pyrite. b Muscovite
and baryte inclusion in quartz of the Tverrdal kyanite quartzite. c Albite inclusion in the Nedre
yvollen pegmatite quartz. d Inclusion of altered muscovite in the Nesodden quartz. e Inclusion
of altered muscovite in the Kvalvik quartz. f Inclusion of altered muscovite in the Svanvik quartz
91
Fig. 4.14 Cathodoluminescence images of quartz. a Dull luminescent, weakly zoned quartz
grain bulged into and replaced bright luminescent quartz by grain boundary migration. The arrow
indicates the growth direction. Melkfjell quartzite. b Dull luminescent, weakly zoned quartz
grains replaced bright luminescent quartz by grain boundary migration. Tverrdal kyanite
quartzite. c Biotite inclusions in bright and homogeneous luminescent quartz in the Nedre
yvollen pegmatite. d Strongly overprinted quartz from the Nesodden vein. The bright
luminescent areas are remains of the primary, unaltered hydrothermal quartz (pqz). The tiny,
bright dots are inclusions of mica-like composition. e Strongly overprinted quartz from the
Kvalvik deposit. The bright luminescent areas are remains of the primary, unaltered hydrothermal
quartz (pqz). The white dots are tiny (\1 lm) mica inclusions. The CL structures are similar to
those in the Nesodden quartz. f Strongly altered quartz from the Svanvik vein with relic domains
of primary hydrothermal quartz (pqz). The altered quartz is cross-cut by a younger partially
healed quartz vein
Al and Li contents in the newly formed quartz. The Melkfjell quartzite was
exposed to pressures between 3 and 5 kbars (Stephens et al. 1985) and temperatures of 466528C (Mller and Koch-Mller 2009 and this study) and, thus, the
LOD
0.80
(example)
Melkfjell quartzite
\0.16
1.18
0.89
2.19
1.25
(0.51)
Average
1.13 0.73
Kyanite quartzites
Gullsteinberg
0.81
\0.8
\0.8
Average
\0.80
Skjomen
1.84
1.80
1.70
Average
1.78 0.07
Tverrdal
\0.80
0.86
\0.80
Average
\0.82
Nasafjell
\0.80
\0.80
\0.80
\0.10
\0.10
\0.10
\0.10
\0.10
\0.10
\0.10
\0.10
\0.10
\0.10
\0.10
\0.10
\0.10
\0.10
\0.10
0.25
\0.13
\0.13
\0.13
\0.13
(0.21)
\0.15
0.10
\1.3
\1.3
\1.3
\1.3
\1.3
\1.3
\1.3
\1.3
\1.3
\1.3
\1.3
\1.3
\1.3
\1.3
\1.3
1.95
0.72
0.50
0.47
0.35
(0.43)
0.80 0.66
1.3
25.2
18.6
27.7
23.8 4.7
13.8
29.2
13.1
18.7 9.1
15.0
17.5
10.0
14.2 3.5
10.2
9.9
15.1
6.9
11.2
\6.6
7.1
\6.6
(25.3)
\7.7
6.0
0.9
Ti
10.49
10.22
9.64
10.12 0.44
3.55
4.72
9.30
5.86 3.04
13.61
10.28
8.02
10.64 2.81
2.70
2.07
2.60
\2.3
7.87
\2.3
8.76
\2.3
2.09
\2.3
5.59
\2.3
7.75
(\2.3) (28.86)
\2.3
6.41 2.68
0.5
\3.1
1.0
\3.1
\0.5
\3.1
0.7
\3.1
\0.7
3.8
\0.5
5.3
17.4
3.9
\0.5
4.3 0.9 \6.1
\3.1
\0.5
\3.1
\0.5
\3.1
\0.5
\3.1
\0.5
\3.1
\0.5
\3.1
0.8
\3.1
\0.5
4.5
\1.8
6.1
2.5
3.4
(3.3)
\3.7
3.1
Table 4.2 Trace element concentrations (lg g-1) of quartz crystals determined by LA-ICP-MS
Li
Be
B
Al
P
K
\0.20
0.24
\0.20
\0.21
\0.20
\0.20
\0.20
\0.20
\0.20
\0.20
\0.20
\0.20
\0.20
\0.20
0.24
0.12
0.32
0.32
0.34
0.19
(0.11)
0.26 0.1
0.2
Mn
0.40
0.10
Ge
\0.40
\0.40
\0.40
\0.40
\0.40
2.07
\0.40
\0.96
\0.40
\0.40
\0.40
\0.40
\0.40
\0.40
2.51
(continued)
0.29
0.24
0.27
0.27 0.03
0.40
0.52
0.53
0.48 0.07
0.44
0.48
0.38
0.44 0.05
0.36
0.48
0.57
\1.20 0.54
\1.20 0.81
\1.20 0.82
\1.20 0.99
\1.20 1.00
(\1.20) (0.70)
\1.20 0.83 0.19
Fe
92
A. Mller et al.
Be
Al
Average
\0.80
\0.10
\1.3
11.7 2.9
Nedre yvollen pegmatite quartz
4.89
0.36
1.82
14.4
4.08
0.24
1.50
30.4
5.87
0.33
1.44
21.4
4.83
0.23
1.52
16.4
4.89
0.36
1.82
14.4
4.08
0.24
1.50
30.4
Average
4.92 0.74
0.29 0.06 1.57 0.17
20.6 7.1
Nesodden hydrothermal quartz vein
4.26
\0.10
1.42
16.7
3.62
\0.10
1.25
14.6
6.91
\0.10
1.64
29.8
7.44
\0.10
1.25
31.4
3.25
\0.10
0.81
18.4
5.58
\0.10
1.47
13.7
7.29
\0.10
1.40
23.4
5.32
\0.10
1.46
11.0
7.34
0.14
1.28
12.1
6.34
\0.10
1.55
15.0
7.83
\0.10
1.39
18.4
4.66
\0.10
1.40
28.9
5.31
\0.10
2.23
6.9
5.22
\0.10
1.51
12.0
(3.53)
(\ 0.10)
(1.80)
(61.3)
(4.16)
(\ 0.10)
(1.53)
(149.2)
\2.8
1.59
\2.8
2.18
\2.8
2.02
\2.8
2.03
\2.8
1.74
\2.8
2.42
\2.8
3.62
\2.8 \1.60
\2.8
1.66
\2.8
1.76
\2.8
1.94
12.9
3.45
\2.8
1.96
\2.8
2.02
(20.8) (2.32)
(91.5) (4.45)
\5.1
\5.1
\5.1
\5.1
\5.1
\5.1
\5.1
5.4
\5.1
\5.1
\5.1
\5.1
\5.1
\5.1
(\ 5.1)
(\ 5.1)
5.46
1.04
4.47
\1.00
5.46
1.04
\2.99
\3.2
\3.2
\3.2
\3.2
\3.2
\3.2
\3.2
\2.4
\2.4
\2.4
\2.4
\2.4
\2.4
\2.4
2.46 0.34
Ti
\0.6
\3.1
\0.40
\0.40
\0.40
\0.40
\0.40
0.44
0.47
0.70
\0.40
\0.40
\0.40
0.43
0.47
\0.40
(\ 0.40)
(\ 0.40)
\0.50
\0.50
\0.50
\0.50
\0.50
\0.50
\0.50
\0.21
Mn
1.25
0.47
0.70
0.36
1.25
0.47
0.70 0.40
0.47 0.11
Ge
(continued)
\1.20 1.02
\1.20 0.99
\1.20 0.86
\1.20 1.04
\1.20 1.08
\1.20 0.97
\1.20 1.14
\1.20 0.89
\1.20 0.91
\1.20 0.92
\1.20 1.00
\1.20 0.94
\1.20 0.92
\1.20 1.16
(1.26) (1.13)
(11.41) (1.05)
\1.00
\1.00
\1.00
\1.00
\1.00
\1.00
\1.00
\1.10
Fe
Be
\1.3
\1.3
\1.3
\1.3
1.9
8.3
5.7
5.3 3.2
\3.1
\3.1
4.7
\3.6
\5.1
\5.1
\5.1
\5.1
\5.1
\5.1
\5.1
(\ 5.1)
\5.1
16.5
12.8
15.1
14.3
16.7
20.8
13.6
(59.7)
15.7 2.7
1.71
1.59
1.47
1.20
1.14
1.70
2.06
(1.73)
1.55 0.32
(101.2)
(\ 5.1)
(134.8)
(\ 5.1)
18.02 7.6 \5.1
Al
(1.99)
(1.94)
1.43 0.3
B
(3.24)
(3.34)
2.14 0.64
Ti
\0.5
\0.5
\0.5
\0.5
1.10
\0.90
1.36
\1.12
\2.7
2.69
\2.7
3.09
\2.7 \1.60
\2.7 \1.6 0
\2.7
1.77
\2.7
1.92
\2.7
2.63
(21.0) (3.30)
\2.7 \2.19
(13.1)
(35.5)
\3.5
\0.20
\0.20
\0.20
\0.20
\0.40
\0.40
\0.40
\0.40
\0.40
0.53
\0.40
(\ 0.40)
\0.40
(0.59)
(0.50)
\0.44
Mn
Ge
\0.40
\0.40
\0.40
\0.40
0.18
0.15
0.23
0.19 0.04
\1.23 0.79
\1.23 0.91
\1.23 0.77
\1.23 0.84
\1.23 0.91
\1.23 0.83
\1.23 0.64
(\ 1.23) (1.00)
\1.23 0.81 0.09
(1.20) (1.01)
(1.20) (1.13)
\1.20 0.99 0.09
Fe
Data on kyanite quartzites and Svanvik are from Mller and Koch-Mller (2009) and from Melkfjell from Wanvik (2009). Note that the limits of detection
(LOD) are different for different analytical sequences. LODs given are examples of the kyanite quartzite measurement sequence. Analyses in parenthesis are not
considered in the calculation of the average value because these analyses are superimposed by mineral inclusions, such as rutile (Ti) and mica (Al, K)
(4.55)
(\ 0.10)
(3.56)
(\ 0.10)
Average
5.74 1.48 \0.10
Kvalvik hydrothermal quartz
3.32
\0.10
3.33
\0.10
3.22
\0.10
3.63
\0.10
4.12
\0.10
7.06
\0.10
5.47
\0.10
(3.69)
(\ 0.10)
Average
4.23 1.45 \0.10
Svanvik hydrothermal quartz vein
2.12
\0.10
2.11
\0.10
2.17
\0.10
Average
2.14 0.03 \0.10
94
A. Mller et al.
95
96
A. Mller et al.
Fig. 4.16 a Locations of kyanite quartzite occurrences in Norway. b Geological map of the
kyanite quartzite occurrence at Gullsteinberg in Solr according to Jakobsen and Nielsen (1977)
considerably by this process, from about 72 lg g-1. The expelled Ti may have
crystallised as sub-micron rutile needles which are common in the Melkfjell quartz
(Fig. 4.12a). From an economic point of view the disadvantage of the widespread
grain-boundary migration is the enclosure of minor and accessory minerals by
quartz, resulting in a high density of mineral inclusions in the quartz.
97
98
A. Mller et al.
Table 4.3 Represenstative whole rock analyses (XRF) of kyanite quartzites in Norway
Gullstein-berg Knsberg Kjeksberg Sormbru Tverrdal Skjomen Nasafjell
Major elements (wt.%)
76.09
SiO2
Al2O3 17.7
Fe2O3 1.42
0.32
TiO2
MgO
0.07
CaO
0.01
Na2O \0.1
0.63
K2O
MnO \0.01
0.02
P2O5
LOI
2.32
total
98.58
Trace elements (lg g-1)
Ba
422
Ga
11
Zn
17
Cu
\10
Ni
\5
Co
5
Ce
\10
La
\10
Nd
\10
W
47
Pr
\10
Mo
\5
Nb
21
Zr
213
Y
\5
Sr
26
Rb
22
Th
\5
Pb
\10
Cr
68
V
\10
As
8
Hf
\10
S
0.32
75.89
19.39
0.71
0.23
0.05
\0.01
\0.1
0.05
\0.01
0.04
1.74
98.11
82.54
12.76
1.67
0.48
\0.01
\0.01
\0.1
\0.01
\0.01
0.05
1.17
98.63
77.37
13.78
1.41
0.32
0.07
0.02
0.2
3.64
\0.01
0.04
2.38
99.24
82.67
14.28
0.13
0.44
0.06
0.01
\0.1
0.15
\0.01
0.03
0.62
98.33
77.68
20.8
0.09
0.18
0.03
\0.01
\0.1
0.05
\0.01
0.02
0.21
99.05
76.27
20.84
0.54
0.18
0.05
0.01
\0.1
0.32
\0.01
0.07
0.56
98.78
74
\10
8
\10
\5
8
\10
\10
\10
24
\10
\5
15
119
\5
47
\5
7
\10
17
19
\5
\10
0.32
1722
\10
10
10
\5
\5
\10
11
\10
57
\10
\5
16
251
\5
92
\5
\5
\10
27
29
5
\10
0.36
646
\10
13
\10
5
\5
104
63
50
35
14
7
13
227
7
73
78
12
21
65
23
\5
\10
0.32
399
\10
7
\10
\5
8
10
13
\10
31
\10
\5
16
310
\5
26
5
8
\10
\10
25
\5
14
\0.1
\10
31
15
\10
\5
\5
\10
\10
\10
48
\10
\5
33
73
\5
5
\5
\5
\10
42
21
5
\10
\0.1
285
11
7
\10
\5
6
84
65
30
27
11
5
22
221
8
69
10
24
10
19
27
17
\10
\0.1
quartzite and metasediments are tectonically intercalated within the volcanic rocks.
The rocks are assumed to have ages between 1.91 and 1.88 Ga (Korneliussen and
Sawyer 1989) and were metamorphosed at amphibolite facies conditions, at 6 kbar
and ca. 575C (Sawyer 1986). The age of the prograde metamorphism is unknown.
99
100
A. Mller et al.
101
Fig. 4.17 Location of the Nedre yvollen pegmatite mine, 1.5 km WNW of Drag. The hammer
and pick symbols indicate locations of former feldspar mines in the Drag pegmatite cluster. The
area shown in the map is within the Proterozoic Tysfjord granite
K-feldspar and mica (Figs. 4.18, 4.10c). The top of the pegmatite consisted of a
plagioclase-rich cupola which has been mined (Neumann 1952). The quartz core
has a marginal zone of smoky quartz. The body is about 30 9 60 m in horizontal
section and approximately 80 m deep. The occurrence of yttrofluorite masses
(up to 2 m in size) within the marginal part of the quartz core is a feature of the
pegmatite. Yttrofluorite is a greenish white fluorite with variable contents of REE.
Other accessories are native bismuth, chalcopyrite, galenite, cassiterite, euxenite(Y) and calcite (Husdal 2008). According to the pegmatite classification by Cerny
and Ercit (2005) the pegmatite is a Niobium-Yttrium-Fluorine (NYF)-type
pegmatite of the rare element REE class. Despite the amphibolite-facies overprint,
the pegmatite body shows no obvious signs of deformation at the macro scale.
Mining history. Mining of feldspar in the Drag area started in 1907. The Nedre
yvollen pegmatite was discovered during construction of a mine railway in 1909
and in the same year the mining of the pegmatite started in an open pit. Feldspar
mining continued, with interruptions, until the 1930s. The total feldspar production
was about 12,000 tons, resulting in a 30-m deep open pit. The deposit was drilled
in 1975 and 1979 by NGU and about 175,000 tons of massive quartz was proven
102
A. Mller et al.
Fig. 4.18 The Nedre yvollen pegmatite. a Horizontal section through the central part of the
Nedre yvollen pegmatite (modified from Norwegian Crystallites 2011). b Vertical E-W section
of the pegmatite according to mli and Lund (1979)
below the open pit (mli and Lund 1979). Mining of quartz, underground, started
in 1987 and continued, with interruptions, until Norwegian Crystallites AS overtook the mine in 1996. Since then the pegmatitic quartz has been the major source
of the HPQ produced by Norwegian Crystallites AS.
Quartz petrography. The average quartz crystal size in the pegmatite core is
about 6 mm. The crystals are clear, and fluid and mineral inclusions are very rare.
The few inclusions identified comprise (in order of frequency): biotite
(10200 lm), albite (2050 lm), K-feldspar (1050 lm), silicate melt inclusions
(550 lm) and calcite (1050 lm; Figs. 4.12c, 4.13c, 4.14c). The quartz grains
commonly show no indications of metamorphic overprint such as undulatory
extinction or sub-grain formation (recrystallisation). The shear stress which
affected the pegmatite core was focused in thin shear zones resulting in cleavagelike shear planes cross-cutting the quartz body. The CL of quartz is very homogeneous and very intense at initial electron exposure (Fig. 4.14c). The CL intensity
decreases rapidly during the first seconds of electron beam exposure. Micro cracks
healed by secondary dull-luminescent quartz are very rare.
Quartz crystal chemistry. The quartz is characterised by low Ti (mean 3.0
lg g-1) and moderately low Al (mean 20.6 lg g-1). The average Li content is 4.9
lg g-1 and B is 1.6 lg g-1. It is the only investigated deposit which contains
quartz with Be concentrations above the limit of detection (mean 0.29 lg g-1).
103
The average quartz composition suggests high-purity quality despite the fact that
individual analyses plot close to the concentration limits for HPQ.
Economic assessment and remarks on HPQ formation. The Nedre yvollen
pegmatite is one of the few deposits in the world from which HPQ has been
produced for more than a decade. The large, massive quartz body, the large crystal
size and the homogeneous chemistry with very low trace element concentrations
have made it a highly economic, world-class HPQ deposit.
The pegmatitic quartz core of the Nedre yvollen pegmatite consists of primary
magmatic HPQ which was generally not or only weakly affected by secondary
recrystallisation due to shearing related to the Caledonian metamorphism. However, it cannot be ruled out that the metamorphic temperature, which was presumably in the range of 420450C and the pressure of 23 kbar (Bjrklund 1989),
may have had an effect on the trace element distribution in the quartz.
Figure 4.5b shows that NYF-type pegmatite melts, such as the Nedre yvollen
pegmatite, have the potential to crystallise HPQ directly from silicate melts. Thus,
the process leading to HPQ crystallisation in the Nedre yvollen quartz is very
different from HPQ genesis in the quartzites discussed above.
104
A. Mller et al.
Fig. 4.19 Simplified geological map according to Solli and Nordgulen (2006) with locations of
the Nesodden and Kvalvik quartz occurrences in the Hardangerfjord area. The extension of the
Hardanger Shear Zone (HSZ) is according to Fossen and Hurich (2005)
Quartz petrography. The crystal sizes of the Nesodden quartz are highly variable, ranging from \1 lm to about 3 cm with an average of about 3 mm. The
quartz is locally strongly recrystallised due to post-crystallisation deformation
related to late Caledonian tectonics. The quartz shows a high porosity at microscale partially due to the high content of fluid inclusions. The fluid inclusions are
liquid-rich and occasionally contain precipitated solids, presumably crystals of
halite and sylvite (Fig. 4.12d). Inclusions of muscovite (1200 lm) are very
common (Fig. 4.13d): occasionally, inclusions of calcite are present (120 lm).
CL imaging shows that the Nesodden quartz is strongly altered due to recrystallisation and fluid-driven overprint. Domains of primary, unaltered quartz occur
as bright luminescent relicts and comprise less than 25 vol.% of the investigated
quartz samples (Fig. 4.14d). CL reveals tiny inclusions of calcite and feldspar-like
composition along healed micro cracks (Fig. 4.14d).
Quartz crystal chemistry. The quartz from Nesodden has low Ti (mean 2.1
lg g-1), moderately low Al (mean 18.0 lg g-1) and Ge (mean 1.0 lg g-1) and
relatively high Li (mean 5.7 lg g-1) contents. The average B content is about 1.4
lg g-1. Four of the quartz analyses listed in Table 4.2 show high Al ([60 lg g-1)
together with high K ([13 lg g-1) indicating that mica-like micro- inclusions
contaminated the analyses. These analyses are not included in the calculation of
the average values, because they do not reflect the content of lattice-bound trace
elements. Because of the high average Li content, the Nesodden quartz is not a
HPQ in the strict sense. However, the experience of the authors in processing of
105
Fig. 4.20 Detailed geological map of the Nesodden quartz vein (modified from Ihlen and Mller
2011)
other quartz deposits, suggests that parts of the interstitially bound Li+ could
possibly be removed during chemical quartz treatment.
Economic assessment and remarks on HPQ formation. The Nesodden quartz
vein is a moderate-sized, massive quartz deposit. Relatively high Li and B concentrations, a high content of high-salinity fluid inclusions, common enclaves of
granitic gneiss and micro-inclusions of muscovite, feldspar and calcite are the
challenges in processing HPQ from the deposit.
Fluids responsible for the quartz vein crystallisation were presumably mobilised
during Caledonian extensional shearing related to the motion of the nearby HFZ.
The LA-ICP-MS analyses suggest that the primary crystallised hydrothermal
quartz contained low concentrations of lattice-bound trace elements. Subsequent
alteration and recrystallisation associated with shearing and fluid flux may have
contributed to further purification of the quartz as indicated by the low CL
intensity of the secondary quartz.
106
Table 4.4 XRF bulk rock
analyses of quartzites from
Kvalvik
A. Mller et al.
Sample 58906
Major elements (wt.%)
SiO2
93.9
2.44
Al2O3
Fe2O3
0.36
0.06
TiO2
MgO
0.06
CaO
0.01
0.24
Na2O
1.16
K2O
MnO
0.01
\0.01
P2O5
LOI
0.21
total
98.5
Trace elements (lg g-1)
Ba
112
Cr
8.5
Ga
3.7
Nb
1.7
Ni
\2
Pb
19
Rb
36
Sr
9
Th
\4
Y
8.2
Zn
1.4
Zr
60
Sample 58908
96.9
0.98
0.29
0.03
0.17
\0.01
\0.1
0.36
\0.01
\0.01
0.24
99.0
28
\4
2.6
1.3
\2
17
15
\1
4.1
4.1
1.3
68
107
Fig. 4.21 Detailed geological map of the Kvalvik quartz deposit (modified from Ihlen and
Mller 2011)
quartzite, but cross-cutting, irregular vein structures are common as well. The zone
of vein-infiltrated quartzite extends about 350 m in NWSE direction and up to
120 NE-SW (Fig. 4.21). It is approximately 40 m thick and plunges 2025 N to
the NE. The volume of the zone corresponds to 1.4 mill m3 (350 9 120 9 40 m).
Ca. 0.3 million m3 (120 9 120 9 20 m) have been mined. The average percentage of hydrothermal veins in this volume is about 25 vol.% resulting in inferred
remaining resources of about 0.7 million tons snow-white quartz. The percentage
increases from ca. 1015 vol.% in the SW to 6080 vol.% in the NE. The NW part
of the deposit was mined for quartz to supply the ferrosilicon smelter at Bjlvefossen. The open pit mine is about 120 9 120 m in size and the SE wall is about
40 m high.
Quartz petrography. The crystal size of the Kvalvik quartz ranges from \1 lm
to about 2 cm with an average of about 2 mm. The quartz is cut by micro-scale
shear zones consisting of micro-crystalline, recrystallised quartz. Locally the
quartz is completely recrystallised due to late Caledonian deformation. The content of liquid-rich fluid inclusions (250 lm) is very high. The fluid inclusions
often contain a gas bubble and occasionally precipitated salt crystals, presumably
108
A. Mller et al.
halite (Fig. 4.12e). The high fluid inclusion content and common micro-scale
cavities result in a high micro-porosity. Inclusions of muscovite (150 lm) are
very common (Fig. 4.13e) and occasionally small inclusions of calcite (2-20 lm)
occur.
CL imaging shows that the primary hydrothermal quartz is strongly altered. The
domains of altered quartz have much lower CL intensity than the primary quartz.
The relicts of primary unaltered quartz comprise about 2030 vol.%. Bright
luminescent sub-micron inclusions (\1 lm) of feldspar-like composition occur
along healed micro cracks (Fig. 4.14e). The CL structures observed are similar to
those in the Nesodden quartz.
Quartz crystal chemistry. The Kvalvik quartz contains low average Ti (2.2
lg g-1), relatively low Al (15.7 lg g-1) and moderately high Li (4.2 lg g-1)
contents. In general, the chemistry is similar to that of the Nesodden quartz.
However, the Kvalvik quartz has slightly lower Al and Li contents compared to the
Nesodden quartz.
Economic assessment and remarks on HPQ formation. The hydrothermal
quartz veins at Kvalvik form a medium-size quartz deposit. The chemistry of the
quartz crystals suggests that a HPQ product could possible be produced from
the deposit. The major challenge will be the separation of the vein quartz from the
feldspar- and mica-bearing quartzitic host. The mica inclusions and sub-micron
inclusions of feldspar will presumably add to the concentrations of lattice-bound
Al, K, Na, and Ca of the quartz.
Quartz chemistry, mineral and fluid inclusion inventory and CL intensity and
structures are similar to those of the Nesodden quartz, suggesting a similar genesis
and source for the quartzforming fluids. The processes of HPQ formation are
analogous to those of the Nesodden deposit as well.
109
Fig. 4.22 Simplified geological map of the Pasvik area with the location of the Svanvik quartz
vein (modified from Siedlecka and Nordgulen 1996)
contact between the Svanvik Complex and the PechengaVarzuga Greenstone Belt
is just 2.2 km E of the new quartz mine at Svanvik.
The hydrothermal quartz vein at Svanvik was discovered in 1984 by the
geologist Mogens Marker (University of Copenhagen) during regional mapping of
the Pasvik area. Between 1985 and 1987 the deposit was sampled, mapped and
drilled by NGU (Wanvik 1988, 1989a, 1989b). Strongly deformed chlorite schist,
with a thickness of up to 10 m, occurs on both sides of the quartz vein (Fig. 4.23).
The vein and the chlorite schist are part of an E-W striking shear zone within the
granitic Svanvik gneiss. The vein was emplaced at an initial stage of shear-zone
formation, presumably during the Early Proterozoic (*1.75 Ga), when the
PechengaVarzuga Greenstone Belt was sandwiched between two colliding continents (Melezhik and Sturt 1994). Wanvik (1989a) suggested that the Svanvik
quartz vein is up to 20 m wide and at least 500 m long and 50 m deep, which
corresponds to about 1 million tons of quartz.
Formation of the vein is associated with strong chloritisation of the host rocks.
During vertical movements in the shear zone, calcic (carbon-rich) fluids entered
extensional fractures crossing the vein and calcite crystallised together with minor
chlorite and epidote. The veins are up to 0.5 m wide. In addition there are small
K-feldspar-quartz veins, up to 2 cm in width. Contact-parallel sheets and layers
(shear planes) of chlorite schist, 0.5 mm to 10 cm in thickness, occur in a 0.52 m
110
A. Mller et al.
wide zone along the vein margins. Late-stage re-activation of the shear zone
caused minor brecciation at the quartz vein contacts and the formation of open
cracks, which were filled with white milky quartz, amethyst, and clear rock
crystals up to 4 cm in size. The primary hydrothermal quartz at Svanvik contains
macroscopic ([1 mm) inclusions of calcite, chlorite, epidote, K-feldspar, chlorite
schist, granitic gneiss, clay (in cavities where calcite has been dissolved), and iron
oxides.
Quartz petrography. Crystal sizes in the Svanvik quartz are highly variable and
range from several centimetres down to \1 lm depending on the intensity of
shearing and associated recrystallisation. Large quartz grains ([2 mm) are commonly elongate and have undulatory extinction and recrystallised margins. The
average crystal size is about 23 mm. Mineral inclusions [5 lm are relatively
rare. Muscovite (10100 lm) represents the most common mineral inclusion
phase (Fig. 4.13f). Fluid inclusions are very common. The aqueous inclusions
(260 lm) contain a gas bubble and occasionally halite crystals, indicating high
salinities ([15 wt.%) in the entrapped fluids (Fig. 4.12f). In addition, traces of
sylvite were detected in decrepitated fluid inclusions by EDX.
The CL intensity of the Svanvik quartz is low (Fig. 4.14f). Structures visualised
by CL indicated a very intense alteration, recrystallisation and brecciation of the
primary hydrothermal quartz due to multiply reactivated shearing. In addition, tiny
(\1 lm) inclusions of calcitic and feldspar-like composition are revealed by CL in
healed micro cracks and recrystallised domains.
Quartz crystal chemistry. Quartz from Svanvik has very low Li (mean 2.1
lg g-1), Al (mean 5.3 lg g-1), Ti (mean 1.1 lg g-1) and Fe (mean 0.2 lg g-1)
concentrations. In general, the Svanvik quartz has the lowest total content of
lattice-bound trace elements (*15 lg g-1) compared to the other deposits discussed in this study.
Economic assessment and remarks on HPQ formation. The Svanvik deposit is a
medium-sized, massive quartz deposit with outstanding pure quartz chemistry,
with a total trace element content of about 15 lg g-1. For that reasons Norwegian
Crystallites AS started seasonal mining of the vein 2 years ago.
However, the multiple overprint of the quartz vein resulted in crystallisation of
additional quartz generations mainly at the vein contacts (Fig. 4.23). These
younger quartz generations form small volumes but have higher trace element
concentrations than the major vein quartz and may thus contaminate the quartz
product (Mller unpublished data). Transecting calcite-chlorite-epidote and
K-feldspar-quartz veins, macroscopic inclusions ([1 mm) of calcite, chlorite,
epidote, K-feldspar, chlorite schist, granitic gneiss, clay, iron oxides and microscopic inclusions of calcite are the challenges for HPQ production from this
deposit.
LA-ICP-MS analyses suggest that the hydrothermal quartz had already initially
very low concentrations of lattice-bound trace elements. Subsequent, multiple
shearing and associated fluid flux resulted in intense recrystallisation and fluiddriven alteration. The latter is evident by the high content of secondary fluid
inclusions. CL imaging indicates that about 8070 vol.% of the primary
111
hydrothermal quartz was affected and altered by these events. These secondary
processes might have contributed to the additional purification of the quartz.
112
A. Mller et al.
113
References
Adachi T, Hokada T, Osanai Y, Toyoshima T, Baba S, Nakano N (2010) Titanium behavior in
quartz during retrograde hydration: occurrence of rutile exsolution and implications for
metamorphic processes in the Sr Rondane Mountains, East Antarctica. Polar Sci 3:222234
mli R, Lund B (1979) Diamantboringer Nedre yvollen kvartsforekomst, Drag i Tysfjord. NGU
rapport 1771, 6 pp
Andresen A, Tull JF (1986) Age and tectonic setting of the Tysfjord gneiss granite, Efjord, North
Norway. Norsk Geologisk Tidsskrift 66:6980
Bambauer HU, Brunner GO, Laves F (1962) Wasserstoff-Gehalte in Quarzen aus Zerrklften der
Schweizer Alpen und die Deutung ihrer regionalen Abhngigkeit. Schweizerische Mineralogische und Petrographische Mitteilungen 42:221236
Bartovic S, Beane R (2007) Analysis of blue color in quartz grains from cushing formation, peaks
island, maine. In: Abstracts GSM spring meeting 2007, GSM Newsletter 36/2:4
Beurlen H, Mller A, Silva D, Da Silva MRR (2011) Petrogenetic significance of trace-element
data analyzed with LA-ICP-MS in quartz from the Borborema pegmatite province,
northeastern Brazil. Mineral Mag 75:27032719
Bibikova EV, Ihlen PM, Marker M (2001) Age of the hydrothermal alteration leading to garnetite
and kyanite pseudo-quartzite formation in the Khizovaara segment of the late Archean Keret
Greenstone Belt, Russian Karelia. EUG XI Strasbourg, 812 April 2001. J Conf Abstr 6:277
Bjrklund L (1989) Geology of the AkkajaureTysfjordLofoten traverse, N. Scandinavian
Caledonides. Ph D thesis, Chalmers Tekniska Hgskola och Gteborgs Universitet, publ. A
59, 214 pp
Blankenburg H-J, Gtze J, Schulz H (1994) Quarzrohstoffe. Deutscher Verlag fr Grundstoffindustrie, Leipzig-Stuttgart, 296 pp
Breiter K, Mller A (2009) Evolution of rare-metal granitic magmas documented by quartz
chemistry. Eur J Mineral 21:335346
Brouard S, Breton J, Girardet G (1995) Small alkali metal clusters on (001) quartz surface:
adsorption and diffusion. J Mol Struct (Theochem) 334:145153
114
A. Mller et al.
115
Haus R (2005) High demands on high purityprocessing of high purity quartz and diatomite.
Industrial Minerals October 2005, pp 6267
Hertweck B, Niedermayr G, Beran A (2003) OH zoning ion alpine quartz from Austria. European
Geophysical Society (CD-Rom) Vol. 5, EGS-AGU-EUG Joint Assembly, 6th11th April
2003, Nice, France, 08506
Husdal T (2008) The minerals of the pegmatites within the Tysfjord granite, northern Norway.
Bergverksmuseets Skrift 38:528
Hyrsl J, Niedermayr G (2003) Magic world: inclusions in quartzGeheimnisvolle Welt:
Einschlsse im Quarz. Bode Verlag GmbH, Haltern, 240 pp
Ihlen PM (2000) Utilisation of sillimanite minerals, their geology, and potential occurrences in
Norwayan overview. NGU Bulletin 436:113128
Ihlen PM, Mller A (2011) Forekomster av hyren kvarts langs Hardangerfjorden. NGU Rapport
2009.024, Trondheim, 69 pp
Ingdal SE, Torske T, Kvale A (2001) Bergrunnskart Jondal 1315 4, M 1:50000. Geological
Survey of Norway, Trondheim
IOTA (2011) IOTA high purity quartz. http://www.iotaquartz.com/techiota4data.html Accessed
20 May 2011
Jacamon F, Larsen RB (2009) Trace element evolution of quartz in the charnockitic Kleivan
granite, SW Norway: the Ge/Ti ratio of quartz as an index of igneous differentiation. Lithos
107:281191
Jakobsen BM, Nielsen E (1977) Kyanit kvartsit projektet 19761977. Laboratorienrapport.
Endogen Laboratorium. Geologisk Institut Aarhus Universitet, Aarhus, Denmark, 19 pp
Jourdan A-L, Vennemann TW, Mullis J, Ramseyer K, Spiers CJ (2009) Evidence of growth and
sector zoning in hydrothermal quartz from Alpine veins. Eur J Mineral 21:219231
Jung L (1992) High purity natural quartz. Part I: High purity natural quartz for industrial use. Part
II: High purity natural quartz markets for suppliers and users. Quartz Technology, Liberty
Corner , p 657
Korneliussen A, Sawyer EW (1989) The geochemistry of lower proterozoic mafic to felsic
igneous rocks, rombak window, North Norway. NGU Bull 415:721
Larsen RB, Henderson I, Ihlen PM, Jacamon F (2004) Distribution and petrogenetic behaviour of
trace elements in granitic quartz from South Norway. Contributions Mineral Petrol
147:615628
Larsson D (2001) Transition of granite to quartz-kyanite rock at Hlsjberg, southern Sweden:
consequence of acid leaching and later metamorphism. GFF 123:237246
Leeder O, Thomas R, Klemm W (1987) Einschlsse in Mineralen. VEB Deutscher Grundstoffverlag, Leipzig, 180 pp
Levchenkov OA, Levsky LK, Nordgulen , Dobrzhinetskaya, Vetrin VR, Cobbing J, Nilsson LP,
Sturt BA (1995) UPb zircon ages from Srvaranger, Norway, and the western part of the
Kola Peninsula, Russia. NGU Special Publication 7:2947
Luckscheiter B, Morteani G (1981) The H contents of quartz from Alpine veins from the penninic
rocks of the central and western tauern window (Austria/Italy). Tschermaks MineralogischPetrologische Mitteilungen 28:223228
Maschmeyer D, Lehmann G (1983) A trapped-hole center causing rose coloration of natural
quartz. Zeitschrift fr Kristallographie 163:181196
McLaren AC, Cook RF, Hyde ST, Tobin RC (1983) The mechanism of the formation and growth
of water bubbles and associated dislocation loops in synthetic quartz. Phys Chem Miner
9:7994
Meinhold G (2010) Rutile and its applications in the earth sciences. Earth Sci Rev 102:128
Melezhik VA, Sturt BA (1994) A review of the general geology and history of the development
of the early Proterozoic Polmalk-Pasvik-Pecheng Imandra/Varzuga-UstPonoy Greenstone
Belt. Earth Sci Rev 36:205241
Miyoshi N, Yamaguchi Y, Makino K (2005) Successive zoning of Al and H in hydrothermal vein
quartz. Am Mineral 90:310315
116
A. Mller et al.
Monecke T, Kempe U, Gtze J (2002) Genetic significance of the trace element content in
metamorphic and hydrothermal quartz: A reconnaissance study. Earth Planetary Sci Lett
202:709724
Mller A, Koch-Mller M (2009) Hydrogen speciation and trace element contents of igneous,
hydrothermal and metamorphic quartz from Norway. Mineral Mag 73:569583
Mller A, Seltmann R, Behr HJ (2000) Application of cathodoluminescence to magmatic quartz
in a tin granitecase study from the Schellerhau Granite Complex, Eastern Erzgebirge,
Germany. Mineralium Deposita 35:169189
Mller A, Kronz A, Breiter K (2002a) Trace elements and growth patterns in quartz: a fingerprint
of the evolution of the subvolcanic Podlesi Granite System (Krune Hory, Czech Republic).
Bull Czech Geol Surv 77:135145
Mller A, Lennox P, Trzebski R (2002b) Cathodoluminescence and micro-structural evidence for
crystallisation and deformation processes of granites in the Eastern Lachlan Fold Belt (SE
Australia). Contributions Mineral Petrol 143:510524)
Mller A, Wiedenbeck M, van den Kerkhof AM, Kronz A, Simon K (2003a) Trace elements in
quartza combined electron microprobe, secondary ion mass spectrometry, laser-ablation
ICP-MS, and cathodoluminescence study. Eur J Mineral 15:747763
Mller A, Ren M, Behr H-J, Kronz A (2003b) Trace elements and cathodoluminescence of
igneous quartz in topaz granites from the Hub Stock (Slavkovsky0 Les Mts, Czech Republic).
Mineral Petrol 79:167191
Mller A, Breiter K, Seltmann R, Pcskay Z (2005a) Quartz and feldspar zoning in the Eastern
Erzgebirge pluton (Germany, Czech Republic): evidence of multiple magma mixing. Lithos
80:201227
Mller A, Wanvik JE, Kronz A (2005b) Norwegian kyanite quartzitespotential resources of
high purity quartz? NGU Report 2005.039, Trondheim, Norway, 70 pp
Mller A, Williamson BJ, Smith M (2005c) Origin of quartz cores in tourmaline from Roche
Rock, SW England. Mineral Mag 69:381401
Mller A, Ihlen PM, Wanvik JE, Flem B (2007) High-purity quartz mineralisation in kyanite
quartzites, Norway. Mineralium Deposita 42:523535
Mller A, Ihlen PM, Kronz A (2008a) Quartz chemistry in polygeneration Sveconorwegian
pegmatites, Froland, Norway. Eur J Mineral 20:447463
Mller A, Wiedenbeck M, Flem B, Schiellerup H (2008b) Refinement of phosphorus
determination in quartz by LA-ICP-MS through defining new reference material values.
Geostand Geoanal Res 32(3):361376
Mller A, Behr H-J, van den Kerkhof AM, Kronz A, Koch-Mller M (2010a) The evolution of
late-Hercynian granites and rhyolites documented by quartza review. Earth Environ Sci
Trans Royal Soc Edinburgh 100:185204
Mller A, Herrington R, Armstrong R, Seltmann R, Kirwin DJ, Stenina NG, Kronz A (2010b)
Trace elements and cathodoluminescence of quartz in stockwork veins of Mongolian
porphyry-style deposits. Mineralium Deposita 45:707727
Neumann H (1952) Feltspat forekomster i Tysfjorddistriktet. NGU Bergarkivrapport nr. 5208
Nordgulen (1999) Geologisk kart over Norge, Berggrunnskart Hamar, M 1: 250.000. Geol Surv
Norway, Trondheim
Northrup CJ (1997) Timing structural assembly, metamorphism, and cooling of the Caledonian
nappes in the Ofoten-Efjorden area, north Norway: Tectonic insights from U-Pb and 40Ar/
39Ar geochronology. J Geol 105:565582
Norwegian Crystallites AS (2011) http://norcryst.no/. Accessed 21 Jan 2011
Parker RB (1962) Blue quartz from the Wind River Range, Wyoming. Am Mineral 47:12011202
Passchier CW, Trouw RAJ (2006) Microtectonics. Springer, Heidelberg 366 pp
Penniston-Dorland SC (2001) Illumination of vein quartz textures in a porphyry copper ore
deposits using scanned cathodoluminescence: grasberg igneous complex, Irian Jaya,
Indonesia. Am Mineral 86:652666
Pfenninger H (1961) Diffusion von Kationen und Abscheidung von Metallen in Quarz unter
elektrischer Feldeinwirkung. PhD Thesis, University Zrich
117
Ramseyer K, Mullis J (1990) Factors influencing short-lived blue cathodoluminescence of quartz. Am Mineral 75:791800
Richter DK (1971) Fazies- und Diagenesehinweise durch Einschlsse in authigenen Quarzen.
Neues Jahrbuch fr Geologie und Palontologie Monatshefte 10:604622
Roedder E (1984) Fluid inclusions. Reviews in mineralogy, vol. 12. Mineralogical Society of
America, Washington, 644 p
Rusk BG, Lowers HA, Reed MH (2008) Trace elements in hydrothermal quartz: relationships to
cathodoluminescence textures and insights into vein formation. Geology 36:547550
Sawyer E (1986) Metamorphic assemblages and conditions in the Rombak basement window.
NGU Rapport 88.116, Trondheim, Norway, 11 pp
Seifert W, Rhede D, Thomas R, Frster H-J, Lucassen F, Dulski P, Wirth R (2011) On the origin
of igneous blue quartz: inferences from a multi-analytical study of submicron mineral
inclusions. Mineral Mag 75:25192534
Shepherd TJ, Rankin AH, Alderton DHM (1985) A practical guide to fluid inclusion studies.
Blackie and Sons, Glasgow 239 pp
Siebers FB (1986) InhomogeneVerteilung von Verunreinigungen in gezchteten und natrlichen
Quarzen als Funktion derWachstumsbedingungen und ihr Einflu auf kristallphysikalische
Eigenschaften. PhD Thesis, Ruhr-Universitt Bochum, 133 pp
Siedlecka A, Nordgulen (1996) Geologisk kart over Norge, berggrunnskart Kirkenes, M 1:250
000. Geological Survey of Norway, Trondheim, Norway
Sigmond EMO (1998) Geologisk kart over Norge; Berggrunnskart OddaM 1:250.000.
Geological Survey of Norway, Trondheim, Norway
Simon K (2001) Does dD from fluid inclusion in quartz reflect the original hydrothermal fluid?
Chem Geol 177:483495
Simpson DR (1977) Aluminum phosphate variants in feldspars. Am Mineral 62:351355
Solli A, Nordgulen (2006) Bedrock map of Norway and the Caledonides in Sweden and
Finland. Scale 1: 2 000 000. Geological Survey of Norway, Trondheim
Stephens MB, Gustavson M, Ramberg IB, Zachrisson E (1985) The Caledonides of central north
Scandinaviaa tectonostratigraphic overview. In: Gee DG, Sturt BA (eds) The Caledonide
OrogenScandinavia and Related Areas. Wiley, New York, pp 135162
Thomas S-M (2008) Wasserstoff in nominell wasserfreien Mineralen. PhD thesis. TU Berlin, D
83, Berlin, Germany, 134 pp
Thomas R, Webster JD, Davidson P (2006) Understanding pegmatite formation: the melt and
fluid inclusion approach. In: Webster JD (ed) Melt inclusions in plutonic rocks. Mineralogical
Association of Canada, Short Course Series 36:189210
Tveten E, Lutro O, Thorsnes T (1998) Geologisk kart over Norge, berggrunnskart lesund, 1:
250.000. Geological Survey of Norway, Trondheim, Norway.
Van den Kerkhof AM, Hein UF (2001) Fluid inclusion petrography. Lithos 55:2747
Van den Kerkhof AM, Mller A (1999) Fluid inclusion re-equilibration and trace element
redistribution in quartz: observations by cathodoluminescence microscopy. ECROFI XV 1999
Abstracts and Program, Potsdam, Terra Nostra 99(6):161162
Van den Kerkhof AM, Kronz A, Simon K, Scherer T (2004) Fluid-controlled quartz recovery in
granulite as revealed bycathodoluminescence and trace element analysis (Bamble sector,
Norway). Contributions Mineral Petrol 146:637652
Wanvik JE (1988) Svanvik kvartsforekomst i Pasvik, Sr-Varanger kommune. NGU Rapport
87.081, Trondheim, Norway, 18 pp
Wanvik JE (1989a) Statusrapport 1989 for underskelse av Svanvik kvartsforekomst. NGU
Rapport 89.078, Trondheim, Norway, 17 pp
Wanvik JE (1989b) Sluttrapport for underskelse av Svanvik kvartsforekomst. NGU Rapport
89.165, Trondheim, Norway, 9 pp
Wanvik JE (1998) Kyanite investigations in Tverrdalen, Surnadal. NGU Rapport 98.080,
Trondheim, Norway, 24 pp
Wanvik JE (2001) Kvartsressurser i Nordland. NGU Rapport 2001.020, Trondheim, Norway,
103 pp
118
A. Mller et al.
Chapter 5
119
120
G. Morteani et al.
5.1 Introduction
The demand for high purity quartz (HPQ) is increasing worldwide (Moore 2005,
Haus 2005, Mller et al. 2007). This increase is driven mainly by the demand as
raw material for special applications in the high-tech industry. The main suppliers
for high purity quartz are Unimin (USA), the Quartz Corporation (USA) and
Norwegian Crystallites (Norway). Unimin and Quartz Corporation mines weakly
metamorphosed alaskite, Norwegian Crystallites mines the a zoned Drag pegmatite. In Russia hydrothermal quartz vein deposits that seem to be suitable for the
production of high purity quartz are investigated at Ust-Puiva/Saranpaul/Tyumen
Oblast (Subpolar Ural) (Burlakov 1995, 1999).
High purity quartz is characterized by less than 50 ppm impurities (e.g. Harben
2002). Lattice-bound trace elements which either substitute Si4+ (i.e. Al3+, Fe3+,
Ti4+, Ge4+, B3+, P5+) or occur at interstitial channel position such as Li+, K+, Na+,
H+, Fe2+ (e.g. Weil 1984, 1993) are the most common ones. They are the quality
determining impurities because they are difficult up to impossible to eliminate by
mineral processing (e.g. Jung 1992). Minor solid inclusions, typically rutile, mica
and feldspars, that carry additional Ca, Na, K, Mg, Al, Sr, Rb, Sm, Nd and Ti
(Rossmann et al. 1987), are amenable to elimination by mineral processing
techniques. Fluid inclusions are potential sources for Na, K, Ca, Cl and different gas
species such as CO2, N2, H2S, CH4 and higher hydrocarbons. However the components contained by fluid inclusions may be removed by milling and subsequent
leaching, thermal treatment and calcination (e.g. Haus 2005). Mineral processing,
that may include in addition to grinding, sieving and magnetic separation also
thermal and chemical treatment by highly reactive acids such as hydrofluoric acid,
can be very expensive. Thus, in the mined raw quartz material the content of
intracrystalline impurities should be as low as possible to save treatment costs.
The evaluation of quartz deposits as potential sources of high purity quartz
including the design of an optimal mineral processing flow sheet requires a reliable
determination of the different impurities that are carried by quartz with a suitable
combination of methods in a reasonable time and at tolerable costs.
The studied pegmatitic quartz veins belong to the hundreds of pegmatites
dotting the whole Sierra de Comechigones (Argentina) ranging from differentiated
and zoned ones to pegmatitic quartz veins with only subordinate K-feldspar and
mica contents. The study area is given in Fig. 5.1. The pegmatitic veins of the
Sierra de Comechigones belong to the large pegmatite province of the Sierras
Pampeanas (Herrera 1968; Morteani et al. 1995). Specifically the pegmatitic quartz
121
Fig. 5.1 Geological sketch map of the Sierras Pampeanas according to Martino (2003), Steenken
et al. (2006) and Siegesmund et al. (2010)
veins with white fine-grained quartz that are hosted in the Sierra de Comechigones
by the mylonites of the Guacha Corral shear zone, get increasing interest as
potential suppliers of high purity quartz to the electronic industry.
In order to get a quick and reliable information on the potential of the different
quartz veins for the production of high purity quartz a combination of LA-ICP-MS,
122
G. Morteani et al.
Fig. 5.2 Left: boudinaged and faulted quartz vein set in the schistosity of the country rock.
Right: landscape with subhorizontal pegmatitic quartz veins
5.2 Geology
The Sierra de Comechigones is the southernmost mountain range in the Sierras
de Cordoba. The Sierras de Cordoba are part of the Sierras Pampeanas of central
Argentina. The Sierra de Comechigones consists of the Sierra de Comechigones
metamorphic complex, which is a sequence of gneisses, migmatitic gneisses,
melt-depleted and melt-enriched diatexites, granulites, amphibolites, marbles and
peridotites (Otamendi et al. 1999) with granitoid rocks (Rapela and Shaw 1979;
Rapela et al. 1982, 1990). During the Cambrian the rocks of the Sierra de
Cordoba underwent Pampean orogeny with a Barrovian-type metamorphism with
peak P, T-conditions of 800900C and 8.59 kbar (Martino et al. 2009, 2010;
Steenken et al. 2010, 2011; Otamendi et al. 1998; Martino 2003; Rapela and
Shaw 1979). Such PT conditions are sufficient to generate local partial melts
(Sims et al. 1998). The metamorphic peak in the Sierra de Comechigones
occurred at 550540 Ma (Siegesmund et al. 2010). The P, T conditions of the
mylonitisation that produced the Guacha Corral shear zone are 540590C and
36 kbar (Simpson et al. 2003; Whitmeyer and Simpson 2003; Martino 2003).
The Guacha Corral shear belt is the most prominent tectonic feature of the Sierra
de Comechigones (Fig. 5.1).
The quartz veins crop out as mostly sheared subhorizontal intrafolial irregular
folded bodies with thickened fold hinges and flattened, often boudinaged limbs.
The size of the bodies ranges from centimetres up to 7 m thickness (Fig. 5.2).
123
5.3 Methods
5.3.1 Bulk Chemical Composition by ICP-MS
Sample aliquots (2030 g) were at first washed in double rinsed deionised water and
subsequently cleaned for three hours in concentrated HNO3 and deionised water at
about 100C. Finally the samples were again rinsed with deionised water and dried in
a dust free cabinet desiccator. The cleaned samples were totally dissolved by
HF- microwave digestion in a closed vessel and the major and trace element concentration of the solution measured by ICP-MS (Perkin Elmer Elan 6000).
5.3.3 Cathodoluminescence
Cathodoluminescence (CL) measurements were done on carbon-coated, doublesided polished thick sections with a hot cathode CL microscope (HC1-LM,
cf., Neuser et al. 1995). The CL system was operated at 14 kV accelerating
voltage and a current density of about 10 lA/mm2. Luminescence images were
captured on-line during CL operations using a Peltier cooled digital
124
G. Morteani et al.
125
(Hmmerli and Diamond 2009; Eichinger et al. 2010). Before crushing, the
sample chamber was evacuated, flushed twice with He to avoid any air
contamination and filled with He to around 200 mbar. During the crushing
process the equipment was heated to 150C to avoid gas sorption on the freshly
crushed quartz surfaces. After crushing the device was directly fitted to a gas
chromatograph and the pressure and concentrations of normal gases (CO2, N2,
Ar, O2) and hydrocarbons were measured by GC-WLD and GC-FID (Shimadzu
GC-17A), respectively. The volume of the extracted gas species per gram of
quartz sample were calculated with respect of the weight of the sample, the gas
pressure, the total volume of the piston-cylinder device and the temperature. Air
contamination was quantified by the oxygen concentration and the results were
corrected according to it.
5.4 Results
Due to the ongoing exploration for high purity quartz in the Sierra de Comechigones
the analytical results are identified only by sample numbers without a reference to
specific quartz deposits/veins.
126
G. Morteani et al.
5.4.1 Petrography
The quartz samples of all studied occurrences showed under polarized light a
mortar texture caused by a strong deformation with dynamic recrystallisation
(Fig. 5.3). The fine-grained intergranular quartz mortar displays equilibrium grain
boundaries documenting a post-deformational static recrystallisation. The high
amount of fluid inclusions (see Sect. 5.4.5.1), and the absence of solid inclusions
identifiable by optical microscopy is remarkable.
5.4.2 Cathodoluminescence
The granular mortar texture of the quartz samples becomes also evident by cathodoluminescence (CL) investigations. All samples exhibit similar luminescence
behaviour. Generally quartz shows an initial bluish luminescence, which turns into
brownish-violet during electron irradiation (Fig. 5.4). The transient CL behaviour
is also visible in the time-dependent CL spectra (Fig. 5.4). The spectrum is
dominated initially by a blue emission band around 500 nm, which is typical for
pegmatite quartz (Gtze et al. 2005). This CL emission band can be related to
alkali compensated trace element centres in the quartz structure (Ramseyer and
Mullis 1990; Gtze et al. 2005).
The large quartz crystals are rimmed by a seam of higher CL luminosity. This
rim corresponds to the fine-grained quartz forming the mortar between the coarse
quartz grains. In the coarser quartz grains the CL reveals several fluid trails with
mostly higher CL intensity. The coarse quartz grains and the quartz forming the
mortar showed in spite of different CL luminosity the same spectral characteristics.
Therefore, the types of lattice defects are similar, but the higher CL intensity
indicates higher defect densities in the mortar quartz.
127
Fig. 5.4 Microphotographs of the quartz sample 15681 in transmitted light (TL), polarized light
(Pol), cathodoluminescence (CL) and time-dependent CL spectrum
128
Table 5.1 Trace element
concentrations by ICP-MS in
two representative quartz
samples from two different
quartz veins (d.l. = detection
limit)
G. Morteani et al.
Element
Unit
Detection limit
15656
15680
Li
Na
K
Rb
Cs
Ca
Mg
Sr
Ba
Al
Ga
As
Se
Ti
Zr
Nb
La
Ce
Fe
SUM
Chemical quality
ppm
ppm
ppm
ppm
ppm
ppm
ppm
ppm
ppm
ppm
ppm
ppm
ppm
ppm
ppm
ppm
ppm
ppm
ppm
ppm
0.5
100
100
0.2
0.05
100
0.01
0.2
1
100
0.1
0.1
0.1
5
1
0.1
0.1
0.1
1.6
0.7
200
100
0.3
0.06
200
\0.01
1.4
0.5
300
0.4
d.l.
d.l.
20
d.l.
0.2
0.2
0.4
d.l.
883
low
d.l.
200
d.l.
0.3
0.1
d.l.
\0.01
0.3
d.l.
300
0.3
d.l.
0.3
26
3
0.4
0.1
0.3
d.l.
532
low
indicating the absence of mica inclusions. The concentrations of Al are 300 ppm
for both samples and that of Ti are 20 and 26 ppm (Table 5.1). Both Al and Ti
contents can be referred to solid inclusions of feldspar and rutile, but alternatively
to lattice-bound Al and Ti.
Both samples contain traces of Ga, Rb, Cs, Sr, Nb, La and Ce. In contrast, Li,
Ba, Se, and Zr occur as trace elements only in one of the two samples. As and Fe
are below the detection limit (Table 5.1). According to Harben (2002) both quartz
samples have to be classified as low quality ones.
129
Table 5.2 Chemical composition by LA-ICP-MS spot analysis of two selected quartz samples
from two different quartz veins (d.l. = below detection limit)
Element
Unit
Limit of 15656
15680
detection
Rim quartz Coarse quartz Rim quartz Coarse quartz
Li
ppm
0.55
Be
ppm
0.06
B
ppm
1.00
Na
ppm
5.5
Al
ppm
6.0
P
ppm
3.0
K
ppm
3.2
Ca
ppm
14.0
Ti
ppm
1.20
Mn
ppm
0.50
Fe
ppm
1.00
Ge
ppm
0.17
SUM
ppm
Chemical quality of quartz
0.87
0.10
1.23
d.l.
26.10
d.l.
5.5
15.4
6.84
d.l.
d.l.
0.48
56.52
Medium
0.61
0.20
2.57
d.l.
14.80
d.l.
d.l.
d.l.
5.55
d.l.
d.l.
0.42
24.15
High
d.l.
d.l.
d.l.
d.l.
62.92
4.27
7.49
d.l.
d.l.
0.64
1.35
0.43
77.10
Medium
1.23
0.15
d.l.
d.l.
23.38
9.19
d.l.
d.l.
d.l.
3.03
d.l.
1.12
38.10
High
grained rim quartz grains show elevated Al, Ca, Ti and K contents compared to the
analyses of the coarse grains (Table 5.2).
In the coarse and fine-grained quartz samples 15680 the Ti is below the
detection limit of 1.2 ppm. In the fine grained quartz the Al, K and Fe contents in
the rim quartz are higher as compared to the coarse quartz. The reverse can be
observed for Li, P, Mn and Ge (Table 5.2). According to Harben (2002) and
Mller et al. (2007) the total trace element content of less than 50 ppm classifies
the coarse quartz grains of both samples given in Table 5.2 as high purity quartz
whereas the fine grained quartz has to be classified only as one of medium quality.
130
G. Morteani et al.
Fig. 5.5 a Three-phase (Laq, Lcar, Vcar) inclusions (population 1) located in a pseudosecondary
trail; b Two-phase inclusions (population 2) located in a secondary plane crosscutting an earlier
three-phase inclusions trail; c,d Higher hydrocarbon bearing inclusion (sample 15656, population
3) as seen within a 200 lm thick quartz section under transmitted light (c) and UV-light (d). Only
the liquid phase fluoresces and is indicative of the presence of hydrocarbons higher than C1
131
Inclusion population
CO2 (mol%)
N2 (mol%)
H2S (mol%)
1
1
2
2
99.3
97.7
47.7
98.9
0.7
2.3
52.3
1.1
\0.1
\0.1
n.d.
n.d.
([97.7 mol%) but low amounts of N2 (\2.3 mol%) and traces of H2S
(\0.1 mol%) are always present (Table 5.3).
Population 2: In the fluid inclusions of population 2 the gas bubble has a strong
tendency to change position under the Raman laser beam. Raman spectroscopy
reveals in the fluid inclusions the presence of a mixture of CO2 and N2. The
concentrations of CO2 and N2 vary significantly between 1.1 and 52 mol% for N2
and 47.7 and 98.9 mol% for CO2 (Table 5.3).
132
Table 5.4 Chemical
composition of the solutes
included in fluid inclusions in
quartz of two selected
samples (ppm = lg/gQtz,
n.a. = not analysed)
G. Morteani et al.
Cations
Sodium (Na+)
Potassium (K+)
Calcium (Ca2+)
Magnesium (Mg2+)
Lithium (Li+)
Strontium (Sr2+)
Barium (Ba2+)
Aluminum (Al3+)
Beryllium (Be3+)
Germanium (Ge4+)
Phosphorus (P)
Titanium (Ti3+)
Anions
Fluoride (F-)
Chloride (Cl-)
Bromide (Br-)
Sulfate (SO42-)
Total dissolved solids (TDS)
15656
15680
ppm
ppm
ppm
ppm
ppm
ppm
ppm
ppm
ppm
ppm
ppm
ppm
11.8
2.5
12.2
0.1
0.02
0.1
0.04
1.1
n.a.
n.a.
n.a.
n.a.
72.7
11.8
14.5
6.4
0.09
0.4
1.00
10.9
\0.01
\0.003
0.2
0.4
ppm
ppm
ppm
ppm
ppm
0.02
19.2
0.07
0.8
48.0
1.21
122.8
\0.05
14.5
257.0
like feldspars or micas. Concentration of sulfate is 0.8 and 14.5 ppm (Table 5.4).
The detected sulfate is probably formed during the extraction procedure by the
oxidation of extracted H2S, which was observed in inclusions of population 1 (cf.
Sect. 5.4.5.2). Traces of Li, Sr, Ba, Ti and partly Ge, P, F and Br are also present
(Table 5.4).
133
Table 5.5 Composition of the gases from fluid inclusions (ll/g) in quartz of two selected
samples from different occurrences, gas volumina are calculated without water vapour
Sample
15656
15680
Normal gases
Argon (Ar)
Nitrogen (N2)
Carbondioxide (CO2)
Saturated hydrocarbons
Methane (CH4)
Ethane (C2H6)
Propane (C3H8)
n-Butane (n-C4H10)
i-Butane (i-C4H10)
i-Pentane (i-C5H12)
n-Pentane (n-C5H12)
Hexane (C6H14)
Unsaturated hydrocarbons
Ethene (C2H4)
Propene (C3H6)
Cis(1)-Butene (cis-C4H8)
Trans-Butene (trans-C4H8)
Pentene (C5H10)
Hydrocarbons tot
Gas total
ll/g
ll/g
ll/g
0.15
26.5
31.2
0.16
14.7
65.0
ll/g
ll/g
ll/g
ll/g
ll/g
ll/g
ll/g
ll/g
0.0888
0.0071
0.0043
0.0019
0.0050
\0.0001
0.0024
0.0011
0.0238
0.0017
\0.0001
0.0006
0.0006
\0.0001
\0.0001
\0.0001
ll/g
ll/g
ll/g
ll/g
ll/g
ll/g
ll/g
0.0063
0.0094
0.0041
0.0001
0.0004
0.131
58.0
0.0019
\0.0001
\0.0001
\0.0001
\0.0001
0.03
79.9
Fig. 5.6 Concentrations of the individual gas species liberated from pegmatitic quartz veins as
function of the gas species (left) and as function of the individual samples (right); HCtot = total
volume of extracted hydrocarbons
134
G. Morteani et al.
5.5 Discussion
In the studied samples the LA-ICP-MS analyses reveal that the fine grained quartz
found as a mortar between coarse quartz grains is the predominant carrier of
impurities (Table 5.2). Already without any mineral preparation the coarser quartz
crystals fit with 24.15 and 38.10 ppm of impurities the criteria for high purity
quartz. Bulk analysis by total digestion and subsequent ICP-MS analysis only
would have missed this point, which is important for the planning of mineral
separation tests and flow sheet. Analytical results and microscopy suggest that the
impurities in quartz were expelled by the deformation and metamorphic recrystallisation leading to an enrichment of the impurities in the fine grained intragranular quartz mortar.
A comparison of the solution ICP-MS and the LA-ICP-MS data shows that as a
whole the content in impurites determined by ICP-MS after total dissolution of the
quartz samples is much higher than that determined by LA-ICP-MS spot analysis.
This indicates that most of the impurities are bound to solid or fluid inclusions and
not to the quartz lattice.
The results obtained from conventional fluid inclusion investigation and the
results obtained by aqueous leaching and gas liberation do complement another.
But it has to be considered that the extraction methods are comparatively quicker
and cheaper methods that can produce reliable information on the bulk composition and total amount of the fluids stored in the inclusions, a key information for
the economic and technical evaluation of quartz material. Nevertheless, a fluid
inclusion study can be indicated for the estimation of the crystallisation pressure
and temperature of quartz from a P,T plot of the isochores obtained after microthermometry and Raman spectroscopy investigations. Such data might be
helpful to understand the genesis of the present pegmatitic quartz veins and
influence the prospection strategy.
5.6 Conclusion
Our multi-technique investigations approach shows that the pegmatitic quartz veins
of the Sierra de Comechigones are rightly in the focus of the exploration for high
purity quartz. The here given first results show that a combination of careful field
135
References
Bottrell SH, Yardley B, Buckley F (1988) A modified crush-leach method fort he analysis of fluid
inclusion electrolytes. Bulletin de Mineralogie 111:279290
Burke EAJ (2001) Raman microspectrometry of fluid inclusions. Lithos 55:139158
Burlakov EV (1995) Dodo: Alpine Klfte im Polar-Ural. Lapis, 20, 1326
Burlakov EV (1999) The Dodo deposit (subpolar Urals, Russia). Mineral Rec 30:427442
Burruss RC (2003) Petroleum fluid inclusions, an introduction. In: Samson IM, Anderson AJ,
Marshall DD (eds) Fluid inclusions: analysis and interpretation. Mineralogical Association
Canada, Short Course Series 32, 159174
Eichinger F, Meier D, Hmmerli J, Diamond LW (2010) Stable isotope signatures of gases
liberated from fluid inclusions in Bedrock at Olkiluoto. Posiva Working Report 201088,
Posiva Oy, Olkiluoto, Finland. (www.posiva.fi)
Flem B, Larsen RB, Grimstvedt A, Mansfeld J (2002) In situ analysis of trace elements in quartz
by using laser ablation inductively coupled plasma mass spectrometry. Chem Geol 182:
237247
Gtze J, Pltze M, Trautmann T (2005) Structure and luminescence characteristics of quartz from
pegmatites. Am Mineral 90:1321
Hmmerli J, Diamond LW (2009) Fluid inclusions in basement rocks at Olkiluoto, Finland, and
their implications for a planned nuclear waste repository. In: ECROFI-XX, Granada, Spain,
pp 109110
Harben PW (2002) The industrial mineral handybooka guide to markets, specifications and
prices. Industrial mineral information, Worcester Park, United Kingdom, 4th edn. p 412
Haus R (2005) High demands on high purity. Ind Mineral 10:6267
Herrera AO (1968) Geochemical evolution of zoned pegmatites of Argentina. Econ Geol 63:
1329
Jung L (1992) High purity natural quartz. Part I: High purity natural quartz for industrial use. Part
II: High purity natural quartz markets for suppliers and users. Quartz Technology. Liberty
Corner, New Jersey, p 657
Martino R (2003) Las fajas de deformation ductil de la Sierras Pampeanas de Crdoba: una
resea general. Revista de la Association Geologica Argentina 58:549571
Martino RD, Guereschi AB, Sfragulla JA (2009) Petrology, structure and tectonic significance of
the Tuclame banded schists in the Sierras Pampeanas of Crdoba and its relationships with the
metamorphic basement of northwestern Argentina. J South Am Earth Sci 27:280298
136
G. Morteani et al.
Martino RD, Guereschi AB, Anzil A (2010) Metamorphic and tectonic evolution at 313600 S
across a deep crustal zone from the Sierra Chica of Cordoba, Sierras Pampeanas, Argentina.
J South Am Earth Sci 30:1228
Moore P (2005) High-purity quartz. Ind Minerals 455:5357
Morteani G, Preinfalk C, Spiegel W, Bonalumi A (1995) The Achala granitic complex and the
pegmatites of the Sierras Pampeanas (Northwest Argentina): A study in differentiation. Econ
Geol 90:636647
Mller A, Ihlen PM, Wanvik JE, Flem B (2007) High-purity quartz mineralisation in kyanite
quartzites, Norvay. Miner Deposita 42:523535
Munz IA (2001) Petroleum inclusions in sedimentary basins: systematics, analytical methods and
applications. Lithos 55:195212
Neuser RD, Bruhn F, Gtze J, Habermann D, Richter DK (1995) Kathodolumineszenz: Methodik
und Anwendung. Zentralblatt fr Geologie und Palontologie Teil I, H 1(2):287306
Otamendi JE, Nullo FE, Patio Douce AE, Fagiano M (1998) Geology, mineralogy and
geochemistry of syn-anatectic granites from the Achiras complex, Crdoba, Argentina; some
petrogenetic and geodynamic implications. J South Am Earth Sci 11:407423
Otamendi JE, Patino Douce AE, Demichelis AH (1999) Amphibolite to granulite transition in
aluminous greywackes from the Sierra de Comechingones, Crdoba, Argentina. J Metamorphic Geol 17:415434
Ramseyer K, Mullis J (1990) Factors influencing short-lived blue cathodoluminescence of alphaquartz. Am Mineral 75:791800
Rapela CW, Shaw DM (1979) Trace and major element models of granitoid genesis in the
Pampean Ranges, Argentina. Geochim Cosmochim Acta 43:11171129
Rapela CW, Heaman LM, McNutt RH (1982) Rb-Sr Geochronology of granitoid rocks from the
Pampean Ranges, Argentina. J Geol 90:574582
Rapela CW, Toselli A, Heaman L, Saavedra J (1990) Granite plutonism of the Sierras
Pampeanas, An inner cordilleran Paleozoic arc in the southern Andes. Geol Soc Am 241:77
90 Special paper
Rossman GR, Weis D, Wasserburg GJ (1987) Rb, Sr, Nd and Sm concentrations in quartz.
Geochim Cosmochim Acta 51:23252329
Siegesmund S, Steenken A, Martino RD, Wemmer K, Lopez de Luchi M, Frei R, Presnyakov S,
Guereschi A (2010) Time constraints on the tectonic evolution of the Eastern Sierras
Pampeanas (Central Argentina). Int J Earth Sci 99:11991226
Simpson C, Law RD, Gromet LP, Mir R, Northrup CJ (2003) Paleozoic deformation in the
Sierras de Crdoba and Sierra de la Minas, eastern Sierras Pampeanas, Argentina. J South Am
Earth Sci 15:749764
Sims JP, Ireland TR, Camacho A, Lyons P, Pieters PE, Skirrow RG, Stuart-Smith PG, Mir R
(1998) U-Pb, Th-Pb, and ArAr geochronology from the southern Sierras Pampeanas:
implications for the Paleozoic tectonic evolution ofr the western Proto-Andean Margin of
Gondwana, vol. 142, Geol Soc London Spec Publ 259281
Steenken A, Lpez de Luchi MG, Martinez Dopico C, Drobe M, Wemmer K, Siegesmund S
(2011) The Neoproterozoic-early Paleozoic metamorphic and magmatic evolution of the
Eastern Sierras Pampeanas: an overview. Int J Earth Sci 100:465488
Steenken A, Siegesmund S, Lpez de Luchi MG, Frei R, Wemmer K (2006) Neoproterozoic to
Early Palaeozoic events in the Sierra de San Luis: implications fort he Famatinian
geodynamics in the Eastern Sierras Pampeanas (Argentina). J Geol Soc 163:965982
Steenken A, Wemmer K, Martino RD, Lopez de Luchi M, Guareschi A, Siegesmund S (2010)
Post-Pampean cooling and uplift of the Sierras Pampeanas in the west of Crdoba (Central
Argentina). N Jb Geol Paleont Abh 256:235255
Tarantola A, Diamond LW, Stnitz H (2010) Modification of fluid inclusions in quartz by
deviatoric stress I: Experimentally induced changes in inclusion shapes and microstructures.
Contrib Min Pet 160:825843
Weil JA (1984) A review of electron spin spectroscopy and its application to the study of
paramagnetic defects in crystalline quartz. Phys Chem Mineral 10:149165
137
Weil JA (1993) A review of the EPR spectroscopy of the point defects in aa-quartz: The decade
19821992. In: Helms CR, Deal BE (eds) Physics and chemistry of SiO2 and the Si-SiO2
interface 2. Plenum Press, New York, pp 131144
Whitmeyer SJ, Simpson C (2003) High strain-rate deformation fabrics characterize a kilometresthick Paleozoic fault zone in the Eastern Sierras Pampeanas, central Argentina. J Struct Geol
25:909922
Wopenka B, Pasteris JD, Freeman JJ (1990) Analysis of individual fluid inclusions by Fourier
transform infrared and Raman microspectroscopy. Geochim Cosmochim Acta 54:519533
Chapter 6
139
140
R. Scholz et al.
6.1 Introduction
Quartz is an important mineral resource in Brazil. The first ever mentioned finding
of quartz crystals was in 1797 in Cristalina, Gois State, by explorers of gold and
emerald. The exploration of quartz intensified at the beginning of the Second
World War with strong participation of the United States Geological Survey
USGS (Campbell 1946; Johnston and Butler 1946) due to increasing demand of
crystal quartz (Arcoverde and Schobbenhaus 1991). Today Brazil is the leading
producer of gemological natural quartz crystals in the world (Drummond 2009),
with potential of growth in the production of metallurgical quartz and the glass
industry.
With a territory of up to 8,514,876.6 km2 and a complex geological evolution,
the environments for formation of quartz mineralizations are manifold. The
Brazilian quartz deposits occur in four different geological settings: (1) Neoproterozoic granitic pegmatites, (2) Neoproterozoic hydrothermal veins, (3) Mesozoic
basaltic sheets (geodes filled by amethyst and agate in hydrothermal conditions)
and (4) Cenozoic secondary deposits. A synthesis of the main Brazilian deposits
and their genetic aspects are presented in this study (Fig. 6.1 and Table 6.1).
Brazil has one of the largest resources of industrial quartz in the world, as well
as China, Madagascar, South Africa, Canada and Venezuela. The Brazilian
resources of quartz lascas were estimated in 17 Mt in 1980 decade, and around
60% of them are to be found in the Minas Gerais State (Alecrim 1982). After this
report, only in 2010 official data of estimated and measured resources were published. In 2005, total resources of industrial sand and lascas were measured up to
2,400 Mt. (MME 2010). The world resources of large natural crystals occur almost
exclusively in Brazil and, in smaller volumes, in Madagascar (DNPM 2006).
In Brazil the production of quartz is predominant of small mining companies
and of informal miners, both in the production of lascas and single crystals.
Electronic grade crystals are rare and their production is sporadic. The absence of
technological training does not allow the addition of value to the mineral good in
the stages of extraction and processing (DNPM 2006). Official production between
1996 and 2005 was measured in 1,143,497 tons of lascas and up to 40 Mt of
industrial sands (MME 2010), however the data is questionable.
The most important Brazilian industrial quartz resources are related to hydrothermal veins and secondary deposits. Transparent quartz crystals from Brazilian
pegmatites show minor importance for metallurgical and technological industries,
141
Fig. 6.1 Location and genetic types of the Brazilian quartz deposits. Localities marked on the
map are described in Table 6.1
however in the last 10 years the production is increasing, mainly for the usage as
gemstone including color treatment.
142
R. Scholz et al.
Table 6.1 Genetic aspects and type of mineralization of the Brazilian quartz deposits. Localities
marked on the map
Locality
Type of deposit
Type of quartz mineralization
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
Fronteira SudesteQuara
and Livramento
Salto do JacuSoledade
Alto Uruguai
Santa Catarina
Descalvado
Perube and So Vicente
Santa Maria de Jetib
Conselheiro Pena District
Gouveia
Sopa
Buenpolis
Serra do Cabral
Padre Paraso
Araua
Itamb
Gro Mogol
Brejinho das Ametistas
Chapada dos Veadeiros
Chapada dos Guimares
Novo Horizonte-Ibitira
Oliveira dos Brejinhos
Grota do Coxo
Pau DArco
Alto Bonito
Parelhas
Currais Novos
Geodes in basalts
Geodes in basalts
Geodes in basalts
Hydrothermal vein
Sedimentary
Sedimentary
Pegmatite
Pegmatite
Hydrothermal vein
Hydrothermal vein
Hydrothermal vein
Hydrothermal vein
Pegmatite
Pegmatite
Pegmatite
Hydrothermal vein
Hydrothermal vein
Hydrothermal vein
Hydrothermal vein
Hydrothermal vein
Hydrothermal vein
Hydrothermal vein
Sedimentary
Hydrothermal vein
Pegmatite
Pegmatite
Amethyst
Amethyst
Silex
Industrial sands
Industrial sands
Amethyst
Gemological quartz and crystals
Gemological and industrial quartz
Gemological and industrial quartz
Gemological and industrial quartz
Gemological and industrial quartz
Gemological quartz and crystals
Gemological quartz and crystals
Industrial quartz
Amethyst
Amethyst
Industrial quartz
Industrial quartz
Quartz with rutile and hematite inclusions
Gemological quartz
Amethyst
Amethyst
Amethyst
Industrial and rose quartz
Industrial and rose quartz
exploited are located in the municipalities of Picu, Pedra Lavrada, Nova Palmeira,
Carnaba dos Dantas, So Vincente do Serid, Juazeirinho, Parelhas, Currais
Novos and Equador. More than 750 of them occur as intrusion in garnet-cordierite
and/or sillimanite-biotite schists of the Serid Formation (Da Silva et al. 1995).
The BPP is historically important for its tantalum/tin ore, beryl, and minerals
for the ceramics and glass industry (quartz, feldspar and mica). It is also wellknown for its small production of gems, including aquamarine, morganite, garnet,
euclase and elbaite. Recently, it became famous worldwide for the production of
the Cu-bearing elbaite commercially known as Paraba Tourmaline.
The pegmatites of BPP were first classified by Johnston and Butler (1946) as
being either homogeneous or heterogeneous basing on their internal structure. The
homogeneous pegmatites are usually sterile, generally concordant, without internal
zoning. Heterogeneous pegmatites are rich in BeLi-Ta-Sn-minerals and present
internal zoning that show variable fabric and mineralogical composition.
143
Fig. 6.2 Brazilian quartz deposits: a core of rose quartz in the Alto do Feio pegmatite, Pedra
Lavrada, Paraba; b quartz vein at Gouveia in Minas Gerais; c industrial hydrothermal quartz
mine hosted in quartzite, Gouveia, Minas Gerais; d amethyst geode in basalt, Soledade, Rio
Grande do Sul (Gtze photo)
Quartz has been produced predominantly from zoned pegmatites that have a
massive quartz core. The quartz is either milky, rose, hyaline or smoked and is
extracted to feed, mainly glass, ceramic and the building industries. Secondarily,
perfect crystals of quartz, which are more appropriate for electronic components
and trade as gems and jewellery, have been found.
In the last 4 years most of the production of rose quartz comes from two
complex pegmatites, the Taboa and Alto do Feio pegmatites (Fig. 6.2a). The
pegmatites are situated, respectively, in the states of Rio Grande do Norte and
Paraba. Taboa and Alto do Feio zoned pegmatites have a core of rose and milky
quartz.
Generally, quartz lascas obtained by the miners are bought by larger companies
that refine the selection and sell it with more than 99% silica (Si02) and less than
0.01% Fe203 and 0.1% humidity to industries. The selling price of the lascas have
varied greatly in the last 10 years, but its average was around $ 0.23 per kilo FOB.
Due to the high degree of informality of the mines, no one has, to date, reliable
data of production, sale and export of quartz mined from this province, although
estimates of such data are available.
144
R. Scholz et al.
145
146
R. Scholz et al.
counties (Cassedanne 1991; Couto 2000) where, in the last 20 years, quartz
crystals with abundant rutile and/or hematite inclusions (rutilated quartz) have
acquired major commercial importance as gems and collection specimens. The
main part of the mineralization is hosted in the contact zone of volcanic beds with
siliciclastic Espinhao rocks. Giant crystals measuring several meters in length in
the c-axis direction have been discovered in this area.
In the region of Oliveira dos Brejinhos, quartz occurs as clear, transparent and
colorless crystals, some with hematite platelets. Quartz extraction occurs directly
from the veins, or more rarely in proximal colluvial deposits. Most other deposits
within this location are exploited in the same way, including those in the region of
northern Espinhao Range, also belonging to the Espinhao Supergroup. Prominent deposits, which in addition yield citrine quartz, are located in Brotas de
Macabas area (Couto 2000).
Amethyst deposits are widespread in the Bahia state and occur mainly in two
areas: Brejinho das Ametistas-Serra do Salto (Caetit and Licnio de Almeida
counties), and the Grota do Coxo mine (Jacobina). Occurrences in Brejinho das
Ametistas are secondary deposits, where amethyst chips and pebbles are erratically
scattered in the colluvial surface cover (Couto 2000). Primary mineralization
occurs in fractures cutting through quartzites of the Espinhao Supergroup in the
Brejinho das Ametistas and in the Serra do Salto (Salto Mountain, a local name for
the Espinhao Range) in the vicinity of the town Licnio de Almeida. In these
areas mineralizations are related to fracture zones that cut fin- to medium- grained
sandstone. Amethyst occurs as druses and geodes, in vugs, veins and fractures, or
as dissemination of crystals in sandstone. Most amethyst batches examined in this
region show good euhedral formation and uniform purple color.
In the Grota do Coxo mine, which is located in the Jacobina county, amethyst
crystals usually occur as druses and geodes filling fractures in the quartzites of the
Rio do Ouro Formation (Paleoproterozoic, Jacobina Group) in the eastern region
of the Espinhao Range (Cassedanne and Cassedanne 1979; Cassedanne 1991).
The crystals are variously colored from reddish violet to velvet violet, and display
face edges from 1 to 30 cm in length. Druses of amethyst crystals are sometimes
over half a meter wide, and contain brownish, usually opaque crystals (Cassedanne
and Cassedanne 1979).
147
quartzite. The amethyst crystals occur in veins, dikes and geodes. Large cavities,
sometimes connected, can be larger than 150 m.
148
R. Scholz et al.
149
150
R. Scholz et al.
being in iron fillings, burned lime or sand as a way to change their colors.
In addition, Kluge (1860) provided an extensive discussion of dyeing agates. But it
was only at the beginning of the twentieth century that the extensive work of Max
Bauer entitled Edelsteinkunde (first published in 1896) became a reference on
gemstone enhancing. In this work he describes the loss of the smoky color of
quartz or its conversion to yellow quartz, and the change of amethyst into citrine
after thermal treatments, along with dyeing of cracked and microcrystalline quartz,
the removal and restoration of the yellow color in tigers-eye among several other
processes. At that time, the quality of the lapidating industry had also reached a
stage where the optimal light reflection in the facets was perfected, as a way to
enhance the color of several gemstones and therefore improve their value. With the
advent of radioactivity, discovered at the end of the nineteenth and at the beginning of the twentieth centuries, its application to change the gemstone colors was
almost immediate. In his book Die Knstlichen Edelsteine (1926), Michel reported
the irradiation of diamonds, corundum, tourmalines, topazes, kunzites and quartz.
Besides the faceting technology, the development of gemstone irradiation became
the main process to increase the value of gemstones by enhancing their colors.
Nowadays, with superficial knowledge about the impurities related to the colors
and the formation of color centers, as well as the development of color-enhancing
processes, several recipes, in particular for quartz, topaz and tourmaline, using a
combination of irradiation (mostly gamma, but also highly-energetic electrons and
neutrons from linear accelerators and nuclear reactors) and thermal annealing, are
globally applied by the gemstone industry despite the fact that most of the recipes
for quartz, like those presented in (Nassau 1984), are empiric.
151
152
R. Scholz et al.
Fig. 6.3 Color varieties of Al-rich quartz treated with gamma irradiation and thermal treatment:
a green-gold quartz; b red-brown quartz; c brown quartz; and d black morion quartz
153
Fig. 6.4 Color varieties of Fe-rich quartz treated with gamma irradiation and/or thermal
treatment: a amethyst; b citrine (burned-amethyst); c prasiolite; and d snow-white quartz
Nassau (1984) distinguished two groups of a-quartz with different color treatment routes based on the primary impurity content: the amethyst group with
dominantly Fe impurities and the smoky quartz group with dominantly Al
impurities. Based on the Fe and Al impurities, which are responsible for the
majority of colors and color centers within each group, we suggest a refined
system of color treatment, in accordance with (Rykart 1995). Figure 6.5 shows the
typical routes for color enhancing of Brazilian quartz according to our work
Pinheiro et al. (1999), and several other authors, e.g. Nassau (1984), Nunes et al.
(2009), Nunes and Lameiras (2005), and Drummond et al. (2010). The requirements in terms of impurities, the treatment process, the most accepted color center
models from the literature and final colors with ordinary names for the gemstones
treated are given in this roadmap.
Very pure colorless quartz does not respond to any irradiation and heating
process. All colors produced are directly or indirectly related to impurities. If the
quartz is Fe-rich and its concentration is by far higher than that of Al, the main
color produced by irradiation is the violet color typical of natural amethysts. The
origin of this color has been controversial for a long time. Although the relationship between the violet colors with the Fe-content is known since Goethe
(1982), it was explicitly recognized in 1925 by Holden and von Klemm (Holden
1925). Lehmann and Bambauer (1973), as well as Lehmann (1975, 1977, 1978)
154
R. Scholz et al.
proposed a color center model in which the impurity Fe3+ substituting Si4+ loses an
electron, after natural irradiation, to an interstitial Fe3+ forming Fe4+ substitutional
and Fe2+ interstitial, whereas the violet color should be related to the Fe4+ center.
They also observed that the depth of the color depends on the right balance of
substitutional Fe3+ and interstitial Fe3+ and that the presence of hydrogen hindered
the violet color formation due to its reaction with substitutional Fe3+. Later,
Nassau (1984) revised this model, suggesting that the substitutional Fe4+ is
responsible for the violet color and proposed that no interstitial Fe3+ was necessary
for the formation of Fe4+. Hassan (1972) and Cohen (1975) proposed a controversial model in which only the interstitial Fe3+ is responsible for the violet color.
Later, Cohen (1985) acknowledged the interstitial Fe4+ as the violet color center
even though he controversially proposed that it could only be produced after
recombination of the Al-related smoky quartz color center. Meanwhile Weil
(1984, 1993, 1994) and Cox (1976), using electron paramagnetic resonance
spectroscopy (EPR), investigated several Fe-related centers in a-quartz. In particular, Cox (1976) observed Fe4+, and also investigated its optics (Cox 1977),
reinforcing the Fe4+ model for the amethyst color center. Recent results based on
the Mssbauer effect (Dedushenko et al. 2004) and X-ray absorption spectroscopy
(XAS) (Benedetto et al. 2010) seem to confirm this model. Therefore, where, in the
scheme above, the violet color appeared, we related it to the Fe4+ color center.
In samples containing Al3+ which substitutes Si4+, the trivalent Al is charge
compensated by monovalent ions like H+, Li+ and Na+, with concentrations of
[Al] C [H] ? [Li] ? [Na] (e.g. Bambauer 1961). Under gamma irradiation, an
electron ejects from an oxygen in the vicinity of the Al impurity and is trapped by
one of the charge compensating ions, neutralizing them (Nassau 1983). The
resulting AlO bound hole h+, [Al3+O4/h+]4-, which is isoelectronic to the normal
[SiO4]4- unit, is the color center responsible for the grey-smoky to black colors
(Nassau 1984). The H0 center is not stable at room temperature, and thus only Li
and Na can stabilize the smoky color center. As a result for the formation of smoky
quartz, it is required that [Li] ? [Na] [ [H] (Rykart 1995). It is well-known that
potassium is also a common interstitial charge compensator for Al, however, its
role on the stabilization of color centers is by far less understood (Rykart 1995).
Apart from these two color centers much less is known about the other colored
varieties of treated quartz. It is known, however, the presence of Fe2+ and the high
OH-content (in quartz grown in reducing-conditions) (Nassau 1984) for the
greenish Prasiolite varieties (Rose and Lietz 1954; Schuman 2007), the presence of
Fe3+ and Fe2O3 nanoparticles or a colloidal dispersion of hydrous ferric oxide in
citrine (Holden 1923); and the correlation of the yellow colors (Nassau and
Prescott 1975) in Al-rich quartz, with the Li-content. For other color varieties such
as the sky-blue quartz, the causes are still a matter of discussion (Nunes et al.
2009). Recently it has been discovered that long-term thermal treatment of OHrich quartz at temperatures above the a-/b-quartz transition (573800C) can result
in opal-like milky quartz (opaline quartz) with a slight blue tint to completely
opaque snow-white quartz (Krambrock et al., 2010, personal communication).
Fig. 6.5 Summary of all possible color-enhancing routes for Brazilian quartz that involves gamma
radiation and heat treatment. See text for details. The thick lines represent the treatments done
commercially
156
R. Scholz et al.
Despite the lack of knowledge about the relationship of some defects and optical
properties, the treatment routes for a large variety of colors are well established, at
least empirically, and details such as radiation dose are also known (Pinheiro et al.
1999). On the other hand, the correlation of the treatment result, i.e. color, with
particular quartz occurrences and the origin, for example most pegmatite quartz
yielding green-gold color (Drummond et al. 2010), are also well known. In order to
improve the color-enhancing processes for quartz and to limit losses, systematic
research of the spectroscopic properties and atomistic structure of color centers and
transition metal impurities, as well as their charge compensation mechanisms and
radiation/thermally-induced solid state reactions are necessary.
References
Agricola G (1556/1955) De Natura Fossilium, Special paper 65. Geogogical Society of America,
New York. De Re Metallica, Basel
Alecrim JD (1982) QuartzoAspectos Gerais, Mineralogia, Geologia. In: Recursos Minerais do
Estado de Minas Gerais. Metamig, Belo Horionte, pp 199202
Arcoverde WL, Schobbenhaus C (1991) Geologia do quartzo. In: Schobbenhaus C et al (eds)
Principais Depsitos Minerais do Brasil. Braslia, Ed. DNPM/CPRM, IV(C), pp 315324
Bambauer HU (1961) Spurenelementgehalte und c-Farbzentren in Quarzen aus Zerrklften der
Schweizer Alpen. Schweiz Mineral Petrogr Mitt 41(2):335369
Bauer M (1860/1968) Edelsteinkunde; published in English as precious stones, vol 2. Dover
Publications, New York
157
Bellieni G, Comin-Chiaramonti P, Ernesto M, Melfi AJ, Pacca IG, Piccirillo EM (1984) Flood
basalt to rhyolite suites in the southern Paran plateau (Brazil): paleomagnetism, petrogenesis
and geodynamic implications. J Petrol 25:579618
Benedetto F, DAcapito F, Fornaciai G, Innocenti M, Montegrossi G, Pardi LA, Tesi S,
Romanelli MA (2010) Fe K-edge XAS study of amethyst. Phys Chem Mineral 37(5):283289
Biringuccio V (1942) Pirotechnia. MIT Press, Cambridge
Boyle R (1672) An essay about the origin and virtues of gems. Hafner Pub. Co., New York
Caly ER (1927) The stockholm papyrus. An English translation with brief notes. J Chem Edu
4(8):9791002
Campbell DF (1946) Quartz crystal deposits in the state of Goiaz, Brazil. Econ Geol
41(8):773799
Cassedanne JP (1991) Tipologia das jazidas brasileiras de gemas. In: Schobbenhaus C et al (eds)
Principais Depsitos Minerais do Brasil. Braslia, Ed. DNPM/CPRM, pp 1752
Cassedanne JP, Cassedanne JO (1979) La mine damthyste de La Grota do Coxo: une mervelle
inconnue. Revist Gemmologie AFG 59:25
Cellini B (1967) The treatises of Benvenuto Cellini on goldsmithing and sculpure. Dover
Publications, New York
Chaves, MLSC, Karfunkel J, Quemnur JJ (1997) Depsitos de quartzo da regio de Batatal
(Diamantina, Minas Gerais). In: Simpsio de Geologia de Minas Gerais, vol 9. Anais, Ouro
Preto, pp 103104
Chaves MLSC, Favacho-Silva MD (2000) Ocorrncias singulares de quartzo gemolgico
(ametista, fum e citrino) na Serra do Espinhao, Minas Gerais. Revista da Escola de Minas
53:181186
Chaves MLSC (2007) O megaveio de quartzo da Serra da Catinga (Datas/Gouveia, MG).
Geocincias 26:109117
Cohen AJ (1975) On the color centers of iron in amethyst and synthetic quartz: a reply. Am
Mineral 60:338339
Cohen AJ (1985) Amethyst color in quartz, the result of radiation protection involving iron. Am
Mineral 70:11801185
Collyer TA, Mrtires RAC (1991) O depsito de ametista do Alto Bonito, municpio de Marab,
Par. In: Principais depsitos minerais do Brasil, vol 4A. Gemas e rochas ornamentais.
DNPM, Rio de Janeiro, pp 287293
Collyer TA, Mrtires RAC, Machado JIL (1991) O depsito de ametista de Pau Drco,
municpio de Conceio do Araguaia, Par. In: Principais depsitos minerais do Brasil, vol
4A. Gemas e rochas ornamentais. DNPM, Rio de Janeiro, pp 295302
Corra TE, Koppe JC, Costa JFCL, Moraes MAL (1994) Caracterizao geolgica e critrios de
prospeco de depsitos de ametista tipo Alto Uruguai, RS. Anais XXXVIII Congresso
Brasileiro de Geologia, Cambori, Santa Catarina 2:137138
Couto P (2000) Gemologic map of the state of Bahiaexplanatory text. CPRM/SMM/DNPM/
SICM Covenor, Salvador, p 76
Cox RT (1976) ESR of an S = 2 center in amethyst quartz and its possible identification as D4
ion Fe4+. J Phys C Solid State Phys 9(17):33553361
Cox RT (1977) Optical absorption of D4 ion Fe4+ in pleochroic amethyst quartz. J Phys C Solid
State Phys 10(22):46314643
Da Silva MRR, Hll R, Beurlen H (1995) Borborema pegmatitic province: geological and
geochemical characteristics. J South Earth Sci 8(3/4):355364
De Boot AB (1609/1961) Germmarum et Lapidum Historia, cited in J.R. Partington. A history of
chemistry, vol 2. Macmillan, London, pp 101102
Dedushenko SK, Makhina IB, Marin AA, Mukhanov VA, Perfiliev YD (2004) What oxidation
state of iron determines the amethyst colour? Hyperfine Interact 165(1):417422
DNPM (2006) Sumrio Mineral. Ministrio de Minas e EnergiaDepartamento Nacional de
Produo Mineral. p 304
DNPM (2011) Data from the Departamento Nacional de Produo Mineral, Instituto Brasileiro de
Gemas e Metais Preciosos (IBGM)
158
R. Scholz et al.
159
Pedrosa-Soares AC, Chaves M, Scholz R (2009) Eastern Brazilian pegmatite province. In: 4th
international symposium on granitic pegmaites, field trip guide, p 28
Pedrosa-Soares AC, Noce CM, Wiedemann CM, Pinto CP (2001) The Araua-West Congo
orogen in Brazil: an overview of a confined orogen formed during Gondwanland assembly.
Precambr Res 110:307323
Pedrosa-Soares, AC, Alkmim FF, Tack L, Noce CM, Babinski M, Silva LC, Martins-Neto MA
(2008) Similarities and differences between the Brazilian and African counterparts of the
Neoproterozoic Araua-West Congo orogen. In: Pankhurst RJ, Trouw RAJ., Brito Neves BB,
De Wit MJ (eds) West Gondwana: pre-cenozoic correlations across the South Atlantic
Region. Geological Society, London, Special Publications 294, pp 153172
Pedrosa-Soares AC, Campos CP, Noce C, Silva LC, Novo T, Roncato J, Castaeda C, Queiroga
G, Dantas E, Dussin I, Alkmim F (2011) Late neoproterozoicCambrian granitic magmatism
in the Araua orogen (Brazil), the Eastern Brazilian Pegmatite Province and related mineral
resources. In: Sial AN et al (eds) Granite-related ore deposits. Geological Society, London,
Special Publications 350, pp 2551
Pinheiro MVB, Lameiras ES, Krambrock K, Karfunkel J, da Silva JB (1999) The effect of the
gamma-irradiation dose combined with heat on the color enhancement of colourless quartz.
Aust Gemol 20:7
Porta JB (1955) Natural magic. Basic Books, New York
Proust D, Fontaine C (2007) Amethyst-bearing lava flows in the Paran Basin (Rio Grande do
Sul): cooling, vesiculation and formation of the geodic cavities. Geol Mag 144(1):5365
Reichmann F (2011) Personal communication from the Operational Director from the CBEEMBRARAD
Rose H, Lietz J (1954) Ein grn verfarbbarer Amethyst. Naturwissenschaften 41:448
Rykart R (1995) Quarz-MonographieDie Eigenheiten von Bergkristall, Rauchquarz, Amethyst,
Chalcedon, Achat. Opal und anderen Varietten, Ott Verlag Thun, Schwiz
Saadi A (1995) A geomorfologia como cincia de apoio ao planejamento urbano em Minas
gerais. Geonomos 5(2):14
Schmitt JCC, Camatti C, Barcellos RC (1991) Depsitos de ametista e gata no estado do Rio
Grande do Sul. In: Principais depsitos minerais do Brasil, vol. 4A. Gemas e rochas
ornamentais. DNPM, Rio de Janeiro, pp 271286
Schuman W (2007) Gemstones of the world, 3 Rev. Exp. edition. Sterling, New York
Scopel RM, Gomes MEBG, Formoso MLL, Proust D (1998) Derrames portadores de ametistas na
regio de Frederico Westphalen-IraiPlanalto-Ametista do Sul, RS-Brasil. Congresso
Uruguaio de Geologia 2, Actas, pp 243252
Theodorovicz AMG, Francisconi O, Ferrari AP (1991) Depsitos de areia industrial do litoral
paulista entre So Vicente e PerubeSo Paulo. In: Schobbenhaus C et al (eds) Principais
Depsitos Minerais do Brasil. Braslia, Ed. DNPM/CPRM, IV(C), pp 357363
Thomas R, Blankenburg HJ (1981) Erste Ergebnisse ber Einschlussuntersuchungen an Quartzen
aus Achatmandeln und Kugeln basischer und sauerer Vulkanite. Z Geol Wiss 9:625633
Van Schmus WR, Brito Neves BB, Williams IS, Hackspacher PC, Fetter AH, Dantas EL,
Babinski M (2003) The Serido Group of NE Brazil, a late Neoproterozoic pre- to syncollisional basin in West Gondwana: insights from SHRIMP U-Pb detrital zircon ages and
Sm-Nd crustal residence (TDM) ages. Precambr Res 127:287327
Weil JA (1984) A review of electron spin spectroscopy and its application to the study of
paramagnetic defects in crystalline quartz. Phys Chem Mineral 10:149165
Weil JA (1993) A review of the EPR spectroscopy of the point defects in a-quartz. The decade
19821992. In: Helms CR, Deal BE (eds) The physics and chemistry of SiO2 and the Si-SiO2
interface 2. Plenum Press, New York, pp 131144
Weil JA (1994) EPR of iron centers in silicon dioxide. Appl Magn Reson 6:116
Wyckoff D (1967) Albertus Magnus: book of minerals. Clarendon Press, Oxford
Chapter 7
7.1 Introduction
Quartz is one of the most abundant minerals in the Earths crust and is an
important material for not only traditional industries such as construction but also
high-tech electronic, optical and solar-energy applications. Point defects in quartz,
Z. Li Y. Pan (&)
Department of Geological Sciences, University of Saskatchewan,
Saskatoon, SK S7N 5E2, Canada
e-mail: yuanming.pan@usask.ca
161
162
Z. Li and Y. Pan
including those associated with silicon or oxygen vacancies, have long been the
subject of intense research (Weeks 1956, 1963; Silsbee 1961; Isoya et al. 1981;
Jani et al. 1983; Mashkovtsev et al. 2007; Botis et al. 2005, 2008; Perlson and Weil
2008; Weeks et al. 2008; Nilges et al. 2008, 2009; Pan et al. 2008, 2009;
Mashkovtsev and Pan 2011), because they are known to exert important (and often
adverse) effects on material properties and device performance and have wide
applications from radiation dosimetry to geochronology, mineral exploration and
quantum computer (e.g., Arends et al. 1963; Ikeya 1993; Stoneham et al. 2003;
Pan et al. 2006; Hu et al. 2008). For example, the oxygen-vacancy-related E0 1
center, first observed by using electron paramagnetic resonance (EPR) spectroscopy (Weeks 1956), is a fundamental and prototype defect in quartz and other
SiO2-based materials (e.g., glasses, thin films and nanoparticles) and has been
investigated by numerous experimental and theoretical studies (e.g., Weeks 1956;
Silsbee 1961; Feigl et al. 1974; Yip and Fowler 1975; Griscom 1980; Jani et al.
1983; Rudra and Fowler 1987; Edwards et al. 1988; Edwards and Fowler 1990;
Boero et al. 1997, 2003, 2004; Pacchioni et al. 1998; Blchl 2000; Carbonaro et al.
2001; Sulimov et al. 2002; Chadi 2003; Mysovsky et al. 2004; Stesmans et al.
2008; Perlson and Weil 2008; Weeks et al. 2008; Usami et al. 2009; Griscom 2011).
Silsbee (1961) first determined one strong and two weak 29Si hyperfine matrices
29
A( Si) for the E0 1 center in neutron-irradiated quartz at room temperature
(Table 7.1). Silsbee (1961) also noted another weak 29Si hyperfine structure with a
splitting of 0.07 mT in spectra measured with the magnetic field parallel to the
crystal c axis, where the strong and two weak 29Si hyperfine splittings are 40, 0.9
and 0.8 mT, respectively. Jani et al. (1983), in a combined single-crystal EPR and
electron nuclear double resonance (ENDOR) study, clarified the reference coordinate system and re-determined the electronic Zeeman matrix g and the three
A(29Si) matrices of the E0 1 center to higher precisions (Table 7.1), allowing the
principal axes of these parameter matrices to be compared with bond directions in
the quartz lattice (see also Perlson and Weil 2008).
One of the most salient features of the E0 1 center in quartz is its close association with the aluminum-associated oxygen hole center [AlO4]0 (Jani et al. 1983;
Toyoda and Hattori 2000; Usami et al. 2009). For example, Jani et al. (1983)
reported a correlation between the growth of the E0 1 center and the decay of the
[AlO4]0 center and suggested that the former formed from trapping of the unpaired
spin ejected from the latter. Toyoda and Hattori (2000) found that the gamma-rayinduced E0 1 center in a natural quartz sample that was annealed at 450C reaches
saturation at the dose of 200 kGy, whereas the [AlO4]0 center in this sample
continues to grow with further irradiation. Usami et al. (2009) confirmed the
observation of Toyoda and Hattori (2000) and found that the number of the E0 1
center in quartz samples containing small amounts of the [AlO4]0 center is small
and not all oxygen vacancies are converted. These results led Usami et al. (2009)
to conclude that the number of electrons ejected from the [AlO4]0 center plays an
important role in the formation of the E0 1 center in quartz.
Semi-empirical calculations have been credited for the widely cited singleoxygen-vacancy model for the E0 1 center in quartz and amorphous SiO2 (Feigl
1,271
1,091
1,091
27.43
22.00
22.00
25.81
20.66
20.66
2.0a
67
\
\
39
\
\
55
\
\
h()
-10
\
\
46
\
\
-157
\
\
/()
-1269.72
-1094.53
-1095.02
-27.53
-22.14
-22.27
-26.01
-20.91
-21.04
Ak/h
114.1
132.1
128.3
140.7
104.6
125.5
58.9
35.0
104.4
h()
229.7
115.9
340.4
284.5
33.1
133.9
260.9
111.4
179.9
/()
1434.5
1248.5
1248.5
61.9
45.3
45.9
67.3
51.3
50.3
11.5
10.2
9.6
Ak/h
113.0
159.0
101.3
141.6
119.3
127.0
50.7
25.3
93.3
131.4
99.6
148.7
h()
228.3
70.4
318.5
281.8
50.3
157.3
252.9
78.6
168.6
347.3
186.6
83.3
/()
-1272
-1109
-1109
-39.2
-33.2
-33.2
-42.0
-36.4
-36.4
2.8
0.0
0.0
Ak/h
112
\
\
149
\
\
56
\
\
127
\
\
h()
Blchl (2000)
29
233
\
\
281
\
\
267
\
\
13
\
\
/()
-1399.3
-1221.9
-1220.5
-30.3
-25.8
-25.5
-32.5
-27.7
-27.5
6.2
3.6
3.9
Ak/h
M(2004)
M(2004) is Mysovsky et al. (2004), in which the directions of the principal hyperfine axes were not available
Silsbee (1961) reported a third weak 29Si hyperfine splitting of 0.07 mT in spectra measured with the magnetic field parallel to the crystal c axis
Weak3
Weak2
Weak1
Strong
Ak/h
Silsbee (1961)
164
Z. Li and Y. Pan
et al. 1974; Yip and Fowler 1975; Rudra and Fowler 1987; Edwards et al. 1988;
Edwards and Fowler 1990). For example, MINDO/3 calculations by Feigl et al.
(1974) and Yip and Fowler (1975), using a Si2O6 cluster, gave the positively (+1)
charged oxygen vacancy V(O)+ model with the unpaired spin largely localized on
one of the two asymmetrically relaxed [SiO3] groups. Rudra and Fowler (1987),
using larger Si8O7 and Si5O16 clusters, improved this model and showed that the Si
atom containing little unpaired spin relaxes toward or even passes through the
basal oxygen plane to form the planar and puckered configurations. The puckered
configuration was adopted and supported by several subsequent studies using
larger clusters (Snyder and Fowler 1993), including the embedded-cluster treatment (Giordano et al. 2007) and Car-Parrinello molecular dynamics (Boero et al.
1997), both of which explicitly consider the valence electrons only. For example,
molecular dynamics calculations by Boero et al. (1997) supported the puckered
configuration, but their calculated 29Si hyperfine coupling constants differ by
[20% for the strong 29Si hyperfine structure and as much as 200% for the two
weak 29Si hyperfine structures (Table 7.1). First-principles calculations by Blchl
(2000) better reproduced the values for the strong 29Si hyperfine structure but
yielded values [40% too large for the two weak 29Si hyperfine structures
(Table 7.1).
However, Fiorentini and Carbonaro (1997) suggested that +1 charge state of the
oxygen vacancy is not stable and therefore can not be a candidate E0 1 center.
Fiorentini and Carbonaro (1997) proposed the -3 charge state of the oxygen
vacancy as an alternative model but did not calculate 29Si hyperfine parameters for
comparison with the experimental data. Carbonaro et al. (1997) noted that the
ground state of the oxygen vacancy in undoped quartz is the neutral charge state,
whereas moderate p- and n-type dopings favor the +3 and -3 states, respectively.
Similarly, Chadi (2003) suggested that the +1 charge state of the oxygen vacancy
is not stable and proposed V(SiO3)+ and V(SiO4)+ as possible models for the E0 1
center in quartz. On the other hand, periodic first-principles calculations by
Carbonaro et al. (2001) provided compelling evidence for thermodynamic stability
of the positively charged oxygen vacancy in quartz. Mysovsky et al. (2004)
recalculated the puckered configuration in amorphous SiO2 with the embeddedcluster method and noted that all calculated 29Si hyperfine values are overestimated by *10% (Table 7.1). Obviously, significant questions remain about the
structural model for the E0 1 center in quartz and amorphous SiO2 (Weeks et al.
2008; Griscom 2011). Also, the close association between the E0 1 and [AlO4]0
centers in both synthetic and natural quartz, as documented by EPR studies (Jani
et al. 1983; Toyoda and Hattori 2000; Usami et al. 2009), has not been addressed
by any previous theoretical calculations.
Accordingly, we have conducted new periodic first-principles calculations with
all-electron Gaussian-type basis sets to further investigate the geometry and
electronic structure of the E0 1 center in quartz. In particular, our calculations
re-evaluate the traditional V(O)+ model but focus on a new tri-vacancy model
containing a silicon and two oxygen vacancies. These models are evaluated on the
basis of their abilities in reproducing the experimental 29Si hyperfine parameters in
165
oxygen atom is connected to two silicon atoms with a short bond (SB) of 1.607
and a long bond (LB) of 1.611 at room temperature. This difference in the SiO
bond distances, which becomes even smaller at low temperatures (Le Page et al.
1980), has been suggested by Feigl et al. (1974) to be responsible for the asymmetric relaxation that pushes the unpaired spin to localize on the Si0 atom on the
SB site of the oxygen vacancy, while leaves Si1 on the LB side almost spin free
(Fig. 7.1a).
166
Z. Li and Y. Pan
Fig. 7.1 Local structure (a) and electron density (b) of the E0 1 center in quartz calculated from
the single-oxygen-vacancy V(O)+ model. Labels of silicon atoms are similar to those in Jani et al.
(1983). Contours are at intervals of 0.005 e/bohr3 and from -0.01 to 0.415 e bohr-3
7:1a
7:1b
Here subscript i takes xx, yy, zz, or simply 1, 2, 3. The EPR experiments usually
report Ai defined as Ai = aiso ? Ti. l0 is the permeability of vacuum, gN is the
29
Sis nuclear g-factor and ge is the electronic g-factor, bN and be are the nuclear
and Bohr magnetons, respectively, \ qspin(rA) [ is the expectation value of the
spin density at the nucleus A (29Si) at point rA, rA = |r 2 rA|, rAi = (r rA)i, the
167
V(O) with Al
V(SiO2)+
Si0Si1
Si1OB
\Si1O[
Si0Si1
AlOB
\AlO[
Si0Si1
Si1OB
\Si1O[
Si0Al
AlOB
\AlO[
Si0Al
AlOB
\AlO[
4.606
3.480
1.587
3.986
1.988
1.692
4.543
3.369
1.623
4.277
4.296
1.722
4.929
2.742
1.670
4.32
1.91
/
/
/
/
/
/
/
/
/
/
/
/
/
4.06
1.81
/
/
/
/
/
/
/
/
/
/
/
/
/
Refc
Refd
4.358
1.852
/
/
/
/
/
/
/
/
/
/
/
/
/
4.46
1.82
/
/
/
/
/
/
/
/
/
/
/
/
/
168
Z. Li and Y. Pan
Figure 7.1b shows that the spin density on Si0 is 0.818 e and that small spin
densities (0.028, 0.043 and 0.053 e) on its three bonded oxygen atoms are also
apparent, which are important in accounting for the two weak 29Si hyperfine
tensors (Silsbee 1961; Jani et al. 1983). The calculated spin densities on Si1, Si2,
Si3 and Si4 are 0.012, 0.012, 0.010 and 0.009 e, respectively. Table 7.3 shows that
our calculated 29Si hyperfine parameters are in better agreement with experimental
EPR results than previous theoretical calculations (Boero et al. 1997; Blchl 2000;
Mysovsky et al. 2004). Nevertheless, the two weak 29Si hyperfine tensors are still
not well predicted (Table 7.3). Also, significant discrepancies are apparent
between the calculated and experimental directions of the principal hyperfine axes
(Table 7.3). Moreover, the calculated principal values for the third weak 29Si
hyperfine structure are overestimated (Table 7.3).
Our calculation for the single-oxygen-vacancy model with an Al impurity at the
Si1 site improves the calculated hfcc for two of three weak 29Si hyperfine structures (Table 7.3), suggesting an important role of Al impurity for the E0 1 center.
-1221.5
-1081.0
-1081.8
-34.3
-28.8
-28.7
-18.3
-14.3
-14.8
-14.1
-10.9
-10.5
124.2
125.0
126.6
118.0
143.4
111.5
38.0
52.6
95.6
102.5
126.9
39.6
h()
221.5
103.1
341.8
302.7
18.6
140.6
288.3
119.8
145.5
79.8
59.3
154.2
/()
-1230.4
-1066.7
-1066.8
-26.8
-21.9
-21.7
-17.0
-12.8
-13.0
-1.4
-1.1
-1.1
Ak/h
109.4
104.4
155.5
106.6
100.6
160.0
74.1
44.5
130.2
104.6
132.9
36.5
h()
V(O) with Al
216.9
121.7
357.4
271.8
118.5
127.4
241.3
117.1
206.4
86.6
42.6
130.9
/()
-1243.3
-1076.6
-1077.1
-29.9
-24.0
-24.3
-27.8
-22.2
-22.5
3.3
2.6
1.6
Ak/h
V(SiO2)+
See text for details about the structural models investigated in this study
Weak3
Weak2
Weak1
Strong
Ak/h
V(O)+
115.5
145.2
112.5
149.1
91.6
118.3
55.3
42.5
111.4
123.7
84.1
34.3
h()
234.5
101.9
335.9
277.2
27.2
123.5
261.4
122.4
187.1
70.9
37.0
178.2
/()
-1197.0
-1030.0
-1030.4
-28.9
-23.1
-23.2
-9.8
-6.7
-7.2
2.4
1.2
1.2
Ak/h
127.3
132.7
124.6
148.0
113.9
110.1
23.5
76.6
109.0
124.9
52.2
57.0
h()
236.6
101.9
347.0
265.6
40.5
139.8
260.1
136.9
222.3
72.5
15.3
135.5
/()
Table 7.3 29Si hyperfine coupling constants (MHz) and directions of the E0 1 center in quartz from periodic PBE0 calculations
29
Si
One-oxygen-vacancy model
Tri-vacancy model
-1282.7
-1114.0
-1114.5
-32.1
-26.0
-26.1
-23.4
-18.5
-18.7
-3.5
-2.7
-2.5
Ak/h
117.7
133.8
123.5
144.2
103.0
122.7
42.8
58.0
115.1
103.6
15.5
82.7
h()
234.6
114.4
344.9
279.3
27.9
126.4
260.6
128.1
201.0
59.8
30.5
148.0
/()
170
Z. Li and Y. Pan
Fig. 7.2 Local structure (a) and electron density (b) of the E0 1 center in quartz calculated from
the tri-vacancy model without an Al impurity. Note that labels of Si0, Si2, Si3 and Si4 are
equivalent to those in Fig. 7.1, but Si1 is not. Also note the presence of the unexpected spin on
the O1 atom. Contours are at intervals of 0.005 e/bohr3 and from -0.01 to 0.415 e/bohr3
Our tri-vacancy model calculations started with the removal of the two oxygen
atoms and the central Si atoms connected by the long bonds, giving the Si0 and
Si00 atoms a dangling short bond each (Fig. 7.2a). The optimized structure turns
out to be very interesting, because it does not form the expected two E0 1 centers
(i.e., biradical or triplet state; Bossoli et al. 1982; Mashkovtsev et al. 2007;
Mashkovtsev and Pan 2011). Rather, the Si00 atom is relaxed to link to a fourth
), and
oxygen atom (O2) from a neighboring SiO4 group (i.e., Si00 O2 = 2.162
the unpaired spin localizes on the O1 atom of another SiO4 group at the Si1 site
(Fig. 7.2b). Nevertheless, the Si0 atom contains 0.846 e, hence an E0 1 center.
In fact, the calculated hyperfine parameters from this model, including the
directions of the principal axes, best reproduce the experimental data (Jani et al.
1983), especially for the two difficult weak 29Si hyperfine structures (Table 3).
in this tri-vacancy model is
Interestingly, the relaxed Si0Si1 distance of 4.543
only 0.06 shorter than the asymmetrically relaxed SiSi pair in the V(O)+ model
(Table 7.2), indicating similar geometry and spin structure between these two
models to account for their similar hyperfine coupling constants. However, the Si1
atom from this tri-vacancy model contains an appreciable amount of the unpaired
spin (-0.017 e), giving rise to significant hyperfine constants Ak/h of 41.1, 37.4
and 39.3 MHz, which have not been observed in EPR experiments (Silsbee 1961;
Jani et al. 1983). These positive hyperfine values indicate their origin from spin
polarization due to the strong spin on the oxygen atom. We tested the possibility of
removing the unexpected spin on the oxygen atom, but the fully optimized
structure has the unpaired spin delocalized from Si0, hence inconsistent with the
E0 1 center.
Further calculations were made for this tri-vacancy model with an Al atom at
the Si1 site. This procedure removed the unexpected spin on the O1 atom that
, while the other three
forms a weak bond with the Al atom at a distance of 1.905
AlO bonds are *1.722 (Fig. 7.3; Table 7.2). This structure gives even better
directions of the principal hyperfine axes, but the principal values for one of the
weak 29Si hyperfine structures are not well reproduced (Table 7.3). The geometry
171
Fig. 7.3 Local structure (a) and electron density (b) of the E0 1 center in quartz calculated from
the tri-vacancy model with an Al atom at the Si1 site. Atom labels are similar to those in Fig. 7.2.
. Contours are at intervals of
Note that the AlO4 group has an elongated AlO1 bond of 1.906
0.005 e/bohr3 and from -0.01 to 0.415 e/bohr3
of this AlO4 group, including the elongated AlO bond, is closely comparable to
those of the well-established [AlO4]0 and [AlO4/M+]+ centers in quartz (Nuttall
and Weil 1981; Walsby et al. 2003; To et al. 2005; Botis and Pan 2009, 2011).
Another interesting feature is that the O1 and O2 atoms are relaxed to form an
(Fig. 7.3a), which is close to the peroxy type
electron-rich OO bond of 1.48
(Edwards and Fowler 1982; Nilges et al. 2009; Pan et al. 2009). This electron-rich
OO bond represents an ideal precursor for trapping a hole to form superoxide
(O2-) radicals (Nilges et al. 2009; Pan et al. 2009). Such superoxide (O2-) radicals, which are close analogues of the peroxy radical and the non-bridging oxygen
hole center (NBOHC) in amorphous SiO2 (Friebele et al. 1979; Griscom and
Friebele 1981; Edwards and Fowler 1982; Uchino et al. 2001), have been found in
association with the E0 1 center in both artificially irradiated quartz and natural
quartz from uranium deposits (Hu et al. 2008; Nilges et al. 2008, 2009; Pan et al.
2008, 2009). Therefore, our calculations of the new tri-vacancy model with an Al
impurity not only better reproduces the experimental EPR data for the E0 1 center in
quartz (Table 7.3) but can now account for its close association with the [AlO4]0
and O2- centers (Jani et al. 1983; Toyoda and Hattori 2000; Nilges et al. 2008,
2009; Pan et al. 2008, 2009; Usami et al. 2009). Further calculations for these pairs
of electronhole paramagnetic centers (i.e., E0 1 and [AlO4]0 vs. E0 1 and O2-) in
quartz are currently underway.
Additional calculations were made for this tri-vacancy model by allowing the O1
atom to relax away from the Al atom. The resulting configuration with an AlO3 group
(Fig. 7.4a) is 1.061 eV more stable than that with the AlO4 group. This AlO3 group is
also slightly puckered (3) and contains a fourth nearest oxygen atom at a distance of
(Table 7.2), which is considerably shorter than the Si1O distance of
2.742
B
3.480 Afrom
our V(O)+ calculation. This AlO3 group represents a precursor for
trapping an electron to form the Al-equivalent E0 1 center, which has been reported to
occur in irradiated vitreous silica (Brower 1979) but has not yet been found in quartz.
also occurs in this
Figure 7.4 shows that the peroxy O1O2 bond of 1.45
172
Z. Li and Y. Pan
Fig. 7.4 Local structure (a) and electron density (b) of the E0 1 center in quartz calculated from
the tri-vacancy-model with an Al atom at the Si1 site. Atom labels are similar to those in Figs. 7.2
and 7.3. Note that the O1 atom is relaxed away from the AlO3 group. Contours are at intervals of
0.005 e/bohr3 and from -0.01 to 0.415 e/bohr3
configuration and that the spin density on Si0 spreads toward this bond. Moreover,
which is
the distance between Si0 and the middle of this OO bond is only 3.91 A,
shorter than the Si0Al distance (Table 7.2). Therefore, the electrons on the OO
bond may reduce the electron deficiency of Si0 and thus stabilize this configuration.
This may explain the stability of the E0 1 center in quartz and other SiO2-based
materials (Weeks 1956; Silsbee 1961; Griscom 1980; Jani et al. 1983; Weeks et al.
2008; Perlson and Weil 2008). This variant of the tri-vacancy model with an Al
impurity also well reproduces the experimental 29Si hyperfine parameters in both
principal values and directions (Table 7.3).
Acknowledgments We thank Prof. Jens Gtze for invitation to the special session on quartz at
BHT2011, Drs. Robert Mckel and Michael Pltze for manuscript review and handling, and
Natural Science and Engineering Research Council (NSERC) of Canada for financial support. All
calculations in this research have been enabled by the use of Westgrid computing resources,
which are funded in part by the Canadian Foundation for Innovation, Alberta Innovation and
Science, BC Advanced Education, and the participating research institutions. Westgrid equipment
is provided by IBM, Hewlett Packard and SGI.
References
Adamo C, Barone V (1999) Toward reliable density functional methods without adjustable
parameters: the PBE0 model. J Chem Phys 110:61586170
Arends J, Dekker AJ, Perdok WG (1963) Color centers in quartz produced by crushing. Phys Stat
Sol 3:22752279
Blchl PE (2000) First-principles calculations of defects in oxygen-deficient silica exposed to
hydrogen. Phys Rev B 62:61586179
Boero M, Pasquarello A, Sarnthein J, Car R (1997) Structure and hyperfine parameters of E0 1
centers in a-quartz and in vitreous SiO2. Phys Rev Lett 78:887890
173
174
Z. Li and Y. Pan
175
Toyoda S, Hattori W (2000) Formation and decay of the E0 1 center and its precursor. Appl Rad
Isotop 52:13511356
Uchino T, Takahashi M, Yoko T (2001) Structure and generation mechanism of the peroxyradical defect in amorphous silica. Phys Rev Lett 80:45604563
Usami T, Toyoda S, Bahadur H, Srivastava AK, Nishido H (2009) Characterization of the E0 1
center in quartz: role of aluminum hole centers and oxygen vacancies. Phys B 404:38193823
Walsby CJ, Lees NS, Claridge RFC, Weil JA (2003) The magnetic properties of oxygen-hole
aluminum centers in crystalline SiO2. VI. A stable AlO4/Li centre. Can J Phys 81:583598
Weeks RA (1956) Paramagnetic resonance of lattice defects in irradiated quartz. J Appl Phys
27:13761381
Weeks RA (1963) Paramagnetic spectra of E0 2 centers in crystalline quartz. Phys Rev
130:570576
Weeks RA, Magruder RH, Stesmans A (2008) Review of some experiments in the 50 year saga
of the E0 center and suggestions for future research. J Non-Cryst Solids 354:208216
Yip KL, Fowler WB (1975) Electronic structure of E0 1 centers in SiO2. Phys Rev B
11:23272338
Chapter 8
Gamma-Irradiation Dependency
of EPR and TL-Spectra of Quartz
Michael Pltze, Dieter Wolf and Matthias R. Krbetschek
M. Pltze (&)
ClayLab, Institute for Geotechnical Engineering,
ETH Zurich, 8093 Zurich, Switzerland
e-mail: michael.ploetze@igt.baug.ethz.ch
D. Wolf
Institute of Mineralogy, TU Bergakademie Freiberg,
Brennhausgasse 14, 09596 Freiberg, Germany
M. R. Krbetschek
Institute for Applied Physics, TU Bergakademie Freiberg,
Leipziger Str 23, 09596 Freiberg, Germany
e-mail: quatmi@mailserver.tu-freiberg.de
177
178
M. Pltze et al.
as electron traps for the TL peaks at 150200 C/330340 nm, 200 C/510 nm and
280 C/470510 nm, whereas oxygen-vacancy-centres for the first peak and the
[AlO4]0-centres for the other peaks are working as recombination sites.
8.1 Introduction
In the past 50 years a large number of paramagnetic centres in quartz have been
detected and crystallographically described using EPR. The results of more than 200
publications about this topic were included in the reviews of Weil 1984 and 1993.
Structural defects are arising from impurities as well as vacancies in the
crystalline structure. There are different mechanisms of trace-element input into
the quartz. Trace elements in quartz are incorporated into the crystal structure or
bound to microinclusions. Due to its structure, quartz is considered to incorporate
only small amounts of impurities in its crystal lattice. The most elements are bound
to fluid inclusions and to microscopic mineral inclusions (Blankenburg et al. 1994;
Gtze et al. 1997, 2004). The structural incorporation of Al, Ge, Ti, Ga, Fe or P
into the Si tetrahedron is well investigated. The charge deficit is compensated by
alkali ions (Li, Na, K) or hydrogen (or Cu, Ag, etc.), which are distributed in
structural channels parallel to the c-axis.
Paramagnetic centres can be formed by incorporation of a priori paramagnetic
ions such as Mn2+, Fe3+ or REE in the proper valence state or by ionizing radiation
which produces charged electron or electron hole centres on structural defects or
atomic vacancies (Table 8.1). Such radiation defects can be used for dating or
radiation dose measurements (Ikeya 1993). The spectral analysis of the cathodoluminescence (CL) emission in combination with electron paramagnetic resonance
(EPR) measurements and/or trace-element analysis allows to study extrinsic (trace
elements) or intrinsic (lattice defects) point defects in the quartz structure. The
defects causing the different CL emissions in quartz often reflect the specific
physicochemical conditions of crystal growth and therefore, can be used as a
signature of genetic conditions (Gtze et al. 2001, 2004, 2005; Botis et al. 2005).
The Al centre is caused by substitution of Al3+ for Si4+ with an electron hole
trapped by a nonbonding 2p orbital at one of the four nearest O2- ions, forming an
O- type centre. This was the paramagnetic centre discovered first in quartz by
Griffiths et al. (1954). The precursor state for this centre is the diamagnetic [AlO4/
M+]0 associated with an adjacent charge compensating cation M+ (H+, Li+).
During irradiation at 295 K the M+-ion may diffuse away yielding the paramagnetic [AlO4]0 (Nuttall and Weil 1981). The EPR of the centre is observable at
temperatures lower than 150 K and was measured at 70 K, at which the line
intensity is maximum. For polycrystalline samples the EPR spectrum of [AlO4]0
exhibits a multiline spectrum of superimposed 6-line hyperfine patterns (Fig. 8.1).
+ 0
The [TiO4 /M ] centres are produced by irradiation of the diamagnetic precursor
0
3+
[TiO4] (Ti , i.e., electron centre at Ti4+) which is formed by substitution of Ti4+
for Si4+ at the Si position, where charge compensation is achieved by a proton or
179
Table 8.1 Classification of selected paramagnetic centres in natural quartz (notation after Weil
1984)
Impurities
Vacancies
Substitutional
for Si4+
Interstitial
[TiO4]- ] [TiO4/M+]0
[GeO4]- [GeO4/M+]
[AlO4]0
[FeO4]0
[H4O4]+ [H3O4]0
[FeO4/M+]0 (S)
[FeO4]-
Li+ ion at a channel position nearby (Wright et al. 1963; Bershov 1970; Rinneberg
and Weil 1972). These centres can be observed between 5 and 120 K. The
spectrum of the polycrystalline sample exhibits a characteristic pattern that
consists of three well-separated doublets for the [TiO4-/H+]0 and quartets for
[TiO4-/Li+]0 (Fig. 8.1). For both centres two types can be distinguished (Okada
et al. 1971; Rinneberg and Weil 1972): One stable at temperatures below 100 K
and the other one stable at room temperature. Only the type which is stable at room
temperature is observable after c-irradiation at 295 K.
The Fe-related centres has been shown to be the uncompensated [FeO4]- centre
(Mombourquette et al. 1986), and the compensated so called S centres [FeO4/M+]0,
where M = H+, Li+ and Na+ (Stegger and Lehmann 1989a, b; Mombourquette
et al. 1989).
180
M. Pltze et al.
365
[20
420490
100
100
100
110
180
[200
200
220
230
380
470
380, 470
280, 330
230
470
280
300
300
330
330
350
340
375
380
440480
470
380
450
470
620
181
Authors
Serebrennikov et al. (1982)
Lysakov et al. (1969)
McKeever (1984)
McKeever (1991)
McKeever (1991)
Martini et al. (1995)
Martini et al. (1985)
Arnold (1976)
Jani et al. (1983)
Marfunin (1979)
Ti4+
[AlO4]0
Li+, Na+
Medlin (1963)
Arnold (1976)
Batrak (1958)
?
[AlO4]0
?
Mejdahl (1986)
Hashimoto et al. (1987)
Hashimoto et al. (1987);
Ichikawa (1967)
8.2 Experimental
Spectrally resolved TL and EPR measurements have been carried out on quartz
from the tin tungsten deposit Ehrenfriedersdorf (Germany), located in the exocontact of a Hercynian LiF granite intrusion. The samples are quartz from veins
(Q-2, Q-51) and from the wall rock gneiss (Q-101). The second sample set consists
of quartz from a drilling profile in the Eldzhurtinskiy Granite (Northern Caucasus,
Russia) (Lyakhovitch and Gurbanov 1992).
The quartz grains were carefully crushed to a grain size\200 lm and sieved in the
fraction of 100200 lm. This fraction was treated with HCl to remove carbonates
and with HF to remove feldspars, cleaned with distilled water, dried at 25 C, and
hand-picked under a binocular microscope in order to get pure quartz samples.
The separated quartz samples were analysed for trace-element content by
atomic emission and absorption spectroscopy.
Before irradiation, the samples were heated at 400 C for 5 h to anneal the
paramagnetic centres formed by natural irradiation. After thermal treatment the
182
M. Pltze et al.
samples were divided in aliquots and c-irradiated with a 60Co source at room
temperature in separate doses from 70 Gy to 1.5 9 107 Gy and a dose rate of appr.
1.5 9 102 Gy/h. For doses higher than 1 9 103 Gy the dose rate was 1 9 104 Gy/h.
The sample sets were stored at 250 K after the irradiation for 2 weeks.
The paramagnetic centres of polycrystalline quartz samples were investigated
by EPR at X-band frequencies (9.2 GHz) at 20, 70, and 295 K. The spectra were
recorded by a Varian E-line spectrometer. The sample weight was 20 mg. The
influence of technical parameters such as modulation amplitude, microwave
power, temperature, scan time etc. on the spectra was checked for the optimal
settings for recording the spectra. The sample temperature was controlled with a
low temperature unit based on a helium gas flow device (Oxford ESR 900A).
These settings (modulation field HM = 1 G, microwave power p = 7 mW,
temperature T = 70 K for [AlO4]0 and [TiO4-/Li+]0 centres) were kept constant
throughout all the measurements to allow correct comparison between the signal
intensities of different spectra. The intensity of [TiO4-/M+]0 centres was measured using the same conditions as for Al centres. The concentration of the
paramagnetic centres was determined as peak to peak or peak to base intensity at
the analytical lines (Moiseev 1985). The specific peak positions of the paramagnetic centres were drawn from simulated spectra with the data from literature (Nettar and Villafranca 1985). The variation of intensity detected by
repeated measurements of selected analytical lines was up to 10%. The concentration of Al centres was quantified using a reference sample with known
[AlO4]0 concentration (Moiseev 1985). All other centres were calculated in
relative amounts.
Spectrally resolved TL measurements were carried out in the temperature range
from 50 to 350 C (heating rate 2 K/s) with a laboratory made CCD-camera based
high sensitivity TL/OSL-spectrometer (Rieser et al. 1994). With a holographic
concave grating the spectrum range of 200800 nm is dispersed onto a N2(liq)
cooled Astromed CCD-chip. To obtain the whole spectral data to make a temperature/wavelength 3-D-plot 40 short exposures were made. Further data processing on a PC puts these slices with the corresponding temperature and displays
the data as 3-D-plots or contour maps.
183
Table 8.3 Content of selected trace elements (ppm) in the quartz samples
Sample
Al
Ti
Ge
Fe
Na
Li
146
750
66
300
5
47
\1
29
176
157
103
407
10.8
2.4
\2.4
47.0
0.9
0.7
nd
nd
20
30
24
46
were detected (Figs. 8.1 and 8.4). The [AlO4]0 and [TiO4-/Li+]0 were the most
intense detected centres whereas the [GeO4-/M+]0 centres show only very weak
intensities. The EPR-spectra of the various vein quartzes (Q-2, Q-51) are relatively
similar. The quartz from the wall rock (Q-101) with the lowest impurity content
shows lower intensity of [AlO4]0 and no [TiO4-/Li+]0 centres. Consequentially,
the quartz with the highest content of impurities (granite quartz 34-2/90) shows
also the highest EPR intensities. The differences in [AlO4]0 centre concentrations
and the occurrence of [TiO4-/Li+]0 between the samples indicate their different
genesis. The higher amounts of the centres in irradiated samples from veins can be
correlated with the higher amounts of Al and Ti substituted for Si. The difference
may be indicative for the specific H and Li concentrations in the respective
hydrothermal solutions.
The investigation of the relationship between the EPR intensity of paramagnetic
centres and the c-irradiation dose reveal different behaviour for different paramagnetic centres. Generally, c-irradiation caused a transformation of diamagnetic
precursor centres into these paramagnetic centres resulting in an increase of the EPR
intensity. The intensity of EPR-spectra of the [AlO4]0 centre increases with the
c-dose and reaches saturation at a dose of about 1 9 106 Gy (Fig. 8.2). This irradiation behaviour corresponds with the time dependent decreasing intensity of the
380 nm cathodoluminescence emission (Gtze et al. 2001). The related diamagnetic
Al/Li+ defect is transformed into the paramagnetic Al-defect during the e- irradiation. The [TiO4-/Li+]0 spectrum shows the highest intensity already after irradiation with 5 9 103 Gy (Fig. 8.3). Further irradiation leads to a decrease of the centre
concentration (Pltze and Wolf 1996). This behaviour was described as radiogenic
annealing (Moiseev 1985). Another explanation could by the diffusion of charge
compensating M+ ions. In this case the intensity of the uncompensated Ti-centre
should increase with decreasing intensity of the compensated centres. However, this
could not be observed. Only after a strong c-irradiation ([19106 Gy) some varieties
of electron hole centres were detected in EPR at 295 K (Figs. 8.4, 8.5). This irradiation behaviour corresponds with the time instable behaviour of the 620 nm
cathodoluminescence emission (Gtze et al. 2001).
Thermoluminescence spectra. The TL-spectra for the different vein quartz
samples are, like in EPR, very similar. Because of the spectrometer sensitivity no
thermoluminescence is observable until a c-dose of 700 Gy. Three main peaks are
detected in the spectra after an irradiation with 19104 Gy: 150200 C/
330340 nm, 200 C/510 nm and 280 C/470510 nm (Fig. 8.6). The quartz from
the wall rock (Q101) shows in the EPR no [TiO4-/Li+]0 centres and in TL only
184
M. Pltze et al.
185
Fig. 8.4 EPR powder spectrum of vein quartz after 107 Gy c-irradiation with simulated spectra
of varieties of Si vacancy related electron hole centres (centre D: Maschmeyer and Lehmann
1983, centre O3-: adapted Nilges et al. 2009)
Fig. 8.5 Radiation dependence of Si vacancy related electron hole centres (sum)
186
M. Pltze et al.
60
Co 1 9 104 Gy)
60
Co 1 9 104 Gy)
Ti-centres could not be detected. Furthermore, the total Ge content and the concentration of paramagnetic [GeO4-/M+]0 centres in our samples is very low.
Therefore, the [TiO4-/Li+]0 centres can be suggested as possible electron traps for
the TL-peaks at 150200 C/330340 nm, 200 C/510 nm and 280 C/470510
nm. This suggestion for the latter peak is in good agreement with results from
Marfunin (1979) who related this peak also to Ti- and Ge-centres. For the TL-peak
at 150200 C/330340 nm oxygen-vacancy centres are working as recombination
sites (Serebrennikov et al. 1982, Rink et al. 1993) and for the TL-peaks at 200 C/
187
60
Co 1 9 106 Gy)
510 nm and 280 C/470510 nm the [AlO4]0 centres (Jani et al. 1983, McKeever
1984, 1991, Rink et al. 1993).
8.4 Conclusion
The most intense detected paramagnetic centres are [AlO4]0, [TiO4/Li+]0, and after
heavy irradiation different electron hole varieties. The EPR intensity of different
centres is related to the impurity content and reflects different formation conditions.
The investigation of the relationship between the EPR intensity of paramagnetic
centres and the c-irradiation dose shows an increase of the EPR intensity, but reveal
different saturation behaviour. The [AlO4]0 centre reaches saturation at a dose of
about 19106 Gy. The [TiO4-/Li+]0 spectrum, however, shows the highest intensity
already after irradiation with 59103 Gy and a decrease after higher irradiation dose.
The saturation intensity in EPR is related to the chemical impurity concentration.
Because of the same c-dose response in TL and the very low [GeO4-/M+]0 centre
concentration, the [TiO4-/Li+]0 centres are suggested as possible electron traps for
the TL-peaks at 150200 C/330340 nm, 200 C/510 nm and 280 C/470510 nm.
Acknowledgments The presented studies were carried out with the support of the Deutsche
Forschungs gemeinschaft DFG (grant Wo 489/1). We acknowledge S.S. Hafner (University
Marburg) for permission to carry out the EPR measurements. We thank S.M. Sukharjevski
(University St. Petersburg) for stimulating discussions and Y. Pan (University of Saskatchewan)
for helpful suggestions.
188
M. Pltze et al.
References
Arnold GW (1976) Thermoluminescence of ion-implanted SiO2. In: Chernaw F (ed) Ion
implantation in semiconductors. Plenum Press, New York
Batrak EN (1958) A model for the colour and emission centres in quartz. Kristallografiya
3:626627 (in Russian)
Bershov LV (1970) Isomorphism of Ti in natural minerals. Izv AN SSSR Ser Geol n12:4754
(in Russian)
Blankenburg HJ, Gtze J, Schulz H (1994) Quarzrohstoffe. Dt. Verl. f. Grundstoffind, LeipzigStuttgart
Botis SM, Nokhrin SM, Pan Y, Xu Y, Bonli T (2005) Natural radiation-induced damage in
quartz. I. Correlations between cathodoluminescence colors and paramagnetic defects. Can
Mineral 43:15651580
Botis SM, Pan Y, Nokhrin S, Nilges MJ (2008) Natural radiation-induced damage in quartz III. A
new ozonide radical in drusy quartz from the Athabasca basin, Saskatchewan. Can Mineral
46:125138
Cordier P, Weil JA, Howarth DF, Doukhan JC (1994) Influence of the (4H)Si defect on
dislocation motion in crystalline quartz. Eur J Mineral 6:1722
Furetta C (2010) Handbook of thermoluminescence. World Scientific Publishing, Singapore
Gtze J, Pltze M, Trautmann T (2005) Structure and luminescence characteristics of quartz from
pegmatites. Am Mineral 90:1321
Gtze J, Pltze M, Graupner T, Hallbauer DK, Bray CJ (2004) Trace element distribution in
pegmatite quartz: a combined study by ICP-MS, Electron Spin Resonance (ESR), Capillary
Ion Analysis (CIA) and Gas Chromatography (GC). Geochim Cosmochim Ac 68:37413759
Gtze J, Pltze M, Habermann D (2001) Origin, spectral characteristics and practical applications
of the cathodoluminescence (CL) of quartza review. Miner Petrol 71:225250
Gtze J, Pltze M (1997) Investigation of trace-element distribution in detrital quartz by electron
paramagnetic resonance (EPR). Eur J Mineral 9:529537
Griffiths JHE, Owen J, Ward IM (1954) Paramagnetic resonance in neutron-irradiated diamond
and smoky quartz. Nature 173:439442
Hashimoto T, Yokosaka K, Habuki H (1987) Emission properties of TL from natural quartz. Nucl
Tracks Radiat Meas 13:5766
Ichikawa Y (1967) Thermoluminescence of natural quartz irradiated by X-rays. Jap J Appl Phys
7:220226
Ikeya M (1993) New applications of electron spin resonance: dating, dosimetry and microscopy.
World Scientific Publishing, Singapore
Jani MG, Halliburton LE, Kohnke EE (1983) Point defects in crystalline SiO2: Thermally
stimulated luminescence above room temperature. J Appl Phys 54:63216328
Kuhn R, Trautmann T, Singhvi AK, Krbetschek MR, Wagner GA, Stolz W (2000) A study of
thermoluminescence emission spectra and optical stimulation spectra of quartz from different
provenances. Radiat Meas 32:653657
Lyakhovich VV, Gurbanov AG (1992) Geochemistry and conditions of formation of Eljurtinsky
massif (North Caucasus). Geokhimiya 6:800-812 (in Russian)
Lysakov VS, Serebrennikov AI, Solntsev VP (1969) Sources and spectra of the thermoluminescence for natural quartz crystals. Zh Prikl Spektroskopii 11:757760 (in Russian)
Marfunin AS (1979) Spectroscopy luminescence and radiation centers in minerals. Springer,
New York
Martini M, Sibilia E, Spinolo G, Vedda A (1985) Pre-dose, TSL and a.c. conductance
interrelation in quartz. Nucl Tracks Radiat Meas 10:497502
Martini M, Paleari A, Spinolo G, Vedda A (1995) Role of [AlO4]0 centers in the 380 nm
thermoluminescence of quartz. Phys Rev B 52:138142
Maschmeyer D, Lehmann G (1983) New hole centers in natural quartz. Phys Chem Miner
10:8488
189
Mashkovtsev RI, Shcherbakova MYA, Solntsev VP (1978) EPR of radiation hole centres in
a-quartz. In: Sobolev VS (ed) Rentgenografiya i spectroscopiya mineralov. Nauka, Novosibirsk
pp 7886
McKeever SWS (1984) Thermoluminescence in quartz and silica. Radiat Prot Dosim 8:8198
Mashkovtsev RI, Howarth DF, Weil JA (2007) Biradical states of oxygen-vacancy defects in
alpha-quartz. Phys Rev B 76:214114
McKeever SWS (1991) Mechanisms of thermoluminescence production: some problems and a
few answers? Nucl Tracks Radiat Meas 18:512
Medlin WL (1963) Thermoluminescence in quartz. J Chem Phys 38:11321143
Mejdahl V (1986) Thermoluminescence dating of sediments. Radiat Prot Dosim 17:219227
Moiseev BM (1985) Natural radiation processes in minerals. Nedra, Moskau (in Russian)
Mombourquette MJ, Tennant WC, Weil JA (1986) EPR study of Fe3+ in alpha-quartz: a
reexamination of the so-called I-center. J Chem Phys 85:6879
Mombourquette MJ, Minge J, Hantehzadeh MR, Weil JA, Halliburton LE (1989) Electronparamagnetic resonance study of Fe3+ in alpha-quartz: hydrogen-compensated center. Phys
Rev B 39:40044008
Nettar D, Villafranca JJ (1985) A program for electron-paramagnetic-res powder spectrum
simulation. J Magn Res 64:6165
Nilges MJ, Pan Y, Mashkovtsev RI (2008) Radiation-damage-induced defects in quartz. I. Singlecrystal W-band EPR study of hole centers in an electron-irradiated quartz. Phys Chem Miner
35:103115
Nilges MJ, Pan Y, Mashkovtsev RI (2009) Radiation-induced defects in quartz. III. Single crystal
EPR, ENDOR and ESEEM study of a peroxy radical. Phys Chem Miner 36:6173
Nuttall RHD, Weil JA (1981) The magnetic properties of the oxygen-hole aluminium centers in
cristalline SiO2. I. [AlO4]0. Can J Phys 59:16961707
Okada M, Rinneberg H, Weil JA, Wright PM (1971) EPR of Ti3+ centers in alpha-quartz. Chem
Phys Lett 11:275276
Orlenov PO (1984) Stable paramagnetic centres in natural quartz: method of concentration
measurement in powder. Mineral Zh 6:1724 (in Russian)
Pan Y, Nilges MJ, Mashkovtsev RI (2008) Radiation-induced defects in quartz. II. Single-crystal
W-band EPR study of a natural citrine quartz. Phys Chem Miner 35:387397
Pan Y, Nilges MJ, Mashkovtsev RI (2009) Radiation-induced defects in quartz: a multifrequency
EPR study and DFT modelling of new peroxy radicals. Mineral Mag 73:519535
Pan Y, Hu B (2009) Radiation-induced defects in quartz. IV. Thermal properties and
implications. Phys Chem Miner 36:421430
Pltze M, Wolf D (1996) EPR and TL-spectra of quartz: radiation dependency of the
[TiO4-/Li+]0-centre. Eur J Mineral 8(suppl.1):227 (in German)
Rieser U, Krbetschek MR, Stolz W (1994) CCD-camera based high sensitivity TL/OSLspectrometer. Radiat Meas 23:523528
Rink WJ, Rendell H, Marseglia EA, Luff BJ, Townsend PD (1993) Thermoluminescence spectra
of igneous quartz and hydrothermal vein quartz. Phys Chem Miner 20:353361
Rinneberg H, Weil JA (1972) EPR studies of Ti3+-H+ centers in X-irradiated alpha-quartz.
J Chem Phys 56:20192028
Serebrennikov AI, Valter AA, Mashkovtsev RI, Shcherbakova MYA (1982) The investigation of
defects in shock-metamorphosed quartz. Phys Chem Miner 8:153157
Stegger P, Lehmann G (1989a) The structures of three centers of trivalent iron in alpha-quartz.
Phys Chem Miner 16:401407
Stegger P, Lehmann G (1989b) Dynamic effects in a new substitutional center of trivalent iron in
quartz. Phys Stat Sol B151:5559
Usami T, Toyoda S, Bahadur H, Srivastava Ak, Nishido H (2009) Characterization of the E1 center
in quartz: role of aluminium hole centers and oxygen vacancies. Phys B 404:38193823
Vyatkin SV, Koshchug DG, Makhotin SS (2007) Various recombination kinetics of Al centers in quartz
from the Elbrus volcano and the Eldzhurtinsky granite rocks. Appl Magn Reson 32:333344
190
M. Pltze et al.
Weeks RA (1956) Paramagnetic resonance of lattice defects in irradiated quartz. J Appl Phys
27:13761381
Weil JA (1984) A review of electron spin spectroscopy and its application to the study of
paramagnetic defects in crystalline quartz. Phys Chem Miner 10:149165
Weil JA (1993) A review of the EPR spectroscopy of the point defects in alpha-quartz: the decade
19821992. In: Helms CR, Deal BE (eds) The physics and chemistry of SiO2 and the Si-SiO2
interface 2. Plenum Press, New York, pp 131144
Wright PM, Weil JA, Buch T, Anderson JH (1963) Titanium colour centres in rose quartz. Nature
197:246248
Chapter 9
Abstract Although quartz is one of the most abundant minerals in many rock
types, it has not been the focus of in situ quantitative chemical analysis by electron
microprobe for a long time. This was simply due to its high purity. Since
cathodoluminescence observations reveal a great variety of complex structures
within quartz, in situ chemical analysis methods like laser ablation inductively
coupled plasma mass spectrometry (LA-ICPMS), secondary ion mass spectrometry (SIMS), and electron microprobe (EMP) applied to quartz have received
increasing interest from geoscientists. Although the concentrations of many trace
elements in quartz are far below the detection limits of an electron microprobe,
Al, K, Ti, and Fe are, amongst others, suitable candidates for quantification. The
advantage of EMP analysis over other methods is its high spatial resolution
combined with high accuracy. Monte Carlo simulations of the elements listed
above in a quartz matrix indicate sampling depths of \2.7 lm for 99% of the
acquired X-ray photons. Sampling volumes range from 75 to 250 lm3, and depend
on excitation energy and defocusing of the electron beam. Unfortunately, beaminduced damage of the quartz lattice limits the use of high beam currents and
focused beams. Irradiation induced damage strongly influences the low energy
X-ray lines like Al-Ka. The beam sensitivity of the various quartz samples needs
to be frequently tested and the analysis protocol must be adapted according to this
signal behaviour. To minimise the effect, we propose dividing the measurement of
Al into several subsets. Furthermore, exact investigation of the curvature of the
background signal is required to avoid systematic errors. Secondary fluorescence
191
192
A. Kronz et al.
9.1 Introduction
Quartz occurs in many rock types. It is very stable during weathering and thus
dominates the mineralogical composition of most sedimentary rocks. The genesis of
quartz covers a large temperature range: from magmatic conditions to moderate-T
hydrothermal conditions to low-T conditions during authigenetic formation in sedimentary rocks. Hence it is an ideal candidate to record the petrological conditions of
its formation as well as indicate its source for provenance studies. Unfortunately,
quartz is a very pure mineral. This is why it has not been the focus micron-scale
analysis by electron microprobe (EMP) in the early microprobing decades.
The electron microprobe is not suitable for analysis of trace concentrations down to
a sub-ppm level (given here as mass ppm, lg g-1), as possible when using secondary
ion mass spectrometry (SIMS) or laser-ablation inductively coupled plasma mass
spectrometry (LA-ICPMS). Besides, light elements with an atomic number below that
of Na are difficult to analyse by EMP with correspondingly poor detection limits. Only
a few traces in quartz exceed a concentration of 1 lg g-1 in quartz (Gtze 2009, and
references therein). This limits the number of measureable elements to Al, K, Ti and
Fe. Na, which is frequently present in quantities above the detection limits of an
electron microprobe, is difficult to analyse for reasons that will be explained below.
Other elements only exceptionally reach a concentration which allows them to be
analysed by microprobe. Those that are listed above the 10 lg g-1 level in the literature either occur only in rare cases in higher concentrations in quartz (P, Ge) or appear
to be mostly concentrated in micro-inclusions (Na, Cl, K, Ca, Rb, Sr, Ba, Gtze 2009
and references therein). High purity quartz is of increasing importance in industry. The
high purity requirements for quartzused for fibre optics, optical glass, or as raw
material for solar panelsprecludes the mining of many quartz deposits. From this
point of view, it is important to know how certain trace elements are distributed in the
raw material (e.g. Mller et al. 2005). Quartz in which traces are concentrated in fluid
193
or mineral inclusions can be more easily purified than quartz where those same trace
elements are bound in the crystal lattice. Thus, analytical methods of high spatial
resolution used in combination with bulk trace-analysis play a crucial role in deciphering the distribution of trace elements in quartz. One of the most prominent
applications for the use of quartz as a petrological indicator is the single-phase thermometer TitaniQ (developed by Wark and Watson 2006, refined by Thomas et al.
2010 and again revised by Huang and Audtat 2011). Beside other newly developed
single-phase thermometers like Zr in rutile (Zack et al. 2004; Watson et al. 2006;
Tomkins et al. 2007), Ti in zircon (Watson et al. 2006), and Zr in sphene (Hayden et al.
2008), it extends the capabilities to evaluate formation temperatures from quartz in
magmatic, metamorphic and hydrothermal systems.
The analysis of trace elements in quartz, eventually combined with cathodoluminescence techniques, offers not only new insights in metamorphic studies or the
reconstruction of the genesis of deposits (e.g. Takahashi et al. 2008; Mller et al.
2010a) but also promises to be a useful tool for provenance studies, especially for the
reconstruction of orogenic cycles (Owen 1991; Bernet and Bassett 2005).
One of the outstanding characteristics of electron microprobe analysis is its nearly
non-destructive behaviour (aside from the sample consumed because of demanding
requirements on sample preparation). In contrast to SIMS and LA-ICPMS, no
material is consumed during the measurement process itself. In theory, this would
allow analyses that are not limited by sample loss during prolonged analysis. But, in
reality, beam induced damage of certain phases place practical restrictions on this
concept. Other disadvantages of EMP analysis are generally low signal strength and
poor signal/noise ratio compared to SIMS or LA-ICPMS. This contribution
exclusively reports methodological details of the analysis of quartz by electron
microprobe and will not discuss results of different case studies. For those we refer to
the literature (Mller et al. 2002, 2003a, b, 2005, 2006, 2008a, b, 2010a, b; Takahashi
et al. 2008; van den Kerkhof 2004a, b). All analyses were conducted on a JEOL JXA
8900 RL instrument at the Department of Geochemistry, Geowissenschaftliches
Zentrum of the University Gttingen. The electron probe is equipped with 5 wavelength dispersive spectrometers (WDS), an energy dispersive (EDS)-system, a
panchromatic cathodoluminescence detector detecting the wavelength range from
200 to 900 nm, as well as secondary and backscattered electron detectors. One
spectrometer is a so-called H-type equipped with LIF and PET analysing crystals
(LIFH, PETH) of Johansson-type geometry (Johansson 1933) and a smaller
Rowland-circle (R = 100 mm). This leads to an approx. 57 times higher performance, but at the expense of a somewhat poorer spectral resolution compared to the
normal spectrometers equipped with LIF/PET (Rowland circle, R = 140 mm).
194
A. Kronz et al.
(a)
MC-simulation
Emitted photons [ (0)=1]
1.8
1.6
1.4
1.2
1.0
0.8
0.6
0.4
0.2
0.0
0.0
(b)
MC-simulation
Emitted photons [%]
Al, K, Ti, Fe
1000 g/g
each, in SiO
K Ti
20 kV
15 kV
Ti
Fe K
Al
Al
Fe
0.5
1.0
1.5
2.0
2.5
3.0
3.5
100
90
80
70
60
50
40
30
20
10
0
0.0
Al
Fe
Ti
Ti
Al, K, Ti, Fe
1000 g/g
each, in SiO2
Al
Fe
20 kV
15 kV
0.5
Depth [m]
1.0
1.5
2.0
2.5
3.0
3.5
Depth [m]
(c)
MC-simulation
Emitted photons [%]
100
90
80
70
60
50
40
30
20
10
0
20 kV
foc.
Ti
Al, K, Ti, Fe
1000 g/g
each, in SiO
Al
Fe
15 kV
foc.
15 kV
5m
20
20 kV
5m
40
60
80
100
120
140
160
Fig. 9.1 a Calculated /(qz)-curves for the trace elements Al, K, Ti and Fe assuming a
concentration of 1000 lg g-1 each in pure SiO2 (Drouin et al. 2007). For the calculations
accelerating voltages of 15 and 20 kV, a take-off angle of 40, and a density of 2.65 g/cm3 was
chosen. 2 9 105 electron trajectories were calculated. b Cumulative emission depth of Al, K, Ti
and Fe from quartz (1000 lg g-1 each). c Cumulative interaction volume in quartz simplified to a
nearly hemispheric excitation, calculated for maximally focused beam and a beam diameter of
5 lm respectively. (modified after Mller et al. 2003b)
depths and excited volumes for different elements in a certain phase. We have
simulated the distribution of the Ka-X-ray signals of Al, K, Ti and Fe, each with a
concentration of 1,000 lg g-1 in a pure quartz matrix at accelerating voltages of
15, 20, and 25 kV, respectively, for maximally focused and defocused electron
beam conditions (Fig. 9.1, Table 9.1, see also Mller et al. 2003b). For analysing
Fe-Ka, a minimum excitation energy of 15 kV is required. Although high accelerating voltages lead to strong signals, energies[20 keV are not recommended for
the analysis of trace elements in matrices, such as quartz, that contain predominantly light elements. Here the excited volume expands to several hundred lm3.
This counteracts the advantage using an electron microprobe: its good spatial
resolution. Furthermore, signal drift arising from beam-induced damage also
increases with beam energy. Using an accelerating voltage of 20 kV, 99% of the
Ka X-rays are emitted from a sample depth in the range of 2.5 (Fe-Ka) to 2.75
(Al-Ka) lm (Fig. 9.1a, b). Calculating the volume from which 99% of the X-ray
photons are released for different beam diameters, we obtain 3244 lm3 for a
maximum focused beam and 128151 lm3 for a beam set to a diameter of 5 lm
195
Table 9.1 Results of the Monte-Carlo simulation (Drouin et al. 2007) calculated for
1000 lg g-1 Al, K, Ti and Fe in pure SiO2-Matrix respectively (density of 2.65 g/cm3)
99% Cumulative emission
15 kV
20 kV
25 kV
from:
Max. focus 5 lm Max. focus 5 lm Max. focus 5 lm
Al-Ka
K-Ka
Ti-Ka
Fe-Ka
Depth (lm)
Volume (lm3)
Depth (lm)
Volume (lm3)
Depth (lm)
Volume (lm3)
Depth (lm)
Volume (lm3)
1.68
11
1.60
9.1
1.51
7.7
1.32
4.5
76
71
66
53
2.74
44
2.71
42
2.64
39
2.46
32
151
148
142
128
3.98
131
4.07
141
4.04
138
3.87
122
286
300
294
274
Calculations were performed for 15, 20 and 25 kV using either a maximum focused beam or a
beam diameter of 5 lm. The values show the depths [lm] and the sampling volumes (lm3 )
when 99% of the emitted X-rays are reached
Fig. 9.2 Schematic illustration of typical sampling volumes of EMP, SIMS and LA-ICPMS in
quartz (modified after Mller et al. 2003b)
(Fig. 9.1c). Comparing this to the typical sampling volumes of SIMS and
LA-ICPMS, the EMP samples a volume at least one order of magnitude smaller
than SIMS and three orders of magnitude smaller than LA-ICPMS (see Mller
et al. 2003b, Fig. 9.2), although this depends on the ablation conditions used for
LA-ICPMS. However, when the laser is focused down to less than 20 lm diameter, analytical sensitivity decreases accordingly. Figure 9.3 illustrates an electron
beam impact of approximately 7 lm in diameter, in comparison to the typical
ablation of SIMS and LA-ICPMS (Mller et al. 2003b). We recommend using a
multi-analytical approach, combining LA-ICPMS or SIMS with EMP, if possible.
Especially small-scale micro-inclusions lead to characteristic offsets in the binary
196
A. Kronz et al.
EMP
EMP
197
198
2000
A. Kronz et al.
(a)
Bg-
1800
1600
qtz, 14m
1400
SiO2-glass
1200
1.84
1.86
1.88
1.90
1.92
1.94
1.96
2500
1.98
2.00
2.02
2000
Bg+
1500
1000
2.50
2.55
2.60
2.65
2.70
2.75
800
2.85
2.90
2.95
3.00
Bg-
700
2.80
600
500
400
3.50
average
error 2-sigma
3.55
3.60
3.65
3.70
3.75
3.80
3.85
3.90
[]
Al-K 20 kV 120 nA, TAP
(b)
(c)
Intensity [cps/A]
7000
Intensity [cps/A]
4000
Si3N4
6000
Si, integral
5000
3500
Bg-
Bg-
3000
4000
3000
2000
2500
qtz 14 m
SiO2 glass 14m
1000
7.6
2000
average
error2-sigma
7.8
8.0
Al-K
SiO2 glass 14m
8.2
8.4
[]
8.6
8.8
9.0
1500
7.6
narrow
background
setting
average
error 2-sigma
7.8
8.0
Bg+
Difference: 54gg-1Al
8.2
8.4
qtz 14 m
8.6
8.8
9.0
[]
199
b Fig. 9.4 a Scans of the continuum around the K-Ka, Ti-Ka and Fe-Ka X-ray lines on quartz and
SiO2-glass respectively. b Long-duration scan in the vicinity of the Al-Ka emission line on
different Si-bearing phases (pure Si, Si3N40.15 mass% Al-, quartz and SiO2-glass). Note the
blurred signals in the left part of the spectra, which occur in all species. For explanation see text.
c Detail of Fig. 9.4b: curvature of the background signal around the Al-Ka emission line leads to
a systematic underestimation of concentration by 54 lg g-1 Al, if background positions are set to
0.23 (for JEOL spectrometer, R = 140 mm, TAP-crystal: 2.5 mm). Although a narrower
background setting of 0.092 (1 mm) does not eliminate the effect, the systematic error is
reduced to insignificance
Table 9.2 LA-ICPMS analysis of quartz reference materials carried out at the Norwegian
Geological Survey, Trondheim by A. Mller. Values are given in (lg g-1)
Sample
LOD
QzGA2
QzGA1
QzTU
Qz_MGB
QzGCh
el/mass
3-r
n=3
n=3
n=3
n=3
n=3
Li7
Be9
B11
Al27
P31
K39
Ti47
Mn55
Fe56
Ge74
Rb85
Sr88
Ba138
Pb208
U238
0.467
0.25
0.304
2.12
4.37
1.921
0.1
0.615
3.974
0.544
0.176
0.201
0.003
0.003
0.002
1.33
3.612
1.574
5.96
\4.37
<1.92
<0.1
2.577
<3.97
\0.54
\0.18
\0.20
0.02
0.02
\0.002
3.124
3.124
0.693
24.55
\4.37
4.79
<0.1
\0.62
<3.97
1.264
\0.18
\0.20
0.024
0.079
\0.002
1.832
\0.25
0.382
8.03
\4.37
3.61
<0.1
2.544
<3.97
\0.54
0.224
\0.20
0.022
0.016
\0.002
3.713
\0.25
2.758
24.31
\4.37
2.367
<0.1
1.088
<3.97
0.746
\0.18
\0.20
0.01
0.018
\0.002
2.64
2.846
0.897
11.31
\4.37
3.00
<0.1
\0.62
<3.97
\0.54
\0.18
0.198
0.008
0.014
\0.002
Kalceff et al. (2000). To estimate the amount of the beam induced damage for trace
analytics, a series of exposure experiments on different quartz crystals and
SiO2-glass were conducted using an accelerating voltage of 20 kV and a beam
current (Faraday) of 80 nA: The beam diameter was varied from maximum
focussed to 3.5, 7 and 14 lm, respectively. A carbon coating of approximately
20 nm was applied. SiO2-glass, synthetic quartz (QzTu, Collection of the
Fachbereich Geowissenschaften, University of Tbingen) that had been cut parallel and perpendicular to its c-axis, and two natural samples (QzGA2: Gamsberg,
Namibia and QzGCh: Gotthard, Switzerland) were irradiated. All of these samples
have relatively low trace element contents: Al: 612 lg g-1, K: \4 lg g-1 Ti:
200
A. Kronz et al.
Fig. 9.5 Surface morphology of SiO2-glass and various quartz samples after 12 min irradiation.
An accelerating voltage of 20 kV was used and the beam current was set to 80 nA. foc.:
maximum beam focusing. For the used beam conditions the beam diameter is approx. 0.4 lm,
according to a diagram provided by JEOL company for the JXA 8900RL EMP), quartz synth.:
hydrothermal synthetic quartz sample QzTu, = c-axis: beam parallel to c-axis. Quartz nat.
GCH: natural quartz from Gotthard/Switzerland, beam parallel to the c-axis Quartz nat. GA2:
Gamsberg/Namibia, beam parallel to the c-axis
\0.1 lg g-1 and Fe: \4 lg g-1. Compositions analysed by LA-ICPMS are listed
in Table 9.2. Figure 9.5 shows the spots resulting from the measurements,
visualized using secondary electron images. To avoid any disturbances resulting
from charging effects, which might occur when irradiated by high beam energies,
the samples were coated again after the experiment by a layer of 10 nm AuPd
(60/40), a procedure which additionally allows a superior secondary-electron image
quality. All of the quartz crystals, synthetic and natural, display comparable visual
effects after irradiation, independent from the crystallographic orientation, however
SiO2-glass behaves differently. All of the quartz crystals exhibit a volume expansion
201
202
A. Kronz et al.
(a)
Intensity [cps/A]
1600
Fe-K,LIFH
1500
1400
1300
Ti-K,PET
1200
1100
1000
900
800
700
K-K,PET
600
500
00:00
02:00
04:00
06:00
08:00
10:00
12:00
(b)
Intensity [cps/A]
216000
Si-K
quartz
212000
increased defocusing
initial
208000
SiO2 glass 5m
204000
average
error 2- sigma
200000
SiO2 glass foc
196000
00:00
02:00
04:00
06:00
08:00
10:00
12:00
203
b Fig. 9.6 Signal variations in SiO2-glass and various quartz samples during 12 min irradiation.
Accelerating voltage was set to 20 kV and a beam current of 80 nA was used. None of the tested
samples contain any trace elements above 10 lg g-1. a At the K-Ka (3.312 keV, k = 3.742 ,
PET-crystal), Ti-Ka (4.508 keV, k = 2.749 , PET) and Fe-Ka (6.398 keV, k = 1.937 ,
LIFH) spectrometer-positions no significant signal drift can be observed during irradiation in
quartz. (Note: the sample does not contain any amount of these elements in concentrations larger
than 5 lg g-1, hence the values represent pure background continuum.) b Si-Ka-X-ray line
(1.739 keV, k = 7.13 ), PET-crystal. During irradiation the Si-Ka-signal decreases in
SiO2-glass, whereas quartz shows the opposite behaviour. Using a focused beam, the signal
increases within the first 3 min. For a beam diameter of 14 lm, nearly constant signals could be
obtained both for SiO2-glass and quartz. c Al-Ka-X-ray line (1.486 keV, k = 8.34 ), TAPcrystal. The X-ray signal at the Al-Ka position in SiO2-glass decreases with time. The effect is
stronger when a focused beam is used. Quartz behaves in an opposite way: a strong increase
within the first two minutes of irradiation is observable. Use of an increasingly defocused beam
minimizes the effect. At 14 lm beam diameter no measurable shift occurs. The synthetic quartz
(Qz-Tu) was mounted both parallel and perpendicular to its c-axis. No orientation effects were
observable at any beam diameter. Also no differences between synthetic and natural quartz could
be detected. d The absorbed current decreases in all quartz samples with time, when irradiated.
Defocusing the beam minimizes the effect, but even at 14 lm beam-diameter no constant
conditions are obtained. SiO2-glass behaves stably for beam diameters larger than 3.5 lm. The
effect is coupled to the increasing amorphisation of the sample, which produces also an increased
backscatter electron signal under irradiation
204
A. Kronz et al.
(c)
Intensity[cps/A]
Al-K
2000
1900
qtzfoc.
average
error2-sigma
1800
1700
qtzsynth.foc.
=c
qtzsynth.foc.
+c
qtz3.5m
1600
SiOqtz7m
2 glass5m
1500
initial
1400
qtz14m
SiO 2 glass5m
1300
SiO 2 glassfoc.
1200
00:00
02:00
04:00
06:00
08:00
10:00
12:00
Irradiationtime[mm:ss]
(d)
7.2E-08
7.0E-08
qt z 14 m
6.8E-08
6.6E-08
qt z 7 m
6.4E-08
6.2E-08
qt z synt h. 3.5 m + c
qt z nat . 3.5 m
6.0E-08
qt z synt h. 3.5 m = c
5.8E-08
5.6E-08
qt z nat . foc.
5.4E-08
qt z synt h. foc. +c
5.2E-08
5.0E-08
qt z synt h. foc. =c
4.8E-08
00:00 01:00 02:00 03:00 04:00 05:00 06:00 07:00 08:00 09:00 10:00 11:00 12:00
Irradiation time [mm:ss]
205
Fig. 9.7 Backscattered electron (BSE) image in composition mode of the 12 min irradiated
(20 kV, 80 nA) spots of different size in quartz. Note the intensity increase with decrease in beam
diameter
T [C]
500
450
400
350
max.
focused
300
1 m
250
200
150
3.5 m
100
7 m
50
14 m
0
0
50
100
150
200
250
300
350
400
450
500
Fig. 9.8 Temperature increase in quartz calculated after the empirical formula given by Castaing
(1960) for different beam diameters at 20 kV accelerating voltage and 80 nA beam current. Heat
conductivity of quartz is perpendicular to the c-axis: 7.25 (W/m/K), parallel to the c-axis (not
shown): 13.2 (W/m/K) (Kleber 1985)
206
A. Kronz et al.
towards
spectrometer
Ti-K :
secondary
fluorescence
top view:
Ti-K :
trace in
quartz
electronbeam
towards
spectrometer
adjustable
sample
rotation
scan
direction
TiO 2
qtz
TiO 2
quartz
Fig. 9.9 Schematic illustration of a test sample for evaluating the effect of secondary
fluorescence (SF)
207
Table 9.3 Composition of Minerals used for secondary fluorescence tests near phase boundaries
Sanidine
Ilmenite
Magnetite
Mass%
Sd_X7Ef
Il-MO4
Mt-PAR
Eifel Germany
Monastery South Africa
Parinacota Chile
SiO2
TiO2
Al2O3
FeO
MnO
MgO
CaO
K2O
Na2O
SrO
Rb2O
BaO
64.3
NA
18.1
0.14
NA
NA
\0.03
14.33
1.5
0.11
0.016
0.89
\0.09
49.6
0.95
37.6
0.19
10.54
\0.03
\0.02
\0.03
NA
NA
NA
\0.08
3.78
0.85
87.5
1.01
0.5
\0.05
\0.04
\0.04
NA
NA
NA
Sd_X7Ef: Sanidine Volkesfeld, Eifel, Germany (courtesy of Gerhard Wrner, Geowissenschaftliches Zentrum, Universitt Gttingen)
Il-MO4: Ilmenite Monastery South Africa (courtesy of Thomas Zack, University Mainz,
Germany)
Mt-PAR: Magnetite, Parincota volcano, Chile, courtesy of John Hora, Geowissenschaftliches
Zentrum, Universitt Gttingen
NA not analysed
Elements used for fluroscence tests highlighted in bold
208
A. Kronz et al.
(a)
Conc. [mass-%]
0.015
Quartz-Sanidine
Al parallel to spectrom.-1TAP
Al
Al perpendicular to spectrom.-1TAP
2-sigma detection limit
0.010
0.005
0.000
0
20
40
60
80
Distance [m]
(b)
Quartz-Sanidine
Conc. [mass-%]
0.015
0.010
0.005
0.000
0
20
40
60
80
Distance [m]
Fig. 9.10 Results of the experiments that test the effect of secondary fluorescence. On the left
concentrations of two scans respectively, along the line of sight of the spectrometer and
perpendicular to it versus distance to the phase boundary are shown. Error bars are 2-sigma
values of the error by counting statistics. Detection limits are 2-sigma values of the background
noise. On the upper right the direction of the profiles for each mineral-couple are visualized as
BSE images. On the lower right an element map of the respective element +WD-spectrometer
combination is shown in beamscan-mode on the pure phase producing a certain SF (TiO2,
sanidine, ilmenite). Hence the focusing behaviour of each element-spectrometer-crystal
combination is obvious relative to the marked scan profiles. a Quartz-sanidine couple: Al-Ka
on TAP, spectrometer-1. b Quartz-sanidine couple: K-Ka on PET, spectrometer-3 c Quartz-TiO2
(snyth.) couple: Ti-Ka on PET, spectrometer-4 d Quartz-magnetite couple: Fe-Ka on LIFH,
spetrometer-5
Figure 9.10 ad summarizes the results of the SF tests. For all measured
Ka-X-ray lines, a significant disturbing influence of the respective adjacent majorelement-phase can be demonstrated. Although Al-Ka (1.486 keV) might be
excited by the characteristic Si-Ka-line (1.739 keV) the effect is not stronger than
the other higher energy X-ray lines. For Al and K excited in an adjacent sanidine,
the analysable influence of SF ends at a distance of approx. 40 lm (Al-Ka) and
60 lm (K-Ka), respectively (Fig. 9.10a, b) where the signal reaches the detection
209
(c)
Conc. [mass-%]
Quartz-TiO2
0.100
0.22 0.13
0.090
Ti
Ti parallel to spectrom.-4PET
Ti perpendicular to spectrom.-4PET
0.080
0.070
0.060
0.050
0.040
0.030
0.020
0.010
0.000
20
40
60
80
(d)
Conc. [mass-%]
Quartz-Ilmenite
0.050
0.045
Fe
0.040
0.035
0.030
0.025
0.020
0.015
0.010
0.005
0.000
20
40
60
80
210
A. Kronz et al.
Quartz-Magnetite
Conc. [mass-%]
0.250
Fe
0.200
0.150
0.100
0.050
0.000
0
20
40
60
80 100 120
Distance [m]
140
160
180
Fig. 9.11 Observation of SF in a natural quartz (Parinacota dacite, Chile) near a magnetite
inclusion: Fe-Ka on LIFH, spectrometer-5: along and perpendicular to its line of sight
211
single measurement
stepwise measurement
blank standard
0.025
0.020
0.015
0.010
0.005
0.000
-150 -100 -50
50
Fig. 9.12 Net intensity versus calculated concentration of Al2O3 in various quartz samples (see
also Mller et al. 2003a). For the uniform quartz matrix the ZAFfactor remains constant, leading
to an ideal linear relationship between net intensities and calculated concentrations. Negative netcounts are produced by subtraction of too large background values and were set by the software to
zero-concentration. For a first set of measurements negative net countrates of -130 cps/lA
equals to an underestimation of approx. 100 lg g-1 Al2O3, although the background curvature
was already corrected. This was due to the large signal shift during measurement: when measuring
in one single step (300 s on peak and thereafter 150 s. on each background) the continuous signal
increase subtracts a too large background value. Subdividing the Al-measurement into 5
distinctive steps of 60 s on peak and 30 s. on each background respectively minimizes the
systematic error (stepwise measurement). A nearly Al-free quartz (GA2, blank-reference;
6 lg g-1 Al) was used to confirm the zero-concentration
1 p
Nbg Nbg
2
where:
NLOD is the error in total counts
r is the sigma factor depending on the applied confidence value (usually 3)
Nbg- an Nbg+ are the total accumulated counts on each background signal.
After transforming the error of total counts into a count-rate (ILOD) this value
can simply be calculated into a concentration (CLOD) by applying the certain
matrix correction procedure. Because of the strong linear behaviour of net-counts
of a certain trace element in an overall uniform matrix (quartz), a simple linear
equation can be applied, after deducing the calibration factor from multiple
measurements (Fig. 9.12):
CLOD lg g1 factor ILOD
212
A. Kronz et al.
50
100
90
40
Al 2 O3 TAP
TiO2 PET
TiO2 PETHS
80
70
60
30
20
60
20
50
20
50
gg - 1
gg - 1
gg - 1
gg - 1
gg - 1
gg - 1
Al
Al
Ti
Ti
Ti
Ti
TAP
TAP
PET
PET
PETHS
PETHS
50
40
20
30
20
10
10
0
0
0
200
400
600
800
1000
200
400
600
800
Counttime [sec] on peak
1000
Fig. 9.13 a Calculated detection limits by counting statistics (2 sigma of the background noise)
for Al and Ti as a function of counting time. b Calculated relative errors by counting statistics (2sigma) for Al and Ti as a function of counting time at different concentration levels. For Ti the
values are calculated for a normal spectrometer equipped with a PET analyser crystal and for a
high sensitive spectrometer (PETH)
213
Table 9.4 Analysis protocol for trace elements (Al, K, Ti, Fe) in quartz: recommended beam
conditions are: Beam-current/Faraday): 80120 nA; Accelerating voltage: 20 kV; beam diameter: 714 lm (if appropriate)
Spectrometer
1
2
3
4
5
(a) Analysis protocol for analysing the elements (AL, K, Ti, Fe) in quartz
1. Sequence:
Element
Si
Si
K
Ti
Fe
X-ray line
Kb
Kb
Ka
Ka
Ka
Crystal
TAP
TAP
PET
PET
LIFH
Peaktime
60
60
360
360
360
Backgr.-time
30
30
180
180
180
Hematite
Standard
Quartz
Quartz
Sanidine
TiO2, synth
2. to 6. Sequence:
Element
Al
Al
X-ray line
Ka
Ka
Crystal
TAP
TAP
Peaktime
60
60
Backgr.-time
30
30
Al2O3, synth
Standard
Al2O3, synth
Al-tot: 5
K: 11
Ti: 18
Fe: 16
LOD (lg g-1)
(b) Analysis protocol for Ti(Al) in quartz, used for the TitaniQ thermometer (Wark and Watson
2006; Thomas et al. 2010)
1. Sequence:
Element
Ti
Si
Ti
Ti
Ti
X-ray line
Ka
Kb
Ka
Ka
Ka
Crystal
PET
TAP
PET
PET
PETH
Peaktime
360
60
360
360
360
Backgr.-time
180
30
180
180
180
Quartz
TiO2, synth
TiO2, synth
TiO2, synth
Standard
TiO2, synth
2. to 6. Sequence:
Element
Al
X-ray line
Ka
Crystal
TAP
Peaktime
60
Backgr.-time
30
Standard
Al2O3, synth
Al: 9
Ti-tot: 4
LOD (lg g-1)
dependant on counting time for the EMP in Gttingen. In Fig. 9.13b relative errors
are calculated by error propagation of peak and background measurements for
single measurements of Al and Ti at different concentration levels.
214
A. Kronz et al.
215
It is always useful to plot net-counts (normalized) versus the calculated concentrations for different correction procedures, either Uqz (Armstrong 1995) or
the common ZAF-correction procedures (Fig. 9.12). Because of the strong
linear behaviour due to the uniform matrix in quartz, linear offline calibrations
can be applied. If calibration factors are determined for an individual spectrometer, multiple counting of a certain element on different spectrometers can
be calculated individually: If systematic differences occur, this may point to
SF-effects. Systematic negative net raw-counts will point to signal drift effects
and could be eliminated by subdividing a measurement sequence into subsets
(Fig. 9.12).
Using large beam currents and high energies in a trace program can cause
problems when analysing the major elements like Si in quartz, due to detector
overflow or offsets in dead-time correction. For quartz we do not recommend to
adopt a calculated SiO2-value (fixed or as difference to 100%) because it
obscures the quality of the analysis. Thus, the major elements should be analysed
in a separate sequence using a smaller beam current, which is not possible in all
microprobe types, or alternatively, a weaker X-ray line needs to be chosen. For Si
we recommend using Si-Kb or the second order diffraction of Si-Ka.
Acknowledgments We are grateful to Herbert Ngele from Windhoek who provided the
QzGA1 and QzGA2 quartz crystals from Gamsberg, Namibia. The constructive review by
A. Renno helped to improve the manuscript.
References
Armstrong JT (1995) CITZAF: a package of correction programs for the quantitative electron
microbeam X-ray analysis of thick polished materials, thin films, and particles. Microbeam
Anal 4:177200
Bastin GF, Loo FJJ, Vosters PJC, Vrolijk JWGA (1983) A correction procedure for characteristic
fluorescence encountered in microprobe analysis near phase boundaries. Scanning 5:172183
Bastin GF, Loo FJJ, Vosters PJC, Vrolijk JWGA (1984) An iterative procedure for the correction
of secondary fluorescence effects in electron-probe microanalysis near phase boundaries.
Spectrochimica Acta 39B:15171522
Bernet M, Bassett K (2005) Provenance analysis by single-quartz-grain SEM-CL/optical
microscopy. J Sed Res 75(3):492500
Castaing R (1960) Electron-probe microanalysis. Adv Electr Electron Phys 13:317386
Dalton JA, Lane SJ (1996) Electron microprobe analysis of Ca in olivine close to grain
boundaries: the problem of secondary X-ray fluorescence. Am Mineral 81:194201
Donovan JJ, Lowers HA, Rusk BG (2011) Improved electron probe microanalysis of trace
elements in quartz. Am Mineral 96:274282
Drouin D, Couture AR, Joly D, Tastet X, Aimez V, Gauvin R (2007) CASINO V2.42a fast and
easy-to-use modelling tool for scanning electron microscopy and microanalysis users.
Scanning 29:92101
Escuder JA, Salvat F, Llovet X, Donovan JJ (2010) Numerical correction for secondary
fluorescence across phase boundaries in EPMA. IOP Conf. Series: Mater Sci Eng 7:012008.
doi:10.1088/1757-899X/7/1/012008
216
A. Kronz et al.
Fialin M, Rmy H, Richard C, Wagner R (1999) Trace element analysis with the electron
microprobe: new data and perspectives. Am Mineral 84:7077
Fournelle JH, Kim S, Perepezko JH (2005) Monte Carlo simulation of Nb Ka secondary
fluorescence in EPMA: comparison of PENELOPE simulations with experimental results.
Surf Interface Anal 37:10121016
Fournelle JH (2007) The problem of secondary fluorescence in EPMA in the application of the
Ti-in-Zircon geothermometer and the utility of PENEPMA Monte Carlo program. Microsc
Microanal 13(supplement 2):13901391
Gtze J (2009) Chemistry, textures and physical properties of quartzgeological interpretation
and technical application. Mineral Mag 73:645671
Hayden LA, Watson EB, Wark DA (2008) A thermobarometer for sphene (titanite). Contrib
Mineral Petrol 155:529540
Huang R, Audtat A (2011) A critical look at the titanium-in-quartz (TitaniQ) thermobarometer.
Mineral Mag 75:1065
Jercinovic MJ, Williams ML, Lane ED (2008) In situ trace element analysis of monazite and
other fine-grained accessory minerals by EPMA. Chem Geol 254(34):197215
Johansson T (1933) ber ein neuartiges, genau fokussierendes Rntgenspektrometer. Z Physik
82:507529
Kleber W (1985) Einfhrung in die Kristallographie. VEB Verlag Technik, Berlin
Llovet X, Galan G (2003) Correction of secondary X-ray fluorescence near grain boundaries in
electron microprobe analysis: application to thermobarometry of spinel Lherzolithes. Am
Mineral 88:121130
Luvizotto GL, Zack T, Meyer HP, Ludwig T, Triebold S, Kronz A, Mnker C, Stockli D,
Prowatke S, Klemme S, Jacob DE, von Eynatten H (2008) Rutile crystals as potential trace
element and isotope mineral standards for microanalysis. Chem Geol 261:346369
Mller A, Kronz A, Breiter K (2002) Trace elements and growth patterns in quartz a fingerprint
of the evolution of the subvolcanic Podles Granite System (Krusn hory Mts., Czech
Republic). Bull Czech Geol Surv 77(2):135145
Mller A, Ren M, Behr H-J, Kronz A (2003a) Trace elements and cathodoluminescence of
igneous quartz in topas granites from the Hub Stock (Slavkovsky Les Mts., Czech Republic).
Mineral Petrol 79:167191
Mller A, Wiedenbeck M, van den Kerkhoff AM, Kronz A, Simon K (2003b) Trace elements in
quartza combined electron microprobe, secondary ion mass spectrometry, laser-ablation
ICP-MS, and cathodoluminescence study. Eur J Mineral 15:747763
Mller A, Ihlen PM, Kronz A (2005) Potential recources of quartz and feldspar raw material in
Srland IV: relationship between quartz, feldspar and mica chemistry and pegmatite type.
Geol Surv Norway. NGU report 2005.075, p 94
Mller A, Seltmann R, Halls C, Siebel W, Dulski P, Jeffries T, Spratt J, Kronz A (2006) The
magmatic evolution of the lands end pluton, Cornwall, and associated pre-enrichment of
metals. Ore Geol Rev 28:329367
Mller A, Seltmann R, Kober B, Eklund O, Jeffries T, Kronz A (2008a) Compositional zoning of
Rapakivi feldspars and coexisting quartz phenocrysts. Can Mineral 46:14171442
Mller A, Ihlen PM, Kronz A (2008b) Quartz chemistry in polygeneration Sveconorwegian
pegmatites, Froland, Norway. Norway Eur J Mineral 20:447463
Mller A, Herringon R, Armstrong R, Seltmann R, Kirwin DJ, Stenina NG, Kronz A (2010a)
Trace elements and cathodoluminescence of quartz in stockwork veins of Mongolian
porphyry-style deposits. Mineralium Deposita 45:707727
Mller A, Kerkhof AM, Behr H-J, Kronz A, Koch-Mller M (2010b) The evolution of lateHercynian granites and rhyolites documented by quartza review. Earth and environmental
transactions of the Royal Rociety of Edinburgh. Earth Environ Sci 100:185200
Nasdala L, Kronz A, Hanchar JM, Tichomirowa M, Davis DW, Hofmeister W (2006) Effects of
radiation damage on back-scattered electron imaging of natural zircon. Am Mineral
91:17391746
217
Chapter 10
Abstract In situ micro analysis of ultra trace element composition of quartz using
laser ablation inductively coupled plasma mass spectrometry (LA-ICP-MS) allows
rapid screening of lattice bound impurities of potential high-purity quartz
resources and samples for petrological research without the need to remove solid
and liquid inclusions by expensive dressing techniques prior to chemical analysis.
Information of the analysed trace element content can not only be used to determine the economic quality of quartz but also the conditions of quartz crystallisation and the origin of the quartz-forming fluids and melts. The main purpose of
this paper is to describe an efficient and precise analytical method for estimating
the concentrations of lattice-bound trace elements in quartz. The best choice of
instrument is considered to be a double focusing sector field inductively coupled
plasma mass spectrometry ICP-MS that provides high sensitivity and a mass
resolution high enough to separate K from its interferences. The ICP-MS should be
coupled to a 193 nm excimer laser, a femto second laser or a similar ablation
system. The following elements are included in the standard analytical protocol
applied at the Geological Survey of Norway (NGU): Al, B, Be, Ca, Cr, Fe, Ga, Ge,
K, Li, Mg, Mn, Na, P, Rb, Sb, Sr, Ti and Zn. Any element with isotopes that can be
ionised in an Ar plasma can easily be included if suitable reference materials are
available. External calibration was done using the international reference materials
NIST SRM 610, 612, 614, 616 and 1830 from the National Institute of Standards
and Technology (NIST), BCS 313/1 from the Bureau of Analysed Samples (BAS)
and the certified reference material pure substance No. 1 silicon dioxide SiO2
from the Federal Institute for Material Research and Testing, Berlin, Germany
(BAM). To improve the lower limit of quantification and analytical uncertainty at
219
220
10.1 Introduction
Although quartz is a common rock-forming mineral in silica-saturated lithologies,
the speciation and concentration of lattice-bound trace elements in quartz is
potentially informative. However, trace element studies of quartz are limited
compared to other rock-forming minerals (e.g., Perny et al. 1992; Larsen et al.
2000a, b; Gtze et al. 2001; Mller and Welch 2009 and references therein).
Because high-purity quartzi.e. quartz with less than 50 lg g-1 lattice-bound
trace elements (e.g., Harben 2002)is a valuable commodity used in the manufacture of various high technology products, there is an industrial need for its
chemical characterisation on the lg g-1 down to ng g-1 level.
Conventional chemical analysis of quartz requires extensive laboratory preparations. e.g., flotation, handpicking and sample digestion. Even then, contamination during preparations and incomplete removal of microscopic inclusions can
lead to spurious results. Laser ablation combined with high resolution inductively
coupled plasma mass spectrometry (LA-HR-ICP-MS) provides an approach to the
analysis of quartz at the lg g-1 to ng g-1 level with minimal sample preparation
and subsequent contamination. State-of-the art laser ablation equipment allows
high resolution optical control of the analysed quartz domain to avoid the analyses
of contaminating micro fluid- and mineral inclusions. Alternative micro-analytical
methods are, for example, secondary ion mass spectrometry (SIMS) and electron
probe micro analysis (EMPA). Advantages and disadvantages of these methods
compared to quartz analysis by LA-ICP-MS are described in Mller et al. (2003).
Quartz has an exceptionally strong atomic configuration of SiO bonds which
allow only a minimum of other elements into its structure. However, minute
amounts of substitutional and interstitial impurities can be incorporated into the
atomic lattice and these elements are classified as lattice-bound impurities. Substitutional impurities replace Si4+ in the SiO tetrahedra which make up the quartz
lattice. Interstitial impurities, mostly small monovalent ions that fit into structural
10
221
10.2 Instrumentation
10.2.1 Inductively Coupled Mass Spectrometry
There are several manufacturers of ICP-MS instruments, which all display particular advantages and disadvantages. ICP-MS instruments can be divided in two
main groups; high mass resolution instruments (HR-ICP-MS) and quadrupole
instruments (Q-ICP-MS). HR-ICP-MS can provide eligible mass resolutions above
300 m/Dm, up to around 9,000 Dm/m, while Q-ICP-MS has a mass resolution
below or around 300 m/Dm. Resolution in mass spectrometry is normally given
222
Fig. 10.1 Analysis of P in NIST SRM 610 (National Institute of Standards and Technology) at
medium resolution 3500 m/Dm with adjacent interferences, 12C18O1H, 14N16O1H (shown in
figure), 28Si1H3, 16O15N (shown in figure), 29Si1H2 and 1H30Si (shown in figure) which need a
mass resolution of 933, 967, 1162, 1457, 1840 and 3951 respectively to be completely resolve
from the target isotope
according to the 10% valley definition, where IUPAC defines Dm as the mass
difference between two resolved peaks, with a 10% valley.
Performing multi-isotope laser ablation on limited sample material, speed is
essential and Q-ICP-MS are regarded as the fastest instruments since they can scan
the whole mass range (2260 Dalton) electrically. The Q-ICP-MS mass analyzer
consists of four cylindrical rods onto which are applied both RF and DC electrical
fields. The ions are separated by velocity determined by their energy and mass. If a
different U/V ratio is applied different m/z is allowed to pass through the quadruple
rods to the detector. HR-ICP-MS instruments are slower since the magnetic field
strength has to be varied when isotopes within a large mass range is measured.
However, Q-ICP-MS has a mass resolution which does not allow interference
free measurements of e.g. P, Cr, Fe. A mathematical interference correction has to
be applied to the measured intensity of an isotope based upon the measurement of
another isotope of the interfering element/species. Using the known natural isotopic abundances of the interfering element/species the contribution to the isotope
of interest is calculated, correcting for the isobaric overlap. In most cases this will
be successful. However, in some cases, such as P in quartz, where interferences at
the mass resolution 300 m/Dm occur with 14N16O1H, 12C18O1H, 28Si1H and 1H30Si
(the required mass resolutions are 967, 933, 1162 and 3951 m/Dm, respectively, to
resolve completely the target isotope from the interferences; Fig. 10.1) many
corrections (with some being sample-dependent) have to be performed making it
very difficult to achieve good reproducible analysis. In cases like P the application
HR-ICP-MS has a major advantage to perform analyses at high mass resolution to
10
223
224
laser with a femtosecond laser and concluded that by using a femtosecond laser the
calibration could be extended due to reduced matrix dependence and because it
grants the ability to use non-matrix matched standards. In Ferndez et al. (2007) an
overview is given to show the state of art in fs-LA-ICP-MS.
Improved detection limits and ablation performance may also be achieved by
performing the laser ablation under a He atmosphere, using a HeAr mixture as the
aerosol carrier (Gnther and Heinrich 1999). Different gases can be added to the
coolant gas or to the sample gas to improve sensitivity and the stability of
the plasma e.g. a small quantity of N2 into the sample gas has the effect of
enhancing the sensitivity of the heavy elements (after Hirata and Nesbitt 1995).
At NGU we have experienced that the addition of small amounts methane
(CH4) to the sample gas during liquid analysis increased the sensitivity and the
plasma stability. The same observation has been made by Rodushkin et al. (2005).
Recently, we have also experimented in adding methane to the sample gas during
laser ablation analysis and experienced a significant improvement in the analysis
of some elements, e.g. P (Fig. 10.3). Comparing the two spectra in Fig. 10.3, one
with and one without methane in the sample gas, one can see that the sensitivity is
doubled by adding methane. However, adding nitrogen, helium, methane or other
10
225
Fig. 10.3 In the diagram on the left the ablation was performed with He as transport gas in the
sample chamber and Ar as makeup sample gas. In the diagram on right a small amount of
methane CH4 was added to the Ar sample gas resulting in higher signal intensity. Please note the
difference in scaling between the two diagrams
gases to the sample gas may also add new interferences that have to be considered
e.g. adding nitrogen to the sample gas may cause interference from 14N++ on 7Li+.
226
Table 10.1 Operating parameters of the ICP-MS and key method parameters
Laser parameters
Wavelength
193 nm
Pulse width
\5 ns
Irradiance (power density)
1.22 GW/cm2
Fluence (energy density)
6.11 J/cm2
Spot size
30100 lm
Laser repetition rate/pulse rate
1020 Hz
Sample helium flow rate
0.61.1 l/min
ICP-MS parameters
Plasma power
9741100 W
Auxiliary gas flow
1.071.2 l/min
0.91.1 l/min
Sample gas flow
0.030.06 l/min
Methane (0.5% CH4 in Ar)
Cone
High performance Ni
CD-2 guard electrode
Yes
Data collection
Scan type
E-scan
No. of scans
1520
Mass range
20% of isotope peak
Low mass resolution (LR)
80100% of isotope peak
Medium mass resolution (MR)
60
Samples per peak
30
Low mass resolution (LR)
0.024 s
Medium mass resolution (MR)
0.048 s
Segment duration (analysis time per isotope per scan)
Low mass resolution (LR)
Medium mass resolution (MR)
point and the aerosole is transported from this cup through a tube to the ICPMS
plasma. Commonly, raster ablation with a laser diameter of 50 lm is performed.
The resulting raster is about 100 9 350 lm in size, applying the measuring
parameters listed in Table 10.1. In the case of optimal ablation the depth of the
raster is 1020 lm (Fig. 10.2a). In the case of sample outbreak the crater can be up
to 100 lm in depth (Fig. 10.2b). Detailed petrographic examination of the 300-lm
thick, polished sections prior to analysis and high-resolution optical control during
laser ablation allowed the selection of ablation areas free of mineral and fluid
micro inclusions ([0.1 lm). In addition the samples are examined by scanning
electron microscope cathodoluminescence (SEM-CL) in order to check the
chemical homogeneity of the ablation area. Natural quartz may show small-scale
(\500 lm wide) growth zoning, alteration structures or healed fractures with
variable trace element content and, therefore, are visualised by SEM-CL. If these
structures lie in the sampling area and are smaller than the ablation raster, the trace
element concentrations of the structures will be averaged.
The ICP-MS instrumental operating parameters were tuned on 139La by ablation of the reference material NIST SRM 612 (National Institute of Standards and
Technology) to give maximum sensitivity and stability. The oxide formation levels
10
227
were monitored by the ratio of thorium oxide and thorium, as this is one of the
most easily formed oxides. The measurement conditions were therefore optimised
for the oxide ion ratio formation less than 0.7% (ThO+/Th+ \ 0.7%).
The following elements are included in the standard analytical protocol at NGU:
Al, B, Be, Ca, Cr, Fe, Ga, Ge, K, Li, Mg, Mn, Na, P, Rb, Sb, Sr, Ti and Zn. Other
elements like e.g. Th and U can easily be included the only limitation of which
elements that can be included is the availability of suitable reference materials and
the ionization energy of the element should be below 12 eV. The existence of
spectroscopic interferences required the use of variable mass resolutions. Li, Be, B,
Mn, Ge, Rb, Sr and Sb were analysed at low mass resolution (m/Dm = 300), but
Na, Mg, Al, P, Ti, Ca, Cr, Ti, Fe, Zn and Ga required medium mass resolution (m/
Dm & 3500) for interference free measurements. For the analysis of potassium a
resolution of 5689 (m/Dm) is necessary to avoid the interference from 38Ar1H+
which means that K usually should be measured in high mass resolution mode.
However, by careful tuning of the instrument and selection of the integration area of
the K peak, it is also possible to achieve appropriate measurements also in medium
mass resolution mode (Fig. 10.4). This was done to save time and make it possible
to perform the analysis on smaller ablation areas. Typical analysis time per sample,
including the elements given above, is less than 50 s.
The purpose of the internal standard is to average out variations in ablation and
plasma instabilities. It is difficult (and in most cases impossible) to find suitable
228
standards that have the same matrix as the samples. It is therefore always necessary to use an internal standard during laser ablation work. Usually the concentrations of the major elements are known or can be determined through other
analytical techniques. During quartz analysis, it is usual to choose one of the three
silicon isotopes as an internal standard. The isotope 29Si was chosen as an internal
standard at low mass resolution and 30Si was used at medium and high mass
resolution. 30Si and 29Si have the lowest abundance of the three silicon isotopes,
which is an advantage in this context to avoid tripping of the detector. However,
30
Si has a spectroscopic interference with the molecular ion 14N16O+ at low mass
resolution, which can be produced in significant amounts in the plasma. In medium
and high mass resolution 30Si is not affected by any spectroscopic interference.
The interferences on isotope 29Si are the less abundant molecular ions 13C16O+ and
15 14 +
N N and it is therefore more suitable as an internal standard at low mass
resolution. If methane is used to enhance sensitivity and stability of the plasma
30
Si will be the best suitable internal standard in low mass resolution since it is not
affected by any carbon containing interferences.
10
229
(BAS), and the certified reference material pure substance no. 1 Silicon dioxide
SiO2 from BAM are also used.
The NIST SRM 1830 was cut into four sections of approximately 10 9 10 mm.
The surface was abraded with diamond paste to enhance absorption of the laser
light and to remove any contaminants at the outer surface. BCS 313/1 and BAM
no.1 SiO2 were melted in a tungsten crucible under He/H2 atmosphere at 2000C.
Both standards were then mounted in epoxy and polished until a planar surface
was obtained. Subsequently, the samples were rinsed with de-ionised water.
10:1
where M is the usual sample median. If observations are randomly sampled from a
normal distribution, MAD does not estimate r, the standard deviation, but z0.75r,
where z0.75 is the 0.75 quantile of the standard normal distribution. MAD can be
rescaled so that it estimates r by (Wilcox 1997):
MADN
MAD
MAD
z0:75
0:6745
10:2
xi M
MADN
10:3
230
Ii
XSi
ISi
10:4
In Eq. (10.4) Ii and ISi are corresponding intensities for the analysed element and
reference element (Si), and XSi is the concentration of the reference element. A
weighted linear regression model was used for the calculation of the curve
parameters. The weighted factors, wi, were based on the square root of intensity for
both analysed and reference element (Si), see Eqs. (10.5) and (10.6).
r
1 1
10:5
sYi / ~sYi Yi
Ii ISi
.
1 ~s2Yi
wi P h i.
10:6
1 2
~sYi n
In Eqs. (10.5) and (10.6), sY is the uncertainty in Yi, ~sY represents the expected
relationship for uncertainties in Yi, wi is the weighted factor, and n is the number of
standards used in the regression analysis. The regression analyses are similar to
those used by degrd et al. (1998); further details can be found therein.
10
231
Isotopes
Detection limit,
LOD (lg g-1)
0.4
0.15
0.35
0.06
0.1
0.05
0.02
0.05
14
3
6
1.4
5.3
35
1.4
0.33
1.2
0.07
0.25
(\0.1)
(\1)
\0.2
\1
(\0.007)
(0.013)
(0.021)
\2
\0.5
8.7
(\1)
0.48
0.42
1.3
0.62
\1.3
(\0.002)
Li
Be
11
B
55
Mn
74
Ge
85
Rb
88
Sr
121
Sb
23
Na
24
Mg
27
Al
31
P
39
K
44
Ca
47
Ti
56
Fe
68
Zn
69
Ga
9
3^
r
S
10:7
Typical detection limits are given in Table 10.2. To improve the analytical uncertainty at low concentrations, it is important to have calibration curves with well
defined intercepts. This can be achieved by using certified standards with trace element concentrations lower than the BAM no. 1 SiO2 or a sample blank, if available.
The use of He as carrier gas instead of Ar may give a decrease in LOD by a
factor of up to two (Jacob 2006). Typical detection limits for Ba,Pb, Th, U and Cr
are given by Flem et al. (2002).
10.7 Precision
Routine control of precision and accuracy using independent standards is an
important part of the analytical protocol. In this case the international reference
materials, BR-K1 and BR-FR2 from Breitlnder, Germany, have been used
232
(Flem et al. 2002). In addition NIST SRM 612 is monitored repeatedly throughout
the sequence to document instrumental drift.
10.8 Discussion
For international standards and reference materials, the concentration of the
internal standard is provided on the accompanying certificate. In samples, however, the concentration of the internal standard is unknown. If the samples are
high-purity quartz with\50 lg g-1 total trace element content, it may be assumed
that the silicon oxide concentration is equal to approximately 99.995 wt.% and
relate all the elements analysed to this concentration. Otherwise, the concentration
of the internal standard can be found prior to analysis by the use of electron probe
micro analyses (EPMA; e.g., Mller et al. 2003).
The Ti isotope with the highest abundance, 48Ti, could not be used for the analysis
because of isobaric interference from 48Ca, present in the matrix of the reference
materials NIST SRM 610, 612, 614 and 616 (*12 wt.% CaO). NIST SRM 612 is
certified with 59.1 lg g-1 Ti equal to 37 lg g-1 48Ti (73.8% abundance) while it
contains 163.4 lg g-1 48Ca (0.19% abundance). Therefore, we have chosen to use
47
Ti, despite the low abundance of 7.28%. The measurements were done in medium
mass resolution (MR) because interferences at low mass resolution occur with the
molecular ions 7Li40Ar+ and 15N16O16O+. Due to the high CaO content in the reference materials NIST SRM 610616, the isotope 24Mg produces a signal which
interferes with that of 48Ca2+, which would affect the calibration curve. The samples,
however, do not have the same matrix as the NIST SRM 612616 reference materials, and the lattice-bound impurity level of Ca in natural quartz is usually
\5 lg g-1. It was therefore necessary to avoid the interference of 48Ca2+ on 24Mg by
analysing Mg in MR. The main interferences on 52Cr are the common argides
12 40
C Ar+ and 36Ar16O+, and 40Ar16O+ is the main interference on 56Fe, all of which
are resolved at MR. The main advantage with medium and high mass resolution is
that interferences like these can be avoided. This advantage has to be considered
against the disadvantage in less sensitivity, possibilities of mass drift and the need for
a larger ablation volume (Flem et al. 2002).
In Table 10.3, examples of in situ trace element analysis performed at NGU is
presented for the elements Li, Be, B, Mn, Ge, Rb, Sr, Na, Al, P, K, Ca, Ti, Fe, Ga
for pegmatite quartz applying the analytical parameters described above. The
analysis were performed in connection with a mapping project of regional quartz
occurrences in Troms, northern Norway. Two different crystals (A and B) were
analysed in each polished thick section in order to check the consistency of the
results. Consentration spices of Na (analytical results above detection limit)
indicate that NaCl-bearing fluid inclusions were hit during the ablation process and
that it was too large to be removed by the outlier identification procedure. Also the
lower limit of detection at that specific analysis-day is given in the Table 10.3, if
compared with the LODs given in Table 10.2 most of the elements has similar
11.0
12.6
Nieidajavri
14.4
13.9
Nieidajavri
12.7
11.6
Kfjordbergan 1.95
3.49
Istinden
5.62
6.53
Signalhyda
6.11
6.35
Trnvatnet
0.76
0.69
Trnvatnet
2.79
3.25
Holmevatnet
LOD
0.17
0.30
0.31
0.12
0.16
\0.08
\0.08
0.16
0.11
\0.08
\0.08
\0.08
0.10
\0.08
0.24
0.32
1.44
5.30
3.08
2.77
3.78
3.26
2.25
2.02
2.82
2.50
1.40
\1.28
\1.28
\1.28
\1.28
\1.28
\0.07
\0.07
\0.07
\0.07
0.33
\0.07
\0.07
\0.07
0.38
\0.07
0.08
0.08
\0.07
0.07
0.14
0.17
1.62
2.75
2.96
2.37
2.14
2.08
1.75
1.66
2.13
2.02
2.40
2.47
0.22
0.42
0.69
0.75
0.03
\0.03
\0.03
\0.03
0.04
\0.03
\0.03
\0.03
\0.03
\0.03
0.04
\0.03
\0.03
0.03
0.05
\0.03
0.05
0.03
\0.02
\0.02
0.06
\0.02
0.06
\0.02
0.04
0.04
\0.02
\0.02
0.02
0.08
\0.02
0.03
\6.4
\6.4
\6.4
8.5
15.9
\6.4
21.2
\6.4
\6.4
\6.4
\6.4
\6.4
\6.4
\6.4
\6.4
\6.4
34.6
49.7
44.5
54.2
61.4
49.1
45.4
22.7
38.7
33.9
16.1
17.8
30.2
17.2
14.9
14.0
1.31
2.05
1.33
1.38
3.13
\1.0
4.28
3.78
1.62
6.71
2.39
1.44
3.90
3.70
2.38
\1.0
\2.3
\2.3
\2.3
\2.3
\2.3
\2.3
\2.3
\2.3
\2.3
\2.3
\2.3
\2.3
\2.3
\2.3
\2.3
\2.3
\29
\29
\29
\29
\29
\29
\29
\29
\29
\29
\29
\29
\29
\29
\29
\29
6.90
10.4
1.97
3.17
9.09
5.52
8.43
7.59
9.84
11.1
5.38
4.79
6.79
5.98
5.03
4.42
\0.52
\0.52
\0.52
\0.52
\0.52
\0.52
\0.52
\0.52
\0.52
\0.52
\0.52
\0.52
\0.52
\0.52
\0.52
\0.52
\0.01
\0.01
\0.01
\0.01
\0.01
\0.01
\0.01
\0.01
\0.01
\0.01
\0.01
\0.01
\0.01
\0.01
\0.01
\0.01
Several elements are analysed, both in low mass resolution (LR) and medium mass resolution (MR). Lower limit of detection (LOD) at that specific analysis day is also
given
Sample ID
60,447-A
60,447-B
60,449-A
60,449-B
60,450-A
60,450-B
60,468-A
60,468-B
60,470-A
60,470-B
60,471-A
60,471-B
60,472-A
60,472-B
60,473-A
60,473-B
7
9
11
55
74
85
88
23
27
31
39
44
47
56
69
Li
Be
B
Mn
Ge
Rb
Sr
Na
Al
P
K
Ca
Ti
Fe
Ga
(LR)
(LR)
(LR)
(LR)
(LR)
(LR)
(LR)
(MR)
(MR)
(MR)
(MR)
(MR)
(MR)
(MR)
(MR)
-1
-1
-1
-1
-1
-1
-1
-1
-1
-1
-1
-1
-1
-1
lg g
lg g
lg g
lg g
lg g
lg g
lg g
lg g
lg g
lg g
lg g
lg g
lg g
lg g-1
lg g
0.22
0.08
1.28
0.07
0.07
0.03
0.02
6.4
6.0
1.0
2.3
29
0.66
0.52
0.01
Table 10.3 In situ laser ablation ICPMS analysis (LA-ICP-MS) of pegmatite quartz samples from Troms, Norway
10
In Situ Analysis of Trace Elements
233
234
LODs but some like B is considerably worse in Table 10.3. The reason for this
can be found in instrumental choices of e.g. laser spot size and laser energy at the
laser and plasma power and gas flows at the ICPMS.
To get good-quality in situ analysis of quartz several criteria have to be considered; primarily all of the hardware equipment, the choice of ICP and laser ablation
unit, but as important is also the well defined and matrix-matching standards and
reference materials. At NGU multi-standard calibration is applied permanently.
Since the concentration level of most elements in natural quartz samples is in a range
close to the detection limit and only a very few silicate glass and quartz standards
with concentrations in this range are available, a weighted regression model is used to
give more importance to the lower standards when calculating the regression line.
For these reasons single-point calibration (gas-blank ? a standard with high element
concentrations) is in our opinion not suitable for precise quartz analysis and, therefore, has not been applied at NGU (see also Mller et al. 2008). Since the concentration level of most elements in natural quartz samples is in a range close to the
detection limit, a weighted regression model is used to give more importance to the
lower standards when calculating the regression line.
10.9 Conclusion
We have shown that the ICP-MS with an attached laser system is very suitable
method for the in situ analysis of lattice-bound elements at ultra trace level in single
quartz crystals. In our experience the best choice of instrument is considered to be a
double focusing sector field ICP-MS that provides high sensitivity and mass
resolution high enough to separate most of the possible interferences. The ICP
should be coupled to an 193-nm excimer laser, a femto second laser or a similar
ablation system. The micro optical system of the laser enables the optical control of
the sampling area to be ablated in order to avoid the contamination of the analysis
by fluid and mineral micro inclusions. By using the LA-ICP-MS, most elements in
the periodic table may be analysed in solid quartz samples at lg g-1 to ng g-1
levels. In most cases, the availability of suitable standard materials comprises the
main limitation in the analyses of certain elements.
Acknowledgments We are grateful to Ben Snook and Ian Henderson who improved the English
language of the manuscript.
References
Beurlen H, Mller A, Silva D, Da Silva MRR (2011) Petrogenetic significance of trace-element
data analysed with LA-ICP-MS in quartz from the borborema pegmatite province,
Northeastern brazil. Mineral Mag 75(5):27032719
Breiter K, Mller A (2009) Evolution of rare-metal granitic magmas documented by quartz
chemistry. Eur J Mineral 21:335346
Fanderlik I (1991) Silica glass and its application. Elsevier, Amsterdam
10
235
236
Chapter 11
Cathodoluminescence Microanalysis
of the Defect Microstructures of Bulk
and Nanoscale Ultrapure Silicon Dioxide
Polymorphs for Device Applications
Marion A. Stevens-Kalceff
M. A. Stevens-Kalceff (&)
School of Physics and Electron Microscope Unit,
University of New South Wales, Sydney, NSW 2052, Australia
e-mail: Marion.Stevens-Kalceff@unsw.edu.au
237
238
M. A. Stevens-Kalceff
11.1 Introduction
Silicon dioxide is an important material in many technologically important
applications including for example, optics, optical fiber and silicon semiconductor
device technology, etc. The optical, electrical and mechanical properties are
dependent on the presence of defects (imperfections and impurities). Characterization of the defect microstructure of silicon dioxide (SiO2) allows the performance of these applications to be optimised. In addition the defect structure of
natural silicon dioxide polymorphs (e.g. crystalline and metamict quartz) can give
useful insight into the geological processes associated with their genesis. Despite
efforts over many years, aspects of the defect structure of silicon dioxide polymorphs remain controversial.
Cathodoluminescence (CL) is the non incandescent luminescent emission from
a material during electron irradiation. The emission of photons in luminescence
processes is due to electronic transitions involving localised band gap states and/or
the conduction band and/or valence band. In an electron microscope, the focused
beam of electrons enables spatially resolved information about the defect structure
of a specimen to be obtained. CL techniques provide nondestructive, sensitive,
high resolution methods of assessing the relative concentration and distribution of
defects in SiO2 polymorphs. The systematic analysis of a range of pure silicon
dioxide polymorphs and specimens for which the impurity concentrations are
known provides a good foundation for the characterization of the defect microstructure of more complex specimens such as natural specimens containing
impurities (e.g. Al, Ti, Fe, Ge). Cathodoluminescence (CL) techniques have been
successfully used for many years in mineralogy and petrology applications, Pagel
et al. (2000), Gotze and Kempe (2008) but no single technique can provide a
complete analysis. For example, some diamagnetic defects associated with luminescent emissions (e.g. neutral oxygen vacancy Skuja (2000)) are not directly
detectable using Electron Paramagnetic Resonance (EPR) techniques. Conversely
some paramagnetic defects, which have been characterised by EPR, are not
associated with luminescent processes (e.g. peroxy radical Nishikawa et al.
(1990)), although their presence may sometimes be inferred as a precursor to an
observed emission. A range of complementary analyses are therefore necessary for
a more comprehensive understanding of the defect structure of silicon dioxide
polymorphs.
Silicon dioxide polymorphs are known to be sensitive to ionizing and energetic
irradiation. In particular, defect centers may be generated and/or modified by
electron irradiation. Atomic displacements from the tetrahedral silicon dioxide
lattice sites can result from elastic scattering via momentum transfer processes
(Hobbs et al. 1994). For example electron beam energies of 200 or 65 keV are
necessary for the removal of a silicon or oxygen atom respectively (Pfeffer 1985).
Electron-beam induced atomic displacements and modification of the low pressure
tetrahedral silicon dioxide microstructure can also occur as a result of inelastic
scattering via radiolytic processes. Radiolysis is an electronic process that results
11
239
240
M. A. Stevens-Kalceff
Table 11.1 Maximum concentration of major impurities in the as-received Silicon dioxide
polymorph specimens investigated by Cathodoluminescence spectroscopy
-OH
Al
Cl
Other
SiO2 polymorph
(e.g. Ca Na K Li)
Synthetic high purity a-SiO2
(Sawyer z cut a-quartz)
Type I a-SiO2 (Infrasil)
Type II a-SiO2 (Herasil)
Type III a-SiO2 (Spectrosil)
Type IV a-SiO2 (Spectrosil WF)
50, 300 and 900 nm thick
Dry Thermal SiO2 on Si(100)
400 nm thick buried SiO2 sandwiched
between the 220 nm Si top layer
and the Si(100) substrate. (SIMOX
400)
\300 ppm
\2 ppm
\2 ppm
\5 ppm
\15 ppm
\1 ppm
\180 ppm \50 ppm
\4 ppm
*1000 ppm *0.1 ppm *50 ppm \1 ppm
*10
*0.1 ppm *180
\1
Nominally pure
(Impurity concentrations depend on substrate)
Si nanoparticles and clusters (form during oxygen ion
implantation and high temperature annealing)
Impurity concentrations also depend on substrate
(Sawyer Research Products Inc. 1994), (Heraeus Quarzglas 1995), (Saint-Gobain Quartz Ltd
1997), (MTI Corporation 2010), (IbIS Technology Corporation 2000) Impurity and native defects
may provide precursor states for irradiation-induced defects. Substitutional and interstitial species
may exist in the specimens as a result of the method of synthesis. Depending on the specimen
temperature, interstitial species may include both atomic and molecular/ion species of O, H, Cl,
(e.g. O2- , ClO2, OH- , H+ , HCl etc.) depending on the specimen type
in a-SiO2 but ranges between 120180 in the a-SiO2 (Wright 2000). The density
of a-SiO2 is 2.7 g.cm-3 while the density of the more open structured amorphous
a-SiO2 is typically 2.2 g.cm-3. A range of crystal and amorphous silicon dioxide
specimens were investigated and are described below. Their impurity concentrations are compared and summarised in Table 11.1.
Single crystal bulk synthetic z-cut premium ultra-pure a-quartz produced by
hydrothermal processes was investigated (Sawyer a-quartz (Sawyer Research
Products Inc.)). A range of homogeneous bulk amorphous silicon dioxide (a-SiO2)
specimens were also investigated including anhydrous (Type I) and hydrated fused
quartz (Type II), and hydrous (Type III) and anhydrated (Type IV) silica glass.
Pure fused quartz SiO2 glasses are produced by melting natural quartz. Type I
a-SiO2 is obtained by electric melting of natural quartz crystals under vacuum,
while Type II a-SiO2 is produced by H2O2 flame fusion of natural quartz, which
results in significant residual -OH concentration. Type III a-SiO2 and Type IV aSiO2 synthetic fused silica are produced by chemical oxidation of SiCl4 to form
SiO2 via vapor-phase hydrolysis or electrically fused respectively. The cation
impurities in Types III and IV are substantially lower than in Types I and II
because of the high purity of SiCl4 used in the processing (Hench and Vasconcelos
1990). However, -OH concentrations of up to 1,000 parts per million (ppm) may
be present in Type III silica due to flame hydrolysis, while unreacted residual
chlorine ions in concentrations of up to a few hundred ppm are retained in both
Types III and IV a-SiO2. Particular methods of manufacture are proprietary processes and the descriptions above are general, however commercially available
11
241
242
M. A. Stevens-Kalceff
11.3 Results
Tetrahedral silicon dioxide polymorphs may be modified by electron irradiation
(Stevens-Kalceff and Phillips 1995a), (Stevens-Kalceff 2000), (Stevens-Kalceff
1998). It is therefore necessary to monitor the effect of electron irradiation on the
CL spectra from SiO2 polymorphs during data collection. Irradiation-induced
effects can be minimised by reducing the electron power density delivered to the
electron beam-specimen interaction volume. This can be achieved by limiting the
electron beam current, increasing the volumes of the specimen from which the CL
is generated and/or exposing the specimen intermittently to the electron beam.
To minimise electron irradiation effects, the CL data presented in this report were
generated by a 10 keV, 4.2 nA electron beam scanned over 2500 lm2 areas of the
surfaces of the range of synthetic silicon dioxide polymorphs described in
Table 11.1, at a rate of 0.2 s/frame (*1.23 9 106 pixels/frame).
CL spectra have been collected as a function of wavelength k (nm), and following subtraction of dark current have been corrected for instrument response.
The response of the components within the CL system (e.g. gratings, detectors,
etc.) is not uniform across the detected wavelengths. For example, the sensitivity
of the CCD camera used to detect the CL spectra presented in this report is
maximised at wavelengths *650750 nm, but reduces to \30% of the maximum
sensitivity for wavelengths [950 and \450 nm. Thus instrument response corrections are important prior to componentizing spectra when CL emission wavelengths are at the low sensitivity limits of CL detection systems and in particular
for materials with broad emissions such as SiO2 polymorphs. As the instrument
broadening function is negligible with respect to the natural width of the SiO2
emission components, a simple spectral decomposition rather than deconvolution
is sufficient (Stevens-Kalceff 2009). Cathodoluminescence spectra must also be
converted from wavelength k (nm) to energy E (eV) prior to peak fitting/componentization, as SiO2 CL emission components are approximately Gaussian in
energy space:
11:1
E hc=k and IE k2 =hc Ik
Figure 11.1a shows typical CL spectra from bulk synthetic a-SiO2 (quartz)
excited with a normal incidence electron beam of energy 10 keV and current 4.2
nA scanned over 2500 lm2 surface regions as a function of increasing electron
beam irradiation exposure time (i.e. dose) ranging from 50 to 500 s. The 10 keV
electrons penetrate up to *1 lm into silicon dioxide, and therefore the CL is
generated from a volume of up to *2500 lm3. Similarly, CL spectra from the
bulk, thin film and buried amorphous silicon dioxide specimens have been collected under identical conditions (10 keV, 4.2 nA electron beam scanned over
2500 lm2 regions as a function of increasing electron beam irradiation exposure
ranging from 50 to 500 s). All spectra have been corrected for instrument response
and converted from wavelength to energy space to enable fitting of Gaussian
11
243
1 R2 expakt
1 R2 exp2akt
11:2
where R(k) is the reflectivity calculated from the wavelength dependent refractive
index n(k):
n1 2
R
11:3
n1
These formulae assume a number of simplifications. It is assumed for example
that the interfaces and surface are smooth with no reflective or absorptive losses.
Equations 11.2 and 11.3 (e.g. see Bhat et al. (2008)) and the optical properties of
silicon Green (2008) have been used to calculate the transmission of photons
through a 220 nm thick silicon layer (see Fig. 11.2). More than 99% of photons of
244
M. A. Stevens-Kalceff
(a)
(b)
(c)
500s
500 s
50s
500 s
50 s
Energy (eV)
50 s
Energy (eV)
(d)
Energy (eV)
(e)
(f)
500 s
500 s
500 s
50 s
50 s
50 s
Energy (eV)
Energy (eV)
(g)
Energy(eV)
(h)
(i)
500 s
500 s
500 s
3
4
Energy(eV)
50 s
50 s
50 s
Energy (eV)
Energy (eV)
Fig. 11.1 Typical CL spectra from a single crystal z-cut a-SiO2; b bulk Type I a-SiO2; c bulk
Type II a-SiO2; d bulk Type III a-SiO2; e bulk Type IV a-SiO2; f 50 nm thick dry thermal a-SiO2
on Si (001); g 300 nm thick dry thermal a-SiO2 on Si (001); h 900 nm thick dry thermal a-SiO2 on
Si (001); i in situ 400 nm thick buried oxide layer in Si (001): SIMOX. These SiO2 polymorphs
have been irradiated with a 10 keV, 4.2 nA electron beam as a function of increasing irradiation
dose for t = 50, 100, 150, 200, 150, 300, 350, 400, 450 and 500 s. The spectra have been corrected
for instrument response, and in the case of (i) the in situ 400 nm thick buried oxide layer in Si
(001), have also been corrected for the effects of optical absorption by the silicon top layer for
energies \3.2eV
energy \1.35 eV (i.e. [920 nm), will be transmitted through the 220 nm thick
silicon top layer (Stevens-Kalceff 2011). However photons of energy [3.2 eV
(i.e. \390 nm) will be absorbed by the 220 nm silicon top layer and therefore will
not be detected (Stevens-Kalceff 2011). Normal components of the emergent CL
emission are attenuated by 220 nm of the silicon top layer, as shown in Fig. 11.2.
11
245
Wavelength (nm)
1000
250
500
Normalized transmission
1.0
220 nm of Silicon
0.5
0.0
1
Energy (eV)
246
M. A. Stevens-Kalceff
11.4 Discussion
Consistent with previous studies, CL emission from tetrahedral silicon dioxide
polymorphs is observed to change as a function of electron beam exposure due to
irradiation-induced defect generation (radiolysis) and transformation (see
Fig. 11.1). Electron irradiation can also result in electrostatic charging effects and
temperature dependent effects, and CL microanalysis allows the in situ monitoring
of the formation and transformation of irradiation sensitive defects.
The radiative transition between two different linearly coupled electronic states
of a point defect in a vibrating lattice produces a characteristic emission profile at
low temperatures. The shapes of luminescence bands are dependent on the magnitude of electronphonon coupling. If there is weak defect-host lattice interaction,
the spectral profile consists of a purely electronic transition (zero phonon line;
ZPL) and a series of equally spaced, lower energy, overlapping phonon replicas
related to fundamental lattice vibration frequencies. The lower energy phonon
replicas broaden successively and overlap to form an asymmetric sideband which
is described as a Pekarian envelope (Henderson and Imbusch 1989). In the case of
strong coupling the resultant emission bands are broad and approximately
Gaussian in profile (Henderson and Imbusch 1989). Many SiO2 defects are
characterised by strong defect-lattice coupling resulting in relatively large Stokes
shifts (e.g. self trapped exciton (Itoh et al. 1994) with *6 eV Stokes shift in
a-SiO2 Ismail-Beigi and Louie (2005)) or larger Huang Rhys factors (e.g. E0 centre
with S = 13.2, Palma et al. (1996) and homogeneous broadening which can range
up to 1 eV (Skuja 2000). Homogeneous broadening results from strong coupling
between electronphonon transitions and is the same for all related defects
throughout the specimen. CL emissions associated with defects with strong
electronphonon coupling, are therefore fundamentally broad, Gaussian in profile
and generally without fine structure even at cryogenic temperatures, due to
homogeneous broadening. Inhomogeneous broadening results from local disorder
in the SiO2 host lattice at the defect site. Inhomogeneous broadening is minimal
for ultrapure synthetic a-SiO2, but can become significant in natural and metamict
quartz. For defects in amorphous silicon dioxide SiO2 the main contribution to the
peak widths of luminescence emissions is usually associated with homogeneous
broadening of *0.21 eV (Skuja 2000) while the magnitude of inhomogeneous
broadening is typically up to *0.1 eV (Skuja 2000). CL is produced by relatively
high energy non-selective excitation, which may be associated with radiolytic
radiation damage of SiO2 and enhanced inhomogeneous broadening of CL emissions. The corrected experimental CL spectra from the SiO2 polymorphs in
Fig.11.1 have broad overlapping emission components which have been fitted with
multi-parameter Gaussian functions using a non linear least squares curve fitting
algorithm. The minimum numbers of statistically significant components have
been fitted to each spectrum comparing v2 factors (Stevens-Kalceff 2009).
11
(b)
Normalized integrated
CL intensity
(a)
247
4
1.9 eV
100
Energy (eV)
200
300
400
500
Fig. 11.3 a Typical example of a componentised CL spectrum from single crystal z-cut a-SiO2
showing the fitted Gaussian component for an electron irradiation exposure of t = 50 s, from
Figs. 11.1a. b Normalised integrated CL emission intensities of the fitted Gaussian spectral
component from Fig. 11.1a as a function of electron irradiation dose for single crystal z-cut
a-SiO2. The integrated CL intensity is normalised to t = 50 s electron irradiation exposure
248
M. A. Stevens-Kalceff
irradiation also induces other CL emissions associated with electron irradiationinduced defects in a-SiO2 (quartz) including the Oxygen Deficient Centers (e.g.
at *4.3 eV/290 nm), and the radiative recombination of Self Trapped Excitons
involving E10 centers (e.g. at *2.7 eV/460 nm). The quartz CL emission energy,
peak width and identifications are summarised in Table 11.2.
The spectra in Fig. 11.1a confirm that a-SiO2 is sensitive to electron beam
irradiation, even for low dose electron beam irradiation. The observation of
changes in the CL spectrum due to the modification of electron irradiation sensitive defects can provide additional useful information. The integrated intensities
of each Gaussian emission component may be plotted as a function of irradiation
exposure. These emission response plots summarise the irradiation-induced
changes in the intensities of the CL emission bands in a spectrum. In Fig. 11.3b the
integrated CL emission intensity of the fitted Gaussian spectral component normalised to t = 50 s (see Fig. 11.3a) from each of the spectra in Fig. 11.1a is
plotted as a function of electron beam irradiation time for bulk synthetic a-SiO2
(quartz). The unresolved multi-component *1.9 eV emission increases in intensity consistent with electron beam induced modification of the defect structure of
a-quartz.
The data in Figs. 11.1a and 11.3 and previous studies show that electron irradiation induces defects in a-quartz (Hobbs and Pascucci 1980) and natural quartz
(Rusk et al. 2006). At higher doses the irradiation-induced disorder can locally
transform the crystal to an amorphous environment (i.e. bond lengths and bond
angles are distorted in the local vicinity of irradiation-induced defects). For
example stationary electron beam irradiation can produce localised amorphised
outgrowths on the surface of quartz (Stevens-Kalceff and Phillips 1995b),
(Stevens-Kalceff et al. 1996), (Stevens-Kalceff 2000). Localised amorphous
environments may occur within natural crystal quartz due to the presence of
defects. The defect structure and associated CL spectra from amorphous silicon
dioxide polymorphs may give insight into the microstructure of amorphised and
metamict natural quartz.
535
460
565
590
2.69 0.03
2.7 0.92
(0.37 0.03) (0.35 0.01)
2.2 0.1
2.2 0.15
(0.43 0.02) (0.4 0.05)
O3:Si-OH
2.15 0.01 (0.20 0.01) Si/SiO2
650
Si/H
Si/SiO2
760
(continued)
Si nano-clusters
Oxidised silicon nano-cluster
ODC(II): 2-fold coordinated Si
Table 11.2 Experimental emission energies and full width at half maximum of low dose CL emission components observed from synthetic pure crystal and
amorphous silicon dioxide polymorphs in comparison with in situ buried oxide layer in silicon
Wave- CL energy (eV)
Association/Identification
length FWHM, eV
(nm)
Bulk
Dry thermal Buried a-SiO2 layer
Bulk
a-SiO2
SiO2
a-SiO2
Silicon dioxide (defect free)
O3:Si-OSi:O3
11
Cathodoluminescence Microanalysis of the Defect Microstructures
249
4.50.15
(0.50.7)
275
4.450.10
(0.60.75)
Dry thermal
SiO2
*3c
O2=Si:
O3:SiSi:O3
O3:Al:MOSi:O3
O3:Si-OSi:O3
Association/Identification
2500 lm2 regions of each SiO2 polymorph have been irradiated with a 10 keV, 4.2 nA electron beam for between t = 50500 s. In the case of the buried
oxide layer, photons of energy greater than *3.2 eV are not transmitted through the *220 nm silicon top layer. The maximum uncertainties (experimental
and fitting) are indicated
a
2+ unresolved components at low dose (Stevens-Kalceff 2009)
b
Evidence for extra component in Type II and Type III (hydrated) a-SiO2 at low doses
c
Emission profile is truncated for energies [3.2 eV due to reduction in transmission through Si
3.45 0.10
(0.8 0.05)
365
415
250
M. A. Stevens-Kalceff
11
251
for Type I-IV a-SiO2, but the relative and absolute emission intensities are
different.
The four common CL emissions observed from bulk amorphous SiO2 (see
Fig. 11.4) are due to native point defect centers associated with the silicon dioxide
tetrahedral structure (O3:Si-OSi:O3) and include non-bridging oxygen hole
centers at 1.9 0.02 eV (NBOHC: O3:Si-O) where () represents an unpaired
electron. The homogenous broadening of the NBOHC in amorphous SiO2 has been
investigated using site selective photoluminescence excitation and transient
spectral hole-burning techniques revealing spectral fine structure corresponding to
a Huang-Rhys factor of *1.5 (Skuja et al. 1995). This spectral fine structure is not
resolved in the CL experiments which reveal an approximately Gaussian peak
shape for the *1.9 eV emission (FWHM *0.18 eV). CL is produced by relatively high energy non-selective excitation provided by an energetic (keV) electron
beam, and may be associated with radiation damage and enhanced inhomogeneous
broadening, resulting in the approximately Gaussian profile of the *1.9 eV CL
emission. The broad CL emission (FWHM *0.4 eV) observed at 2.2 0.1 eV is
associated with the radiative recombination of the self trapped exciton (e.g. STE:
O3:Si-OSi:O3? O3:Si-OSi:O3). The STE is a correlated electronhole
pair localised in a self-induced lattice distortion (Williams and Song 1990).
Radiative STE recombination from SiO2 is characterised by lifetimes typically of
order ms (Tsai and Griscom 1991), a large Stokes shift, (Ismail-Beigi and Louie
2005) and an intrinsically broad, approximately Gaussian band profile due to
relatively strong defect-lattice coupling (Williams and Song 1990). Note, for
example that the E0 defect (Si:O3), which is a component of the STE, has a
Huang-Rhys factor which has been determined to be *13.2 (Palma et al. 1996).
CL emissions associated with Oxygen Deficient Centers (ODC) are observed at
2.69 0.03 eV and 4.5 0.15 eV. The variation of peak energies (4.44.65 eV)
and peak widths (0.50.7) between the different bulk a-SiO2, are possibly associated with the concentration of defect precursors introduced during synthesis, and
resultant sensitivities to energetic ionising radiation and inhomogeneous broadening. A variety of oxygen deficient type defects have been proposed for amorphous SiO2 (Skuja 1998, 2000). Oxygen Deficient Centers (ODC) in amorphous
SiO2 are known as ODC(I) and ODC(II). The ODC (I) has been associated with
the relaxed neutral oxygen vacancy (i.e. O3:SiSi:O3) with a covalent SiSi
bond and reduced atom spacing of *2.5 , and a luminescent emission
at *4.4 eV (Nishikawa et al. 1994), (Pacchioni and Ierano 1997). A possible
configuration for the ODC(II) has been proposed, consistent with damage induced
by energetic ionizing radiation, and is known as the two-fold coordinated silicon
defect (O2=Si:) (Skuja et al. 1984), (Griscom 1991), (Skuja 1994). Theoretical
simulations have predicted luminescent emissions associated with the silica
ODC(II) at both 2.6 eV (T1?S0) and 4.6 eV (S1?S0) (Pacchioni and Ierano
1998). In addition, to the four common CL emissions, an emission
at *3.4 0.1 eV (*365 nm) is observed in Type I and II a-SiO2 CL spectra (see
Figs. 11.1b, c, 11.4(i)a and (ii)a). This radiation sensitive CL emission is due to
low concentrations of Al impurities in natural quartz raw material used in the
252
M. A. Stevens-Kalceff
(a)
(b)
Normalized integrated
CL intensity
1.9 eV
2.2 eV
2.7 eV
3.4 eV
4.5 eV
3
2
1
0
200
300
400
500
Normalized integrated
CL intensity
100
Energy (eV)
1.9 eV
2.2 eV
2.7 eV
3.4 eV
4.5 eV
3
2
1
0
Energy (eV)
x10
200
300
400
500
Normalized integrated
CL intensity
100
1.9 eV
2.2 eV
2.7 eV
4.5 eV
3
2
1
0
Energy (eV)
100
200
300
400
500
Normalised integrated
CL intensity
1.9 eV
2.2 eV
2.7 eV
4.5 eV
3
2
1
0
Energy (eV)
100
200
300
400
500
Fig. 11.4 a Typical example of a componentised CL spectrum from amorphous bulk (i) Type I
a-SiO2; (ii) Type II a-SiO2; (iii) Type III a-SiO2; and (iv) Type IV a-SiO2 polymorphs showing
the fitted Gaussian components for an electron irradiation exposure of t = 50 s, from Figs 11.1b,
c, d, e. b Normalised integrated CL emission intensities of the fitted Gaussian spectral component
from Figs 11.1b, c, d, e as a function of electron irradiation dose for (i) Type I a-SiO2; (ii) Type II
a-SiO2; (iii) Type III a-SiO2; and (iv) Type IV a-SiO2 polymorphs. The integrated CL intensity is
normalised to t = 50 s electron irradiation exposure
11
253
manufacture of Type I and II a-SiO2. The *3.4 eV may be associated with the
charge compensated substitutional Al3+ alkali ion defect which has been reported
in amorphous and natural crystalline SiO2 polymorphs; O3:(Al3+:M+)-OSi:O3 where M+ is typically Li+, Na+, K+ or H+ which has been characterised
using evidence from EPR (Halliburton et al. 1981), thermally stimulated luminescence, (Alonso et al. 1983), and cathodoluminescence experiments (Ramseyer
and Mullis 1990), (Stevens-Kalceff 1998). Electron irradiation tends to dissociate
the charge compensating hydrogen or alkali ion from substitutional aluminium,
attenuating the *3.4 eV CL emission (Ramseyer and Mullis 1990), (StevensKalceff and Phillips 1995a), (Gorton et al. 1997), (Gotze et al. 2001). The CL
emission energies, peak widths and identifications for bulk amorphous SiO2
polymorphs are summarised in Table 11.2.
The CL energies and peak widths are generally similar but the relative CL
intensities are different for each bulk a-SiO2 polymorph. For example, in comparison with the 1.9 eV emission the intensities of the 2.2, 2.7, 3.4 and *4.5 eV
CL emissions are greater for Types I and II a-SiO2 than for Type IV and in
particular for Type III a-SiO2. This is consistent with the generally higher concentrations of impurities in Type I and Type II a-SiO2, noting that chlorine
impurities are not directly associated with any of the observed CL emissions from
a-SiO2. The peak widths of the 2.2 and 2.7 eV CL emissions from Type I a-SiO2
are generally slightly broader (i.e. by *0.05 eV) than the peak widths of CL
emissions from the other bulk a-SiO2 polymorphs. The broader emissions from
Type I a-SiO2 are likely to be due to enhanced inhomogeneous broadening
resulting from local disorder in the SiO2 host lattice. The electron beam radiation
responses of each emission component in each spectrum are different consistent
with their association with a range of independent defect centers (see Fig. 11.4(i)b,
Fig. 11.4(ii)b, Fig. 11.4(iii)b and Fig. 11.4(iv)b). Electron irradiation of bulk
a-SiO2 generally results in enhancement of the CL emissions because of the
increase in defect concentration. For example the *2.7 eV emission associated
with oxygen deficient defects is observed to increase significantly in all cases as a
function of dose. The *4.5 eV CL emission which is also associated with oxygen
deficient defects, is not as sensitive to electron irradiation in Types II and III
a-SiO2 (that have significant concentrations of OH impurities) in comparison with
the *4.5 eV CL emission in Types I and IV a-SiO2 (with negligible OH concentration). See Table 11.1. Note that for bulk Type III a-SiO2, the very low
intensity of the *4.5 eV emission results in a large uncertainty in the estimation
of the associated normalised integrated CL intensities: The modification of
the *4.5 eV emission from Type III a-SiO2 by the low dose electron irradiation is
not significant within experimental uncertainty (see Fig 11.4(iii)b). Mobile OH
interstitials may anneal some types of oxygen deficient defects. The electron
irradiation response of Type I-IV a-SiO2 polymorphs is influenced by trace levels
of impurities including hydrogen (e.g. H, OH), aluminium, alkali ions and chlorine. The *3.4 eV CL emission associated with aluminium impurities in Type I
and II a-SiO2 is attenuated due to the irradiation-induced disassociation of the
charge compensating hydrogen or alkali ion from substitutional aluminium (Al3+).
254
M. A. Stevens-Kalceff
11
255
(b)
5
Normalized integrated
CL intensity
CL intensity (arb.units)
(a)
1.9 eV
2.2 eV
2.7 eV
4.5 eV
4
3
2
1
0
200
300
400
500
Normalized integrated
CL Intensity
100
Energy (eV)
1.9 eV
2.2 eV
2.7 eV
4.5 eV
4
3
2
1
0
0
100
200
300
400
500
Energy (eV)
Normalized integrated
CL intensity
1.9 eV
2.2 eV
2.7 eV
4.5 eV
4
3
2
1
0
Energy (eV)
100
200
300
400
500
Fig. 11.5 a Typical example of a componentised CL spectrum from (i) 50 nm (ii) 300 nm and
(iii) 900 nm thick dry thermal a-SiO2 on Si (001) showing the fitted Gaussian components for an
electron irradiation exposure of t = 50 s, from Figs. 11.1f, g, h. b Normalised integrated CL
emission intensities of the fitted Gaussian spectral component from Figs. 11.1f, g, h as a function
of electron irradiation dose for (i) 50 nm (ii) 300 nm and (iii) 900 nm thick dry thermal a-SiO2 on
Si (001). The integrated CL intensity is normalised to t = 50 s electron irradiation exposure
256
M. A. Stevens-Kalceff
sites. Disorder can be localised in the vicinity of defects (point defects, interfaces,
surfaces, etc.) and at sites of strain and irradiation damage. Inhomogeneous
broadening is particularly evident in CL from the 50 nm thick dry thermal a-SiO2,
where the effects of interface and surface states are more significant in comparison
with defects in the volume of the thin film. In contrast, the CL peak energies, peak
widths and electron irradiation responses of the 300 nm and in particular the
900 nm thick dry thermal a-SiO2, approach that of the bulk Type IV a-SiO2. It is
also noted that in comparison with the 1.9 eV CL emission, the intensities of the
2.2, 2.7, and *4.5 eV CL emissions are greater for the 50 nm thick dry thermal
a-SiO2 than for the 300 or 900 nm thick dry thermal a-SiO2. This is consistent with
generally higher concentrations of defects (and higher surface/volume and interface/volume ratio) in the thinner dry oxide. It is also noted that the variation in the
energy and peak width of the *4.5 eV emission observed in the CL spectra from
the bulk a-SiO2 Type I-IV a-SiO2 specimens is not observed in the dry Thermal
SiO2 specimens which are synthesised using the same process and have the
same impurity concentrations. The variety of responses of CL emission intensities
to electron irradiation shown in Figs. 11.4, 11.5, 11.6 are consistent with defect
generation and, in particular, oxygen vacancy generation in electron irradiated
amorphous SiO2.
11
(a)
(b)
Normalized integrated
CL intensity
257
1.5
1.0
0.5
1.65 eV
1.9 eV
2.15 eV
2.35 eV
2.7 eV
0.0
2
Energy (eV)
100
200
300
400
500
Fig. 11.6 a Typical example of a componentised CL spectrum from in situ 400 nm thick buried
oxide layer in SIMOX on Si (001) showing the fitted Gaussian components for an electron
irradiation exposure of t = 50 s, from Fig. 11.1(i). b Normalised integrated CL emission intensities
of the fitted Gaussian spectral component from Fig. 11.1(i) as a function of electron irradiation
dose for in situ 400 nm thick buried oxide layer in SIMOX on Si (001). The integrated CL intensity
is normalised to t = 50 s electron irradiation exposure
Oxygen vacancies are known to be incorporated into the buried oxide during
the fabrication process. Following oxygen ion implantation, the subsequent hightemperature anneal results in an oxygen reduced oxide due to out-diffusion of
oxygen into the silicon and/or in-diffusion of a reducing species from the silicon
(Nishikawa et al. 1999). Silicon nanoparticles are formed within the buried oxide
layer during the post implantation, high-temperature anneal (Afanasev and
Stesmans 1999), (Nishikawa et al. 1999). Investigations of the charge trapping
properties indicate smaller Si clusters (\1 nm diameter) are found to be approximately uniformly distributed through the buried oxide and are associated with
confined exciton states in Si nanocrystals (Revesz and Hughes 1997), (Afanasev
et al. 1996a). In comparison, larger (*34 nm diameter) Si clusters have been
found near the interface between the buried oxidesilicon substrate (Afanasev
et al. 1996a) and silicon top layerthe buried oxide (Nishikawa et al. 1999).
The fabrication of a buried oxide structure via the SIMOX process produces
significant residual strain which is maximised at the upper and lower Si-SiO2
interfaces (Camassel et al. 2001). Residual compressive strain is observed in the
buried oxide layer between the two SiSiO2 interfaces. Residual tensile strain is
observed across the silicon top layer from the upper Si-SiO2 interface and
extending from the lower SiSiO2 interface into the silicon substrate(Camassel
et al. 2001). The formation or transformation of SiO2 defects which involve the
structural rearrangement or relaxation of the surrounding oxide network may be
hindered in the confined strained oxygen deficient buried oxide layer. The strained
confined buried oxide is also likely to contain significant concentrations of surface
or interface defects associated with the SiSiO2 interfaces and the Si nanoclusters
(which have large surface to volume ratios).
258
M. A. Stevens-Kalceff
The CL data in Figs. 11.1i, 11.6a and b have been collected under low dose
conditions. High dose CL spectroscopy of the buried oxide also reveals a minimum of six Gaussian CL emissions components with similar wavelengths and
peak widths, but different relative intensities. The origins of these emissions have
previously been described in detail (Stevens-Kalceff 2011). Silicon nanoparticles
formed within the buried oxide layer during the post implantation, high temperature anneal are associated with the *1.65 eV (*750 nm) see Figs. 11.1i and
11.6a. Luminescence from nanostructured silicon is believed to be strongly
dependent on surface states and surface passivation (Saeta and Gallagher 1997),
(Wolkin et al. 1999), (Fauchet et al. 1998). The *1.65 eV CL emission is related
to either Si nanoparticleSiO2 interface defects or passivated quantum confined
silicon nanoparticles (Afanasev and Stesmans 1999), (Saeta and Gallagher 1997),
(Wolkin et al. 1999), (Vasiliev et al. 2002) or competition between these two
recombination mechanisms (Godefroo et al. 2008). During electron irradiation,
radiolytic dissociation of the passivating species from the silicon nanoparticles will
result in the attenuation of the 1.65 eV CL emission.
An unresolved low intensity emission component is observed at 1.9 eV
(FWHM *0.17 eV) from the buried oxide layer: The 1.9 eV (650 nm) CL
emission may be partially resolved in high dose experiments only when the
adjacent CL components have been sufficiently attenuated by the electron beam
irradiation (Stevens-Kalceff 2011). The fitted 1.9 eV CL emission energy and peak
width are consistent with the 1.9 eV CL emission (full width half maximum;
FWHM *0.18 eV) observed from bulk a-SiO2 observed under the same conditions and associated with the native NBOHC defect. CL emissions at 2.15 eV
(FWHM 0.20 eV) and 2.35 eV (FWHM 0.18 eV) are observed in CL spectra from
the buried oxide layer (see Figs. 11.1i and 11.6a). The peak widths of both the
2.15 eV (590 nm) and 2.35 eV (535 nm) emissions from the buried oxide layer are
half that observed under identical excitation conditions for the 2.2 eV (*565 nm)
CL emission from bulk SiO2 which is attributed to the radiative combination of the
STE (Stevens-Kalceff 2011). Thus neither the 2.15 or 2.35 eV CL emissions from
the buried oxide layer are likely to be associated with the characteristically broad
STE. In addition the residual strain within the buried oxide layer will reduce
the probability of radiative relaxation of the SiO2 tetrahedral structure and
therefore radiative recombination of the STE is unlikely (Stevens-Kalceff 2011).
The absence of a CL component associated with the STE is also consistent with
complementary data showing enhanced concentrations of E type defect centers
(e.g. Si:O3) in SIMOX (Revesz and Hughes 1997), (Warren et al. 1993) as the
radiative STE recombination would reduce the concentration of E defects (e.g.
O3:Si-OSi:O3? O3:Si-OSi:O3+hmSTE) (Stevens-Kalceff 2011). The
2.15 eV emission has been observed only in CL spectra from buried SiO2 to date
(Stevens-Kalceff 2011). It is most sensitive to electron irradiation and may possibly be associated with (hydrogen) passivated silicon nano-clusters. Electron
beam induced dissociation of the passivating species from the silicon nanoparticles
will result in the attenuation of the 2.15 eV CL emission. Previous photoluminescence (PL) investigations have removed the silicon top layer to enable efficient
11
259
excitation and detection of luminescence from the buried oxide: Etch-back photoluminescence experiments have revealed a 2.4 eV PL emission associated with
silicon nanoclusters localised at the interface between the silicon top layer and the
buried oxide. The CL emission observed at 2.35 eV (FWHM 0.18 eV) has a
similar emission energy and peak width to the 2.4 eV PL emission.
The peak width of the CL emission from the buried oxide layer at 2.7 eV
(FWHM 0.23 eV) is significantly less than the 2.7 eV (FWHM 0.35 eV) emission
observed from bulk a-SiO2 under identical conditions. Due to strong defect lattice
coupling, the peak width of the *2.7 eV emission from bulk a-SiO2 is intrinsically broad (*0.30 eV) even at 5 K (Stevens-Kalceff 1998). (Reducing specimen
temperature assists in the resolution of spectral components due to reduction in the
CL peak widths with a decrease in thermal broadening and enhancement of CL
intensities associated with reduced diffusion of competitive non-radiative centers).
In addition, the formation of ODC (II) defects which is associated with significant
structural rearrangement is likely to be hindered in the strained confined buried
oxide layers (Nishikawa et al. 1999). Therefore in contrast to the situation
observed in bulk a-SiO2 e-irradiated under the same conditions, the 2.7 eV CL
emission from the buried oxide layer is unlikely to be associated with the ODC
(II). Insight into the possible origin of the CL emission observed at *2.7 eV from
the buried oxide layer may be obtained from complementary investigations.
Photoionization spectroscopy of silicon nanoparticles in SiO2 attribute defects at
oxidised clusters of silicon to luminescent emission observed at 2.42.8 eV
(Afanasev and Stesmans 1999). This is consistent with photoluminescence data
from etch-back experiments of buried oxide in silicon which attribute a 2.62.8 eV
PL emission to Si clusters found near the buried oxidesilicon interface
(Nishikawa et al. 1999). The 2.7 eV CL emission is therefore likely to be associated with oxidised silicon clusters located in strained buried oxide near the SiO2
Si interface. The 2.7 eV CL emission is also attenuated by electron irradiation but
both low and high dose (Stevens-Kalceff 2011) experiments show, that it is not as
irradiation-sensitive as the other CL emissions associated with Si clusters in the
buried oxide (i.e. 1.65, 2.15 and 2.35 eV emissions; see Fig. 11.6b).
The peak profile of the low intensity CL emission observed at *3 eV (415 nm) is
strongly influenced by reduced transmission of photons through the Si top layer (e.g.
see Fig. 11.2). Despite the correction of the SIMOX CL spectra over the wavelength
range 1.43.2 eV, for transmission through the silicon top layer, it is difficult to
accurately determine the energy, peak width or irradiation response of the *3 eV
(415 nm) CL emission due to uncertainties introduced by the steep drop off in photon
transmission through the silicon top layer at these energies. Insight into the possible
origins of the *3 eV CL is again provided by comparison with other investigations.
Photoionization experiments of Si particles in SiO2 attribute energy levels 3.1 eV
below the SiO2 conduction band to a hydrogen complexed oxygen vacancy in SiO2
(Afanasev and Stesmans 1999) while a 3.1 eV PL emission is attributed to excess
silicon near the SiO2Si substrate interface (Nishikawa et al. 1999). In addition,
inclusions of a dense coesite-like crystalline phase have also been reported near the
SiO2Si substrate interface (Afanasev et al. 1996b; 1997) which is a location of high
260
M. A. Stevens-Kalceff
residual compressive strain (Camassel et al. 2001). Natural coesite can form as a
result of high pressure. Coesite crystallites exhibit a broad CL emission at *3 eV
(Trukhin et al. 2003). Thus the *3 eV CL emission may be associated with excess
silicon defects and/or crystalline coesite inclusions near the SiO2 Si substrate
interface. The CL emission energies, peak widths and identifications from the buried
oxide layer are summarised in Table 11.2.
During low dose irradiation, the CL emissions associated with defects in the
buried oxide initially increase indicating an electron irradiation-induced increase
in defect concentrations, however as dose increases, all CL emissions associated
with the buried oxide layer are attenuated. This is in contrast to the (low and high
dose) situation in bulk a-SiO2 where generally the CL emission components
increase in intensity as a function of irradiation. This attenuation of all CL
emissions associated with the buried oxide layer corresponds to the electron beam
induced modification of the defects within the buried oxide and at the SiSiO2
interfaces rather than an annealing or reduction in the concentration of the defects.
At higher doses all CL emissions from the buried oxide attenuate and volume loss
occurs from the irradiated specimen, which is consistent with the breakdown of the
SiO2 structure in the buried oxide layer (Stevens-Kalceff 2011). In bulk, thin film
and buried silicon dioxide, charge trapping at existing and radiation induced
defects will result in an irradiation-induced electric field. Charge trapping in
irradiated buried oxides is enhanced in comparison with other forms of SiO2
(Paillet et al. 1995), due to the higher defect density in the strained buried oxide
layer and at the Si-SiO2 interfaces. The irradiation-induced localised electric field
is enhanced therefore contributing to the electron irradiation-induced breakdown
of the buried oxide layer.
11.9 Conclusions
A range of synthetic pure silicon dioxide polymorphs typically used in device
applications have been systematically investigated using cathodoluminescence
spectroscopy techniques. Bulk single crystal silicon dioxide (quartz), bulk Type
IIV amorphous silicon dioxide, dry amorphous thermal silicon dioxide thin films
on silicon (001) and buried strained amorphous silicon dioxide layer in silicon
(001) have been investigated (see Table 11.1) under the same low dose excitation
conditions. CL emissions associated with local SiO2 defects are fundamentally
broad due to strong electronphonon coupling. The influence of homogeneous
broadening may become significant in amorphous polymorphs of silicon dioxide.
A single broad multicomponent emission (FWHM 0.4 eV) at approximately
1.9 eV (650 nm) is observed from a-SiO2 and is attributed to the non-bridging
oxygen hole centre (NBOHC) with contribution from at least two precursors.
CL emissions from bulk Type IIV amorphous SiO2 and dry thermal amorphous SiO2 thin films are identified with a range of native defect centers associated
with point defects in the a-SiO2 tetrahedral structure. These characteristic CL
11
261
emissions are associated with oxygen deficient defects, (e.g. oxygen vacancies),
non bridging oxygen defects and self trapped excitons, however the relative
intensities, peak widths and/or irradiation kinetics differ between each bulk and
thin film a-SiO2 polymorph. The CL emissions from the bulk and thin film
amorphous SiO2 polymorphs include the non-bridging oxygen hole centre
(NBOHC) at 1.9 eV (650 nm); the radiative recombination of the self trapped
exciton (STE) at 2.2 eV (565 nm); and Oxygen Deficient Centers (ODC) at 2.7 eV
(460 nm) and *4.5 eV (275 nm). A CL emission at *3.4 eV (365 nm) is
observed from Type I and II a-SiO2 and is attributed to be charged compensated
substitutional Al3+:M+ defect (where M+ is typically Li+, Na+, K+ or H+). See
Table 11.2.
The CL emission from buried amorphous SiO2 is significantly different from
bulk and thin film amorphous SiO2 polymorphs. In general, the CL emissions from
the buried oxide layer cannot simply be associated with analogous amorphous
silicon dioxide native point defects. The Separation by IMplantation of OXygen
(SIMOX) fabrication process produces buried oxides confined within the silicon
wafer with residual compressive strain maximised at the two Si-SiO2 planar
interfaces. The concentration of defects is enhanced at surfaces and interfaces. The
formation of the native SiO2 defects which involve the relaxation of the surrounding oxide network is hindered in the confined strained buried oxide, while
the formation of silicon nano-clusters and crystalline coesite platelets is facilitated.
CL emission from the confined strained buried oxide is dominated by defects
associated with large surface to volume ratio of nanoscale silicon clusters and their
interfaces. Following suitable corrections for optical absorption, CL microanalysis
allows the defect microstructure of the buried oxide layer to be investigated in situ
without removal of the silicon top layer. A minimum of six visible CL emission
components from the in situ buried oxide are observed (see Table 11.2). The
1.65 eV (760 nm) emission is associated with passivated silicon nano-clusters and
SiO2-Si particle surface states. Similarly the 2.15 eV (590 nm) and 2.35 eV
(535 nm) emissions are associated with surface states of silicon nano-clusters near
the Siburied oxide interfaces. The 2.7 eV (460 nm) emission is associated with
the oxidised silicon nano-clusters. The peak profile of the nominally *3 eV
(415 nm) emission is distorted by reduced transmission through the silicon top
layer, preventing the peak width and energy from being more accurately determined. The *3 eV emission is associated with a silicon excess defect (e.g. H
complexed oxygen vacancy) and/or coesite crystalline platelets near the SiO2Si
substrate interface. The unresolved 1.9 eV CL emission is possibly associated with
the NBOHC (See Table 11.2).
The most significant physical processes contributing to the changes in the CL
spectra are defect generation via radiolysis and local modification due to highly
localised electric fields produced by charge trapping at defects within the SiO2.
The rate of generation/modification of electron irradiation-induced effects can be
reduced by decreasing the electron beam power density. Electron irradiationinduced dose dependent changes in the CL emission intensities as shown in
Figs. 11.4, 11.5, 11.6 are consistent with defect generation and transformation.
262
M. A. Stevens-Kalceff
References
Afanasev VV, Revesz AG, Hughes HL (1996a) Confinement phenomena in buried oxides of
SIMOX structures as affected by processing. J Electrochem Soc 143:695700
Afanasev VV, Stesmans A (1999) Photoionization of silicon particles in SiO2. Phys Rev B
Condens Matter 59:20252034
Afanasev VV, Stesmans A, Revesz AG, Hughes HL (1997) Structural inhomogeneity and silicon
enrichment of buried SiO2 layers formed by oxygen ion implantation in silicon. J Appl Phys
82:21842199
Afanasev VV, Stesmans A, Twigg ME (1996b) Epitaxial growth of SiO2 produced in silicon by
oxygen ion implantation. Phys Rev Lett 77:4206
Alonso PJ, Halliburton LE, Kohnke EE, Bossol RB (1983) X-ray induced luminescence in
crystalline SiO2. J Appl Phys 54:53695375
Bhat R, Dutta PS, Guha S (2008) Crystal growth and below-bandgap optical absorption studies in
InAs for non-linear optic applications. J Cryst Growth 310:19101916
Billeb A, Grieshaber W, Stocker D, Schubert EF (1997) Microcavity effects in GaN epitaxial
films and in Ag/GaN/sapphire structures. Appl Phys Lett 70:27902792
Camassel J, Falkovsky LA, Planes N (2001) Strain effect in silicon-on-insulator materials:
Investigation with optical phonons. Phys Rev B Condens Matter 63:1880
Cazaux J (1986) Some considerations on the electric field induced in insulators by electron
bombardment. J Appl Phys 59:14181430
Egerton RF, Li P, Malac M (2004) Radiation damage in the TEM and SEM. Micron 35:399409
Fauchet PM, Tsybeskov L, Zacharias M, Hirschman K (1998) Nanocrystalline silicon/amorphous
silicon dioxide superlattices. Mater Res Soc Symp Proc 485:4959
Godefroo S, Hayne M, Jivanescu M, Stesmans A, Zacharias M, Lebedev OI, van Tendeloo G,
Moshchalkov VV (2008) Classification and control of the origin of photoluminescence from
Si nanocrystals. Nat Nanotechnol 3:174178
Gorton NT, Walker G, Burley SD (1997) Experimental analysis of the composite blue
cathodoluminescence emission in quartz. J Lumin 72:669671
Gotze J, Kempe U (2008) A comparison of optical microscope- and scanning electron
microscope-based cathodoluminescence (CL) imaging and spectroscopy applied to geosciences. Mineral Mag 72:909924
Gotze J, Plotze M, Habermann D (2001) Origin, spectral characteristics and practical applications
of the cathodolumimescence (CL) of quartza review. Mineral Petrol 71:225250
11
263
264
M. A. Stevens-Kalceff
Chapter 12
Abstract Recent results on the cathodoluminescence (CL) and the trace element
composition of quartz are the starting point to review the properties of quartz from
different origin. CL-spectroscopy revealed five emission bands to be important in
quartz additionally to one at approx. 340 nm which has been reported in the
literature: the first one in the near-UV at 395 nm, the second in the blue range of
the spectrum at 450 nm, the third at 505 nm (greenish blue), the forth at 570 nm
(greenish yellow) and the last one in the red range of the spectrum at 650 nm. The
bands at 395 and 505 nm are characterised by a strong decrease of intensities
during irradiation while the band at 650 nm increases with increasing dose. This
phenomenon is very common in quartz grown from aqueous solutions while
magmatic quartz may show more stable luminescence emission. Trace element
analyses display also differences in the composition between these two groups of
quartz. Aluminium, Li and H have been found to be most important in authigenic,
hydrothermal and metamorphic quartz but magmatic quartz is generally enriched
in Ti. Germanium, Fe, B and Na is present at low levels in all quartz samples.
A strong linear correlation between Al and Li indicates combined incorporation in
[AlO4|Li+]-defects. A high unstable intensity at 395 nm has been observed especially in Al-rich quartz. In these samples, the luminescence commonly attenuates
completely. However, different quartz samples show different correlation with Al.
This result puts doubt on the interpretation that Al-related centres are the only
reason for the near-UV emission. The emission band at 505 nm which also shows
T. Gtte (&)
Institute of Geosciences, Goethe-University Frankfurt,
Altenhferallee 1, 60438 Frankfurt, Germany
e-mail: goette@em.uni-frankfurt.de
K. Ramseyer
Institute of Geological Sciences, University Bern, Baltzerstr. 1+3,
3012 Bern, Switzerland
265
266
12.1 Introduction
Quartz is one of the most important and most abundant minerals in the Earths
crust and forms in a great variety of physicochemical environments in continental
rocks (Fig. 12.1). Hence, it has been a matter of great interest to Earth scientists,
but also to physicists and chemists. So the properties of quartz have been dealt
with in numerous studies in different research fields. Although the luminescence
has also been intensively investigated and used for various purposes (e.g., visualisation of growth fabrics, age dating), many details on the defect structure and
luminescence centres in quartz are poorly understood.
The emission of light after excitation of solid material by means of high-energy
electrons is termed cathodoluminescence (CL). In quartz, the CL ranges from near
ultra-violet light (350 nm) to near infrared light (750 nm). The luminescence
properties of quartz are related to the physicochemical conditions at the time of
crystal growth and the history thereafter (Gtze 2001). In general, the defect
structure (and also the luminescence properties) of quartz precipitating from an
aqueous solution will differ from that of quartz crystallized from a silicate melt or a
supercritical fluid. Moreover, quartz which experienced elevated temperatures after
crystallization (e.g., metamorphosis between 200 and 800C for several tens of
millions of years or [1014 s) probably has different defect structure than unannealed quartz with a more imperfect lattice. Therefore, the luminescence properties may provide useful information about the formation and subsequent history
of quartz. It is worthwhile mentioning that irradiation of quartz with high energy
particles (e.g., electrons, protons, ions) has an important effect on point defects such
as [AlO4-|M+]0 or some intrinsic defects. The excitation of luminescence by particles changes the type and concentration of point defects in quartz. This effect is
well shown by EPR taken from quartz as collected in nature compared with those
after annealing or irradiation (e.g., Botis et al. 2005; Gtze et al. 2005).
Commonly, quartz is nearly pure SiO2 and for that reason little attention has
been paid to its geochemical investigation for a long time. However, recent
improvements of analytical techniques with high spatial resolution and low limits
of detection have stimulated numerous studies on the trace element geochemistry
of quartz (Larsen et al. 2004; Landtwing and Pettke 2005; Gtze et al. 2004;
Larsen et al. 2009; Mller and Koch-Mller 2009; Lehmann et al. 2011;
12
267
Fig. 12.1 Phase diagram of quartz (a- and b-quartz) with different characteristic environments of
quartz formation. The parental phases which are present in the Earths crust are shown in outlined
text
Gtte et al. 2011). The focus has been on magmatic quartz which is commonly
enriched in Ti (Larsen et al. 2004; Mller 2000; Gtze et al. 2004) and on
hydrothermal quartz providing higher Al-, Li-, Na- and K-concentrations
(Bambauer et al. 1961; Monecke et al. 2002; Landtwing and Pettke 2005; Gtte
et al. 2011). But systematic variation in Ge and P-concentrations has also been
reported (Larsen et al. 2004, 2008). Different mechanisms influence the trace
element composition. Elements that are incompatible to most magmatic minerals
such as Ge, P or B are enriched in the melt at the beginning of crystallisation and
can be incorporated in quartz formed from a highly differentiated magma (Larsen
et al. 2004). Quartz grown from aqueous solutions is more variable in its trace
element composition and can be enriched in Al, Li and H, while Ti is commonly
absent below temperatures of about 400C. The trace element incorporation also
influences the real structure and vice versa, because impurities provide additional
electrons or electron traps.
This study aims at reviewing the spectral characteristics of the luminescence
emission of quartz and to compare it to the trace-element composition and to
intrinsic point defects. Many results reviewed in this paper are published elsewhere (Mller 2000; Monecke et al. 2002; Larsen et al. 2004; Gtze et al. 2004;
Landtwing and Pettke 2005; Lehmann et al. 2011; Gtte et al. 2011), but some
new data are included in this article. A complete list of the samples to which this
work refers is given in Table 12.1. A set of seven crystalline samples from New
Zealand (five metamorphic rocks and two plutonic rocks, Preusser et al. 2006)
have been investigated in detail by trace element analysis (LA-ICP-MS) and
cathodoluminescence (SEM-CL). A second set of 14 magmatic rocks was
investigated in an optical CL-microscope, and trace elements in these samples
268
Table 12.1 List of samples which where analysed for this study
Sample
Lithology
Location
Authigenic quartz
Bo 14
WrgI 17
Reference
Ruhr-sandstone
Sollingsandstone
Sollingsandstone
Bochum (W-Germany)
Wrgassen (Germany)
Rock crystal
Rock crystal
Gigerwald (Switzerland)
Gtte et al. (2011)
Rohdenhaus (W Germany) Gtte et al. (2011)
Biotite gneiss
Augengneiss
Greenshist
Quartz
segregation
Quartz tectonite
New
New
New
New
Saar-Nahe (Germany)
Saxony (Germany)
Rosia Montana (Romania)
Saar-Nahe (Germany)
Steiermark (Austria)
Tumasvlakte (Namibia)
Gtte
Gtte
Gtte
Gtte
Gtte
Gtte
Gt29
Gt30
Gt31
Rhyolite
Rhyolite
Rhyolite
Dacite
Quartz trachyte
Two-micagranite
Granite
Granite
Granite
Gt32
Gt33
Gt34
Gt35
Gt36
Gt37
Gt38
Gt40
Granite
Granite
Tonalite
Granodiorite
Quartz diorite
Diorite
Granodiorite
Rhyolite
Kh
Hydrothermal
quartz
Gig 1b
Roh 2
Metamorphic
quartz
FJ1
FJ2
OK2
TAR1
TAR2
Magmatic quartz
Gt22
Gt23
Gt24
Gt26
Gt27
Gt28
Karlshafen (Germany)
Zealand
Zealand
Zealand
Zealand
Alps
Alps
Alps
Alps
Preusser
Preusser
Preusser
Preusser
Gtte
Gtte
Gtte
Gtte
Gtte
Gtte
Gtte
Gtte
et
et
et
et
and
and
and
and
and
and
and
and
and
and
and
and
and
and
al.
al.
al.
al.
(2006)
(2006)
(2006)
(2006)
Richter
Richter
Richter
Richter
Richter
Richter
Richter
Richter
Richter
Richter
Richter
Richter
Richter
Richter
(2006)
(2006)
(2006)
(2006)
(2006)
(2006)
(2006)
(2006)
(2006)
(2006)
(2006)
(2006)
(2006)
(2006)
were obtained by proton micro-probe. In the following sections, the cathodoluminescence properties and the trace-element composition will be briefly discussed, before the correlation of distinct emission bands with the defect structure
of quartz will be reviewed.
12
269
12.2 Methods
12.2.1 CL-Microscopy and -Spectroscopy
The SEM-CL-system GATAN CL3 mounted to an EVO 50 Zeiss SEM at the
Institute of Geological Sciences at the University of Bern has been used for the
CL-spectroscopy of hydrothermal, authigenic and metamorphic quartz. The SEM
is equipped with a grating of 150 lines/mm and a PIXIS CCD-detector for the need
to detect simultaneously the entire spectral range comprising the visible, near-UV
and near-IR light (i.e., 320860 nm). The wavelength calibration was done with a
Hg-lamp. A Faraday cup for accurate measurements of the beam current has been
installed. An acceleration voltage of 14 kV and a beam current of about 5 nA have
been used for the CL-spectroscopy and twenty spectra each with an exposure time
of 10 s were recorded. The electron beam was defocussed to avoid rapid
destruction of the crystal lattice. The resulting spot size was about 95 lm2
resulting in a dose rate of 0.7 lA mm-2.
The optical CL-microscope HC1-LM at the Institute for Geology, Mineralogy
and Geophysics of the Ruhr-University Bochum has been used for the CL-analyses
of 14 magmatic samples (for technical details cf. Neuser et al. 1996). The
microscope is equipped with a grating spectrograph of EG&G Princeton Research
Instruments comprising a grid with 150 lines/mm and a CCD-camera for recording
of luminescence spectra. The spatial resolution of the spectroscope is approx.
30 lm and an acceleration voltage of 14 kV, exposure time of 20 s per spectrum
and a beam current density of 5 lA mm-2 were used for the analyses. A sequence
of 10 spectra was recorded with this instrument.
A computer software was implemented to analyse the CL-spectra and nonlinear least square fitting of Gaussian, Lorentzian and Voigtian curves to the
recorded data. The procedure ensures optimal deconvolution of overlapping
emission bands in quartz spectra. All spectra were converted into the energy-space
for fitting but spectra are presented in the wavelength-space in the figures because
this is more common in geosciences.
270
12
271
Fig. 12.2 CL-spectra of various magmatic quartz samples (black: initial spectrum, grey: final
spectrum after 200 s), a Quartz trachyte (Gt 27, Steiermark, Austria), b Rhyolite (Gt 23: Saxony,
Germany), c Granite (Gt 29: kl. Spitzkoppe, Namibia), d Granite (Gt 33: Asilomar State Park,
USA)
for increasing emission, respectively, with the measured intensity I, the modelled
unstable part Iinst, the stable part Istbl, the decay constant of defects j and k,
respectively, the dose rate R (in lA mm-2) and the irradiation time t (Fig. 12.4,
Gtte et al. 2011). The decay-constants j and k depend on the technical
equipment and need to be determined for each experimental set-up. For the
combination of an EVO-Zeiss SEM and a Gatan CL-system, Gtte et al. (2011)
determined values of 31 12 mm2 lA-1 s-1 and 34 13 mm2 lA-1 s-1 for the
395 nm-band and the 505 nm-band, respectively, and 4.5 3 mm2 lA-1 s-1 for
the 650 nm band. The slight difference in the former bands is within the range of
error, but the red band increases with a significantly slower rate. This indicates that
there is no common mechanism for both phenomena as suggested by King et al.
(2010). The emission band at 570 nm has been found to be dominant only in the
hydrothermal quartz from Rohdenhaus which shows fabrics of fast crystallisation
(Gtte 2004; Gtte et al. 2011).
272
Fig. 12.3 CL-spectra of quartz formed from aqueous solutions samples (black: initial spectrum,
grey: final spectrum after 200 s), a quartz cement in the Finefrau-Sandstone (Upper
Carboniferous, W-Germany), b Al-rich quartz cement in the Finefrau-Sandstone (Upper
Carboniferous, W-Germany), c quartz cement in the Solling-Sandstone (Lower Triassic,
Germany), d metamorphic quartz (Alps, New Zealand)
Authigenic quartza
Finefrau Z2
Finefrau Z3
Solling Z2
Solling Z3
Karlshafen Z2
Hydrothermal quartza
Gig1b 1
Gig1b 2
Gig1b 3
Roh2 1
Roh2 2
Roh2 3
Roh2 4
Roh2 5
Metamorphic quartza
FJ1
FJ2
OK2
Tar1
Tar2
Magmatic quartzb
Gt22
Gt23
Gt24
Gt26
Gt27
Gt28
Gt29
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
1.8
0.9
5.8
7.2
7.9
21
23
21
30
21
(0.3)
(0.2)
(4.1)
(5.7)
(1)
1.9 (1.1)
1.2 (0.3)
1.2 (0.2)
9 (3)
n.d.
40 (5)
13 (2.7)
9.5 (1.5)
650 (170)
82 (8)
94 (3)
360 (180)
27 (37)
3,000 (250)
1,770 (490)
270 (90)
(4)
(2)
(1)
(14)
(16)
6.3 (1.1)
4.7
n.d.
n.d.
20
92 (76)
430 (170)
17 (10)
83 (23)
36 (27)
(24)
(95)
(10)
(7)
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
3.6
6.2
2.6
3.4
7.9
(5.7)
(12.7)
(0.9)
(3.8)
(7.6)
2.5 (1.7)
4.5 (0.7)
2.5 (1)
35 (14)
2.7 (3.8)
126 (65)
75 (22)
38 (10)
16
19
29
50
58
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
69 (73)
270 (560)
43 (2.6)
160 (140)
57 (38)
1,590 (510)
140 (19)
214 (9)
520 (245)
39 (49)
6,800 (250)
2,200 (540)
420 (120)
750 (430)
2,100 (370)
100 (29)
300 (35)
160 (125)
Al
lmol/mol
Na
Li
(0.9)
(0.5)
(1)
(0.6)
(1)
(4)
(22)
(55)
(16)
(7)
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
13
11
21
23
12
95
46
34
34
37
64
34
34
23 (4)
28 (7)
17 (3)
17 (4)
25(5)
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
530
n.d.
n.d.
n.d.
n.d.
n.d.
38
n.d.
n.d.
51
93 (59)
83 (32)
n.d.
410 (220)
n.d.
33 (20)
128
n.d.
66
54
56
69
23
n.d.
5.1 (9.8)
2.4 (3.1)
2.3 (0.4)
0.7 (0.9)
0.23 (0.1)
0.7 (0.4)
0.5 (0.2)
0.7 (0.3)
n.d.
n.d.
45
n.d.
n.d.
57 (50)
1.4 (0.5)
8.5
8.5
23 (13)
Ti
n.d.
34
n.d.
n.d.
34
30
n.d.
190
14 (27)
n.d.
19 (14)
4 (1.7)
5 (7)
2 (1)
2 (1)
n.d.
n.d.
117
n.d.
n.d.
20 (13)
65 (27)
25
8.4
600 (410)
Fe
(0.6)
(0.6)
(0.1)
(0.2)
(0.2)
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
0.4
0.4
0.2
0.2
0.7
(0.1)
(0.4)
(0.05)
(0.1)
(0.1)
4.8 (0.9)
1.2 (0.1)
1.2 (0.1)
3.4 (0.9)
1.8 (0.1)
15 (10)
4.5 (0.7)
4.3 (0.1)
1.4
1.9
0.7
0.7
1.1
Ge
0.23
0.19
0.09
0.12
0.21
0.05
0.12
1.9 (0.4)
5.5 (5.5)
3 (1.6)
10 (7.3)
0.6 (0.1)
122 (29)
9.5 (0.3)
37 (5.9)
6.9
2 (0.8)
26 (7.5)
16
3.4 (0.6)
5.7
65 (38)
7.2 (8.1)
n.d.
n.d.
390 nm
Iinst
lmol/mol lmol/mol lmol/mol lmol/mol lmol/mol cts/nA/s
(0.5)
(0.9)
(1.5)
(0.8)
(0.1)
-0.01
0.25
0.14
-0.01
0.13
0.07
-0.11
1.8
3.8
5.5
3.4
0.4
12 (2.4)
2.6 (0.2)
7.4 (0.9)
6.6
4.8 (1.1)
4.1 (0.6)
17
10 (0.2)
2.6
6.5 (2.3)
5.4 (3.1)
n.d.
n.d.
Istbl
cts/nA/s
0.3
0.23
0.05
0.2
0.34
0.14
0.08
5.3 (2.3)
15 (13)
2.6 (2.1)
18 (12)
0.4 (0.3)
30 (6.8)
26 (4.8)
51 (15)
15
3.4 (0.2)
3.5
28
n.d.
0.5
13 (8)
5.9 (3.5)
n.d.
n.d.
505 nm
Iinst
cts/nA/s
(0.8)
(1.6)
(0.8)
(1.1)
(0.2)
-0.03
0.29
0.19
0.14
0.18
0.06
0.22
2.6
5.1
4.4
6.6
0.6
3.9 (0.8)
4.3 (0.2)
9.9 (0.9)
9.4
5.7 (2.4)
2.3
n.d.
n.d.
1.1
2.1 (0.6)
3.6 (2)
n.d.
n.d.
Istbl
cts/nA/s
0.11
0.64
0.79
0.2
0.29
0.1
0.13
0.27
0.06
0.02
0.28
0.22
0.21
0.22
1.4(0.01)
8.7 (13)
37 (32)
3.5 (1.3)
0.8 (0.2)
3.2
3.2 (1)
11 (0.7)
13
8.9 (3.3)
n.d.
n.d.
13 (0.6)
9
0 (4.2)
4.3 (0.9)
4.7
n.d.
Istbl
cts/nA/s
2.5 (0.4)
10 (16)
9.4 (3.4)
7.2 (1.5)
1.6 (0.4)
1.7
6.8 (0.4)
12 (5.4)
8.6
9.6 (4.6)
0.9 (0.5)
2.7
26 (3.8)
15
1.7 (6.5)
4.9 (1)
8.5
n.d.
650 nm
Iinst
cts/nA/s
Table 12.2 Results of trace element analyses and CL-spectroscopy for the samples from Table 12.1. Standard deviations are given in brackets. Please note
that Al and Li cannot be analysed with the proton microprobe
12
273
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
29
33
n.d.
33
77
92
52
55
n.d.
Ti
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
18
n.d.
Fe
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
Ge
0.13
0.21
0.03
0.07
0.15
0.14
0.25
0.15
0.09
390 nm
Iinst
lmol/mol lmol/mol lmol/mol lmol/mol lmol/mol cts/nA/s
Gt30
Gt31
Gt32
Gt33
Gt34
Gt35
Gt37
Gt38
Gt40
Al
lmol/mol
Na
Li
-0.03
0.03
0.14
0.08
-0.06
-0.11
0.17
0.03
0.03
Istbl
cts/nA/s
0.16
0.34
0.2
0.17
0.23
0.12
0.35
0.23
0.06
505 nm
Iinst
cts/nA/s
0.11
0.23
0.23
0.19
-0.08
-0.08
0.3
0.08
0.11
Istbl
cts/nA/s
0.12
0.12
0.19
0.13
0.16
0.07
0.27
0.1
0.13
650 nm
Iinst
cts/nA/s
0.26
0.3
0.29
0.27
0.14
0.12
0.24
0.15
0.72
Istbl
cts/nA/s
274
T. Gtte and K. Ramseyer
12
275
Fig. 12.4 Decreasing luminescence in the near-UV (a, b) and greenish blue (c, d) range of the
spectrum and increasing luminescence at 650 nm in hydrothermal quartz form Rohdenhaus
(a, b) and Gigerwald (c, d, modified from Gtte et al. 2011)
is not completely satisfied in most samples if only Li and Na are considered. This
indicates the presence of significant amounts of AlOH-defects even in magmatic
quartz. Li/Al-ratios are high (0.50.8) in hydrothermal and pegmatitic quartz, but
are significantly lower (0.10.5) in magmatic, metamorphic and authigenic quartz.
Boron is commonly low in quartz with maximum values between 5 and
25 lmol mol-1. Hydrothermal, metamorphic and pegmatitic quartz are slightly
enriched in B, which shows no correlation to Al. Tourmaline, the most important
B-mineral in magmatites, is a common constituent of many granitic pegmatites
underlining the incompatible character of B. It is therefore enriched in pegmatitic
rocks compared to plutonic or volcanic quartz. Enhanced concentrations of B can
also be found in agates (Gtze, pers. com.). The concentration of Fe in quartz
shows non-systematic variation and is assumed to be affected by submicroscopic
inclusions. Otherwise, contents as high as 190 lmol mol-1 (in sample FJ1) cannot
be explained. However, it is probably present on single lmol mol-1 level in many
magmatic or metamorphic quartz samples. Its occurrence in hydrothermal quartz
probably depends on the regional geological setting.
Ti-concentration is generally below 50 lmol mol-1 and is highest in the plutonic and volcanic quartz. Trace contents of Ti have been found in hydrothermal
quartz formed in Alpine fissures at high temperature (Gig1b, B350C) but it is
generally absent in hydrothermal quartz, which formed at significantly lower
temperatures (Roh2, B130C, Table 12.2). Titanium detected in authigenic quartz
cement is interpreted to indicate submicroscopic mineral inclusions or contamination with detrital quartz during the analysis (cf. Lehmann et al. 2011). It has long
been known, that Ti is a common impurity in magmatic quartz but might form
exsolutions of rutile during cooling. Recently, it has been found to correlate
strongly with temperature (Wark and Watson 2006).
276
Fig. 12.5 Correlation of monovalent cations (Li ? Na ? K) and Al in different quartz (solid line:
1:1-correlation). Authigenic quartz: unpublished data from Gtte (open circles, grey circles) and
data from Lehmann et al. (2011, black dots); hydrothermal quartz: data from Gtte et al. (2011, grey
and white circles) and Landtwing and Pettke (2005, black circles); metamorphic quartz: this study
(white circles) and data from Monecke et al. (2002, black circles); pegmatitic quartz: from Gtze
et al. (2005); plutonic quartz: from Larsen et al. (2004); volcanic quartz: from Mller (2000)
12
277
278
Fig. 12.6 Linear correlation between the unstable intensity of the 390 nm-band and the Alconcentration (left) and the Li-concentration (right, solid lines: linear regression). Data for hydrothermal quartz from Gtte et al. (2011), data of authigenic and metamorphic quartz in Table 12.2
defect centre in quartz is the luminescence centre. But since this emission band is
lacking in many quartz samples with high Al and Li concentrations this interpretation is highly questionable. In contrast, Botis et al. (2005) observed a 350 nm
(3.5 eV) emission in naturally irradiated quartz from uranium deposits and
assumed that a silicon vacancy-hole centre is responsible for this band because
the emission band and the centre disappear after annealing the quartz for 24 h at
550600C. However, it should be noted that these annealing conditions are
similar to those of Al-centres compensated by single charged cations (Ramseyer
and Mullis 1990).
Alonso et al. (1983) firstly proposed that the emission at 380 nm (3.1 eV) is
attributed to an Al-defect in quartz and, indeed, the unstable part Iinst of this band
displays a good correlation with the Al-concentration in many cases (Fig. 12.6).
Aluminium is incorporated in [AlO4|M+]-defects in quartz (Rossman 1994) that
are destroyed by ionizing irradiation and form [AlO4]0- and M0-centres, which are
no luminescence centres. This model has been the preferred interpretation for
short-lived emission in hydrothermal and pegmatitic quartz up to now (Perny et al.
1992; Gtze et al. 2005; King et al. 2010). Perny et al. (1992) found short-lived
emission (i.e., substantial decrease of the intensity during continuous electron
irradiation) in hydrothermal quartz from the Swiss Alps containing more than
50 ppm Al, but without being directly correlated to the Al-concentration. Similarly, Gtte et al. (2011) found different correlation for different quartz samples
and proposed that only Li-compensated Al-centres may contribute to the shortlived luminescence. However, the uncertainty of the correlation is too high and the
discrepancies in published data are too large to accept this correlation as a prove
for the physical relationship of the [AlO4|Li+]-defect and the 380 nm-band. Luff
and Townsend (1990) and Gorton et al. (1997) reported a significant emission at
380/390 nm in ultra-pure synthetic quartz. Both authors excluded the relation to an
Al-defect but the observed growth related intensity change is a clear evidence of a
defect incorporated during growth according to Luff and Townsend (1990).
12
279
In essence, the CL emission in the near UV and specifically the short-lived type
at 390 nm, shows clear relations with [AlO4|M+]0 defects (for more details see
Ramseyer and Mullis 1990; Gtze et al. 2005; Gtte et al. 2011), but other defect
centres may also emit in this spectral range. In the case of TL, McKeever et al.
(1985) and Yang and McKeever (1990) suggest that the emission at 380 nm for the
110C TL peak is correlated with the release of electrons from the [GeO4]-centres and recombination with holes trapped at [AlO4]0 and [H3O4]0-centres.
280
(160 K) and only plays a minor role at room or elevated temperatures (Itoh et al.
1988). A broad emission band with several side-bands has been observed in luminescence spectra of silica nano-particles. This band has been assigned to SiH-groups
at the surface (Glinka et al. 1999). A convincing correlation of these emission bands
with a distinct defect in the bulk crystal has not been found up to now and some
further research is required in this case. However, it should be noticed that different
defects might contribute to the luminescence at approx. 500 nm.
12
281
Fig. 12.7 Increasing intensity at 650 nm with increasing dose in unaltered (P1) and hydrothermally altered synthetic quartz (P2, black: initial, grey: final). Data from Dersch (2001)
282
2000). Kuzuu et al. (1993) assigned the emission at 650 nm and a related
absorption band at 260 nm (4.8 eV) in silica glass to defects which are related to
oxygen excess (free O2-molecules as well as NBOHC). Recently, Botis et al.
(2005) argued that the high thermal stability (up to the a-quartztridymite
transformation temperature of 867C) of the centre emitting at 650 nm is unrelated
with the presence of NBOHCs as the latter centre is only stable below 500C
whereas the emission band has been detected even in quartz that was heated to
600C. But however, this is not necessarily contradictory because NBOHC might
be generated from peroxy-defects which are still present at this temperature during
the measurement.
A strong relationship between the luminescence at 650 nm and irradiation by
a-particles has repeatedly been reported. In irradiation experiments, yellowish to
reddish luminescing zones or halos are produced which show increased intensities
at 650 nm (Komuro et al. 2002; Krickl et al. 2008). In naturally irradiated samples,
quartz crystals show rims of yellowish to brown luminescence adjacent to the
radioactive material (e.g., Gtze 2009). However, experiments with proton beams
revealed a more complex interaction with the defect structure, because other bands
(e.g., at 580 nm) are also affected.
An additional band was reported in some cases in the near infrared at 710 nm
(1.74 eV, Kempe et al. 1999). It is likely caused by the incorporation of Fe3+.
In essence, CL is a valuable tool to decipher the origin and thermal history of
quartz. Furthermore, some of the luminescence centres are partly related to trace
elements commonly incorporated in natural quartz (i.e., Al, Li, H, Fe, etc.).
12.5 Conclusions
Cathodoluminescence investigation and trace element analyses revealed very
complex properties of natural quartz. Obviously, the presence of impurities in the
quartz lattice affectsat least partlythe luminescence emission. These interactions lead to a large variety of different luminescence colours in different quartz
samples. The most important results can be summarised as follows:
Luminescence spectra of quartz are composed of emission bands at 395, 450,
505, 570 and 650 nm. All emission bands have high width at half maximum
resulting in a strong overlap of distinct bands. Deconvolution therefore is the
first step for a reliable interpretation.
The emission bands at 395, 505 and 650 nm are unstable and show decreasing
intensity (395, 505 nm) or increasing intensity (650 nm) with ongoing electron
irradiation.
Impurity-related point defects are common in quartz, although the trace element
content is generally low. Aluminium, Li and H are the most prominent elements
incorporated in quartz which grew at low temperatures from an aqueous
12
283
Acknowledgments We wish to thank Thomas Pettke (Bern) and Jan Meijer (Bochum) for
technical support with the LA-ICP-MS and the proton microprobe, respectively. Frank Preusser
(Bern) kindly provided crystalline quartz samples from New Zealand. We gratefully acknowledge
the financial support from the German Research foundation (DFG, Go 1089/3-1). We also thank
an anonymous reviewer, Jens Gtze (Freiberg), and Robert Mckel (Freiberg) for their suggestions on an earlier draft of the manuscript.
References
Alonso PJ, Halliburton LE, Kohnke EE, Bossoli RB (1983) X-ray-induced luminescence in
crystalline SiO2. J Appl Phys 54(9):53695375
Bambauer HU (1961) Spurenelemente und c-Farbzentren in Quarzen aus Zerrklften der
Schweizer Alpen. Schweiz Miner Petrogr Mitt 41:335369
Bernstein LR (1985) Germanium geochemistry and mineralogy. Geochim Cosmochim Acta
49:24092422
Botis S, Nokhirn SM, Pan A, Xu Y, Bonli Th, Sopuck V (2005) Natural radiation-induced
damage in quartz. I. correlations between cathodoluminescence colors and paramagnetic
defects. Can Mineral 43:15651580
Demars C, Pagel M, Deloule E, Blanc P (1996) Cathodoluminescence of quartz from sandstones:
Interpretation of the UV range by determination of trace element distributions and fluidinclusion P-T-X properties in authigenic quartz. Am Mineral 81(78):891901
Dersch O (2001) Wasseraufnahme von Quarz: Grundlage fr eine Methode zur Datierung
archologischer Quarzartefakte. Ph.D. Thesis Goethe-University Frankfurt, 250 S, Frankfurt
am Main, Germany
Glinka YD, Lin S-H, Chen Y-T (1999) The photoluminescence from hydrogen-related species in
composites of SiO2 nanoparticles. Appl Phys Lett 75(6):778780
Glinka YD, Lin S-H, Hwang LP, Chen YT (2000) Photoluminescence from mesoporous silica:
similarity of properties to porous silicon. Appl Phys Lett 77(24):39683970
Gorton NT, Walker G, Burley SD (1997) Experimental analysis of the composite blue
cathodoluminescence emission in quartz. J Lumin 724:669671
Gtte T (2004) Petrographische und geochemische Untersuchungen zu den postvariszischen
Mineralisationen im devonischen Massenkalk des nordwestlichen rechtsrheinischen Schiefergebirges unter besonderer Bercksichtigung der Kathodolumineszenz, Ph.D. Thesis,
University of Bochum, Bochum, Germany, pp 186
Gtte Th, Richter DK (2003) Late Palaeozoic and Early Mesozoic hydrothermal events in the
northern Rhenish Massif: results from fluid inclusion analyses and cathodoluminescence
investigations. J Geochem Explor 7879:531535
Gtte Th, Richter DK (2006) Cathodoluminescence characterization of quartz particles in mature
arenites. Sedimentology 53(6):13471359
284
12
285
Chapter 13
Mineralogy, Geochemistry
and Cathodoluminescence of Authigenic
Quartz from Different Sedimentary Rocks
Jens Gtze
Abstract Authigenic quartz is present in different sedimentary rocks of NorthEastern Germany. Single crystals of euhedral quartz were detected in the Permian
(Zechstein) salt deposit of Roleben, in quartz nodules within Triassic sandstone
layers (Chirotherien sandstone, Bunter) from Jena, and Tertiary lignite deposits in
the Leipzig region (Zwenkau, Cospuden). Mineralogical and geochemical investigations revealed that the authigenic quartz crystals from the different geological
units differ in morphology (habit), characteristic inclusions, trace-element geochemistry and cathodoluminescence properties. Accordingly, the results allow not
only to clearly distinguish between authigenic and detrital quartz, but also between
authigenic quartz from different sedimentary environments. Authigenic quartz
from Zechstein salt deposits shows characteristic euhedral forms dominated by
rhombohedral faces or a combination of rhombohedral and prism faces, and
mineral inclusions (halite or anhydrite) in dependence on the saliniferous facies.
The crystals exhibit a blue luminescence, which can be related to a broad emission
band at 450 nm. The authigenic quartz crystals from the Bunter sandstone
are often intergrown, forming aggregates of several mm up to cm in size. At least
three growth zones can be distinguished: spherulithic growth starting from calcite
inclusions, quartz with complex internal CL structure, and a homogeneous outer
zone with no visible luminescence. The second zone exhibits a cathodoluminescence pattern similar to that of agate with three emission bands at 650, 580 and
450 nm. Authigenic quartz from Tertiary lignites is characterized by doubly terminated crystals with prism and rhombohedral faces. Intergrowth of two or more
crystals was observed. The CL is dominated by a transient emission band at
650 nm, which increases in intensity during electron irradiation. The crustal
J. Gtze (&)
Institute of Mineralogy, TU Bergakademie Freiberg,
Brennhausgasse 14, 09596 Freiberg, Germany
e-mail: goetze@mineral.tu-freiberg.de
287
288
J. Gtze
signature of all quartz REE distribution patterns and high contents of Al and Fe
indicate the origin of the silica-bearing fluids from weathering solutions and do not
show any influence of hydrothermal fluids. On the other hand, elevated concentrations of Na, K, Mg, Ca, and B can probably be related to the influence of
saliniferous fluids during quartz precipitation. Although the specific physicochemical conditions may have been different for the various occurrences, the data
suggest a formation of the authigenic quartz crystals during early diagenesis.
13.1 Introduction
Quartz is one of the most important minerals in the earths crust occurring in large
amounts in sedimentary rocks. Macrocrystalline quartz, chalcedony as well as
poorly and non-crystalline silica (opal-A, opal-CT) play a central role in the
composition and diagenesis of sediments. Therefore, investigations of quartz are
extensively used for the evaluation of provenance in arenites (e.g., Zinkernagel
1978; Zuffa 1985; Seyedolali et al. 1997; Bahlburg and Floyd 1999; Gtze and
Zimmerle 2000; Boggs et al. 2002; Richter et al. 2003; Bernet and Basset 2005).
On the other hand, secondary quartz neophormism in form of quartz overgrowth
cements, fracture fillings and silicification of fossil remains is a common feature in
sedimentary and diagenetic environments.
In particular authigenic quartz cement plays a major role in controlling the
hydraulic properties and quality of reservoir sandstones (e.g., McBride 1989;
Bjrlykke and Egeberg 1993; Wordan and Morad 2000). The application of
cathodoluminescence (CL) to sandstone petrology by Sippel (1968) and Zinkernagel
(1978) first revealed the spectacular difference between detrital quartz and authigenic quartz cements. More recent studies showed that the amount of diagenetic
quartz may be quantified by image analysis (Evans et al. 1994). The combination of
CL with other analytical methods can provide important information about polyphase quartz cementation and the burial history of sandstones (e.g., Sippel 1968;
Houseknecht 1991; Hartmann et al. 2000). For instance, CL microscopy with fluid
inclusion studies is ideal to correlate quartz cements with the formation of other
authigenic minerals or dissolution events, and to gain information concerning temperature and salinity of fluids during precipitation (e.g., Burley et al. 1989;
Walderhaug 1990, 1994). On the other hand, timing of cementation and evaluation of
temperature and precipitation mechanisms is supported by high precision in situ d18O
analysis (e.g., Lyon et al. 2000; Hiatt et al. 2007; Kelly et al. 2007). Authigenic quartz
plays also an important role in more intensively tectonized sandstones with crackseal structures during processes of brittle deformation and healing (e.g., Milliken
and Laubach 2000; Markowitz and Milliken 2003).
In contrast to the widespread occurrence and significance of ordinary authigenic
quartz overgrowths, isolated euhedral quartz crystals seem to be a relatively
uncommon form of diagenetic silica formation in sediments. Authigenic quartz
13
289
formation is described in different geological formations and from various sedimentary environments. A few publications document the occurrence of authigenic
quartz crystals in soil (Dixon and Weed 1989), limestones and carbonates
(Black 1949; Richter 1971; Molenaar and deJong 1987; Miik 1995; Chavetz and
Zhang 1998; Liu et al. 2004; Evans and Elmore 2006), sulphate rocks (Richter 1971;
Friedman and Shukla 1980), salt deposits (Grimm 1962; Nachsel 1969; Sedletskyi
1971; Fabricius 1987; Fruth and Blankenburg 1992), bituminous coal and lignite
(Baker 1946; Hoehne 1954; Leskevich 1959; Ruppert et al. 1985; Soong and Blattner
1986; Botz et al. 1986; Fruth and Blankenburg 1992) or oil deposits and bitumen
veins (Fchtbauer 1961; Parnell et al. 1996). In most of these investigations,
inclusions and/or isotope data have been used for the reconstruction of diagenetic
conditions.
Considering these results, the present study focused on the comparison of
authigenic quartz crystals from contrasting sedimentary environments. Therefore,
the investigation includes authigenic quartz from Zechstein salt deposits, nodular
quartz from Triassic sandstones, and single authigenic quartz crystals from Tertiary lignite seams. The different sedimentary quartz occurrences were analyzed
and compared concerning their characteristic mineralogical and geochemical
properties, and the conditions of formation.
290
J. Gtze
Fig. 13.1 Location of investigated quartz samples in different sedimentary units of North-eastern
Germany and schematic profiles of the relevant sediment units (stratigraphic classification
according to Heynke and Znker 1970; Langbein 1974; Bellmann 1986). 1 Zechstein salt,
Roleben, 2 Bunter sandstone, Jena, 3 Tertiary lignites, Zwenkau and Cospuden
silicified lenses and roots occur, which consist of small euhedral quartz crystals
(Bellmann 1986). These crystals are also distributed within the surrounding coal.
After careful crushing, washing and sieving of the quartz containing lignite samples, the authigenic quartz crystals were hand picked under a binocular microscope.
The selected quartz crystals of all samples were cleaned by ultrasonic agitation
in distilled water and prepared for the different analytical methods. For SEM
studies, single crystals were arranged on a sample holder. Polished thin sections
were made for polarizing and CL microscopy.
The analytical procedure started with polarizing microscopy using a ZEISS
Axio Imager A1 m microscope. The material was additionally studied by SEM
(JEOL 6400 with EDX detector) to detect variations in grain size and morphology
of the quartz crystals and identify microinclusions. CL measurements were done
on carbon-coated, polished thin sections using a hot cathode CL microscope
HC1-LM (Neuser et al. 1995). The system was operated at 14 kV accelerating
voltage and a current density of about 10 lA/mm2. Luminescence images were
captured on-line during CL operations using a peltier-cooled digital videocamera (KAPPA 961-1138 CF 20 DXC). CL spectra in the wavelength range
380900 nm were recorded with an Acton Research SP-2356 digital triple-grating
spectrograph with a Princeton Spec-10 CCD detector that was attached to the CL
13
291
microscope by a silica-glass fibre guide. CL spectra were measured under standardized conditions (wavelength calibration by a Hg-halogen lamp, spot width
30 lm, measuring time 10 s).
Separated aliquots of the quartz samples were investigated for trace-element
composition by ICP-MS analysis. The sample material was carefully crushed and
again separated by hand picking under a binocular microscope. The separated
fractions were treated with distilled water to remove adhering particles and then air
dried. 400500 mg milled sample material (B30 lm grain size) were digested in a
glassy carbon vessel with 5 ml concentrated HF and 3 ml concentrated HNO3 at
50C and analyzed using a Perkin Elmer Sciex Elan 5000 quadrupole instrument
with a cross-flow nebuliser and a rhyton spray chamber. Details of the preparation
and analytical conditions are summarized in Monecke et al. (2000).
Investigations by XRD and Micro-Raman were carried out to check the samples
concerning other silica modifications than a-quartz. X-ray powder diffraction
analysis was carried out by means of an XRD 7 diffractometer using Cu-Ka
radiation and a secondary graphite monochromator. Samples were scanned with 2h
step sizes at step times of 10 s per step. The Rietveld algorithm BGMN was used
for refinement. Raman spectra were obtained by means of a JOBINYVON T
64000 Raman spectrometer with OLYMPUS microscope and macro-sample
chamber. Spectra were excited by the 514.5 nm line of an Ar+ laser with
0.51.5 mW beam power.
292
J. Gtze
Fig. 13.2 SEM micrographs of typical morphologies of the investigated authigenic quartz
samples. a Isolated euhedral quartz crystal from the kieseritic hard salt of the Zechstein salt
deposit Roleben (Germany); the crystal has dominant rhobohedral faces and almost no prism
faces; b Authigenic quartz crystal from the anhydritic hard salt of the Zechstein salt deposit
Roleben (Germany) showing a combination of rhombohedral and prism faces; c/d Aggregate of
authigenic quartz from a temporary outcrop of the Chirotherien sandstone (Middle Bunter) in
Jena (Thuringia, Germany); the close-up (d) of the outer part shows well developed crystal faces;
e Intergrowth of two isolated euhedral quartz crystals from the Tertiary lignite deposit Zwenkau
(Germany); f Authigenic quartz crystals from the Tertiary lignite deposit Cospuden (Germany)
with typical intergrowth (twinning?) of individual crystals
saliniferous environments are reflected in the mineral inclusions within the authigenic quartz crystals. Whereas the quartz from the kieseritic hard salt only contains
halite inclusions, the quartz crystals from the anhydritic hard salt exclusively inherit
microinclusions of calcium sulphate.
13
293
The authigenic quartz from the Bunter sandstone has a completely different
appearance. In contrast to common diagenetic quartz overgrowths in sandstones,
the crystals of nodular quartz may reach several mm or even cm in size, and
isolated euhedral crystals are more or less absent. Instead, the crystals are intergrown and form aggregates (Fig. 13.2c, d). Polarizing microscopy revealed that
the quartz shows zonal growth (Fig. 13.3c). Crystallization started with spherulitic
growth and the formation of chalcedonic quartz, later changing into well developed crystals. A similar sequence has been described for quartz in silicified
evaporites (e.g., Milliken 1979; Maliva 1987). The different stages are clearly
detectable by CL microscopy. Minute inclusions of calcite are included in the
central part of the aggregates. The quartz sample was checked by XRD and microRaman analyses, and only a-quartz was found as silica modification. This is in
contrast to the associated reddish chalcedony (carneol), where moganite could be
detected as a second silica phase besides a-quartz.
The authigenic quartz from the Tertiary lignite seams mainly consists of doubly
terminated euhedra, varying from some 100 lm to some mm in size. Within the
coal, the crystals are small and isolated, whereas also thin lenses and layers of
clusters and aggregates exist. Again, the crystals contain no detectable nucleus and
the morphology is dominated by prism and rhombohedral faces. The majority of
crystals shows elongated forms with well developed prism faces (Fig. 13.2e, f),
although few crystals with subordinate prism faces were also found. Moreover,
twinning and intergrowth of euhedral crystals is common. Investigations in
transmitted light revealed that inclusions of well preserved organic material are
common (Fig. 13.3d, e) and microinclusions of calcite sometimes occur.
13.3.2 Cathodoluminescence
Investigations by CL microscopy and spectroscopy showed that the quartz crystals
from the different sedimentary environments show significant differences in their
luminescence behaviour. Walker and Burley (1991) concluded that different
luminescence characteristics of authigenic quartz in sediments may reflect different conditions of formation.
The authigenic quartz crystals from the saline environment exclusively exhibit
a dark blue luminescence colour (Fig. 13.3a, b). Inclusions of halite (kieseritic
hard salt) and anhydrite (anhydritic hard salt), respectively, are clearly detectable
due to their brighter luminescence. Spectral CL measurements revealed that the
CL emission consists of a single broad band in the blue centred at ca. 450 nm
(2.75 eVFig. 13.4a). This luminescence emission can be related to the twofold
coordinated silicon on an oxygen vacancy (Fitting et al. 2001). Other typical CL
emission bands of quartz are lacking.
In contrast, the luminescence of the nodular quartz from the Bunter sandstone is
more complex. At least three different growth zones are detectable with CL. The
core of the quartz aggregates exhibits a yellow-orange CL with inclusions of
294
J. Gtze
13
295
b Fig. 13.3 Pairs of micrographs from polarizing and cathodoluminescence microscopy of the
authigenic quartz samples (uniform scale bar for all Figures). a Quartz crystal from the kieseritic
hard salt of the Zechstein deposit Roleben; the quartz has dark blue CL and brightly luminescing
microinclusions of halite; the euhedral shape of the crystal is slightly corroded. b Authigenic
quartz from the anhydritic hard salt of the Zechstein deposit Roleben; the quartz shows dark
blue CL; inclusions of anhydrite are detectable by their brighter luminescence. c Quartz aggregate
from the Middle Bunter sandstone of Jena; at least three growth zones are detectable: (1) a central
part with spherulitic growth and microinclusions of calcite, (2) a second zone with yellowbrownish CL and a complex internal structure, and (3) an outer zone of well crystallized quartz
without visible luminescence. d Intergrowth of two authigenic quartz crystals from the Tertiary
lignite deposit Zwenkau; the quartz exhibits an initial CL colour of brownish-violet; residuals of
organic matter with preserved cell structures are visible; e Authigenic quartz from the Tertiary
lignites of Cospuden with brownish luminescence after ca. 1 min of electron irradiation
296
J. Gtze
Fig. 13.4 Cathodoluminescence spectra of investigated authigenic quartz samples. a The quartz
from the Zechstein salt has a characteristic blue emission caused by a broad emission band at
450 nm. b The quartz from Jena (Bunter sandstone) shows a complex CL spectrum with at least
three emission bands: a main band at 650 nm and two additional bands at 580 and 450 nm.
c/d The quartz crystals from Tertiary lignite deposits (Zwenkau, Cospuden) have a dominant
emission band at 650 nm. Time-resolved CL measurements (d) revealed that the 650 nm
luminescence emission increases during electron irradiation
13.3.3 Geochemistry
The trace-element data of the investigated quartz samples are summarized in
Table 13.1. Because of the limited sample material of quartz from the salt deposits,
analyses could only be realized for the quartz from the kieseritic hard salt.
In general, trace elements in quartz are either incorporated into the crystal
lattice or contained in fluid or mineral inclusions. Because of the small ionic radius
and the high charge of the Si4+ ion, only a few elements are substitutionally
incorporated into quartz. The most common ion is Al3+; the structural incorporation of Ge, Ti, Ga, Fe or P (partially with charge compensating cations) was also
proved (Weil 1984). Other elements are preferentially bound on microinclusions
(Gtze et al. 2004).
The contents of most trace elements in the investigated authigenic quartz
samples are low. Some trace elements (e.g., Ga, Pb, Sn, V, Co, Cr, Mn, Nb, Ni, Zr)
were not detectable in all analyzed quartz samples. Only some specific elements
are enriched, which reflect both the specific physico-chemical environment of
13
297
Table 13.1 Trace element contents (ppm) of authigenic quartz from different occurrences
QR
QCh
QZ
QCo
Al
Au
B
Ba
Ca
Cs
Fe
Ge
K
Li
Mg
Na
Rb
Sr
Ti
Th
U
770
0.043
10
\7.7
146
0.108
105
nd
1240
60
640
7360
1.11
nd
nd
0.075
\0.08
21.9
68
2.83
351
103
2.10
26.5
0.97
6.09
53.0
0.05
0.86
7.24
1.32
104
17
7.33
580
45.7
0.04
40.7
0.80
15.1
88.2
0.11
0.89
1.88
0.08
0.02
Y
La
Ce
Pr
Nd
Sm
Eu
Gd
Tb
Dy
Ho
Er
Tm
Yb
Lu
nd
1.020
\0.29
0.038
\0.013
0.006
0.034
\0.009
0.0448
0.0226
0.0695
0.0051
0.0163
0.0041
0.0013
0.0056
0.0012
0.0059
0.0016
0.0051
0.0009
0.0053
0.0011
0.1103
1.8929
2.6064
0.2412
0.2894
0.0435
0.0046
0.0259
0.0044
0.0211
0.0048
0.0139
0.0018
0.0099
0.0015
79.4
nd
6.13
28.1
8.8
53.1
3.57
8.21
43.1
0.06
0.26
9.26
0.0490
0.3711
0.4364
0.0406
0.0401
0.0046
0.0009
0.0048
0.0010
0.0061
0.0018
0.0064
0.0011
0.0070
0.0013
298
J. Gtze
Fig. 13.5 Chondrite-normalized REE distribution patterns of authigenic quartz samples from
different sediments (normalization according to data of Mason 1979). QChBunter sandstone,
Jena; QZTertiary lignites, Zwenkau; QCoTertiary lignites Cospuden
within the primary fluid-inclusions of quartz from the Zechstein salt deposits.
Therefore, these elements were entrapped during the authigenic quartz formation
within the saline environment. This conclusion is also supported by the elevated
content of B, since authigenic boracite was found together with quartz in the
insoluble residues of the salts. Additionally, soluble metal salts (e.g., Fe, Al, Au)
could be transported and precipitated simultaneously with Si.
Looking at the trace-element composition of the authigenic quartz from nodules
in the Bunter sandstone, at first the elements Al, Ca, Na and K are contained in
elevated concentrations. Assuming a silica source from SiO2-saturated pore fluids,
these elements could probably originate from the silica supply during alteration
and dissolution of silicate minerals (e.g., feldspar minerals). After release and
migration within the silica bearing solutions, these elements could be incorporated
into the authigenically forming quartz. The contents of some specific trace elements (e.g., B, Ge, U) are also remarkable. These elements show a similar geochemical behaviour like Si and thus, can be transported in the fluid together with
silicon. This is interesting in so far as these trace elements are also characteristic
elements in the composition of agates (Gtze et al. 2001b).
The trace-element contents in the two samples from Tertiary lignites are in
general low, except the already mentioned enrichment in Al and Fe. Elevated
contents of Ca can be related to the occurrence of minor inclusions of calcite.
Additionally, the elements Mg and Na are preferentially concentrated in fluid
inclusions. Highly saline MgCl2(NaCl)-solutions were detected in authigenic
quartz crystals from Tertiary lignite (Fruth and Blankenburg 1992).
13
299
300
J. Gtze
13
301
with the replacement of carbonate nodules by silica during the highly saline
Roethian transgression (Langbein 1974). The mixing of saline water with silicabearing pore fluids resulted in the dissolution/replacement of carbonate and precipitation of silica. The trace element data indicate that the SiO2 in the pore
solutions derives from the alteration and decomposition of silicate material in the
sediment; no hydrothermal silica input could be detected.
Whereas most of the silica rapidly crystallized as chalcedony (carneol) and
chalcedonic pore cement, lowering of the SiO2 concentration favoured the crystallization of euhedral quartz crystals in cavities. The microscopic investigations
showed that the formation of quartz took place in a multi-step process. Starting on
a nucleus of carbonate, spherulithic quartz crystallized first. The formation of
chalcedony is promoted both by the replacement of pre-existing carbonate and
probably high silica concentrations. The following phanerocrystalline quartz zone
seems to be the prosecution of chalcedony fibres. The high defect density
(characteristic 580 nm CL emission band) and the occurrence of sector zoning
detected by CL may result from rapid crystallization under non-equilibrium conditions. Only the outer part of the quartz aggregates probably crystallized directly
from an undersaturated fluid with respect to silica.
Milliken (1979) discussed similar crystallization sequences of quartz from
silicified evaporates. She concluded that microcrystalline quartz formed earliest in
fluids related to sea water, whereas megaquartz crystallized later from meteoric
fluids. Applied to the quartz nodules from the Bunter sandstone, the changing
quartz microtexture could be an indication for a changing ratio of pore fluids of
saline and meteoric origin, and decreasing silica content.
In conclusion, specific microchemical conditions were necessary for the formation of the euhedral authigenic quartz crystals.
302
J. Gtze
The perfect euhedral shape of most authigenic quartz crystals in lignites, the
absence of any detrital nucleus and the inclusions of organic matter and calcite
support the theory of an authigenic formation within the coal seams. They may
have formed by crystallization from silica-rich solutions migrating in the pore
space and along fissures during an early stage of diagenesis. The presence of
organic matter can increase the silica solubility (e.g., stabilization by protecting
colloids). The diagenesis of the organic matter accompanied by CO2 production
lowers the pH and results in the precipitation of quartz (Zajic 1969). On the other
hand, the inflow of saliniferous pore fluids during the transgression of the sea may
have promoted the precipitation of quartz. This idea is supported by the detected
high salinity of the fluid inclusions in the authigenic quartz crystals (Fruth and
Blankenburg 1992).
The temperatures of quartz formation can be estimated from results of fluid
inclusion and oxygen isotope studies, respectively. Oxygen isotope analysis of
separated authigenic quartz crystals from a subbituminous coal seam (Charlston,
New Zealand) gave a result of 26.5 0.5 % d18O (SMOW), which the authors
interpreted with a maximum temperature of formation of about 65C (Soong and
Blattner 1986). This is in accordance with homogenization temperatures of quartz
from Tertiary lignites in Germany, which gave minimum temperatures of crystallization between 39 and 60C (Fruth and Blankenburg 1992). These temperatures are consistent with burial to only shallow depths given normal geothermal
gradients.
13.4 Conclusions
The results of the present study illustrate that quartz authigenesis is common in
different sedimentary environments. The formation of quartz crystals often went
under similar conditions but, nevertheless, resulted in characteristic properties of
authigenic quartz in dependence on the specific physico-chemical conditions and
the sedimentary environment, respectively.
Most authigenic quartz crystals are suspended within the rock matrix. Microscopic studies (SEM, CL) show that they have a perfect euhedral shape and no
identifiable nucleus. Therefore, nucleation of authigenic quartz on detrital quartz
grains, as it is dominant for secondary silica overgrowths in sandstones, does not
play any role in the formation of isolated authigenic quartz crystals. The morphology of the quartz crystals (dominating crystal faces) strongly depends on
the environment of formation. Moreover, the microinclusions and also the traceelement composition of authigenic quartz reflect the specific geochemical
environment during quartz growth.
CL microscopy and spectroscopy revealed that the quartz crystals from different
sediments show completely different luminescence behaviour. Quartz from the salt
deposit is characterized by a deep blue CL with an emission band at 450 nm.
Authigenic nodular quartz from the Bunter sandstones shows zonal growth with
13
303
References
Bahlburg H, Floyd PA (1999) Advanced techniques in provenance analysis of sedimentary rocks.
Sedimentary Geology, vol 124. Elsevier, Amsterdam
Baker G (1946) Microscopic quartz crystals in brown coal, Victoria. Am Mineral 31:2230
Bellmann HJ (1986) Zur Genese der verkieselten Hlzer und Braunkohlenquarzite im Raum
Leipzig. Zeitschrift geologische Wissenschaften 13:699704
Bennett P (1991) Quartz dissolution in organic-rich aqueous systems. Geochim Cosm Acta
55:17811797
Bennett P, Siegel DI (1987) Increased solubility of quartz in water due to complexing by organic
compounds. Nat 326:684686
Bernet M, Bassett K (2005) Provenance analysis by single-quartz-grain SEM-CL/optical
microscopy. J Sed Res 75:492500
Bjrlykke K, Egeberg PK (1993) Quartz cementation in sedimentary basins. Am Assoc Petrol
Geol Bull 77:15381548
304
J. Gtze
Black WW (1949) An occurrence of authigenic feldspar and quartz in Yoredale limestones. Geol
Mag 86:129
Boogs S Jr, Kwon Y-I, Goles GG, Rusk BG, Krinsley D, Seyedolali A (2002) Is quartz
cathodoluminescence a reliable provenance tool? A quantitative examination. J Sed Res
72:408415
Botz RW, Hunt JW, Smith JW (1986) Isotope geochemistry of minerals in Australian bituminous
coal. J Sed Petrol 56:99111
Burley SD, Mullis J, Matter A (1989) Timing diagenesis in the Tartan Reservoir (UK North Sea):
constraints from combined cathodoluminescence microscopy and fluid inclusion studies.
Marine Petrol Geol 6:98120
Chavetz HS, Zhang J (1998) Authigenic euhedral megaquartz crystals in a Quaternary dolomite.
J Sediment Res 68:9941000
Dixon JB, Weed SB (1989) Minerals in soil environments. Soil Science Society of America,
Madison
Evans MA, Elmore RD (2006) Fluid control of localized mineral domains in limestone pressure
solution structures. J Struct Geol 28:284301
Evans J, Hogg AJC, Hopkins MS, Howarth RJ (1994) Quantification of quartz cements using
combined SEM, CL, and image analysis. J Sediment Petrol A64:334338
Fabricius J (1987) Natural Na-K-Mg-Cl solutions and solid derivatives trapped in euhedral quartz
from Danish Zechstein salt. Chem Geol 61:95112
Fitting H-J, Barfels T, Trukhin AN, Schmidt B (2001) Cathodoluminescence of crystalline and
amorphous SiO2 and GeO2. J Non-Cryst Solids 279:5159
Friedman GM, Shukla V (1980) Significance of authigenic quartz euhedra after sulphates;
example from the Lockport formation (Middle Silurian) from New York. J Sediment Res
50:12991304
Fruth M, Blankenburg H-J (1992) Charakterisierung von authigenen idiomorphen Kohle- und
Salinarquarzen durch Einschlussuntersuchungen. Neues Jahrbuch Mineralogie, Abhandlungen
165:5364
Fchtbauer H (1961) Zur Quarzneubildung in Erdllagersttten. Erdl und Kohle 14:169173
Gtze J (1998) Geochemistry and provenance of the Altendorf feldspathic sandstone in the
Middle Bunter of the Thuringian basin (Germany). Chem Geol 150:4361
Gtze J, Zimmerle W (2000) Quartz and silica as guide to provenance in sediments and
sedimentary rocks. Contributions to Sedimentary Geology, vol 21E. Schweizerbartsche
Verlagsbuchhandlung, Stuttgart, p 91
Gtze J, Pltze M, Fuchs H, Habermann D (1999) Defect structure and luminescence behavior of
agate-results of electron paramagnetic resonance (EPR) and cathodoluminescence (CL)
studies. Mineral Mag 63:149163
Gtze J, Pltze M, Habermann D (2001a) Origin, spectral characteristics and practical
applications of the cathodoluminescence (CL) of quartz: a review. Miner Petrol 71:225250
Gtze J, Tichomirowa M, Fuchs H, Pilot J, Sharp Z (2001b) Geochemistry of agates: a trace
element and stable isotope study. Chem Geol 175:523541
Gtze J, Pltze M, Graupner T, Hallbauer DK, Bray C (2004) Trace element incorporation into
quartz: a combined study by ICP-MS, electron spin resonance, cathodoluminescence,
capillary ion analysis and gas chromatography. Geochim Cosmochim Acta 68:37413759
Grimm W-D (1962) Idiomorphe Quarze als Leitmineralien fr salinare Fazies. Erdl und Kohle
15:880887
Hartmann BH, Juhsz-Bodnr K, Ramseyer K, Matter A (2000) Polyphased quartz cementation
and its sources: a case study from the Upper Paleocoic Haishi Group sandstones; Sultanate of
Oman. IAS Special Publications 29:253270
Heynke A, Znker G (1970) Zur Ausbildung und Leitbankgliederung des Stafurtsteinsalzes im
Sdharz-Kalirevier. Zeitschrift angewandte Geologie 16:344356
Hiatt EE, Kyser TK, Fayek M, Polito P, Holk GJ, Riciputi LR (2007) Early quartz cements and
evolution of paleohydraulic properties of basal sandstones in three Paleoproterocoic
13
305
continental basins: evidence from in situ d18O analysis of quartz cements. Chem Geol
238:1937
Hoehne K (1954) Zur Neubildung von Quarz in Kohlenflzen. Neues Jahrbuch Geologie
Palonthologie, Abhandlungen 99:209220
Houseknecht DW (1991) Use of cathodoluminescence petrography for understanding compaction, quartz cementation, and porosity in sandstones. In: Baker CE, Kopp OC (eds)
Luminescence microscopy: quantitative and qualitative aspects. SEPM, Dallas, pp 5966
Kelly JL, Fu B, Kita NT, Valley JW (2007) Optically continuous silcrete quartz cements of the
St. Peter sanstone: high precision oxygen isotope analysis by ion microprobe. Geochim
Cosmochim Acta 71:38123832
Langbein R (1974) Zur Petrologie der Karneole des thringischen Chirotherien Sandsteins
(Solling-Folge). Chemie der Erde 33:301325
Leskevich LE (1959) Quartz crystals in coal (in Russian). Doklady Akademii Nauk SSSR
124(3):575577
Liu X, Wang S, Zhang F (2004) Fission track dating of authigenic quartz in red weathering crusts
of carbonate rocks in Guizhou province (Chinese with English Abstract). Acta Geol Sinica
78:11361142
Lyon IC, Burley SD, McKeever PJ, Saxton PJ, Macaulay JM (2000) Oxygen isotope analysis of
authigenic quartz in sandstones: a comparison of ion microprobe and conventional analytical
techniques. In: Worden RH, Morad S (eds) Quartz cementation in sandstones. Blackwell
Science, Oxford, pp 299316
Maliva RG (1987) Quartz geodes: Early diagenetic silicified anhydrite nodules related to
dolomitization. J Sediment Res 57:10541059
Markowitz A, Milliken KL (2003) Quantification of brittle deformation in burial compaction,
Frio and Mount Simon formation sandstones. J Sediment Res 73:10071021
Mason B (1979) Cosmochemistry, Part I. Meteorites. In: Fleischer M (ed) Data of geochemistry,
U.S. Geological Survey professional papers, 440-B1, 132 p
McBride EF (1989) Quartz cement in sandstones: a review. Earth Sci Rev 26:69112
Milliken KL (1979) The silicified evaporate syndrometwo aspects of silicification history of
former evaporate nodules from Southern Kentucky and Northern Tennessee. J Sediment
Petrol 49:245256
Milliken KL, Laubach SE (2000) Brittle deformation in sandstone diagenesis as revealed by
scanned cathodoluminescence imaging with application to characterization of fractured
reservoirs. In: Pagel M, Barbin V, Blanc P, Ohnenstetter D (eds) Cathodoluminescence in
geosciences. Springer, Berlin, pp 225243
Miik M (1995) Authigenic quartz crystals in the Mesozoic and Paleogene carbonate rocks of the
Western Carpathians. Geologica Carphatica 46:227239
Molenaar N, deJong AFM (1987) Authigenic quartz and albite in Devonian limestones: origin
and significance. Sedimentology 34:623640
Monecke T, Bombach G, Klemm W, Kempe U, Gtze J, Wolf D (2000) Determination of trace
elements in quartz standard UNS-SpS and in natural quartz by ICP-MS. Geostand Newslett
24(1):7381
Nachsel G (1969) Idiomorphe Quarze und Vertaubungen im Kaliflz Stafurt des SdharzKalireviers. Z Angew Geol 15:420425
Neuser RD, Bruhn F, Gtze J, Habermann D, Richter DK (1995) Kathodolumineszenz: Methodik
und Anwendung. Zentralblatt fr Geologie und Palontologie Teil I H 1/2:287306
Parnell J, Carey PF, Monson B (1996) Fluid inclusion constraints on temperatures of petroleum
migration from authigenic quartz in bitumen veins. Chem Geol 129:217226
Richter DK (1971) Fazies- und Diagenesehinweise durch Einschlsse in authigenen Quarzen.
Neues Jahrbuch Geologie Palonthologie, Monatshefte H 10:604622
Richter DK, Gtte Th, Gtze J, Neuser RD (2003) Progress in application of cathodoluminescence (CL) in sedimentary petrology. Miner Petrol 79:127166
Ruppert LF, Cecil CB, Stanton RW, Christian RP (1985) Authigenic quartz in the Upper Freeport
coal bed, west-central Pennsylvania. J Sediment Res 55:334339
306
J. Gtze
Schneider W (1986) Phytogene Verkieselungen in der mioznen Braunkohle und deren Aussagen
fr Stratigraphie, Fazies und Flzgenese. Zeitschrift geologische Wissenschaften 14:153162
Sedletskiy VI (1971) Some features of authigenic quartz formation in evaporate basins (in
Russian). Geologika Geofizika 5:7277
Seyedolali A, Krinsley DH, Boggs S Jr, OHara PF, Dyavik H, Goles GG (1997) Provenance
interpretation of quartz by scanning electron microscope-cathodoluminescence fabric
analysis. Geology 25:787790
Siegel GH, Marrone MJ (1981) Photoluminescence in as-drawn and irradiated silica optical
fibers: An assessment of the role of nonbridging oxygen defect centers. J Non-Cryst Solids
45:235247
Sippel RF (1968) Sandstone petrology, evidence from luminescence petrography. J Sediment
Petrol 38:530554
Soong R, Blattner P (1986) Biterminal authigenic 18O-enriched quartz in a subbituminous coal
seam, Charleston, New Zealand. N Z J Geol Geophys 29:141145
Spiro B, Rozenson I (1982) Formation and properties of authigenic minerals in bituminous
calcareous shales, Ghareb Formation, Israel. Can Mineral 20:2939
Walderhaug O (1990) A fluid inclusion study of quartz-cemented sandstones from offshore midNorwaypossible evidence for continued quartz cementation during oil emplacement.
J Sediment Petrol 60:203210
Walderhaug O (1994) Temperature of quartz cementation in Jurassic sandstones from the
Norwegian continental shelfevidence from fluid inclusions. J Sediment Res A64:311323
Walker G, Burley S (1991) Luminescence petrography and spectroscopic studies of diagenetic
minerals. In: Barker CE, Kopp OC (eds.) Luminescence microscopy and spectroscopy:
Qualitative and quantitative applications. SEPM, Tulsa, pp 8396
Walther H, Gtze J (1994) Zur Bildung von Quarziten und authigenen Quarzen in Braunkohlen.
Eur J Mineral Beiheft 1(6):301
Weil JA (1984) A review of electron spin spectroscopy and its application to the study of
paramagnetic defects in crystalline quartz. Phys Chem Miner 10:149165
Wood SA (1990) The aqueous geochemistry of the rare-earth elements and yttrium. Chem Geol
88:99125
Wordan RH, Morad S (2000) Quartz cementation in oil field sandstones: a review of the key
controversies. In: Worden RH, Morad S (eds.) Quartz cementation in sandstones. Blackwell
Science, Oxford, pp 120
Zajic JM (1969) Microbial biogeochemistry. Academic Press, New York
Zinkernagel U (1978) Cathodoluminescence of quartz and its application to sandstone petrology.
Contrib Sedimentol 8:169
Zuffa GG (1985) Provenance of arenites. NATO ASI series C 148, Reidel Publ. Co., Boston, 393 p
Chapter 14
B. Rusk (&)
Economic Geology Research Unit, School of Earth and Environmental Sciences,
James Cook University, Townsville 4811, Australia
e-mail: brian.rusk@jcu.edu.au
307
308
B. Rusk
from quartz analysis can therefore be used to fingerprint the origin of quartz and
make some inferences about the pressure, temperature, and fluid compositional
changes that accompany hydrothermal quartz precipitation.
14.1 Introduction
Quartz is one of the most abundant minerals in the crust and is the dominant gangue
mineral in many hydrothermal ore deposits, where it forms under a wide range of
temperature and pressure conditions from fluids of diverse origins and compositions. Genetic models of ore deposit formation rely heavily on information obtained
from fluid inclusions trapped in and isotopic data derived from vein quartz. The
interpretation of such data relies upon our ability to understand the timing relationships between the analyzed quartz and the other minerals of interest. Whereas
optical petrography offers some insight into these textural relationships (Adams
1920; Spurr 1926; Dowling and Morrison 1989; Vearncombe 1993; Dong et al.
1995), scanning electron microscopecathodoluminescence (SEM-CL) reveals
textures in quartz that are not observable by optical petrography (Sippel 1968,
Smith and Stenstrom 1965). These unique textures illustrate the sequential history
of quartz precipitation, dissolution, fracturing, recrystallization, and/or subsequent
quartz growth. When deciphered, such textures have important implications for
unraveling the physical and chemical history of a hydrothermal system.
Cathodoluminescent textural variations result from defects in the crystal lattice,
many of which are related to the incorporation of trace elements (Marshall 1988;
Ramseyer et al. 1988; Stevens-Kalceff and Phillips 1995; Gtze et al. 1999, 2001;
Walker 2000; Stevens-Kalceff et al. 2000; Gtze 2009). The trace element variations and the resultant CL textures reflect the rate of crystallization, fluid/magma
composition, pressure, or temperature, as well as any subsequent metamorphic
and/or deformational history after quartz formation (Perny et al. 1992; Mullis et al.
1994; Lowenstern and Sinclair 1996; Ramseyer and Mullis 2000; Gtze et al.
2001; Monecke et al. 2002; Larsen et al. 2004; Breiter and Mller 2009). Currently
little is known about the specific conditions that are reflected by quartz trace
elements, however an increasing number of studies of natural and synthetic quartz
are aimed at quantifying the effect of crystal growth rate, temperature, pressure,
and fluid composition on the incorporation of trace elements in quartz (Wark and
Watson 2006; Rusk et al. 2008; Thomas et al. 2010; Gtte et al. 2011). Because
quartz precipitates throughout much of the history of hydrothermal systems, the
interpretation of CL textures and quartz trace elements greatly enhances our ability
to interpret the evolving physical and chemical conditions of hydrothermal fluid
flow in the crust.
This paper provides an overview of CL textures observed and trace elements
measured in quartz from a variety of hydrothermal environments. We compare CL
textures, spectra, and trace elements among hydrothermal ore deposits and show
14
309
that these data can be used to interpret the physical and chemical conditions of
vein formation, to correlate hydrothermal events across a broad region, to relate
fluid inclusion and isotopic data to specific mineralization events, and to fingerprint ore deposit types. This overview is based on existing literature as well as on
trace element and SEM-CL studies of *300 vein samples from *50 hydrothermal ore deposits, with a focus on porphyry-Cu (-Mo-Au) deposits, epithermal
deposits, and orogenic Au deposits. Comparison of textures and trace elements
among quartz veins formed under different pressure and temperature conditions
helps to constrain the most significant factors that affect CL textures and trace
element incorporation. In addition to understanding geologic processes, this work
has implications for understanding the formation of high purity quartz, which is
required for the manufacture of many high-tech products such as silicon wafers
used in computers and photovoltaic solar cells.
310
B. Rusk
14
311
Fig. 14.2 a SEM-CL image of a zoned phenocryst form the Tamagawa Tuff. Euhedral growth
zones of oscillating CL intensity are typical of volcanic quartz phenocrysts; b Phenocryst from an
intrusive porphyrytic rock from Summitville Colorado. Growth zones are slightly more complex
than in typical volcanic rocks and embayments are common; c Quartz from intrusive igneous
rocks, may or may not display euhdral growth zonations (Nimrod granite, OR). CL-dark fractures
and spiders are common in plutonic quartz; d Metamorphic quartz typically displays nearly
homogenous or slightly mottled texture, indicating annealing of original CL textures. This sample
is from the Packsaddle schist, Texas
hydrothermal veins have complex histories that involve multiple stages of fluid
flow, quartz precipitation, dissolution, fracturing, and/or recrystallization. These
textures provide critical petrographic evidence that can be used to (1) Deconvolute
the chronologic sequence of vein forming events; (2) To relate specific generations
of quartz to specific hydrothermal events or fluid inclusion populations or (3) To
infer distal and cryptic expressions of hydrothermal systems, (Boiron et al. 1992;
Valley and Graham 1996; Wilkinson and Johnston 1996; King et al. 1997;
312
B. Rusk
Fig. 14.3 SEM-CL photomosaic of a chalcopyrite-bearing quartz vein from the porphyry-CuMo deposit in Butte, Montana. CL image shows an early CL-bright generation of quartz that is
cut by later CL-darker quartz. The later CL-dark quartz both cuts and overgrows the earlier CLbright quartz. Where the later quartz grew into open space, it developed euhedral growth zones
and filled a void along with chalcopyrite (black mineral in upper left corner). The later euhedral
crystals and the earlier CL-bright crystals are all cut by a later generation of CL-dark fractures,
which correspond to trails of fluid inclusions
14
313
(4) CL-dark bands; (5) spider and cobweb texture; (6) rounded cores with overgrowths; (7) microbrecciation; (8) rounded or wavy concentric zonation; and (9)
homogeneous (or slightly mottled) texture (Fig. 14.4). The above CL textures
result from the initial (primary) conditions of quartz crystallization followed by
secondary processes such as dissolution, fracturing, overgrowth, and recrystallization that modify the original textures. The presence of numerous primary and
secondary CL textures in many veins from many hydrothermal ore deposits
indicates that the veins have a complex history involving multiple fluid incursions
under varying pressure and temperature conditions. Interpretation of the conditions
under which these textures form leads to improved models of hydrothermal systems and hydrothermal fluid flow in the Earths crust.
314
B. Rusk
Fig. 14.4 SEM-CL textures in hydrothermal quartz are much more variable and complex in
hydrothermal quartz than in any other quartz type. a Euhedral growth zones are the most common
texture observed in hydrothermal vein quartz (Tokatea epithermal deposit, NZ); b Spherical
texture, only observed in epithermal veins (McLaughlin epithermal deposit, CA); c Mosaics of
fine grained equant crystals of vein quartz from Refugio porphyry Au deposit, Peru. Grains within
the quartz mosaic show internal wavy growth zonation; d Early CL-bright quartz cut by later CLdarker quartz with euhedral growth zones, cut by CL-dark fractures (Bata Hijau, Indonesia
porphyry Cu deposit); e Spider and cobweb texture, typical of porphyry type deposits (Henderson
porphyry MO deposit, CO); f CL-bright cores with CL-darker overgrowths. This texture has been
recognized in multiple porphyry-type deposits and is attributed to dissolution, possibly caused by
fluid cooling through the zone of retrograde quartz solubility (Gaby Sur, Chile); g Wavy
concentric banding, seen only in quartz from porphyry-type deposits, and likely indicating some
post-crystallization elemental diffusion (Butte, Montana); h Microbreccia texture showing early
CL-bright quartz (Q1) and pyrite fractured and infilled by CL-darker Q2, which is subsequently
overgrown by slightly brighter euhedral Q3. All of these quartz generations are then fractured by
late CL-dark fractures (Q4) (Cordilleran base metal lodes, Butte, MT); I Homogenous CL texture,
typical of orogenic Au veins (Valdez Creek, Alaska). Cp = chalcopyrite, Ep = epoxy, Ab =
albite, WR = wallrock, Py = pyrite
14
315
Fig. 14.5 Textures typical of porphyry copper deposits (a, b, c), epithermal deposits (d, e, f), and
orogenic Au deposits (g, h, i). The consistent suite of textures in each deposit type makes CL
imaging useful for fingerprinting the origin of quartz in samples where the origin is not known
(stream samples, soil samples, etc.). a Four generations of quartz (labeled Q1 to Q4)
demonstrating superposition of multiple hydrothermal events. CL-brightest Q1 is igneous quartz
in the wall rock. Q1 is overgrown by CL-bright mosaic quartz of Q2, which is dissolved and
overgrown by CL-darker quartz of Q3. Q2 and Q3 are cut by a late CL-dark fracture of Q4 (Los
Pelambres, Chile); b Mosaics of euhedrally zoned quartz (Grasberg, PNG); c Quartz displaying
many features typical of porphyry copper deposits including inward-oriented euhdral growth
projecting into chalcopyrite, splatter and cobweb texture, and CL-dark bands (healed fractures)
(El Teniente, Chile); d Chalcedonic bands intergrown with bands that are slightly more
crystalline (Creede, CO, USA); e A fairly simple inward-oriented euhedral quartz crystal
(Comstock Lode, NV, USA); f Microcrystalline quartz with irregular concentric and spherical
patterns (McLaughlin, CA, USA); g Homogenous texture in quartz with some feldspar in the vein
(Valdez Creek, AK, USA). h Mostly homogenous CL-dark quartz (Norseman, Aus); i Slightly
mottled to homogenous texture of quartz with pyrite (Carson Hill, CA, USA). Cp = chalcopyrite,
Ep = epoxy, Ab = albite, WR = wallrock, Py = pyrite
316
B. Rusk
trails of secondary fluid inclusions. Where CL-dark bands coincide with trails of
secondary fluid inclusions, the inclusions are primary with respect to the dark
quartz in which they lie. Thus if the event with which the fractures are associated
can be determined through CL imaging, then the inclusions may be considered
primary with respect to that event. In a few samples multiple generations of
fracturing are obvious based on cross-cutting relations of CL-dark fractures of
varying CL intensities. Some CL-dark bands are filled with inwardoriented
euhedral quartz crystals indicating fracturing, then dilation, fluid infiltration, and
quartz precipitation. In some examples CL-dark bands end in, or grade into
euhedral quartz crystals of similar CL intensity, which indicates that the fluid that
flowed through the fracture creating the band is the same fluid from which crystals
precipitated into fluid-filled space (Fig. 14.3 for example).
Recrystallization of quartz is reflected by homogenous or mottled CL texture
and low contrast rounded ghosty growth zones. However, in some cases, it is
difficult to infer whether the quartz initially precipitated with little or no CL texture
or whether a previously existing CL-texture was subdued by annealing. Homogenous texture is characteristic of quartz from metamorphic rocks (Seyedolali et al.
1997; Boggs et al. 2002), where the lack of texture is due to annealing of CL
textures resulting from redistribution of lattice defects and trace element concentrations under metamorphic conditions (Sprunt et al. 1978; Spear and Wark
2009). Alternatively, mottled and homogenous textures may be a primary precipitation feature, indicating little internal lattice variation in quartz grains that
grew under stable conditions at relatively high temperature and pressure, as suggested for pegmatitic quartz (Gtze et al. 2005). Homogenous and mottled CL
textures are common in orogenic Au samples and in high temperature veins with
potassic alteration from some porphyry copper deposits.
Mineral replacement is another possible secondary texture, but the CL textures
resulting from mineral replacement may vary widely. Mineral replacement is most
obvious in the case of pseudomorphs, where the original crystal shape is preserved,
such as bladed calcite crystals replaced by quartz, which is common in some
epithermal ore deposits. Leach et al. (2005) conclude that interlocking euhedral
quartz crystals with multiple crystallographic orientations formed from the
replacement of barite by quartz in the Red Dog PbZn deposit of AK.
Evidence for dissolution includes rounded or embayed CL-bright quartz cores
with euhedrally zoned CL-dark overgrowths. This texture forms where early
CL-bright quartz is fractured, pervaded, and dissolved by hydrothermal fluids,
followed by precipitation of CL-dark quartz into the pore-space created by fracturing and dissolution. Dissolution is also inferred where early quartz has a
rounded contact against younger quartz, which is almost always CL-darker.
In some instances growth zones within early quartz are truncated by growth zones
in the later overgrowth quartz, which have different zonation patterns. Evidence
for dissolution is most convincing when the CL-bright cores are rounded or
embayed against the later CL-darker quartz, and when the late quartz displays
euhedral growth zones indicating growth into open space.
14
317
318
B. Rusk
Electron microprobe offers the highest spatial resolution and smallest sampling
volume of any of the above analytical techniques. This is critical in many samples
where CL features, such as cross cutting quartz generations or euhedral growth
zones, are only a few microns or tens of microns in width. However detection
limits, typically in the range of several tens of ppm using long counting times and
other techniques (see Donovan et al. 2011), are generally too high to detect all but
the most abundant quartz trace elements. Typically the only elements quantifiable
by microprobe are Al, K, and sometimes Ti and Fe. Important quartz trace elements Li, K, Na, P, and H cannot typically be analyzed via electron microprobe.
The main benefit of LA-ICP-MS is the ability to quantify greater than 30 elements simultaneously and achieve detection limits in the range of tens of ppb to a
few ppm for many elements. However, in order to obtain such low detection limits,
the volume of consumed sample material using laser ablation is much larger than by
microprobe. Spot sizes [35 l in diameter are required to obtain low detection
limits for multiple elements. Because the interaction volume is so large with a laser,
the accidental analysis of fluid, mineral, and melt inclusions is always a risk,
especially in hydrothermal quartz, which can be rich in mineral and fluid inclusions.
The accidental ablation of inclusions may be obvious if they are large, however it
may be very difficult to determine whether an analyzed element is present in the
quartz structure or as micro- or nano-inclusions. Like with EPMA, H cannot be
analyzed using LA-ICP-MS, however most other elements of interest in quartz can.
Only a few studies exist that show the usefulness of trace element analysis in
quartz using SIMS. The spatial resolution of SIMS is intermediate between EPMA
and LA-ICP-MS, with a typical spot size of *1015 l. Currently the main
drawback of this technique is that no standards exist that can be used to quantify
quartz trace element data. SIMS analyses can still be useful in showing element
distribution in the form of qualitative trace element maps (Lehmann et al. 2009).
The main advantage of SIMS is that it is the only instrument that can be used to
map H distribution. If appropriate standards were developed or characterized,
quantitative SIMS analyses of quartz would have the combined advantages of low
detection limits and high spatial resolution (c.f. Behr et al. 2011).
Of the above techniques, Micro-FTIR is the only one able to quantify H
concentration and speciation. However other trace elements cannot be simultaneously quantified and there are a number of complications in data quantification
(Miyoshi et al. 2005; Mller and Koch-Mller 2009; Thomas et al. 2009). The
resolution of this technique is on the order of 50 l and it requires special sample
preparation and inclusion-free quartz. The main advantage of this technique in
studies of quartz trace elements is that it is the only in situ technique available for
the quantification of H in quartz.
14
319
Wanvik 2012). Even though it is pure relative to most other minerals, hydrothermal
quartz contains several important elements. As with CL textures, concentrations of
trace elements in hydrothermal quartz vary more widely than do quartz trace elements concentrations in other geologic environments, reflecting the wide range of
physical and chemical conditions under which hydrothermal quartz forms.
Of the trace elements that are commonly identified in quartz, Al is typically the
most abundant, and concentrations vary widely from a few ppm up to
*5,000 ppm. Al is the only trace element that is ever present in concentrations
greater than a thousand ppm. Other trace elements present in concentrations
between a few tens and a few 100 ppm in quartz may include Ti, Li, K, Sb, Fe, Ca,
Na, P. In addition to these, H, B, Ge, Ga, Sn, Ba, Cs, and As, may be present in the
concentration range of a few 100 ppb to a few ppm in quartz from various
environments (Penniston-Dorland 2001; Monecke et al. 2002; Flem et al. 2002;
Breiter and Mller 2009; Mller et al. 2003; Larsen et al. 2004; Allan and Yardley
2007; Rusk et al. 2008). In addition, some studies indicate that Cl and H2O may be
structurally bound in quartz (Allan and Yardley 2007; Stenina 2004).
Lithium varies from a few ppm to a few 100 ppm in hydrothermal quartz and is in
the range of *300 ppm in low temperature quartz where Al is in the range of
3,000 ppm. Numerous studies show that the most common correlation among quartz
trace elements is the correlation between Al and Li. Although typically slightly less
abundant, K and Na are also common in hydrothermal quartz and range from a few
ppm up to a few 100 ppm and tend to correlate positively with Al concentration (for
example Perny et al. 1992). However, in quantifying concentrations of Na and K in
hydrothermal quartz, one must be cautious of the presence of fluid inclusions, which
may contain several orders of magnitude more Na and K than quartz. The accidental
analysis of volumetrically minor quantities of fluid inclusions will have significant
impacts on the measured concentrations of these elements. The common observation
of positive correlations between Al and these monovalent cations, provides evidence
for a coupled a substitution in quartz where Al3+ and Li+, K+, or Na+ substitute for
Si4+, thus maintaining charge balance. In most quartz however, the molar ratio of Al/
(Na ? K ? Li) is significantly greater than one and H+ is most commonly suggested
to make up the remainder of the charge balance, as it has been shown to be abundant
in some quartz (Miyoshi et al. 2005, Lehmann et al. 2009). Considering all of the
potentially significant trace elements in quartz, to maintain charge neutrality, the
total of the trivalent cations Fe and Al should be equivalent to the total of monovalent
cations plus P5+. However in the few cases where H+ is measured along with Al and
the other cations, the total of H, Li, Na, K, and P are greater than Al and Fe by a factor
of 1.52 (Mller and Koch-Mller 2009). The correlation between Al and the
monovalent cations is displayed most strongly by low temperature hydrothermal
quartz, such as that from detrital overgrowths on sedimentary rocks, alpine fissure
quartz, and quartz from Mississippi Valley Type and epithermal ore bodies (Jourdan
et al. 2009; Lehmann et al. 2009; Rusk et al. 2011; Gtte et al. 2011) (Fig. 14.6).
In higher temperature quartz, the correlation commonly exists but is weaker, likely
owing to the relatively rapid and variable diffusion rates among Al and the
monovalent cations in quartz (c.f. Cherniak 2010).
320
B. Rusk
Fig. 14.6 SEM-CL image and LA-ICP-MS trace element maps of quartz from the McLaughlin,
CA epithermal deposit. Al concentrations are correlated with Li, Na, K, and Cs and are negatively
correlated with Ga. Ti is below detection limit. Although the CL image that corresponds with the
mapped area was lost in an unfortunate hardware malfunction, the CL image shown here is only
millimeters away from the location of the original trace element map, and the CL textures shown
here are nearly identical to those observed in the location of the trace element map. The boxed area
in the CL image closely approximates the CL texture where the trace element maps were obtained
Titanium is also common in hydrothermal quartz, and ranges from a few ppb to
a few 100 ppm. Titanium concentrations correlate with Al concentrations in some
cases, but as a whole in hydrothermal quartz, there is not a consistent correlation
between Ti and Al. Antimony is not typically measured or reported in quartz,
however it is present in concentrations of several 100 ppm in some inclusion-free
hydrothermal quartz from some low temperature (\300C) ore deposits (Rusk
et al. 2011). Phosphorous, B, Fe, Cl, and Ca are present in various quantities up to
a few 100 ppm in some hydrothermal quartz, but no correlations among these
elements or between them and any other elements have been recognized. Tin, Ge,
Ga, and As are typically present in concentrations below *5 ppm and also do not
consistently correlate with each other or with any other element.
14
321
Fig. 14.7 CL images of hydrothermal vein quartz and corresponding Al and Ti maps from a Los
Pelambres; and b El Teniente, porphyry Cu deposits, Chile. Trace element maps were acquired
by electron microprobe (cf. Rusk et al. 2008). Titanium concentrations correlate very closely with
CL intensity. The correlation between Al and CL intensity is also close
322
B. Rusk
Fig. 14.8 Monochromatic SEM-CL image with a false color spectral CL image. The
wavelengths of the CL emissions at various points are shown. Ti concentrations measured by
electron microprobe are also shown. There is a very close correlation between CL intensity and Ti
concentration. CL-bright quartz is characterized by Ti concentrations in the range of
*80160 ppm. Here the quartz shows a strong emission band in the range of 420480 nm as
well as another band in the range of 600 nm. The later, CL-darker quartz that is related to sulfide
deposition (not shown) contains less than 10 ppm Ti. In this quartz, the CL emissions are
dominated by the broad 600 nm peak and the 420480 nm peak is barely visible. These spectral
emissions and Ti concentrations are typical of vein quartz from porphyry copper deposits
including Butte, Bingham, Oyu Tolgoi, and El Slvador (Landtwing and Pettke
2005; Rusk et al. 2006; Mller et al. 2010; Donovan et al. 2011). Here the same
relationship is shown for vein quartz from the Los Pelambres and El Teniente
porphyry Cu deposits (Figs. 14.7 and 14.8). Other elements such as Fe, Ge, P, and
Sb do not show systematic relationships to CL intensity variations.
The strong correlation between CL intensity and Ti concentration in hydrothermal quartz, where Ti concentrations are above several ppm, suggests that the
substitution of Ti4+ for Si4+ in the quartz structure leads to increased CL intensity.
This hypothesis is supported by comparisons between quartz trace element concentrations and spectral maps of quartz CL emissions from numerous hydrothermal ore deposits. The 450 nm (*2.7 eV) emission band dominates in Ti-bearing
quartz (Fig. 14.8), such as from porphyry copper deposits, but it is small or absent
in low temperature quartz (\*300C), such as that from epithermal ore deposits
(Fig. 14.9). An emission band in this spectral range has also been shown to be
important in Ti-bearing igneous quartz (Mller et al. 2002; Wark and Spear
(2005). Further, in vein quartz from porphyry type deposits, the intensity of the
450 nm emission band correlates with Ti concentration such that in CL-bright
quartz with tens or hundreds of ppm Ti, the 450 nm peak dominates, but in
CL-darker quartz, containing only a few ppm Ti, the 450 nm band is present but
14
323
Fig. 14.9 Monochromatic SEM-CL image, Al map, false color CL image, and CL spectra of
hydrothermal breccia-infill quartz from the Jerritt Canyon Carlin-type Au deposit, NV. The CLbright euhedral quartz contains \100 ppm Al, and \2 ppm Ti. This quartz is characterized by a
broad peaks at 600 nm. CL-dark euhedral quartz is characterized by [3,000 ppm Al (and also
\2 ppm Ti) and the spectral emission is weak, but has a broad peak around 605 nm (not shown).
The CL-dark chert, which is a breccia fragment, shows similar luminescence characteristics to the
late CL-bright euhedral quartz, with a peak *605 nm
subordinate and a broad peak at 600 nm dominates. The broad 600 nm peak was
observed in hydrothermal quartz of all origins, regardless of temperature of formation, and dominates in low temperature quartz, such as that found in epithermal
deposits. This peak has not been recognized before and is interpreted here to be a
composite of overlapping peaks at *580 nm and *620650 nm, both of which
are common in quartz (see Gtte et al. 2011; Stevens-Kalceff 2009). A peak at
380 nm is common in Al-rich quartz and diminishes rapidly upon interaction with
the electron beam (Gtte et al. 2011; Stevens-Kalceff 2012). This peak was not
observed here, probably resulting from CL imaging by SEM prior to obtaining
spectral analyses, thus reducing the intensity of this peak.
324
B. Rusk
Fig. 14.10 Titanium versus Aluminum concentrations in hydrothermal quartz from epithermal,
orogenic Au, and porphyry-type deposits. The different deposit types can be distinguished from
one another based on trace element concentrations alone. Quartz trace elements reflect the
physio-chemical conditions of quartz growth and subsequent modifications
between these elements with an Al/Ti ratio between *1 and 10. In distinct contrast, quartz from epithermal ore deposits always contains less than about 3 ppm Ti
and Al concentrations range widely between *20 and 4,000 ppm Al. Al/Ti ratios
range widely between *100 and *10,000. Orogenic Au quartz is intermediate
between these two with Al concentrations typically between 100 and 1,000 and Ti
concentrations between 1 and 10 ppm. Al/Ti ratios vary from 10 to 100.
The systematic variation in trace element concentration among quartz crystallized under different physical and chemical conditions is evidence that the trace
elements incorporated in quartz reflect geologic conditions of formation, and
subsequent modifications. The relations between most quartz trace elements and
the conditions of formation are not well constrained, but recent studies suggest that
pressure, temperature (Thomas et al. 2010), fluid composition (Perny et al. 1992;
Rusk et al. 2008; Lehmann et al. 2011), and crystallization rate (Lowenstern and
Sinclair 1996) all influence quartz trace element composition. A major aim of
recent studies has been to determine quantitative links between the conditions of
quartz formation, and the resultant trace element endowment of the quartz. For
example the experimentally calibrated Ti in quartz thermobarometer, TitaniQ
(Wark and Watson 2006; Thomas et al. 2010), is commonly applied to understand
pressure and temperature variations in metamorphic, igneous, and hydrothermal
systems. In hydrothermal quartz, combined application of CL petrography and the
Ti in quartz thermobarometer (Figs. 14.7 and 14.8), shows that in many mineralized veins from porphyry copper deposits, chalcopyrite is related to late CL-dark
quartz that cuts earlier CL-bright quartz. Early CL bright quartz typically contains
between about 50 and 200 ppm Ti, whereas later sulfide-related CL-dark quartz
typically contains \10 ppm Ti. Rutile is common in early CL-bright quartz, but
less common in volumetrically minor later CL-dark quartz. Application of the Ti
14
325
14.4 Conclusions
Scanning electron microscope-cathodoluminescence textures provide unique
insights into the origin of hydrothermal quartz. These textures distinguish the
chronologic relations among quartz generations and can be used to relate fluid
inclusions, mineral precipitation, and isotopic data to specific mineralization
events. The CL textures and related trace element fluctuations in quartz reflect the
fluid pressure, temperature, and composition at the time of precipitation, as well as
the rate of fluctuation of these variables. Together CL textures and quartz trace
element compositions can be used to fingerprint the deposit type from which the
quartz likely originated. Since quartz is ubiquitous in hydrothermal ores and
resilient against physical and chemical weathering, this discrimination technique
can readily be applied to exploration for hydrothermal ore deposits.
Acknowledgments I would like to thank Mark Reed, David Krinsley, Alan Koenig, Heather
Lowers, Yi Hu, and Kevin Blake for stimulating discussions about quartz and quartz analysis. I
also thank Thomas Gtte and Jens Gtze for constructive reviews, which improved the content
and presentation of the manuscript.
References
Adams SF (1920) A microscopic study of vein quartz. Econ Geol 15:623664
Allan MM, Yardley BWD (2007) Tracking meteoric water infiltration into a magmatic
hydrothermal system: A cathodoluminescence, oxygen isotope, and trace element study of
quartz from Mt. Leyshon. Australia Chem Geol 240:343360. doi:10.1016/j.chemgeo.2007.
03.004
326
B. Rusk
14
327
Ioannou SE, Gtze J, Weiershuser L, Zubowski SM, Spooner ETC (2004) Cathodoluminescence
characteristics of Archean VMSrelated quartz: Noranda, Ben Nevis, and Matagami districs,
Abitibi subprovince. Canada Geochem Geophys Geosys. doi:10.1029/2003GC000613
Jacamon F, Larsen RB (2009) Trace element evolution of quartz in the charnockitic Kleivan
granite, SW Norway: the Ge/Ti ratio of quartz as an index of igneous differentiation. Lithos
107:281191
Jourdan A, Vennemann TW, Mullis J, Ramseyer K (2009) Oxygen isotope sector zoning in
natural hydrothermal quartz. Min Mag 73:615632
Kanaori Y (1986) A SEM cathodoluminescence study of quartz in mildly deformed granite from
the region of Atotsugawa fault, central Japan. Tectonophys 131:133146
King EM, Barrie CT, Valley JW (1997) Hydrothermal alteration of oxygen isotope ratios in
quartz phenocrysts, Kidd Creek mine, Ontario: magmatic values are preserved in zircon. Geol
25:10791082
Landtwing M, Pettke T (2005) Relationships between SEMcathodoluminescence response and
trace element composition of hydrothermal vein quartz. Am Min 90:122131
Landtwing M, Pettke T, Halter WE, Heinrich CA, Redmond PB, Einaudi MT, Kunze K (2005)
Copper deposition during quartz dissolution by cooling magmatic hydrothermal fluids: the
Bingham porphyry. Earth Planet Sci Lett 235:229243
Larsen RB, Henderson I, Ihlen PM, Jacamon F (2004) Distribution and petrogenetic behaviour of
trace elements in granitic quartz from South Norway. Contrib Mineral Petrol 147:615628
Larsen RB, Jacamon F, Krontz A (2009) Trace element chemistry and textures of quartz during
the magmatic-hydrothermal transition of Oslo Rift granites. Min Mag 73:691707
Leach DL, Marsh E, Emsbo P, Rombach CS, Kelley KD, Anthony M (2005) Nature of
Hydrothermal Fluids at the Shale-Hosted Red Dog Zn-Pb-Ag Deposits, Brooks Range,
Alaska. Econ Geol 99:14491480
Lehmann K, Berger A, Gtte T, Ramseyer K, Wiedenbeck M (2009) Growth related zonations in
authigenic and hydrothermal quartz characterized by SIMS-, EPMA-, SEM-CL- and SEMCC-imaging. Min Mag 73:633643
Lehmann K, Pettke T, Ramseyer K (2011) Significance of trace elements in syntaxial quartz
cement, Haushi group sandstones, Sultanate of Oman. Chem Geol 280:4757
Lowenstern JB, Sinclair WD (1996) Exsolved magmatic fluid and its role in the formation of
comblayered quartz at the Cretaceous Logtung W-Mo deposit, Yukon Territory, Canada.
Trans Roy Soc Edinburgh Earth Sci 87:291303
Lubben J (2004) Silicification across the Betze-Post carlin-type Au deposit; clues to ore fluid
properties and sources, northern Carlin Trend, Nevada. Unpublished masters thesis,
University of Nevada, Las Vegas
Machel HG, Burton EA (1991) Factors governing cathodoluminescence in calcite and dolomite,
and their implications for studies of carbonate diagenesis. In: Barker CE, Kopp OC (eds)
Luminescence microscopy and spectroscopy: qualitative and quantitative applications: SEPM
short course, Tulsa, 25:3757
Marshall DJ (1988) Cathodoluminescence of geological materials. Unwin Hyman, Boston
Miyoshi N, Yamaguchi Y, Makino K (2005) Successive zoning of Al and H in hydrothermal vein
quartz. Am Mineral 90:310315
Monecke T, Kempe U, Gtze J (2002) Genetic significance of the trace element content in
metamorphic and hydrothermal quartz: A reconnaissance study. Earth Planet Sci Let
202:709724
Mller A, Koch-Mller M (2009) Hydrogen speciation and trace element contents of igneous,
hydrothermal and metamorphic quartz from Norway. Min Mag 73:569583
Mller A, Wanvik J (2012) Petrological and chemical characterisation of high-purity quartz
deposits with examples from Norway, This volume
Mller A, Lennox P, Trzebski R (2002) Cathodoluminescence and micro-structural evidence for
crystallization and deformation processes of granites in the Eastern Lachlan Fold Belt (SE
Australia). Cont Min Pet 143:510524
328
B. Rusk
Mller A, Wiedenbeck M, Van den Kerkhof AM, Kronz A, Simon K (2003) Trace elements in
quartz: a combined electron microprobe, secondary ion mass spectrometry, laserablation
ICPMS, and cathodoluminescence study. Eur J Min 15:747763
Mller A, Breiter K, Seltman R, Pecskay Z (2005) Quartz and feldspar zoning in the Eastern
Erzgebirge pluton (Germany, Czech Republic): Evidence of multiple magma mixing. Lithos
80:201207
Mller A, Herrington R, Armstrong R, Seltman R, Kirwin D, Stenina N, Kronz A (2010) Trace
elements and cathodoluminescence of quartz in stockwork veins of Mongolian porphyry-style
deposits. Mineralium Deposita 45:707727
Mullis J, Dubessy J, Poty B, ONeil J (1994) Fluid regimes during late stages of a continental
collision: Physical, chemical, and stable isotope measurements of fluid inclusions in fissure
quartz from a geotraverse through the Central Alps, Switzerland. Geochim Cosmochim Acta
58:22392267
Pagel M, Barbin V, Blanc P, Ohnenstetter D (eds) (2000) Cathodoluminescence in geoscience.
Springer, Berlin, Heidelberg, New York, p 514
Penniston-Dorland SC (2001) Illumination of vein quartz textures in a porphyry copper ore
deposit using scanned cathodoluminescence: grasberg igneous complex, Irian Jaya, Indonesia.
Am Min 86:652666
Peppard BT, Steele IM, Davis AM, Wallace PJ, Anderson AT (2001) Zoned quartz phenocrysts
from the rhyolitic Bishop Tuff. Am Min 81:10341052
Perny B, Eberhardt P, Ramseyer K, Mullis J, Pankrath R (1992) Microdistribution of Al, Li, and
Na in alpha quartz: possible causes and correlation with short-lived cathodoluminescence. Am
Min 77:534544
Ramseyer K, Mullis J (2000) Geologic application of cathodoluminescence of silicates. In: Pagel
M, Barbin B, Blanc C, Ohnstetter (eds) Cathodoluminescence in geosciences. Springer,
Berlin, pp 177191
Ramseyer K, Baumann J, Matter A, Mullis J (1988) Cathodoluminescence colours of alphaquartz. Mineral Mag 52:669677
Redmond PB, Einaudi MT, Inan EE, Landtwing MR, Heinrich CA (2004) Copper deposition by
fluid cooling in intrusioncentered systems: new insights from the Bingham porphyry ore
deposit, Utah. Geol 32:217220
Richter DK, Gtte T, Gtze J, Neuser RD (2003) Progress in application of cathodoluminescence
in sedimentary petrology. Min Pet 79:127166
Rusk B (2009) Insights into hydrothermal processes from cathodoluminescence and trace
elements in quartz. In: Williams PJ, Rusk B, Oliver N (eds) Smart sciences for exploration
and mining, proceedings of the 10th Biennial SGA meeting. Townsville, pp 749751. https://
www.e-sga.org/index.php?id=231&tx_commerce_pi1[showUid]=1650&tx_commerce_pi1
[catUid]=43&cHash=3feeb527d1
Rusk BG, Reed MH (2002) Scanning electron microscopecathodoluminescence of quartz
reveals complex growth histories in veins from the Butte porphyry copper deposit, Montana.
Geol 30:727730
Rusk B, Reed M, Dilles J, Kent A (2006) Intensity of quartz cathodoluminescence and trace
element content of quartz from the porphyry copper deposit in Butte, Montana. Am Min
91:13001312
Rusk B, Lowers H, Reed M (2008) Trace elements in hydrothermal quartz; relationships to
cathodoluminescent textures and insights into hydrothermal processes. Geol 36:547550
Rusk B, Koenig A, Lowers H (2011) Visualizing trace element distribution in quartz using
cathodoluminescence, electron microprobe, and laser ablation inductively coupled plasma
mass spectrometry. Am Mineral 96:703708
Sekine K (2003) Development of fracture and fluid migration in granite during uplift and
emplacement. Dissertation, Tohoku University, p 256
Seyedolali A, Krinsley DH, Boggs S, OHara PF, Dypvik H, Goles GG (1997) Provenance
interpretation of quartz by scanning electron microscope-cathodoluminescence fabric
analysis. Geol 25:783786
14
329
Sippel RF (1968) Sandstone petrology, evidence from luminescence petrography. J Sed Pet 38:
530554
Smith JV, Stenstrom RC (1965) Electron-excited luminescence as a petrological tool. J Geol
73:627635
Spear FS, Wark DA (2009) Cathodoluminescence imaging and titanium thermometry in
metamorphic quartz. J Metamorph Geol 27:187205
Sprunt ES, Dengler LA, Sloan D (1978) Effects of metamorphism on quartz cathodoluminescence. Geol 6:305308
Spurr JE (1926) Successive banding around rock fragments in veins. Econ Geol 21:519537
Stenina NG (2004) Water-related defects in quartz. Bullet Geosci 79:251268
Stevens-Kalceff M (2009) Cathodoluminescence microcharacterization of point defects in aquartz. Min Mag 73:585605
Stevens-Kalceff MA (2012) Cathodoluminescence microanalysis of the defect microstructures of
bulk and nanoscale ultrapure silicondioxide polymorphs for device applications. (this volume)
Stevens-Kalceff MA, Phillips MR (1995) Cathodoluminescence microcharacterization of the
defect structure of quartz. Phys Rev B 52:31223134
Stevens-Kalceff MA, Phillips MR, Moon AR, Kalceff W (2000) Cathodoluminescence
microcharacterization of silicon dioxide polymorphs. In: Pagel M, Barbin B, Blanc C,
Ohnstetter D (eds) Cathodoluminescence in geosciences, Springer, Berlin, pp 193224
Tarashchan AN, Waychunas G (1995) Interpretation of luminescence spectra in terms of band
theory and crystal field theory. Sensitization and quenching, photoluminescence, radioluminescence, and cathodoluminescence. In: Marfunmin AS (ed) Advanced mineralogy 2,
Methods and instrumentations: results and recent developments. Springer, Berlin, pp 124135
Thomas SM, Koch-Mller M, Reichart P, Rhede D, Thomas R, Wirth R (2009) IR calibrations
for water determination in olivine, r-GeO2 and SiO2 polymorphs. Phys Chem Miner
36:489509
Thomas JB, Watson EB, Spear FS, Shemella PT, Nayak SK, Lanzirotti A (2010) TitaniQ under
pressure: the effect of pressure and temperature on the solubility of Ti in quartz. Contrib
Mineral Petrol 160:743759
Valley JW, Graham CM (1996) Ion microprobe analysis of oxygen isotope ratios in quartz from
Skye granite: healed microcracks, fluid flow, and hydrothermal exchange. Contrib Min Pet
124:225234
Vearncombe JR (1993) Quartz vein morphology and implications for formation depth and
classification of Archaean gold-vein deposits. Ore Geol Rev 8:407424
Walker G (2000) Physical parameters for the identification of luminescence centres in minerals.
In: Pagel M, Barbin B, Blanc C, Ohnstetter D (eds) Cathodoluminescence in geosciences.
Springer, Berlin, pp 2340
Wark DA, Spear FS (2005) Titanium in quartz: cathodoluminescence and thermometry. Geochim
Cosmochim Acta Suppl 69:A592
Wark DA, Watson BE (2006) TitaniQ: a titanium in quartz geothermometer. Contrib Min Pet
152:743754. doi:10.1007/s00410-006-0132-1
Wark DA, Hildreth W, Spear FS, Cherniak DJ, Watson EB (2007) Pre-eruption recharge of the
Bishop magma system. Geol 35:235238
Wilkinson JJ, Johnston JD (1996) Pressure fluctuations, phase separation, and gold precipitation
during seismic fracture propagation. Geol 24:395398
Wilkinson JJ, Boyce AJ, Earls G, Fallick AE (1999) Gold remobilization by lowtemperature
brines: evidence from the Curraghinalt gold deposit, northern Ireland. Econ Geol 94:289296
Chapter 15
Abstract Quartz represents one of the most widespread minerals and is widely
used in geosciences to reconstruct physic-chemical conditions of rock and mineral
formation. However, interpretation of analytical data may be limited by the ability
of quartz to regenerate during secondary alteration processes occurring under
metamorphic or hydrothermal conditions. This behaviour distinguishes quartz
from most minerals commonly associated with. Primary genetic information is
obliterated during quartz regeneration. This includes features related to the real
structure of quartz, but also to fluid and mineral inclusions. The present contribution examines examples covering various fields of mineral research, namely the
genetic interpretation of trace element content in quartz, quartz provenance
analysis using cathodoluminescence (CL) colour imaging, and the analysis of
mineral and fluid inclusions in quartz. It is demonstrated in all cases that care
331
332
U. Kempe et al.
15.1 Introduction
Quartz is one of the most common rock-forming minerals in the earth crust. Due to
its widespread occurrence and formation under various geological conditions, the
behaviour and real structure of quartz are an important focus of geoscientific
research. Quartz properties are studied in petroleum exploration and exploitation, in
mineral exploration, and in manufacturing industries using quartz as a raw material.
Despite the low variability in the main chemical and structural parameters, a
wide range of analytical methods and approaches is now routinely used in
investigations on quartz. These include trace element analysis (Bambauer 1961;
Dennen 1966; Novgorodova et al. 1984; Watt et al. 1997; Monecke et al. 2002;
Mller et al. 2003; Gtze et al. 2004; Larsen et al. 2004; Landtwing and Pettke
2005; Rusk et al. 2006; Breiter and Mller 2009; Whiting et al. 2010; Agangi et al.
2011; Gtte et al. 2011; Rusk et al. 2011), oxygen isotope analysis (Blatt 1987;
Girard and Deynoux 1991; Vennemann et al. 1992; Barton et al. 1992; Onasch and
Vennemann 1995; Williams et al. 1997; Jourdan et al. 2009), electron paramagnetic resonance (EPR) spectroscopy (Gtze and Pltze 1997; Gtze et al. 1999;
Botis et al. 2005, 2006), cathodoluminescence (CL) imaging (Seyedolali et al.
1997; Watt et al. 1997; Gtze 2000; Penniston-Dorland 2001; Rusk and Reed
2002; Larsen et al. 2004; Rusk et al. 2011), and CL spectroscopy (Claffy and
Ginther 1959; Ramseyer and Mullis 1990; Perny et al. 1992; Gtze 2000; Gtze
et al. 2001; Lehmann et al. 2009). Quartz is also used together with associated
minerals in the context of fluid inclusion analysis (e.g., Roedder 1984; Samson
et al. 2003) and melt inclusion studies (e.g., Frezzotti 2001; Webster 2006).
Genetic interpretation of data obtained by these techniques often relies on the
assumption that quartz is a resistant repository for genetic information and that
trace element content, CL characteristics, or melt and fluid inclusion inventory
directly relate to the environment of quartz formation. This view is based, at least
in part, on the relatively high hardness of quartz, the lack of a distinct cleavage,
and its resistance to weathering (e.g., Fchtbauer et al. 1982; Zuffa 1985;
Hallbauer 1992; Bodnar 2003b; Thomas et al. 2005).
However, growing evidence in the recent literature suggests that this assumption
is not always valid. Under some circumstances quartz is more susceptible to secondary alteration than many associated minerals. Alteration often proceeds without
significant destruction or modification of the quartz grains and aggregates. Quartz
shows a conspicuous tendency to structural regeneration. For instance, recrystallization, ductile deformation, healing of cracks or continuation of crystal growth
immediately after brittle deformation, as well as formation of secondary fluid
inclusions are common phenomena observed in quartz in magmatic, metamorphic,
15
333
and hydrothermal environments (e.g., Rusk and Reed 2002; Landtwing and Pettke
2005; Mller et al. 2010). Therefore, information obtained from the investigation of
quartz cannot always be simply related to the processes of primary quartz growth.
The behaviour of quartz under these conditions appears to differ from many other
common minerals.
The present study provides an overview over advantages and pitfalls of some of
the analytical techniques commonly used in the study of quartz. Four different case
studies are presented that either focused directly on the investigation of quartz (trace
element analysis and CL colour imaging) or considered the relationships of quartz to
associated minerals (fluid inclusion analysis and analysis of mineral sequences in
hydrothermal vein deposits). Although all four case studies had different objectives
and employed different analytical techniques, they all demonstrate that quartz does
not represent a stable repository of genetic information under metamorphic and
hydrothermal conditions but can be used for reconstruction of the alteration history
of the mineral.
334
U. Kempe et al.
15
Li
Al
Sr
Rb
Mn
335
Metamorphic
quartz
ED-1 (ppm)
Metamorphic quartz
in the alteration halo
ED-520 (ppm)
Vein and
pegmatite
quartz (ppm)
0.4
14.5
0.22
0.05
0.4
0.9
121
0.19
0.29
0.5
1560
150640
0.060.09
0.81.2
12
minerals and rocks from tin deposits (Monecke et al. 2011 and references therein).
Samples from the endocontact have a strong negative europium anomaly while
vein quartz from the exocontact displays a positive europium anomaly (Haler
et al. 2005).
The quartz sample from the metamorphic lens outside the alteration halo displays typical characteristics of metamorphic quartz with low contents of lithium,
aluminium, manganese, strontium, and rubidium. The sample close to the granite
contact has trace element characteristics that are intermediate between metamorphic quartz and quartz from the ore veins with elevated lithium, aluminium, and
rubidium concentrations (Monecke et al. 2002; cf. Table 15.1). The two quartz
samples collected from the wall rocks at different distances to the granite intrusion
also have distinct REE patterns (Fig. 15.1a). The metamorphic quartz sample
collected distal to the intrusion (ED-1) shows a distribution pattern typical of
metamorphic quartz samples with an enrichment of the light REEs over heavy
REEs, a positive cerium anomaly, and no significant anomaly of europium. In
contrast, the sample collected close to the granite (ED-520) has a small positive
cerium anomaly and a distinct positive europium anomaly. The heavy REEs are
distinctly enriched.
The observed differences in the trace element abundances suggest that the
metamorphic quartz lens close to the granite was overprinted by hydrothermal
fluids related to the tin mineralisation without visible changes in the metamorphic
fabrics. At the same time, the metamorphic quartz located at a larger distance was
not notably affected by hydrothermal alteration. Both samples show similar
microfabrics and mineral associations with the quartz being intergrown with
biotite, muscovite, minor garnet, albite, apatite, ilmenite, and zircon. The only
difference is the occurrence of rare inclusions of fluorite and of euhedral tourmaline in the altered quartz. Fluorite and tourmaline are also associated with the
tin mineralisation in the ore zones.
The conclusion that the metamorphic sample collected close to the granite
contact was affected by hydrothermal alteration is also supported by CL microscopy. The metamorphic quartz sampled away from the granite contact showed
mostly stable brownish CL typical of metamorphic quartz, while the hydrothermally altered quartz has a transient bluish CL that is replaced by brownish CL
during electron irradiation. Such behaviour is characteristic for some types of
hydrothermal quartz (Fig. 15.2 see below).
336
U. Kempe et al.
Fig. 15.1 REE distribution patterns of quartz from lenses in the metamorphic host rocks and of
fluorite from an early tourmaline veinlet (Ehrenfriedersdorf tin deposit, Erzgebirge, Germany).
a Comparison of REE distribution patterns of quartz from metamorphic quartz lenses distal
(ED-1) and proximal (ED-520) to the granite intrusion and ore zones. b REE distribution pattern
for the hydrothermal fluid, calculated from the difference between the patterns of the distal and
proximal quartz samples, in comparison to fluorite from an early tourmaline veinlet (ED-268).
ICP-MS data of quartz from Monecke et al. (2002). ICP-MS data of fluorite from Dulski and
Kempe (unpublished)
Fig. 15.2 Optical microscope-based cathodoluminescence (CL) images of quartz from lenses
hosted by the metamorphic wall rocks of the Ehrenfriedersdorf tin deposit, Germany. a Brownish
and relatively homogeneous CL of quartz collected distal to the granite and tin ore zones (ED-1).
b Remnants of transient bluish CL of quartz sample collected close to the granite (ED-520). Note
the heterogeneous distribution of colour intensity. Bright bluish spots in the lower left corner of
the image are small inclusions of fluorite. Both images were taken after 1 min of irradiation time.
Initial strong bluish transient CL in the hydrothermally altered quartz (ED-520) could be
observed, but not imaged due to the comparably long exposure time required. Scale bar is 0.3 mm
Assuming that the REE patterns of both metamorphic quartz lenses investigated
were identical prior to hydrothermal alteration of the quartz lens located proximal
to the granite, the differences between both patterns are a direct reflection of the
15
337
REE signature of the hydrothermal fluids. The REE pattern of the hydrothermal
fluids reconstructed this way (except for cerium which is partly removed during
alteration) shows a depletion of the light REEs, strong enrichment of the heavy
REEs, and a positive europium anomaly (Fig. 15.1b). The calculated REE pattern
of the hydrothermal fluid is similar to that of fluorite occurring in cassiteritebearing tourmaline veinlets, which represent the earliest set of hydrothermal veins
formed during ore formation (Monecke et al. 2000b, 2011). The REE pattern of the
fluorite has been interpreted to record the trace element signature of the pristine
ore-forming hydrothermal fluids prior to extensive interaction with the metamorphic host rocks (Gavrilenko et al. 1997; Monecke et al. 2000b, 2011).
It is important to note that the observed differences between the two quartz
samples from the metamorphic wall rocks could not have been interpreted
correctly without integration of observations made by optical microscopy, scanning electron microscopy, and cathodoluminescence microscopy. Hydrothermal
alteration of quartz not only resulted in distinct changes in the trace element
signature, but also modified the CL properties of the quartz. The occurrence of new
accessory minerals provides additional evidence corroborating the conclusion that
the metamorphic quartz lens close to the granite contact was affected by a process
of hydrothermal alteration. The CL observations suggest that the entire volume of
the metamorphic quartz lens was affected by this process. Although changes in the
trace element contents such as those of the REEs may be assigned to changes in
the fluid inclusion inventory, variations in the contents of trace elements commonly located in the crystal structure and changes in the CL characteristics suggest
that alteration was not restricted to the formation of secondary fluids or the
absorption of elements to internal and external surfaces. Hydrothermal alteration
changed the quartz characteristics without visible deformation or destruction of the
quartz lens.
338
U. Kempe et al.
appears brownish under CL, but (4) authigenic quartz mostly exhibits weak or no
visible CL but may luminesc in the UV range. Finally, hydrothermal quartz
(5) often (but by far not always) shows transient greenish to blue and sometimes
yellowish luminescence (Gtze 2000; Gtze and Zimmerle 2000; Gtze et al. 2001
and references therein).
However, recent work indicates that such a rigid classification based on
empirical evidence may be an oversimplification. The CL colours of quartz of
different origins may show more variability, requiring either changes to the classification scheme or independent tests establishing the validity of the interpreted
provenance (e.g., Boggs et al. 2002). In addition, post-depositional processes may
affect CL characteristics. A study on quartz from the gold-bearing reefs of the
Witwatersrand, South Africa, provides evidence that secondary processes can
indeed modify the CL behaviour of quartz in sedimentary rocks.
Over the past decades, the genesis of the gold mineralisation in the reefs has
been the focus of much debate. Traditionally, the gold deposits were regarded as
conglomerate-hosted paleoplacers (e.g. Robb and Meyer 1990; Minter et al. 1993).
However, abundant evidence points to the importance of hydrothermal activity
(e.g., gold morphology and composition, formation of pyrophyllite and chlorite,
association with sulphides and uraninite with high Th, Y, and REE contents;
Barnicoat et al. 1997; Vollbrecht et al. 2002), implying that the gold is of
hydrothermal origin (e.g., Phillips and Myers 1989; Barnicoat et al. 1997; Phillips
and Law 2000; Law and Phillips 2005). Considering the detrital nature of some
minerals and the fact that the host rocks are conglomerates, most of recent authors
prefer a modified placer model that assumes a detrital genesis of quartz, pyrite,
gold, and uranium with a late hydrothermal overprint and remobilisation of at least
some of the gold (Robb and Meyer 1991; Frimmel and Gartz 1997; Gartz and
Frimmel 1999; Vollbrecht et al. 2002; Frimmel et al. 2005; Schaefer et al. 2010
and references therein). Some authors proposed possible relationships between the
gold and the Vredefort impact event (Kamo et al. 1996; Gartz and Frimmel 1999;
Gibson and Reimold 1999).
To constrain the provenance of quartz from the reefs, Vollbrecht et al. (1996,
2002) used colour imaging by CL microscopy. These studies focused on the sand
fraction (\5 mm) and cements in conglomerates from reefs in the ore fields.
Vollbrecht et al. (1996, 2002) found that mono- or polycrystalline quartz displaying bluish CL colours and with rare growth zoning is the most abundant type.
This prevailing quartz type was interpreted to be magmatic in origin. It may have
been derived from plutons, pegmatites, and (rare) felsic porphyries located in the
hinterland of the Witwatersrand basin. Brownish luminescent mono- and polycrystalline grains were less abundant, but still common, and were interpreted to
have been derived from metamorphic rocks. Hydrothermal quartz identified by
short-lived luminescence was found only in rare cases. Vollbrecht et al. (2002)
also recognised the occurrence of secondary alteration features introduced during
ductile and brittle deformation (undulate extinction, deformation lamella, pressure
solution, fragmentation, and shearing) and hydrothermal alteration (replacement,
alteration rims, and rims from radioactive irradiation).
15
339
A study on quartz pebbles from mineralised and barren parts of the Dominion,
Vaal, Ventersdorp Contact, and Black Reefs was carried out by Poutivtsev (2001),
Poutivtsev et al. (2001), and Kremenetsky et al. (2005). These studies employed
the same CL techniques as used by Vollbrecht et al. (1996, 2002) but also included
trace element analyses and analytical scanning electron microscopy. The trace
element characteristics of quartz from the pebbles were found to be similar to
those from mesothermal gold vein deposits. In addition, it was noted that cements
in the mineralised parts of the reefs contain pyrrhotite, a sulphide commonly found
in such gold deposits, and is absent in the barren parts of the reefs (Poutivtsev et al.
2001).
The CL imaging of the quartz pebbles yielded results that were generally
comparable to the observations on the sand fraction made by Vollbrecht et al.
(1996, 2002). The most common quartz CL was stable bluish to violet. Primary
growth zoning was occasionally found. Quartz showing a brownish luminescence
was less abundant, but still common. Transient bluish luminescence, with the CL
colour changing to brown during continuous electron irradiation, was observed
only in one quartz sample. In the gold-bearing conglomerates, orange luminescent
rims on the pebbles were observed that formed as a result of radioactive irradiation
of quartz. Rims with radiation-induced CL were less common in the unmineralised
parts of the reefs.
Based on the observations by Vollbrecht et al. (1996, 2002), Poutivtsev (2001),
Poutivtsev et al. (2001), and Kremenetsky et al. (2005), it can be concluded that
quartz pebbles and quartz from the sand fraction are mainly magmatic in origin
with subordinate but significant amounts derived from metamorphic rocks.
However, more detailed inspection of the mineralised conglomerates from the
Witwatersrand revealed that deformation and widespread hydrothermal alteration
modified the CL behaviour of the quartz (Fig. 15.3). Larger quartz grains and
pebbles showed evidence for the occurrence of processes of grain size reduction,
in particular along their margins. Recrystallisation and development of subgrains
(Fig. 15.3a, b) contributed to the overall grain size reduction. Bluish luminescent
quartz grains commonly showed deformation lamellae that are clearly recognisable under CL as narrow, parallel zones of brownish CL (Fig. 15.3c). In addition,
zones of microbreccia are observed (Fig. 15.3c, d). Brittle and plastic quartz
deformations are thus clearly documented. The reduction in grain size and the
formation of subgrains and quartz fragments was accompanied by a decrease in CL
intensity and by a change in CL colour from bluish through dull violet to brownish.
Evidence for changes in CL colour as a result of post-depositional fluid flow is
provided by the occurrence of networks of secondary fluid inclusions trails that are
surrounded by large halos and irregular trails of brownish CL (Fig. 15.3a, c, e).
Embayed grain boundaries that formed by processes of quartz dissolution are often
paralleled by rims of orange and zoned orange/bluish luminescence (Fig. 15.3c, e).
As noted by previous workers, this change in CL colour is probably related to
radioactive irradiation of the quartz during circulation of uranium-rich hydrothermal fluids. Similar zones of modified CL are also observed around inclusions
of uranium-rich minerals.
340
U. Kempe et al.
Fig. 15.3 Optical microscope-based CL (left) and transmitted light crossed polar (right) images
of quartz pebbles from reefs of the Witwatersrand, South Africa. a, b Fluid trails (left side,
brownish CL) in a fragmented quartz pebble; recrystallisation and formation of smaller grains and
subgrains (right side; bluish, dull violet to brownish CL); dissolution embayed grain boundaries,
sometimes with orange luminescent rims from radioactive irradiation. Sample W63A. Goldbearing conglomerate from the Ventersdorp Contact Reef, East Drifontein; c, d Deformation
lamellae (upper left and right), fluid trails (left, right, brownish CL, orange CL around inclusions
in the trails) and local brecciation (middle, brownish CL) in a quartz pebble; orange luminescent
rims from radioactive irradiation. Sample 20-1a. Barren conglomerate from the Vaal Reef,
Johannesburg, Vickers Road; e, f Fragmentation at the rim of a quartz pebble and fluid inclusion
trails (brownish CL). Sample BR1/2. Barren conglomerate from the Black Reef near Klerkskraal.
Figures 15.4cf after Poutivtsev (2001). Scale bar is 0.3 mm
15
341
Obviously, the texture of the quartz pebbles, their CL properties, trace element
contents, and fluid inclusion characteristics were changed under ductile and brittle
deformation and during hydrothermal alteration. These observations may satisfactorily explain the hydrothermal trace element signature of the quartz pebbles
observed by previous workers (Poutivtsev 2001; Poutivtsev et al. 2001; Kremenetsky et al. 2005). The identified ductile and brittle deformation features
(Vollbrecht et al. 1996, 2002; this work) possibly correspond to thrust-related
fracture networks observed by Jolley et al. (1999) and Gartz and Frimmel (1999)
for the Ventersdorp Contact Reef. Together with these observations made at the
macro- and microscales, new findings may open ways to a more comprehensive
interpretation of the provenance of the conglomerates and their response to
deformation by thrusting and/or a meteorite impact.
The textural relationships found for the quartz pebbles suggest that the
brownish CL colour of quartz from the Witwatersrand is not primary but also does
not result simply from low-grade regional metamorphism. A similar transformation from bluish/violet to brownish CL was also observed for quartz in goldbearing brecciated quartz bodies from Muruntau (Uzbekistan). In analogy to
samples from the Witwatersrand, the vein quartz at Muruntau was affected by
ductile/brittle deformation and hydrothermal alteration at elevated temperatures
(Graupner et al. 2000). Furthermore, observations by Sprunt et al. (1978) demonstrate that even the bluish CL of quartz is not always related to a magmatic
origin of the quartz as high-grade metamorphism can also produce this CL colour.
Additional evidence for the ambiguous nature of blue and brownish CL of quartz
were presented by Boggs et al. (2002) and Spear and Wark (2009).
Summarising, the results of the CL study on quartz pebbles from the Witwatersrand basin show that provenance analysis is not possible using this technique in
cases where quartz experienced regional or dynamic metamorphism and/or intense
hydrothermal alteration. This conclusion is consistent with suggestions made by
Boggs et al. (2002).
342
U. Kempe et al.
Fig. 15.4 Summary of fluid inclusion investigations on minerals from the quartz-wolframite vein
deposits Nuurijn gol and Khovd gol, Mongolian Altai. Several fluid inclusion types may be
distinguished which occur as primary, pseudo-secondary, and secondary inclusions in wolframite,
quartz, fluorite, scheelite, and siderite. Note the rare occurrence of Type Ia inclusions as primary
inclusions in quartz from Nuurijn gol and of Type II inclusions in quartz from Khovd gol. See
text for further explanation. Data from Enchbat (2007)
Based on textural criteria, fluid inclusions in minerals are conveniently classified as primary, pseudosecondary, and secondary inclusions (Roedder 1984;
Van den Kerkhof and Hein 2001; Bodnar 2003a). As shown by these authors, a
correct classification may be possible only for a small part of the fluid inclusion
inventory as unequivocal textural evidence may be missing. Additional limitations
are imposed by the fact that quartz often contains a large variability and high
quantity of fluid inclusions assemblages. Interpretation of fluid inclusion data may
also be limited due to the occurrence of reequilibration phenomena during
hydrothermal alteration (e.g., Bakker 2009). A study on two tungsten deposits in
the Mongolian Altai in the western Mongolia (Enchbat et al. 1999; Enchbat 2007)
highlights some of the pitfalls and difficulties arising in the study of fluid
inclusions in quartz.
Enchbat et al. (1999) and Enchbat (2007) studied the fluid inclusion inventory
of various vein minerals occurring in the quartz-wolframite veins of the Nuurijn
gol and Khovd gol deposits to better constrain the conditions of tungsten mineralisation in the Mongolian Altai. Based on field observations and textural evidence
collected in thin section, the sequence of mineral formation in the veins was
determined (Fig. 15.4). Two fluid inclusion types could be distinguished in both
15
343
deposits: (1) low (to intermediate) saline, two-phase inclusions with NaCl being
the dominant salt (Type I) and (2) low (to intermediate) saline, two- and threephase CO2-bearing inclusions with NaCl also being the dominant salt (Type II).
Based on the degree of filling and the related homogenisation temperatures Th, the
first type may be further subdivided into Type Ia (filling below 90%) and Type Ib
(filling above 90%; Fig. 15.4). Raman spectroscopy showed that Type II fluid
inclusions sometimes contain traces of N2 in the vapour phase (Enchbat 2007).
Comparison of vein quartz with the other vein minerals shows that quartz
contains the highest total abundance of fluid inclusions. The quartz is also typified
by more complex fluid inclusion characteristics (Fig. 15.4). It proved to be difficult
to distinguish between inclusions in quartz formed during primary growth or
secondary processes. Fluid inclusion assemblages of both CO2-bearing and CO2barren types (Type I and Type II, respectively) appear to have been entrapped
during primary quartz growth. This finding is somewhat unexpected as the general
importance of CO2 in wolframite precipitation is widely discussed in the literature
(Higgins 1980, 1985; Campbell et al. 1988; Graupner et al. 1999). The role of both
fluid types in ore formation is therefore fundamental for a correct understanding of
hydrothermal processes in both deposits.
To explain the apparently primary occurrence of both inclusion types in quartz,
several hypotheses were tested. There were no direct indications for heterogeneous
fluid entrapment as all inclusions in one assemblage are of the same type. Alternatively, the occurrence of primary and pseudosecondary inclusions of both types
could be the result of fluid evolution that paralleled quartz formation. However,
investigation on a vein profile from Khovd gol failed to provide support for such a
hypothesis (Enchbat 2007). It was further tested whether textural relationships
were misinterpreted with a resulting incorrect classification of the inclusions to be
primary or pseudosecondary. The nature of inclusions in quartz was evaluated by
CL imaging. As shown by Roedder (1984), Boiron et al. (1992), Graupner et al.
(2000), and Van den Kerkhof and Hein (2001) as well as by the study of quartz
pebbles of the Witwatersrand discussed above, optical CL microscopy or scanning
electron microscopy often reveals complex networks of secondary alteration trails
in quartz to which fluid inclusions are confined and which may therefore be
recognised as secondary in origin. No clear indications for a secondary origin of
fluid inclusions were found by CL imaging for quartz from the Khovd gol deposit.
The occurrence of fluid inclusions in planes parallel to visible growth zones of a
crystal is usually a reliable criterion for a primary origin (e.g., Roedder 1984;
Fig. 15.4 and 15.5). However, as shown by Goldstein ( 2001) for carbonates,
secondary alteration can sometimes proceed along certain growth zones resulting
in secondary formation of fluid inclusions. A study on the Muruntau gold deposit
in Uzbekistan demonstrated that similar situations may also occur in quartz
(Graupner et al. 2000). However, no indication for a secondary formation of
inclusions arranged along primary growth zones could be found in quartz from
Nuurijn gol and Khovd gol (Fig. 15.5).
344
U. Kempe et al.
Fig. 15.5 Microphotographs of fluid inclusion assemblages in quartz from Khovd gol,
Mongolian Altai. a Typ I fluid inclusions arranged parallel to growth zones in quartz;
b Close-up image of a group of fluid inclusions of Type I. Modified from Enchbat (2007)
15
345
that CO2 was a major component of the fluids precipitating wolframite, which is
not supported by direct investigations on wolframite. Results from similar studies
on quartz and wolframite from deposits in Cornwall, Germany, Portugal, New
Mexico, and Peru by Campbell and Robinson-Cook (1987), Campbell et al.
(1988), Campbell and Panter (1990), and Lders (1996) also demonstrated that
inclusions in both minerals may be distinct. Differences between the fluid inclusions in quartz and ore minerals are not restricted to tungsten deposits as recently
shown by Kouzmanov et al. ( 2010). These authors documented similar relationships for quartz and pyrite from the Rosia Poieni porphyry deposit, Romania.
In the context of the present study, the question arises why fluid inclusion
assemblages in quartz can be so complex. Several important aspects need to be
considered. Although quartz behaves brittle during deformation accompanying
vein formation, vein quartz commonly shows features related to dissolution,
recrystallisation, and repeated growth. This behaviour of quartz during vein
formation is favourable for the repeated formation of fluid inclusions in the mineral
by crack-healing. However, by far not all fluid inclusion trails observed in quartz
may simply be related to brittle deformation. Large alteration zones containing
fluid inclusions are often revealed by CL imaging (e.g., Rusk and Reed 2002;
Fig. 15.3). It appears possible that the quartz structure may yield pathways for
penetration of fluids without previous fragmentation or dissolution. Recent
investigations on re-equilibration (e.g., stretching, leakage, decrepitation, necking
down) of fluid inclusions in quartz by Bodnar (2003b) and Bakker (2009) support
and necking down such a view.
346
U. Kempe et al.
1999). For instance, Kempe et al. (1999) reported on the secondary formation of
euhedral albite laths within K-feldspar and quartz grains in hydrothermally altered
alkaline rocks. Similarly, euhedral tourmaline crystals in hydrothermally altered
quartz occurring within the metamorphic host rocks from the Ehrenfriedersdorf tin
deposit clearly postdate quartz formation (see above). An additional example of
such a textural relationship between vein quartz and euhedral ore minerals is
provided by the South Crofty mine near Pool, Cornwall (Gavrilenko et al. 1998).
The quartz-wolframite veins at South Crofty contain pseudomorphs of scheelite
after wolframite, which are frequently overgrown by euhedral scheelite
(Fig. 15.6a). Euhedral scheelite grains that are fully enclosed by quartz also occur
abundantly in the veins. Crystal faces are always developed at the grain boundaries
with quartz (Fig. 15.6a). The scheelite forming the euhedral crystals and the
scheelite within the pseudomorphs display different CL characteristics and internal
structures, suggesting variations in the mechanisms of crystal growth. The CL
spectra of both scheelite types are distinct (Gavrilenko et al. 1998). Visible
oscillatory growth zoning in the euhedral scheelite crystals forming the overgrowth are the result of dynamic processes common during primary crystal growth
(cf. Shore and Fowler 1996 and references therein). These observations along with
the occurrence of scheelite as inclusions in quartz may be taken as evidence for the
following sequence of events: (1) crystallisation of wolframite, (2) brittle deformation and formation of pseudomorphs of scheelite after wolframite, (3) growth of
euhedral scheelite as an overgrowth, (4) precipitation of quartz surrounding and
enclosing the tungsten minerals.
Closer examination, however, suggests a different interpretation. Some scheelite crystals enclosed in quartz are arranged in trails. These trails commonly appear
to have developed along fractures in the quartz (Fig. 15.6a) and must, therefore,
15
347
348
U. Kempe et al.
metamorphic quartz must have affected its real structure through the formation of
new ionic and electron defects which define the CL behaviour of quartz (Gtze
et al. 2001; Larsen et al. 2009).
The case studies at the Ehrenfriedersdorf tin deposit, Germany, and the Witwatersrand, South Africa, show that the CL colour characteristics of quartz can
change in response to deformation and hydrothermal alteration. Due to potential
post-depositional changes in CL colour, this property of quartz cannot always be
used to unequivocally constrain provenance of quartz grains in sediments and
sedimentary rocks. The results of CL investigations need to be tested against data
obtained from complementary techniques. In the case of the conglomerates from
the Witwatersrand basin, CL colour imaging of quartz from the large pebbles and
the sand-sized fractions suggests that the quartz has been largely derived from
magmatic and metamorphic rocks. However, more detailed CL studies, compounded by trace element analysis, suggest that the brownish CL interpreted to be
indicative for a metamorphic origin of some quartz is in fact introduced during
post-depositional deformation and hydrothermal alteration. The conclusion that
brown CL of quartz can result from secondary processes is in accordance with
earlier observations made on brecciated and altered gold quartz veins at the
Muruntau gold deposit, Uzbekistan (Graupner et al. 2000). Furthermore, it appears
possible that even the bluish CL, which is widely regarded to be indicative for a
magmatic origin of the quartz, is not of primary origin. Although in a different
geological environment, dark blue CL in quartz has been shown to develop in
response to a high-grade metamorphic overprint (e.g., Sprunt et al. 1978; Boggs
et al. 2002). Based on the evidence, it is suggested here that the CL characteristics
of quartz from the Witwatersrand reefs are probably of secondary origin and
should, therefore, not be used to interpret primary provenance of the conglomerate.
The formation of secondary fluid inclusions in quartz is a common phenomenon, providing evidence for fluid flow through quartz crystals after their formation
(Van den Kerkhof and Hein 2001; Rusk and Reed 2002). As demonstrated in this
study and in the literature, CL imaging can be used to identify pathways of fluid
flow through quartz crystals (Van den Kerkhof and Hein 2001; Rusk and Reed
2002), including those that are not marked by secondary fluid inclusions (this
study). Recent studies also suggest that primary fluid inclusions hosted by quartz
may be affected by secondary processes through re-equilibration (Van den Kerkhof
and Hein 2001; Bodnar 2003a, b; Bakker 2009). These changes occur without
obvious quartz degeneration as brittle deformation, fragmentation, or dissolution
(i.e. they occur when quartz remained in solid state). Fluid inclusion data can only
be interpreted correctly if primary, pseudosecondary, and secondary fluid inclusion
assemblages can be confidently distinguished, which is not always trivial or
possible without the help of advanced analytical techniques. Moreover, establishment of paragenetic relationships between quartz and ore minerals is critical.
The nature of the ore-forming fluid can only be constrained if it can be demonstrated that the quartz and ore minerals formed contemporaneously. The case study
on the quartz-wolframite vein deposits in the Mongolian Altai showed that the
results of fluid inclusion studies on quartz alone would have suggested that CO2
15
349
350
U. Kempe et al.
1983; Doukhan and Trpied 1988). Formation of small water inclusions in the
synthetic quartz was observed after the experiments (Doukhan and Trpied 1988).
Experimental investigations on diffusion and weakening using synthetic quartz
provide critical insights into the mechanisms leading to changes in water content
and mechanical behaviour of natural quartz, the formation of sub-grain boundaries,
and the origin of fluid inclusions at elevated temperatures and/or pressure. One
important point is that water is not only confined to the structural channels but can
also hydrolyse siliconoxygen bounds (Griggs and Blacic 1965). The conditions at
which diffusion and hydrolytic weakening of quartz are observed in the experiments correspond to the temperaturepressure ranges at which transformation of
quartz is observable in nature.
In summary, this contribution shows that quartz represents a mineral that is
quite susceptible to recrystallisation and structural changes under metamorphic
and hydrothermal conditions. This behaviour of quartz may be effectively used to
identify and monitor secondary alteration processes. At the same time, care needs
to be taken to ensure that primary and secondary characteristics are correctly
identified and distinguished. Future research on quartz would undoubtedly benefit
from a better integration of analytical data obtained by complementary techniques.
Acknowledgments We gratefully acknowledge reviews by Torsten Graupner and Brian Rusk,
which helped us to significantly improve the manuscript. We thank Dieter Wolf for useful
discussion on the behaviour of quartz. The trace element analyses would not have been possible
without the analytical support by Gisela Bombach and Werner Klemm.
References
Agangi A, McPhie J, Kamenetsky VS (2011) Magma chamber dynamics in a silicic LIP revealed
by quartz: the Mesoproterozoic Gawler Range Volcanics. Lithos 126:6883
Agel A, Petrov I (1990) Substitutional aluminium in the quartz lattice as indicator for the
temperature of formation. Eur J Mineral 2 (Bh.1):144 (abstract, in German)
Bakker RJ (2009) Reequilibration of fluid inclusions: bulk diffusion. Lithos 112:277288
Bambauer HU (1961) Spurenelementgehalte und c-Farbzentren in Quarzen aus Zerrklften der
Schweizer Alpen. Schweiz Mineral Petrol Mitt 41:335369
Barnicoat AC, Henderson IHC, Knipe RJ, Yardley BWD, Napier RW, Fox NPC, Kenyon AK,
Mutingh DJ, Strydom D, Winkler KS, Lawrence SR, Cornford C (1997) Hydrothermal gold
mineralization in the Witwatersrand basin. Nature 386:820824
Barton JM, Wenner DB, Hallbauer DK (1992) Oxygen isotopic study of the nature and
provenance of large quartz and chert clasts in gold-bearing conglomerates of South Africa.
Geology 20:11231126
Blatt H (1987) Oxygen isotopes and the origin of quartz. J Sediment Petrol 57:373377
Bodnar RJ (2003a) Introduction to fluid inclusions. In: Samson I, Anderson A, Marshall D (eds)
Fluid inclusions: analysis and interpretation. Mineral Assoc Canada, Short Course vol 32,
Quebec, pp 18
Bodnar RJ (2003b) Reequilibration of fluid inclusions. In: Samson I, Anderson A, Marshall D
(eds) Fluid inclusions: analysis and interpretation. Mineral Assoc Canada, Short Course vol
32, Quebec, pp 213230
15
351
Boggs S, Kwon YI, Goles GG, Rusk BG, Krinsley D, Seyedolali A (2002) Is quartz
cathodoluminescence color a reliable provenance tool? A quantitative examination. J Sediment
Res 72:408415
Boiron MC, Essarraj S, Sellier E, Cathelineau M, Lespinasse M, Poty B (1992) Identification of
fluid inclusions in relation to their host microstructural domains in quartz by cathodoluminescence. Geochim Cosmochim Acta 56:175185
Botis S, Nokhrin SM, Pan Y, Xu Y, Bonli T (2005) Natural radiation-induced damage in quartz.
I. Correlations between cathodoluminescence colors and paramagnetic defects. Can Mineral
43:15651580
Botis S, Pan Y, Bonli T, Xu Y, Zhang A, Nokhrin S, Sopuck V (2006) Natural radiation-induced
damage in quartz. II. Distribution and implications for uranium mineralization in the
Athabasca basin, Saskatchewan, Canada. Can Mineral 44:13871402
Bottrell SH, Yardley B, Buckley F (1988) A modified crush-leach method for the analysis of fluid
inclusion electrolytes. Bull Minral 111:279290
Breiter K, Mller A (2009) Evolution of rare-metal granitic magmas documented by quartz
chemistry. Eur J Mineral 21:335346
Campbell AR, Panter KS (1990) Comparison of fluid inclusions in coexisting (cogenetic?)
wolframite, cassiterite, and quartz from St. Michaels Mount and Cligga Head, Cornwall
England. Geochim Cosmochim Acta 54:673681
Campbell AR, Robinson-Cook S (1987) Infrared fluid inclusion microthermometry on coexisting
wolframite and quartz. Econ Geol 82:16401645
Campbell AR, Robinson-Cook S, Amindays C (1988) Observation of fluid inclusions in
wolframite from Panasqueira, Portugal. Bull Minral 111:251256
Charoy B, Noronha F (1996) Multistage growth of a rare-element, volatile-rich microgranite at
Argemela (Portugal). J Petrol 37:7394
Cherniak DJ (2010) Diffusion in quartz, melilite, silicate perovskite, and mullite. In: Zhang Y,
Cherniak DJ (eds) Diffusion in minerals and melts. Rev Mineral Geochem 72:735756
Claffy EW, Ginther RJ (1959) Red-luminescing quartz. Am Mineral 44:987994
Dennen WH (1966) Stoichiometric substitution in natural quartz. Geochim Cosmochim Acta
30:12351241
Dennen WH, Blackburn WN, Quesada A (1970) Aluminium in quartz as a geothermometer.
Contrib Mineral Petrol 27:332342
Doukhan JC, Trpied L (1988) Plastic deformation of quartz single crystals. Bull Minral
108:97123
Enchbat D (2007) Die erzbildenden Fluide in den Au-W-Mineralisationen des Mongolischen
Altai: Untersuchungen zu Fluideinschlssen und Spurenelementchemismus von Erz- und
Gangmineralen. Freiberg Forschungsh C520, TU Bergakademie Freiberg, 101 pp
Enchbat D, Kempe U, Dandar S, Wolf D (1999) Fluid inclusion characteristics and trace element
chemistry of ore and vein minerals from the Au-W deposits in Altai tectonic zone of
Mongolian Altai. In: Lders V, Schmidt-Mumm A, Thomas R (eds) ECROFI XV: European
current research on fluid inclusions: abstracts and program, vol 99(6). Alfred-WegenerStiftung, Potsdam, Terra nostra, pp 9394
Frezzotti ML (2001) Silicate-melt inclusions in magmatic rocks: applications to petrology. Lithos
55:273299
Frimmel HE, Gartz VH (1997) Witwatersrand gold particle chemistry matches model of
metamorphosed, hydrothermally altered placer deposits. Mineral Deposita 32:523530
Frimmel HE, Groves DI, Kirk J, Ruiz J, Chesley J, Minter WEL (2005) The formation and
preservation of the Witwatersrand gold fields, the worlds largest gold province. Econ Geol
100th Anniversary Volume:769797
Fchtbauer H, Leggewie R, Gockeln C, Heinemann C, Schrder P (1982) Methoden der
Quarzuntersuchung, angewandt auf mesozoische und pleistozne Sandsteine und Sande. N Jb
Geol Palont Mh 193210
Gartz VH, Frimmel HE (1999) Complex metasomatism of an Archean placer in the Witwatersrand
basin, South Africa: the Ventersdorp Contact reefa hydrothermal aquifer? Econ Geol 94:689706
352
U. Kempe et al.
15
353
Lonergan L, Wilkinson JJ (eds) Fractures, fluid flow and mineralization. Geol Soc, vol 155.
Spec Publ, London, pp 153165
Jourdan AL, Vennemann TW, Mullis J, Ramseyer K (2009) Oxygen isotope sector zoning in
natural hydrothermal quartz. Mineral Mag 73:615632
Kamo SL, Reimold WU, Krogh TE, Colliston WP (1996) A 2.023 Ga age for the Vredefort
impact event and a first report of shock metamorphosed zircons in pseudotachylitic breccias
and granophyre. Earth Planet Sci Lett 144:369387
Kempe U, Dandar S, Getmanskaya TI, Wolf D (1994) The tungsten-antimony mineralization
(Focussed on new occurrences in the Mongolian Altai). In: Seltmann R, Kmpf H, Mller P
(eds) Metallogeny of Collisional Orogens. Czech Geological Survey, Prague, pp 301308
Kempe U, Gtze J, Dandar S, Habermann D (1999) Magmatic and metasomatic processes during
formation of the Nb-Zr-REE deposits Khaldzan Buregte and Tsakhir (Mongolian Altai):
indications from a combined CL-SEM study. Mineral Mag 63(2):165177
Kouzmanov K, Pettke T, Heinrich CA (2010) Direct analysis of ore-precipitating fluids:
combined IR microscopy and LA-ICP-MS study of fluid inclusions in opaque ore minerals.
Econ Geol 105:351373
Kremenetsky AA, Maksimyuk IE, Yushko NA, Kempe U, Poutivtsev M (2005) Trace elements in
quartz from conglomerates in the Witwatersrand Basin (South African Republic) and its role
in the understanding of the deposit formation. In: Burenko EK, Kremenetsky AA (eds)
Prikladnaya Geokhimiya. IMGRE, Moscow, vyp. 7, 1 87100 (in Russian)
Landtwing MR, Pettke T (2005) Relationships between SEM-cathodoluminescence response and
trace-element composition of hydrothermal vein quartz. Am Mineral 90:122131
Larsen RB, Henderson I, Ihlen PM, Jacamon F (2004) Distribution and petrogenetic behaviour of trace
elements in granitic pegmatite quartz from South Norway. Contrib Mineral Petrol 147:615628
Larsen RB, Jacamon F, Kronz A (2009) Trace element chemistry and textures of quartz during
the magmatic hydrothermal transition of Oslo Rift granites. Mineral Mag 73:691707
Law JDM, Phillips GN (2005) Hydrothermal replacement model for Witwatersrand gold. Econ
Geol 100th Anniversary Volume:799811
Lehmann K, Berger A, Gtte T, Ramseyer K, Wiedenbeck M (2009) Growth related zonations in
authigenic and hydrothermal quartz characterized by SIMS-, EPMA-, SEM-CL- and SEMCC-imaging. Mineral Mag 73:633643
Lehmann K, Pettke T, Ramseyer K (2011) Significance of trace elements in syntaxial quartz
cement, Haushi Group sandstones, Sultanate of Oman. Chem Geol 280:4757
Lders V (1996) Contribution of infrared microscopy to fluid inclusion studies in some opaque
minerals (wolframite, stibnite, bournonite): metallogenic implications. Econ Geol 91:14621468
McLaren AC, Cook RF, Hyde ST, Tobin RC (1983) The mechanisms of formation and growth of
water bubbles and associated dislocation loops in synthetic quartz. Phys Chem Mineral 9:7994
Minter WEL, Goedhart M, Knight J, Frimmel HE (1993) Morphology of Witwatersrand gold
grains from the Basal Reef: evidence for their detrital origin. Econ Geol 88:237248
Miyoshi N, Yamaguchi Y, Makino K (2005) Successive zoning of Al and H in hydrothermal vein
quartz. Am Mineral 90:310315
Monecke T, Bombach G, Klemm W, Kempe U, Gtze J, Wolf D (2000a) Determination of trace
elements in the quartz reference material UNS-SpS and in natural quartz samples by ICP-MS.
Geostandard Newlett 24:7381
Monecke T, Monecke J, Mnch W, Kempe U (2000b) Mathematical analysis of rare earth
element patterns of fluorites from the Ehrenfriedersdorf tin deposit, Germany: evidence for a
hydrothermal mixing process of lanthanides from two different sources. Mineral Petrol
70:235256
Monecke T, Kempe U, Gtze J (2002) Genetic significance of the trace element content in
metamorphic and hydrothermal quartz: a reconnaissance study. Earth Planet Sci Lett
202:709724
Monecke T, Kempe U, Trinkler M, Thomas R, Dulski P, Wagner T (2011) Unusual rare earth
element fractionation in a tin-bearing magmatic-hydrothermal system. Geology 39:295298
354
U. Kempe et al.
Mller A, Herrington R, Armstrong R, Seltmann R, Kirwin D, Stenina NG, Kronz A (2010) Trace
elements and cathodoluminescence of quartz in stockwork veins of Mongolian porphyry-style
deposits. Mineral Deposita 45:707727
Mller A, Wiedenbeck M, Van den Kerkhof AM, Kronz A, Simon K (2003) Trace elements in
quartza combined electron microprobe, secondary ion mass spectrometry, laser-ablation
ICP-MS, and cathodoluminescence study. Eur J Mineral 15:747763
Novgorodova MI, Veretennikov VM, Boyarskaya RV, Drynkin VI (1984) Geochemistry of trace
elements in gold-bearing quartz. Geochem Int 21:101113
Onasch CM, Vennemann TW (1995) Disequilibrium partitioning of oxygen isotopes associated
with sector zoning in quartz. Geology 23:11031106
Penniston-Dorland SC (2001) Illumination of vein quartz textures in a porphyry copper ore
deposit using scanned cathodoluminescence: Grasberg Igneous Complex, Irian Jaya,
Indonesia. Am Mineral 86:652666
Perny B, Eberhardt P, Ramseyer K, Mullis J, Pankrath R (1992) Microdistribution of Al, Li, and
Na in a quartz: possible causes and correlation with short-lived cathodoluminescence. Am
Mineral 77:534544
Phillips GN, Law JDM (2000) Witwatersrand gold fields: geology, genesis, and exploration. Rev
Econ Geol 13:439500
Phillips GN, Myers RE (1989) The Witwatersrand goldfields: Part II. An origin for
Witwatersrand gold during metamorphism and associated alteration. Econ Geol Mon
6:598608
Pltze M (1995) EPR investigations of quartz, scheelite and fluorite from high-thermal tracemetal mineralization (in German). PhD thesis, TU Bergakademie Freiberg, 141 p
Poutivtsev M (2001) Bestimmung der Spurenelementgehalte in Konglomeratquarzen aus
vererzten und unvererzten Reefs der Au-U-Lagersttte Witwtersrand (Sdafrikanische
Republik) im Vergleich mit Konglomeratquarzen aus Vorkommen in Karelien (Russland)
Unpubl Diploma thesis, TU Bergakademie Freiberg, 76 pp
Poutivtsev M, Kempe U, Gtze J, Monecke T, Wolf D, Kremenetsky AA (2001) Cathodoluminescece and trace element characteristics of quartz pebbles from the Witwa-tersrand, South
Africa. In: Cathodoluminescence in geosciences: new insights from CL in combination with
other techniques, Abstracts. TU Bergakademie Freiberg, pp 101102
Ramseyer K, Mullis J (1990) Factors influencing short-lived blue cathodoluminescence of
a-quartz. Am Mineral 75:791800
Richter DK, Gtte T, Gtze J, Neuser RD (2003) Progress in application of cathodoluminescence
(CL) in sedimentary geology. Mineral Petrol 79:127166
Robb LJ, Meyer FM (1990) The nature of the Witwatersrand hinterland: conjectures on the
source area problem. Econ Geol 85:511536
Robb LJ, Meyer FM (1991) A contribution to recent debate concerning epigenetic versus
syngenetic mineralization processes in the Witwatersrand basin. Econ Geol 86:396401
Roedder E (1984) Fluid inclusions. Reviews in Mineralogy, vol 12. Mineralogical Society of
America, Washington, 646 pp
Rusk BG, Reed MH (2002) Scanning electron microscopecathodoluminescence analysis of
quartz reveals complex growth histories in veins from the Butte porphyry copper deposit,
Montana. Geology 30:727730
Rusk BG, Koenig A, Lowers HA (2011) Visualizing trace element distribution in quartz using
cathodoluminescence, electron microprobe, and laser ablation inductively coupled plasma
mass spectrometry. Am Mineral 96:703708
Rusk BG, Lowers H, Reed MH (2008) Trace elements in hydrothermal quartz; relationships to
cathodoluminescent textures and insights into hydrothermal processes. Geology 36:547550
Rusk BG, Reed MH, Dilles JH, Kent AJR (2006) Intensity of quartz cathodoluminescence and
trace element content of quartz from the porphyry copper deposit in Butte, Montana. Am
Mineral 91:13001312
Samson I, Anderson A, Marshall D (2003) Fluid inclusions: analysis and interpretation. Mineral
Assoc, Canada, Short Course vol 32, Quebec, 374 pp
15
355
Schaefer BF, Pearson DG, Rogers NW, Barnicoat AC (2010) ReOs isotope and PGE constraints on
the timing and origin of gold mineralisation in the Witwatersrand basin. Chem Geol 276:8894
Seyedolali A, Krinsley DH, Boggs S, OHara PF, Dypvik H, Goles GG (1997) Provenance
interpretation of quartz by scanning electron microscope-cathodoluminescence fabric
analysis. Geology 25:787790
Shore M, Fowler AD (1996) Oscillatory zoning in minerals: a common phenomenon. Can
Mineral 34:11111126
Srensen BE, Larsen RB (2009) Coupled trace element mobilisation and strain softening in
quartz during retrograde fluid infiltration in dry granulite protoliths. Contrib Mineral Petrol
157:147161
Spear FS, Wark DA (2009) Cathodoluminescence imaging and titanium thermometry in
metamorphic quartz. J metamorphic Geol 27:187205
Sprunt ES, Dengler LA, Sloan D (1978) Effects of metamorphism on quartz cathodoluminescence. Geology 6:305308
Taylor RP (1992) Petrological and geochemical characteristics of the Pleasant Ridge
zinnwalditetopaz granite, Southern New Brunswick, and comparisons with other topazbearing felsic rocks. Can Mineral 30:895921
Thomas R, Frster HJ, Rickers K, Webster JD (2005) Formation of extremely F-rich hydrous
melt fractions and hydrothermal fluids during differentiation of highly evolved tin-granite
magmas: a melt/fluid-inclusion study. Contrib Mineral Petrol 148:582601
Van den Kerkhof AM, Hein UF (2001) Fluid inclusion petrography. Lithos 55:2747
Vennemann TW, Kesler SE, ONeil JR (1992) Stable isotope compositions of quartz pebbles and
their fluid inclusions as tracer of sediment provenance: implications for gold- and uraniumbearing quartz pebble conglomerates. Geology 20:837840
Vollbrecht A, Oberthr T, Ruedrich J, Weber K (2002) Microfabric analyses applied to the
Witwatersrand gold- and uranium-bearing conglomerates: constraints on the provenance and
post-depositional modification of rock and ore components. Mineral Deposita 37:433451
Vollbrecht A, Ruedrich J, Weber K, Oberthr T (1996) Gefgekundliche Untersuchungen an
Gerllquarzen der Witwatersrand-Lagersttte in Sdafrika. Z Angew Geol 42:156161
Wark DA, Watson EB (2006) TitaniQ: a titanium-in-quartz geothermometer. Contrib Mineral
Petrol 152:743754
Wark DA, Hildreth W, Spear FS, Cherniak DJ, Watson EB (2007) Pre-eruption recharge of the
Bishop magma system. Geology 35:235238
Watt GR, Wright P, Galloway S, McLean C (1997) Cathodoluminescence and trace element
zoning in quartz phenocrysts and xenocrysts. Geochim Cosmochim Acta 61:43374348
Webster JD (2006) Melt inclusions in plutonic rocks. Mineral Ass Can, Short courses vol 36,
Quebec, 237 pp
Weil JA (1984) A review of electron spin spectroscopy and its application to the study of
paramagnetic defects in crystalline quartz. Phys Chem Mineral 10:149165
Weil JA (1993) A review of the EPR spectroscopy of the point defects in a-quartz: the decade
19821992. In: Helms CR, Deal BE (eds) Physics and Chemistry of SiO2 and the Si-SiO
interface 2. Plenum Press, New York, pp 131144
Whiting KL, Rusk B, Spandler C, Dimond A, Emsbo P (2010) Insights into the origin of Charters
Towers Warrior Vein system from fluid inclusions and quartz trace elements. EGRU
Newsletter, School of Earth & Environmental Science, Economic Geology Research Unit,
James Cook University, Australia, vol 8, pp 1214
Williams LB, Hervig RL, Bjrlykke K (1997) New evidence for the origin of quartz cements in
hydrocarbon reservoirs revealed by oxygen isotope microanalyses. Geochim Cosmochim
Acta 61:25292538
Zinkernagel U (1978) Cathodoluminescence of quartz and its application to sandstone petrology.
Contrib Sedimentol 8:196
Zuffa GG (1985) Provenance of arenites. NATO ASI series C 148. Reidel Publ. Co., Boston, p 393
Index
A
Alaskite, 12, 18
Al (aluminum), 32, 37, 40, 74, 76, 78, 129,
151, 253, 278, 282283, 333335
Amorphous, 3, 5, 162, 164, 171, 239, 248262
Analytics, 3745, 8485, 123129, 191215,
219234, 269, 289, 317
Authigenic, 89, 13, 82, 268, 270, 272276,
281, 287303, 338
Autoclave, 20, 55
Application, 2, 9, 11, 18, 3135, 193, 238
B
Beam, 76, 84, 131, 194206, 214, 223, 238,
282, 323
Brazil, 30, 139156, 199
Brick, 5458, 6062, 69
Building industry, 5369, 143
C
Calibration, 84, 123124, 211, 214, 224,
228230, 269
Cathodoluminescence (CL), 69, 85, 91,
123124, 126, 237262, 270,
293296, 307325, 337341
Cement, 14, 16, 54, 5960, 270, 272, 275, 288,
338
Centers, 41, 43, 149, 153154, 161172,
177188, 237262
Chalcedony, 3, 5, 7, 13, 288, 300301
Chemical analysis. See analytics
Chemistry, 3, 18, 57, 5961, 76, 83, 88, 92,
105, 108, 110, 128129, 131133,
183, 199, 217, 231, 240, 271, 273,
277, 296299
D
Damage, 89, 193, 197206, 214, 251
Defects, 37, 7375, 100, 156, 178, 181, 183,
203, 237262, 277283
Deposit, 3031, 6769, 85111, 119135,
139149
Dislocation, 3, 6, 48
Detection limit. limit
of detection
E
E center, 5, 161172, 181, 248, 254, 258
Emission, 67, 10, 183, 206, 238, 249250,
270, 277283, 295296, 302, 309,
320323
EPR, 40, 161172, 177188, 266
F
Fluorescence, 39, 76, 124, 130,
206208
357
358
F (cont.)
Fluids, 89, 35, 8182, 106, 109, 112, 144,
147, 298303, 312, 316, 334,
336337, 341345
Flint, 13, 67
Foundry industry, 1819, 54, 69
G
Gamma irradiation, 5, 150155, 162,
177188, 192193,
238, 270
Gemstone, 139156
Genesis, 10, 18, 85135, 183,
287303, 338
Geochemistry. see also chemistry, 266,
287303
Germany, 89, 14, 17, 19, 56, 58, 207, 268,
290, 333337
Glass industry, 12, 19, 54, 142, 148
Growth. crystal growth
H
Hydrothermal, 7, 1112, 18, 20, 36, 91, 9394,
103, 107112, 120, 140, 142143,
145148, 193, 221, 267268, 273,
275, 280, 307325, 333337
alteration, 281, 333342, 347348
I
Imaging, 85, 239, 309, 320, 337341, 348
Impurities, 89, 12, 18, 3233, 35, 39, 49,
7382, 120, 134, 153, 178180,
220221, 238, 240, 253, 267, 277,
280, 282, 309, 334
Inclusions, 11, 37, 72, 75, 7781, 9091,
85111, 152, 178, 192, 318
fluid, 11, 4245, 76, 81, 129134, 288,
308, 325, 334, 341345,
347350
liquid, 4245, 300
melt, 8182
mineral, 11, 3942, 8182, 210, 250, 259,
293295, 345347
Industry, 2, 12, 1618, 3135, 5369,
142, 148
Industrial application, 2, 9, 11, 18, 3135, 193,
238
In situ, 76, 84, 111, 191215, 219234, 241,
288, 317
Intergrowth, 1415, 99, 291295, 345347
Iota, 18, 39, 73, 8283
Index
K
Keatite, 3
L
Lascas, 18, 21, 30, 140, 143, 156
Laser Ablation ICP-MS, 12, 39, 7576, 8485,
92, 111, 123, 127129, 134,
192196, 199, 219234, 274,
317318, 320
Lattice, 4, 810, 3940, 43, 7277, 81, 83, 85,
111113, 120, 126, 134, 162, 178,
197, 220221, 234, 238, 246, 251,
259, 279, 282, 308, 316317, 333
Lechatelierite, 3, 5, 21
Lighting industry, 3334
Limit of detection, 40, 76, 99, 102, 128129,
192, 208214, 224, 231, 234, 269,
317318
Luminescence, 34, 610, 99, 126127,
177188, 237262, 265283,
287303, 337341
M
Macrocrystalline, 4, 288, 299
Magmatic, 1213, 1719, 78, 103, 192,
267268, 271, 273, 279, 282283,
333, 337338, 348
Mass spectrometry. See also laser ablation
ICP-MS, 39, 76, 192, 334
Melanophlogite, 3
Metamorphic, 7, 1213, 1719, 21, 36, 77, 82,
96, 112, 122, 134, 146, 267268,
272276, 279, 283, 308311, 324,
333337, 347350
Microanalysis, 41, 228, 237262
Microprobe (electron microprobe), 191215,
228, 273274, 318, 321,
322, 334
Microinclusion, 4, 1011, 178, 290, 294297,
302
Microstructure, 5, 9, 14, 17, 237262
Mineralogy, 25
Mining, 46, 64, 101, 111, 140, 147
Moganite, 3, 5, 293
Mortar, 54, 6364, 69, 134
Mortar texture, 126
N
Nomenclature
SiO2 rocks, 1216
SiO2 system, 3
Index
Non-bridging oxygen hole center (NBOHC),
5, 7, 171, 247, 249, 251, 254, 260,
280, 295
Norway, 30, 85111, 120
O
Opal, 3, 5, 1315, 154, 288, 299300
Optical fibers, 32, 35
Optical microscopy, 40, 45, 85, 126, 337
Oxygen, 125, 154, 180, 206, 241
Oxygen vacancy, 57, 161172, 178181,
185186, 238, 250251, 256, 259,
261262, 278, 293, 295
P
Paramagnetic centers, 5, 40, 165, 171,
177187
Pegmatite, 2, 7, 12, 1819, 21, 3637, 71, 78,
81, 8687, 8992, 100103, 112,
120, 126, 139144, 156, 184, 221,
224, 232, 233, 270, 275, 334335,
338
Petrography, 87, 98, 101, 103, 107, 110, 126,
129, 308, 324, 344
Photovoltaic industry, 3234
Plutonic rocks, 267
Point defect, 26, 161, 178, 237, 246, 248,
251, 254, 256, 260261, 266267,
272, 277, 282283
Porcellanite, 2, 13
Postmagmatic, 12, 1718
Precipitation, 13, 15, 288, 300302, 307, 308,
311313, 316, 324325, 341,
343344, 347, 349
Preparation, 19, 47, 60, 64, 119120, 134135,
151, 193, 220221, 228, 291, 318,
334
Processing, 3031, 35, 39, 4549, 61, 6465,
67, 7172, 75, 8485, 99, 106, 111,
113, 120, 140, 150151, 214, 240
Properties
chemical, 3, 2021
physical, 24, 148
typomorphic, 13
Provenance, 910, 192193, 221, 288, 331,
337338, 341, 348
Q
Quartz
authigenic, 9, 82, 268270, 273, 275276,
287303, 338
359
deposits, 3031, 36, 71113, 120, 125,
139156, 192
formation, 4, 267, 297303, 308, 324, 332,
341, 343, 346347
high-purity, 9, 1112, 16, 1819, 35,
71113, 119, 219220, 232
hydrothermal, 2, 7, 9, 11, 1819, 71, 77,
79, 9192, 94, 103113, 120, 143,
145148, 221, 266268, 270271,
273, 275276, 278, 307325,
334335, 338
igneous, 7, 7879, 82, 315, 321322
Iota, 18, 39, 73, 8283
magmatic, 1719, 265, 267269, 271, 273,
275, 281
metamorphic, 7, 1213, 17, 19, 21, 36, 82,
112, 265, 268270, 272273,
275276, 278279, 281, 283,
310311, 334337, 348
milky, 4, 4547, 109, 145, 154
pebbles, 63, 66, 146, 148, 337341, 343,
348
postmagmatic, 12, 1718
sand, 2, 9, 1619, 21, 36, 5369, 72, 8384,
139140, 142, 147148, 150, 156
sedimentary, 18, 289
smoky, 36, 48, 100, 145, 149154
synthetic, 2, 7, 9, 2021, 35, 145, 199201,
203, 247, 277278, 281, 308, 325,
349350
varieties, 4, 150151
vein, 12, 18, 36, 108112, 142, 182183,
185187, 307325, 334335, 341,
343347, 349
volcanic, 78, 210, 270, 275276, 291,
309, 311
Quartzite, 2, 1214, 1620, 36, 6668, 71, 82,
85100, 106107, 112113, 141,
143, 145148
R
Rare earth elements, 333
REE, 10, 101, 178, 288, 298299, 303,
333338
Raw materials, 13, 912, 1621, 3031, 37,
46, 5369, 72, 8183, 111, 120,
148, 192, 251, 332
Real structure, 34, 265, 267, 332, 348
Refinement, 30, 73, 75, 112113
Resources, 29, 99, 103, 107, 139140, 148,
156, 219, 221
Rocks
magmatic, 12, 17, 267, 345
360
R (cont.)
metamorphic, 12, 13, 19, 82, 148, 221, 267,
316, 337339, 348
igneous, 311
plutonic, 267
sedimentary, 23, 1213, 16, 19, 21, 145,
192, 287303, 319, 338, 348
siliceous, 2, 1314, 21
siliciclastic, 1317, 21, 289
volcanic, 17, 9697, 145, 311, 337
S
Sandstone, 2, 9, 13, 1617, 63, 89, 146, 148,
268, 270, 272, 287290, 292293,
295303
Sediment, 1314, 1617, 56, 6768, 97, 108,
144146, 221, 280, 288, 290, 293,
298303, 348
Semiconductor, 9, 11, 1718, 29, 31, 3335,
49, 83, 238
Seifertite, 3
Self-trapped exciton, 7, 237, 246, 248, 251,
254, 261, 279
Silanol group, 6, 74, 295, 303
Silcrete, 1416, 20
Silica, 15, 9, 1215, 1721, 29, 3132,
3436, 45, 49, 60, 6263, 82, 143,
171, 220, 228, 240, 251, 280281,
288, 291, 293, 297303
Siliceous sinter, 1314
Slicification, 14, 280, 288
SiO2
amorphous, 35, 123, 162, 164, 171, 237,
251, 253254, 256, 260262
classification, 23
high-purity, 1819
modifications, 24
minerals, 121
Index
polymorphs, 25, 237238, 242, 244, 246,
252253, 261262
rocks, 13, 1219, 21
system, 3
varieties, 4
Spatial resolution, 191, 193, 237, 266, 269,
317318, 333
Spectroscopy, 4, 9, 12, 4, 45, 74, 119, 122,
124, 130131, 134, 154, 177,
181182, 240, 258, 259260, 265,
269270, 293, 302, 332, 343
Standards, 18, 31, 36, 39, 62, 220, 222, 224,
226, 228232, 234, 318
Stishovite, 3
T
Texture, 15, 21, 40, 99, 126, 148, 301, 303,
307325, 341
Thermoluminescence, 177, 180, 183
Trace elements, 912, 19, 36, 3840, 71,
7378, 8283, 85, 90, 92, 98, 100,
103, 106, 111113, 120, 123,
125126, 128129, 178, 181183,
191215, 219234, 265283, 287,
291, 296298, 301, 307309, 316,
317325, 331337, 339, 341,
347349
Treatment
chemical, 4749, 120
physical, 45
thermal, 6, 45, 4950, 81, 120, 149154,
181
Tridymite, 3, 5, 7, 20, 282
Tripoli, 3, 14
X
XRD, 291, 293