Crystallization and Phase Stability of CaSO4 and CaSO4-based Salts

Article (PDF Available)inMonatshefte fuer Chemie/Chemical Monthly 134(5):693-719 · April 2003with 4,745 Reads
DOI: 10.1007/s00706-003-0590-3
Abstract
Calcium sulfate occurs in nature in form of three different minerals distinguished by the degree of hydration: gypsum (CaSO4·2H2O), hemihydrate (CaSO4·0.5H2O) and anhydrite (CaSO4). On the one hand the conversion of these phases into each other takes place in nature and on the other hand it represents the basis of gypsum-based building materials. The present paper reviews available phase diagram and crystallization kinetics information on the formation of calcium sulfate phases, including CaSO4-based double salts and solid solutions. Uncertainties in the solubility diagram CaSO4–H2O due to slow crystallization kinetics particularly of anhydrite cause uncertainties in the stable branch of crystallization. Despite several attempts to fix the transition temperatures of gypsum–anhydrite and gypsum–hemihydrate by especially designed experiments or thermodynamic data analysis, they still vary within a range from 42–60°C and 80–110°C. Electrolyte solutions decrease the transition temperatures in dependence on water activity. Dry or wet dehydration of gypsum yields hemihydrates (α-, β-) with different thermal and re-hydration behaviour, the reason of which is still unclear. However, crystal morphology has a strong influence. Gypsum forms solid solutions by incorporating the ions HPO42−, HAsO42−, SeO42−, CrO42−, as well as ion combinations Na+(H2PO4)− and Ln3+(PO4)3−. The channel structure of calcium sulfate hemihydrate allows for more flexible ion substitutions. Its ion substituted phases and certain double salts of calcium sulfate seem to play an important role as intermediates in the conversion kinetics of gypsum into anhydrite or other anhydrous double salts in aqueous solutions. The same is true for the opposite process of anhydrite hydration to gypsum. Knowledge about stability ranges (temperature, composition) of double salts with alkaline and alkaline earth sulfates (esp. Na2SO4, K2SO4, MgSO4, SrSO4) under anhydrous and aqueous conditions is still very incomplete, despite some progress made for the systems Na2SO4–CaSO4 and K2SO4–CaSO4–H2O.
Monatshefte f
uur Chemie 134, 693–719 (2003)
DOI 10.1007/s00706-003-0590-3
Invited Review
Crystallization and Phase Stability
of CaSO
4
and CaSO
4
Based Salts
Daniela Freyer
and Wolfgang Voigt
Institut f
uur Anorganische Chemie, TU Bergakademie Freiberg, D-09596 Freiberg, Germany
Received December 17, 2002; accepted January 10, 2003
Published online April 3, 2003 # Springer-Verlag 2003
Summary. Calcium sulfate occurs in nature in form of three different minerals distinguished by the
degree of hydration: gypsum (CaSO
4
2H
2
O), hemihydrate (CaSO
4
0.5H
2
O) and anhydrite (CaSO
4
).
On the one hand the conversion of these phases into each other takes place in nature and on the other
hand it represents the basis of gypsum-based building materials. The present paper reviews available
phase diagram and crystallization kinetics information on the formation of calcium sulfate phases,
including CaSO
4
-based double salts and solid solutions.
Uncertainties in the solubility diagram CaSO
4
–H
2
O due to slow crystallization kinetics particularly
of anhydrite cause uncertainties in the stable branch of crystallization. Despite several attempts to x
the transition temperatures of gypsumanhydrite and gypsumhemihydrate by especially designed
experiments or thermodynamic data analysis, they still vary within a range from 42–60
Cand
80–110
C. Electrolyte solutions decrease the transition temperatures in dependence on water activity.
Dry or wet dehydration of gypsum yields hemihydrates (-, -) with different thermal and re-
hydration behaviour, the reason of which is still unclear. However, crystal morphology has a strong
influence.
Gypsum forms solid solutions by incorporating the ions HPO
4
2
,HAsO
4
2
,SeO
4
2
,CrO
4
2
,as
well as ion combinations Na
þ
(H
2
PO
4
)
and Ln
3 þ
(PO
4
)
3
. The channel structure of calcium sulfate
hemihydrate allows for more flexible ion substitutions. Its ion substituted phases and certain double
salts of calcium sulfate seem to play an important role as intermediates in the conversion kinetics of
gypsum into anhydrite or other anhydrous double salts in aqueous solutions. The same is true for the
opposite process of anhydrite hydration to gypsum. Knowledge about stability ranges (temperature,
composition) of double salts with alkaline and alkaline earth sulfates (esp. Na
2
SO
4
,K
2
SO
4
,MgSO
4
,
SrSO
4
) under anhydrous and aqueous conditions is still very incomplete, despite some progress made
for the systems Na
2
SO
4
–CaSO
4
and K
2
SO
4
CaSO
4
–H
2
O.
Keywords. Calcium sulfate phases; Gypsum; Hemihydrate; Solubility; Polyhalite.
Corresponding author. E-mail: daniela.freyer@chemie.tu-freiberg.de
Dedicated to Professor H. Gamsj
aager on occasion of his 70
th
birthday
Introduction
Calcium sulfate in form of anhydrite or gypsum represents the most abundant
sulfate mineral in nature. Evaporitic deposits contain also various amounts of
double or triple salt minerals like syngenite, K
2
SO
4
CaSO
4
H
2
O, glauberite,
Na
2
SO
4
CaSO
4
, or polyhalite, K
2
SO
4
MgSO
4
2CaSO
4
2H
2
O. In connexion with
its application as binder or building material much attention has been paid to the
hydrationdehydration processes of calcium sulfate under various conditions.
A number of important industrial processes like the wet limestone-gypsum ue-
gas desulfurization (FGD), production of phosphoric acid or phosphate fertilizers,
desalination of brackish or seawater, hydrometallurgical production of zinc and
copper, and recovery of natural gas and oil, are accompanied by crystallization
of calcium sulfate phases. For more than half a century efforts have been directed
toward the control of growth rate and morphology of gypsum crystals formed in
these processes or the prevention of its growth (anti-scaling). Crystallization and
transformation of the calcium sulfate phases are inuenced in a complex manner
by temperature, pressure, dissolved electrolytes or organics, and the presence of its
own or other minerals. Knowledge of the respective phase diagrams and solubility
data belong to the necessary prerequisites to investigate crystallization processes.
Together with kinetic information and structural relations between important
phases, a basis for mechanistic understanding and control of crystallization would
be provided. It is the aim of this review to summarize such information and to point
out deciencies. Thereby, the scope is broadened to related systems and conditions,
where certain calcium sulfate containing phases can be formed.
System CaSO
4
H
2
O
Phases and Structures
In contact with water three phases of calcium sulfate can crystallize: gypsum,
anhydrite, and hemihydrate. Their main structural features are illustrated in
Fig. 1. Common structural motif in all CaSO
4
phases are chains in the form
[CaSO
4
CaSO
4
], where sulfate tetrahedra are coordinated through oxygen
atoms with two neighbouring Ca ions in chain direction.
Gypsum is distinguished by a marked layer structure with perfect cleavability
parallel to (010), where the sheets of coordinated water are located. Vibrational
spectroscopy revealed a pressure-induced phase transition in gypsum structure at
approx. 56 GPa caused by water disorder [1].
In the hemihydrate structure the Ca
2þ
SO
4
2
-chains run along the c-axis and
form channels of approx. 4 A
˚
diameter hosting the water molecules (Fig. 1b). Sym-
metry is near a threefold screw axis. Deviations are obviously correlated with water
content as discussed below. Water molecules can obviously be replaced by other
small molecules like methanol [2]. By means of careful drying of hemihydrate the
water can be removed almost quantitatively, which yields the so-called soluble an-
hydrite (also denoted as AIII phase or -CaSO
4
) with hexagonal symmetry.
The structure of the thermodynamic stable orthorhombic anhydrite (AII,
insoluble anhydrite) contains the CaSO
4
Ca chains oriented in direction of the
694 D. Freyer and W. Voigt
shorter c-axis. In a- and b-direction these chains are corner-linked as illustrated in
Fig. 1c.
Solubility Diagram Thermodynamic Force of Transformations,
Temperature Dependence
In Fig. 2 the solubility curves of the three phases gypsum, anhydrite, and hemi-
hydrate are plotted in the temperature range 0200
C at saturation pressure. At a
given temperature the solid phase with the lowest solubility represents the stable
phase. At low temperatures this is gypsum, at high temperatures it is anhydrite.
Hemihydrate remains metastable at all temperatures. Due to slow crystallization
kinetics the solubility determinations of a given solid can be extended into the
metastable temperature range when nuclei of the stable phase are absent. Intersec-
tions of these curves yield the transition temperatures gypsumanhydrite and gyp-
sumhemihydrate. Both temperatures are of considerable importance for the
Fig. 1. Structures of a) gypsum, b) hemihydrate, and c) insoluble anhydrite with view in c-direction
Crystallization and Phase Stability of CaSO
4
695
geology of evaporites (gypsumanhydrite) or the production and application of
gypsum products (gypsumhemihydrate). As can be seen from Fig. 2 the data
are considerably scattered and thus the intersection points vary depending on
selected data.
Discussion of the transition temperature gypsumanhydrite has a long history
with changing opinion about its correct value. Most of the difculties in this re-
spect arise from the fact that anhydrite does not crystallize in water with measur-
able rate at temperatures below 70
C, even in presence of anhydrite seed crystals.
Thus, the solubility equilibrium of anhydrite cannot be proved by approaching
from both sides, that is, from under- and supersaturation.
In Fig. 3 the relevant part of the solubility diagram is enlarged and it can be
seen that the borderlines of the data points yield transition temperatures from about
2552
C.
However, the rst value proposed by Van t Hoff et al. [9] at T¼63
C was even
higher and was derived from dilatometric investigations and thermodynamic
considerations. This value was essentially accepted until the review on solubility
equilibria of the oceanic salt systems prepared by DAns [3]. Hill [6] and Posnjak
[10] determined the solubility carefully and argued that the transition temperature
should be much lower, that is at 422
C. In a revised critical discussion
DAns et al. [4] supported this view by including their own solubility data and
Fig. 2. Literature data of solubility of gypsum, anhydrite, and hemihydrate in the temperature range
0200
C at saturation pressure
Fig. 3. Relevant enlarged part of the solubility diagram (Fig. 2) for transition temperature gypsum
anhydrite, values of Hill [6] and Posnjak [10] are marked
696 D. Freyer and W. Voigt
thermodynamic modelling. The differential dissolution enthalpies of gypsum de-
rived from the temperature dependence of solubility agreed within 1.6 kJmol
1
with enthalpy measurements by Lange et al. [11], which DAns et al. [4] took as
additional support for the lower transition temperature. In 1967 Hardie [12] pub-
lished results of an extensive work to decide the question of gypsumanhydrite
conversion. He used a reaction approach. Gypsum, anhydrite, or their mixtures
were suspended in sulfuric acid or sodium sulfate solutions of a certain water
activity at constant temperatures between 2070
C for 20 to 300 days. After the
experiments X-ray patterns were recorded and from the intensity ratio of one
typical reex the degree of conversion was determined. An extrapolation of a plot
conversion temperature versus water activity from solutions with lower water activ-
ities to pure water yielded a transition temperature of 582
C. Although DAns
[7] replied very decisively against the statement of Hardie [12], it seems that this
temperature is presently considered as more or less correct. In a more recent
thermodynamic analysis of the system CaSO
4
NaClH
2
O Raju et al. [13] claim
agreement with this value. For the calculation of the Gibbs energy, DG, of reaction
(I) the authors used solely thermodynamic standard data from sources [14] and
[15].
CaSO
4
2H
2
OðsÞ ð CaSO
4
ðsÞþ2H
2
OðlÞðIÞ
DG passes zero at T¼60
C. Unfortunately, it was not exactly stated which data
from which source had been selected. Nevertheless Knacke and Gans [16]
pointed out that they corrected the standard entropy of anhydrite in [17] by
1.6 Jmol
1
K
1
in order to t the transition temperature gypsumanhydrite
of 55.52
C, determined by themselves in a particular experiment. Thus, the
transition temperature calculated from tables of standard thermodynamic data is
not independent on solubility data.
In our opinion, the presently preferred transition temperatures between 55
60
C have no more justication than the lower values between 4245
C. From
both experimental studies [12, 16] higher transition temperatures are not beyond
doubt. The high extrapolated temperature in the water activity-temperature plot
used by [12] is xed by ve experiments between 50 and 55
C, in which reaction
of anhydrite to gypsum was observed, however, with one exception not to com-
pletion. There were also runs with unexpected results. From the description of the
experimental part it becomes clear that the water activity was not constant during
the experiment, since approx. 100 g of sulfate were obviously suspended in approx.
100 g sulfuric acid solution. Thus, depending on the reaction (hydration or dehy-
dration) solution concentration and hence the water activity will change.
Knacke and Gans [16] designed a clever experiment exploiting the fact that in a
mixed suspension of anhydrite and gypsum the solubility is exclusively controlled
by gypsum with the much faster crystallization kinetics. Thus, they were able to
agitate anhydrite over a period of 3 months at temperatures near the assumed
transition point gypsumanhydrite. Above the transition temperature gypsum is
metastable and its solubility is higher than that of anhydrite. Consequently, addi-
tion of gypsum causes dissolution of gypsum, which was detected by a correspond-
ing increase of electrical conductivity. The opposite effect, decrease of electrical
conductivity, was observed below the transition temperature. In this way the
Crystallization and Phase Stability of CaSO
4
697
transition temperature was xed at 55.51.5
C. However, in their gure the con-
ductivity of the solution obviously increases over the whole period of the experi-
ment (two and three months, respectively). Assuming proportionality between
conductivity and dissolved CaSO
4
concentration the increase is more than 6%.
Reported anhydrite dissolution kinetic data [4, 6] show that constant concentration
within 12% of the value is reached after 2030 days. Also in cases where experi-
ments lasted up to several months [6] no further concentration change was found.
Thus, the reason for the conductivity increase remains unclear and consequently
the conclusions drawn from the experiments must be questioned.
In Fig. 4 the crossing region of the solubility curve of gypsum and hemihydrate
is shown. Within the scatter of the solubility data the possible transition temper-
ature gypsumhemihydrate covers a range from less than 80 to nearly 110
C. From
their dilatometric and tensiometric experiments Van t Hoff et al. [9] derived 106
C
as transition temperature. This value was accepted until Posnjak [10] critically
analysed the work of Va nt Hoff et al. [9]. Posnjak [10] derived a transition tem-
perature of 971
C from solubility data and supported this value by means of a
particular experiment, in which he observed complete conversion of gypsum into
large hemihydrate crystals in pure water at a temperature below 100.5
C within
two days.
Influence of Electrolytes on Transformation Temperature
There are numerous data on the solubility of gypsum in electrolyte solutions [18,
19]. More recent solubility determinations and discussion of trends can be found in
[2023]. In general the addition of non-common ion electrolytes enhances the
solubility and can reach the tenfold value of the pure gypsum solution. With in-
creasing electrolyte concentration the gypsum solubility surpasses a maximum
value. The decrease of solubility at high electrolyte concentrations correlates ob-
viously with the hydration ability of the electrolyte [23]. Most frequently the in-
uence of NaCl was investigated. In this case both gypsum and anhydrite solubility
were studied. In Fig. 5 data for T¼25
C are plotted. The two solubility curves
intersect at a NaCl content of approximately 4 molkg
1
H
2
O. Below this con-
centration gypsum represents the stable solid, above anhydrite. In a solution satu-
rated with NaCl the transition temperature gypsumanhydrite was estimated to
Fig. 4. Relevant enlarged part of the solubility diagram (Fig. 2) for the metastable transition tem-
perature of gypsumhemihydrate
698 D. Freyer and W. Voigt
18
C [4, 12]. Different authors more or less agree about this temperature. The
reason may be seen in a faster equilibration in concentrated salt solutions.
For a comparison of the effects of various electrolytes on the shift of the
gypsumanhydrite transition only the water activity of the solutions has to be
considered, given that the interaction between electrolyte and calcium sulfate does
not yield new solid phases. From the dehydration reaction (I) and the appropriate
expression of the Gibbs energy of this reaction, DG [Eq. (1)], it follows that the
transition temperature decreases with decreasing water activity.
DG ¼ DG
ø
2RT lna
W
¼ 0 ð1Þ
For a given temperature function of DG
ø
the iterative solution of Eq. (1) pro-
vides the transition temperature as a function of water activity. Applying the stan-
dard data used by Raju et al. [13] a curve as given in Fig. 6 is obtained. The vertical
line depicts the water activity of a saturated NaCl solution (a
w
¼0.75), which is
quite temperature independent. Thus, with the DG
ø
function of Raju et al. [13] a
transition temperature of about 32
C is calculated, which is much too high. Re-
correcting the 1.6 Jmol
1
K
1
entropy adjustment introduced by Knacke and
Gans [16] to the standard data of Barin et al. [17] results in a transition temperature
closer to the accepted value (see Fig. 6).
The inuence of electrolytes on the transition temperature gypsumhemihy-
drate can be discussed in a similar way. For a saturated NaCl solution a transition
Fig. 5. Solubility of gypsum and anhydrite in dependence on sodium chloride concentration
at T¼25
C
Fig. 6. Transition temperature gypsumanhydrite derived from Raju et al. [13]
Crystallization and Phase Stability of CaSO
4
699
temperature of 76
C was estimated from vapour pressure data in [4], which at least
does not contradict the experimentally observed conversion of gypsum into hemi-
hydrate at 83 [4] and 75
C [24]. With solutions of lower water activity, such as of
MgCl
2
(aq.) or concentrated strong acids (e.g., HNO
3
) hemihydrate can be formed
from gypsum near room temperature, which is exploited in preparative work. How-
ever, to our knowledge no systematic study or thermodynamic analysis of the
hemihydrate formation in solutions of low water activity and temperatures has
been undertaken.
Pressure Dependence
The solubility of all CaSO
4
phases increases with pressure. Monnin [25, 26] per-
formed a detailed thermodynamic analysis of the pressure effect on solubility in-
cluding the interaction with major sea water components over a broad temperature
range. Another thermodynamic model of the pressure effect at 25
C was developed
by Krumgalz et al. [27]. Increase of solubility with pressure is higher for anhydrite
than for gypsum, which must cause a shift in transition temperature gypsum
anhydrite. Mac Donald [28] estimated an increase by 8 K if pressure is raised by
500 bar. No such estimations exist for the formation of hemihydrate. High pressure
experiments of Kr
uuger et al. [29] up to 6 GPa and 400
C gave no evidence for
enhanced thermodynamic stability of hemihydrate with respect to anhydrite.
Crystallization of Gypsum
Nucleation, Crystallization Kinetics, and Morphology
Induction period measurements with an optical technique [30] at relative super-
saturations of 14 and temperatures between 2590
C resulted in an apparent en-
ergy of activation for nucleation (30 kJmol
1
) and an interfacial tension of
37 mJm
2
. Supersaturation was generated by mixing Na
2
SO
4
and CaCl
2
solu-
tions of appropriate concentrations. Logarithmic plots of induction time versus
supersaturation allowed to distinguish between homogeneous and heterogeneous
nucleation mechanisms.
He et al. [31] determined the induction times by means of turbidity combined
with sampling of solution and titrating Ca
2þ
concentration with EDTA in NaCl
solutions up to 6 molkg
1
. Whereas Lancia et al. [30] found no temperature
dependence of the interfacial tension, He et al. [31] reported increasing values
with temperature from 39 mJm
2
at 25
Cto64mJm
2
at 90
C. NaCl has a
strong inuence on gypsum nucleation [3133]. Interestingly, the variation of in-
duction times with NaCl concentration is opposite to the trend of gypsum solubility
in dependence on NaCl concentration [31]. That is, shortest induction times are
observed at 3 molkg
1
, where the solubility of gypsum passes a maximum, and
therefore concentration supersaturation is lowest. It could be demonstrated by the
authors that the induction period is correlated with the interfacial tension, which
decreases with solubility. An activation energy of nucleation of 53 kJmol
1
was
calculated from the temperature dependence at 3 M NaCl.
700 D. Freyer and W. Voigt
In a subsequent paper of He et al. [34] on growth kinetics no induction period
was reported when seeding with 0.54 g of gypsum per kg of water. Hina et al.
[35] investigated the crystallization kinetics of gypsum in dependence on the molar
Ca
2þ
=SO
4
2
ratio in solution. The growth rate decreased with increase of this ratio.
With KCl as supporting electrolyte the effect was stronger than with NaCl. In NaCl
solutions the growth rate was generally lower than in KCl solutions. The effect
does not correlate with variations in the thermodynamic driving force with solution
compositions. Sodium ion adsorption was suggested as a reason for kinetic retar-
dation. More drastic growth retardation is reported when La
3þ
,Ce
3þ
, and Eu
3þ
ions are present [36]. Lanthanum, which has the largest effect, reduces the growth
rate by a factor of 10 already at concentrations of 310
4
M.
In the presence of inhibiting additives like phosphonates (ENTMP, TENTMP)
an induction period is observed [37] even in seeded solutions. Growth kinetics after
the induction period is nearly the same as in solutions without additives and no
change in morphology could be detected. From these facts it was concluded that
the inhibitors are adsorbed at the gypsum surface and incorporated by overgrowth.
The kinetics of gypsum crystallization from anhydrite suspended in water at
1040
C was investigated by Kontrec et al. [38]. Under the conditions chosen (few
grams of solid per dm
3
solution) transformation kinetics was dependent on both
dissolution kinetics of anhydrite and growth of gypsum. The authors derived a
kinetic model for this interdependence from their experimental data.
Inuence of PANa on homogeneous and heterogeneous nucleation was inves-
tigated by Boisvert et al. [39]. They found blocking of gypsum nucleation as the
controlling factor on the kinetics of hemihydrate to gypsum conversion.
Bertoldi [40] tested a long list of organic and inorganic additives on their in-
uence on crystal size and shape, however, no denite conclusions had been drawn
with respect to certain additives. Konak [41] carried out similar experiments but did
not report much about experimental details. For Separan (polyacrylamide) he stat-
ed a pronounced retardation in crystallization and growth of aggregates consisting
of long, needle-shaped crystals [41]. Effects of additives on morphology and tex-
ture were also examined by Amathieu et al. [42]. They tested the inuence of
malonic, tartaric, and polyacrylic acid, as well as sodium tripolyphosphate, sodium
laurylsulfate, dodecylammonium chloride, and ammonium sulfate. Attempts to
correlate texture changes with strength of the nal gypsum product is not
straight-forward. Thus, large crystals can generate large and also low strengths.
In a more recent study [43] the inuence of malic acid, citric acid, tartaric acid, and
adipic acid on gypsum crystallization kinetics and morphology from suspended
hemihydrate was investigated. Also the adsorption of adipic and rac-malic acid was
measured by capillary zone electrophoresis. Largest effects were observed for cit-
ric acid and malic acid. The results were explained by positional matching of the
acids oxygen with the Ca
2 þ
distances on the (120) and (111) crystallographic
planes on gypsum.
From texture analysis Follner et al. [44] conclude that after setting of -hemi-
hydrate at a low water=gypsum ratio there is a tendency of parallel orientation of the
(010) faces of gypsum with (100) faces of hemihydrate. No preferred orientation of
gypsum formed from hemihydrate at water=solid ratios between 0.50 to 1.50 was
observed in time-resolved synchrotron X-ray powder diffraction experiments [45].
Crystallization and Phase Stability of CaSO
4
701
Hydration of anhydrite into gypsum is accelerated by certain acids, bases, and
salts. Alkali sulfates are the most effective. In dilute solutions hydration proceeds
via transient complexes. The formation of complexes is believed to be a surface
ionic transfer process. Depending upon temperature and concentration double salts
can be formed [46].
Ion Substitution in Gypsum and Related Phases
Kushnir [47] investigated the co-precipitation of Na
þ
,K
þ
,Mg
2þ
, and Sr
2þ
with
gypsum and determined distribution coefcients in dependence on concentration
between 3050
C. No extensive incorporation is observed for Na
þ
,K
þ
, and
Mg
2þ
with values of D (x
G
=x
Brine
) in the range of 10
5
to 310
4
but a corre-
lation with the growth rate of gypsum is found. The value for Sr
2þ
is between 0.2
to 0.7, which is much higher, but because of the low solubility of SrSO
4
the ab-
solute amounts of incorporated Sr
2þ
are very small. Uptake of Cd
2þ
during gyp-
sum crystallization from dilute solutions was determined but no specic effects
could be found [48].
Extensive solid solutions are formed with HPO
4
2
ions. The mineral brushite,
CaHPO
4
2H
2
O, is structurally very similar to gypsum, although proton ordering
lowers the symmetry to Ia [49], whereas the gypsum structure is described with
I2=c. Thus, an isodimorphic series of mixed crystals was found from inspection of
IR spectra and X-ray powder patterns [50]. According to these authors up to 70
mass-% CaHPO
4
2H
2
O can be incorporated into the gypsum structure. Conse-
quently the mineral ardealite, CaSO
4
CaHPO
4
4H
2
O, should be considered as a
gypsum based mixed crystal formulated as Ca(SO
4
)
1-x
(HPO
4
)
x
2H
2
O with x 0.5
[50]. This view is supported by the site occupation factors of 0.5 for sulphur and
phosphorus in a crystal structure analysis [51].
At a pressure of 2.1 GPa pure brushite undergoes a phase transition [52]. In a
more recent crystallization study in the gypsumbrushite system stable mixed
crystals were only obtained from solutions with at least 30% phosphate. From
solutions with 1020% phosphate mixed crystals appeared as the rst phase, then
gypsum precipitated and the mixed crystals dissolved. When gypsum crystallized
alone it did not contain any detectable amounts of phosphate [53]. Kinetic experi-
ments have shown that brushite may serve as an effective nucleator in gypsum
crystallization [54].
The dehydration of CaHPO
4
2H
2
O differs from that of gypsum, an adequate
hemihydrate is not known [55]. Monosodium phosphate, NaH
2
PO
4
, can also be
incorporated into the gypsum lattice [56]. Calcium hydrogen arsenate also forms a
dihydrate with a gypsum analog structure [57]. However, no detailed studies of
arsenate uptake by gypsum are available.
The dihydrate of calcium selenate crystallizes also in a gypsum type structure
with space group I2=a [58] and lattice constants very similar to gypsum, however,
its solubility is more than ten times higher [59]. Thermal dehydration is similar to
gypsum with formation of an intermediate hemihydrate at approx. 137
C [60]. One
should expect a continuous series of isomorphic mixed crystals between these two
dihydrates, however, experimental investigations about this subject could not be
found in literature.
702 D. Freyer and W. Voigt
Double substitution of Ca
2þ
and SO
4
2
by trivalent lanthanide and phosphate
is realized in the mineral churchite, [Y
(1-x)
(Gd, Dy, Er)
x
]PO
4
2H
2
O, which is also
isomorphous with gypsum [61].
Interestingly, the crystal structure of CaCrO
4
2H
2
O is quite distinct from gyp-
sum [62], although there exists an unstable monoclinic polymorph, which is de-
scribed as isomorphous with gypsum [63].
Another mineral related to gypsum represents the recently discovered rapid-
creekite [64]. Lattice symmetry and dimensions are different compared with gyp-
sum. However, its structure can be derived by replacing every second sulfate row in
gypsum lattice by carbonate anions.
Calcium Sulfate Hemihydrate and Related Phases
Formation, Morphology, and Properties
Hemihydrate occurs in nature as the mineral bassanite. Recently, it was also dis-
covered in statoliths of a deep-sea medusa [65]. Generally hemihydrate can be
prepared from gypsum by drying at enhanced temperatures, the product obtained
by this way is denoted as -hemihydrate [6671]. The dehydration reaction is
described as a topotactic solid state reaction, thereby it is suggested that the orig-
inal [010] or [001] crystal axis of gypsum becomes the new [001] of the hemihy-
drate [72, 73]. In the dry system a stable T-p
H
2
O
region exists for the hemihydrate,
which was carefully investigated recently [74, 75].
In aqueous solutions, hemihydrate can be crystallized as a metastable phase,
because of the low crystallization rate of the stable anhydrite. The products pre-
pared from aqueous solutions are denoted as -hemihydrate. Normally, a suspen-
sion of gypsum is heated above the transformation temperature. The latter depends
on the water activity as discussed earlier. Passing the transition temperature, gyp-
sum in an aqueous suspension will be converted spontaneously into hemihydrate,
which crystallizes in typical aggregates of hexagonal columns. The transformation
of gypsum into -hemihydrate is described as a topotactic solid state reaction in
case of gypsum single crystals only. The transformation takes place under retention
of the crystallographic c-axis [76, 77]. For polycrystalline gypsum the dominant
mechanism is the dissolution of gypsum and generation of a supersaturation with
respect to hemihydrate followed by nucleation of hemihydrate and growth of nuclei
to macroscopic crystals from solution or on the surface of mother gypsum crystals
without directional correlation [76, 78]. In dependence on the history of the gyp-
sum used (formation, crystal size, impurities, ...) the mechanism of crystallization
is supplemented by the topochemical reaction in a certain degree. A deeper under-
standing of governing mechanisms is necessary to control crystal form and sizes as
an important issue in the production of special gypsum binder, where the control of
hemihydrate morphology is effected by organic and inorganic additives. A review
about signicance and operating mode of different additives is given by Koslowski
et al. [80]. More insight into the mechanistic steps could lead to a necessary com-
prehension of the effects of additives and their exploitation. Investigations on
homogenous hemihydrate nucleation would be desirable in order to differentiate
between solution sustained crystallization and the inuence of hetero surfaces.
Crystallization and Phase Stability of CaSO
4
703
Such effects, e.g. the epitaxial growth of hemihydrate on uorapatite was observed
by Dorozhkin [79]. Different electrolytes show also inuences on the -hemihy-
drate morphology, where at least the lowered transition temperature gypsumhemi-
hydrate represents one factor one has to consider. Table 1 gives examples for
hemihydrate formation experiments carried out by Fl
oorke [66].
The -hemihydrate differs in a number of properties (heat of solution, specic
surface, solubility, effects in thermal analysis) from the -form. There exist nu-
merous papers dealing with these differences. In Table 2 some important articles
are listed together with the investigated properties and main conclusion.
The dehydration of - and -hemihydrate takes place between 100200
C
depending on water vapour pressure followed by formation of soluble AIII and
its transformation to insoluble AII, which occurs at different temperatures for -
und -hemihydrate. The formation of AII is observed at temperatures between
200600
C [66, 68, 70, 88, 9294]. For -hemihydrate Kuntze [69], Powell
[82], and Cliffton [91] observed a weak exothermic effect at 350375
C. The
occurrence of an exothermic effect for -hemihydrate at lower temperatures close
to the dehydration is discussed controversially. Budnikov et al. [95] did not observe
such an effect. In contrast, Kuntze [69] monitored an exothermic effect below
250
C. Powell [82] and Cliffton [91] found the exothermic effect between
Table 1. Transformation of gypsum into hemihydrate in acid or salt solutions [66]
Electrolyte T=
C
HNO
3
(60%) 50, 80
concentrated NaCl-solution 80
concentrated MgCl
2
-solution 55
Table 2. Investigations on - and -hemihydrate properties
References Investigated Main conclusion
properties,
methods
[68, 8183] x-ray no basic differences of - und -HH
a)
weaker reexes for -HH due to crystal size and lattice defect
[8486] surface very high specic surface of -HH compared to -HH
[81, 8790, 91] IR no differences between - und -HH
weaker bands for -HH
[66, 68, 70, 82, thermal controversial results (see text below)
88, 9195] behaviour
[96, 97] hydration heat lower heat of hydration for -HH than for -HH, the latter shows
a wide variation in its hydration values
[98] structure no differences
[44] differences between - und -HH
[99] -HH monoclinic and -HH trigonal
a)
HH¼hemihydrate
704 D. Freyer and W. Voigt
163255
C depending on the water vapour pressure. Powell [82] noticed different
magnitudes of the effect for samples with different initial weights. The slower
transformation at higher temperatures of the -form is explained with a near rela-
tionship to AII [82, 92] but this assumption seems to be questionable.
Our own investigations on - and -hemihydrate show that the morphology
(changes in surface conditions) of the samples has a large inuence on the thermal
behaviour [100]. Crystals of bres (Fig. 7) show an apparently higher thermal
stability. The dehydration takes place at some higher temperatures (100160
C)
compared with disk-like crystals (90125
C) short along c-axis (Fig. 9). In
Figs. 79. Hemihydrate morphologies
Crystallization and Phase Stability of CaSO
4
705
this morphology the water molecules have a short outlet way. The hemihydrate
transformations were monitored by in situ Raman spectroscopy. A fast transfor-
mation to AIII was observed for stick-morphology (Fig. 8) in connexion with a
sharp exothermic effect. Hemihydrate bres transform very slowly into AIII, there-
fore the exothermic effect is not detectable. According to the exothermic effect for
-hemihydrate above 320
C the topotactical transformation to AII seems to be
retarded in a certain manner by the pseudomorphosis to gypsum. Analogously,
in the aqueous system the transformation of gypsum to AII is not a direct reaction.
The intermediate formation of -hemihydrate is the preferred reaction way. So, the
topotactical transformation of a well crystallized -hemihydrate compared with a
gypsum-pseudomorphic -hemihydrate is promoted. Probably, the contradictory
discussions in literature result from different hemihydrate morphologies, among
other things.
There are different conclusions in literature regarding water content and sym-
metry of hemihydrate. Kuzel et al. [98], Bushuev et al. [101], Abriel et al. [102],
and Oetzel et al. [75] proposed that the water content and thus the symmetry varied
depending on water vapour pressure (CaSO
4
xH
2
O, 0.5x0.8).
The so-called subhydrates are described with a trigonal space group P3
1
21 [94,
98, 102, 103]. For 0.5<x<1 a statistically disordered distribution of water mol-
ecules allows a deection from their positions to reach the necessary distances. The
pronounced anisotropy of the O
H
2
O
in the trigonal cell shows this fact [103]. For
x¼0.5 an ordered distribution of water molecules with a monoclinic cell is reached
by doubling of the c-axis length [98, 104] whereas literature presents various
structures (Table 3). Our earlier investigation [24] conrms a monoclinic lattice
of the hemihydrate based on low temperature Raman spectroscopy.
The transition hemihydratesubhydrate is demonstrated by small changes in
the powder diffraction pattern [98, 104], also by Oetzel et al. [75]. They determined
the vapour pressuretemperature condition in the solidgas equilibrium for its
Table 3. Structure of CaSO
4
0.5H
2
O proposed in literature
References Lattice Comment
[66, 105] orthorhombic
[66, 102],
[44] trigonal deviations from the trigonal sym-
metry are small and that the different
variants of occupying the crystallo-
graphically possible water positions
cannot be distinguished by X-ray
methods at present
[105] hexagonal
[98, 101, 104, 106],
[107] monoclinic, C2orI2 strongly pseudo-trigonal, small devia-
tions from trigonal structure arise
from water molecules ordering inside
the channels
706 D. Freyer and W. Voigt
existence. The term subhydrate is often used for hemihydrate without specied
water content.
Ion Substitution Sodium, Potassium, Strontium, and Rare Earth Cations
Sodium and Potassium
First hints for the inclusion of sodium ions in hemihydrate were given by Hill et al.
[108]. A so-called sodium pentasalt, Na
2
SO
4
5CaSO
4
3H
2
O, is obtained when gyp-
sum is stirred in sodium sulfate containing solutions at 75
C. Depending on tem-
perature conversion is completed after 4 days up to one week [108, 109]. With the
aim to prepare hemihydrate in sodium chloride solution a content of 2.32.9%
sodium was found in the chloride free CaSO
4
0.5H
2
O which is described by similar
properties as the sodium pentasalt [96, 110].
The similarity of the X-ray powder diffraction pattern with the hemihydrate
already suggests a hemihydrate structure, where Ca
2 þ
is statistically substituted by
Na
þ
. According to this explanation of the X-ray patterns one Na
þ
ion replaces
1=6ofCa
2 þ
in the lattice. For charge compensation a second Na
þ
is located
statistically inside the structure. The discussions differ between a location within
the CaSO
4
chain-matrix by distortions of sulfate tetrahedra or vacancy occupation
and locations inside the water channel (Fig. 10) [89, 94, 110, 111].
Reisdorf et al. [94] denote Na-pentasalt as Na-polyhalite, although the name
sodium polyhalite has been used already for a salt with the composition
3=5Na
2
SO
4
2=5K
2
SO
4
5CaSO
4
3H
2
O[112113], which also has hemihydrate
structure. Accepting this structural scheme one could expect a continuous series of
solid solutions CaSO
4
0.5H
2
O(Na
2x
Ca
6-x
)SO
4
0.5H
2
O [111]. Attempts to prepare
such solid solutions from Na
2
SO
4
or NaCl-containing solutions ended all the time
very close to the pentasalt composition Na
2
Ca
6
(SO
4
)
3
0.5H
2
O or at sodium contents
below Na
0.5
Ca
5.75
(SO
4
)
3
0.5H
2
O [109]. Thermal analyses support the view that so-
dium ions are located in the channels along the c-axis. As can be seen in Fig. 11 the
dehydration temperature of the hemihydrate increases with sodium content. The high-
est dehydration temperature is observed for the potassium containing salt, which was
rst prepared by Autenrieth [112] and which he named ‘‘Na-polyhalite’’.Inthision
substituted hemihydrate the sodium content is substituted partly by potassium. A
proposed structure based on the X-ray diffraction pattern by Gudowius et al. [113]
ascribed this salt also to the hemihydrate structure type. Its thermal behaviour ts into
this view. The potassium ions substitute sodium ions in the channel, which causes a
stronger hindrance for water molecules to leave along the channel [114].
Fig. 10. Projection along the channel in hemihydrate structure (A) beside the analogous projection
in the proposed pentasalt structure (B) [94]
Crystallization and Phase Stability of CaSO
4
707
Because of the possible substitution of ‘‘channel’’-sodium ions by potassium a
substitution by ammonia ions in the same way can be expected in the hemihydrate
structure. Crystal structure analysis from a small twin with sodium pentasalt com-
position yielded a superstructure of hemihydrate whereby one sodium position is
found in the channel near the fourth unoccupied water position beside the sodium
substituted Ca(2) position in the chain-matrix. For structural reasons the Ca(1)
position is not substituted [109] (Fig. 12). Thus higher sodium contents in the
hemihydrate enforce structural distortions, which gives rise to a discontinuity in
solid solution formation. However, more extensive investigations will be necessary
to determine the correlation between crystallographic parameters and composition
of solid and aqueous solution.
Preparation of solid solutions between CaSO
4
0.5H
2
O and SrSO
4
0.5H
2
O and
its identication by means of the X-ray patterns was reported [115, 116]. Precipi-
tation of the sulfates from 0.5 M nitrate solution by adding 20% sulfuric acid at
90
C resulted in a maximum of 14 mol-% SrSO
4
in CaSO
4
0.5H
2
O [116].
Rare Earth Cations
The ionic radii of the trivalent rare earth cations of the light lanthanides (LaEu)
are very similar to Ca
2 þ
, for Ce
3 þ
the values are nearly identical. Thus, isomor-
phic substitution of Ca
2 þ
should be possible if suitable ions for charge compen-
sation are available. Bushuev et al. [117] investigated the incorporation of Ce
3 þ
into the hemihydrate according to two different substitution schemes I and II.
2Ca
2þ
, Ce
3þ
þ Na
þ
ðIÞ
Ca
2þ
þ SO
4
2
, Ce
3þ
þ PO
4
3
ðIIÞ
Fig. 11. TG=DTA of hemihydrate phases according to the compositions IV [109, 114]; I
CaSO
4
0.5H
2
O (6CaSO
4
3H
2
O); II 0.23Na
2
SO
4
5.77CaSO
4
2.95H
2
O; III 0.9Na
2
SO
4
5.1CaSO
4
2.85H
2
O; IV 1.01Na
2
SO
4
4.99CaSO
4
2.90H
2
O; V (0.350.05)K
2
SO
4
(0.550.05)Na
2
SO
4
(50.05)CaSO
4
(30.05)H
2
O
708 D. Freyer and W. Voigt
Precipitation of the hemihydrates from nitrate solutions of suited compositions
by adding sulfuric or phosphoric acid at 7095
C yielded continuous series of
solid solutions CaSO
4
0.5H
2
ONa
0.5
Ce
0.5
(SO
4
)0.5H
2
O and CaSO
4
0.5H
2
O
CePO
4
0.5H
2
O. Existence of a solid solution phase was conrmed by recording
the X-ray diffraction patterns and determination of the lattice constants, which
varied smoothly with composition of the solid phase. Mixed crystals containing
Na
0.5
Ce
0.5
(SO
4
)0.5H
2
O were of more isometric habitus. The hemihydrate analog
structure was conrmed by crystal structure analysis for Na
0.5
Ce
0.5
(SO
4
)0.5H
2
O
[118a] and Na
0.5
La
0.5
(SO
4
)0.5H
2
O [118b]. Sorption kinetics and distribution co-
efcients of Ce
3þ
and Eu
3þ
in solutions of 7.5 M H
3
PO
4
and hemihydrate at 90
C
have been determined using radioactive isotopes of the metal ions [119].
The complete series of isomorphous mixed crystals NaLn(SO
4
)
2
H
2
O(Ln¼Y,
La, Ce, ..., Yb, Pu) was prepared and characterised by X-ray diffraction [120]. A
thorough thermal characterisation (TG, DTA, IR, X-ray) can be found in [121].
Since NaPu(SO
4
)
2
H
2
O is isomorphous to the corresponding rare earth sulfate mon-
ohydrates [120] the incorporation of Pu(III) into CaSO
4
phases with hemihydrate
structure can be expected. Interestingly, no incorporation of Ce
3þ
into gypsum was
observed. The ability of hemihydrate to incorporate lanthanide ions on one side and
the inability of gypsum for this substitution on the other side is exploited for extrac-
tion of rare earth elements in phosphoric acid production from apatitic ores [122].
Anhydrite
In aqueous solutions, the crystallisation of anhydrite is the most difcult of all
calcium sulfate phases and reports about its formation at temperatures below
Fig. 12. Projection across the channel in Na-pentasalt structure [109]
Crystallization and Phase Stability of CaSO
4
709
100
C are contradictory. There exists agreement that below 90
C no spontaneous
formation of anhydrite occurs. Dissolved salts facilitate anhydrite formation and
solidliquid equilibration. Hill [123] attempted to prepare anhydrite for dissolution
experiments by boiling gypsum in electrolyte solutions. The nal product after 13
days boiling was mostly hemihydrate. Hill [123] obtained successful conversion to
anhydrite within 3 days in 5% K
2
SO
4
or 1520% H
2
SO
4
solution. Crystal sizes of
anhydrite were in the range of 2030 mm. At 90.5
C gypsum was transformed
through hemihydrate into anhydrite within 10 days in solutions containing NaCl
or NaClMgCl
2
with chloride concentrations higher than 2.8 moldm
3
[124].
Ostroff [124] postulated that supersaturation with respect to hemihydrate is re-
quired to form anhydrite. Thus, in concentrated NaCl solution at 50 and 70
C
anhydrite was not formed corresponding to the fact that in saturated NaCl solution
hemihydrate is formed only above 80
C [4]. Hardie [12] argues that anhydrite is
not formed through a solution-precipitation mechanism, because addition of anhy-
drite has no accelerating effect, also citing similar results from Zen [125]. How-
ever, both authors reported unexpected experimental results of gypsum formation
from anhydrite under conditions where anhydrite should have been stable. Hydra-
tion of anhydrite to gypsum is also kinetically hindered. Without gypsum seed
crystals anhydrite can be agitated in solutions several months without any change
[12, 16].
Calcium Sulfate-Based Double Salts
Double Salts with Sodium Sulfate
In the anhydrous system the appearance of the compounds 4Na
2
SO
4
CaSO
4
,
Na
2
SO
4
CaSO
4
(the mineral glauberite), and 2Na
2
SO
4
CaSO
4
is discussed contro-
versially. Calcagni et al. [126] suppose a 3Na
2
SO
4
CaSO
4
phase with a melting
point at 949
C. Different versions of phase diagrams exist [127130]. The reasons
have to be seen in the tendency of high temperature forms of sodium sulfate to
coexist at low temperature in metastable equilibrium. Thus, the phase diagrams
reect kinetic situations but not equilibrium. In Freyer et al. [131] the formation
and transformation of metastable into stable phases is investigated thoroughly and
the equilibrium phase diagram is derived (Fig. 13). Glauberite is the only stable
anhydrous compound of sodium and calcium sulfate and will decompose above
520
C. The phase 4Na
2
SO
4
CaSO
4
[127] is part of a solid solution series of the
hexagonal form of Na
2
SO
4
(I). Thermal effects given in earlier phase diagrams at
230280
C arise from reproducible hexagonal , monoclinic transformations in
the metastable region of the phase diagram. The hexagonal form of Na
2
SO
4
(I)
represents the only stable solid solution. 2Na
2
SO
4
CaSO
4
[129] indicates a com-
position of another solid solution series of Na
2
SO
4
(I), which is metastable at all
temperatures. Depending on the CaSO
4
content three metastable forms of solid
solutions can be obtained by quenching of melts to room temperature. A transfor-
mation scheme into the stable phases was proposed [131].
Glauberite is also obtained from aqueous solutions as are the metastable hy-
drates 2Na
2
SO
4
CaSO
4
2H
2
O(‘‘labile salt’’) and Na
2
SO
4
5CaSO
4
3H
2
O, the so-
called sodium pentasalt already discussed before [46, 108, 132, 133]. Hill et al.
710 D. Freyer and W. Voigt
[108] were able to determine the corresponding metastable solubility equilibria at
25, 35, 50, and 75
C (Fig. 14). The same phases occur in the system NaCl
Na
2
SO
4
CaSO
4
H
2
O [134]. The ‘‘labile salt’’ was discovered beside glauberite in
1857 by Fritzsche [135]. It crystallizes with considerably slow conversion kinetics
to the stable phase glauberite [108, 136]. Further, no reproducible ways of forming
intermediate hydrates are described in Vasilevskaya [137], Druzhinin and Lopina-
Shendrik [138], Fridman [139], and Lopina-Shendrik [140]. Emons et al. [141]
postulated the existence of an orthorhombic solid solution causing the different
hydrate stoichiometries in literature. Denite conclusions could not be drawn until
Fig. 13. Phase diagram of the system Na
2
SO
4
CaSO
4
Fig. 14. System Na
2
SO
4
CaSO
4
H
2
Oat75
C [108]
Crystallization and Phase Stability of CaSO
4
711
now since the very tiny needles are difcult to separate from the mother liquor
without decomposition. From combined optical, thermoanalytical and lattice con-
stant determinations Emons et al. [141] conclude that mixed crystals with CaSO
4
:
Na
2
SO
4
ratios between 1:1.5 up to more than 1:>1.7 can be formed. The water
content of the hydrate 1.6Na
2
SO
4
CaSO
4
xH
2
O is given with x¼1.5.
The following minerals are known: eugsterite, 2Na
2
SO
4
CaSO
4
2H
2
O [142]
identical with the ‘‘labile salt’’; glauberite; hydroglauberite, 5Na
2
SO
4
3CaSO
4
6H
2
O [144], also obtained as solid phase in the hexary oceanic salt system at
isothermal equilibrations at 25 and 35
C for a long time [144, 145]; wattevillite,
Na
2
SO
4
CaSO
4
4H
2
O [146], and a hydrate composition of Na
2
SO
4
2CaSO
4
3H
2
O
was discovered [147]. Also, the different mineral compositions indicate a possible
solid solution series. The occurrence of all these minerals in paragenesis with other
oceanic salts like gypsum, bassanite, thenardite, halite, astracanite, and glauberite
makes an exact composition determination very difcult.
The formation of double salts has been investigated by hydration processes of
quenched metastable phases 2Na
2
SO
4
CaSO
4
and 4Na
2
SO
4
CaSO
4
with water
vapour (see above). Hydration of 2Na
2
SO
4
CaSO
4
leads to the formation of
glauberite and thenardite (Na
2
SO
4
V), preceded by the intermediate formation of
‘‘labile salt’’. The phase 4Na
2
SO
4
CaSO
4
reacts to hydroglauberite and thenardite
monitored by gravimetry and X-ray diffraction [148]. Hydration of the known
metastable Na
2
SO
4
CaSO
4
solid solutions in solidgas (water vapour) reactions
as demonstrated in [148] could possibly reveal more details about the hydrate
stoichiometries.
Double Salts with Potassium Sulfate
From the phase diagrams [127, 149153] of the anhydrous system no consistent
picture can be derived about the possible stoichiometries of compounds. The ex-
istence of K
2
SO
4
2CaSO
4
as the only stable double salt is described uniformly but
with disagreements regarding the melting point [127, 149151]. Mukinov et al.
[152] reported a possible phase K
2
SO
4
3CaSO
4
and K
2
SO
4
CaSO
4
, which should
exist below 780
C. Golubeva et al. [153] reported a 2K
2
SO
4
3CaSO
4
compound
with the melting point of 1020
C. To resolve the disagreement Rowe et al. [154]
performed phase equilibrium studies at high temperatures and found the phase
K
2
SO
4
2CaSO
4
(calcium langbeinite) only, which agrees with [127, 149151].
Calcium langbeinite is well characterised. The thermal effect observed at
936
C remains unclear. According to [154] it does not represent a presumed in-
version to a high temperature -K
2
SO
4
2CaSO
4
denoted in respective phase dia-
grams in Refs. [127, 149, 151]. At room temperature K
2
SO
4
2CaSO
4
has an
orthorhombic lattice and transforms at 200
C into the cubic langbeinite structure
(K
2
SO
4
2MgSO
4
¼parent compound) [155]. At high temperatures both phases can
form solid solutions with each other as was shown during the thermal decompo-
sition of polyhalite [156].
Anhydrous potassiumcalcium double sulfates do not crystallize from aqueous
solutions below 200
C. In the system K
2
SO
4
CaSO
4
H
2
O up to nearly 200
C
only the double salts K
2
SO
4
CaSO
4
H
2
O (syngenite) and K
2
SO
4
5CaSO
4
H
2
O
(goergeyite) appear. The concentration range of syngenite along the solubility
712 D. Freyer and W. Voigt
isotherms becomes smaller with increasing temperature and that of goergeyite
enlarges at the same time [123]. Recently, from determinations of the solubility
equilibria in this system up to 200
C it was shown that the maximum temperature
of existence of syngenite in contact with aqueous solutions is limited to about
180190
C. At 200
C and high potassium sulfate concentrations an anhydrous
double sulfate occurs for the rst time. The composition was determined as
K
2
SO
4
CaSO
4
. The monoclinic space group C2=c and Z¼8 formula units
with the lattice parameters a¼7.510(2) A
˚
, b¼21.856(4) A
˚
, c¼9.237(2) A
˚
and
¼113.24
were determined. The double salt is isotypic with the palmierite
(K
2
SO
4
PbSO
4
) and considered as its calcium analog [157]. This again raises the
question of the existence eld of K
2
SO
4
CaSO
4
in the anhydrous system K
2
SO
4
CaSO
4
.
Syngenite and goergeyite type double salts are also known with NH
4
þ
or Rb
þ
instead of K
þ
.AsDAns et al. [136] pointed out variation of the monovalent ion in
the series K
þ
NH
4
þ
Rb
þ
Cs
þ
decreases the upper temperature limit of hydrate
stability in aqueous solutions. Rubidium syngenite exists only below 42
C, for
cesium sulfate no hydrous double salt with CaSO
4
is known. On the other hand
the anhydrous dicalcium salt Cs
2
SO
4
2CaSO
4
of cubic langbeinite type is formed
from aqueous solutions already below 0
C [136]. In Table 4 the comparable double
salts containing alkaline metal and ammonium ions are listed. With Li
2
SO
4
only
Table 4. Calcium sulfate double salts with alkaline metal and ammonium sulfate
Existence conditions of X
2
SO
4
CaSO
4
H
2
O, X
2
SO
4
5CaSO
4
H
2
O and anhydrous phases
X (Syngenite type) (Goergeyite type) X
2
SO
4
CaSO
4
X
2
SO
4
2CaSO
4
Li ––
Na (Na
2
SO
4
5CaSO
4
3H
2
O, from 29
C [136] in
hemihydrate structure) aqueous system, at
520
C thermal
decomposition [131]
K until about 190
C in from 40
C (31.8
C at 200
C in aqueous until 1011
Cin
aqueous system [157], [158]) enlarging system [157] anhydrous system [154]
thermal decomposition stability eld with
at 270
C temperature [123, 157]
NH
4
below 2590
Cin 17110
Cin above 76
C [136] (62
C
aqueous system [136] aqueous system [136] [159]) in aqueous
system
Rb below 042
C [136] ––above 20
C [136], until
1043
C in anhydrous
system [127]
Cs –– below 0
C, with
increasing temperature
more stable [136], until
959
C in anhydrous
system [127]
Crystallization and Phase Stability of CaSO
4
713
one hydrous double salt phase, x Li
2
SO
4
CaSO
4
3H
2
O, is discussed where the
mole number x is varying [160]. This phase crystallizes from aqueous solution
at enhanced temperatures. At 25
C only the single hydrates Li
2
SO
4
H
2
O and
CaSO
4
2H
2
O crystallize in the ternary system Li
2
SO
4
CaSO
4
H
2
O [161].
Double Salts with Alkaline Earth Metal Sulfates
In aqueous systems nothing is known about double salt formation between
CaSO
4
and alkaline earth metal sulfate, whereas Rowe et al. [162] report a
3MgSO
4
CaSO
4
compound in the melting diagram of the system MgSO
4
CaSO
4
.
Smith et al. [163] recently detected a compound 2MgSO
4
CaSO
4
in a ue gas lter
cake and characterised the double salt by IR spectroscopy and X-ray diffraction.
Polyhalite
The mineral polyhalite represents a triple salt K
2
SO
4
MgSO
4
2CaSO
4
2H
2
O with
wide spread occurrence in evaporitic rock salt formations with average contents of
13%. The hydrate water is coordinated at the magnesium ion [164] and therefore
water is lost only when heating above 250
C [156, 165]. The salt can be prepared
by reaction of gypsum with appropriate solutions in the ternary system K
2
SO
4
MgSO
4
H
2
O at temperatures above 70
C [3, 158]. At lower temperatures poly-
halite crystallization becomes slow, at room temperature crystallization was not
observed under laboratory conditions. Thus, limits of solution composition for the
stable existence of polyhalite at low temperatures could only be proved by disso-
lution experiments. The most recent investigation of solubility equilibria in the
hexary oceanic salt system including the stability eld of polyhalite between
35110
C can be found in Refs. [166, 167]. Surprisingly, DAns [158] reported
a much easier crystallization of the analogous triple salts containing Cu and Ni
instead of Mg. He synthesized also the triple salts with the cation combinations K
ZnCa, NH
4
CuCa, and NH
4
CdCa and supposed also the substitution by the
divalent ions Fe
2þ
,Mn
2þ
,Co
2þ
, and monovalent ions Rb
þ
and Cs
þ
. Later the
triple salt with the KCuCa combination was found as a mineral in Chile and was
named Leightonite [168]. Cell dimensions are very similar to polyhalite, however,
there is some confusion about the exact crystal symmetry [169].
Conclusions and Outlook
The chemistry of calcium sulfate is dominated by the phases gypsum,
CaSO
4
2H
2
O, hemihydrate, CaSO
4
0.5H
2
O, and anhydrite, CaSO
4
. The extremely
slow crystallization kinetics of anhydrite in aqueous solutions at temperatures be-
low 70
C prevent reliable establishment of solubility equilibria from supersaturat-
ed solutions. As a consequence uncertainty remains in the anhydrite solubility
curve, which gives rise to the corresponding uncertainty of the gypsumanhydrite
conversion temperature in water with best estimates varying between 4260
C.
The calcium sulfate phases give an instructive example that determination of reli-
able solubility data requires consideration of the substance and phase specic crys-
tallization and transformation kinetics.
714 D. Freyer and W. Voigt
Hemihydrate can be formed by dehydration of gypsum in dry solid state or in
aqueous solutions at enhanced temperatures. The resulting hemihydrates (-, -
hemihydrate) show different thermal and hydration characteristics, which cannot
be traced back to specic structural features of - and -hemihydrate. Despite the
importance of knowledge of the - and -hemihydrate content in industrial gyp-
sum products its quantitative determination relies exclusively on thermal analysis.
Because of the importance for gypsum-based binder and building materials a
number of studies on crystallization and setting kinetics of gypsum have been
performed with emphasis on retarding and accelerating effects of additives. No
general mechanistic model has been developed until now, which can explain most
of the effects observed with respect to kinetics and crystal morphology.
Systematic kinetic investigations on hemihydrate formation in aqueous solu-
tions are missing up to now. Obviously, hemihydrate represents an important in-
termediate phase during transformation of gypsum into anhydrite. Its open one
dimensional channel structure makes ion substitution easier than for gypsum. This
yields solid solutions or nearly stoichiometric compounds like the sodium penta-
salt. Ion substituted hemihydrates seem to play a role in the crystallization mech-
anism of anhydrite or double salts like glauberite, goergeyite, or polyhalite
precipitating from oceanic salt solutions or brines.
In situ techniques such as FT-Raman spectroscopy or time-resolved X-ray dif-
fraction available for the detection of crystallising phases will provide detailed
mechanistic insights in future. Much more emphasis should be given to the crystal
chemistry, phase equilibria, and formation kinetics of related phases discussed in
this review.
Isomorphous ion substitution (e.g. PO
4
3
, Ln
3 þ
) could be exploited as struc-
tural probes in CaSO
4
phases applying techniques like optical spectroscopy or
nuclear magnetic resonance. On the other hand foreign ion distribution studies
between solution and solid sulfate phases would provide the currently lacking
quantitative information for assessment procedures of disposal of inorganic toxic
and nuclear waste in rock salt formations.
References
[1] Knittel E, Phillips W, Williams Q (2001) Phys Chem Min 28: 630
[2] Reisdorf K, Abriel W (1988) Zement-Kalk-Gips 41: 356
[3] DAns J (1933) Die L
oosegleichgewichte der Systeme der Salze ozeanischer Salzablagerungen.
Kali-Forschungs-Anstalt GmbH, Berlin Verlagsgesellschaft f
uur Ackerbau MBH, Berlin SW11
[4] DAns J, Bredtschneider D, Eick H, Freund H-E (1954) Kali u Steinsalz 9:17
[5] Sborgi U, Bianchi C (1940) Gazz Chim Ital 70: 823
[6] Hill AE (1937) J Am Chem Soc 59: 2242
[7] DAns J (1968) Kali Steinsalz 5: 109
[8] Bock E (1961) Can J Chem 39: 1746
[9] Vant Hoff JH, Armstrong EF, Hinrichsen W, Weigert F, Just G (1903) Z phys Chem 45: 257
[10] Posnjak E (1938) Amer J Sci 5(35A): 247
[11] Lange E, Monheim J (1925) Z phys Chem A150: 349
[12] Hardie LA (1967) Amer Min 52: 171
[13] Raju KUG, Atkinson G (1990) J Chem Eng Data 35: 361
[14] Parker VB, Wagman DD, Evans WH (1971) NBS Tech Note No 270:6
Crystallization and Phase Stability of CaSO
4
715
[15] Wagman DD, Evans WH, Parker VB, Schumm RH, Halow I, Bailey SM, Churney KL, Nuttal
RL (1982) J Phys Chem Ref Data 11 Suppl No 2
[16] Knacke O, Gans W (1977) Z Phys Chem 104:41
[17] Barin I, Knacke O, Kubaschewski O (1977) Thermochemical Properties of Inorganic Sub-
stances. In: Springer-Verlag Berlin, Heidelberg, New York
[18] Linke WF, Seidell A (1965) Solubilities of Inorganic and Metal Organic Compounds, 4th Ed,
Vol 1, 2, Amer Chem Soc, Washington DC
[19] Pelsh AD (1973) Handbook of experimental solubility data in multi component water-salt
systems. vol 1 (1973), vol 2 (1975), Publ Khimiya, Leningrad
[20] Marshall WL, Slusher R (1973) J Chem Thermodyn 5: 189
[21] Kalyanaraman R, Yeattts LB, Marshall WL (1973) J Chem Thermodyn 5: 891
[22] Kruchenko VP, Beremzhanov BA (1980) Zh Neorg Khim 25: 3076
[23] Kruchenko VP (1985) Zh Neorg Khim 30: 1566
[24] Freyer D (2000) Zur Phasenbildung und -stabilit
aat im System Na
2
SO
4
CaSO
4
H
2
O. Dis-
sertation, TU Bergakademie Freiberg
[25] Monnin C (1990) Geochim Cosmochim Acta 54: 3265
[26] Monnin C (1999) Chem Geol 153: 187
[27] Krumgalz BS, Starinsky A, Pitzer KS (1999) J Solution Chem 28: 667
[28] Mac Donald GJF (1953) Amer J Sci 251: 884
[29] Kr
uuger R-R, Abriel W (1990) Z Naturforsch 45B: 1221
[30] Lancia A, Musmarra D, Prisciandaro M (1999) AIChE J 45: 390
[31] He S, Oddo JE, Tomson MB (1994) J Coll Interfac Sci 162: 297
[32] Prisciandaro M, Lancia A, Musmarra D (2001) AIChE J 47: 929
[33] Prisciandaro M, Lancia A, Musmarra D (2001) Ind Eng Chem Res 40: 2335
[34] He S, Oddo JE, Tomson MB (1994) J Coll Interfac Sci 163: 372
[35] Hina A, Nancollas GH (2000) In: Alpers ChN, Jambor JL and Nordstrom DK (ed) Reviews in
Mineralogy & Geochemistry; Sulfate Minerals, Crystallography, Geochemistry, and Envi-
ronmental Signicance. Mineralogical Society of America, vol 40. Washington, DC, p 277
[36] De Vreugd CH, Witkamp GJ, van Rosmalen GM (1994) J Cryst Growth 144:70
[37] Liu ST, Nancollas GH (1975) J Colloid Interfac Sci 52: 593
[38] Kontrec J, Kralj D, Brececvic L (2002) J Cryst Growth 240: 203
[39] Boisvert J-P, Domenech M, Foissy A, Persello J, Mutin J-C (2000) J Cryst Growth 220: 579
[40] Bertoldi GA (1978) Zement-Kalk-Gips 31: 626
[41] Konak AR (1976) Krist Tech 11:13
[42] Amathieu L, Boistelle R (1987) Chem -IngTech 59: 858
[43] Badens E, Veesler St, Boistelle R (1999) J Cryst Growth 198=199: 704
[44] Follner S, Wolter A, Preusser A, Indris S, Silber C, Follner H (2002) Cryst Res Technol 37:
1075
[45] Solberg C, Hansen S (2001) Cem Concr Res 31: 641
[46] Conley RF, Bundy WM (1958) Geochim Cosmochim Acta 15:57
[47] Kushnir J (1980) Geochim Cosmochim Acta 44: 1471
[48] Witkamp GJ, Rosmalen GM (1991) J Cryst Growth 108:89
[49] Curry NA, Jones DW (1971) J Chem Soc A: 3725
[50] Aslanian S, Stoilova D, Petrova R (1980) Z anorg allg Chem 465: 209
[51] Sakae T, Nagata H, Sudo T (1978) Am Mineral 63: 520
[52] Xu J, Butler IS, Gilson DFR (1999) Spectrochim Acta A55: 2801
[53] Rinaudo C, Lanfranco AM, Boistelle R (1996) J Cryst Growth 158: 316
[54] Hina A, Nancollas GH, Grynpas M (2001) J Cryst Growth 223: 213
[55] Zdukos AT, Vaimakis TK (1985) Russ J Inorg Chem 30: 1124
[56] Haerter M (1971) Tonind-Ztg 95:9
[57] Ferraris G, Jones DW, Yerkess I (1971) Acta Cryst B27: 349
716 D. Freyer and W. Voigt
[58] Kr
uuger R-R, Abriel W (1991) Acta Cryst C47: 1958
[59] Selivanova NM, Shneider VA (1959) Izv Vyssh Ucheb Zav Khim Khim Tekhnol 2: 651
[60] Selivanova NM, Shneider VA (1958) Nauch Dokl Vyssh Shk Khim Khim Tekhnol 664
[61] Kohlmann M, Sowa H, Reithmayer K, Schulz H, Kr
uuger R-R, Abriel W (1994) Acta Cryst
C50: 1651
[62] Ben Amor M, Louer M, Le Marouille JY (1982) CR Seances Acad Sci 2: 294
[63] Gmelins Handbuch der Anorganischen Chemie. (1962) 8 Au, vol 52, Verlag Chemie GmbH,
Weinheim
[64] Cooper MA, Hawthorne FC (1996) Can Mineral 34:99
[65] Tiemann H, S
ootje I, Jarms G, Paulmann C, Epple M, Hasse B (2002) J Chem Soc, Dalton
Trans, 1266
[66] Fl
oorke OW (1952) Neues Jb Mineral Abh 4: 189
[67] Eipeltauer E (1958) Tonind-Ztg 6: 264
[68] McAdie HG (1965) Can J Chem 42: 792
[69] Kuntze RA (1965) Can J Chem 43: 2522
[70] Murat M, Comel C (1971) Tonind-Ztg 95:29
[71] Wiedemann HG (1975) Z Anal Chem 276:21
[72] Hummel H-U, Abdussaljamow B, Fischer H-B, Stark J (2001) Zement-Kalk-Gips 54: 272
[73] Heide K (1969) Silikattech 20: 232
[74] Kurpiers K (1970) Untersuchungen der Entw
aasserung von Gips bei niedrigen Wasserdampf-
partialdr
uucken. Dissertation TU Clausthal 1970
[75] Oetzel M, Heger G, Koslowski T (2000) Zement-Kalk-Gips 53: 354
[76] Bobrov BS, Romaschkov AB, Andreva EP (1987) Zh Neorg Khim 23: 497
[77] Schwotzer M, Weidler PG, N
uuesch R (2002) 3. Marburger Gipstagung Phillips-Universit
aat
Marburg
[78] Pritzel C, Trettin R (2002) 3. Marburger Gipstagung Phillips-Universit
aat Marburg
[79] Dorozhkin SV (1996) Scanning 18: 119
[80] Koslowski Th, Ludwig U (1999) Zement-Kalk-Gips 5: 274
[81] Goto M (1966) Aust J Chem 19: 313
[82] Powell DA (1958) Nature 182: 792
[83] Morris RJ (1963) Nature 198: 1298
[84] Tpiollier M, Guilhot B (1976) Cement and Concrete Res 6: 507
[85] Galtier P, Soustelle M, Guilhot B (1983) Cement and Concrete Res 13: 703
[86] Kuzel H-J (1987) N Jb Min, Abh 156: 155
[87] Robert J, Morris J (1963) Anal Chem 35: 1489
[88] Wiegel E, Kirchner HH (1966) Ber Dtsch Keram Ges 43: 718
[89] Lager GA, Armbruster Th, Rotella FJ, Jorgensen JD, Hinks DG (1984) Am Min 69: 910
[90] Bensted, Prakash S (1968) Nature 219:60
[91] Cliffton JR (1971) Phys Sci 232: 125
[92] Lehmann H, Rieke K (1973) Tonind-Ztg 97: 157
[93] Lehmann H, Rieke K (1974) Tonind-Ztg 98:81
[94] Reisdorf K, Abriel W (1987) N Jb Min, Abh 157:44
[95] Budnikov PO, Kosyreva ZS (1953) Voprosy Petrograf i Minera Akad Nauk SSSR 2: 342
[96] Eipeltauer E (1956) Zement-Kalk-Gips 9: 501
[97] Southard JC (1940) Ind Eng Chem 32: 442
[98] Kuzel H-J, Hauner M (1987) Zement-Kalk-Gips 12: 628
[99] Bushuev NN, Borisov VM (1982) Russ J Inorg Chem 27: 341
[100] Schneider J, Freyer D, Voigt W (2002) 3. Marburger Gipstagung Phillips-Universit
aat Marburg
[101] Bushuev NN (1982) Zh Neorg Khim 27: 610
[102] Abriel W, Nesper R (1993) Z Krist 205:99
[103] Abriel W (1983) Acta Cryst C36: 956
Crystallization and Phase Stability of CaSO
4
717
[104] Bezou C, Nonat A, Mutin J-C, Christensen AN, Lehmann MS (1995) J Solid State Chem 117:
165
[105] Frik M, Kuzel H-J (1982) Fortschr Miner 60:79
[106] Gallitelli P (1933) Periodico Mineral Roma 4:1
[107] Ballirano P, Maras A, Meloni S, Caminiti R (2001) Eur J Mineral 13: 985
[108] Hill AE, Will JH (1938) J Amer Chem Soc 60: 1647
[109] Freyer D, Reck G, Bremer M, Voigt W (1999) Monatsh Chem 130: 1179
[110] Powell DA (1962) Austr J Chem 15: 868
[111] Sugimoto K (1958) Asahi Garasu Kenkyu Hokoku 8:32
[112] Autenrieth H (1958) Kali u Steinsalz 2: 181
[113] Gudowius E, von Hodenberg R (1979) Kali u Steinsalz 7: 501
[114] Freyer D, Ziske S, Voigt W (2002) Freib Forschh E3: 127
[115] Takahashi S, Seki M, Setoyama K (1993) Bull Chem Soc Jpn 66: 2219
[116] Bushuev N, Nabiev AG (1988) Zh Neorg Khim 33: 2962
[117] Bushuev NN, Nabiev AG, Petropavlovskii IA, Smirnova IS (1988) Zh Neorg Khim 61: 2153
[118a] Blackburn AV, Gerkin RE (1995) Acta Cryst. C51: 2215
[118b] Blackburn AV, Gerkin RE (1994) Acta Cryst. C50: 835
[119] Melikhov IV, Berdonosova DG, Fadeev VV, Burlakova EV (1991) Zh Prikl Khim 64: 334
[120] Lyer PN, Natarajan PR (1989) J Less-Common Met 146: 161
[121] Kolcu O, Zumreoglu-Karan B (1994) Thermochim Acta 240: 185
[122] Koopman C, Witkamp GJ (2000) Hydrometal 58:51
[123] Hill AE (1934) J Am Chem Soc 56: 1071
[124] Ostroff AG (1964) Geochim Cosmochim Acta 28: 1363
[125] Zen E-A (1965) J Petrol 6: 124
[126] Calcagni G, Mancini G (1910) Atti Linc 19 II: 426
[127] M
uuller H (1910) N Jb Min, Beilagebd 30:1
[128] Komissarova LN, Plyushev VE, Stepina SB (1955) Tr Mosk Inst Tonkoi Khim Tekhnol 5:3
[129] Speranskaja EI, Baraskaja IB (1961) Zh Neorg Khim 6: 1392
[130] Bandaranayake PWSK, Mellander B-E (1988) Solid State Ionics 26:33
[131] Freyer D, Voigt W, K
oohnke K (1998) Eur J Solid State Inorg Chem 35: 595
[132] vant Hoff JH (1905) Ber Berl Akad 478
[133] Barre M (1911) Ann Chim Phys 24: 162
[134] Rogosowskaya MS, Konontschuk TI, Lukjanowa NK (1980) Zh Neorg Khim 25: 1095
[135] Fritzsche J (1857) J prakt Chem 72: 291
[136] DAns J, Schreiner O (1909) Z anorg allg Chem 62: 129
[137] Vasilevskaya AG (1959) Izvest Sib Otd Akad Nauk SSSR 1: 76; CA 53 (1959) 71687
[138] Druzhinin S, Lopina-Shendrik MD (1962) Izv Akad Nauk Kirg SSR Ser Estestr i Tekknol
Nauk 4: 61; CA 59 (1963) 400552
[139] Fridman YD, Zinovev AA, Bogdanovskaya RZ (1953) Tr Inst Khim Kirgiz Filial Akad Nauk
SSSR 5: 49; CA 50 (1956) 87989
[140] Lopina-Shendrik MD (1958) Tr Molodykh Nauchn Rabot Akad Nauk Kirg SSR: 43; CA 55
(1961) 50869
[141] Emons H-H, Seyfarth H-H, Stegmann E (1971) Krist u Tech 6:85
[142] Vergouen L (1981) Am Min 66: 632
[143] Sljusareva MN (1969) Zap vses mineralog Obsc 98:59
[144] Gu S, Lin H (1985) Kexue Tongbao 30: 1375
[145] Gu S, Lin H (1986) Kexue Tongbao 31: 624
[146] Fejer E, Cressey G (1988) British Museum, London, England, UK, JSPDS Grant-in-Aid,
Report
[147] Hodenberg R, Miotke F-D (1983) Kali Steinsalz 8: 374
[148] Freyer D, Fischer St, K
oohnke K, Voigt W (1997) Solid State Ionics 96:29
718 D. Freyer and W. Voigt
[149] Graham W (1913) Z Anorg Chem 81: 257
[150] J
aanecke E, M
uuhlh
aauser W (1936) Z anorg allgem Chem 228: 241
[151] Bellanca A (1942) Periodico Mineral (Rome) 13:21
[152] Mukinov SM, Krylova NI, Bergman AG (1949) Tr Inst Khim Akad Nauk Uz SSR Inst Khim
Obshch i Neorg Khim 2:94
[153] Golubeva MS, Bergman AG (1956) Zh Obshch Khim 26: 328
[154] Rowe JJ, Morey GW, Hansen ID (1965) J Inorg Nucl Chem 27:53
[155] Morey GW, Rowe JJ, Fournier RO (1964) J Inorg Nucl Chem 26:53
[156] Fischer St, Voigt W, K
oohnke K (1996) Cryst Res Technol 31:87
[157] Freyer D, Voigt W (2003) accepted for publication in Cosmochim Geochim Acta
[158] DAns J (1908) Chem Ber 41: 1776
[159] Hill AE, Yanick NS (1935) J Am Chem Soc 57: 645
[160] Petrova MI, Alymkulova KS, Dzhashakueva BK, Kydynov MK (1981) Deposited Doc,
VINITI: 2357, USSR; CA 97 (1982) 616657
[161] Kydynov MK, Petrova MI (1965) Zh Prikl Khim 38: 2590
[162] Rowe JJ, Morey GW, Silber CC (1967) J Inorg Nucl Chem 29: 925
[163] Smith DH, Seshadri KS, (1999) Spectrochimica Acta A(55): 795
[164] Schlatti M, Sahl K, Zemann A, Zemann J (1970) Tschermaks Mineralog Petrogr Mitt 14:75
[165] Jockwer N (1981) Kali Steinsalz 8: 126
[166] Kropp E, Beate R, Grosch Ch, Kranz M, Holldorf H (1988) Freib Forschh A764:42
[167] Kropp E, Holldorf H (1988) Freib Forschh A764:67
[168] Palache Ch (1938) Am Mineral 23:34
[169] Hawthorne FC, Krivovichev SV, Burns PC (2000) In: Alpers ChN, Jambor JL, Nordstrom DK
(eds) Reviews in Mineralogy & Geochemistry; Sulfate Minerals, Crystallography,
Geochemistry, and Environmental Signicance. Mineralogical Society of America, vol 40
Washington, DC, p 1
Crystallization and Phase Stability of CaSO
4
719
  • ... Salt reaction products may also decompose as temperature is raised (Table 4). Hydrated minerals such as sulfate phases that contain structural H 2 O commonly dehydrate in a step-wise manner (e.g., King and McSween 2005;Hyde et al. 2011;Freyer and Voigt 2003). For example, when subjected to increasing temperature (or decreasing relative humidity), MgSO 4 ·nH 2 O species may dehydrate to form a series of variably hydrated decomposition products (Chipera and Vaniman 2007;Steiger et al. 2011), before finally producing MgSO 4 that at high decomposes to MgO and SO 3 (Stern and Weise 1966;Rowe et al. 1967;Du 2000;Stern 2001). ...
  • ... Salt reaction products may also decompose as temperature is raised (Table 4). Hydrated minerals such as sulfate phases that contain structural H 2 O commonly dehydrate in a step-wise manner (e.g., King and McSween 2005;Hyde et al. 2011;Freyer and Voigt 2003). For example, when subjected to increasing temperature (or decreasing relative humidity), MgSO 4 ·nH 2 O species may dehydrate to form a series of variably hydrated decomposition products (Chipera and Vaniman 2007;Steiger et al. 2011), before finally producing MgSO 4 that at high decomposes to MgO and SO 3 (Stern and Weise 1966;Rowe et al. 1967;Du 2000;Stern 2001). ...
    Article
    High temperature gas–solid reactions interactions are ubiquitous in and on the Earth and other planetary bodies (Fig. 1). In these natural systems, gases are elusive. Like a hurricane’s wind that leaves a path of destruction but no trace of wind after the event, the gas that initiated a gas–solid reaction may be conspicuously absent. Because gases effectively escape from geologic systems, gas–solid reactions may only be recorded by their stable, or metastable, solid reaction products. In this chapter, we seek to unravel the evidence and approaches that can be used to examine gas–solid reactions, especially through their solid reaction products. We define the major types of gas–solid reactions, paying particular attention to surface-mediated reactions. Factors that influence gas–solid reactions are presented followed by models for transfer of mass, energy and momentum. We then examine case studies in which different experimental, analytical and modeling approaches have been used to investigate gas–solid reactions. We highlight the evidence for these reactions such as unusual phase assemblages including salts (e.g., sulfates and chlorides), chemical anomalies and textural features.
  • ... Previous studies have provided important insights into the geochemical processes affecting the CaSO 4 − H 2 O system, for example; by determining stability fields of anhydrite and gypsum under various pressure, tem- perature, and concentration conditions (e.g., Blount & Dickson, 1973;Dai et al., 2017;Freyer & Voigt, 2003;MacDonald, 1953;Marsal, 1952;Partridge & White, 1929), which can be used to characterize the critical conditions that may lead to sulfate dissolution or precipitation. Under low-pressure and low-temperature conditions, which normally prevail in Gypsum Keuper formations relevant to geothermal applications and tunneling, the solubility of anhydrite is higher than that of gypsum. ...
    Article
    Swelling of clay-sulfate rocks often causes large problems in geotechnical applications such as tunneling. The primary mechanism inducing the increase in rock volume is the chemical transformation of anhydrite to gypsum, which is triggered by the ingress of groundwater. In the present study, a novel conceptual and numerical modeling approach is developed that emphasizes the effect of groundwater flow in conjunction with the associated availability of water and changing geochemical conditions on the chemical transformation of anhydrite to gypsum. A reactive transport model was developed and hydraulic, reactive, and solute transport as well as mechanical model parameters were estimated through an inversion process, constrained by geodetic ground heave measurements from a study site in Staufen, Germany. The conceptual model of the swelling process was implemented numerically through a dual-domain modeling approach, whereby the mobile domain accounts for solute transport along discontinuities, and the immobile reactive domain represents the matrix. A mass transfer process accounts for diffusive and/or capillary water transport into the matrix, where the rate-limited transformation of anhydrite to gypsum takes place. The model calculates heave at the land surface depending on water inflow, the transformation of anhydrite into gypsum and the local stress conditions exerted by overburden pressure. The results show that the proposed reactive transport modeling approach is suitable to quantify the observed swelling-induced heave at the study site with a plausible parameterization. The study also highlights that diffusion is a decisive factor for the effective rate of anhydrite dissolution and, therefore, the overall chemical transformation process.
  • ... [47], the system CaSO 4 -H 2 O has been examined by several authors both experimentally and theoretically based on the thermodynamic concept of Gibbs free energy (see [48], for a review). Nonetheless, most of the studies are outdated and many of them fail to provide all the necessary information. ...
  • Technical Report
    Full-text available
    Tunnels through geological formations containing anhydrite and swelling clays are very difficult to maintain due to constant floor heave from the combined action of the swelling clay and the transformation of anhydrite to gypsum (ATG). This can even occur during the construction phase, and requires expensive, repeated tunnel closures for maintenance. Normally, controlling this problem is carried out by geotechnical solutions. In this research, carried out by Prof. Robert J. Flatt at ETH Zurich, and detailed in this report FGU2012/001, from the Tunnel Research Working Group, a potential material solution is investigated, and a useful experiment for better decoupling the two phenomena at the heart of the floor heave is illustrated. This experiment uses a “thermodynamic switch” to decouple the swelling processes at the lab scale by using temperature to effectively shut off the ATG reaction. The implications of this study show the potential of a method for better studying these phenomena at the lab scale, which is often a very long-term and costly project. From these results, both clay swelling inhibitors and ATG inhibitors could be assessed. Results demonstrated that clay swelling inhibitors showed almost no effect at inhibiting early stage swelling, however, for long term clay swelling inhibition they could still show potential. On the other hand, the ATG inhibitors, well known from the gypsum board industry, showed great success, and their potential to be used in drilling fluids is highlighted. Their mechanism of action was also investigated and determined to be primarily dissolution inhibition of anhydrite, although nucleation and growth inhibition could be at play, depending on the inhibitor. These results could aid in further investigations on how to control ATG. A final larger scale combined experimental and computational study was carried out on a model swelling rock to give some insight on how to carry out an on site borehole study, with the most important conclusion being that it is essential to carry out proper characterization of the tunnel invert beforehand, and to consider the reaction-diffusion front of any inhibitor applied, in order to best optimize borehole spacing.
  • Article
    Here we reported a method to simultaneously control the particle size and morphology of α‐CaSO4·½H2O (α‐HH) prepared from flue gas desulfurization gypsum by adjusting the succinic acid concentration and glycerol content under mild conditions. Succinic acid controlled the crystal morphology by adsorption onto α‐HH surfaces, and glycerol controlled the crystal particle size, in which an increase in the maximal relative supersaturation (S max) and nucleation rate of α‐HH was hypothesized to cause the change in α‐HH particle size. Then, based on the method, α‐HH with different particle sizes but with almost the same morphology was prepared, and the influence of the crystal particle size on the mechanical strength of the α‐HH pastes was explored. With decreasing α‐HH particle size from about 26 to 5 μm, the dry compressive strength of the pastes made from the α‐HH decreased remarkably from 68.02 to 34.85 MPa, which was ascribed to an increase in the internal porosity of the pastes. This article is protected by copyright. All rights reserved.
  • Article
    Full-text available
    Fly ash usually contains a considerable amount of toxic elements that can be leached into the environment, thereby easily leading to serious contaminations. In this work, the leaching behaviors of poisonous elements including boron (B), phosphorus (P), vanadium (V), chromium (Cr), arsenic (As), selenium (Se), molybdenum (Mo), antimony (Sb), and tungsten (W) from fly ash were explored by sequential extraction. Importantly, the associations of these elements in fly ash were discussed based on their leaching and X-ray absorption near-edge structure (XANES) results. From the XANES results, it was observed that V(IV), Cr(III), As(V), Se(IV), and W(IV) were their main states of existence in fly ash. In terms of leaching results, large amounts of Mo and W were leached into pure water, which indicated their high mobilities. Furthermore, the occurrence of Mo in fly ash was mainly in the form of oxides, and W had complex associations including WX4 (X can be monovalent anions), its reduction state or association with the elements that can be oxidized, and existence in silicates. B was as easily released into the environment as Mo and W. It can have several associations with the other cations, such as Ca²⁺, Na⁺, and Mg²⁺, and occurs in silicates. In contrast, most of the Cr and Sb were locked in silicates, indicating that they were very stable in fly ash. In addition, P, V, and As can exist within the structure of silicates as well. However, a considerable amount of them leached in the reduction step with a low pH. Hence, they can be associated with Ca²⁺, Na⁺, Mg²⁺, or Fe³⁺. In terms of Se, oxidation processes played an important role in controlling its leaching because of the oxidation of Se(IV) to Se(VI). Calcium selenite should be the predominant form of Se in fly ash.
  • Article
    Within the last two decades, a large number of articles were published that report, as the authors claim, ‘metastable phase equilibria’, or ‘metastable solubilities’. The main objective of these studies was to carry out experiments under conditions closely meeting those in solar evaporation ponds or industrial evaporation–crystallization processes. Occasionally, such studies were subject to controversial discussion and criticism and the question was raised whether metastable equilibrium data are worth being published at all. This paper provides a critical discussion of such evaporation experiments. The thermodynamic background of stable and metastable solubility and typical experimental difficulties in solubility determinations are discussed in detail. We also demonstrate that the knowledge of metastable equilibria is very useful in different research areas such as geochemistry or industrial application of solubility equilibria. Finally, it is shown that so-called ‘isothermal evaporation method’ used in the above-mentioned studies does not yield stable or metastable solubilities.