Geochimica et Cosmochimica Acta, Vol. 68, No. 10, pp. 2197–2205, 2004
Copyright © 2004 Elsevier Ltd
Printed in the USA. All rights reserved
0016-7037/04 $30.00 ⫹ .00
Pergamon
doi:10.1016/j.gca.2003.12.001
Thermochemistry of jarosite-alunite and natrojarosite-natroalunite solid solutions
CHRISTOPHE DROUET,1 KATRINA L. PASS,1 DIRK BARON,2 SARA DRAUCKER,2 and ALEXANDRA NAVROTSKY1,*
1
2
Thermochemistry Facility and NEAT ORU, University of California at Davis, Davis, California 95616, USA
Department of Physics and Geology, California State University at Bakersfield, Bakersfield, California 93311, USA
(Received July 10, 2003; accepted in revised form December 2, 2003)
Abstract—The thermochemistry of jarosite-alunite and natrojarosite-natroalunite solid solutions was investigated. Members of these series were either coprecipitated or synthesized hydrothermally and were characterized by XRD, FTIR, electron microprobe analysis, ICP-MS, and thermal analysis. Partial alkali substitution
and vacancies on the Fe/Al sites were observed in all cases, and the solids studied can be described by the
general formula K1-x-yNay(H3O)xFezAlw(SO4)2(OH)6 –3(3-z-w)(H2O)3(3-z-w). A strong preferential incorporation of Fe over Al in the jarosite/alunite structure was observed. Heats of formation from the elements, ⌬H°f,
were determined by high-temperature oxide melt solution calorimetry. The solid solutions deviate slightly
from thermodynamic ideality by exhibiting positive enthalpies of mixing in the range 0 to ⫹11 kJ/mol. The
heats of formation of the end members of both solid solutions were derived. The values ⌬H°f ⫽ ⫺3773.6 ⫾
9.4 kJ/mol, ⌬H°f ⫽ ⫺4912.2 ⫾ 24.2 kJ/mol, ⌬H°f ⫽ ⫺3734.6 ⫾ 9.7 kJ/mol and ⌬H°f ⫽ ⫺4979.7 ⫾ 7.5
kJ/mol were found for K0.85(H3O)0.15Fe2.5(SO4)2(OH)4.5(H2O)1.5, K0.85(H3O)0.15Al2.5(SO4)2(OH)4.5(H2O)1.5,
Na0.7(H3O)0.3Fe2.7(SO4)2(OH)5.1(H2O)0.9, and Na0.7(H3O)0.3Al2.7(SO4)2(OH)5.1(H2O)0.9 respectively. To our
knowledge, this is the first experimentally-based report of ⌬H°f for such nonstoichiometric alunite and natroalunite
samples. These thermodynamic data should prove helpful to study, under given conditions, the partitioning of Fe
and Al between the solids and aqueous solution. Copyright © 2004 Elsevier Ltd
have been described (Brophy et al., 1962; Härtig et al., 1984;
Alpers et al., 1992; Long et al., 1992; Stoffregen et al., 2000).
Brophy et al. (1962) and Härtig et al. (1984) studied mixing in
synthetic solids.
Jarosites and alunites crystallize in the R3̄m space group that
can be described by a hexagonal unit cell (Hendricks, 1937;
Menchetti and Sabelli, 1976). Deficiencies in iron and aluminum have been reported (Hendricks, 1937; Parker, 1954; Ripmeester et al., 1986; Drouet and Navrotsky, 2003). Charge
neutrality for such nonstoichiometric samples is conserved by
protonation of hydroxyl groups in the structure, leading to
structural water molecules (Ripmeester et al., 1986). Recently,
Drouet and Navrotsky (2003) have reported, in the case of
K-Na-H3O jarosites, that although such “excess water” was
part of the structure, it was only weakly bound. Partial substitution of hydronium for potassium or sodium in A-sites is often
observed for both synthetic and natural samples (Kubisz, 1970;
Dutrizac and Kaiman, 1976; Ripmeester et al., 1986).
A limited set of data is available on the thermochemistry of
stoichiometric jarosites and alunites (see Stoffregen et al.
(2000) for review), and very few on nonstoichiometric samples
or their solid solutions (Bohmhammel et al., 1987). Yet, thermodynamic data for these phases are required to draw accurate
phase diagrams in the Fe-Al-S-O-H system. Such data are also
needed to constrain the conditions under which natural samples
formed. Other applications in the geochemistry field consist of
estimating aqueous concentrations from the chemical compositions of the solids present and vice versa.
In this work, we focused on the determination of the enthalpies of formation of jarosite-alunite and natrojarosite-natroalunite solid solutions. This study was performed by solution
calorimetric experiments carried out in a molten oxide solvent
at 700°C, on highly characterized synthetic samples.
1. INTRODUCTION
Jarosites and alunites are members of the alunite supergroup
of minerals and can be described by the ideal formula
AB3(SO4)2(OH)6, with B-sites being occupied by Fe3⫹ and
Al3⫹. Although numerous cations can occupy the A-site, potassium and sodium are incontestably the most commonly
encountered in natural solids of this group as well as in synthetic phases (Dutrizac and Jambor, 2000).
Natural specimens have been found in many places on Earth,
some dating from several million years (Hofstra et al., 1999;
Stoffregen et al., 2000). These recent to ancient occurrences
raise interest in the archeology and paleoclimatology fields. As
stated by Stoffregen et al. (2000), “the multiple crystallographic sites available for measurement of stable isotopes in
alunite and jarosite make weathering-related samples desirable
candidates for paleoenvironmental studies.” These materials
can indeed lead to important information about the geochronology and climate of the area in which they are found.
Jarosite is common in acid rock drainage environments and
results from the weathering of sulfide ore deposits. The formation of hydrothermal jarosite phases (Dutrizac and Jambor,
2000; Stoffregen et al., 2000), including the so-called “sour gas
jarosites” (Lueth et al., 2000; Lueth and Rye, 2003; Lueth et al.,
2003), is also documented in the literature. Alunite occurs in
weathering, lacustrine, and hydrothermal environments as an
alteration product of Al-containing minerals such as feldspars
(Stoffregen and Alpers, 1992; Stoffregen et al., 2000). Natural
solids with intermediate compositions between jarosite and
alunite, as well as occurrences of these two minerals together,
* Author to whom correspondence should be addressed (anavrotsky@
ucdavis.edu).
2197
2198
C. Drouet et al.
Table 1. Composition of starting synthesis solutions.
Synthesis solution composition
Sample
ID
Vol. (mL)
K2SO4 (g)
Na2SO4 (g)
Fe2(SO4)3 (g)
Al2(SO4)3 18H2O (g)
10.5
6.2
3.6
3.1
1.3
40.5
45.7
49.0
49.5
51.5
K series: jarosite-alunite solid solution
II-2*
100
4.4
0
II-10
100
4.4
0
II-7
100
4.4
0
II-4
100
4.4
0
II-5
100
4.4
0
Na series: natrojarosite-natroalunite solid solution
C-1
100
0
3.1
C-2
100
0
5.7
C-3
100
0
5.0
H-1
2
0
0.41
H-2
2
0
0.44
H-3
2
0
0.60
Alunite-natroalunite solid solution
T-1
90
0.84
0.60
T-2
90
0.24
1.20
T-3
90
0.14
1.35
Synthesis method
Brophy
Brophy
Brophy
Brophy
Brophy
et
et
et
et
et
al.
al.
al.
al.
al.
(1962)
(1962)
(1962)
(1962)
(1962)
3.6
8.4
4.8
0.15
0.07
0
2.0
3.1
6.2
0.31
0.58
0.60
Coprecipitation
Coprecipitation
Coprecipitation
Hydrothermal
Hydrothermal
Hydrothermal
0
0
0
4.8
4.8
5.4
Stoffregen and Cygan (1990)
Stoffregen and Cygan (1990)
Stoffregen and Cygan (1990)
* Same sample names as those given by Brophy et al. (1962).
2. EXPERIMENTAL
2.1. Synthesis
2.1.1. Jarosite-alunite solid solution
Five compositions in the solid solution between jarosite,
KFe3(SO4)2(OH)6, and alunite, KAl3(SO4)2(OH)6, were synthesized
following a procedure described by Brophy et al. (1962). Varying
amounts of K2SO4, Fe2(SO4)3 and Al2(SO4)3 䡠 18H2O were placed in
100 mL of 0.2 N H2SO4 and kept at 100°C and 1 atm for three days
using refluxing condensers. The compositions of the starting solutions
are summarized in Table 1. After three days, the resulting products
were separated by filtering, washed thoroughly with ultrapure water (18
Mohm.cm), and dried at 110°C for 24 h.
2.1.2. Natrojarosite-natroalunite solid solution
Six members of the solid solution between natrojarosite,
NaFe3(SO4)2(OH)6, and natroalunite, NaAl3(SO4)2(OH)6, were synthesized from varying weights of reagent grade Na2SO4, Fe2(SO4)3 and
Al2(SO4)3 䡠 18H2O using two synthesis methods (Table 1). The first
synthesis route consisted of a coprecipitation of the above reactants at
90°C for 4 h, with sulfuric acid 0.1 N. The precipitate was then filtered,
washed with deionized water, and dried at 130°C for 4 h. In the second
method, the samples were obtained by hydrothermal synthesis at
200°C, in a sealed silica glass vial. After 2 d, the furnace was allowed
to cool down, and the precipitate was then filtered, washed and dried as
above. Note that sodium sulfate was added in excess due to difficulty
encountered in obtaining high-sodium contents.
2.2. Characterization
Powder X-ray diffraction patterns were recorded on a Rigaku BiPlane diffractometer for the jarosite-alunite and the alunite-natroalunite
series and on a Scintag PAD-V diffractometer for the natrojarositenatroalunite series. In all cases, Cu K␣ radiation was used (␭ ⫽
1.54056 Å). Before each experiment, the diffractometer was calibrated
with quartz as a standard.
The chemical composition of the jarosite-alunite and alunite-natroalunite samples was determined by ICP-MS. Small amounts of the precipitates were digested in 20% trace-metal grade HNO3 and analyzed for K,
Na, Al, Fe, and S using a Perkin Elmer Elan 6100. The instrument was
calibrated using NIST-traceable standards. The uncertainty associated with
such measurements is 5%. The composition of the natrojarosite-natroalunite samples was drawn from electron microprobe analysis. The measurements were carried out with a Cameca SX-50 apparatus. Each sample
was pressed into a 5 mg pellet and immobilized with epoxy resin. The
microprobe results were used to determine the potassium, sodium, iron,
aluminum, and sulfur contents, assuming that each unit formula contains
two sulfate groups. The standard deviation on each determination (not
reported here for the sake of clarity) was close to 5%. For all experiments,
total weight percents of the analyzed elements are in good agreement with
expected totals.
Thermal analyses (TG/DSC) were performed on a Netzsch STA 449
C, with argon as the carrier gas. The heating rate was 10°C/min, and the
gas phase was constantly monitored by Fourier transform infrared
spectroscopy using a Bruker EQUINOX 55 spectrometer (MCT detector). Transmission infrared spectra were recorded on the same spectrometer, but using a DTGS detector. Pellets of 13 mm diameter (150
mg) were pressed under 300 bar for 1 min. The spectrometer was
flushed with nitrogen to avoid contamination. Spectra were collected in
the 400 – 4000 cm⫺1 range with a resolution of 4 cm⫺1. A baseline
correction was made before interpretation.
2.1.3. Alunite-natroalunite solid solution
2.3. High-temperature Calorimetry
To cross-check the results obtained for the jarosite-alunite and natrojarosite-natroalunite solid solutions, we prepared three solids from the
alunite-natroalunite solid solution, with varying Na/K ratios. These
compounds were synthesized according to the procedure used by
Stoffregen and Cygan (1990) and the Na/K ratios that we selected were
similar to those used by Parker (1962). The starting solutions (see Table
1) were placed in a sealed vessel and heated to 150°C for 48 h. The
precipitates were filtered, washed with ultrapure water, and dried for
24 h at 110°C. The resulting solids were then heated to 300°C for 1 h.
High temperature drop solution calorimetry was carried out at 700°C
in a custom-built Tian-Calvet twin calorimeter. The technique was
described in detail elsewhere (Navrotsky, 1977 and 1997). In this work,
sodium molybdate, 3Na2O 䡠 4MoO3, was used as solvent.
The pelletized samples (5–15 mg) were dropped into a platinum
crucible containing the solvent, located in the hot zone of the calorimeter (700°C). The end of the reaction was judged by the return of the
baseline to its initial value. The final state in these calorimetric experiments is a dilute solution in 3Na2O 䡠 4MoO3 of the sulfates of Na, K,
Jarosite-alunite thermochemistry
Fig. 1. XRD patterns for samples C-3 and H-3, PDF standards for
natrojarosite and natroalunite.
Al and Fe3⫹ (as ions in the melt) and gaseous H2O, all at 700°C.
Indeed, preliminary studies on simple sulfates and jarosite phases
(Majzlan et al., 2002; Drouet and Navrotsky, 2003) have shown that all
the sulfur was retained in the melt rather than emitted in the gas phase.
During the experiments, oxygen is flushed through the gas space
above the melt (⬃35 mL/min) and bubbled through the solvent (⬃5
mL/min) to maintain oxidizing conditions, stir the melt, and remove the
evolved water.
Throughout the text, the notation ⌬H°f will represent the enthalpy of
formation from the elements of a given species. For heats of formation
determined from the oxides/sulfates, the corresponding notation will be
⌬H°f,ox.
3. RESULTS AND DISCUSSION
3.1. Characterization
The X-ray diffraction patterns obtained with the samples
precipitated in this work were compared to the PDF (powder
2199
diffraction file) standard files 22– 0827 for jarosite, 71–1776 for
alunite, 36 – 0425 for natrojarosite, and 41–1467 for natroalunite. Figure 1 is related to samples C-2 and H-3, which are
close in composition to natrojarosite and natroalunite respectively. In all cases, the Fe-rich samples could easily be identified as jarosite group minerals, and Al-rich samples as alunite
group minerals. Evolution of the XRD patterns from jarositelike to alunite-like was observed as the aluminum content
increased, as expected for solid solutions.
The unit cell parameters were determined for all samples
(Table 2). For both the jarosite-alunite and natrojarositenatroalunite solid solutions, the unit cell parameter “a” was
found to increase monotonically with the iron content. In contrast, the parameter “c” remained almost constant for these
series, close to 17.1 Å for the potassium-containing samples
and 16.6 Å for the Na samples, regardless of the Fe/Al ratio.
These results are consistent with those reported by Stoffregen
et al. (2000) for Fe-Al substitution in jarosite-like species.
The unit cell parameters obtained for the samples from the
alunite-natroalunite solid solution (Table 2) show, on the contrary, constant “a” values, close to 6.98 Å, but increasing “c”
values in the range 16.7–17.2 Å as the sodium content increases. This is in accord with reported data on alunites and
natroalunites (Stoffregen et al., 2000), and the “a” and “c”
variations observed here also agree with earlier results on
A-site potassium-sodium substitution in jarosites (Drouet and
Navrotsky, 2003).
The chemical compositions of the solids synthesized were
determined from ICP-MS and/or electron microprobe analyses
(Table 2). The K ⫹ Na content of all solids was lower than 1
mol/2 mol of sulfate, implying some substitution of hydronium
ions for potassium and sodium. For the sodium-containing
samples, the hydronium content was not a simple function of
the amount of sodium in the reactant mixture. It was affected by
Table 2. Chemical composition and unit cell parameters of the samples prepared in this work.
Chemical compositiona
Sample
ID
K⫹
Na⫹
H3O⫹b
Fe3⫹
K series: jarosite-alunite solid solution
II-2
0.79
0
0.21
1.99
II-10
0.79
0
0.21
1.61
II-7
0.83
0
0.17
1.32
II-4
0.89
0
0.11
1.19
II-5
0.99
0
0.01
0.63
Na series: natrojarosite-natroalunite solid solution
C-1
0
0.68
0.32
2.66
C-2
0
0.69
0.31
2.60
C-3
0
0.70
0.30
2.44
H-1
0
0.66
0.34
1.24
H-2
0
0.68
0.32
0.74
H-3
0
0.66
0.34
0
Alunite-natroalunite solid solution
T-1
0.70
0.08
0.22
0
T-2
0.48
0.29
0.23
0
T-3
0.24
0.59
0.17
0
a
Unit cell parameters
Al3⫹
SO42⫺
OH⫺c
H2Oc
a (Å)
c (Å)
0.44
0.79
1.12
1.41
2.19
2
2
2
2
2
4.29
4.20
4.32
4.80
5.46
1.71
1.80
1.68
1.20
0.54
7.248 ⫾ 0.003
7.209 ⫾ 0.003
7.175 ⫾ 0.004
7.146 ⫾ 0.003
7.064 ⫾ 0.004
17.087 ⫾ 0.006
17.090 ⫾ 0.007
17.108 ⫾ 0.011
17.090 ⫾ 0.008
17.049 ⫾ 0.009
0.07
0.07
0.20
1.46
1.88
2.65
2
2
2
2
2
2
5.17
5.00
4.91
5.09
4.86
4.97
0.83
1.00
1.09
0.91
1.14
1.03
7.327 ⫾ 0.001
7.328 ⫾ 0.001
7.316 ⫾ 0.001
7.137 ⫾ 0.003
7.048 ⫾ 0.002
6.987 ⫾ 0.001
16.654 ⫾ 0.002
16.654 ⫾ 0.002
16.653 ⫾ 0.002
16.631 ⫾ 0.007
16.623 ⫾ 0.005
16.635 ⫾ 0.002
2.69
2.63
2.74
2
2
2
5.08
4.89
5.23
0.92
1.11
0.77
6.988 ⫾ 0.001
6.985 ⫾ 0.036
6.981 ⫾ 0.003
17.123 ⫾ 0.002
16.905 ⫾ 0.088
16.776 ⫾ 0.008
Normalized composition, containing two SO42 ions per unit formula. Standard deviation on each value is close to 5%.
Derived from K⫹ Na⫹ H3O ⫽ 1 in A-site.
c
Derived from Fe ⫽ 3-z and OH ⫽ 6-3z and H2O ⫽ 3z.
b
2200
C. Drouet et al.
Fig. 2. Ternary diagram Fe/Al/B-site vacancies, for samples from the
K and Na series and derived end members. Dashed line corresponds to
Fe⫹Al ⫽ 2.5 and dotted line to Fe⫹Al ⫽ 2.7.
the Fe/Al ratio: the amount of H3O⫹ decreasing for higher iron
contents. In a similar way, the sum Fe ⫹ Al was generally
smaller than 3 mol/2 mol of sulfate, indicating B-site vacancies
and the associated substitution of water molecules for part of
the OH- groups (protonation). The amount of such vacancies
decreased with increasing iron content, especially for samples
synthesized hydrothermally.
We focused our syntheses on the preparation of two series of
Fe-Al compounds. Samples from the first series were members
of the jarosite-alunite solid solution having similar potassium
contents and amounts of B-site vacancies. To a first approximation, the only compositional variable was the iron-to-aluminum ratio. This series will be referred to as the “potassium
series.” The second family of compounds was the parallel to the
first one, but with sodium replacing potassium (i.e., natrojarosite-natroalunite solid solution), therefore leading to a “sodium series.” The general chemical formulas of the potassium
and sodium series can be approximated to K0.85(H3O)0.15
Fe2.5-wAlw(SO4)2(OH)4.5(H2O)1.5 and Na0.7(H3O)0.3Fe2.7-w
Alw(SO4)2(OH)5.1(H2O)0.9, although some variations in the K
series are observed. Figure 2 shows the composition of the
samples synthesized in this work for both series, in the plane
Fe/Al/B-sites vacancies.
The composition of the intermediate solids from the potassium series (jarosite-alunite solid solution) shows that Fe is
incorporated preferentially over Al into the crystal structure.
This is apparent from comparison of the Al/Fe ratios in the
starting solutions and the resulting solids. For example, the
synthesis solution for sample II-4 has an Al/Fe mole ratio of
⬃10:1, but leads to a precipitate with an Al/Fe ratio close to
1.2:1 (Tables 1 and 2). Such preferential incorporation of Fe
has been reported by Brophy et al. (1962) for similar compositions.
The Fe-rich samples from the sodium series were synthesized by coprecipitation. We experienced some difficulty in the
preparation of Al-rich samples with this technique, even when
using elevated Al/Fe ratios in the starting solutions. We succeeded in preparing Al-rich samples in the natrojarosite-
natroalunite series by using a hydrothermal synthesis route. As
a general observation, Fe was incorporated preferentially to Al
in B-sites, as for the potassium series. Hydrothermal synthesis
enabled us to synthesize Al-rich samples with Al/Fe ratios
close to the target compositions. These findings are in agreement with the observations made by Brophy et al. (1962)
showing that increasing the temperature and the pressure and
lowering the acidity of the starting solutions may lead to higher
Al/Fe ratios.
The solids were analyzed by transmission FTIR spectroscopy. Infrared spectra for jarosite species have been reported
and discussed elsewhere (Powers et al., 1975; Baron and
Palmer, 1996; Drouet and Navrotsky, 2003). In particular, the
FTIR spectra obtained here for Fe-rich samples were similar to
those obtained with K-Na-H3O jarosites in the literature. In the
case of Al-rich samples, however, slight differences were observed, especially below 600 cm⫺1. Figure 3 gives the FTIR
spectrum obtained with our sample H-3, that is close to natroalunite, as compared to a typical spectrum obtained for the natrojarosite sample Na0.71(H3O)0.29Fe2.63(SO4)2(OH)4.89(H2O)1.11
during a previous study (Drouet and Navrotsky, 2003).
The number of absorption bands is the same for both spectra:
only the position and, to some extent, the intensity of some of
these bands differ. This is indeed often the case for solid
solutions, where the progression from one structure to the next
is related to ion substitutions in equivalent crystallographic
sites. It was, therefore, possible to assign the absorption bands
observed here by comparison with the natrojarosite sample
(Fig. 3). The differences observed between the two spectra in
the region 400 – 600 cm⫺1 are most likely due to changes in
octahedra vibrations between FeO6 and AlO6. Indeed, absorption bands assigned to FeO6 vibrations were reported at 477,
513, and 575 cm-1 for jarosite (Baron and Palmer, 1996), and
similar bands have been observed for the jarosite-natrojarosite
solid solution (Drouet and Navrotsky, 2003). In the present
case, three bands are seen below 600 cm⫺1 for Al-rich samples
(Fig. 3), at ca. 490, 515, and 592 cm⫺1. Taking into account the
above, these bands are most likely assignable to AlO6 octahedra vibrations. Although the band at 592 cm⫺1 is surprisingly
intense, the positions of these three bands (490, 515, and 592
cm⫺1) are close to those (486, 528, and 595 cm⫺1) reported by
Powers et al. (1975) for alunite.
In the region above 3000 cm⫺1, the broad band corresponding to OH stretching (3456 cm⫺1) appears at higher frequencies
Fig. 3. FTIR spectra of sample H-3 (this work) and a natrojarositelike sample from Drouet and Navrotsky (2003).
Jarosite-alunite thermochemistry
2201
deed been confirmed by Menchetti and Sabelli (1976), in the
case of jarosite/alunite: these authors showed that Al-OH bond
lengths in alunites were close to 1.879 Å as compared to 1.975
for Fe-OH in jarosites.
It is, moreover, possible to estimate, in a semiquantitative
way, the enthalpy of decomposition of the (dehydrated) natrojarosite/natroalunite samples from integration of peak 1, after
calibration of the DSC. The corresponding values of ⌬Hdecomp
range between 193 and 235 kJ/mol (Table 3). ⌬Hdecomp becomes more endothermic as the Al content increases, confirming that the decomposition of the jarosite/alunite network requires more energy as the samples become richer in Al. Thus,
the higher decomposition temperature for Al-rich samples has
a thermodynamic basis.
3.2. Calorimetry
Fig. 4. TG/DSC curves for sample C-2 and evolution of DSC peak
maxima with Al content.
than for K-Na jarosites (ca. 3360 –3390 cm⫺1, Drouet and
Navrotsky, 2003). This is in accord with literature data for a
K-Na alunite sample (Okada et al., 1982) as well as with
previous results comparing alunite and jarosite spectral features
(Powers et al., 1975).
Thermal analyses of the samples showed the presence of
“excess water,” indicated by a peak close to 250°C (Fig. 4).
Also, the two intense endothermic peaks that are typical during
jarosite thermal decomposition were also observed here for
both the K series and the Na series. Considering earlier works
on the thermal decomposition of jarosites (Kubisz, 1971;
Drouet and Navrotsky, 2003; and references therein) and alunites (Pysiak and Glinka, 1981; Bohmhammel et al., 1987;
Stoffregen and Alpers, 1992), one can assign the first of these
intense peaks (peak 1) in the range 400 – 600°C to dehydroxylation, leading among other phases to yavapaiite-like compounds (AB(SO4)2, A ⫽ K, Na, B ⫽ Fe, Al). The second
intense peak (peak 2), at 700 – 820°C, can be attributed to the
thermal decomposition of such yavapaiite-type structures.
The temperature dependence of these DSC peaks was followed as a function of the Fe/Al ratio in the sodium series,
which is less documented than the potassium series in the
literature. As shown in Table 3, the temperature of the maxima
of these peaks increases as the samples become richer in
aluminum, and differences of up to 113°C are seen between
Al-rich and Fe-rich samples. This is in agreement with earlier
results (Brophy et al., 1962), showing that the dehydroxylation
of alunites requires higher temperatures than the dehydroxylation of jarosites. Note that the same general trend is observed
for Fe and Al oxyhydroxides (Wolska et al., 1992; Kloprogge
et al., 2002; Mitov et al., 2002): the higher the Al content, the
higher the decomposition temperatures. According to Brophy et
al. (1962), these observations may be indicative of stronger
bonds formed between Al and hydroxyl groups. This has in-
The enthalpies of formation from the elements (⌬H°f) of the
samples prepared in this work were obtained from high-temperature oxide melt calorimetric experiments as was done earlier for K-Na-H3O jarosites (Drouet and Navrotsky, 2003). The
values of ⌬H°f were determined from experimental drop solution enthalpies (⌬Hds) by application of the thermodynamic
cycle described in Table 4.
In this cycle, x, y, z, and w respectively stand for the
hydronium, sodium, iron, and aluminum contents in the solids,
the potassium content being 1-x-y. As was mentioned above,
the sum z ⫹ w was generally lower than 3, leading to modifications in the OH sites as compared to the theoretical composition AB3(SO4)2(OH)6 where B represents Fe or Al. In this
regard, the general chemical formula of these compounds is
K1-x-yNay(H3O)xFezAlw(SO4)2(OH)6-3(3-z-w)(H2O)3(3-z-w).
In addition to the enthalpies of drop solution of these solids,
the thermodynamic cycle given in Table 3 involves the phases
K2SO4, Na2SO4, ␣-Fe2O3 (hematite), ␣-Al2O3 (corundum),
SO3(g), and H2O(l). Thermodynamic data for these compounds
were taken from the literature (Robie and Hemingway, 1995;
Majzlan et al., 2002; Drouet and Navrotsky, 2003), and they are
given in Table 5. Numerical values for the enthalpies of drop
solution ⌬Hds measured in this work, as well as the derived
enthalpies of formation ⌬H°f, have also been included in this
table. Note that each of these ⌬H°f values have been determined using the actual x, y, z, and w values corresponding to
each sample. Therefore, the uncertainties in ⌬H°f take into
account the variations in chemical composition as well as the
uncertainties on the experimental determination of ⌬Hds.
Table 3. DSC peak maxima for the natrojarosite-natroalunite solid
solution, and enthalpy of decomposition.
Tmax (°C)
Sample
Fe/Al
Peak 1
Peak 2
⌬Hdecomp (kJ/mol)a
C-1
C-2
C-3
H-1
H-2
H-3
38.0
37.1
12.2
0.9
0.4
0
435.6
438.1
444.1
494.4
507.7
548.3
730.3
729.2
731.2
750.8
769.2
812.7
202
193
225
224
235
218
a
Estimated from area of DSC peak 1.
2202
C. Drouet et al.
Table 4. Thermodynamic cycle used for the determination of the enthalpy of formation of K-Na-H3O alunite-jarosites.
Reactiona
⌬H (kJ/mol)b
(1) K1ⴚxⴚyNay(H3O)xFezAlw(SO4)2(OH)6ⴚ3(3ⴚzⴚw)(H2O)3(3ⴚzⴚw) (s, 298) 3 (1⫺x⫺y)/2 K2O (soln, 973) ⫹ y/2 Na2O
(soln, 973) ⫹ z/2 Fe2O3 (soln, 973) ⫹ w/2 Al2O3 (soln, 973) ⫹ 2 SO3 (soln, 973) ⫹(15⫹3x⫺3z⫺3w)/2 H2O (g, 973)
(2) K2SO4 (s, 298) 3 K2O (soln, 973) ⫹ SO3 (soln, 973)
(3) Na2SO4 (s, 298) 3 Na2O (soln, 973) ⫹ SO3 (soln, 973)
(4) 2 K (s, 298) ⫹ S (s, 298) ⫹ 2 O2 (g, 298) 3 K2SO4 (s, 298)
(5) 2 Na (s, 298) ⫹ S (s, 298) ⫹ 2 O2 (g, 298) 3 Na2SO4 (s, 298)
(6) ␣-Fe2O3 (s, 298) 3 Fe2O3 (soln, 973)
(7) 2 Fe (s, 298) ⫹ 3/2 O2 (g, 298) 3 ␣-Fe2O3 (s, 298)
(8) ␣-Al2O3 (s, 298) 3 Al2O3 (soln, 973)
(9) 2 Al (s, 298) ⫹ 3/2 O2 (g, 298) 3 ␣-Al2O3 (s, 298)
(10) SO3 (g, 298) 3 SO3 (soln, 973)
(11) S (s, 298) ⫹ 3/2 O2 (g, 298) 3 SO3 (g, 298)
(12) H2O (1, 298) 3 H2O (g, 973)
(13) H2 (g, 298) ⫹ 1/2 O2 (g, 298) 3 H2O (1, 298)
Formation of a (K, Na)-jarosite-alunite:
(12) (1⫺x⫺y) K (s, 298) ⫹ y Na (s, 298) ⫹ z Fe (s, 298) ⫹ w Al (s, 298) ⫹ 2 S (s, 298) ⫹ (14⫹x)/2 O2 (g, 298) ⫹
(15⫹3x⫺3z⫺3w)/2 H2 (g, 298) 3 K1ⴚxⴚy Nay (H3O)x FezAlw (SO4)2 (OH)6ⴚ3(3ⴚzⴚw) (H2O)3(3ⴚzⴚw) (s, 298)
⌬Hfo (sample) ⫽ ⫺⌬H1 ⫹ (1⫺x⫺y)/2 ⌬H2 ⫹ y/2 ⌬H3 ⫹ (1⫺x⫺y)/2 ⌬H4 ⫹ y/2 ⌬H5 ⫹ z/2 ⌬H6 ⫹ z/2 ⌬H7 ⫹ w/2 ⌬H8
⫹ w/2 ⌬H9 ⫹ (3⫹x)/2 ⌬H10 ⫹ (3⫹x)/2 ⌬H11 ⫹ (15⫹3x⫺3z⫺3w)/2 ⌬H12 ⫹ (15⫹3x⫺3z⫺3w)/2 ⌬H13
⌬Hds (sample)
⌬Hds (K2SO4)
⌬Hds (Na2SO4)
⌬H°f (K2SO4)
⌬H°f (Na2SO4)
⌬Hds (␣-Fe2O3)
⌬H°f (␣-Fe2O3)
⌬Hds (␣-Al2O3)
⌬H°f (␣-Al2O3)
⌬Hds (SO3(g))c
⌬H°f (SO3(g))
⌬Hds (H2O(1))
⌬H°f (H2O(g))
⌬H°f (sample)
s ⫽ solid; g ⫽ gas; soln ⫽ in solution (in sodium molybdate).
⌬Hds, ⌬H°f and ⌬Hhc are respectively the drop solution enthalpy, standard enthalpy of formation, and heat content.
c
⌬Hds (SO3(g)) as determined by Majzlan et al. (2002).
a
b
As the ratio Fe/Al decreases, ⌬H°f varies from ⫺3963 to
⫺4778 kJ/mol for the potassium series, and from ⫺3760 to
⫺4980 kJ/mol for the sodium series. In other words, the enthalpies of formation from the elements become less exothermic as samples become richer in iron. However, because the
heats of formation (from the elements) of corundum and hematite themselves differ by a factor two (Table 5), it is interesting to calculate the enthalpies of formation from the oxides/
sulfates, ⌬H°f,ox. For the K series, ⌬H°f,ox is defined here as the
enthalpy accompanying the reaction (at 298 K):
Table 5. Thermodynamic data for the solid phases used in this work (uncertainties calculated from thermodynamic cycle using two standard
deviations of the mean).
Compound
⌬H°f (kJ/mol)d
⌬Hds (kJ/mol)
Na2SO4
155.7 ⫾ 2.3 (6)
⫺1387.8 ⫾ 0.4
K2SO4
153.4 ⫾ 1.8 (8)a
⫺1437.7 ⫾ 0.5b
a
␣-Fe2O3
95.0 ⫾ 1.8 (8)
⫺826.2 ⫾ 1.3b
␣-Al2O3
105.0 ⫾ 1.2 (8)f
⫺1675.7 ⫾ 1.3b
SO3(g)
⫺205.8 ⫾ 3.7c
⫺395.7 ⫾ 0.7b
H2O(1)
⫺285.8 ⫾ 0.0b
Jarosite-alunite solid solution: K0.85 (H3O)0.15 Fe2.5-w Alw (SO4)2 (OH)4.5 (H2O)1.5
II-2
513.0 ⫾ 8.0 (6)
⫺3963.4 ⫾ 10.4
II-10
521.9 ⫾ 5.3 (6)
⫺4118.0 ⫾ 8.4
II-7
533.2 ⫾ 6.0 (6)
⫺4270.1 ⫾ 8.7
II-4
544.3 ⫾ 6.8 (6)
⫺4410.3 ⫾ 9.2
II-5
574.5 ⫾ 4.5 (6)
⫺4778.4 ⫾ 7.6
Natrojarosite-natroalunite solid solution: Na0.7 (H3O)0.3 Fe2.7-w Alw (SO4)2 (OH)5.1 (H2O)0.9
C-1
472.5 ⫾ 5.3 (6)
⫺3760.2 ⫾ 8.7
C-2
491.7 ⫾ 2.1 (7)
⫺3776.8 ⫾ 7.2
C-3
488.4 ⫾ 2.9 (6)
⫺3826.8 ⫾ 7.4
H-1
547.4 ⫾ 9.6 (3)
⫺4417.5 ⫾ 11.7
H-2
563.3 ⫾ 4.0 (6)
⫺4606.3 ⫾ 7.7
H-3
612.1 ⫾ 3.3 (9)
⫺4979.7 ⫾ 7.5
Alunite-natroalunite solid solution: K0.8-x Nax (H3O)0.2 Al2.7 (SO4)2 (OH)5.1 (H2O)0.9
T-1
593.4 ⫾ 4.4 (6)
⫺4996.4 ⫾ 7.9
T-2
604.9 ⫾ 4.1 (10)
⫺4974.7 ⫾ 7.7
T-3
609.5 ⫾ 4.0 (6)
⫺5023.0 ⫾ 7.6
a
a
⌬H°f,ox (kJ/mol)e
⫺377.2 ⫾ 10.1
⫺384.4 ⫾ 8.1
⫺394.0 ⫾ 8.5
⫺403.7 ⫾ 9.1
⫺430.0 ⫾ 7.5
6.8 ⫾ 8.6
10.4 ⫾ 5.3
10.9 ⫾ 6.0
10.1 ⫾ 6.8
7.7 ⫾ 4.5
⫺361.2 ⫾ 8.5
⫺380.4 ⫾ 6.9
⫺376.4 ⫾ 7.2
⫺429.1 ⫾ 11.5
⫺442.9 ⫾ 7.5
⫺487.9 ⫾ 7.2
8.8 ⫾ 7.2
⫺10.4 ⫾ 2.1
⫺0.4 ⫾ 2.9
5.3 ⫾ 9.6
11.0 ⫾ 4.0
1.7 ⫾ 3.3
Drouet and Navrotsky (2003). In parentheses are the numbers of experiments performed.
Robie and Hemingway (1995).
c
Majzlan et al. (2002).
d
Enthalpy of formation from the elements. Uncertainties are two standard deviations of the mean (experimental).
e
Enthalpy of formation from the oxides/sulfates.
f
This work.
b
⌬Hmix (kJ/mol)
b
Jarosite-alunite thermochemistry
2203
By definition, the enthalpy of mixing for an intermediate composition KFe3-wAlw(SO4)2(OH)6 between the end-members
jarosite, KFe3(SO4)2(OH)6, and alunite, KAl3(SO4)2(OH)6, is
given by the equation:
⌬Hmix(Fe3-wAlw) ⫽ ⌬H°f(Fe3-wAlw)
⫺ (3-w)/3 . ⌬H°f (jarosite) ⫺ w/3 . ⌬H°f (alunite)
(5)
or, equivalently:
⌬Hmix(Fe3-wAlw) ⫽ (3-w)/3 . ⌬Hds (jarosite)
⫹ w/3 . ⌬Hds (alunite) ⫺ ⌬Hds(Fe3-wAlw)
Fig. 5. ⌬Hds and ⌬H°f for the potassium series (jarosite-alunite solid
solution) and for end members (this work, Drouet and Navrotsky,
2003).
0.425 K2SO4(s) ⫹ (2.5-w)/2 Fe2O3(s)
⫹ w/2 Al2O3(s) ⫹ 3.975 H2O(l) ⫹ 1.575 SO3(g) 3
K0.85(H3O)0.15Fe2.5-wAlw(SO4)2(OH)4.5(H2O)1.5(s)
(1)
Likewise, for the Na series, the reaction is
0.35 Na2SO4(s) ⫹ (2.7-w)/2 Fe2O3(s)
⫹ w/2 Al2O3(s) ⫹ 3.9 H2O(l) ⫹ 1.65 SO3(g) 3
Na0.7(H3O)0.3Fe2.7-wAlw(SO4)2(OH)5.1(H2O)0.9(s)
(2)
These enthalpies can be derived from the enthalpy of drop
solution of the samples and the thermodynamic cycle in Table
4 by the equations:
⌬H°f,ox (K series) ⫽ ⫺⌬Hds(sample) ⫹ 0.425 ⌬Hds(K2SO4)
⫹ (2.5-w)/2 ⌬Hds(Fe2O3) ⫹ w/2 ⌬Hds(Al2O3)
⫹ 3.975 ⌬Hds(H2O(l)) ⫹ 1.575 ⌬Hds(SO3(g))
(3)
and
⌬H°f,ox (Na series) ⫽ ⫺⌬Hds(sample) ⫹ 0.35 ⌬Hds(Na2SO4)
(6)
The use of Eqn. 6 is preferred to that of Eqn. 5 because the
magnitude of the heats of formation is often very large; in
addition, the drop solution enthalpies are the experimental
values directly measured here, and thus calculation using Eqn.
6 gives smaller uncertainties.
Application of Eqn. 6 to the potassium and sodium series
studied in this work requires the knowledge of the enthalpies of
drop solution of four end members: jarosite, natrojarosite,
alunite, and natroalunite. However, because our samples show
deviations from the theoretical stoichiometric formulas, as
many natural samples do, we consider here end members with
A- and B-site contents close to those of the samples studied.
Values of the enthalpies of drop solution (in sodium molybdate) of the jarosite K0.72(H3O)0.28Fe2.60(SO4)2(OH)4.80(H2O)1.20
(⌬Hds ⫽ 504.1 ⫾ 3.9 kJ/mol) and the natrojarosite
Na0.68(H3O)0.32Fe2.83(SO4)2(OH)5.49(H2O)0.51 (⌬Hds ⫽ 477.7
⫾ 5.0 kJ/mol) have been reported in a previous study (Drouet
and Navrotsky, 2003). Taking into account their chemical
compositions, these compounds can be chosen as the nonstoichiometric jarosite and natrojarosite end members needed here.
The remaining two end members, namely nonstoichiometric
alunite K0.8(H3O)0.2Al2.7 (SO4)2(OH)5.1(H2O)0.9 and natroalunite Na0.8(H3O)0.2Al2.7(SO4)2(OH)5.1(H2O)0.9, were
extrapolated from the linear variation of ⌬Hds with the
sodium content for the alunite-natroalunite solid solution
K0.8-xNax(H3O)0.2Al2.7(SO4)2(OH)5.1(H2O)0.9 (samples T-1,
T-2, T-3, Fig. 7). The numerical values are ⌬Hds ⫽ 593.3 ⫾ 2.9
kJ/mol for the K-Al end member and ⌬Hds ⫽ 616.4 ⫾ 5.9
kJ/mol for the Na-Al end member.
⫹ (2.7-w)/2 ⌬Hds(Fe2O3) ⫹ w/2 ⌬Hds(Al2O3)
⫹ 3.9 ⌬Hds(H2O(l)) ⫹ 1.65 ⌬Hds(SO3(g))
(4)
Numerical values of ⌬H°f,ox, obtained by application of Eqn. 3
and Eqn. 4, are listed in Table 5. For both series, these values
show that, like ⌬H°f from the elements, the heats of formation
from the oxides/sulfates are more endothermic for Fe-rich
samples. These results indicate that the effect of Al/Fe ratio on
⌬H°f is not solely due to the difference in heats of formation of
hematite and corundum.
Figures 5 and 6 depict the variation of the enthalpies of drop
solution and of formation (from the elements) for the potassium
and sodium series respectively, as a function of the Al content
in B-sites. In both cases, a deviation from linearity is observed
and the curves can best be fitted to second-degree polynomials.
This trend points out that these solid solutions depart from
thermodynamic ideality, and it is indicative of non-zero enthalpies of mixing, ⌬Hmix.
Fig. 6. ⌬Hds and ⌬H°f for the sodium series (natrojarosite-natroalunite solid solution) and for end members (this work, Drouet and
Navrotsky, 2003).
2204
C. Drouet et al.
This linearity of ⌬Hds versus Na content for the alunitenatroalunite solid solution is not surprising as it parallels what
was observed in previous work (Drouet and Navrotsky, 2003)
on the jarosite-natrojarosite solid solution. This suggests that
the enthalpies of mixing for substitutions in A-sites are negligible regardless of the nature of the cation located in B-sites. In
contrast, a recent study (Drouet et al., in press) on the solid
solution between jarosite and its chromate analog,
KFe3(CrO4)2(OH)6, showed non-zero (negative) enthalpies of
mixing for the substitution of chromium (VI) for sulfur.
The above values of ⌬Hds for the four end members were
then used to calculate the enthalpies of mixing for all intermediate compositions in the potassium and the sodium series
using Eqn. 6. Numerical values of ⌬Hmix are given in Table 5.
Despite fairly large uncertainties, mostly originating from experimental uncertainties on the ⌬Hds values, these enthalpies of
mixing are unambiguously different from zero and mostly
range from 0 to ⫹11 kJ/mol. These positive enthalpies of
mixing suggest a tendency toward clustering and exsolution.
It is possible to extrapolate the heats of formation, from the
elements, of the end members of the K and Na series from the
curves ⌬Hds vs. Al content (Table 6) and application of the cycle in Table 4. For the potassium series, the extrapolation gives
⌬H°f ⫽ ⫺3773.6 ⫾ 7.7 kJ/mol for K0.85(H3O)0.15Fe2.5(SO4)2
(OH)4.5(H2O)1.5 and ⌬H°f ⫽ ⫺4912.2 ⫾ 7.1 kJ/mol for
K0.85(H3O)0.15Al2.5(SO4)2(OH)4.5(H2O)1.5 (Fig. 5). For the sodium series, a similar extrapolation gives ⌬H°f ⫽ ⫺3734.6 ⫾
8.6 kJ/mol for Na0.7(H3O)0.3Fe2.7(SO4)2(OH)5.1(H2O)0.9 (Fig.
6). The value ⌬H°f ⫽ ⫺4979.7 ⫾ 7.5 kJ/mol obtained for the
sample H-3 can be taken as the heat of formation of the
end-member Na0.7(H3O)0.3Al2.7(SO4)2(OH)5.1(H2O)0.9 due to
its chemical composition. The values obtained here for the
K-Fe and Na-Fe end members compare well to earlier results
on K-H3O and Na-H3O jarosites (Drouet and Navrotsky, 2003).
However, these extrapolations are based, for the K-Fe, K-Al
and Na-Fe end members, on the assumption that the potassium
and sodium series can be described by the formulas
K0.85(H3O)0.15Fe2.5-wAlw(SO4)2(OH)4.5(H2O)1.5 and Na0.7(H3O)0.3
Fe2.7-wAlw(SO4)2(OH)5.1(H2O)0.9 respectively. As mentioned above,
some variations to these formulas are observed in the samples,
which leads to additional uncertainty for the extrapolation of
Table 6. Equation of curves ⌬Hds ⫽ f (Al mole fraction) for the K
and Na series.
y ⫽ A ⫹ Bx ⫹ Cx2, y ⫽ ⌬Hds and
x ⫽ Al mole fraction
K series:
Na series:
A ⫽ 504.4, B ⫽ 44.8, C ⫽ 78.2
A ⫽ 479.9, B ⫽ 113.9, C ⫽ 39.3
the heats of formation of the corresponding end members. The
parabolic curves ⌬Hds vs. Al mole fraction are described by
second-order polynomials y ⫽ A ⫹ Bx ⫹ Cx2 (Table 6).
Variation in x (chemical composition), noted ⌬x, will therefore
lead to a variation in y, ⌬y. The samples studied differ from the
ideal series formulas by their Fe⫹Al content as well as their
K⫹Na content. One can include these two sources of variation
in chemical composition by calculating ⌬x as ⌬x ⫽ 共⌬A2
⫹ ⌬B2 兲1/ 2 where ⌬A and ⌬B are the differences between the
actual A-site and B-site contents, respectively, and their values
in the series formulas. ⌬y can then be derived from ⌬x by
differentiation of the equation y ⫽ A ⫹ Bx ⫹ Cx2 , giving:
⌬y ⫽ (B ⫹ 2Cx) ⌬x . This leads, for the end members, to
additional uncertainty on ⌬Hds of ⫾2.1 kJ/mol for the K-Fe end
member, ⫾18.5 kJ/mol for K-Al, and ⫾1.4 kJ/mol for Na-Fe.
The corresponding additional uncertainties on the heats of
formation ⌬H°f are ⫾1.7 kJ/mol for K-Fe, ⫾17.1 kJ/mol for
K-Al, and ⫾1.1 kJ/mol for Na-Fe, as determined from the
uncertainties applying in the the thermodynamic cycle of Table
4. Therefore, the final values for the extrapolated heats of
formation of these end members are ⌬H°f ⫽ ⫺3773.6 ⫾ 9.4
kJ/mol for K0.85(H3O)0.15Fe2.5(SO4)2(OH)4.5(H2O)1.5, ⌬H°f
⫽ ⫺4912.2 ⫾ 24.2 kJ/mol for K0.85(H3O)0.15Al2.5
(SO4)2(OH)4.5(H2O)1.5, and ⌬H°f ⫽ ⫺3734.6 ⫾ 9.7 kJ/mol for
Na0.7(H3O)0.3Fe2.7(SO4)2(OH)5.1(H2O)0.9. Note that the higher
uncertainty obtained for the nonstoichiometric alunite end
member (K-Al) is due to the higher scatter observed for Al-rich
samples, especially sample II-5, in Figure 2. To our knowledge,
this is the first report of thermochemical data on nonstoichiometric (K, Na)-alunite-jarosite phases.
4. CONCLUSIONS
Fig. 7. ⌬Hds and ⌬H°f for the alunite-natroalunite solid solution
K0.8-xNax(H3O)0.2Al2.7(SO4)2(OH)5.1(H2O)0.9 (samples T-1, T-2, T-3
and H-3).
The thermochemistry of the jarosite-alunite and natrojarosite-natroalunite solid solutions was investigated by oxide
melt solution calorimetry. The enthalpies of formation from the
elements of 11 samples from these series and 3 samples from
the alunite-natroalunite solid solution series were determined.
⌬H°f becomes more endothermic with the iron content. Slightly
positive enthalpies of mixing were found, indicating that these
solid solutions involving B-site Fe/Al substitutions are not
thermodynamically ideal.
The heats of formation of the end members of the series were
extrapolated from the experimental data. These thermodynamic
data should enable the estimation/prediction of the heats of
formation of natural (K-Na-H3O)-jarosite-alunite samples, encountered, for example, in acid rock drainage environments,
which generally depart from the AB3(SO4)2(OH)6 stoichiometry. Also, under given conditions, the partitioning of Fe and Al
between the solids and the aqueous solution may be estimated
Jarosite-alunite thermochemistry
using these thermodynamic data and aqueous speciation programs, which were beyond the scope of the present paper.
The strong affinity of the jarosite-alunite structure for Fe,
observed in this work, is consistent with observed metal uptake
in industrial applications using jarosite precipitation, e.g., in
zinc hydrometallurgy, to eliminate iron impurities from raw
minerals (Dutrizac and Jambor, 2000).
Acknowledgments—This work was supported by the U.S. Department
of Energy (grant DE-FG0397SF 14749). The authors also wish to thank
greatly Dr. C. Alpers for a thorough review of this contribution.
Associate editor: D. Wesolowski
REFERENCES
Alpers C. N., Rye R. O., Nordstrom D. K., White L. D., and Bi-Shia
King (1992) Chemical, crystallographic and stable isotopic properties of alunite and jarosite from acid-hypersaline Australian lakes.
Chem. Geol. 96, 203–226.
Baron D. and Palmer C. D. (1996) Solubility of jarosite at 4 –35 °C.
Geochim. Cosmochim. Acta 60, 185–195.
Bohmhammel K., Naumann R., and Paulik F. (1987) Thermoanalytical
and calorimetric investigations on the formation and decomposition
of some alunites. Thermochim. Acta 121, 109 –119.
Brophy G. P., Scott E. S., and Snellgrove R. A. (1962) Sulfate studies
II. Solid solution between alunite and jarosite. Am. Miner. 47,
112–126.
Drouet C. and Navrotsky A. (2003) Synthesis, characterization, and
thermochemistry of K-Na-H3O jarosites. Geochim. Cosmochim.
Acta 67, 2063–2076.
Drouet C., Baron D., and Navrotsky A. (2003) On the thermochemistry
of the solid solution between jarosite and its chromate analog. Am.
Miner., in press.
Dutrizac J. E. and Jambor J. L. (2000) Jarosites and their applications
in hydrometallurgy. Reviews in Mineralogy and Geochemistry 40,
405– 452.
Dutrizac J. E. and Kaiman S. (1976) Synthesis and properties of
jarosite-type compounds. Can. Mineral. 14, 151–158.
Härtig C., Brand P., and Bohmhammel K. (1984) Fe-Al isomorphie und
strukturwasser in Kristallen von jarosit-alunit typ. Z. Anorg. Allg.
Chem. 508, 159 –164.
Hendricks S. B. (1937) The crystal structure of alunite and the jarosites.
Am. Miner. 22, 773–784.
Hofstra A. H., Snee L. W., Rye R. O., Folger H. W., Phinisey J. D.,
Loranger R. J., Dahl A. R., Naeser C. W., Stein H. J., and Lewchuk
M. T. (1999) Age constraints on Jerritt Canyon and other Carlin-type
gold deposits in the western United States: Relationship to midtertiary extension and magnetism. Economic Geol. 94, 769 – 802.
Kloprogge J. T., Ruan H. D., and Frost R. L. (2002) Thermal decomposition of bauxite minerals: infrared emission spectroscopy of
gibbsite, boehmite and diaspore. J. Mat. Sci. 37, 1121–1129.
Kubisz J. (1970) Studies on synthetic alkali-hydronium jarosites. I.
Synthesis of jarosite and natrojarosite. Mineralogia Polonica 1,
47–57.
Kubisz J. (1971) Studies on synthetic alkali-hydronium jarosites. II:
Thermal investigations. Mineralogia Polonica 2, 51–59.
Long D. T., Fegan N. E., McKee J. D., Lyons W. B., Hines M. E., and
Macumber P. G. (1992) Formation of alunite, jarosite and hydrous
2205
iron oxides in a hypersaline system: Lake Tyrrell, Victoria, Australia. Chem. Geol. 96, 183–202.
Lueth V. W., Rye R. O., and Peters L. (2000) Hydrothermal “sour gas”
jarosite: ancient to modern acid sulfate mineralization events in the
Rio Grande rift. Geological Society of America, 2000 Annual Meeting, Abstracts and Programs, pp A49.
Lueth V. W. and Rye R. O. (2003) Sulfuric and hydrofluoric acid
speleogenesis associated with fluorite-jarosite mineralization along
the Rio Grande rift, New Mexico. New Mexico Geological Society
2003 Spring Meeting, Abstracts and programs and New Mexico
Geology 25, pp 42.
Lueth V. W., Rye R. O., and Peters L. (in press) “Sour gas” hydrothermal jarosite: ancient to modern acid sulfate mineralization in the
southern Rio Grande rift: chemical geology.
Majzlan J., Navrotsky A., and Neil J. M. (2002) Energetics of anhydrite, barite, celestine, and anglesite: A high-temperature and differential scanning calorimetry study. Geochim. Cosmochim. Acta 66,
1839 –1850.
Menchetti S. and Sabelli C. (1976) Crystal chemistry of the alunite
series; crystal structure refinement of alunite and synthetic jarosite.
Neues Jahrbuch fuer Mineralogie. Monatshefte 9, 406 – 417.
Mitov I., Paneva D., and Kunev B. (2002) Comparative study of the
thermal decomposition of iron oxyhydroxides. Therm. Acta 386,
179 –188.
Navrotsky A. (1977) Progress and new directions in high temperature
calorimetry. Phys. Chem. Miner. 2, 89 –104.
Navrotsky A. (1997) Progress and new directions in high temperature
calorimetry revisited. Phys. Chem. Miner. 24, 222–241.
Okada K., Hirabayashi J., and Ossaka J. (1982) Crystal structure of
natroalunite and crystal chemistry of the alunite group. Neues Jahrbuch fuer Mineralogie. Monatshefte 12, 534 –540.
Parker R. L. (1954) Alunitic alteration at Marysvale, Utah. Ph.D.
Thesis, Columbia University.
Parker R. L. (1962) Isomorphous substitution in natural and synthetic
alunite. Am. Miner. 47, 127–136.
Powers D. A., Rossman G. R., Schugar H. J., and Gray H. B. (1975)
Magnetic behavior and infrared spectra of jarosite, basic iron sulfate
and their chromate analogs. J. Sol. Stat. Chem. 13, 1–13.
Pysiak J. and Glinka A. (1981) Thermal decomposition of basic aluminium potassium sulphate. Part I: Stages of decomposition. Thermochimica Acta 44, 21–28.
Ripmeester J. A., Ratcliffe C. I., Dutrizac J. E., and Jambor J. L. (1986)
Hydronium in the alunite-jarosite group. Can. Miner. 22, 773–784.
Robie R. and A. Hemingway B. S. (1995) Thermodynamic properties
of minerals and related substances at 298.15K and 1bar (105 Pascals)
pressure and at higher temperatures. U.S. Geological Survey Bull.
2131.
Stoffregen R. E. and Alpers C. N. (1992) Observations on the unit-cell
dimensions, H2O contents, and ␦D values of natural and synthetic
alunite. Am. Miner. 77, 1092–1098.
Stoffregen R. E. and Cygan G. L. (1990) An experimental study of
Na-K exchange between alunite and aqueous sulfate solutions. Am.
Miner. 75, 209 –220.
Stoffregen R. E., Alpers C. N., and Jambor J. L. (2000) Alunite-jarosite
crystallography, thermodynamics and geochronology. Reviews in
Mineralogy and Geochemistry 40, 453– 479.
Wolska E., Szajda W., and Piszora P. (1992) Determination of solid
solution limits based on the thermal behavior of aluminum substituted iron hydroxides and oxides. J. Therm. Anal. 38, 2115–2122.