Vous êtes sur la page 1sur 10

American Mineralogist, Volume 86, pages 438–447, 2001

Thermodynamics of ion-exchanged and natural clinoptilolite

SANYUAN YANG,1,* ALEXANDRA NAVROTSKY,1 AND RICK WILKIN2


1
Thermochemistry Facility, Department of Chemical Engineering and Materials Science, University of California at Davis, Davis, California
95616, U.S.A.
2
U.S. EPA, National Risk Management Research Laboratory, P.O. Box 1198, Ada, Oklahoma 74821, U.S.A.

ABSTRACT
Natural clinoptilolite (Cpt: Na0.085K0.037Ca0.010Mg0.020Al0.182Si0.818O2 ·0.528H2O) from Castle Creek,
Idaho, and its cation-exchanged variants (Na-Cpt, NaK-Cpt, K-Cpt, and Ca-Cpt) were studied by
high-temperature calorimetry. The hydration enthalpy for all the clinoptilolites is about –30 kJ/mol
H2O (liquid water reference state) at 25 °C. The energetic stabilization effect of hydration on each
clinoptilolite can be largely correlated to its hydration capacity. The higher the average ionic poten-
tial of the extra-framework cations, the larger the hydration capacity of the clinoptilolite. This trend
may be attributed to the small size as well as the efficient water-cation packing of high field strength
cations in the zeolite structure. The hydration properties of these clinoptilolites are compared with
those previously reported in the literature. The dehydration conditions as well as the measurement
direction (dehydration of the initially hydrated sample or rehydration of the dehydrated zeolites) are
important factors to control to obtain consistent thermodynamic properties for hydration.
The standard enthalpy for formation of the clinoptilolites from the constituent elements at 25 °C
based on two framework O atoms was obtained from the calorimetric data: –1117.57 ± 0.95 kJ/mol
Cpt, –1130.05 ± 1.00 kJ/mol Na-Cpt, –1109.49 ± 1.04 kJ/mol NaK-Cpt, –1094.21 ± 1.12 kJ/mol K-
Cpt, and –1153.78 ± 1.07 kJ/mol Ca-Cpt. Their molar entropy was determined by a summation
method based on the thermodynamic properties of the component oxides. Thus the standard free
energy based on two framework O atoms was derived: –1034.01 ± 1.05 kJ/mol Cpt, –1044.19 ± 1.10
kJ/mol Na-Cpt, –1027.26 ± 1.13 kJ/mol NaK-Cpt, –1014.89 ± 1.21 kJ/mol K-Cpt, and –1064.95 ±
1.16 kJ/mol Ca-Cpt.

INTRODUCTION channels: one 8-membered channel and one 10-membered chan-


Heulandite group zeolites, including mineral clinoptilolite nel are parallel with the c-axis, the other 8-membered channel
and heulandite and their synthetic analogues, have the same is parallel with the a-axis. Although HEU-type zeolites have
framework topology (structure code HEU). According to been difficult to synthesize (Chi and Sand 1983; Satokawa and
Coombs et al. (1997), clinoptilolite is distinguished from heu- Itabashi 1997; Zhao et al. 1998), they are the most abundant
landite based on Si/Al ratio, i.e., clinoptilolite if Si/Al > 4 and zeolites in nature (Gottardi and Galli 1985).
heulandite if Si/Al < 4. It is well known that the thermal stabil- The Yucca Mountain site in southeast Nevada is being de-
ity of heulandite group zeolites is strongly influenced by ex- veloped as a geological repository for high-level radioactive
tra-framework cations (Mason and Sand 1960; Mumpton 1960; waste (Kerr 2000). Primary volcanic glass at Yucca Mountain
Boles 1972; Zhao et al. 1998). This property warrants the heu- is partially altered to clinoptilolite, mordenite, analcime, and a
landite group zeolites to be a preferred system for investiga- variety of other zeolites. Clinoptilolite is the most abundant
tion of the cationic effects on thermodynamic (in contrast to species among these zeolites. Over the past decade, numerous
thermal) stability. studies concern evaluation of the thermodynamic properties of
The framework structure of various HEU type species and clinoptilolites and their implications for the suitability of the
the site distributions of the exchangeable cations and water Yucca Mountain site as a nuclear waste repository (e.g., Johnson
molecules have been extensively studied by X-ray and neutron et al. 1991; Glassley et al. 1994; Murphy et al. 1996; Carey
diffraction techniques (e.g., Mason and Sand 1960; Mumpton and Bish 1996, 1997; Chipera and Bish 1997; Wilkin and Barnes
1960; Alietti 1972; Alberti 1975; Koyama and Takéuchi 1977; 1998, 1999; Pabalan and Bertetti 1999; Benning et al. 2000).
Hambley and Taylor 1984; Armbruster 1993; Yang and The role of clinoptilolite as a barrier to radionuclide migration
Armbruster 1996; Yang et al. 1997). There are three types of is an important factor on these considerations.
HEU type zeolites have also been used in many other appli-
cations such as catalysis, soil amendment, formulation of
* E-mail: syyang@ucdavis.edu nonphosphate detergents, contaminant remediation, and nuclear
0003-004X/01/0004–438$05.00 438
YANG ET AL.: THERMODYNAMICS OF CLINOPTILOLITE 439

waste treatment. To understand chemical processes involving TABLE 1. Molar formula of clinoptilolites based on two framework
clinoptilolites in both nature and applied technologies, their O atoms
thermodynamic data are indispensable. The hydration proper- Zeolite Molar formula on TO2 MW (g)
ties of several clinoptilolites have been previously investigated Cpt Na0.085K0.037Ca0.010Mg0.020Al0.182Si0.818O2·0.528H2O 73.7
Na-Cpt Na0.182Al0.182Si0.818O2·0.572H2O 74.4
using HF solution calorimetry (Johnson et al. 1991), immer- NaK-Cpt Na0.097K0.085Al0.182Si0.818O2·0.495H2O 74.4
sion calorimetry (Barrer and Cram 1971; Petrova and Kirov K-Cpt K0.182Al0.182Si0.818O2·0.433H2O 74.8
Ca-Cpt Ca0.084Mg0.008Al0.182Si0.818O2·0.646H2O 75.0
1995; Carey and Bish 1997), and equilibrium measurements
(Carey and Bish 1996; Wilkin and Barnes 1999). There are
considerable disagreements among the reported enthalpies of comparable to the standard data of the constituent oxides and
hydration, which cannot be explained only by the composi- zeolites of different structures.
tional differences of the samples used for the measurements. Ion-exchanged clinoptilolites are denoted as Na-Cpt, NaK-
Most reported thermodynamic properties, such as enthalpy and Cpt, K-Cpt, and Ca-Cpt to reflect their major cations. A pre-
free energy of formation, were measured for a few individual requisite for accurate calorimetric studies is that the water
clinoptilolites rather than for a series showing systematic com- content of each zeolite sample must be maintained at a con-
positional variation (e.g., Hemingway and Robie 1984; Johnson stant and reproducible level. For this reason, the (de)hydration
et al. 1991; Murphy et al. 1996; Wilkin and Barnes 1998; conditions of the zeolite samples and the handling procedure
Benning et al. 2000). At present, it is difficult to parameterize are described in detail.
the thermodynamic properties of clinoptilolite as a function of The hydrated samples are referred to their equilibrium states
composition (Si/Al ratio, extra-framework cation and water with an atmosphere of 50% relative humidity at room tempera-
content) because of the lack of consistent experimental data. ture (25 °C). Accordingly our zeolite samples were placed in a
This investigation focuses on the effect of the exchange- desiccator containing a saturated solution of Ca(NO3)2 for at
able cations on the thermodynamic properties of clinoptilolites. least a week before use. The same condition had been applied
Our objectives are: (1) high-temperature calorimetric measure- to the hydrated clinoptilolites before their water contents were
ments of the energetics of hydration and formation of some determined (Wilkin and Barnes 1998). Our calorimetry labo-
selected clinoptilolites; (2) comparison of the new results with ratory is controlled to have a relative humidity of 50 ± 3% and
those reported in the literature for similar clinoptilolites and a temperature of 25 ± 1 °C (Yang and Navrotsky 2000). The
analysis of the major factors influencing the experimental re- procedure for taking powder sample out of desiccator, making
sults; and (3) identification of trends and systematics of the pellets and weighing them for calorimetric measurements is
thermodynamic properties of clinoptilolite as a function of the always performed in the same controlled area and takes less
exchangeable cations. We studied natural clinoptilolite (Cpt) than 10 minutes.
and its four ion-exchanged variants (Na-Cpt, NaK-Cpt, K-Cpt, The dehydrated samples are prepared as follows. The hy-
and Ca-Cpt). The thermodynamic properties of the same Na- drated zeolite sample is first treated at 400 °C for 9 hours fol-
Cpt, K-Cpt, and Ca-Cpt materials have been previously stud- lowed by 500 °C for 3 hours. Because Ca-Cpt almost completely
ied by solubility measurements (Wilkin and Barnes 1998; loses its XRD crystallinity after such treatment (see Results),
Benning et al. 2000) and hydration titration measurements dehydrated Ca-Cpt was also prepared under less severe condi-
(Wilkin and Barnes 1999), allowing direct comparison of the tions. The dehydrated sample is designated as Ca-CptA after
results obtained from different methods. treatment at 300 °C for 15 hours followed by additional treat-
ment at 400 °C for 3 hours, and Ca-CptB after treatment at 300
EXPERIMENTAL METHODS °C for 7 hours followed by additional treatment at 350 °C for 1
hour and at 400 °C for 1 hour. Dehydration is apparently com-
Clinoptilolite samples: origin, (de)hydration conditions, plete because the mass loss under prolonged heating (500 °C)
and handling procedures or at higher temperature (550 °C) is negligible. However, about
Crystals of natural clinoptilolite (Table 1) denoted as Cpt 1.5% of residual water content is found in Ca-CptB dehydrated
are standard reference material from Castle Creek, Idaho. The under less severe conditions. The residual water content of the
purification and ion exchange procedures and the chemical dehydrated samples is determined as the additional weight loss
analyses were reported elsewhere (Wilkin and Barnes 1998). after using the handling procedure for anhydrous sample (see
Most of the 5% impurities present in the collected sample were below). This procedure is consistent with that for determina-
removed by the repeated purification treatments (Wilkin and tion of the water content of the hydrated samples (Wilkin and
Barnes 1998). The clinoptilolite samples are regarded as pure Barnes 1998).
species here. The chemical formulas of the ion-exchanged and We used specially modified deep Pyrex vials (Yang and
natural clinoptilolites are based on two framework O atoms Navrotsky 2000). The upper inner wall was coated with a layer
(Table 1) for two reasons. First, the thermodynamic data of of paste containing a mixture of NaY zeolite and kaolin clay,
heulandite group zeolites reported in literature are based on 18 to prevent the dehydrated pellets from rehydrating during
(Johnson et al. 1985), 36 (Johnson et al. 1991), or 72 sample handling. Zeolite pellets and their vial containers are
(Hemingway and Robie 1984) oxygen formulas, it is more con- heated together for dehydration. Immediately after heating, the
venient to convert and compare the data based on the same vials containing the zeolite pellets are placed in a desiccator
number of framework oxygen than on widely different num- (desiccant: freshly generated commercial zeolite-type beads).
bers. Second, data based on two oxygen formulas are more Next the desiccator is transferred into a glove-box (moisture
440 YANG ET AL.: THERMODYNAMICS OF CLINOPTILOLITE

< 1 ppm) where the samples are handled, including weighing


and capping the vials. Less than 1 minute elapses between the
time the zeolite sample is removed from the glove-box and the
time it is dropped into the calorimeter. During this period, the
dehydrated zeolite is kept in dry Ar atmosphere in a capped
vial. Air exposure occurs only when the pellet leaves the dry
Ar atmosphere and enters the top end of the dropping tube (~0.1
second or ~5 cm path in free fall). Furthermore, because the
flushing gas (also dry Ar) passing through the calorimeter mixes
with air at the top end of the dropping tube, the dehydrated
sample is actually exposed to an atmosphere drier than ambi-
ent air during this 0.1-second exposure. Our experience indi-
cates that handling highly hygroscopic materials, such as
dehydrated zeolites, is a delicate process and every step must
be carefully controlled to obtain consist and reliable results.

Calorimetry
A custom-built Tian-Calvet twin high temperature micro-
calorimeter operating at 700 °C was used (Navrotsky 1997).
About 30 g of 2PbO·B2O3 solvent was used for sequential drops
of 5–10 pellets, 15–20 mg each, of zeolite. The calorimeter
was calibrated by measurement of comparable mass corundum
pellets dropped into an empty crucible. The experimental pro-
cedure is similar to that used previously (Yang and Navrotsky
2000).
The enthalpy of drop solution of a hydrated zeolite, ∆HDS-hyd, FIGURE 1. Powder XRD patterns of the clinoptilolite samples (Cpt,
can be expressed by the following chemical process, Na-Cpt, NaK-Cpt, and K-Cpt) before and after their dehydration
treatment.
∆HDS-hyd MaM'bAlxSiyO2·nH2O (crystalline, 25 °C) →
0.5aM2O (soln, 700 °C) + bM'O (soln, 700 °C) +
0.5xAl2O3 (soln, 700 °C) + ySiO2 (soln, 700 °C) +
nH2O (g, 700 °C) (1)

where MaM'bAlxSiyO2·nH2O is the composition of the hydrated


zeolite on a TO2 basis [one mole of (Si + Al)], M stands for
monovalent cations such as Na and K, and M’ for divalent cat-
ions such as Ca and Mg.
When the measurement is performed on a dehydrated
sample, ∆HDS-deh, can be expressed by the following chemical
process,

∆HDS-deh MaM'bAlxSiiyO2 (dehydrated phase, 25 °C) →


0.5aM2O (soln, 700 °C) + bM'O (soln, 700 °C) +
0.5xAl2O3 (soln, 700 °C) + ySiO2 (soln, 700 °C) (2)

RESULTS

Characterization of the exchanged clinoptilolite samples


Before and after dehydration, the framework integrity of
the zeolite samples was examined by powder X-ray diffraction
(XRD). The crystallinity established by XRD is mostly retained
after the standard dehydration treatment for all the clinoptilolites
except Ca-Cpt, which becomes a nearly amorphous phase (Figs.
1 and 2). Therefore the dehydrated samples for all the compo-
sitions but Ca-Cpt are considered to be anhydrous clinoptilolite
in their respective cationic forms. The dehydrated sample of FIGURE 2. Powder XRD patterns of the Ca-Cpt before and after
Ca-Cpt is considered as an anhydrous amorphous phase. Calo- dehydration treatments under three different conditions. An asterick
rimetric measurement was nonetheless performed for this (*) marks an impurity phase obtained during dehydration.
YANG ET AL.: THERMODYNAMICS OF CLINOPTILOLITE 441

sample because of the inaccessibility of a completely dehy- The space occupancy of water molecules and cations in the
drated Ca-form of clinoptilolite. Because the energetic differ- fully hydrated clinoptilolites can be estimated by calculation
ence between Na-aluminosilicate framework structures (Yang and Navrotsky 2000). The overall volume occupied by
(Petrovic and Navrotsky 1997) and glass of the same composi- the extra-framework species (i.e., water molecules and cations)
tion (Navrotsky et al. 1980, 1982) is quite small (often less decreases as the average cation size increases and/or the total
than 3 kJ per TO2), the energetic effects associated with the number of cations increases (Table 3). Yang and Armbruster
minor structural changes in the dehydrated Cpt, Na-Cpt, NaK- (1996) concluded that increasing cation radius leads to the de-
Cpt, and K-Cpt are negligible. The enthalpy of the crystal-glass crease of the number of H2O molecules in HEU structure due
transition for the dehydrated Ca-Cpt is less than –3 kJ/mol TO2. to the space limitations. Our results suggest that the change of
To preserve the crystal structure of Ca-Cpt while achieving the volume occupied by cations is only partially compensated
dehydration, lower temperature and shorter heating time were by the change of the volume occupied by water molecules.
also applied. Even though the crystallinity of the zeolite struc- Considering the low average hydration number (water mol-
ture is largely retained based on XRD under these less severe ecules per cation) of the hydrated K-Cpt (only about 2.4, see
dehydration conditions, a small portion of the clinoptilolite in below), one cannot attribute its low occupancy of extra-frame-
both Ca-CptA and Ca-CptB is transformed to another crystal- work space to the water-cation bond strength. Therefore, one
line phase (Fig. 2). Furthermore the dehydration under these possible explanation is that the packing of the extra-framework
new conditions is not complete, about 1.5% water persists in species (i.e., cation and water) is less effective in the
the dehydrated crystal structure of Ca-CptB. clinoptilolite structure for larger extra-framework cations. This
The rehydration capacity of the dehydrated samples was may be expected considering that the size of these extra-frame-
measured by the mass change after equilibration with water work species is not too much smaller than the pore size of the
vapor of 50% relative humidity at room temperature (Table 2). framework.
The data suggest that the thermal treatment during dehydra-
tion reduces the sorption capacity and the structure has changed. Hydration enthalpy
This effect varies with the type of exchanged cations. The de- Drop solution calorimetry was performed for both hydrated
crease of the rehydration capacities for the dehydrated Ca-Cpt and dehydrated clinoptilolites. The calorimetric data are pre-
at the same or less severe conditions is more pronounced than sented both on the original mass basis and on the molar basis
for the samples with other cations. This observation is consis- of two framework O atoms (Table 4). Because the dehydrated
tent with the more pronounced structure changes of the Ca- Ca-Cpt became amorphous, a sample (Ca-CptB) dehydrated at
exchanged clinoptilolite as discussed above, and agrees with less severe conditions was also measured.
the previous conclusion that Ca-Cpt, similar to heulandite, has
lower thermal stability as compared with monovalent cationic TABLE 3. Space occupancy of the extra framework water mol-
clinoptilolite (Mason and Sand 1960; Petrova and Kirov 1995). ecules and cations in the natural and ion-exchanged
clinoptilolites equilibrated with water vapor of 50% rela-
Carey and Bish (1996) investigated the effect of thermal treat- tive humidity at 25 °C
ment on the rehydration capacity of clinoptilolites, and reported
Zeolite n* VH2O (Å3)† r (Å)‡ VCation (Å3) VCation + H2O (Å3)
that reversibility of rehydration for K- and Ca-exchanged Ca-Cpt 0.646 19.37 0.98 0.36 19.73
clinoptilolite could not be retained after heating the sample Na-Cpt 0.572 17.12 1.02 0.81 17.93
above temperatures as low as 250 °C. The irreversibility of Cpt 0.528 15.92 1.07 0.85 16.77
NaK-Cpt 0.495 14.82 1.19 1.36 16.17
dehydration-hydration for their Na-exchanged clinoptilolite K-Cpt 0.433 12.96 1.38 2.00 14.96
occurred at even lower temperature, 220 °C. The onset of irre- Notes: data are based on two framework O atoms.
versibility of dehydration-hydration may indicate some irre- * Molar numbers of water molecules based on two framework O atoms.
† Volume of water based on two framework O atoms. The water density
versible structure change. is taken as 1.0 g/mL.
‡ Population-weighted average ionic radius of the extra-framework cations.

TABLE 2. Water contents of the hydrated and rehydrated


clinoptilolites (both equilibrated with water vapor of 50% TABLE 4. Drop solution enthalpies of the natural and ion-exchanged
relative humidity at 25 °C) before and after treatment at clinoptilolites measured by high temperature calorimetry
400 °C for 9 hours then followed by at 500 °C for 3 hours Zeolite ∆HDS-hyd (Hydrated zeolite) ∆HDS-deh (Deydrated zeolite)
Zeolite H2O wt% H2O wt% DHC (%)* J/g kJ/mol* J/g kJ/mol*
(initial) (rehydration) Ca-Cpt† 1323.3 ± 8.1 (5) 99.31 ± 0.60 587.1 ± 4.1 (5) 37.23 ± 0.26
Ca-Cpt 15.50 6.34 59.1 Ca-CptB‡ – – 707.7 ± 14.1 (7) 45.68 ± 0.91
Ca-CptA† 15.50 11.28 27.2 Cpt 1315.0 ± 5.0 (6) 96.90 ± 0.37 682.1 ± 8.9 (5) 43.78 ± 0.57
Ca-CptB‡ 15.50 11.82 23.7 Na-Cpt 1375.4 ± 6.5 (5) 102.29 ± 0.49 711.8 ± 5.5 (5) 45.61 ± 0.35
Cpt 12.90 10.54 18.3 NaK-Cpt 1305.5 ± 7.4 (6) 97.08 ± 0.55 747.8 ± 10.6 (5) 48.93 ± 0.69
Na-Cpt 13.85 11.23 18.9 K-Cpt 1254.7 ± 9.1 (6) 93.87 ± 0.68 755.7 ± 5.9 (5) 50.64 ± 0.39
NaK-Cpt 12.00 10.95 8.7 Notes: errors are expressed as two standard deviations of the mean,
K-Cpt 10.43 9.72 6.8 value in parentheses is number of measurements.
* Decrease of hydration capacity = [H2O% (initial) – H2O% (rehydrated)/ * On a TO2 basis.
H2O% (initial). † Most of the XRD crystallinity is lost after dehydration (see also Fig. 2).
† Dehydration at 300 °C for 15 hours followed by at 400 °C for 3 hours. ‡ While the XRD crystallinity is largely retained after dehydration under
‡ Dehydration at 300 °C for 7 hours followed by at 350 °C for 1 hour and less severe condition, about 1.5% of water content is found in the dehy-
at 400 ºC for 1 hour. drated zeolite (see text and Fig. 2).
442 YANG ET AL.: THERMODYNAMICS OF CLINOPTILOLITE

The hydration process of a zeolite can be written as dehydrated zeolitic Ca-Cpt and the anhydrous amorphous Ca-
Cpt. The former is, of course, experimentally inaccessible by
∆Hhyd MaM'bAlxSiyO2 (crystalline, 25 °C) + nH2O (liquid, 25 °C) → the present method.
MaM'bAlxSiyO2·nH2O (crystalline, 25 °C) (3)
Enthalpy of formation
where ∆Hhyd is the molar hydration enthalpy based on a TO2 Formation of a hydrated zeolite from its constituent oxides
formula unit relative to liquid water at 25 °C. Then can be generalized by the following chemical reaction

∆Hhyd-z = ∆HDS-deh + n·∆Hwater – ∆HDS-hyd (4) ∆Hf-hyd 0.5aM2O (c, 25 °C) + bM'O (c, 25 °C) + 0.5xAl2O3
(c, 25 °C) + ySiO2 (c, 25 °C)
where ∆Hwater is the heat content for transformation of one mole + nH2O (l, 25 °C) → MaM'bAlxSiyO2·nH2O (zeolite, 25 °C) (5)
of liquid water from 25 °C to gas at 700 °C at 1 bar (Robie and
Hemingway 1995). The (∆Hhyd-z in Table 5) integral molar en- where ∆Hf-hydis the enthalpy of formation, which can be de-
thalpy of hydration of a zeolite is based on a TO2 formula unit rived from the following thermodynamic cycle:
reflecting the stabilization effect of the zeolite structure by
hydration. ∆H6M2O (c, 25 °C) → M2O (soln, 700 °C) (6)
The molar hydration enthalpy based on one mole of water, ∆H7M’O (c, 25 °C) → M'O (soln, 700 °C) (7)
∆Hhyd-w, is calculated by ∆Hhyd-w = ∆Hhyd-z/n (Table 6). This value ∆H8Al2O3 (c, 25 °C) → Al2O3 (soln, 700 °C) (8)
shows the average bond energy between water molecules and ∆H9SiO2 (c, 25 °C) → SiO2 (soln, 700 °C) (9)
the zeolite structure. Note the hydration state of the zeolites. ∆H10H2O (c, 25 °C) → H2O (g, 700 °C) (10)
The hydrated sample has been equilibrated with water vapor at ∆Hf= 0.5a·∆H6 + b·∆H7 + 0.5x·∆H8 + y·∆H9 + n·∆H10 – ∆H1 (11)
50% relative humidity at 25 °C. The dehydrated samples cor-
respond to the standard dehydration conditions applied in this ∆H6, ∆H7, ∆H8, ∆H9, and ∆H10 correspond to the drop solution
work, in which Cpt, Na-Cpt, Na/K-Cpt, and K-Cpt are dehy- enthalpy of the respective oxides and water. These values have
drated clinoptilolites in their respective cation forms. For Ca- been established previously (Table 7). Using Equation 11 and
Cpt, the hydration enthalpies (Table 6) correspond to the the thermodynamic properties for the constituent oxides (Table
energetic differences between the hydrated Ca-Cpt and the de- 7), we obtained the molar enthalpies of formation based on a
hydrated amorphous sample. In all cases, the reference state TO2 formula unit for the hydrated clinoptilolites as well as for
for H2O is liquid water. Because the energetics associated with the dehydrated samples (Table 5). Because water is referred to
the crystal-glass transition of the dehydrated Ca-Cpt is small the liquid state at 25 °C, the difference of the ∆Hf values for
(~3 kJ/mol TO2 as discussed earlier), the hydration enthalpy the hydrated and the dehydrated clinoptilolites is in fact the
obtained for Ca-Cpt can be taken as a reasonable estimate even hydration enthalpy of the zeolite.
if we ignore the energetic difference between the hypothetical The standard molar enthalpy of formation of the exchanged

TABLE 5. Enthalpy of hydration (∆Hhyd-z) and enthalpy of formation (∆Hf) of clinoptilolite at 25 °C


Zeolite n Z/r* Water/Cation† ∆Hhyd-z (kJ/mol‡) ∆Hf from oxides, kJ/mol‡
Hydrated Dehydrated
Ca-Cpt 0.646 2.07 7.0 –17.59 ± 0.77 –14.18 ± 0.67 3.41 ± 0.38§
Cpt 0.572 1.21 3.5 –16.76 ± 0.78 –26.66 ± 0.46 –9.90 ± 0.64
Na-Cpt 0.528 0.98 3.1 –17.29 ± 0.72 –31.40 ± 0.56 –14.11 ± 0.45
NaK-Cpt 0.495 0.86 2.7 –14.02 ± 0.96 –34.88 ± 0.62 –20.86 ± 0.74
K-Cpt 0.433 0.72 2.4 –13.39 ± 0.88 –39.88 ± 0.74 –26.50 ± 0.48
* Population weighted average ionic potential.
† Average number of water molecules per cation.
‡ Based on TO2.
§ Most of the XRD crystallinity is lost after dehydration under the standard conditions (see text and Fig. 2).

TABLE 6. Integral enthalpies of hydration of clinoptilolite (∆Hhyd-w, kJ/mol – H2O 25 °C, based on liquid water) measured in this work and
reported in the literature
Calorimetric methods Equilibrium measurements
Zeolite This work Carey and Johnson et al. Barrer and Cram Wilkin and Barnes Carey and Bish
Bish (1997) (1991) (1971) (1999) (1996)
(Si/Al = 4.5) (Si/Al = 4.3) (Si/Al = 4.2) (Si/Al = 5.0) (Si/Al = 4.5)* (Si/Al = 4.3)
Ca-Cpt –27.25 ± 1.19† –30.3 ± 2.0 – – – –32.92 ± 2.88
Cpt –31.77 ± 1.48 – –23.5 – – –
Na-Cpt –30.24 ± 1.25 –23.4 ± 0.6 – –21.8 –40.8 ± 9.7 –30.19 ± 3.46
NaK-Cpt –28.30 ± 1.94 – – – – –
K-Cpt –30.91 ± 2.03 –22.4 ± 0.8 – – –37.0 ± 7.5 –23.78 ± 1.25
* Same clinoptilolite as used in this work.
† The value may be more exothermic because the energetics of crystal-glass transition of the dehydrated Ca-Cpt is neglected (see text for explanation).
YANG ET AL.: THERMODYNAMICS OF CLINOPTILOLITE 443

TABLE 7. Drop solution enthalpies (∆HDS) of the constituent ox- hydration vs. hydration degree) are about –25 to –30 kJ/mol
ides from room temperature at 25 °C into molten H2O at the highest hydration degrees (60 to 80%) of the differ-
2PbO·B2O3 at 700 °C and their standard formation en-
thalpies (∆H0f) from the constituent elements at 25 °C ent cationic clinoptilolites. Their results show similar trends to
ours but close comparison is difficult because of impurities
Component ∆H0f (kJ/mol) * ∆HDS (kJ/mol)
Al2O3 corundum –1675.7 ± 1.3 107.9 ± 1.0†
(20%) in their samples and lack of experimental details such
SiO2 quartz –910.7 ± 1.0 39.1 ± 0.3† as hydration-dehydration conditions.
Na2O –414.8 ± 0.3 –113.1 ± 0.8‡ Carey and Bish (1997) and Barrer and Cram (1971) used
K 2O –363.2 ± 2.1 –193.7 ± 1.1‡
CaO –635.1 ± 0.9 –17.4 ± 0.9† immersion calorimetry and Johnson et al. (1991) used HF so-
MgO –601.6 ± 0.3 36.5 ± 1.0§ lution calorimetry to measure hydration enthalpies (Table 6).
H2O liquid –285.8 ± 0.1 68.9 ± 0.1* Their ∆Hhyd-w values are considerably less exothermic than ours
* Robie and Hemingway (1995).
† Kiseleva et al. (1996a).
for similar cationic clinoptilolites. The difference in the frame-
‡ Kiseleva et al. (1996b). work Si/Al ratio of the different clinoptilolites may affect the
§ Smelik et al. (1994). water content as well as the ∆Hhyd-w value. However the varia-
tion of the framework Si/Al ratios among the samples is small
and is unlikely to account for the large changes in the ∆Hhyd-w
clinoptilolites from the constituent elements can also be ob- values. Carey and Bish (1997) used a temperature of 250 °C
tained by combining the calorimetric results of this work with under water vapor pressure of 0.4 mbar, Johnson et al. (1991)
the known standard formation enthalpies of the constituent used 350 °C, and Barrer and Cram (1971) used 360 °C under
oxides (Table 7). The molar enthalpy of formation for the hy- vacuum for the dehydration of their samples. Residual water
drated samples including Ca-Cpt measured by drop solution would persist after dehydration at these temperatures as com-
calorimetry will not be affected by the structure change during pared with zeolites treated at higher temperature (500 °C in
dehydration. this study), e.g., Carey and Bish (1997) reported that about 2–
20% of the initial water content remained in their most dehy-
DATA ANALYSIS AND DISCUSSION drated clinoptilolites. Because of the significantly higher
differential molar hydration enthalpy at low coverage, even a
Effect of exchanged cations on hydration of clinoptilolites small amount of residual water may cause the integral molar
For all the ion-exchanged clinoptilolites in this study, the hydration enthalpy to be significantly less exothermic. Indeed,
integral enthalpies of hydration based on one mole of water we obtained only –14.1 ± 1.7 kJ/mol H2O for Ca-CptB, which
(∆Hhyd-w) are about –30 kJ/mol with a span of 3.5 kJ/mol, which is dehydrated under less severe conditions and has a water con-
is close to the error range of the measurements (Table 6). Be- tent of about 1.5 wt% (Table 2). In addition, Carey and Bish
cause the water molecules in zeolites are bonded to different (1997) and Barrer and Cram (1971) measured hydration prop-
sites, the integral hydration enthalpy yields a value averaged erties at 100% (in contrast to 50% in this study) relative hu-
over these sites. The similar ∆Hhyd-w values imply that the aver- midity. Therefore the different dehydration conditions used by
age bonding strength felt by each zeolitic water molecule in the various researchers represent one major factor accounting
these different cationic clinoptilolites is approximately the same. for the different ∆Hhyd-w values.
This hydration behavior was also found for the various ion- Dehydration of HEU zeolites induces cation migration and
exchanged FAU zeolites (Yang and Navrotsky 2000) and a se- structure distortion (Dyer and White 1999; Hambley and Tay-
ries of ion-exchanged CHA zeolites (Shim et al. 1999). This lor 1984; Merkle and Slaughter 1968; Armbruster 1993; Valueva
behavior may result from a balance of two competing effects 1995; Yang and Armbruster 1996; Yang et al. 1997; Zhao et al.
when the extra-framework cations are exchanged. First, the 1997) and sample shrinkage (Kranz et al. 1989; Dell’Agli et
higher the ionic potential of the extra-framework cations, i.e., al. 1999). A high degree of dehydration and/or high tempera-
the smaller their size and/or the higher their charge, the stron- ture treatment often lead to irreversible structural changes or
ger their attraction to water molecules. Second, for a given zeo- even complete structure collapse (Armbruster and Gunter 1991).
lite structure, the higher the ionic potential of the cations, the Heulandites, i.e., Al-rich and/or Ca-rich members of HEU type
more water molecules can be packed in the structure because zeolites, undergo structural collapse by 350 °C (e.g., Gottardi
these cations occupy less space and the water-cation packing and Galli 1985). The decrease of hydration capacity after high
is more efficient. Hence, the stronger attraction of these cat- temperature dehydration or irreversibility of hydration-dehy-
ions to water is shared by a larger number of water molecules. dration reactions at high temperature will give rise to different
Another consequence of these competing effects is that the ef- ∆Hhyd-w values depending on the measurement procedure, i.e.,
fect of the extra-framework cations on the molar enthalpy of dehydration of the initial hydrated zeolite or the hydration of
hydration of a zeolite (∆Hhyd-z) will be largely determined by dehydrated zeolite. Less exothermic ∆Hhyd-w values may be
their effect on the hydration capacity (Table 5). This type of obtained from the latter procedure than from the former one.
behavior of hydration enthalpy may be expected for other open This may be the second major factor for the less exothermic
framework zeolites. ∆Hhyd-w values obtained by Barrer and Cram (1971).
Using immersion calorimetry, Petrova and Kirov (1995) Non-calorimetric methods also give ∆Hhyd-w (Table 6). Carey
measured the integral heats of hydration at various hydration and Bish (1996) used equilibrium water content measurements
degrees for several ion-exchanged clinoptilolites (Si/Al ~5.0). of clinoptilolite-H2O from 25 to 250 °C and 0.2 to 35 mbar
The ∆Hhyd-w values estimated from their plots (integral heats of water vapor pressure. Their ∆Hhyd-w values are similar to ours
444 YANG ET AL.: THERMODYNAMICS OF CLINOPTILOLITE

for Na-Cpt and Ca-Cpt but are less exothermic for K-Cpt.
Wilkin and Barnes (1999) used a pressure titration technique
to 300 °C and PH2O < 30 bars for study of hydration. Their val-
ues have a comparatively large uncertainty but are more exo-
thermic for the same Na-Cpt and K-Cpt than those obtained
here. Although the equilibrium method provides useful infor-
mation about the dehydration behavior as a function of tem-
perature, water vapor pressure PH2O and hydration level, the
derived integral ∆Hhyd-w values may be less certain relative to
calorimetric results for the following reasons. First, these ex-
perimental data were measured only over part of the hydration
range. To obtain an integral enthalpy of hydration by extrapo-
lation, these data are usually fit into a Langmuir-type sorption
model. Such a data processing procedure simplifies or neglects
the heterogeneous nature of zeolite-water reactions because the
properties of zeolitic water vary with the degree of filling of
the channel and cavities, i.e., the hydration level (e.g.,
Hemingway and Robie 1984). Second, the hydration level be-
comes increasingly difficult to measure at low coverage and F IGURE 3. Plot of the molar enthalpies of formation of the
cannot be experimentally controlled at very low coverage. clinoptilolites from the constituent oxides based on two framework O
Nonetheless these low-coverage and high-energy water mol- atoms vs. the population-weighted average ionic potential of their extra
ecules contribute most in terms of enthalpy of molar hydra- framework cations.
tion. Third, as discussed by Barrer and Cram (1971), true
hydration-dehydration equilibrium may not be readily attain- zeolites (Yang and Navrotsky 2000). This systematic trend is a
able at very low hydration levels. Therefore, the direct mea- useful tool for predicting the energetics of ion-exchanged forms
surement of hydration enthalpy by calorimetry may provide in the fast growing field of molecular sieves.
useful constraints on modeling zeolite-water reaction based on In natural HEU species, the cation composition often var-
the equilibrium measurements. ies with Si/Al ratio. Using the systematics of Figure 3, we esti-
In conclusion, a highly dehydrated zeolite is unlikely to have mate other energetic contributions by separating the energetic
a truly intact framework structure and therefore the hydration- effect of cation differences. For example, from the ∆Hf value
dehydration reaction is not ideally reversible. The major fac- of a Siberia heulandite (Ca0.86Na0.37K0.06Al2.14Si6.86·6.1H2O, Si/
tors influencing the measured hydration enthalpy are the Al = 3.2), –26.52 ± 0.54 kJ/mol-TO2, recently measured by
dehydration conditions and the direction of the measurement. Kiseleva et al. (2000), the ∆Hf value of the clinoptilolite coun-
These experimental-procedure-related factors may account for terpart with equivalent cation composition can be estimated to
the different hydration enthalpies measured by calorimetric be – 20.62 kJ/mol-TO2. Therefore, the energetic cost for the Si/
methods in this work and those reported by others (Carey and Al ratio change from 3.2 to 4.5 is estimated to be endothermic
Bish 1997; Johnson et al. 1991; Barrer and Cram 1971). There- at about 5.90 kJ/mol-TO2.
fore explicitly defined experimental conditions for hydration The standard molar enthalpies of formation of the ion-ex-
and dehydration are highly desirable and they should be care- changed clinoptilolites from the constituent elements at 25 °C
fully taken into account when the hydration enthalpy values are in Table 8. We are aware of only one other set of directly
are put into use. For a real zeolite, hydration enthalpy may be measured data for a natural clinoptilolite by Johnson et al.
difficult to define uniquely and unambiguously. (1991) using HF solution calorimetry. The ∆H0f values for the
hydrated and dehydrated clinoptilolites of Johnson et al. is
Effect of extra-framework cations on energetics of –1146.94 ± 0.75 kJ/mol-TO2 and –959.30 ± 0.74 kJ/mol-TO2
clinoptilolites respectively. Because ∆H 0f values are strongly dependent on
The enthalpy of formation of the dehydrated clinoptilolites the composition of the zeolite, we convert their data to the en-
from the constituent oxides becomes less exothermic as the thalpies of formation from the constituent oxides so that we
average ionic potential of the extra-framework cations increases can more easily compare their results with the Figure 3. Using
(Fig. 3). The plot has two linear segments: a steep section at the standard thermodynamic properties of the respective ox-
low Z/r for the monovalent clinoptilolites (K-Cpt, Na/K-Cpt, ides and liquid water (Robie and Hemingway 1995), the corre-
and Na-Cpt) and a shallow section at high Z/r for the mixed sponding enthalpies of formation from the constituent oxides
monovalent/divalent cationic forms (Na-Cpt, Cpt, and Ca-Cpt). and liquid water at 25 °C for Johnson et al.’s hydrated and de-
The trend of the plot can be related to the basicity of the oxides hydrated samples are –26.64 ± 1.11 kJ/mol-TO2 and –12.45
of the corresponding ion-exchanged cations for the acid-base ± 1.11 kJ/mol-TO2, respectively. Placing these experimental
interaction in the substitution reaction, Si4+ → Al3+ + M+ or results on Figure 3 and considering that the slightly lower Si/
0.5M2+ (Navrotsky 1994). A similar shift of the linear depen- Al ratio of their sample compared to ours would cause their
dence of ∆Hf on Z/r from monovalent cationic zeolites to mul- results to be slightly more exothermic as discussed above. Yet
tivalent cationic zeolites was also found for ion-exchanged FAU the ∆Hf value for their hydrated sample agrees reasonably with
YANG ET AL.: THERMODYNAMICS OF CLINOPTILOLITE 445

the trend predicted by the systematics illustrated in Figure 3. Carey and Bish (1996) is also negative. This negative entropy
However, the ∆Hf value for the dehydrated clinoptilolite is off- change suggests that the water present in the hydration spheres
set from the trend of our values considerably, being about 5 kJ/ of the extra-framework cations is more constrained than in the
mol-TO2 more exothermic (Fig. 3). This exothermic offset is liquid state. Negative enthalpies of hydration are compensated
consistent with the less exothermic enthalpy of hydration arising by negative entropies of hydration. This point will be discussed
from their use of a relatively low temperature for dehydration (see further in future communication.
the foregoing discussion about hydration enthalpy). That is, their As discussed by Hemingway and Robie (1984), the heat
dehydrated sample may not have been completely anhydrous. capacity of the average zeolitic water in clinoptilolites will be
somewhat greater than the heat capacity of ice at 100 K and
Entropy and free energy of the ion-exchanged somewhat less than that for pure water at 298 K. So the en-
clinoptilolite at 25 °C tropy contribution of the zeolitic water in clinoptilolites will
Two sets of entropy data for hydrous clinoptilolites (S0zeo, be less than the same amount of pure water. However, there are
corresponding to S0(298 K) – S0(0 K) because of unknown con- some trade-offs for the lower mobility of zeolitic water mol-
figuration disorder) exist (Johnson et al. 1991; Hemingway and ecules as compared to pure water at 25 °C: as the zeolitic water
Robie 1984). The S0zeo value of hydrous heulandite (Johnson et content increases the mobility of the exchangeable cations in-
al. 1985) is also listed in Table 9 as a “reference clinoptilolite” creases (Abdoulaye et al. 2000) and water-cation mixing dis-
with “radical difference in composition.” These three sets of order increases. Relative to pure liquid water and cations in the
S0zeo data were obtained by low temperature adiabatic calori- pure oxides the lower entropy contribution associated with the
metric measurements. The sum of the S0zeo values for the com- lower mobility of zeolitic water molecules is then partially com-
ponent oxides (ΣS0oxide) of the three HEU zeolites is calculated pensated by the larger entropy contribution related to the higher
by using the thermodynamic data tabulated by Robie and mobility of the cations and to the higher mixing disorder. For
Hemingway (1995). The entropy difference between these zeo- the same zeolite structure (HEU) and the same hydration con-
lites and their component oxides (H2O as liquid water), S0zeo – ditions (equilibrated to 50% humidity at 25 °C), the overall net
ΣS0oxide, are between –4 to –6 J/mol·K (Table 9). Apparently the entropy difference between different HEU zeolites and their
S0zeo value of these hydrous HEU zeolites can be related to the components may be expected both to be small and similar in
ΣS0oxide by the following empirical equation: magnitude. Such a rationalization may qualitatively explain the
behavior of the ∆S for the HEU zeolites seen in Table 9.
S0zeo (J/mol·K TO2) = ΣS0oxides – (5.0 ± 1.5) (12) The difference of S0zeo and S0H2O, where S0H2O represents the
entropy of the same quantity of pure water, gives rather close
The correction factor, –5.0 mol·K, and the estimated error, values (last column in Table 9). The parallel values of S0zeo and
±1.5 J/mol·K, are determined from the analysis of the data in S0H2O for the different HEU zeolites suggest that the slightly dif-
Table 9. Despite the different origins of these zeolites and the ferent water contents in these HEU-type zeolites are mainly
wide range of their compositions (such as Si/Al ratio and cat- responsible for their different S0zeo values. For the different struc-
ion constituents and water contents), the entropy difference (S0zeo ture types of zeolites, however, the thermodynamic properties
– ΣS0oxide) is consistently negative about –5.0 mol·K within the of the zeolitic water may be significantly different, e.g., the
errors of ± 1.5 J/mol·K. The entropy of hydration measured by zeolitic water more closely represents the properties of pure

TABLE 8. Thermodynamic properties of the natural and ion-exchanged clinoptilolites at 25 °C


Zeolite ∆Hf0, kJ/mol S0zeo, J/mol·K ∆Gf0, kJ/mol 445 ∆Gf0, kJ/mol S0zeo, J/mol·K
(Calorimetric results (Summation results (This work) (Solubility (Derived data)†
of this work) of this work) measurement)*
Ca-Cpt –1153.78 ± 1.07 82.16 ± 1.50 –1064.95 ± 1.16 –1064.3 ± 1.7 79.8 ± 6.7
Cpt –1117.57 ± 0.95 76.41 ± 1.50 –1034.01 ± 1.05 – –
Na-Cpt –1130.05 ± 1.00 80.46 ± 1.50 –1044.19 ± 1.10 –1044.7 ± 1.7 81.9 ± 6.5
NaK-Cpt –1109.49 ± 1.04 75.91 ± 1.50 –1027.26 ± 1.13 – –
K-Cpt –1094.21 ± 1.12 72.47 ± 1.50 –1014.89 ± 1.21 –1017.8 ± 1.7 82.1 ± 6.8
Notes: All values are based on two framework O atoms.
* Data of Benning et al. (2000) for Ca-Cpt, Wilkin and Barnes (1998) for Na-Cpt and K-Cpt.
† Derived from the experimental data: the ∆Hf0 values in column 2 and the ∆Gf0 values in column 5.

TABLE 9. Comparison of the published experimental entropies (S0zeo) of HEU type zeolites and the sum of the entropies of the constituent
oxides ( ΣS0oxide)
Zeolite Entropy at 25 °C, J/mol·K Effect of water content on entropy (J/mol·K)
S0zeo ΣS0oxide S0zeo-ΣS0oxide n
H2 O S0H2O (liquid H2O) S0zeo-S0H2O
Cpt-I* 82.39 86.48 ± 0.14 –4.09 0.607 42.47 39.92
Cpt-II† 79.79 ± 0.25 85.68 ± 0.13 –5.89 ± 0.28 0.611 42.78 37.01
Heu‡ 85.24 ± 0.09 90.67 ± 0.12 –5.43 ± 0.12 0.667 46.67 38.58
Notes: All values in the table are based on two framework O atoms.
* The original formula of the clinoptilolite (Johnson et al. 1991): Sr0.036Mg0.124Ca0.761Mn0.002Ba0.062K0.543Na0.954Al3.450Fe0.017Si14.533O36·10.922H2O.
† The original formula of the clinoptilolite (Hemingway and Robie 1984): Na0.56K0.98Ca1.50Mg1.23Al6.7Fe0.3Si29O72·22H2O.
‡ The original formula of the heulandite (Johnson et al. 1985): Ba0.065Sr0.175Ca0.585K0.132Na0.383Al2.165Si6.835O18·6.00H2O.
446 YANG ET AL.: THERMODYNAMICS OF CLINOPTILOLITE

water in the FAU type structure (zeolite X) type than in the REFERENCES CITED
phillipsite type structure; clinoptilolite and analcime show even Abdoulaye, A., Zanchetta, J.V., Di Renzo, F., Giuntini, J.C., Vanderschueren, J., and
greater deviations from the properties of pure water than Chabanis, G. (2000) Dielectric properties of faujasites: Comparison between
types X and Y during dehydration. Microporous Mesoporous Materials, 34,
phillipsite does (e.g., Hemingway and Robie 1984; Johnson et 317–325.
al. 1985). Therefore, Equation 12 may have different correc- Alberti, A. (1975) The crystal structure of two clinoptilolites. Tschermaks
tion factors for other types of zeolites. Mineralogische und Petrographische Mitteilungen, 22, 25–37.
Alietti, A. (1972) Polymorphism and crystal-chemistry of heulandites and
Using Equation 12, we calculated the S0zeo values for our clinoptilolites. American Mineralogist, 57, 1448–1462.
hydrous clinoptilolites (Table 8). This calculation is similar to Armbruster, T. (1993) Dehydration mechanism of clinoptilolite and heulandite: Single
the summation procedure used for estimation of the ∆H0f and crystal X-ray study of Na-poor, Ca-, K-, Mg-rich clinoptilolite at 100 K. Ameri-
can Mineralogist, 78, 260–264.
∆G0f values of zeolites from the thermodynamic properties of Armbruster, T. and Gunter, M.E. (1991) Stepwise dehydration of heulandite-
the component oxides or hydroxides (Tardy and Garrels 1974; clinoptilolite from Succor Creek, Oregon, U.S.A.: a single-crystal x-ray study
at 100 K. American Mineralogist, 76, 1872–1883.
Nriagu 1975; La Iglesia and Aznar 1986; Chermak and Rimstidt Barrer, R.M. and Cram, P.J. (1971) Heats of immersion of outgassed and ion-ex-
1989; Mattigod and McGrail 1999). Combining the calculated changed zeolites. In E.M.Flanigen and L.B. Sand, Eds., Molecular sieve zeo-
entropy values with the enthalpy values based on the experi- lites - II, p. 105–131. American Chemistry Society, Washington, D.C.
Benning, L.G., Wilkin, R.T., and Barnes, H.L. (2000) Solubility and stability of
mental measurements of this work, we obtain the standard free zeolites in aqueous solution: II. Calcic clinoptilolite and mordenite. American
energy (∆G0f) of the ion-exchanged clinoptilolites (Table 8). The Mineralogist, 85, 495–508.
Boles, J.R. (1972) Composition, optical properties, cell dimensions, and thermal
small correction factor (–5.0 J/mol·K) used in the calculation stability of some heulandite group zeolites. American Mineralogist, 57, 1463–
of entropy only translates to a contribution of 1.5 kJ/mol-TO2 1493.
to the resulting ∆G0f value. Bowers, T.S. and Burns, R.G. (1990) Activity diagrams for clinoptilolite: Suscepti-
bility of this zeolite to further diagenetic reactions. American Mineralogist, 75,
The standard thermodynamic properties of the ion-ex- 601–619.
changed clinoptilolites are in Table 8. The ∆G0f values agree Carey, J.W. and Bish, D.L. (1996) Equilibrium in the clinoptilolite-H2O system.
well with the respective values for the same Na-Cpt and K-Cpt American Mineralogist, 81, 952–962.
———(1997) Calorimetric measurement of the enthalpy of hydration of
(Wilkin and Barnes 1998) and the same Ca-Cpt (Benning et al. clinoptilolite. Clays and Clay Minerals, 45(6), 826–833.
2000) derived from solubility measurements. The S0zeo values Chermak, J.A. and Rimstidt, J.D. (1989) Estimating the thermodynamic properties
(∆Gf0 and ∆Hf0) of silicate minerals at 298 K from the sum of polyhedral contri-
of these three zeolites can also be calculated from the experi- butions. American Mineralogist, 74, 1023–1031.
mental ∆G0f values of Wilkin and Barnes (1998) and Benning Chi, C. and Sand, L.B. (1983) Synthesis of Na- and K-clinoptilolite end members.
et al. (2000) and the experimental ∆H0f values of this work. The Nature, 304, 255–257.
Chipera, S.J. and Bish, D.L. (1997) Equilibrium modeling of clinoptilolite-anal-
difference between these calculated S0zeo values and those cal- cime equilibria at Yucca Mountain, Nevada, USA. Clays and Clay Minerals,
culated using Equation 12 are of the same magnitude as the 45(2), 226–239.
associated errors (Table 8). Such an agreement further vali- Coombs, D.S., Alberti, A., Armbruster, T., Artioli, G., Colella, C., Galli, E., Grice,
J.D., Liebau, F., Mandarino, J.A., Minato, H., Nickel, E.H., Passaglia, E., Peacor,
dates the use of Equation 12 for the estimation of the S0zeo val- D.R., Quartieri, S., Rinaldi, R., Ross, M., Sheppard, R.A., Tillmanns, E., and
ues for the hydrated clinoptilolite. Vezzalini, G. (1997) Recommended nomenclature for zeolite minerals: report
of the subcommittee on zeolites of the International Mineralogical Association,
Several factors may account for the good agreement be- Commission on New Minerals and Mineral Names. The Canadian Mineralo-
tween our ∆G0f values and those reported by Wilkin and Barnes gist 35, 1571–1606.
(1998) and Benning et al. (2000). Our careful calorimetric Dell’Agli, G., Ferone, C., Mascolo, G., and Pansini, M. (1999) Dilatometry of Na-, K-,
Ca- and NH4-clinoptilolite. Thermochimica Acta, 336, 105–110.
measurement provides high quality data for enthalpy of for- Dyer, A. and White, K.J. (1999) Cation diffusion in the natural zeolite clinoptilolite.
mation of hydrated zeolite. In addition, those results pertain Thermochimica Acta, 340–341, 341–348.
only to the hydrated samples, the uncertainties associated with Glassley, W.E., Bruton, C.J., and Bourcier, W.L. (1994) Testing long-term predic-
tions from hydro-geochemical models. Materials Research Society symposium
dehydration are irrelevant. The oxide summation method pro- proceedings, 333 (Scientific Basis for Nuclear Waste Management XVII), 805–
vides an excellent estimation for entropy of the hydrated zeo- 810.
Gottardi, G. and Galli, E (1985) Natural Zeolites, p. 256-305. Springer, Berlin.
lites. Because the entropy contribution to the Gibbs free energy Hambley, T.W. and Taylor, J.C. (1984) Neutron diffraction studies on natural heu-
of a hydrated zeolite is much smaller than the enthalpy contri- landite and partially dehydrated heulandite. Journal of Solid State Chemistry,
bution, the quality of the derived Gibbs free energy will be 54, 1–9.
Hemingway, B.S. and Robie, R.A. (1984) Thermodynamics properties of zeolites:
primarily determined by the calorimetric enthalpy data. Third, Low temperature heat capacities and thermodynamic functions of phillipsite
the identical zeolite samples used in these measurements favor and clinoptilolite. Estimates of the thermodynamical properties of zeolitic wa-
consistent results. ter at low temperature. American Mineralogist, 69, 692–700.
Johnson, G.K., Tasker, I.R., Jurgens, R., and O’Hare, P.A.G. (1991) Thermody-
The thermodynamic properties of three natural clinoptilolites namic studies of zeolites: Clinoptilolite. Journal of Chemical Thermodynam-
were also reported by Murphy et al. (1996), Johnson et al. ics, 23, 475–484.
Johnson, G.K., Flotow, H.E., O’Hare, P.A.G., and Wise, W.S. (1985) Thermody-
(1991), and also by Bowers and Burns (1990) using the heat namic studies of zeolites: Heulandite. American Mineralogist, 70, 1065–1071.
capacity and entropy data of Hemingway and Robie (1984). Kerr, R.A. (2000) Nuclear waste disposal: Science and policy clash at Yucca Moun-
As discussed by Wilkin and Barnes (1998), the ∆G0f values for tain. Science, 288, 602.
Kiseleva, I., Navrotsky, A., Belitsky, I.A., and Fursenko, B.A. (1996a) Thermo-
these clinoptilolites cannot be readily compared because of the chemistry and phase equilibria in calcium zeolites. American Mineralogist, 81,
compositional difference among the clinoptilolites. 658–667.
———(1996b) Thermochemistry and phase equilibria of natural potassium sodium
calcium leonhardite and its cation-exchanged forms. American Mineralogist,
ACKNOWLEDGMENTS 81, 668–675.
———(2000) Thermochemical study of calcium zeolites – heulandite and stilbite.
We thank Bill Carey, Micky Gunter, and Bruce Hemingway for their help- American Mineralogist, 86, 448–455.
ful comments and discussion. This work was supported by the National Science Koyama, K. and Takéuchi., Y. (1977) Clinoptilolite: The distribution of potassium
Foundation, Grants DMR-97-31782. atoms and its role in thermal stability. Zeitschrift für Kristallographie, 145, 216–
YANG ET AL.: THERMODYNAMICS OF CLINOPTILOLITE 447

239. temperatures. U.S. Geological Survey Bulletin, 2131.


Kranz, R.L., Bish, D.L., and Blacic J.D. (1989) Hydration and dehydration of zeolitic Satokawa, S. and Itabashi, K. (1997) Crystallization of single phase (K, Na)-
tuff from Yucca Mountain, Nevada. Geophysical Research Letters, 16(10), 1113– clinoptilolite. Microporous Materials, 8, 49–55.
1116. Smelik, E.A., Jenkins, D.M., and Navrotsky, A. (1994) A calorimetric study of syn-
La Iglesia, A. and Aznar, A.J. (1986) A method of estimating the Gibbs energies of thetic amphiboles along the tremolite-tschermakite join and the heats of forma-
formation of zeolites. Zeolites, 6, 26–29. tion of magnesio-hornblende and tschermakite. American Mineralogist, 79,
Mason, B. and Sand, L.B. (1960) Clinoptilolite from Patagonia: The relationship 1110–1122.
between clinoptilolite and heulandite. American Mineralogist, 45, 341–350. Shim, S.H., Navrotsky, A., Gaffney, T.R., and MacDougall, J.E. (1999) Chabazite:
Mattigod, S.V. and McGrail, B.P. (1999) Estimating the standard free energy of Energetics of hydration, enthalpy of formation, and effect of cations on stabil-
formation of zeolites using the polymer model. Microporous Mesoporous Ma- ity. American Mineralogist, 84, 1870–1882.
terials, 27, 41–47. Tardy, Y. and Garrels, R.M. (1974) ) A method of estimating the Gibbs energies of
Merkle, A.B. and Slaughter, M. (1968) Determination and refinement of the struc- formation of layer silicates. Geochimica et Cosmochimica Acta, 38, 1101–1116.
ture of heulandite. American Mineralogist, 53, 1120–1138. Valueva, G. (1995) Dehydration behavior of heulandite-group zeolites as a function
Mumpton, F.A. (1960) Clinoptilolite redefined. American Mineralogist, 45, 351– of their chemical composition. European Journal of Mineralogy, 7(6), 1411–
369. 1420.
Murphy, W.M., Pabalan, R.T., Prikryl, J.D., and Goulet, C.J. (1996) Reaction kinet- Wilkin, R.T. and Barnes H.L. (1998) Solubility and stability of zeolites in aqueous
ics and thermodynamics of aqueous dissolution and growth of analcime and solution: I. Analcime, Na-, and K-clinoptilolite. American Mineralogist, 83,
Na-clinoptilolite at 25 °C. American Journal of Science, 296, 128–186. 746–761.
Navrotsky, A. (1994) Repeating patterns in mineral energetics. American Mineralo- ———(1999) Thermodynamics of hydration of Na- and K-clinoptilolite to 300 °C.
gist, 791, 589–605. Physics and Chemistry of Minerals, 26, 468–476.
———(1997) Progress and new directions in high temperature calorimetry revis- Yang, P. and Armbruster, T. (1996) Na, K, Rb, and Cs exchange in heulandite single-
ited. Physics and Chemistry of Minerals, 24, 222–241. crystals: X-ray structure refinements at 100 K. Journal of Solid State Chemis-
Navrotsky, A., Hon, R., Weill, D.F., and Henry, D.J. (1980) Thermochemistry of try, 123, 140–149.
glasses and liquids in the systems CaMgSi2O6-CaAl2Si2O8-NaAlSi3O8, SiO2- Yang, P., Stolz, J., Armbruster, T., and Gunter, M.E. (1997) Na, K, Rb, and Cs ex-
CaAl2Si2O8-NaAlSi3O8, SiO2-Al2O3-CaO-Na2O. Geochimica et Cosmochimica change in heulandite single-crystals: Diffusion kinetics. American Mineralo-
Acta, 44, 1409–1423. gist, 82, 517–525.
Navrotsky, A., Peraudeau, G., McMillan, P., and Coutures, J.P. (1982) A thermody- Yang, S. and Navrotsky, A. (2000) Energetics of Formation and Hydration of Ion
namical study of glasses along the joins silica-calcium aluminate and silica- Exchanged Zeolite Y. Microporous and Mesoporous Materials, 37, 175–186.
sodium aluminate. Geochimica et Cosmochimica Acta, 46, 2039–2047. Zhao, D., Szostak, R., and Kevan, L. (1997) Electron Spin Resonance and Electron
Nriagu, J.O. (1975) Thermodynamical approximations for clay minerals. American Spin Echo Modulation Spectroscopic Studies of Cupric Ion-Adsorbate Interac-
Mineralogist, 60, 834–839. tions in Synthetic Clinoptilolite. The Journal of Physical Chemistry B, 101,
Pabalan, R.T. and Bertetti, F.P. (1999) Experimental and modeling study of ion ex- 5382–5390.
change between aqueous solutions and the zeolite mineral clinoptilolite. Jour- Zhao, D., Cleare, K., Oliver, C., Ingram, C., Cook, D., Szostak, R., and Kevan, L.
nal of Solution Chemistry, 28(4), 367–393. (1998) Characteristics of the synthetic heulandite – clinoptilolite family of zeo-
Petrova, N. and Kirov, D. (1995) Heats of immersion of clinoptilolite and its ion- lites. Microporous and Mesoporous Materials, 21, 371–379.
exchanged forms: A calorimetric study. Journal of Thermal Analysis, 43, 323–
328.
Petrovic, I. and Navrotsky, A. (1997) Thermochemistry of Na-faujasites with vary-
ing Si/Al ratios. Microporous Materials, 9, 1–12. MANUSCRIPT RECEIVED MAY 15, 2000
Robie, R.A. and Hemingway, B.S. (1995) Thermodynamics properties of minerals MANUSCRIPT ACCEPTED DECEMBER 7, 2000
and related substances at 298.15K and 1 bar (105 pascals) pressure and at higher PAPER HANDLED BY RICK RYERSON