Chemical Geology: A A A A A A B B

Chemical Geology 520 (2019) 33–51
Contents lists available at ScienceDirect
Chemical Geology
journal homepage: www.elsevier.com/locate/chemgeo
Chemical weathering of S-type granite and formation of Rare Earth Element T

(REE)-rich regolith in South China: Critical control of lithology
⁎
Wei Fua, , Xiaoting Lia, Yangyang Fenga, Meng Fenga, Zhao Penga, Hongxia Yua,b, Henry Linb
a
Guangxi Key Laboratory of Hidden Metallic Ore Deposits Exploration, Guilin University of Technology, Guilin 541004, China
b
Department of Ecosystem Science and Management, The Pennsylvania State University, University Park, PA 16802, USA
A R T I C LE I N FO A B S T R A C T
Editor: Jérôme Gaillardet REE-rich granite regolith is a significant host for the ion-adsorption type Rare Earth Elements (REE) ore re-
Keywords: sources in South China. The issue of why and how such specific regolith was generated has attracted great
Chemical weathering attentions for both scientific and economic interests. To improve the understanding of critical factors that control
Lithology the formation of REE-rich granite regolith, especially for those related to weathering of the S-type granite, an
S-type granite investigation of granite and its overlying regolith is conducted in South China's largest S-type granite terrain
Accessory minerals (Darongshan, Guangxi). The Hercynian and Indosinian granites outcrop as neighboring lithofaces, but their
REE enrichment overlying regolith show significant differential chemical weathering. Examinations of their representative re-
South China
golith profiles found that the profile from the Hercynian granite is thick and REE-rich, whereas that from the
Indosinian granite is relative thin and REE-poor (12 m vs. 6 m in depth, 929 ppm vs. 226 ppm in REE con-
centration). Given similar climatic, topographic, and vegetative conditions, the difference between two profiles
can be principally traced down to their parent granite lithology. Formation of the thick and REE-rich profile is
closely associated with some specific lithological factors of its parent granite. Contrastingly, it has coarser grain
size, wider microcracks, as well as higher biotite and plagioclase contents, which is expected to be more con-
ductive to enhance water-rock interaction and drive deep weathering. Meanwhile, it contains higher initial REE
concentration (342 ppm vs. 132 ppm) and, more importantly, richer REE-bearing accessory minerals (monazite,
apatite and zircon), which offered dominant REE source to regolith. Especially, apatite (REE = 1549–4413 ppm)
is an critical REE source mineral in supplying mobile REE to the regolith, through which REE have access to be
fixed by clay minerals (mainly kaolinite and illite) and then developed ion-exchangeable form enrichment in the
regolith. These evidences indicate that the granite lithology exerts a principal influence on the formation of thick
and REE-rich regolith through fundamental control of chemical weathering and REE input. Moreover, this study
find that, for the S-type granite particularly, the high P2O5 content (> 0.08 wt%) granite seems more optimistic
in generating ion-adsorption type REE ore (especially LREE) by weathering than previously thought.
1. Introduction ability of the Earth's surface and lead to diversified regolith (Scott and
Pain, 2008). Among silicate rocks outcropped in land surface, granite
Chemical weathering plays an important role in providing nutrients accounts for about 25% of whole Earth's area (Oliva et al., 2003), and
to ecosystems, regulating earth surface environment over geological its weathering represents an important part of global weathering and
time (e.g., Nesbitt and Young, 1982; Ollier et al., 1988; Twidale and Earth surface system. Insight into granite chemical weathering is thus of
Campbell, 1995; Kump et al., 2000; Drever, 2004; Gislason et al., great significance to understand Earth surface formation and evolution.
2009), and even generating ore deposit for mining industry (e.g., Butt Also, the knowledge of granite chemical weathering is closely asso-
et al., 2000; Anand and Paine, 2002; Freyssinet et al., 2005; Anand and ciated with mankind's living within their outcrop regions, especially
Butt, 2010; Butt, 2016; González-Álvarez et al., 2016). The rate and linking to many economic, engineering, and environmental issues (e.g.,
nature of chemical weathering vary widely and are controlled by many Wakatsuki and Matsukura, 2008; Yang et al., 2013). Many studies re-
variables such as parent-rock lithology, topography, climate and bio- lated to granite chemical weathering, in particular from a Critical Zone
logical activity (Goldich, 1938; Lasaga et al., 1994; White and Brantley, science perspective (Brantley et al., 2007; Anderson et al., 2007), are
1995). Lithological differences give rise to the differential weathering increasingly conducted to understand how granite is gradually
⁎
Corresponding author at: College of Earth Sciences, Guilin University of Technology, Jiangan Road No.12, Guilin 541004, China.
E-mail address: fuwei@glut.edu.cn (W. Fu).
https://doi.org/10.1016/j.chemgeo.2019.05.006
Received 20 August 2018; Received in revised form 1 March 2019; Accepted 2 May 2019
Available online 07 May 2019
0009-2541/ © 2019 Elsevier B.V. All rights reserved.
W. Fu, et al. Chemical Geology 520 (2019) 33–51
weathered and its multiple significance in hydrology, geology, ped- impacts on granite chemical weathering as well as REE accumulation in
ology, and ecology by exploring the record of granite regolith and its regolith.
related weathering processes (e.g., Brantley, 2010; Graham et al., 2010;
Dethier and Bove, 2011; Lybrand and Rasmussen, 2014; Liu et al., 2. Geological and geographical setting
2016).
In South China, a large volume of granites are subject to long-term Darongshan granite is a general term in local geology, which is
chemical weathering and developed extensive regolith. Of these actually a large granitic suite distributed in the southeastern part of the
granite-derived regolith, some are found hosting valuable Rare Earth Guangxi (near the China–Vietnam border), South China (Fig. 1-A).This
Elements (REE) sources that can be extracted and applied in high-tech granitic suite is very important and well documented in the South China
industries, which makes them attractive for both scientific and eco- block where extensive Hercynian to Indosinian intrusive and volcanic
nomic interests (Bao and Zhao, 2008; Kynicky et al., 2013; Wang et al., rocks expose with an area of over 10,000 km2. It is well accepted as the
2013; Simandl, 2014; Sanematsu and Watanabe, 2016; Li et al., 2017). largest granitic complex composed of batholiths and plutons typical of
Especially, the southern part of Jiangxi Province is well-known in REE- S-type origin in South China (Chen et al., 2011), nearly accounting for
rich regolith distribution, where the weathering of some specific half of whole distribution area of granitic rocks in Guangxi. Spatially, it
granites have generated the exclusive HREE (from Gd to Y) resources in consists of five major units (Taima, Nadong and Jiuzhou plutons, and
the world (Huang and Wu, 1988; Wu et al., 1990; Yuan et al., 1992; Bao Pubei and Darongshan batholiths), and the extending direction of this
and Zhao, 2008; Yuan et al., 2012; Wang et al., 2013; Xu et al., 2017). granite suite is NE-NEE, parallel to that of Hercynian fold line. Litho-
Based on many pioneering studies that have deepened the under- logically, major rock types in this entire granitic suite include cor-
standing of REE behaviors in weathering environment (e.g., Nesbitt, dierite–biotite granite, hypersthene granitic porphyry, biotite granite,
1979; Henderson, 1984; Braun et al., 1990; Wood, 1990; Nesbitt and and cordierite–biotite monzogranite. A series of geochemistry, geo-
Wilson, 1992; Koppi et al., 1996; Oliva et al., 1999; Patino et al., 2003), chronology studies have been conducted due to its great significance to
a large number of case studies have been conducted in South China to regional geology evolution (Fang et al., 1987; Wang, 1991; Deng et al.,
explore the REE enrichment effect caused by granite chemical weath- 2004; Qi et al., 2007). Results of geochemical and isotopic studies
ering from various perspectives (e.g., Song and Shen, 1987; Huang (Deng et al., 2004), coupled with the presence of abundant Al-rich
et al., 1989; Wu et al., 1990; Zhang, 1990; Ma and Liu, 1999; Bao and minerals, such as cordierite, garnet, hypersthene, almandine, sillima-
Zhao, 2008; Liu and Wu, 2011; Chi et al., 2012; Zhao et al., 2014; Yang nite, and andalusite, suggest that they can be identified as S-type
et al., 2015; Sanematsu and Watanabe, 2016; Xie et al., 2016; Zhao (peraluminous) granite in terms of geochemical classification, or il-
et al., 2017). Current knowledge of the nature and origin of the REE- mentite series granite in terms of mineral classification. The age of
rich granite regolith have been significantly advanced: (1) large var- these granitic rocks are reported in a range of 275 to 230 Ma on the
iation in REE geochemical indices are present within the REE-rich Zircon UePb technique and the whole-rock RbeSr isochron (Deng
granite regolith, which may reflect the complex supergene processes et al., 2004; Charoy and Barbey, 2008). High ISr (> 0.72) and low εNd
and the history of weathering (e.g., Wu et al., 1990; Sanematsu and (t) values (−13.0 to −9.9) indicate that they were originated from
Watanabe, 2016; Li et al., 2017); (2) factors influencing the distribution evolved crustal materials (Qi et al., 2007).
and mobilization of REE are a combination of mineralogical, chemical, This study pays more attention on the north part of Darongshan
and biological parameters, involving in the primary mineral composi- granite suite, i.e., the Darongshan batholiths (Fig. 1-B). There are two
tions of the parent rock, the secondary minerals, the solution com- main types of granitic rock in terms of lithology: fine-grained biotite
plexes, biological recycling, organic matter, and so on (e.g., Zhang, adamellite granite and coarse-grained biotite adamellite granite (Fig. 1-
1990; Bao and Zhao, 2008; Wang et al., 2013; Zhao et al., 2014). Each C). According to regional geology investigation (Guangxi bureau of
factor may act as a major influence on the concentration and behavior Geology and Mineral Exploration, 1985), the fine-grained biotite
of REE, ultimately control the general characteristics of REE geo- granite belongs to Indosinian intrusions, whereas the coarse-grained
chemistry within the profile; (3) several concepts are proposed to define biotite granite is part of Hercynian intrusions.
the REE mineralization related to granite weathering, including ion- Geographically, the study area is located in the southeast of Rong
adsorption type (Wu et al., 1990), weathered crust elution-deposited county of Yulin, Guangxi Autonomous Region, South China. It is a ty-
type (Sanematsu and Watanabe, 2016), or regolith-hosted type ore (Li pical subtropical climate condition with a yearly alternation of wet
et al., 2017); and (4) the distribution of REE resource hosted granite season (from March to August) and dry season (from September to
regolith have been found expanding from South China to comparable April). According to local weather station, the average annual tem-
regions worldwide, such as Laos (Sanematsu et al., 2009), Thailand perature and average annual rainfall are about 21.5 ± 5 °C and
(Sanematsu et al., 2009; Sanematsu et al., 2011, 2013) and United 1580 ± 300 mm/year, respectively. The high temperature and abun-
States (e.g., Foley and Ayuso, 2015; Bern et al., 2017). However, the dant rainfall support lush vegetation over the entire region.
issue of why and how such REE-rich regolith was generated remains Local landscape is a typical of hilly terrain with generally low relief
imperfectly modeled, and many detailed scientific problems have not (100–300 m in elevation). Few moderate relief hills (> 500 m) are
been well answered yet. For example, how does the thick granite re- scatteredly located in the Northeast and Southwest parts. A large
golith with tens of meters depth formed? What speciation of REE does it number of massifs are the dominant geomorphic units. These massifs
present in granite regolith and which type of REE speciation plays are all granite-related, and they occur continuously or locally dissected
dominant role during its migration in granite weathering? What factors by river system. The shape of granite-related massifs is generally shield-
and mechanisms might be responsible for the degree of REE enrichment like with round or flat hilltop. Although the surface of massifs is wavy,
and fractionation during the intensive weathering of granite? their hilltop elevations similarly concentrate at 200–300 m, and there
Our efforts in this study focus on an investigation of REE-rich re- are no obvious stepped elevations, which may reflect the landform of
golith related to weathering of the S-type granite (as defined by this area is relative simple and uniform. In vegetation, this area is
Chappell and White, 1974), which is considered as a significant source mainly covered by pine tree, eucalyptus, and other typical subtropical
rock for generating the ion-adsorption type REE ore. In particular, two shrubs, except part of them are reclaimed for farming.
adjacent granite regolith profiles from the Darongshan area in the
Guangxi Autonomous Region (Fig. 1-A) are selected for comparison. 3. Regolith geology
They are 4 km apart, with similar climatic background, biological (ve-
getation) condition, and topographic setting, but with different parent Field investigations find that the granite regolith in study area is
rock lithology. It helps to elucidate the importance of lithological generally at a range of 5–30 m in thickness, and their depth vary greatly
34
Fig. 1. (A) Simplified distribution map of the granitic rocks in South China, showing the position of Darongshan granite suite; (B) main lithological units of the
Darongshan granite suite; (C) simplified geological map of the study area, showing the locations of the sampled regolith profiles (D1 and D2).
even locally beyond 30 m in some terrains with low-moderate relief

(< 200 m), but the thickness of fine-grained granite regolith is only
ranging from 3 to 10 m. Their representative regolith profiles are shown
in Fig. 2-A and -B respectively.
Similar lithostratigraphic sequences are observed vertically
throughout the coarse-grained and fine-grained granite regolith, despite
of their difference in thickness i.e., the weathering profiles can be
subdivide into several lithostratigraphic units based on the field char-
acteristic in color, texture and mineral composition (Fig. 2-A, B). They
are orderly defined as full-weathered horizon in the upper part, highly-
weathered horizon in the middle part and semi-weathered horizon in
the lower part. Specifically, the full-weathered horizon is light red in
color, soft, porous, and fine-grained in texture. By naked eyes ob-
servation, its mineral composition seems to be totally clay minerals.
The highly-weathered horizon is pale-yellow in color, porous and fine-
grained in texture, and its mineral composition is a mixture of clay
minerals and primary minerals. Fine-grained quartz, as a typical
weathering-resistant mineral, can be easily observed from the coex-
isting clay minerals in this position. Therefore, the presence of quartz
can be taken as a mineral indicator for distinguishing the highly-
weathered horizon with its overlying full-weathered horizon. With
profile downward, the semi-weathered horizon is pale white, loose,
porous and friable. The granitic texture inherited from parent rock is
Fig. 2. Studied profiles derived from two lithologically contrasting parent rocks partly observed in this horizon, and thus this textual feature can be
in areas of the Darongshan batholiths, showing their sampling columns and taken as an indicator for separating the semi-weathered horizon from
division of lithostratigraphic units. (A) The sampling column of the D1 profile; its overlying highly-weathered horizon. The mineral composition of this
(B) the sampling column of the D2 profile. horizon seems rather chaotic, showing mixtures of different proportions
of fine-grained clay minerals and coarse-grained residual primary
with topographic locations. The thick regolith sites commonly situate at fragments. Also, a number of spheroidal granitic stones are randomly
the flatty summit of the granite massifs, and there is a shallower trend observed at the lower part of this horizon.
downslope. Comparatively, the regolith overlying the Hercynian
coarse-grained granite (D1) displays thicker (commonly double thick- 4. Sampling and analytical methods
ness) than that from the Indosinian fine-grained granite (D2). In detail,
the coarse-grained granite may develop regolith about 8–25 m, and Two representative regolith profiles, derived from the Hercynian
35
coarse-grained granite (D1) and the Indosinian fine-grained granite (D2) silicate mineral and REE-bearing minerals analysis, respectively.
respectively, are selected for sampling and further laboratory ex- Analytical error is ≤5% at the ppm level. In-run signal intensity for
aminations. The sampling is conducted orderly from the fresh parent indicative trace elements was monitored during analysis to make sure
rock through the semi-weathered and highly-weathered horizons to the that the laser beam stayed within the phase selected and did not pe-
full weathered horizon, as shown in Fig. 2-A and -B. Totally, 12 samples netrate inclusions.
from the D1 profile (D1-1–D1-12) and 7 samples from the D2 profile (D2- The organic matter was determined applying the dichromate oxi-
1–D2-7) were collected. dation method (Nelson et al., 1996). Briefly, one gram of soil was
For fresh granite samples (D1-12 and D2-7), they were analyzed by oxidized at 150 °C for 1 h with 25 ml of a mix of potassium dichromate
using petrographical observations, quantitative X-ray diffraction and sulfuric acid. Then, a reverse titration of Cr2O72− ions was made
(QXRD), X-ray fluorescence (XRF) and inductively coupled plasma mass using an acidified ferrous ammonium sulfate solution.
spectrometry (ICP-MS). Prior to XRD and bulk geochemical analyses, For the bulk geochemical analyses, the prepared samples were first
the samples were air dried, crushed, pulverized by agate mortar and dried at 70 °C, and then both the dried regolith and fresh bedrock were
sieved through a 200 meshes. In order to study accessary minerals, ground into powder. The powder was baked at 105 °C to remove ad-
electron microprobe (EPMA) and laser ablation inductively coupled sorbed water before analysis. Major elements were measured using X-
plasma mass spectrometry (LA-ICP-MS) are used after making the po- ray fluorescence (XRF). Analytical precision for major elements
lished thin sections. is < 1%, and the determination limits for the major elements are gen-
For the regolith samples (D1-1–D1-10 and D2-1–D2-6), besides the erally better than 30 ppm. Trace elements and REE were measured
routine analyses (QXRD, XRF and ICP-MS), they were also analyzed using X-SERIES ICP-MS. Approximately 100 mg of powdered sample
using scanning electron microscopy (SEM-EDS), organic matter (OM) was digested with 1 ml HNO3 and 2 ml HF in screw-top, PTFE-lined
measurements and sequential extraction experiment (SEE). stainless steel bombs at 190 °C for 48 h. Remove the inner Teflon tank
Most of analytical works, including petrography, QXRD, EMPA, on the hot plate after cooling and the solution was evaporated to dry-
SEM-EDS, LA-ICP-MS and EES were conducted at the Key Laboratory of ness at 165 °C. The solution was then drained and evaporated to dryness
the Guangxi Hidden Metallic Ore Deposits Exploration in China. Except with 1 ml HNO3. This procedure was repeated twice. The final residue
that, the bulk geochemical analyses (XRF + ICP-MS) were carried out was redissolved by adding 5 ml of 6 mol/L HNO3. Subsequently, the
at the Chinese National Research Center of Geoanalysis in Beijing, bomb was resealed and heated at 150 °C for 5 h. After cooling to room
China. Detailed analytical methods of main referred laboratory works temperature, the final solution was diluted to 100 ml by adding distilled
are present as follows: de-ionized water. The reagent blanks were treated following the same
The XRD data were obtained using a Philips X' Pert MPD dif- procedures as the samples. Total analytical errors for trace elements
fractometer. It works on Cu target at 40 kV and 40 mA. The range of 2θ and REE in this study are within ± 6% (1 σ).
scanning was from 5° to 80°. The scan step and step duration were 0.05° A Seven-step sequential extraction was performed to determine the
and 3 s, respectively. A standard grinding and mounting of the sample REE speciation of the regolith samples. Detailed REE extraction pro-
was performed prior to XRD. Qualitative analysis is firstly conducted to cedure is given by Shi et al. (2014). Briefly, the reacted REE at each step
get the mineral compositions of the samples through Highscore soft- is orderly assumed to water soluble fraction (F1) extracted by purified
ware. And then, Rietveld method is further applied to get the quanti- H2O in step 1, and then ion exchangeable fraction (F2) extracted by
tative data of the mineral compositions, which was carried out through H2O + MgCl2 solution in step 2, carbonate-bound or specific adsorption
Topas software (academic V5.1). The core technique of quantitative fraction (F3) extracted by NaAc solution in step 3, humic acid fraction
analysis is full spectrum fitting refinement (Hill and Howard, 1987). It (F4) extracted by Na4P2O7 solution in step 4, iron‑manganese oxide
consists of a numerical simulation of the XRD pattern based on both the fraction (F5) extracted by NH2OH·HCl-HCl solution in step5, strong
instrumental conditions used in data collection and crystallographic organic-bound fraction (F6) extracted by HNO3 + H2O2 + NH4Ac-
information on the phases identified in the sample (Rietveld, 2014). HNO3 solution in step 6, and finally residual fraction (F7) extracted by
Following the suggestions of Andrade et al. (2018), the first refinement HNO3-H2SO4-HCLO4-HF solution in step 7. After each extraction step,
step was to correct for sample displacement with respect to the focal the collected filtrate is tested for REE concentration by ICP-MS and then
circle of the equipment. The next step was to adjust cell parameters for proceeds to next step continuously. The recovery rate of all seven REE
all minerals phases according to their crystallographic information files fractions relative to total REE amount is between 90% and 110%.
(CIF), which is an independent fitting of peak positions and intensity.
Refinement was made in iterative steps until convergence of the cal- 5. Results
culated and observed patterns were reached.
The EMPA analysis was performed on polished sections for de- 5.1. Mineral compositions
termining the accessory minerals in parent rock samples, which is
carried out using JXA-8230 electron microprobe with a wavelength Petrographical observations (Fig. 3-A–C, Fig. 4-A–C) show that both
dispersive system. The measurement conditions were 20 kV accel- of granite parent rock contain similar rock-forming minerals, mainly K-
erating voltage, 15 nA probe current, and 2 μm beam diameter. The feldspars, plagioclase, quartz, and biotite (in places altered to chlorite).
SEM-EDS analyses were conducted for determine the clay minerals in Quantitative XRD analysis (Table S1), however, indicates that the
regolith samples, using ∑IGMA series field emission scanning electron percentage of these rock-forming minerals is rather different between
microscopy developed by as ZEISS Company, equipped with an X-ray the two. The D1 granite contains relative higher plagioclase (33% vs.
spectrometer (EDS). Its resolution is 1.3 nm, acceleration voltage is 23%) and quartz (34% vs. 29%) contents but lower K-feldspar (18% vs.
0.1–30 kV, probe current is 4pa-20na and the maximum magnification 43%) content compared to the D2 granite. Textually, the grain size of
multiple is 1 million times with an error around 3%. Prior to testing, rock-forming minerals in the D1 granite is mainly in the range of
fine particles of the regolith samples were chosen and then fixed on the 2–8 mm and exhibits a typical coarse-grained texture (Fig. 3-A–C),
surface of glass section using epoxy resin. whilst the D2 granite exhibit a fine-grained texture with grain size
In-situ LA-ICP-MS analyses were performed for examining REE ranging in 0.5–3 mm (Fig. 4-A–C).
concentration of accessary mineral phases from the granites in polished Accessory minerals in the D1 granite include zircon, apatite, ilme-
thin section, using Agilent 7500 laser inductively coupled plasma mass nite, monazite, rutile, and minor allanite (Fig. 3-D–I). Notably, apatite
spectrometry equipped with 193 nm laser. The diameter of the ablation is a common accessory mineral in the D1 granite. It has a stumpy or
spot is 32 μm. NIST 610 glass was used as a calibration standard for all hexagonal prism shape up to 0.3 mm in length, and mostly occurs as
samples, with 29Si and 43Ca as internal standards for quantitative inclusions within biotite grains. In the D2 granite, similar accessory
36
Fig. 3. Macro- and micro-photographs of the D1 granite parent rocks. (A) Hand specimen of the coarse-grained D1 granite; (B–C) coarse-grained texture under
microscope from the D1 granite sample, with a large number of microcracks mainly present at the internal of the quartz and K-feldspar grains; (D–I) back-scattered
electron (BSE) images of accessory minerals from the D1 granite; (D–E) presence of apatite, zircon, monazite, rutile and xenotime at the internal of biotite grains; (F)
irregular-shaped apatite, coexisting with columnar zircon and granular-shaped monazite; (G) hexagonal-shaped apatite grains coexisting with monazite and zircon
present at the internal of biotite grain; (H) irregular-shaped allanite with fine grain coarse; (I) tetragonal-shaped xenotime surrounded by K-feldspar and biotite.
(Abbreviations: Qtz-Quartz, Bt-Biotite, Kfs-K feldspar, Pl-Plagioclase, Ap-Apatite, Mnz-monazite, Rt-Rutile, Xen-Xenotime; Zrn-Zircon, Aln-Allanite).
mineral assemblage is observed, including zircon, ilmenite, monazite, 36% → 22% → 50% → 35% → 26%). As a whole, the D1 profile has
rutile and few xenotime (Fig. 4-D–F). In contrast, the accessory minerals more abundant total clay mineral concentration than the D2 profile
observed in the D2 granite show less abundance and smaller grain size (70% vs. 43% in highest value). In addition, although iron and man-
compared to those from the D1 granite. Especially, apatite is rarely ganese minerals could not be deduced from X-ray diffract grams, these
observed in the D2 granite versus D1 granite, which is consistent with minerals are certainly present. They are observed as small spots or thin
lower P2O5 content in the D2 granite than that of the D1 granite in the coatings scattered within the highly-weathered and semi-weathered
following geochemistry description. horizons, especially in the D1 profile.
For the regolith samples, quantitative XRD analyses (Table S1) show
that the mineral compositions of the D1 and D2 profiles are comparable. 5.2. Organic matter, major and trace elements
Quartz, K-feldspar, kaolinite, illite, and in some case including plagio-
clase, constitute the common mineral assemblage in most of the re- The concentration of organic matter (OM), major and trace elements
presentative samples. Biotite is not recognized even in the lower part of data of two studied profiles are shown in Tables 2 and 3. The organic
semi-weathered horizon, indicating that it is disappeared from incipient matter concentration varies broadly with a range of 0.74–9.47 wt% in
stage of granite weathering. Kaolinite and illite, representing two the D1 profile and 0.63–4.89 wt% in the D2 profile. The highest OM
dominant kinds of clay minerals, are prevailing throughout the two value occurs commonly at the top of the profiles, whereas the lowest
studied profiles. The SEM images show that the micromorphology of OM value is found at the semi-weathered horizon. This demonstrates an
illite is characterized by capillary and needle-like (Fig. 5-A, B). Kaoli- increasing trend in OM concentration with both profiles upward, in-
nite is mainly sub-euhedral, with sheet-like, flaky, irregular hexagonal dicating more bio impacts as moving up closer to surface.
plate in shape (Fig. 5-C), and commonly 4–8 μm in grain size. Spatially, All regolith samples show markedly increased concentrations in
illite is intimately coexisting with kaolinite (Fig. 5-D). Results of Al2O3, Fe2O3 and TiO2 than that of the parent rock in both the profiles,
quantitative XRD analysis (Table S1) show that the percentage of coupled with a reduced concentrations in NaO, CaO, SiO2, MgO, and in
kaolinite is increasing with sampling upward both in the D1 profile some cases also a reduced concentration in K2O. Specifically, taking the
(6% → 22% → 45% and the D2 profile (3% → 5% → 6% → 11%). Illite is D1 profile for example, the semi-weathered samples have shown a
not as exactly as the same trend of kaolinite. It shows a general in- significant change in major element composition comparing with
creasing trend in the D1 profile (21% → 14% → 18% → 23% → 32%), parent rock, especially being depleted in Na2O and CaO. A further de-
and a firstly increasing then decreasing trend in the D2 profile (16% → pletion in Na2O, CaO, SiO2 and MgO whilst an enrichment in Al2O3 and
37
Fig. 4. Macro- and micro-photographs of the D1 granite parent rocks. (A) Hand specimen of the fine-grained D2 granite; (B–C) fine-grained texture under microscope
from the D2 granite sample, with microcracks observed in the granite-forming minerals; (D–F) back-scattered electron (BSE) images of accessory minerals from the D2
granite. (D) Presence of zircon, monazite and xenotime at the internal of biotite and K-feldspar grains; (E) granular-shaped monazite surrounded by K-feldspar and
quartz; (F) irregular-shaped xenotime, coexisting with a small amount of needle-shaped rutile and apatite with smaller grain size.
(Abbreviations: Qtz-Quartz, Bt-Biotite, Kfs-K feldspar, Pl-Plagioclase, Ap-Apatite, Mnz-monazite, Rt-Rutile, Xen-Xenotime; Zrn-Zircon).
Fe2O3 appeared in the highly-weathered and full-weathered samples. 5.3. Bulk REE geochemistry
In trace element, depleted Rb, Sr and Ba concentrations are ob-
served in the regolith samples compared to their corresponding parent The concentrations of REE elements of all parent rock and regolith
rocks, whereas Zr, Nb, Hf and Th show increased values in the regolith samples are shown in Tables 1 and 2. Also, Fig. 6 demonstrates the
samples compared to the parent rocks. This is not surprising that these vertical variation of some important REE indices including ΣREE, LREE,
traces elements are of immobile nature, which is in agreement with Ti. HREE, LREE/HREE, δCe and δEu.
However, some expected redox-sensitive trace elements, including Cr, For the total REE contents, it varies widely from 1 × 102 ppm order
As, Sb and U, show no steadily trends along the profiles. scale to 1 × 103 ppm order scale in whole regolith samples, showing
several fold enrichment compared to their parent rocks. Specifically,
Fig. 5. SEM images of clay minerals detected from the regolith samples. (A–B) Morphology of capillary or needle-like illite; (C) morphology of sub-euhedral or sheet-
like kaolinite; (D) occurrence of flaky kaolinite together with capillary illite.
(Abbreviations: Ill-Illite, Kln-Kaolinite).
38
Table 1
Chemical composition of the parent rock and regolith samples in the D1 profile, including organic matter (OM, wt%), major elements (wt%), and trace elements, REE
geochemistry data (in ppm).
Horizon Full weathered horizon Highly weathered horizon Semi-weathered horizon Parent rock
Sample D1-1 D1-2 D1-3 D1-4 D1-5 D1-6 D1-7 D1-8 D1-9 D1-10 D1-12
OM 9.47 2.24 2.22 2.10 2.0 1.74 1.63 1.07 1.19 0.74 –
SiO2 47.02 50.82 61.66 60.30 54.5 51.77 65.31 60.42 60.08 57.83 71.34
Al2O3 29.47 27.80 21.42 21.33 23.8 26.13 18.21 21.16 20.89 22.66 13.15
CaO 0.08 0.08 0.09 0.09 0.1 0.09 0.10 0.10 0.10 0.09 1.25
Fe2O3 7.72 6.01 4.44 4.34 6.9 6.95 3.77 4.08 5.14 5.24 2.91
K2O 2.03 2.79 3.69 6.10 3.7 3.62 6.73 7.76 6.74 5.93 5.39
MgO 0.49 0.50 0.53 0.66 0.7 0.64 0.39 0.52 0.59 0.85 0.90
MnO 0.01 0.01 0.02 0.07 0.1 0.05 0.04 0.05 0.07 0.06 0.04
Na2O 0.02 0.05 0.10 0.31 0.1 0.06 0.25 0.36 0.25 0.16 1.75
P2O5 0.06 0.06 0.06 0.07 0.1 0.07 0.08 0.11 0.09 0.08 0.19
TiO2 0.56 0.66 0.61 0.59 0.9 0.91 0.54 0.41 0.55 0.59 0.46
LOI 11.52 10.13 6.30 5.11 8.0 8.82 3.80 4.32 4.80 5.40 2.14
Total 98.98 98.91 98.92 98.97 98.8 99.11 99.22 99.29 99.30 98.89 99.52
CIA 93.26 90.49 84.66 76.64 86.09 87.39 72.00 72.02 74.66 78.57 –
Sc 13 12 9.1 8.9 14 15 7.3 11 7.8 11 9.0
Rb 170 183 200 328 244 259 311 296 304 277 213
Sr 9.3 13 21 46 27 27 44.3 56 47 32 58
Ba 145 183 277 708 672 665 659 927 699 899 496
Th 33 40 30 41 31 32 25.8 41 36 47 30
U 3.4 3.9 4.1 3.9 4.1 6.2 3.6 4.5 4.3 4.8 3.4
Pb 66 98 119 176 28 32 44.1 52 105 45 34
Nb 14 18 17 17 16 17 13 12 14 13 11
Ta 1.1 1.4 1.3 1.2 1.5 1.6 0.9 0.9 1.0 1.2 0.8
Zr 203 248 323 412 431 465 206 408 306 379 278
Hf 6.2 7.8 9.9 12 13 13 6.5 12 9.6 11 8.4
V 89 74 50 47 90 92 45 32 44 62 38
Co 5.7 6.5 6.3 23 21 10 32 21 25 14 4.9
Ga 36 32 25 28 30 32 23 24 24 31 17
Ge 1.6 1.5 1.4 1.4 1.6 1.9 1.1 1.4 1.3 1.4 1.0
Ti 3302 3920 3552 3745 5256 5511 3231 2597 3227 3612 2599
Cu 9.0 11 9.1 11 22 24 10 35 9.3 19 12
Zn 70 69 67 86 57 74 46 76 60 94 50
La 60 79 76 135 164 136 65 75 77 92 54
Ce 120 174 121 144 275 105 83 119 120 142 106
Pr 14 18 17 29 44 30 13 17 17 20 12
Nd 55 73 70 111 158 122 51 66 66 81 46
Sm 11 14 14 25 34 27 13 14 14 15 9.4
Eu 1.8 1.8 1.9 5.8 7.5 6.2 2.4 1.5 1.9 1.9 0.9
Gd 10 12 11 29 35 35 17 12. 14 13 9.7
Tb 1.6 1.6 1.5 4.8 6.6 6.1 3.6 2.2 2.2 2.0 1.7
Dy 8.7 8.3 8.3 26 27 35 22 14 12 10 11
Ho 1.6 1.4 1.6 4.9 7.4 6.7 4.3 3.2 2.3 1.9 2.6
Er 4.1 3.7 4.6 13 16 18 11 11 6.6 5.1 8.7
Tm 0.6 0.5 0.7 1.7 3.9 2.4 1.5 1.7 0.9 0.7 1.4
Yb 3.5 3.1 4.3 11 15 14 9.2 12 5.9 4.5 9.7
Lu 0.5 0.4 0.6 1.5 3.9 2.1 1.2 1.8 0.9 0.7 1.5
Y 37 32 36 129 132 188 126 84 61 46 68
∑REE 329 421 368 581 929 733 424 433 401 437 342
LREE/HREE 8.48 11.75 9.38 4.87 2.79 3.58 3.26 5.00 6.60 9.12 4.85
δCe 0.99 1.10 0.79 0.54 0.47 0.39 0.66 0.79 0.79 0.78 1.00
δEu 0.50 0.41 0.48 0.65 0.67 0.61 0.50 0.35 0.41 0.39 0.29
the D1 profile contains higher REE concentrations with ΣREE+Y values exhibiting 0.4–0.7 in the D1 profile and 0.1–0.4 in the D2 profile, re-
ranging in 329–929 ppm, which are 1–2.7 folds higher that its parent spectively. Ce anomalies are uniformly negative as well, with an only
rock (342 ppm). High values of ΣREE+Y (beyond 500 ppm) are com- exception of sample D1–2 (δCe = 1.1) in the D1 profile. For each profile,
monly present in the highly-weathered horizon in this profile. For ex- Ce negative anomaly data fluctuated widely in the upper part of the
ample, the D1–5 sample with the highest value (929 ppm) is from the profiles whereas it is relative stable in the lower part of the profiles. The
middle of the highly-weathered horizon. In contrast, the D2 profile most intensive Ce negative anomaly data (δCe = 0.4 in the D1 profile
shows a relative lower REE concentration, with ΣREE+Y values ran- and δCe = 0.3 in the D2 profile) are present in lower part of the highly-
ging in 155–226 ppm, which is only about 1.2–1.7 folds higher than its weathered horizon. In chondrite-normalized diagrams (Fig. 7), the re-
parent rock (132 ppm). For the LREE/HREE ratio, it ranges from 3.3 to golith samples display a right-inclined style REE pattern with typical
11.8 in the D1 profile and from 2.8 to 3.7 in the D2 profile, indicating an valleys of negative Eu anomalies. Such patterns are parallel to that of
essential fractionation between light REE and heavy REE in these their parent rocks but with pronounced valleys of negative Ce anomaly.
granite regolith samples. The highest value of LREE/HREE ratio is in
the highly-weathered horizon whereas the lowest value is in the semi-
weathered horizon. 5.4. Mineral REE geochemistry
Eu anomalies of all regolith samples are characteristically negative,
REE concentrations data of main granite-forming minerals and some
39
Table 2
Chemical composition of the parent rock and regolith samples in the D2 profile, including organic matter (OM, wt%), major elements (wt%), and trace elements, REE
geochemistry data (in ppm).
Horizon Full weathered horizon Highly weathered horizon Semi-weathered horizon Parent rock
Sample D2-1 D2-2 D2-3 D2-4 D2-5 D2-6 D2-7
OM 4.89 1.04 0.97 0.69 2.23 0.63

SiO2 55.86 66.27 70.92 68.33 66.00 67.39 72.58
Al2O3 25.79 19.13 16.05 17.26 18.92 18.04 14.59
CaO 0.09 0.10 0.13 0.23 0.23 0.24 0.24
Fe2O3 3.84 2.12 1.94 2.23 2.51 2.15 1.72
K2O 4.70 6.29 6.24 5.80 5.74 5.59 5.92
MgO 0.31 0.30 0.21 0.21 0.21 0.28 0.21
MnO 0.02 0.01 0.02 0.03 0.05 0.02 0.01
Na2O 0.33 0.50 0.96 2.74 2.15 3.03 2.79
P2O5 0.09 0.09 0.10 0.12 0.13 0.13 0.14
TiO2 0.25 0.16 0.16 0.18 0.18 0.14 0.14
LOI 7.88 4.17 2.66 2.16 3.16 2.27 0.69
Total 99.16 99.14 99.39 99.29 99.28 99.28 99.03
CIA 83.44 73.52 68.65 66.31 69.97 67.06 –
Sc 6.9 5.2 5.0 4.1 6.4 4.4 4.1
Rb 453 580 512 494 506 485 448
Sr 7.6 8.9 11 11 10 12 13
Ba 79 86 95 70 93 79 74
Th 26 22 21 23 30 27 19
U 8.3 6.7 8.9 14 19 18 21
Pb 59 54 37 29 31 29 31
Nb 25 18 19 19 19 17 14
Ta 3.9 3.2 2.9 3.1 2.9 2.8 2.7
Zr 136 129 118 148 185 124 100
Hf 6.0 5.7 5.0 6.5 8.4 5.6 4.6
V 11 3.0 3.5 2.9 3 2.4 2.3
Co 6.5 11 14 12 5.2 7.1 0.8
Ga 35 27 23 25 25 25 201
Ge 2.3 1.8 1.7 1.7 2.0 1.9 1.6
Ti 1349 864 889 901 928 715 684
Cu 1 0.6 5.5 0.6 0.8 0.9 0.6
Zn 49 27 38 36 50 49 34
La 23 45 20 20 28 24 18
Ce 41 23 33 34 58 49 32
Pr 5.8 11 4.8 4.8 6.9 5.9 4.3
Nd 21 40 17 17 25 21 15
Sm 6.2 10 5.1 5.3 7.7 6.2 4.4
Eu 0.4 1.2 0.2 0.2 0.2 0.2 0.2
Gd 7.3 10 6.1 6.4 9.0 7.1 5.1
Tb 1.7 2.0 1.4 1.5 2.0 1.6 1.2
Dy 10 12 8.6 9.5 12 9.3 7.0
Ho 1.8 2.0 1.6 1.7 2.1 1.6 1.2
Er 4.6 5.0 4.2 4.4 5.4 4.2 3.2
Tm 0.6 0.7 0.6 0.6 0.8 0.6 0.5
Yb 4.2 4.3 3.8 4.0 5.0 3.7 2.9
Lu 0.6 0.6 0.5 0.6 0.7 0.5 0.4
Y 50 59 48 49 61 49 36
∑REE 178 226 155 159 223 184 132
LREE/HREE 3.14 3.57 2.99 2.84 3.38 3.74 3.46
δCe 0.86 0.25 0.80 0.85 1.00 0.97 0.87
δEu 0.18 0.35 0.13 0.09 0.08 0.09 0.12
interesting REE-bearing accessory minerals (apatite, monazite and δCe = 0.9–1.0 in D2 granite). Compared with the K-feldspars, the pla-
zircon) in the granites by EMPA and LA-ICP-MS analyses are listed in gioclases show similar characteristic in total REE concentration
Tables 3 and 4. Fig. 8 demonstrates their chondrite-normalized dia- (∑REE = 15.9–103 ppm in D1 granite, ∑REE = 17.9–18 ppm in D2
grams. granite), LREE/HREE fractionation values (LREE/HREE = 8.1–20.1 in
LA-ICP-MS analyses show that the main rock-forming minerals D1 granite, LREE/HREE = 8.5–10.4 in D2 granite) and Ce anomalies
(plagioclase feldspar, K-feldspar, quartz, and biotite) of the D1 and D2 (δCe = 0.5–0.9 in D1 granite, δCe = 0.7–0.8 in D2 granite). However,
granite contain total REE concentration generally at a magnitude of the plagioclases show negligible negative Eu anomalies, which is con-
below 100 ppm individually, which is similar to other examples in sistent with the knowledge that Eu is more compatible between feldspar
previous studies (e.g., Wang et al., 2015; He et al., 2017). Specifically, and high silica magmas than other REE (Streck and Grunder, 1997). A
the total REE concentration of K-feldspars ranges from 23 to 121 ppm lower level of total REE concentration is determined in biotite. Its
and 18 to 23 ppm for D1 and D2 granite respectively. They show strong chondrite-normalized REE pattern shows characteristically LREE en-
LREE enrichment with an average LREE/HREE value of 31.8 and 39.2 richment (LREE/HREE = 9.5–44.4 in D1 granite, LREE/
at D1 granite and D2 granite, respectively. They have negative Eu HREE = 6.8–9.4 in D2 granite) with obviously negative Eu anomaly
anomalies (δEu = 0.5–0.8 in D1 granite, δEu = 0.6 to 0.8 in D2 granite) (δEu = 0.3–0.7 in D1 granite, δEu = 0.3–0.4 in D2 granite).
and negligible Ce anomalies (δCe = 0.8–1.2 in D1 granite, For accessory minerals, monazites are determined by EMPA as the
40
Fig. 6. Vertical variations of some important REE indices throughout the two studied profiles, including total REE content, LREE content, HREE content, LREE/HREE
ratios, Eu anomaly values, and Ce anomaly values. Shaded area highlights the position of the sample with the highest total REE concentration and their corresponding
situations for other REE indices.
most REE-rich phase. They contain Ce as the dominant lanthanide [i.e., 5.5. REE speciations
monazite-(Ce)] and their total REE contents range from 25.9 to 27.8 wt
%. Also, by LA-ICP-MS, apatite is verified having high levels of REE (up Percentages data of seven-step sequential extracted REE (F1–F7)
to 4413 ppm in D1 granite, up to 6534 ppm in D2 granite), especially throughout the D1 and D2 profiles are listed in Table S2 to show the REE
LREE and shows a significant LREE enrichment (LREE/HREE = 6.7 to speciation of the regolith samples. The vertical change of REE specia-
37.5 in D1 granite and 3.8 in D2 granite) and a strong negative Eu tion throughout the D1 and D2 profiles are illustrated in Fig. 9. Also,
anomaly (δEu = 0.1–0.5 in D1 granite and 0.2 in D2 granite). In the REE patterns of all seven fractions are given in Fig. 10 by chondrite
chondrite-normalized diagram, apatite REE shows a typical right-in- normalization.
clined pattern (Fig. 8). REE concentration of zircons range at Results of sequential extraction experiments indicate that REE spe-
1408–1579 ppm, and their REE distribution patterns are typical of ciation varies greatly throughout the regolith profiles and changes as a
strong HREE enrichment with LREE/HREE ≤0.01, which is quite dif- function of weathering products. Large amount of REE are present as
ferent from that of apatite and monazite. Positive Ce anomalies the ion exchangeable fraction (F2) and the residual fraction (F7). For
(δCe = 1.5–6.2) and negative Eu anomalies (δEu ≤0.04) were observed the ion exchangeable fraction REE, it is of economic significance and
in zircons (Fig. 8). can be extracted for utilizable REE resource. This fraction REE is mainly
detected at the upper part of the regolith profiles where clay minerals
are prevailing, and it accounts for 28.3%–65.1% of total amount of
REE. In distribution pattern, the ion-exchangeable fraction REE is
Fig. 7. Chondrite-normalized REE pattern of all parent rock and regolith samples from the D1 profile (left) and the D2 profile (right). Chondrite values are from
McDonough and Sun (1995).
41
Table 3
REE chemistry data of quartz, biotite, K-feldspar, plagioclase, apatite, and zircon analyzed by LA-ICP-MS (ppm) and monazite analyzed by EMPA (wt%) for the D1
granite sample.
Quartz Biotite K-feldspar Plagioclase
La 1.6 1.7 12.0 33.6 3.4 8.9 4.1 6.8 13.0 5.4 7.4 27.2 18.2 9.5 6.5 32.2 9.4 35.6 5.6 4.5
Ce 3.2 3.2 20.1 6.3 6.0 12.0 7.3 11.4 28.4 9.5 13.1 67.1 31.9 15.0 11.0 38.1 14.0 23.2 7.1 5.5
Pr 0.3 0.4 1.8 2.3 0.7 1.1 1.0 1.6 4.0 1.1 1.8 7.5 3.4 1.5 1.1 5.3 2.0 4.0 1.1 0.6
Nd 1.5 1.4 5.6 14.5 3.1 3.1 3.5 5.3 9.7 4.9 6.3 14.7 8.4 4.3 3.1 19.2 7.3 14.1 4.3 2.8
Sm 0.1 0.1 0.5 0.5 0.5 0.3 0.6 0.8 0.7 0.8 0.6 1.6 0.8 0.4 0.3 2.2 1.2 2.3 0.6 0.6
Eu 0.0 0.0 0.1 0.1 0.1 0 0.1 0.2 0.2 0.2 0.2 0.3 0.1 0 0 1.3 0.2 0.6 0.1 0.2
Gd 0.2 0.3 0.5 1.4 0.5 0.3 0.4 0.9 0.8 0.7 0.8 0.8 0.4 0.2 0.1 2.6 0.7 2.0 0.2 0.5
Tb 0.0 0.0 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.2 0.1 0 0 0.2 0.1 0.2 0.1 0.1
Dy 0.1 0.2 0.2 0.5 0.3 0.3 0.5 0.8 0.9 0.5 0.8 0.7 0.3 0.1 0.1 1.1 0.3 1.1 0.3 0.4
Ho 0.0 0.0 0 0.1 0.1 0.1 0.1 0.1 0.2 0.1 0.2 0.2 0.1 0 0 0.2 0.1 0.1 0.1 0.1
Er 0.1 0.1 0.1 0.3 0.2 0.2 0.2 0.5 0.6 0.4 0.5 0.3 0.1 0.1 0 0.5 0.2 0.4 0.1 0.3
Tm 0.0 0.0 0 0 0 0 0 0.1 0.1 0 0.1 0.1 0 0 0 0.1 0 0 0 0
Yb 0.1 0.1 0.1 0.2 0.3 0.2 0.3 0.4 0.4 0.5 0.5 0.5 0.2 0.1 0.1 0.3 0.2 0.4 0.2 0.3
Lu 0.0 0.0 0 0 0 0 0 0.1 0.1 0.1 0.1 0.1 0 0 0 0.1 0 0.1 0 0
∑REE 7.3 7.5 41 60 15 27 18 29 59 24 32 121 64 31 23 103 36 84 20 16
LREE/HREE 13.04 8.92 44.38 22.72 9.49 21.41 9.91 8.85 17.94 8.99 9.30 42.87 55.90 57.26 53.28 19.93 20.07 18.81 20.14 8.12
δEu 0.31 0.41 0.40 0.28 0.65 0.42 0.56 0.83 0.66 0.65 0.77 0.74 0.52 0.54 0.64 1.66 0.75 0.90 0.84 0.93
δCe 1.09 0.99 1.05 0.12 0.96 0.92 0.88 0.84 0.96 0.96 0.88 1.15 1.00 0.97 0.91 0.65 0.79 0.48 0.70 0.85
Apatite Zircon Monazite
La 510 445 493 773 424 717 0 0.1 0.8 5.48 5.15 5.48 5.81 4.85
Ce 1008 892 729 2005 757 1365 0.5 0.9 3.3 13.31 12.63 13.05 13.31 12.69
Pr 123 95 86 243 75 159 0 0.1 0.4 3.02 2.88 3.05 2.92 2.90
Nd 309 304 267 868 226 680 0.6 1.4 2.5 4.87 4.42 5.21 4.77 4.58
Sm 40 32 18 183 20 126 3.1 2.7 4.0 b.d.l. b.d.l. b.d.l. b.d.l. b.d.l.
Eu 2.9 2.5 3.0 4.1 3.3 8.6 0 0.1 0.2 b.d.l. b.d.l. b.d.l. b.d.l. b.d.l.
Gd 18 17 10 123 17 164 30 26 32 b.d.l. b.d.l. b.d.l. b.d.l. b.d.l.
Tb 3.5 2.8 10 13 2.3 20 17 14 14 b.d.l. b.d.l. b.d.l. b.d.l. b.d.l.
Dy 30 15 10 83 12 107 239 210 195 b.d.l. b.d.l. b.d.l. b.d.l. b.d.l.
Ho 3.4 3.1 1.5 17 1.8 22 94 79 72 b.d.l. b.d.l. b.d.l. b.d.l. b.d.l.
Er 9.7 8.4 4.5 41 4.8 64 400 365 329 b.d.l. b.d.l. b.d.l. b.d.l. b.d.l.
Tm 1.4 1.0 0.5 6.3 0.7 17 80 77 67 b.d.l. b.d.l. b.d.l. b.d.l. b.d.l.
Yb 12 7.9 4.5 49 5.0 58 600 688 578 b.d.l. b.d.l. b.d.l. b.d.l. b.d.l.
Lu 1.3 1.3 0.8 5.5 0.9 1.7 105 116 110 b.d.l. b.d.l. b.d.l. b.d.l. b.d.l.
∑REE 2073 1827 1639 4413 1549 3509 1568 1579 1408 27.40 25.94 27.74 27.75 26.01
LREE/HREE 24.82 31.42 37.52 12.11 34.39 6.74 0.00 0.00 0.01 – – – – –
δEu 0.28 0.30 0.62 0.08 0.53 0.18 0.00 0.01 0.04 – – – – –
δCe 0.96 1.01 0.80 1.13 0.96 0.95 6.23 2.62 1.48 – – – – –
b.d.l., below the detection limits.
Table 4
REE chemistry data of quartz, biotite, K-feldspar, plagioclase, apatite, and zircon analyzed by LA-ICP-MS (ppm) and monazite analyzed by EMPA (wt%) for the D2
granite.
Quartz Biotite K-feldspar Plagioclase Apatite Zircon Monazite
La 1.5 1.5 3.8 3.7 3.6 3.7 5.1 6.1 5.2 6.5 5.8 4.9 990 0.5 5.94 3.93 5.15 4.05
Ce 3.1 2.7 6.1 7.0 6.8 6.7 10 10 9.1 11 6.1 6.9 2081 1.2 15.50 10.74 12.85 10.70
Pr 0.4 0.4 0.8 0.8 0.9 0.8 1.4 1.3 1.1 1.1 0.8 0.9 332 0.2 3.20 2.40 3.00 2.70
Nd 1.3 1.0 2.5 3.0 2.9 2.5 4.2 3.2 2.2 3.1 2.6 3.3 1349 1.8 4.59 3.68 4.55 3.81
Sm 0.2 0.3 0.5 0.4 0.5 0.4 0.5 0.3 0.2 0.3 0.5 0.4 385 4.0 b.d.l. b.d.l. b.d.l. b.d.l.
Eu 0 0 0.1 0 0.1 0.1 0.1 0 0 0.1 0.2 0.1 32 0.1 b.d.l. b.d.l. b.d.l. b.d.l.
Gd 0.1 0.2 0.5 0.4 0.5 0.4 0.2 0.1 0.1 0.1 0.5 0.4 454 32 b.d.l. b.d.l. b.d.l. b.d.l.
Tb 0 0 0.1 0.1 0.1 0.1 0 0 0 0 0.1 0.1 63 15 b.d.l. b.d.l. b.d.l. b.d.l.
Dy 0.2 0.2 0.6 0.4 0.4 0.4 0.2 0.2 0.1 0.1 0.5 0.4 395 220 b.d.l. b.d.l. b.d.l. b.d.l.
Ho 0 0 0.1 0.1 0.1 0.1 0 0 0 0 0.1 0.1 70 84 b.d.l. b.d.l. b.d.l. b.d.l.
Er 0.1 0.1 0.3 0.3 0.3 0.2 0.1 0.1 0.1 0.1 0.3 0.3 180 378 b.d.l. b.d.l. b.d.l. b.d.l.
Tm 0 0 0.1 0 0 0 0 0 0 0 0.1 0 23 73 b.d.l. b.d.l. b.d.l. b.d.l.
Yb 0.2 0.1 0.3 0.2 0.2 0.2 0.1 0.1 0.1 0.1 0.3 0.2 158 629 b.d.l. b.d.l. b.d.l. b.d.l.
Lu 0 0 0 0 0 0 0 0 0 0 0.1 0 23 114 b.d.l. b.d.l. b.d.l. b.d.l.
∑REE 7.3 6.5 16 17 17 16 22 22 18 23 18 18 6534 1552 29.37 21.69 25.76 22.16
LREE/HREE 8.00 9.51 6.75 9.35 8.73 9.30 38.25 30.13 38.41 50.18 8.45 10.40 3.79 0.01 – – – –
δEu 0.31 0.34 0.36 0.29 0.36 0.37 0.60 0.63 0.65 0.77 0.97 1.02 0.23 0.02 – – – –
δCe 1.02 0.88 0.87 1.01 0.93 0.93 0.92 0.88 0.94 0.99 0.70 0.81 0.89 0.85 – – – –
b.d.l., below the detection limits.
characterized by a steep right-inclined style with typical valleys of both part of the profiles, especially in the semi-weathered horizon where lots
negative Ce and Eu anomalies by chondrite normalization. For the re- of primary minerals are observed, whereas it may decrease greatly with
sidual fraction REE, it approximately shows an opposite trend as the profiles going upward. Thus, the presence of the residual fraction REE is
above ion exchangeable fraction REE, i.e., it is prevailing at the lower controlled by those residual primary minerals.
42
Fig. 8. Chondrite normalized REE distribution patterns for Quartz, Biotite, K-feldspar, Plagioclase, Apatite and Zircon from the D1 and D2 granites, respectively.
Chondrite values are from McDonough and Sun (1995).
Small amounts of REE are present in carbonate-bound or specific 1993; Anderson et al., 2002), where Cj is the concentration in weight
adsorption fraction REE (F3). It shows about 9.1%–23.4% in percentage percent of element j, Ci is the concentration in weight percent of im-
and slightly changed through the profiles. Since there is no carbonate mobile element i, and the subscripts w and p refer to weathered material
minerals observed in the profiles, this REE fraction can be mainly at- and parent rock, respectively. In this study, Ti is selected as the most
tributed to the specific adsorption mechanism by secondary oxide mi- suitable element for normalization rather than other potential candi-
nerals (Wang et al., 1997). Iron‑manganese oxide fraction (F5) and date elements (Al, Zr, Sc, Th, etc.), because by using it we calculated
strong organic-bound fraction (F6) REE are similar in their percentage negative changes for most other elements. Results of elemental mobility
values. Both of them range in 1.6%–15.1%. Their higher values are calculation demonstrate that most of major elements are distinctly de-
corresponding to the presence of higher OM concentration and iron- pleted throughout the D1 and D2 profiles (Fig. 11-A, E). Take the D1
manganese oxide minerals, respectively. Notably, a positive Ce profile, for example, Alkali elements (Na, K) and alkaline elements (Mg
anomaly is recognized in the chondrite-normalized REE patterns of and Ca) are remarkably depleted and their τ values are mostly less than
both these two fractions. For humic acid fraction (F4), it is gen- −98%. Si and P is also strongly depleted with τ values ranging from
erally < 4%, representing a minor type of REE speciation in granite −57% to 0.01% and from −81% to −7%, respectively. Such element
regolith. loss could be interpreted by breakdown and disappearance of most of
primary granite-forming minerals as well as formation of clay minerals
with part of chemical components leached out of regolith (Weijden and
6. Discussions Middelburg, 1988). Notably, the τ values of Mn may change from ne-
gative to positive with samples downward in both the D1 and D2 pro-
6.1. Comparison of the granite chemical weathering files, indicating that enrichment of Mn present at the lower part of the
two profiles. This is consistent with the fact of secondary Mn-oxides
To evaluate the degree of granite chemical weathering, a large occurrence in the semi-weathered horizon.
number of geochemical approaches have been proposed in previous Both CIA values and mass transfer coefficient (τ) values indicate that
studies (e.g., Nesbitt and Young, 1982; Fedo et al., 1995; Ohta and Arai, the chemical weathering of the D1 profile are substantially intense as a
2007; Sheldon and Tabor, 2009; Nordt and Driese, 2010). Here, con- whole than that of the D2. This inference is in accordance with those
sidering the wide usage in granite regolith study, the chemical index of from the field observations and mineralogical analysis, i.e., the D1
alteration (CIA = [Al2O3 / (Al2O3 + CaO + Na2O + K2O)] × 100%) profile has thicker regolith depth and more abundant clay minerals
suggested by Nesbitt and Young (1982) is applied for quantification of than that of the D2 profile.
the weathering degree of the studied profiles. Calculated CIA values
range between 61.05 and 93.26 D1 profile and from 61.98 to 83.44 in
the D2 profile. 6.2. Lithological influence on the granite regolith development
In addition, element motilities during granite weathering are de-
termined in this study based on the geochemical data normalized with The formation of granite regolith involves the interaction of nu-
respect to immobile element in fresh granite. The equation for calcu- merous geological and environmental factors (e.g., Pye, 1986; Twidale
Cj, w Ci, p
lating elemental mobility coefficient (τ) is τ = C C − 1 (Braun et al., and Campbell, 1995; Butt et al., 2000; Anand and Paine, 2002;
j , p i, w
43
Fig. 9. Corresponding relationships between REE speciation and quantitative mineral compositions throughout the D1 and D2 profiles. Shaded area highlights the
position of sample with the highest ion-exchangeable REE enrichment and its corresponding situations for mineral composition. Note: F1-water soluble fraction, F2-
ion-exchangeable fraction, F3-carbonate-bound or specific adsorption fraction, F4-humicacid fraction, F5-iron-manganese oxide fraction, F6-strong organic-bound
fraction, F7-residual fraction.
(Abbreviations: Qtz-Quartz, Bt-Biotite, Kfs-K-feldspar, Pl-Plagioclase, Ill-Illite, Kln-Kaolinite).
Jimenez-Espinosa et al., 2007; Wakatsuki and Matsukura, 2008). Gen- difference between the D1 and D2 regolith, and instead it could be in-
erally, it starts with the emplacement of the granite body, followed by terpreted from a lithology-controlled point of view.
exposure to a humid subtropical or tropical climate, and finally The control of lithology on chemical weathering and regolith for-
achieved under an essentially continuous weathering episode with re- mation is taken place through some specific lithological factors, invol-
lative stable tectonic background (Butt et al., 2000; Anand and Paine, ving in mineral grain size, mineral texture, mineral composition, as well
2002). as density and width of discontinuities (i.e., joints, faults, and micro-
As for the study area, the age of D1 granite is a bit older than the D2 fractures). These factors are directly related to hydraulic conductivity of
granite. Their overlying regolith, however, seems to be formed at si- granite that controls the weathering rate and intensity through water-
milar weathering episode according to regional geology and paleocli- rock reaction (e.g., Pye, 1986; White et al., 2001; Meunier et al., 2007;
mate literatures (e.g., Guangxi bureau of Geology and Mineral Worthington et al., 2016). Based on comparison of the D1 and D2
Exploration, 1985; Guo, 2004; Li et al., 2007; Qin et al., 2010; Deng granite, a number of specific lithological factors that might have con-
et al., 2014). Specifically, after the intrusion of D1 and D2 granite units, tributed to their differential weathering are identified:
they are thought to be progressively exhumed up to land surface due to Firstly, the D1 and D2 granites contain similar mineral assemblage
regional tectonic uplift in late Cretaceous to Cenozoic and subsequent but with different mineral percentage. For granite and other compar-
denudation processes during the Yanshanian-Himalayan (Guo, 2004; able rocks, even slight differences in mineral composition can lead to
Qin et al., 2010). Starting at least Eocene till to present, the outcropped differential rates of weathering (e.g., Pye, 1986; Blum and Stillings,
Hercynian and Indosinian granites were subject to chemical weathering 1995; White et al., 2001; Graham et al., 2010; Goodfellow et al., 2011;
under a humid and hot climate condition, due to the establishment of Da Silva et al., 2017). Specifically, the D1 granite contains higher per-
Indian summer monsoon (Li et al., 2007; Deng et al., 2014). Also, from centage of biotite compared to the D2 granite (11% vs. 4%). This is
the landform aspect, the geomorphological features of the Hercynian consistent with higher MgO concentrations present in the D1 granite
and Indosinian granite terrains are relative uniform, characterized by than that of the D2 granite (0.90 wt% vs. 0.21 wt%). Biotite is generally
low elevations and hilly topography. No significant elevation and to- considered to be more susceptible to chemical weathering than felsic
pography differences are observed between them. Thus, combined with minerals such as quartz and feldspars in initial stage of weathering
above mentioned similarity in weathering history landform, we spec- (e.g., Allen and Hajek, 1989; Wakatsuki and Matsukura, 2008; Graham
ulate that the exogenic environment factors, neither paleoclimate nor et al., 2010). And notably, it may expand during weathering due to the
landform, are not the main controlling factors that result in the replacement of inter horizon K by hydrated Mg cations (Isherwood and
44
Fig. 10. Chondrite-normalized patterns of each REE fraction determined by the sequential extraction experiment for both the D1 and D2 profile samples. Chondrite
values are from McDonough and Sun (1995).
Street, 1976). This expansion may generate 30%–40% increase in vo- Secondly, the features of microcracks present in the D1 and D2
lume and helps to disaggregate granitic bedrock and clasts (Graham granite are substantially different; that is, the D1 granite shows larger
et al., 2010), or even generating oxidation-driven fracturing width of microcracks than that of the D2 granite as illustrated in Figs. 3-
(Goodfellow et al., 2011), which is a favorable factor for deep weath- B, C and 4-B, C. By statistics, the width of microcracks in the D1 granite
ering of granitic rocks. Besides biotite, the D1 granite contains higher is at a range of 2.6–82.3 μm in microscope observation scale. They are
percentage of plagioclase compared to the D2 granite (33% vs. 23%). mainly present within coarse-grained quartz and K-feldspar (Fig. 3-B,
According to the theoretical weathering order of granite-forming mi- C). In contrast, the width of microcracks in the D2 granite is only at a
nerals, plagioclase weathering follows biotite, and it is less resistant to range of 1.4–18.4 μm (Fig. 4-B, C). Regarding to the origin of these
weathering than K-feldspar (Meunier et al., 2007; Graham et al., 2010; microcracks, it is probably linked to volume expansion by biotite oxi-
Churchman and Lowe, 2012). Such theoretical inference of mineral dation (Graham et al., 2010), or stress gradients induced by heating and
weathering is consistent with our observation of vertical mineral dis- cooling and/or wetting and drying (Winkler, 1977). Whatever their
tribution through the two profiles. That is, with profile upward, pla- origin, the importance of microcracks lies in increasing the perme-
gioclase is always decreased more sharply and ultimately disappeared ability of rocks and allowing chemical reactions to take place at ex-
earlier than K-feldspar, indicating that plagioclases suffered more rapid posed mineral surfaces (Pye, 1986). Mosquera et al. (2000) indicated
weathering than K-feldspar. Once the plagioclase suffered weathering, that the microcracks that cut across grain boundaries and intergranular
it gave rise to the occurrence of secondary clay minerals as illustrated in permeability occurring along grain boundary contacts may generate
Fig. 5. Thus, the higher concentration of total clay minerals (up to 70%) transgranular permeability and account for the main pathway of hy-
in the D1 profile is most likely attributed to the higher percentage of drologic conductivity in granites. Note that quantifying the density of
plagioclase in fresh granite as well as its relative rapid weathering rate. microcracks in granites is also needed to fully understand the potential
In contrast, the less abundant total clay minerals in the D2 profile are influence of microcracks on rock weathering, like using CT scanning
associated with the high amount of K-feldspar in fresh granite and its and image analysis techniques (e.g., Zhu et al., 2011; Heriawan and
relative low weathering rate. Given the higher contents of biotite and Koike, 2015), this study, however, don't obtain relevant data due to
plagioclase in the D1 granite compared to the D2 granite, it is expected limitation of experimental conditions. Even so, from the microcrack
that more dissolution pores, open boundary space, and microfractures width point of view, we can still make a reasonable inference on its
in the D1 granite are expected due to biotite oxidation and plagioclase possible influence on rock weathering due to its importance in reg-
dissolution, and extensive solution channel network may be developed ulating hydrologic permeability. According to the well-known “cubic
through the space of weathered biotite and plagioclase. This may law” (Zimerman and Yeo, 2000) revolving in fluid flow in rocks and
contribute to a higher permeability in the D1 granite than in the D2 porous medium, the volumetric flow rate is proportional to the cubic of
granite with the weathering ongoing. the fracture width (also called fracture aperture). That means a small
45
Fig. 11. Diagrams showing depth of weathering profile versus τ value normalized by Ti. (A–D) Vertical changes of τ values for major elements, ∑REE, ∑LREE and
∑HREE, respectively in the D1 profile; (E–H) vertical changes of τ values for major elements, ∑REE, ∑LREE and ∑HREE, respectively in the D2 profile. Shaded area
highlights the position of the sample with the highest τ value for ∑REE and its corresponding situations for other elements or indices.
change in fracture width could cause a significant change in rate of weathered and up to 12 m deep regolith is developed overlying the D1
flow. In other words, the wider microcracks in granite may support a granite rather than the D2 granite.
positive contribution to an easier and deeper circulation of the meteoric
fluids.
Besides the above two factors, there are undoubtedly other litho- 6.3. Lithological influence on REE enrichment and fractionation
logical factors might have associations with the differential weathering
of the two granites. For example, the grain size of the D1 and D2 granite REE enrichment in regolith varies as a function of both parent rock
is different as well. Theoretically, as mineral grain size decreases and and weathering process (e.g., Nesbitt, 1979; Braun et al., 1990; Marker
specific surface area increases, a fine-grained rock should offer an po- and Deoliveira, 1994; Condie et al., 1995; Braun et al., 1998; Tyler,
tential increased area over which water penetration may occur, and can 2004; Laveuf and Cornu, 2009). However, it is not easy to define which
be expected to weather more rapidly than a coarse-grained rock factor between them is the dominant, and the actual situation may vary
(Berner, 1979; White and Brantley, 1995; Luttge, 2005). However, our case by case. Given the REE mobility coefficient (τ) calculated by the
observed facts are in contradiction to this theoretical inference; that is, mass balance method can reflect REE redistribution during weathering
the coarse-grained D1 granite suffered more intense weathering than (e.g. Sanematsu et al., 2011; Yusoff et al., 2013), we can use it to assess
the fine-grained D2 granite. We hypothesis that the effect of grain size the impact of weathering on REE enrichment in regolith. For this study,
was probably overridden by those of mineral composition and micro- the REE of two studied profiles, in general, have a loss (τ < 0) in the
cracks. upper part of profiles, whilst a gain (τ > 0) at a certain depth nearly
In summarize, given similar weathering history, landform and ve- corresponding to the lower part of the highly weathered horizon and
getation, the differential chemical weathering of the D1 and D2 granite, upper part of the semi-weathered horizon (Fig. 11-B–D, F–H). Con-
particularly in weathering degree and weathering depth, could be ex- trastingly, the D1 profile only has a small higher REE gain (τ = 53%)
plained as a lithology-governed effect. The D1 granite presents wider than that of the D2 profile (τ = 50%), which indicate that the weath-
microcracks and higher contents of biotite and plagioclase than the D2 ering process cannot be regarded as the major factor that caused the
granite, which are expected to generate large area of open space for large difference in REE enrichment (929 ppm vs. 226 ppm) between the
hosting more water and route them into deep depth, causing advection two regolith profiles. Instead, the dominant factor is most likely
of fluids, accelerate hydrolysis, deep oxidation and mechanical dis- stemmed from their parent granites.
aggregation (Navarre-Sitchler et al., 2013). This is why a stronger Parent granite is the primary source of REE to regolith when ex-
cluding the external input (e.g., Alderton et al., 1980; Sawka and
46
Chappell, 1988; Wu et al., 1990; Bao and Zhao, 2008). The bulk geo- The fixation of REE from weathering solution to specific weathering
chemical data reveal that the initial REE content of the two granites is product could also be ascribed to several possible mechanisms, in-
quite different. The D1 granite contains > 2 folds of the starting REE cluding secondary precipitation, adsorption, and ion exchange (Laveuf
contents of the D2 granite (342 ppm vs.132 ppm), which is the con- and Cornu, 2009). The precipitation of secondary REE-rich minerals,
genital advantage of the D1 granite that could generate REE enrichment such as rhabdophane, florencite, and cerianite (e.g., Banfield and
by weathering. Furthermore, in terms of REE distribution in granite Eggleton, 1989; Aubert et al., 2001; Kohler et al., 2005), are not ob-
mineralogy, we verified that the contribution of REE from various served in the current study. Thus, the main mechanism of REE fixation
primary minerals to the whole REE content of the granite is different in this case is most likely to be ion exchange process between clay
(Table S3). Take the D1 granite for example, results show that the main minerals and weathering solutions, and this speculation is supported by
granite-forming minerals, including plagioclase, K-feldspar, quartz and the coupling facts of abundant clay minerals (up to 70%) and high REE
biotite, account for 11.5% of total REE content in granite. In other concentration with a large proportion of ion exchangeable form (up to
words, nearly 88.5% of REE come from the accessory mineral phases. 65.1%) in regolith profile (Fig. 9). In general, Kaolinite and halloysite
This knowledge is consistent with previous studies that the accessory are believed to be main mineral hosts of REE indicated by many cases
mineral phases dominate the source of REE in granite (Alderton et al., studies from South China and comparable regions (Wu et al., 1990; Bao
1980; Sawka and Chappell, 1988; Harlavan and Erel, 2002; Bao and and Zhao, 2008; Sanematsu et al., 2009). The fixation of REE by clay
Zhao, 2008; Braun et al., 2018). Accessory mineral examinations show minerals is generally considered as occurring at permanent negative
that monazite, apatite and zircon are the main REE-bearing minerals. charge centers at the inter-layer sites due to the isomorphous replace-
Among them, apatite is easily subject to weathering especially under ment of Si4+ by Al3+, or variable negative charge centers by exposures
the impact of microorganism (Chen et al., 2006). Experimental studies of oxygen atom from broken SieO and AleO bonds at edge and basal
also indicate that apatite dissolution rates increase with decreasing pH sites for the kaolinite group minerals (Ma and Eggleton, 1999; Chi and
of solution (Harouiya et al., 2007). Theoretically, REE3+ are very Tian, 2007; Li et al., 2017). Such fixation mechanism is greatly de-
compatible in apatite and prefer the Ca2+ site in its crystal structure pendent on the nature of clay minerals and weathering solution pH
(Fleet and Pan, 1995). All these knowledge indicate that apatite is so- environment (Laveuf and Cornu, 2009; Sanematsu and Watanabe,
luble at weathering environment, and it could facilitate REE liberation 2016; Li et al., 2017). In our case, a large number of illite is observed
from the mineral phase and then redistributed into weathering system along with kaolinite. Both kaolinite and illite are negatively charged
(Yusoff et al., 2013; Foley et al., 2014). For this study, both petro- (Sanematsu and Watanabe, 2016), and they have high cation exchange
graphic observation and geochemistry analyses (i.e., the higher P2O5 capacity (CEC) to adsorb those species with positive charges from
content in the D1 granite than that of the D2 granite) indicated that the weathering solution (CEC values of illite range at 10–40 and kaolinite at
D1 granite contain more apatite than the D2 granite. Also, by plotting 3–15, Weaver and Pollar, 1973). Given the ion-exchangeable REE dis-
P2O5 content, τ value for P2O5 versus REE content (in particular the ion- tribution in regolith profile is pH-dependent (e.g., Wu et al., 1989; Bao
exchangeable fraction) throughout the regolith profile (Fig. 12), we and Zhao, 2008), we infer that the variable negative charge on the
observed that there is a greater depletion of P2O5 when the D1 granite surface of clay minerals is the main factor to adsorb REE from weath-
suffered intense weathering. And interestingly the highest depletion of ering solution, which could form an outer-sphere REE complex. In
P2O5 is well correspondent to the position of highest ion-exchangeable contrast, the permanent negative charge from the isomorphous sub-
REE enrichment. This suggests that the weathering of apatite have stitution at the inter-layer sites of clay minerals is considered not very
provided important source to the ion-exchangeable REE enrichment. significant in exchangeable REE cations. In addition, some researchers
As for monazite, in spite of its highest REE concentration over suggest that Mn-oxides may act as an effective adsorptive material re-
several magnitudes than that of the apatite, its REE can be regarded as sponsible for REE fixation in regolith through co-precipitation, ad-
immobile and cannot be activated into weathering system due to sorption, surface complex formation, ion exchange, and penetration of
monazite's robust resistant-weathering nature (Sengupta and Van the lattice (Chao and Theobald, 1976; Cao et al., 2001; Ohta and
Gosen, 2016). Hence, the REE from monazite may only present as re- Kawabe, 2001; Laveuf and Cornu, 2009). For this study, however, al-
sidual form in granite regolith. More abundant occurrence of residual though FeeMn oxides are observed in the D1 profile, SSE analyses
monazite can be inferred in the D1 regolith than that of D2 regolith by suggest that it is not the main REE-carrying mineral due to the small
the higher Th concentration (41 ppm vs. 30 ppm, the highest values in proportion of FeeMn oxides related REE fraction in whole REE spe-
two profiles respectively), because monazite is a potential source of Th ciation of the regolith (Table S2). Hence, the clay minerals (kaolinite
and U in soil and sediments (Wagani et al., 2011). This is consistent and illite) are confirmed as the dominant exchangeable REE carriers,
with higher residual REE in the D1 profile than that of D2 profile. Be- which are responsible for the fixation of activated REE during the
sides monazite, zircon may do similar contribution only to the residual granite weathering in this study.
fraction of REE in regolith. Besides the total REE enrichment, the signature of REE fractionation
After REE is activated from parent rock minerals, a high REE mo- in the two regolith profiles also largely stem from the parent granites.
bility accompanying with other mobile elements scavenge have taken Both of them are characterized by a right-incline REE fractionation
place as indicated by the above mass balance evaluation, especially in pattern (Fig. 7). Specifically, the D1 profile shows a significant REE
the upper part of the studied profiles where a low pH and a rich organic fractionation with LREE/HREE ratio progressively increased from the
matter condition exists (Fig. 11). REE was likely to be transported by bedrock (4.9) to regolith samples (2.8–11.8), whereas the D2 profile
forming bicarbonate complexes or organic matter complex (e.g., Dupre only exhibits a slightly REE fractionation with LREE/HREE ratio from
et al., 1999; Oliva et al., 1999). Organic matter, however, is supposed the bedrock (3.5) to regolith (2.8–3.7). Despite of the various extents in
not to be the dominant factor controlling the REE mobility in this case, fractionation degree, they all show an elevated ratio of LREE/HREE
due to its concentration change in the upper and middle part of the fractionation in the weathered materials relative to the bedrock. The
profiles being not quite consistent with that of the REE. Moreover, preferentially enrichment of LREE over HREE in these S-type granite
given F− bearing primary minerals (e.g., apatite) is observed in the regolith could be interpreted from two aspects. Firstly, it might be at-
studied granites, fluoride complex is another possibility for REE mo- tributed to the inheritance of REE pattern from parent rock, particularly
bility as suggested Wu et al. (1989). Therefore, we speculate that the that from accessory minerals. As above mentioned, apatite is an im-
REE transferring is an integrate effect with diversified forms linked to portant source for whole REE budget in studied granite regolith, and
both parent granite constituents (e.g., F) and exogenic constituents particularly it is characterized by a profound LREE-enriched nature,
(e.g., CO2 or OM), although it is currently not well known about how with LREE/HREE ratio at 6.7–37.5 and a typical right-incline REE
much of a role each possible forms play during their migration. patterns in chondrite-normalized diagram (Fig. 8). This means that
47
Fig. 12. Diagrams showing vertical variations of P concentration and τ values for P versus the ion-exchangeable fraction REE. Shaded area highlights the position of
the sample with the highest ion-exchangeable fraction REE concentration and their corresponding situations for P concentration and τ values for P.
more LREE rather than HREE is expected to be released into regolith formation ion-adsorption type REE mineralization. For instance, from
environment with apatite weathering. Due to the different geochemical the granitic magma evolution point of view, it is proposed that frac-
behaviors of LREE with HREE in the weathering environment, it is often tional crystallization is important for HREE enrichment (e.g., HREE-rich
preferentially scavenged by secondary minerals in regolith (Aubert Zudong pluton is strongly fractionated, Wu et al., 1992) and partial
et al., 2001; Coppin et al., 2002). Our data show that the highest LREE/ melting is important for LREE enrichment (e.g., LREE-rich Heling vol-
HREE fractionation ratio present in the upper part of the profiles, which canic rock or porphyry shows high alkali indicating low degree of
appears to be close coupling with the richest clay mineral location partial melting, Yuan et al., 2012). Hydrothermal alteration is also
(Fig. 9). This coupling implies that clay minerals mainly composed by highly emphasized for the formation of HREE-rich granites (Wu et al.,
kaolinite and illite preferentially scavenge LREE over HREE depending 1992; Wang et al., 2013; Zhao et al., 2014; Sanematsu and Watanabe,
on ionic strength, consistent with the experimental findings of Coppin 2016; Xu et al., 2017). Besides that, from the classification of I-, S-, and
et al. (2002) and Yusoff et al. (2013). Notably, zircon is the other ac- A-type granite, Bao and Zhao (2008) proposed that most of granitic
cessory mineral that have significantly influenced the LREE/HREE rocks favoring supergene REE mineralization are I-type and S-type
fractionation pattern of the regolith. It is detected as the only mineral of granitic rocks with some A-type. However, it seems that there is no
those examined here that has a left-dipping REE distribution (Fig. 8) certain preference in granite genetic type for this type deposit because
and is also quite resistant to weathering, which could explain both the various types of granites could serve as the parent rocks (Li et al.,
relatively flat REE distribution in the residual fraction (Fig. 10) and the 2017).
relative enrichment of LREE in other fractions relative to bulk granite. Even so, we consider that linking the classification of I-, S-, and A-
type granite to the development of ion-adsorption type REE deposit
remain meaningful. It could help to understand where the original
6.4. Implication for the ion-adsorption type REE resource exploration source of REE comes from. This opinion is based on the common
knowledge that different granite types usually contain different acces-
Our comparative study gives insight that the REE accumulation in sory mineral assemblage (Ishihara, 1977; Barbarin, 1999). Specifically,
granite regolith may vary greatly with lithological faces even if under more phosphate minerals may crystallize from S-type melt than those
similar climatic and topographical settings, which is consistent with the from I-type melt during magmatic differentiation, and thus a apatite
knowledge that parent granite lithology plays critical role in generating +monazite+zircon+xenotime REE-bearing mineral assemblage is
ion-adsorption type REE resource (Zhao et al., 2014; Wang et al., 2015; common in S-type granite (e.g., Bea et al., 1994; Bea, 1996; Zhou et al.,
He et al., 2017). As such, many researchers have been trying to illu- 2006). Magmatic allanite and titanite tend to occur in granites of I-type
minate the nature and origin of granitic rocks that favorable for the
48
affinities rather than those of S-type (Bea, 1996), leading to a allanite S-type granite particularly, the high P2O5 content (> 0.08 wt%) granite
+titanite+apatie+zircon REE-bearing mineral assemblage in I-type seems more optimistic in generating ion-adsorption type REE (espe-
granite (e.g., Li et al., 2007; Sanematsu et al., 2015). Contrastingly, F- cially LREE) ore by weathering than previously thought. It may shed
bearing accessory minerals are prevailing in A-type granite (Collins light on further exploration of the ion-adsorption type REE resources in
et al., 1982; Whalen et al., 1987), which lead to the occurrence of a South China and comparable regions worldwide.
fluorite+fluorocarbonates+apatite+monzaite+titanite assemblage in
many A-type granites (e.g., Price et al., 1999; Wang et al., 2014; Wang Acknowledgments
et al., 2015). Hence, based on the potential difference in REE source
accessory minerals, we infer that the weathering of I-, S-, and A-type This research was financially supported by the National Natural
granite may not exactly experience the similar REE enrichment process, Science Foundation of China (41462005), Guangxi Natural Science
and they might have different scenarios in developing ion-adsorption Foundation (2014GXNSFAA118304), and the project of Collaborative
type REE mineralization. Innovation Center for Exploration of Hidden Nonferrous Metal Deposits
In addition, for the S-type granite particularly, this study indicates and Development of New Materials in Guangxi. We acknowledge the
that the issue of P2O5 content in granite should be concerned when support of Zhang Bo from China Nonferrous Metal (Guilin) Geology and
assessing the potential of REE mineralization in terms of parent granite. Mining Co., Ltd. in field investigations. We also would like to thank
Previous studies suggested that the parent granites of low phosphate Prof. Bryan Krapez from Curtin University, Prof. Li Jianwei from China
contents (< 0.08 wt% P2O5) potentially show high possibility of de- University of Geoscience, and Prof. Feng Zuohai from Guilin University
veloping ion-exchangeable REE enrichment (e.g., Berger et al., 2014; of Technology for their assistance. We appreciate careful and ex-
Sanematsu et al., 2015; Bern et al., 2017). However, results of this study haustive comments and revisions of two anonymous journal reviewers,
indicate that greater ion-exchangeable REE content in regolith is de- which improved the manuscript.
rived from the granite containing more P2O5 (D1 granite, 0.19 wt%
P2O5), which seems an opposite result of literature mentioned. Re- Appendix A. Supplementary data
garding to this discrepancy between previous studies and this case, it
could be explained by variation in phosphate mineral types in their Supplementary data to this article can be found online at https://
respective parent granites. For this study, the significant ion-ex- doi.org/10.1016/j.chemgeo.2019.05.006.
changeable REE content in the D1 granite regolith is linked to the oc-
currences of abundant apatite. It is of weatherable nature and con- References
sidered having contributed to the REE liberation and redistribution
process during the granite weathering. In contrast, the literature-men- Alderton, D.H.M., Pearce, J.A., Potts, P.J., 1980. Rare earth element mobility during
tioned cases with high phosphate but poor ion-exchangeable REE granite alteration: evidence from southeast England. Earth Planet 149–165.
Allen, B.L., Hajek, B.F., 1989. Mineral Occurrence in Soil Environments, 2nd edition. Soil
content is linked to residual monazite and secondary REE phosphates, Sci. Soc. Am, pp. 199–278.
which has immobilized REE during weathering (Sanematsu et al., Anand, R.R., Butt, C.R.M., 2010. A guide for mineral exploration through the regolith in
2015). Hence, from the point of P2O5 content in granite, we suggest that the Yilgarn Craton, Western Australia. Aus. J. Earth. Sci. 57, 1015–1114.
Anand, R.R., Paine, M., 2002. Regolith geology of the Yilgarn Craton, Western Australia:
the high P2O5 content (> 0.08 wt%) granite seems more optimistic than implications for exploration. Aus. J. Earth. Sci. 49, 1–163.
previously thought in generating ion-adsorption type REE ore (espe- Anderson, S.P., Dietrich, W.E., Brimhall, G.H., 2002. Weathering profiles, mass-balance
cially LREE) by weathering. Note that, because the REE concentration analysis, and rates of solute loss: linkages between weathering and erosion in a small,
steep catchment. Geol. Soc. Am. Bull. 114, 1143–1158.
of phosphate is related to the source composition and crystallization Anderson, S.P., Von Blanckenburg, F., White, A.F., 2007. Physical and chemical controls
conditions of their host rocks (Teiber et al., 2015; Mao et al., 2016), on the critical zone. Elements 3, 315–319.
great variation of REE concentration in phosphate indeed exist (e.g., Andrade, F.R.D., Polo, L.A., Janasi, V.A., Souza Carvalho, F.M., 2018. Volcanic glass in
Cretaceous dacites and rhyolites of the Paraná Magmatic Province, southern Brazil:
apatite with a broad contents range from 61 to 45,300 ppm, Sha and
characterization and quantification by XRD-Rietveld. J. Volcanol. Geoth. Res. 355,
Chappell, 1999; Belousova et al., 2002). This is one possible reason 219–231.
explaining why apatite-rich granitic rocks are common but its related Aubert, D., Stille, P., Probst, A., 2001. REE fractionation during granite weathering and
ion-adsorption type REE mineralization not. removal by waters and suspended loads: Sr and Nd isotopic evidence. Geochim.
Cosmochim. Ac. 65, 387–406.
Banfield, J.F., Eggleton, R.A., 1989. Apatite replacement and rare earth mobilization,
7. Conclusions fractionation, and fixation during weathering. Clay Clay Miner. 37, 113–127.
Bao, Z., Zhao, Z., 2008. Geochemistry of mineralization with exchangeable REY in the
weathering crusts of granitic rocks in South China. Ore Geol. Rev. 33, 519–535.
Given similar weathering history, topographical setting, and vege- Barbarin, B., 1999. A review of the relationships between granitoid types, their origins
tation condition, the differential chemical weathering of the two granite and their geodynamic environments. Lithos 46, 605–626.
lithofaces is principally attributed to their intrinsic difference in granite Bea, F., 1996. Residence of REE, Y, Th and U in granites and crustal protoliths: im-
plication for the chemistry of crustal melts. J. Petrol. 37, 521–552.
lithology. A combination of several specific lithological factors, in- Bea, F., Pereira, M.D., Corretge, L.G., Fershtater, G.B., 1994. Differentiation of strongly
cluding coarser grain size, greater biotite and plagioclase content, as peraluminous, perphosphorus granites: the Pedrobernardo pluton, central Spain.
well as wider microcracks, is suggested to be responsible for the for- Geochim. Cosmochim. Ac. 58, 2609–2627.
Belousova, E.A., Griffin, W.L., O'Reilly, S.Y., Fisher, N.I., 2002. Apatite as an indicator
mation of thicker regolith.
mineral for mineral exploration: trace-element compositions and their relationship to
The large difference in REE geochemistry between the two con- host rock type. J. Geochem. Explor. 76, 45–69.
trasting profiles is closely associated with initial REE concentration and Berger, A., Janots, E., Gnos, E., Frei, R., Bernier, F., 2014. Rare earth element mineralogy
and geochemistry in a laterite profile from Madagascar. Appl. Geochem. 41,
its distribution in granite mineralogy. Significant REE enrichment
218–228.
prefers to be generated in the regolith derived from the granite with Bern, C.R., Yesavage, T., Foley, N.K., 2017. Ion-adsorption REEs in regolith of the Liberty
high initial REE concentration and rich REE-bearing accessory minerals. Hill pluton, South Carolina, USA: an effect of hydrothermal alteration. J. Geochem.
Monazite, apatite and zircon are verified as main REE source minerals Explor. 172, 29–40.
Berner, R.A., 1979. Mechanism of feldspar weathering-II. Observations of feldspars from
in two granites. Especially, apatite is suggested to be more critical in soils. Geochim. Cosmochim. Ac. 43, 1173–1186.
supplying mobile REE to regolith and further giving rise to the forma- Blum, A.E., Stillings, L.L., 1995. Feldspar dissolution kinetics: chemical weathering rates
tion of ion-exchangeable REE enrichment. of silicate minerals. Rev. Miner. 31, 291–351.
Brantley, S.L., 2010. Weathering rock to regolith. Nature 3, 305–306.
Moreover, this study combined with literature compiling propose Brantley, S.L., White, T.S., Ragnarsdottir, K.V., 2007. The critical zone: where rock meets
that weathering of the I-type, A-type and S-type granite may have dif- life. Elements 3, 307–314.
ferent REE enrichment scenarios in generating the ion-adsorption type Braun, J.J., Pagel, M., Muller, J.P., Bilong, P., Michard, A., Guillet, B., 1990. Cerium
anomalies in lateritic profiles. Geochim. Cosmochim. Ac. 54, 781–795.
REE ore due to their potential variation in REE source minerals. For the
49
Braun, J.J., Pagel, M., Herbillon, A., Rosin, C., 1993. Mobilization and redistribution of Goodfellow, B.W., Hilley, G.E., Schulz, M.S., 2011. Vadose zone controls on weathering
REEs and thorium in a syenitic lateritic profile: a mass balance study. Geochim. intensity and depth: observations from grussic saprolites. Appl. Geochem. 26,
Cosmochim. Ac. 57, 4419–4434. S36–S39.
Braun, J.J., Viers, J., Dupre, B., Polve, M., Ndam, J., Muller, J.P., 1998. Solid/liquid REE Graham, R.C., Rossi, A.M., Hubbert, K.R., 2010. Rock to regolith conversion: producing
fractionation in the lateritic system of Goyoum, east Cameroon: the implication for hospitable substrates for terrestrial ecosystems. GSA Today 20, 4–9.
the present dynamics of the soil covers of the humid tropical regions. Geochim. Guangxi bureau of Geology and Mineral Exploration, 1985. Regional Geology of Guangxi
Cosmochim. Ac. 62, 273–299. Autonomous Region. Geological Publishing House, Beijing (in Chinese).
Braun, J.J., Riotte, J., Battacharya, S., Violette, A., Oliva, P., Prunier, J., 2018. REY-Th-U Guo, T.L., 2004. Meso-Cenozoic Tectono-thermal History of Shiwandashan Basin, South
dynamics in the critical zone: combined influence of reactive bedrock accessory China. Ph. D. thesis. Tongji University.
minerals, authigenic phases, and hydrological sorting (mule hole watershed, South Harlavan, Y., Erel, Y., 2002. The release of Pb and REE from granitoids by the dissolution
India). Geochem. Geophy. Geosy. 19, 1611–1635. of accessory phases. Geochim. Cosmochim. Ac. 66, 837–848.
Butt, C.R.M., 2016. The development of regolith exploration geochemistry in the tropics Harouiya, N., Chaïrat, C., Köhler, S.J., Gout, R., Oelkers, E.H., 2007. The dissolution ki-
and sub-tropics. Ore Geol. Rev. 73, 380–393. netics and apparent solubility of natural apatite in closed reactors at temperatures
Butt, C.R.M., Lintern, M., Anand, R.R., 2000. Evolution of regoliths and landscapes in from 5 to 50 °C and pH from 1 to 6. Chem. Geol. 244, 554–568.
deeply weathered terrain-implications for geochemical exploration. Ore Geol. Rev. He, C., Xu, C., Zhao, Z., Kynicky, J., Song, W., Wang, L., 2017. Petrogenesis and miner-
16, 167–183. alization of REE-rich granites in Qingxi and Guanxi, Nanling region, South China. Ore
Cao, X.D., Chen, Y., Wang, X.R., Deng, X.H., 2001. Effects of redox potential and pH value Geol. Rev. 81, 309–325.
on the release of rare earth elements from soil. Chemosphere 44, 655–661. Henderson, P.ed., 1984. Rare Earth Element Geochemistry. Elsevier, Amsterdam.
Chao, T.T., Theobald, P.K., 1976. The significance of secondary iron and manganese Heriawan, M.N., Koike, K., 2015. Coal quality related to microfractures identified by CT
oxides in geochemical exploration. Econ. Geol. 71, 1560–1569. image analysis. Int. J. Coal Geol. 140, 97–110.
Chappell, B.W., White, A.J.R., 1974. Two contrasting granite types. Pac. Geol. 8, Hill, R.J., Howard, C.J., 1987. Quantitative phase analysis from neutron powder dif-
173–174. fraction data using the Rietveld method. J. Appl. Crystallogr. 20, 467–474.
Charoy, B., Barbey, P., 2008. Ferromagnesian silicate association in S-type granites: the Huang, D.H., Wu, C.Y., 1988. REE geochemistry and mineralization characteristics of the
Darongshan granitic complex (Guangxi, South China). B. Soc. Geol. Fr. 179, 13–27. Zudong and Guanxi Granites, Jiangxi Province. Acta Geol. Sin. 311–328.
Chen, S., Lian, B., Zheng, H.Y., 2006. Research progress of apatite microbial weathering. Huang, D.H., Wu, C.Y., Han, J., 1989. REE geochemistry and mineralization character-
Bull. Mineral. Petrol. Geochem. 25, 77–79. istics of the Zudong and Guanxi granites, Jiangxi Province. Acta Geol. Sin. 2, 139–157
Chen, C.H., Hsieh, P.S., Lee, C.Y., Zhou, H.W., 2011. Two episodes of the Indosinian (English edition).
thermal event on the South China Block: Constraints from LA-ICPMS U-Pb zircon and Isherwood, D., Street, A., 1976. Biotite-induced grussification of the Boulder Creek
electron microprobe monazite ages of the Darongshan S-type granitic suite. Granodiorite, Boulder County, Colorado. Geol. Soc. Am. Bull. 87, 366–370.
Gondwana Res. 19, 1008–1023. Ishihara, S., 1977. The magnetite-series and ilmenite-series granitoids. Min. Geol. 27,
Chi, R., Tian, J., 2007. Review of weathered crust rare earth ores. J. Chin. Rare Earth Soc. 293–305.
25, 641–650. Jimenez-Espinosa, R., Vazquez, M., Jimenez-Millan, J., 2007. Differential weathering of
Chi, R., Tian, J., Luo, X., Xu, Z., He, Z., 2012. The basic research on the weathered crust granitic stocks and landscape effects in a Mediterranean climate, Southern Iberian
elution-deposited rare earth ores. Nonferr. Metal. Sci. Eng. 3, 1–13. Massif (Spain). Catena 70, 243–252.
Churchman, G.J., Lowe, D.J., 2012. Alteration, formation, and occurrence of minerals in Kohler, S.J., Harouiya, N., Chairat, C., Oelkers, E.H., 2005. Experimental studies of REE
soils. In: Handb. Soil Sci, 2nd edition. pp. 1–72. fractionation during water-mineral interactions: REE release rates during apatite
Collins, W.J., Beams, S.D., White, A.J.R., Chappell, B.W., 1982. Nature and origin of A- dissolution from pH 2.8 to 9.2. Chem. Geol. 222, 168–182.
type granites with particular reference to southeastern Australia. Contrib. Mineral. Koppi, A.J., Edis, R., Field, D.J., Geering, H.R., Klessa, D.A., Cockayne, D.J.H., 1996. Rare
Petrol. 80, 189–200. earth element trends and cerium-uranium-manganese associations in weathered rock
Condie, K.C., Dengate, J., Cullers, R.L., 1995. Behavior of rare-earth elements in a pa- from Koongarra, northern territory, Australia. Geochim. Cosmochim. Ac. 60,
leoweathering profile on granodiorite in the front range, Colorado, USA. Geochim. 1695–1707.
Cosmochim. Ac. 59, 279–294. Kump, L.R., Brantley, S.L., Arthur, M.A., 2000. Chemical, weathering, atmospheric CO2,
Coppin, F., Berger, G., Bauer, A., Castet, S., Loubet, M., 2002. Sorption of lanthanides on and climate. Annu. Rev. Earth Pl. Sc. 28, 611–667.
smectite and kaolinite. Chem. Geol. 182, 57–68. Kynicky, J., Smith, M.P., Xu, Cheng, 2013. Diversity of rare earth deposits: the key ex-
Da Silva, Y.J.A.B., do Nascimento, C.W.A., Biondi, C.M., van Straaten, P., de Souza ample of China. Elements 8, 361–367.
Júnior, V.S., daSilva, Y.J.A.B., dos Santos, C.A., deAraújo, J.C.T., 2017. Influence of Lasaga, A.C., Soler, J.M., Ganor, J., Burch, T.E., Nagy, K.L., 1994. Chemical-weathering
metaluminous granite mineralogy on the rare earth element geochemistry of rocks rate laws and global geochemical cycles. Geochim. Cosmochim. Ac. 58, 2361–2386.
and soils along a climosequence in Brazil. Geoderma 306, 28–39. Laveuf, C., Cornu, S., 2009. A review on the potentiality of rare earth elements to trace
Deng, X., Chen, Z., Li, X., 2004. Shrimp U-Pb zircon dating of the Darongshan- pedogenetic processes. Geoderma 154, 1–12.
Shiwandashan granitoid belt in southeastern Guangxi. China. Geol. Rev. 50, Li, X.H., Li, W.X., Li, Z.X., 2007. Re-discussion on the types and tectonic significance of
426–432. the Early Yanshan granite in Nanling. Chin. Sci. Bull. 52, 981–991.
Deng, X.D., Li, J.W., Vasconcelos, P.M., Cohen, B.E., Kusky, T.M., 2014. Geochronology of Li, Y.H.M., Winston Zhao, W., Zhou, M.F., 2017. Nature of parent rocks, mineralization
the Baye Mn oxide deposit, southern Yunnan Plateau: implications for the late styles and ore genesis of regolith-hosted REE deposits in South China: an integrated
Miocene to Pleistocene paleoclimatic conditions and topographic evolution. genetic model. J. Asian Earth Sci. 148, 65–95.
Geochim. Cosmochim. Ac 139, 227–247. Liu, Q., Wu, J., 2011. The Shrimp zircon U-Pb dating and its geological significance on the
Dethier, D.P., Bove, D.J., 2011. Mineralogic and geochemical changes from alteration of volcanic-intrusive rocks in broken crater of Xunwu, Jiangxi province. J. East China
granitic rocks, Boulder Creek catchment, Colorado. Vadose Zone J. 10, 858–866. Inst. Technol. (Natural Science). 34, 231–237.
Drever, J.I., 2004. Surface and ground water, weathering, and soils. Treatise on Liu, W., Liu, C., Brantley, S.L., Xu, Z., Zhao, T., Liu, T., Yu, C., Xue, D., Zhao, Z., Cui, L.,
Geochemistry 5, 626. Zhang, Z., Fan, B., Gu, X., 2016. Deep weathering along a granite ridgeline in a
Dupre, B., Viers, J., Dandurand, J.L., Polve, M., Benezeth, P., Vervier, P., Braun, J.J., subtropical climate. Chem. Geol. 427, 17–34.
1999. Major and trace elements associated with colloids in organic-rich river waters: Luttge, A., 2005. Mineral dissolution kinetics at grain boundaries. Geochim. Cosmochim.
ultrafiltration of natural and spiked solutions. Chem. Geol. 160, 63–80. Ac. 69, A350.
Fang, Q., Feng, J., He, L., 1987. The S-type granite suite in Darongshan, Guangxi Lybrand, R.A., Rasmussen, C., 2014. Linking soil element-mass-transfer to microscale
Province. Acta Petrol. Sinica. 3, 23–35. mineral weathering across a semiarid environmental gradient. Chem. Geol. 381,
Fedo, C.M., Nesbitt, H.W., Young, G.M., 1995. Unraweiling the effects of potassium 26–39.
metasomatism in sedimentary-rocks and paleosols, with implications for paleo- Ma, C., Eggleton, R.A., 1999. Cation exchange capacity of kaolinite. Clay Clay Miner. 47,
weathering conditions and provenance. Geology 23, 921–924. 174–180.
Fleet, M.E., Pan, Y., 1995. Site preference of rare earth elements in fluorapatite. Am. Ma, Y., Liu, C.Q., 1999. Trace elemental geochemistry in chemical weathering - a case
Mineral. 80, 329–335. study from weathering crust on biotite granite, Longnan, Jiangxi. Sci. Bull. 44,
Foley, N.K., Ayuso, R.A., 2015. REE enrichment in granite-derived regolith deposits of the 2433–2437.
Southeastern United States: prospective source rocks and accumulation processes. In: Mao, M., Rukhlov, A.S., Rowins, S.M., Spence, J., Coogan, L.A., 2016. Apatite trace ele-
Symposium on Strategic and Critical Materials Proceedings. ment compositions: a robust new tool for mineral exploration. Econ. Geol. 111,
Foley, N.K., Ayuso, R.A., Bern, C.R., Hubbard, B., Vazquez, J., 2014. REE distribution and 1187–1222.
mobility in regolith formed on granite bedrock, Southeast United States. Acta Geol. Marker, A., Deoliveira, J.J., 1994. Climate and morphological control of rare-earth ele-
Sin. 88, 428–430. ment distribution in weathering mantles on Alkaline rocks. Catena 21, 179–193.
Freyssinet, P., Butt, C.R.M., Morris, R.C., Piantone, P., 2005. Ore-forming processes re- McDonough, W.F., Sun, S.S., 1995. The composition of the Earth. Chem. Geol. 120,
lated tolateritic weathering. Econ. Geol. 681–722 (Economic Geology 100th 223–254.
Anniversary Volume). Meunier, A., Sardini, P., Robinet, J.C., Pret, D., 2007. The petrography of weathering
Gislason, S.R., Oelkers, E.H., Eiriksdottir, E.S., Kardjilov, M.I., Gisladottir, G., Sigfusson, processes: facts and outlooks. Clay Miner. 42, 415–435.
B., Snorrason, A., Elefsen, S., Hardardottir, J., Torssander, P., Oskarsson, N., 2009. Mosquera, M.J., Rivas, T., Prieto, B., Silva, B., 2000. Capillary rise in granitic rocks:
Direct evidence of the feedback between climate and weathering. Earth Planet. Sci. Interpretation of kinetics on the basis of pore structure. J. Colloid. Interf. Sci. 222,
Lett. 277, 213–222. 41–45.
Goldich, S.S., 1938. A study in rock-weathering. J. Geol. 46, 17–58. Navarre-Sitchler, A.K., Cole, D.R., Rother, G., Jin, L., Buss, H.L., Brantley, S.L., 2013.
González-Álvarez, I., Boni, M., Anand, R.R., 2016. Mineral exploration in regolith- Porosity and surface area evolution during weathering of two igneous rocks. Geochim
dominated terrains: Global considerations and challenges. Ore Geol. Rev. 73, Cosmochim. Ac. 109, 400–413.
375–379. Nelson, D.W., Sommers, L.E., Sparks, D.L., Page, A.L., Helmke, P.A., 1996. Total carbon,
50
organic carbon and organic matter. Methods of Soil Analysis Part-chemical Methods Twidale, C.R., Campbell, E.M., 1995. Pre-Quaternary landforms in the low latitude con-
1996, 961–1010. text: the example of Australia. Geomorphology 12, 17–35.
Nesbitt, H.W., 1979. Mobility and fractionation of rare earth elements during weathering Tyler, G., 2004. Rare earth elements in soil and plant systems - a review. Plant Soil 267,
of granodiorite. Nature 279, 206–210. 191–206.
Nesbitt, H.W., Wilson, R.E., 1992. Recent chemical-weathering of basalts. Am. J. Sci. 292, Wagani, I., Pagel, M., Janots, E., Gros, E., 2011. Monazite in the Tim Mersoi basin, Niger:
740–777. provenance and contribution to the uranium budget in siliciclastic sediments. Can.
Nesbitt, H.W., Young, G.M., 1982. Early Proterozoic climates and plate motions inferred Mineral. 49, 487–501.
from major element chemistry of lutites. Nature 299, 715–717. Wakatsuki, T., Matsukura, Y., 2008. Lithological effects in soil formation and soil slips on
Nordt, L.C., Driese, S.D., 2010. New weathering index improves paleorainfall estimates weathering-limited slopes underlain by granitic bedrocks in Japan. Catena 72,
from Vertisols. Geology 38, 407–410. 153–168.
Ohta, T., Arai, H., 2007. Statistical empirical index of chemical weathering in igneous Wang, S.N., 1991. The Darongshan and Shiwandashan Area of Guangxi Province, China.
rocks: a new tool for evaluating the degree of weathering. Chem. Geol. 240, 280–297. Acta Petrol. Sin. 2, 73–80.
Ohta, A., Kawabe, I., 2001. REE (III) adsorption onto Mn dioxide (delta-MnO2) and Fe Wang, L., Wang, Y., Zhang, S., 1997. Speciation of rare earth elements in different types of
oxyhydroxide: Ce(III) oxidation by delta-MnO2. Geochim. Cosmochim. Ac. 65, soils in China. J. Chin. Rare Earth Soc. 15, 64–69.
695–703. Wang, D., Zhao, Z., Yu, Y., Zhao, T., Li, J., Dai, J., Liu, X., He, H., 2013. Progress, pro-
Oliva, P., Viers, J., Dupre, B., Fortune, J.P., Martin, F., Braun, J.J., Nahon, D., Robain, H., blems and research orientation of ion-adsorption type rare earth resources. Rock
1999. The effect of organic matter on chemical weathering: Study of a small tropical Miner. Anal. 32, 796–802.
watershed: Nsimi-Zoetele site, Cameroon. Geochim. Cosmochim. Ac. 63, 4013–4035. Wang, Z., Chen, B., Ma, X., 2014. Petrogenesis of the Late Mesozoic Guposhan composite
Oliva, P., Viers, J., Dupré, B., 2003. Chemical weathering in granitic environments. Chem. plutons from the Nanling Range, South China: implications for W-SN mineralization.
Geol. 202, 225–256. Am. J. Sci. 314, 235–277.
Ollier, C.D., Chan, R.A., Craig, M.A., Gibson, D.L., 1988. Aspects of landscape history and Wang, L., Xu, C., Zhao, Z., Song, W., Kynicky, J., 2015. Petrological and geochemical
regolith in the Kalgoorlie region, Western Australia. BMR J. Aust. Geol. Geophys. 10, characteristics of Zhaibei granites in Nanling region, Southeast China: implications
309–321. for REE mineralization. Ore Geol. Rev. 64, 569–582.
Patino, L.C., Velbel, M.A., Price, J.R., Wade, J.A., 2003. Trace element mobility during Weaver, C.E., Pollar, L.D., 1973. The Chemistry of Clay Minerals. Elsevier, Amsterdam.
spheroidal weathering of basalts and andesites in Hawaii and Guatemala. Chem. Weijden, C.H.V.D., Middelburg, J.J., 1988. Hydrogeochemical factors controlling the
Geol. 202, 343–364. inorganic composition of the river Rhine. Chem. Geol. 70, 18.
Price, J.D., Hogan, J.P., Gilbert, M.C., London, D., Morgan VI, G.B., 1999. Experimental Whalen, J.B., Currie, K.L., Chappell, B.W., 1987. A-type granites: geochemical char-
study of titanite–fluorite equilibria in the A-type Mount Scott Granite; implications acteristics, discrimination and petrogenesis. Contrib. Geol. Mineral Resour. 95,
for assessing F contents of felsic magma. Geology 27, 951–954. 407–419.
Pye, K., 1986. Mineralogical and textual controls on the weathering of granitoid rocks. White, A.F., Brantley, S.L., 1995. Chemical weathering rates of silicate minerals: an
Catena 13, 47–57. overview. In: Chemical Weathering. Rates of Silicate Minerals. vol. 31. pp. 1–22.
Qi, C.S., Deng, X.G., Li, W.X., Li, X.H., Yang, Y.H., Xie, L.W., 2007. Origin of Sangongshan White, A.F., Bullen, T.D., Schulz, M.S., Blum, A.E., Huntington, T.G., Peters, N.E., 2001.
- Shiwandashan S - type granitoid belt from southeastern Guangxi: geochemistry and Differential rates of feldspar weathering in granitic regoliths. Geochim. Cosmochim.
Sr-Nd-Hf isotope constraints. Acta Petrol. Sin. 23, 403–412. Ac. 65, 847–869.
Qin, X.F., Pan, L.Z., Hu, G., Zhou, F.S., Xie, L.F., Li, G.N., 2010. Apatite fission track Winkler, E.M., 1977. Insolation of rocks and stones a hot item. Geology 5, 188–189.
evidence for Mesozoic-Cenozoic differential uplift on both sides of Qinfang tectonic Wood, S.A., 1990. The aqueous geochemistry of the rare-earth elements and Yttrium. 1.
belt in Southern Guangxi. J. Guilin Univ. Technol. 30 (3), 325–331. Review of available low-temperature data for inorganic complexes and the inorganic
Rietveld, H.M., 2014. The Rietveld method. Physicacripta 89, 710–712. REE speciations of natural-waters. Chem. Geol. 82, 159–186.
Sanematsu, K., Watanabe, Y., 2016. Characteristics and genesis of ion adsorption-type Worthington, S.R.H., Davies, G.J., Alexander Jr., E.C., 2016. Enhancement of bedrock
rare earth element deposits. In: Verplanck, P.L., Hitzman, M.W. (Eds.), Reviews in permeability by weathering. Earth-Sci. Rev. 160, 188–202.
Economic Geology. vol. Volume 18. Society of Economic Geologists, Inc., pp. 55–79. Wu, C.Y., Huang, D.H., Guo, Z.X., 1989. REE geochemistry in the weathered crust of
Sanematsu, K., Murakami, Hiroyasu, Watanabe, Y., 2009. Enrichment of rare earth ele- granites, Longnan area, Jiangxi Province. Acta Geol. Sin. 4, 349–362.
ments (REE) in granitic rocks and their weathered crusts in central and southern Laos. Wu, C.Y., Huang, D.H., Guo, Z., 1990. REE geochemistry in the weathered crust of
Bull. Geol. Surv. Japan. 60, 527–558. granites, Longnan area, Jiangxi Province. Acta Geol. Sin. 3, 193–210.
Sanematsu, K., Moriyama, T., Sotouky, L., Watanabe, Y., 2011. Mobility of rare earth Wu, C.Y., Bai, G., Huang, D.H., 1992. Characteristics and significance of HREE-rich
elements in basalt-derived laterite at the Bolaven Plateau, Southern Laos. Resour. granitoids of the Nanling Mountain area. Bull. Chin. Acad. Geol. Sci. 25, 43–58.
Geol. 61, 140–158. Xie, Y.L., Hou, Z., Goldfarb, R.J., Guo, X., Wang, L., 2016. Rare earth element deposits in
Sanematsu, K., Kon, Y., Imai, A., Watanabe, K., Watanabe, Y., 2013. Geochemical and China. Rev. Econ. Geol. 18, 115–136.
mineralogical characteristics of ion-adsorption type REE mineralization in Phuket, Xu, C., Kynicky, J., Smith, M.P., et al., 2017. Origin of heavy rare earth mineralization in
Thailand. Mineral. Deposita 48, 437–451. South China. Nat. Commun. 8, 14598.
Sanematsu, K., Kon, Y., Imai, A., 2015. Influence of phosphate on mobility and adsorption Yang, X.J., Lin, A., Li, X.L., Wu, Y., Zhou, W., Chen, Z., 2013. China's ion-adsorption rare
of REEs during weathering of granites in Thailand. J. Asian Earth Sci. 111, 14–30. earth resources, mining consequences and preservation. Environ. Geol. 8, 131–136.
Sawka, W.N., Chappell, B.W., 1988. Fractionation of uranium, thorium and rare earth Yang, C., Gu, Z., Wang, B., Liu, Y., Luo, Y., Xu, Q., Tan, Y., Tang, Y., Yang, B., 2015.
elements in a vertically zoned granodiorite: implications for heat production dis- Content and mineralization of Xunwu rare earth ore. Bull. Chin. Ceramic Soc. 34,
tribution in the Sierra Nevada batholith, California, U. S. A. Geochim. Cosmochim. 231–233.
Ac. 52, 1131–1144. Yuan, Z.X., Wu, C.Y., Xu, L.M., Ni, Y.X., 1992. The distribution of trace elements in
Scott, K.M., Pain, C.F., 2008. Regolith Science. CSIRO Pub, Dordrecht, Netherlands, granitoids of the Nanling region of China. Geochimica 4, 334–345.
Collingwood, Australia. Yuan, Z.X., Li, J.K., Wang, D.H., et al., 2012. Metallogenetic Regularity of REE Deposit.
Sengupta, D., Van Gosen, B.S., 2016. Placer-type rare earth element deposits. Rev. Econ. Geological publication, Beijing (in Chinese).
Geol. 18, 81–100. Yusoff, Z.M., Ngwenya, B.T., Parsons, I., 2013. Mobility and fractionation of REEs during
Sha, L.K., Chappell, B.W., 1999. Apatite chemical composition, determined by electron deep weathering of geochemically contrasting granites in a tropical setting, Malaysia.
microprobe and laser-ablation inductively coupled plasma mass spectrometry, as a Chem. Geol. 349, 71–86.
probe into granite petrogenesis. Geochim. Cosmochim. Ac. 63, 3861–3881. Zhang, Z., 1990. A study on weathering crust ion adsorption type REE deposits, South
Sheldon, N.D., Tabor, N.J., 2009. Quantitative paleoenvironmental and paleoclimatic China. Contrib. Geol. Mineral Resour. Res. 5, 57–71.
reconstruction using paleosols. Earth-Sci. Rev. 95, 1–52. Zhao, Z., Wang, D., Chen, Z., Chen, Z., Zhang, G., Liu, X., 2014. Zircon U-Pb age, en-
Shi, Y.H., Li, X.B., Yang, Z.P., 2014. Application of inductively coupled plasma-mass dogenic mineralization and petrogenesis of rare earth ore-bearing granite in Longnan,
spectrometry to the speciation analysis of rare earth elements in ion-adsorption rare Jiangxi province. Acta Geol. Sin. 35, 719–725.
earth ore. Metall. Anal. 34, 14–19. Zhao, Z., Wang, D., Chen, Z., Chen, Z., 2017. Progress and research on metallogenic
Simandl, G.J., 2014. Geology and market-dependent significance of rare earth element regularity of ion-absorbtion type REE deposit in the Nanling range. Acta Geol. Sin.
resources. Mineral. Deposita 49, 889–904. 91, 2814–2827 (in Chinese with English abstract).
Song, Y.H., Shen, L.P., 1987. REE geochemistry of the weathered crust of acid volcanic Zhou, X., Sun, T., Shen, W., Shu, L., Niu, Y., 2006. Petrogenesis of Mesozoic granitoids
rocks - an experimental study. Geochemistry 6, 165–176. and volcanic rocks in South China: a response to tectonic evolution. Episodes 29, 26.
Streck, M.J., Grunder, A.L., 1997. Compositional gradients and gaps in high-silica rhyo- Zhu, H.G., Xie, H.P., Yi, C., 2011. CT identification of microcracks evolution for rock
lites of the Rattlesnake Tuff, Oregon. J. Petrol. 38, 133–163. materials. Chin. J. Rock Mech. Eng. 30 (6), 1230–1238.
Teiber, H., Marks, M.A.W., Arzamastsev, A.A., Wenzel, T., Markl, G., 2015. Compositional Zimerman, R., Yeo, I., 2000. Fluid flow in rock fractures: from the Navier-Stokes
variation in apatite from various host rocks: clues with regards to source composition Equations to the cubic law. In: Dynamics of Fluids in Fractured Rock. American
and crystallization conditions. Neues. Jb. Miner. Abh. 151–167. Geophysical Union, Washington DC, pp. 213–224.
51

Chemical Geology: A A A A A A B B

Uploaded by

Copyright:

Available Formats

Chemical Geology: A A A A A A B B

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Chemical Geology: A A A A A A B B

Uploaded by

Copyright:

Available Formats

Chemical Geology 520 (2019) 33–51

Contents lists available at ScienceDirect

Chemical weathering of S-type granite and formation of Rare Earth Element T

even locally beyond 30 m in some terrains with low-moderate relief

Sample D2-1 D2-2 D2-3 D2-4 D2-5 D2-6 D2-7

OM 4.89 1.04 0.97 0.69 2.23 0.63

Apatite Zircon Monazite

b.d.l., below the detection limits.

b.d.l., below the detection limits.

You might also like