J . Phys. Chem. 1992, 96,2645-2652
ESR g values. The ESR line widths largely support these assignments. The particles are concluded to be formed in the cy-cages of the zeolite structure. The zeolite framework modifies the gas-phase electron-transfer energetics. Based on gas-phase energetics three classes are found: normal electron transfer, no electron transfer, and reverse electron transfer. These different classes involve the electron-transfer distance, electron transfer via
2645
the zeolite framework, ion size effect on electrostatic interactions with the oxygen framework, large cation blocking of cy-cage entrances, and ion/atom migration after electron transfer.
Acknowledgment. This research was supported by the National Science Foundation, the Robert A. Welch Foundation, and the Texas Advanced Research Program.
Solvation of Cu( II)in Cu( 11)-Exchanged Synthetic Fluorohectorite, Synthetic Hydroxyhectorlte, Synthetic Beidellite, and Montmorillonite Studied by Electron Spin Resonance and Electron Spin Echo Modulation Jean-Marc Comets, Vittorio Luca, and Larry Kevan* Department of Chemistry, University of Houston, Houston, Texas 77204-5641 (Received: September 25, 1991)
The solvation of exchangeable Cu(I1) cations in montmorillonite, beidellite, fluorohectorite, and hydroxyhectorite has been studied by electron spin resonance (ESR) and electron spin echo modulation (ESEM). Room-temperature ESR spectra of Cu(I1)-doped Mg(I1)-exchanged montmorillonite (CuMg-mont) that was vacuum-dried at room temperature and then either equilibrated at 100%relative humidity or water-soaked showed that the Cu(I1) is coordinated to six water molecules and is freely tumbling within the middle of the interlayer space. The room-temperature ESR spectrum of similarly treated Cu(I1)-doped Mg(I1)-exchanged beidellite (CuMg-beid) contained signals from two Cu(I1) species A and B. Both of these species appear to be the Cu(I1) hexaaquo complex but species A has greater mobility than species B. Species A gives a single symmetric ESR line at room temperature while species B gives an ESR spectrum characteristic of Cu(I1) in axial symmetry. Cu(I1)-doped Mg(I1)-exchanged fluorohectorite (CuMg-fluorohect) and hydroxyhectorite (CuMg-hydroxyhect) under the same conditions gave a room-temperature ESR spectrum both containing one Cu(I1) species having a signal characteristic of Cu(I1) in axial symmetry. The Cu(I1)-water complex in water-soaked samples giving this signal was shown by ESEM to be coordinated to only two water molecules. Therefore, Cu(I1) cations in CuMg-fluorohect that is vacuum-dried prior to rehydration remain bound to the clay layers. Samples of the four smectites that were dried under vacuum at 100 OC and then water-soaked gave room-temperature ESR spectra characteristic of Cu(I1) in an axial environment and ESEM indicates that the Cu(I1) is never fully solvated. The Cu(I1)-exchanged smectite samples do, however, retain the ability to swell with water after dehydration at 100 OC. This suggests that it is not necessary for exchangeable cations to be fully hydrated for smectite interlayers to swell.
Introduction Smectite clay minerals are used as catalysts and catalyst supports, are important in controlling soil mechanics, and are used in drilling fluids for their rheological properties. These properties are in some part controlled by the presence of charge balancing cations between the clay layers and the interaction of these cations with surface sites and adsorbate molecule^.^-^ The structure of water adsorbed on the surfaces of smectites has been s t ~ d i e d ~ - ~ and re~iewed.~Low10-12explained that the swelling of smectites is monitored primarily by the interaction between clay surfaces and vicinal water. Na cations do not diffuse away significantly from the surfaces as the smectite is increasingly hydrated. The same does not appear to be true for divalent exchangeable cations. From Monte Carlo computer simulation techniques, it is concluded that, in the two-layer hydrates of Na(1)- and Mg(I1)-montmorillonite13and vermic~lite,’~ the Mg(I1) cations are positioned in ( I ) Laszlo, P. Science 1987, 235, 1473. (2) Laszlo, P.; Mathy, A. Helv. Chim. Acta 1987, 70, 577. (3) Newman, A. C. D. In Chemistry of Clays and Clay Minerals; Newman, A. C. D., Ed.; Wiley-Interscience: New York, 1987. (4) Clementz. D. M.; Pinnavaia, T. J.; Mortland, M. M. J . Phys. Chem. 1973, 77, 196. (5) McBride, M. B.; Mortland, M. M. Soil Sci. SOC.Am. Proc. 1974,48, 408.
( 6 ) McBride, M. B.; Pinnavaia, T. J.; Mortland, M. M. J. Phys. Chem. 1975, 79,2430. (7) McBride, M. B. Clays Clay Miner. 1976, 24, 211. (8) McBride. M. B. Clays Clay Miner. 1982, 30, 200. (9) Sposito, G.;Prost, R. Chem. Reu. 1982, 82, 553. (IO) Low, P. F.; Margheim, J. F. Soil Sci. SOC.Am. J . 1979, 43, 473. ( I I ) Low, P. F. Soil Sci. SOC.Am. J . 1980, 44, 667. (12) Low, P. F. Soil Sci. SOC.Am. J . 1981, 45, 1074.
the middle of the interlayers and are coordinated to six water molecules, while Na(1) ions in Na(1)-montmorillonite are bound to the basal oxygen surfaces and coordination around the Na(1) is completed by water molecules. It has been shown using electron spin resonance (ESR) and electron spin echo modulation (ESEM) that Ag(1) ions are coordinated to the basal oxygen surfaces of beidellite and probably also of fluorohect~rite.~~J~ Thus it seems that a consistent picture of hydrated cations within the interlayers of smectite clays is emerging. Monovalent cations, with the possible exception of Li’, probably remain coordinated to basal oxygen surfaces irrespective of hydration state while divalent cations are fully solvated when smectites are hydrated but move toward basal surfaces during dehydration. ESR studies of Cu(I1)-doped Mg-montmorillonites and hectorites have shown that in the one-layer hydrate Cu(I1) is coordinated to four water molecules and the Cu(I1)-water complex has tetragonal symmetry with the principal axis perpendicular to the ab plane of the clay layer^.^ The ESR indicates that the Cu(I1) is in axial symmetry even at room temperature. In the two-layer hydrates, the Cu(I1) is octahedrally coordinated to six water molecules. The ESR spectrum of this hydrate is a symmetrical line at room temperature, indicating that the Cu( 11)water complex is able to freely tumble and therefore the g and ~
(13)
~~
Skipper, N. T.; Soper, A. K.; McConnell, J. D. C. J . Chem. Phys.
1991, 94, 5751. (14)
Skipper, N. T.; Refson, K.; McConnell, J. D. C. J . Chem. Phys. 1991,
94, 7434. (15) (16)
Luca, V.; Brown, D. R.; Kevan, L. J . Phys. Chem. 1991, 95, 10065. Luca, V.; Chen, X.; Kevan, L. Chem. Mater. 1991, 3, 1073.
0022-365419212096-2645%03.00/0 0 1992 American Chemical Society
2646 The Journal of Physical Chemistry, Vol. 96, No. 6, 1992
T
M"+XH ~ O
Figure 1. Idealized structure of smectite clays where each layer is composed of three sheets. The center sheet is called an octahedral sheet where the oxygens form an octahedron about a center metal ion. Only the oxygens are shown by open circles. The solid circles are hydroxyls. The two outer sheets in each layer are tetrahedral sheets. The arrows show pseudohexagonal cavities.
hyperfine anisotropies are averaged. Bleam" has shown that the electronegativity pattem above the basal surface of smectites depends on the position (in the tetrahedral versus in the octahedral sheet) at which layer charge is created and according to several s t u d i e ~ , ' the ~ , ~position ~ in the layer lattice at which negative layer charge originates can influence the hydration properties of the smectite clays. In the present work we study the solvation complex of Cu(l1) doped into Mg(1I)-exchanged montmorillonite, synthetic beidellite, synthetic fluorohectorite, and synthetic hydroxyhectorite. A schematic of smectite clay structure is shown in Figure 1 and described in the caption. Montmorillonite is a dioctahedral smectite in which layer charge is created by the substitution of Mg(I1) for AI(II1) in the octahedral sheet. Beidellite is also a dioctahedral smectite but in this case layer charge is created by the substitution of Al(II1) for Si(1V) in the tetrahedral sheet. Fluorohectorite and hydroxyhectorite are trioctahedral smectites in which layer charge is created by the substitution of Li(1) for Mg(I1) in the octahedral sheet. Fluorohectorite contains fluorine atoms on the edges of the octahedrons and in hydroxyhectorite those sites are occupied by hydroxyl groups instead of fluorine atoms. In this study we have compared the type of Cu(I1) solvation complex formed in these clays when they are wet with water and when they are equilibrated at 100%relative humidity and simultaneously the effect of the dehydration prior to rehydration has been examined. The clays have been brought to these hydration states after vacuum drying for 2 h the Cu(I1)-doped smectites at room temperature or at 100 OC.
Experiamatal Section The montmorillonite used was a natural sample STx-1 supplied by The Clay Resource Repository, University of Missouri. Each gram of the crude clay was stirred with 0.05 M HCl to remove carbonate impurities and was filtered and dispersed in 1 M NaCl solution for 12 h. This suspension was filtered and the montmorillonite stirred with 1 M NaCl solution for about 12 h. After this treatment was repeated twice, the clay suspension was dialyzed to remove excess cations. Fluorohectorite was prepared by a similar method to that of Granquist and Pollack.'* A wet gel was prepared with a composition corresponding to the chemical formula Nao.5(Mg2,,,Lio.s)Si,OloF*: The synthas was as follows: Tetraethoxysilane (41.66g, Baker, 99.6%)was added to 100 mL of ethanol and stirred at about 60 OC for 3 h. To this solution was added 25.41 g of MgCIz.6H20 ~~~
~
Bleam, W. F. Clays Clay Miner. 1990, 38, 527. (18) Granquist, W. T.; Pollack, S. S . Am. Miner. 1967, 52, 212. (17)
Comets et al. (Baker, 99.7%) dissolved in 200 mL of deionized water. This ethanol/water solution was stirred for at least 4 h and then an NaOH solution was added dropwise until a final pH of 9 was reached. After the solution was stirred for 12 h the precipitate obtained was filtered and washed. The sol* were resuspended in deionized water and 0.648 g of LiF (Fisher, AR) and 3 g of NaF added. This suspension was stirred for about 12 h and then refluxed for 7 days. X-ray powder diffraction patterns of the suspension removed periodically from the reaction vcssel indicated that crystallization of fluorohectorite was essentially complete after 4 days of reflux. Hydroxyhectorite was prepared by a similar method except that LiCl and NaCl were used in place of LiF and NaF. Beidellite was prepared by a method similar to that of Klo~ r 0 g e e . l ~A gelzowas made with composition corresponding to the Chemid formula Nao.a,AL(si,.,,Ab.s)O~*(~H),.TO 100 g of gel, 8.03 g of NaNOJ was dissolved in 50 mL of deionized water. This solution was added to a solution of 250.4 g of AI(N03),.9H20 in 550 mL of water. To this was added 230 mL of ethanol and then 218.7 g of tetraethoxysilane (Aldrich). This solution was stirred for 3 h to achieve hydrolysis of the tetraethoxysilane and then 230 mL of 25%N H 4 0 H was added to the continuously stirred solution to precipitate the hydroxides. The viscous gel was homogenized in a blender and*subsequentlydried at 80 OC for 14 h, 120 OC for 8 h, 150 OC for 14 h, and finally 400 OC for 4 h. The finely ground dry gel (4 g) and 13 mL of 0.15 M NaOH were placed in a gold tube that was fused at one end. The top end of the tube was fused while the tube was kept cool in an ice bath. The gold tube when seaIed had a final volume of about 24 mL. The sealed tube was weighed and then placed in a 1-L Parr bomb containing water and the bomb was heated at 330 OC for 7 days. At the end of the reaction, the gold tube was weighed to confirm that the tube had not ruptured. The product was dialyzed to remove exions and dried. The X-ray powder diffraction pattern of the synthetic beidellite was recorded on a Philips PW 1840 diffractometer and good agreement was obtained with the d spacings of the hk reflections reported recently by Kloprogge.I9 The four smectites were converted to the Mg(I1)-exchanged form using MgC12 by a similar method used to exchange Na(1) into the montmorillonite. Replacement of some of the Mg(I1) by Cu(I1) was achieved by dispersing a known amount of the Mg(I1)-smectite in deionized water and adding a calculated amount of CuC12. The suspension was stirred overnight and then filtered and the smectite was dialyzed. The dialyzed smectites were dried in air at about 40 OC. Generally, enough CuClz was added to exchange about 5% of Mg(I1) by Cu(I1). Dilute concentrations of paramagnetic Cu(I1) were used to obviate the possibility of spin-exchange interactions between Cu(I1) cations. In some cases samples containing higher Cu(I1) loadings were also prepared. The Cu(I1)-doped Mg(I1)-exchanged smectites were treated in two ways prior to exposure to vapor or liquid water. One set of samples were dried for 2 h at 100 OC undw a dynamic vacuum of 5 mTorr. The second set of samples were dried under a vacuum of 5 mTorr without heating. These two sets of samples were then either water-soaked by adding about 2 mL of water or D 2 0 per milligram of smectite or equilibrated in an atmosphere of 100% H20or D20. Electron spin resonance spectra were recorded at 77 K on a modified Varian E-4 ESR spectrometer. Electron spin echo modulation data were recorded at 4 K on a home-built spectrometer described previously.z' Three-pulse stimulated e c h w were recorded with a a / 2 - ~ / 2 ~ /pulse 2 sequence and the echo was detected as a function of T,the time between the second and third pulses. The time between the first and second pulse, T , was chosen to maximize modulation from 'Li or 2H. Two-pulse echoes (19) Kloprogge, J. T.; Jansen, J. E. H.; Gum,J. W.Clays Clay Miner. 1990, 38.409. (20) Hamilton, D. L.;Henderson, C . M.E.Mineral. Mag. 1968,36,832. (21) Kevan, L.; Bowman, M. K.;Narayana, P. A.; Boeckman, R. K.; Yandanov, V. F.; Tsvetkov, Yu.D. J. Chem. Phys. 1975, 63, 409.
The Journal of Physical Chemistry, Vol. 96, NO. 6, 1992 2647
Solvation of Cu(I1) Cations
TABLE I: ESR Parameters of CuMg-beid and CuMg-mont sample T,K gu" A,
CuMg-mont
CuMgbeid, v-d,b RTd water-soaked
77 2.39 300 g,, = 2.18 2.34 77 2.40 300 gi, = 2.17 2.34
100% RH' d(001) = 1.9 nm CuMgbeid, v-d, 100 "C (low Cu) water-soaked
100% RH d(001) = 1.9 nm CuMgmont, v-d, RT water-soaked
2.34 2.36 77 2.34 2.36
141 2.09 139 2.08
2.07 c 2.07
c
c
7-
CuMg-fluorohect
kA/
2.40 138 2.08 gi, = 2.19 2.40 141 2.08 gi, = 2.19
C
77 2.28 300 C 77 2.27 300 C
100% RH d(001) = 2.0
CuMg-beid
180 2.09
155 2.08 2.07 178 2.07 c 2.08
77 2.37 300 C 77 2.35 300 C 77 300 77 300
145 2.08
150 100 150 100
77
100% R H d(001) = 1.9 nm CuMg-beid, v-d, 100 OC (high Cu) water-soaked
100% RH d(001) = 2.0 nm CuMgmont, v-d, 100 O water-soaked
A
g,
176 2.07 2.08 175 2.06 c 2.08
c
"Units cm-' with estimated error of k5 X cm-I. Vacuum-dried, temperature. Due to paramagnetic impurities, g,,and A,, cannot be determined. dRoom temperature. 'Relative humidity.
200G,
V
Figure 2. ESR spectra at 300 K of Cu(I1)-doped Mg(I1)-exchanged smectites (=5% Cu) vacuum-dried at room temperature, then watersoaked: (a) CuMg-montmorillonite, (b) CuMg-beidellite, (c) CuMgfluorohectorite, and (d) CuMg-hydroxyhectorite.
TABLE II: ESR Parameters of CuMg-fluorohect and CuMghydroxykct sample CuMg-f-htct, v-d,b RTd water-soaked 100% RH' d(001) = 2.0 nm CuMg-f-hect, v-d, 100 OC water-soaked 100% R H d(001) = 2.0 nm CuMg-h-hect, v-d, RT water-soaked 100% RH CuMg-h-hKt, v-d, 100 "C water-soaked
100% RH
TK 77 300 77 300
gl,
AIIO
g,
2.41
120
C
c
2.41
123
C
C
2.08 2.07 2.08 2.08
2.33 2.33 2.33 2.33
176 166 176 167
2.06 2.06 2.06 2.06
77 300 77 300
2.36 2.35 2.34
154 154 160
C
C
2.07 2.07 2.07 2.07
77 300 77 300
2.35 2.35 2.35 2.36
161 147 151 140
2.07 2.07 2.07 2.07
77 300 77
300
"Units cm-l with estimated error of f5 X cm-l. Vacuum-dried, temperature. Due to paramagnetic impurities, g,, and AI, cannot be determined. dRoom temperature. e Relative humidity.
were recorded to observe 'Li modulation in the absence of D20. Simulations were made in terms of N equivalent nuclei at a distance R and isotropic hyperfine coupling A using a spherical approximation.22 N is determined to the nearest integer and R is usually determined to fO.O1 nm. Results 1. Room-Temperature Electron Spin Resonance Spectra.
Room-temperature ESR spectra of vacuum-dried water-soaked smectite samples containing nominally 5% Cu are shown in Figure 2. The ESR parameters of the smectites are summarized in Tables I and 11. Cu(I1)doped Mg(I1)-exchanged montmorillonite (CuMg-mont) gives a single isotropic ESR line with g = 2.18 (22) Narayana, P. A.; Kevan, L. Magn. Reson. Reu. 1983, I , 234.
CuMg-hydroxyhect
b.003,
200G,
2648 The Journal of Physical Chemistry, Vol. 96, No. 6, I992
Comets et al.
TABLE III: Shnuhti-
of 'IBrrc-p.bc ISEM Data for CnMg-beid
and CuMg-mont
field, sample CuMg-beid, v-d, RT waier-soaked 100%RH CuMg-beid, v-d, 100 O C (low Cu) water-soaked 100% RH
CuMg-beid, v-d, 100 OC (high Cu) water-soaked 100%RH CuMg-mont, v-d, RT water-soaked CuMg-mont, v-d, 100 OC water-soaked 100%RH
G
simulation parameters N R,nm A,MHz
3191 3204
8 8
0.27 0.27
0.24 0.26
3215 3314 3219 3304
4 2 4 2
0.29 0.25 0.29 0.26
0.16 0.35 0.16 0.35
3210 3207
6 4
0.27 0.26
0.25
3207
12
0.29
0.30
3215 3196
6 6
0.34 0.33
0.02
0.30
0.05
Figure 4. ESR spectra at 77 K of CuMg-beidellite dehydrated under vacuum at 100 OC and ambient relative humidity with different Cu(I1) contents: (a) 2.5%Cu(II), (b) 5% Cu(II), (c) 10% Cu(II), and (d) 50% Cu(I1).
200G,
-
u
3207
b
z
P O [ \ -1
m 0.8 H = 3207G
z
T = 0.27
fiS
t( 0.4
Figure 6. (a) ESR spectrum at 77 K of CuMg-montmorillonite (4% Cu) dehydrated under vacuum at 100 OC, then water-soaked. (b) Experimental (-) and simulated (---) ESE modulation spectrum of the sample from (a), recorded at 4 K.
I-
z O
J
0
1
2
3
4
5
T, fis
Figure 5. (a) ESR spectrum at 77 K of CuMg-montmorillonite (4%
Cu) vacuum-dried at room temperature, then water-soaked. (b) Experimental (-) and simulated (---) ESE modulation spectrum of the sample from (a), recorded at 4 K. with the amount of Cu(I1) added to the Mg-beid suspension and given as percentage of cation-exchange capacityz3(total amount of cations that can be exchanged in the clay). As the Cu(I1) content increased the species A signal intensity increases relative to the species B signal. At 50% of the exchange capacity only a single broad resonance is observed. 2. Tbree-Puke ESEM Dab. CuMg-mont. The 77 K ESR spectrum of a sample of CuMg-mont that was vacuum-dried before being watcr-aoaked is shown in F w 5a. Only one Cu(I1) species is evident and ESEM data w a reGorded at 3207 G (Figure 5b). The ESEM data could be best simulated with N = 12, R = 0.29 nm, and A = 0.30 MHz (Table 111). This result is consistent with that obtained by Brown and Kwan.% No echo was obtained for the vacuum-dried sample that was equilibrated at 100% relative humidity D20. When the same CuMg-mont sample is dried at 100 OC under vacuum before being water-soaked the 77 K ESR spectrum of (23) Van Olphen, H. An Introduction to Clay Colloid Chemistry; Wiley-Interscience: New York, 1977; p 65. (24) Brown, D.R.;Kevan, L. J. Am. Chem. SOC.1988, 110, 2743.
Figure 6a is obtained. The ESEM data recorded at 3215 G was dramatically different to that recorded for the vacuum-dried sample (Figure 6b)and could be well simulated using N = 6, R = 0.34 nm, and A = 0.02. The CuMpmont sample dried under vacuum and then equilibrated at 100% relative humidity gave an ESEM data that could be well simulated using a similar model to the sample that had been dried at 100 OC and then watersoaked. CuMg-beid-low[Cu). The room-temperature and 77 K ESR spectra of a vacuum-dried water-soaked sample of CuMg-beid containing less than 10% Cu is shown in Figure 7a,b. Despite the fact that two species are observed for this sample at room temperature (Figure 2c), only one Cu(I1) species is observed at 77 K. When the sample is cooled to 77 K the Cu(I1)-water complex is no longer free to tumble and anisotropies are no longer averaged. Therefore, Cu(I1) species A giving the single symmetric line at room temperature give a spectrum characteristic of Cu(I1) in an axial environment at 77 K. This spectrum must be superimpoeed on the spectrum of species B which is observed at room temperature. It is difficult to know how the intensities of species A and B vary with temperature but the fact that a single spectrum is observed at 77 K indicates that there is far less of one species than the other and/or both species have a similar coordination environment. Therefme, ESEM data were recorded at otle field value of 3191 G. This ESEM data is shown in Figure 7c and could be we4l simulated with N = 8, R = 2.73 nm, and A = 0.24 MHz. This indicates that there are four water molecules in the first
The Journal of Physical Chemistry, Vol. 96, No. 6,1992 2649
Solvation of Cu(I1) Cations
-
s’”in
t = 0.27 pS
Shell
z 0
1
N R nm A,,,, MHz 6 0.27 0.25
z
m 0.8
M
11 1
Shell
W
I1 I
2
3
4
I
0
N R nm A,,,, MHz 8 0.27 0.24
1
2
3
4
5
Figure 9. (a) ESR spectrum at 77 K of CuMg-beidellite (>lo% Cu or high [Cu]) dehydrated under vacuum at 100 O C , then water-soaked. (b) Expzrimental (-) and simulated (---) ESE modulation spectrum of the sample from (a), recorded at 4 K.
T, its
Figure 7. ESR spectra, recorded (a) at 300 K and (b) at 77 K of CuMg-beidellite (
