Fe2O3 Reduction
Fe2O3 Reduction
Fe2O3 Reduction
net/publication/257373364
CITATIONS READS
207 8,188
4 authors, including:
Zbigniew Kaszkur
Institute of Physical Chemistry, Polish Academy of Science
114 PUBLICATIONS 1,708 CITATIONS
SEE PROFILE
Some of the authors of this publication are also working on these related projects:
Structure dynamics of nanocrystalline gold during WGS and PROX reactions- new approach to the reaction mechanism. View project
Structure evolution of metal-catalyst during chemical reaction at the surface View project
All content following this page was uploaded by Zbigniew Kaszkur on 18 November 2017.
PII: S0926-860X(10)00264-4
DOI: doi:10.1016/j.apcata.2010.04.003
Reference: APCATA 12529
This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.
*Graphical Abstract
t
in two steps, Fe2O3→Fe3O4 →Fe, in dry 5,8%H2+Ar mixture, but
ip
in three steps, Fe2O3→Fe3O4→”FeO”→Fe, in wet 5,8%H2+Ar+H2O mixture.
cr
us
an
M
ed
pt
ce
Ac
1
Page 1 of 18
*Manuscript
t
ip
Poland
cr
Key words: reduction of iron oxide, temperature-programmed reduction, TPR, XRD, Fe2O3,
us
hematite, magnetite, wustite, iron.
Abstract an
The reduction of Fe2O3 with hydrogen was studied. The thermodynamic analysis of
M
the process implied that temperature-programmed reduction of the oxide should proceed in
three steps, i.e. Fe2O3→Fe3O4→”FeO”→Fe, at X H 2O X H 2 ratio over 0.35, but in two
ed
steps, i.e. Fe2O3→Fe3O4→Fe, below that value. The idea was verified by TPR and XRD
studies. Generally, the examinations confirmed the suggestions. The reduction is three-step
reaction at high X H 2O X H 2 ratio, but two-step reaction at low that ratio. Additionally, it
pt
was revealed that at extremely low X H 2O X H 2 ratio the TP reduction is a one-step reaction,
ce
i.e. Fe2O3→Fe.
Ac
Introduction
The reduction of Fe2O3 with hydrogen was a subject of numerous studies [1-7], but
there still have been differing opinions concerning the reaction route, even for unsupported
samples. Thus, there are TPR studies, which point to the two-step mechanism, i.e.
Fe2O3→Fe3O4→Fe [14], and there are reports indicating the three-step mechanism, i.e.
Fe2O3→Fe3O4→”FeO”→Fe [57]. The aim of this examination is to clarify mechanism of
Fe2O3 reduction. This knowledge is of great importance when TPR method is applied to
characterization of iron supported catalysts.
1
Page 2 of 18
TPR method is a relatively simple technique, but analysis of the experimental results
is often difficult. The problem is that the obtained TPR profile depends not only on the type
of the examined sample, but also on the experimental parameters of the measurements such
as weight of the sample, composition and flow rate of the reducing mixture, and the
temperature ramp rate [2,7]. All these parameters collectively influence two essential
reduction parameters, i.e. H2 and H2O concentrations, which in turn affect the reduction
rate. The problem of H2 depletion in the course of TPR tests has been revealed and
t
ip
extensively discussed by Monti and Baiker [8]. As a result, they proposed that the
experimental parameter of TPR test should be selected so as to get only a small decline of
cr
H2 concentration. In contrast to that question, insufficient attention has still been paid to the
effect of water. Water is the inevitable product in TPR experiment and its concentration
us
generally changes in the course of TPR test from a negligibly small value to a significant
value, i.e. a few orders of magnitude, and may radically change position and shape of the
recorded TPR profile [2, 9-11]. an
M
Experimental
Apparatus
ed
The examinations were carried out in a glass flow system [12], equipped with a
gradientless microreactor [13]. Two H2+Ar mixtures (83.6 and 5.8 mol % H2) of high purity
pt
(99.999 %) were used in this study. The gas stream required was fed to the measuring
system by a selecting valve and, before entering the reactor, it was additionally purified
ce
from traces of oxygen and water in MnO/SiO2 column. The gas stream leaving the reactor
was examined with the TCD cell and the results were collected with a computer-controlled
system.
Ac
The TPR studies were supplemented with XRD examinations. The measurements
were carried out using powder diffractometer D5005 manufactured by Siemens (Bruker
AXS) using Ni-filtered Cu-K radiation. A part of these examinations were performed in
situ in the course of TP reduction using a home made metal camera [14] and a position
sensitive detector CPS120 manufactured by INEL.
The ex situ XRD measurements of some samples were performed when their TP
reduction continued to high temperature. In this case, the reduction was carried out in the
flow system described above, whereupon the sample was cooled down, flushed with a He
2
Page 3 of 18
stream and passivated with small O2 pulses. The XRD studies of these samples were
performed with the use of standard scintillation detector in a step scan mode.
Measurement procedure
The raw material for the TPR tests was Fe2O3 powder p.a. (Merck, No. 4625270).
The XRD phase analysis of the oxide demonstrated hexagonal Fe2O3 phase. Total surface
t
area of the material was 2.2 m2/g, which corresponds to particle size of 0.5 m. Scanning
ip
electron microscope showed a sharp particles distribution with an average diameter of about
cr
0.4 m.
Prior to the TPR test, the examined sample was dried in situ in He stream (400 ºC,
us
30 cm3/min, 30 min). Then the reactor with the sample was cooled down to room
temperature whereupon the stream was replaced with one of the H2+Ar mixtures of 30
an
cm3/min. Generally, the TPR measurements were carried out at 5ºC/min. The stream leaving
the reactor was dried in a trap at –78 ºC and consumption of hydrogen was measured. After
M
that, the reactor with the examined sample was by-passed and calibration of the system was
performed, introducing well-defined H2 pulse into the H2+Ar stream.
A part of the TPR measurements presented in this communication was carried out
ed
with H2+Ar mixtures saturated with water at a selected temperature in order to get a desired
H2O concentration. These experiments are referred to as “wet” in the subsequent parts of
pt
this communication, in contrast with the “dry” experiments where the unsaturated H2+Ar
mixture was used.
ce
3
Page 4 of 18
X H 2O
A 3Fe 2 O 3 H 2 2Fe 3O 4 H 2 O KA
XH2
X H2O
B 0.25 Fe 3O 4 H 2 0.75 Fe H 2 O KB
X H2
X H2O
B1 1.202 Fe 3O 4 H 2 3.808 Fe 0.947O H 2 O K B1
X H2
t
X H2O
Fe 0.947O H 2 0.947 Fe H 2 O K B2
ip
B2
X H2
cr
us
an
M
ed
pt
express the equilibrium state of the system and thin, dotted lines show the equilibrium state
for selected reactions. Hereafter “FeO” denotes wustite, thermodynamically stable
Ac
The relations presented in Fig.1 allow to anticipate the reaction route in the course
of TP reduction of Fe2O3. They show that:
The reduction Fe2O3→Fe3O4, expressed by line a in Fig.1, is thermodynamically
possible even at high X H 2O X H 2 ratio.
4
Page 5 of 18
The reduction Fe3O4→Fe may occur only at sufficiently low X H 2O X H 2 ratio, and the
at X H 2O X H 2 below ca. 0.35, line b’, the reduction should be a one-step reaction,
Fe3O4→Fe, and
at X H 2O X H 2 over ca. 0.35, line b”, the reduction should be a two-step reaction:
Fe3O4→”FeO”→Fe.
t
ip
TPR examinations of Fe2O3
cr
Figure 2 presents preliminary TPR tests of Fe2O3, in which the two reducing H2+Ar
us
mixtures, H2 concentration of 83.6 or 5.8 mol %, and two sizes of the examined sample, 5
or 25 mg, were used. The results demonstrate a strong effect of experimental parameters on
the TPR profiles for Fe2O3 and thereby they justify the diverse opinions on the reaction
an
mechanism. Referring back to the results in Fig.2, it is important that these profiles are
similar to the respective profiles, as reported previously [1-4, 7]. Thus:
M
profile 1, obtained for 5 mg sample and 83.6%H2+Ar mixture is similar to the profiles
obtained by Wimmers et al. for small samples and 67%H2+Ar mixture [2],
profile 2 is much the same as the one obtained by Munteanu et al. [3], and also the one
ed
obtained by Kock et al. [1], and identical with the profile obtained in pure H2 by
Jóźwiak et al. [7],
pt
5%H2+Ar mixture.
Ac
5
Page 6 of 18
t
ip
cr
us
Fig.2.The effect of experimental conditions on TPR profiles recorded for F2O3
an
Closer inspection of the results in Fig.2 leads to two important statements: (i) H2
concentration decreases only slightly in these experiments, which suggests that the change
M
does not affect the reduction rate and (ii) the five-fold increase of the size of the examined
sample increases H2O evolution only two-fold, which, indicates that water strongly retards
ed
Fe2O3 reduction.
Figure 3 presents the effect of water, 0.6% and 1.2% in 5.8%H2+Ar mixture on TP
pt
reduction of 5 mg Fe2O3 sample. In order to attain a better clarity, the profiles are shown
only in terms of H2O concentration/production. The obtained results show a complex effect
ce
of water on the reaction. For the dry test, the TPR profile consists of two peaks of H2O
production, and , while for the wet tests, the peak splits into two separate peaks, 1
Ac
6
Page 7 of 18
t
ip
cr
us
Fig.3. The effect of water on TP reduction of 5 mg Fe2O3 sample
an
M
ed
pt
ce
Ac
7
Page 8 of 18
Supplementary data on Fe2O3 reduction were obtained using the 83.6%H2+Ar
mixture (see Fig.5). In this case even for: (i) large samples of 50 mg, (ii) large temperature
ramp of 10ºC/min and (iii) large concentration of water in the reducing mixture of 2.5%, the
TP tests suggest the two-step reduction route.
t
ip
cr
us
an
M
Fig.5. The temperature-programmed reduction of 50 mg Fe2O3 sample in
ed
83.6%H2+Ar mixture
pt
The TPR results in Figs.3-5 appear to confirm the thermodynamic suggestions from
Fig.1. that: (i) at low X H 2O X H 2 ratio the Fe2O3 reduction occurs as a two-step reaction
ce
8
Page 9 of 18
Table 1. Fraction of water produced during the subsequent stages of Fe2O3 reduction
t
ip
7 5 mg, 1.2%H2O 0,116 0,194 0,689
8 25 mg 0,101 0,899
cr
9 25 mg, 0.6%H2O 0,105 ~0,192 ~0.703
10 25 mg, 1.2% H2O 0,107 0,189 0,704
us
11 50 mg ~0,097 ~0,903
12 50 mg, 2.5%H2O 0,111 0,889
determined size of individual H2O peak with the theoretical values which confirms the two
following routes of Fe2O3 reduction:
at low X H 2O X H 2 ratio the two-step route: Fe2O3 Fe3O4
Fe , and
pt
1 2
Fe2 O3 Fe3O 4 " FeO" Fe .
Moreover, a closer inspection of the profile 7 in Fig.3 implies that pure Fe3O4 and
Ac
“FeO” intermediate phases appeared during the experiment at 500 and 655 ºC, respectively.
XRD study
The examinations were performed for the samples reduced ex situ or in situ in the
way simulating TPR tests of 5 mg Fe2O3 samples. Namely, 15 mg Fe2O3 samples and
three-fold greater flow rate of reducing mixture were used in the reduction in order to attain
possibly the same route of the reaction. The ex situ TP reduction of 15 mg Fe2O3 sample in
the 5.8%+Ar+1.2%H2O mixture (90 cm3/min) gave the same result as the one shown in
9
Page 10 of 18
Fig.3 (curve 7), which indicates the same route of the reduction. At the same time this result
indicated that an external mass transfer does not influence the reduction.
t
ip
which confirms the three-step model of Fe2O3 reduction in the 5.8%H2+Ar+1.2%H2O
mixture.
cr
us
an
M
ed
pt
ce
reduction.
Figure 7 presents the XRD spectra recorded in situ during the TP reduction
simulating the exp.5 in Fig.3. In this case, the three phases are observed, hexagonal Fe2O3,
cubic Fe3O4, and cubic Fe, and no cubic “FeO” phase is recorded, which is in line with
equilibrium thermodynamic suggestions. These results show that at low H2O concentration
(occurring only as a result of Fe2O3 reduction) Fe2O3 reduction follows the two-step
mechanism.
10
Page 11 of 18
t
ip
cr
us
an
Fig.7. The XRD spectra recorded during the TP reduction of Fe2O3 in the
5.8%H2+Ar mixture. The numbers tagging the reflections stand for: 1 hexagonal Fe2O3, 2
– cubic Fe3O4 and 4 cubic Fe. Two large reflections of Fe2O3 and of Fe3O4 (2 theta about
M
35.5 and 61.5 deg) are not marked as their position is close to each other and they do not
show the phase transition clearly.
ed
Figure 8 presents the XRD spectra recorded in situ during the TP reduction
pt
simulating the exp.1 in Fig.2. In this case, only two phases are recorded, hexagonal Fe2O3,
and cubic Fe. These results demonstrate that at low H2O concentration but at high H2
ce
concentration the reduction follows a direct transformation of Fe2O3 into metal iron.
Ac
11
Page 12 of 18
t
ip
cr
us
Fig.8. The XRD spectra recorded during the TP reduction of Fe2O3 in the
83.6%H2+Ar mixture. The numbers tagging the reflections stand for: 1 hexagonal Fe2O3
and 4 – cubic Fe.
an
M
Discussion
The above experimental results indicate that water: (i) does not affect Fe2O3→Fe3O4
ed
reaction, (ii) moderately retards Fe3O4→”FeO” reaction, and (iii) strongly retards
”FeO”→Fe reaction.
pt
The effect of water on the reduction of Fe2O3 appears clearer, if we compare the
reaction quotient, i.e. Q X H 2O X H 2 ratio in the course of the reduction, with the
ce
equilibrium constants calculated for each individual reaction (see Figs.9-11). Generally, the
comparisons confirm the thermodynamic suggestions, however, some experimental results
Ac
seem incompatible with thermodynamic data. Namely, the profile 6 in Fig.9 and profile 9 in
Fig.10 suggests wustite formation below 571 ºC though the thermodynamics predicts the
formation of metallic iron. In relation to this issue, it should be kept in mind that admittedly,
the equilibrium thermodynamic relations in Figs.9-11 show that Fe3O4→Fe reaction is the
most favourable below 571 ºC, but still, the Fe3O4→”FeO” reaction is also possible, and it
is the kinetics to decide which of the two possible reactions will actually proceed.
12
Page 13 of 18
t
ip
cr
us
Fig.9. Equilibrium constant and quotient of Fe2O3 reduction for 5 mg Fe2O3 sample
an
M
ed
pt
ce
Ac
13
Page 14 of 18
t
ip
cr
us
Fig.11. Equilibrium constant and quotient of Fe2O3 reduction for 50 mg Fe2O3
sample
an
The formation of “FeO” phase below 571 ºC was evidenced by ex situ XRD
examination of the sample reduced in the way simulating exp.6 in Fig.3 (also in Fig.9). The
M
TP reduction was terminated at 561 ºC, i.e. 10 ºC below the eutectoid point. The obtained
spectrum (see Fig.12) shows, next to cubic Fe3O4 phase, considerable quantity of “FeO”
ed
phase and traces of Fe phases. Closer analysis of the spectrum shows that the “FeO”
reflections are shifted to low angle by 0.3-0.5 deg in comparison with “FeO” reflections in
Fig.6, which suggests slightly lower content of oxygen in this intermediate.
pt
ce
Ac
14
Page 15 of 18
Fig.12. The XRD patterns of Fe2O3 reduced ex situ in the 5.8%H2+Ar+0.6%H2O
mixture. Terminal temperature of the sample TP reduction was 561 ºC. The numbers
tagging the reflections stand for: 2 – cubic Fe3O4, 3 – cubic “FeO” and 4 cubic Fe.
The formation of “FeO” below eutectoid point implies that Fe3O4→”FeO” reaction
is relatively fast and the appearing " FeO" is immediately reduced to metal iron at small
t
H2O concentration, but it forms a stable phase at large H2O concentration. Regular position
ip
of and 1 peaks in relation to KB1 line (see Figs.9, 10, 11) appears to support the
cr
suggestion. Therefore, it is supposed that the two-step Fe2O3 reduction observed at low
X H 2O X H 2 ratio is in fact the three-step reaction, Fe2O3→Fe3O4→”FeO”→Fe, in which the
us
third step, i.e. ”FeO”→Fe reaction, is fast and thereby “FeO” phase is not recorded by
XRD.
an
The two-step route of Fe2O3 reduction in 83.6%H2+Ar mixture (see Fig.5) is easier
to understand, if we compare quotient for these tests with the equilibrium constants of the
M
reduction (see Fig.11). The comparison demonstrates that the application of rich H2+Ar
mixture strongly favours Fe3O4 reduction. It is particularly significant that the low-
temperature side of the peak coincides with the KB1 line, which implies that
ed
Fe3O4→”FeO” reaction occurs nearly in the equilibrium state and the next reaction, i.e.
”FeO”→Fe, is very fast.
pt
Fe3O4→”FeO” reaction should occur at low temperature of about 400 ºC, i.e. the
temperature close to Fe2O3→Fe3O4 reduction temperature. This explains the one-step
Ac
mechanism of the Fe2O3 reaction, which was demonstrated by TPR test (profile 1 in Fig.2)
and XRD study (Fig.8).
Conclusions
The temperature-programmed reduction of Fe2O3 with hydrogen depends strongly
on a number of experimental parameters, which affect X H 2O X H 2 ratio in the gas phase, the
15
Page 16 of 18
Qualitatively, temperature-programmed reduction of Fe2O3 with hydrogen proceeds
in line with thermodynamic suggestions, i.e. it is three-step reaction at large
X H 2O X H 2 ratio, and two-step reaction at low X H 2O X H 2 ratio.
t
0.35 and at temperature below 571 ºC.
ip
Additionally, it was evidenced that at extremely low X H 2O X H 2 ratio the TR
cr
reduction of Fe2O3 proceeds as a one-step reaction, Fe 2 O 3 Fe .
us
an
M
ed
pt
ce
Ac
16
Page 17 of 18
References
1. J.H.M. Kock, H.M. Fortuin, J.W. Geus, J. Catal. 96 (1985) 261-275.
2. G.J. Wimmers, P. Arnoldy, J.A. Moulijn, J. Phys. Chem. 90 (1986) 1331-1337.
3. G. Munteanu, L. Ilieva, D. Andreeva, Termochim. Acta 291 (1997) 171-177.
4. H.Y. Lin, Y.W. Chen, Ch.P. Li, Termochim. Acta 400 (2003) 61-67.
t
5. A. Venugopal, J. Aluha, D. Mogano, M.S. Scurrel, Appl. Catal. A: Gen. 245 (2003)
ip
149-158.
cr
6. A. Venugopal, M.S. Scurrel, Appl. Catal. A: Gen. 258 (2004) 241-249.
7. W.K. Jóżwiak, E. Kaczmarek. T.P. Maniecki, W. Ignaczak, W. Maniukiewicz,
us
Appl. Catal. A: Gen. 326 (2007) 17-27.
8. D.A.M. Monti, A. Baiker, J. Catal. 83 (1983) 323-335.
9. J. Zieliński, Catal. Lett. 12 (1992) 389-394.
10. J. Zieliński, Catal. Lett. 31 (1995) 47-56.
an
11. J. Zieliński, J. Chem. Soc. Faraday Trans. 93 (1997) 3577-3580.
M
12. J. Zieliński, J. Catal. 76 (1982) 157-163.
13. J. Zieliński, React. Kinet. Catal. Lett. 17 (1981) 69-75.
14. J. Zieliński, A. Borodziński, Appl. Catal. 13 (1985) 305-310.
ed
Verlag, 1973.
17. Powder Diffraction, Joint Committee on Powder Diffraction Standards, ICDD,
ce
1999: Fe2O3 – Card 86-0550, Fe3O4 – Card 85-1436, “FeO” – Card 85-0625, Fe –
Card 851410.
Ac
17
Page 18 of 18