J. Phys. Chem. 1988,92, 337-339
337
lengths from that of atoms in the main site. With respect to this last point, we note that the 535- and 574-nm emission bands mentioned before, and assigned to transitions of Pd in secondary sites, have a similar red shift (-50 meV) with respect to the main bands. On the other hand, we remark that in laser-induced fluorescence the 4d85s2,3F3state is populated directly from the 4d95s1,3D3state (Figure 14), whereas in the broad-band xenon arc excitation (Figure 7 in paper 15) this state is populated from higher energy states having a 4d95p1 electronic configuration. Finally, the absence of any anti-Stokes emission following excitation at 514.5 or 457.9 nm, implies a high stability for the matrix-entrapped 4d95s1,3D3state, which cannot be photolytically transformed to the normal 4dl0,'So ground state. This observation is further strengthened by the photolytic studies of Pd/Ar samples described earlier. Thus following long periods of irradiation at different excitation wavelengths, neither the intensity nor spectral profile of the Pd/Ar atomic absorption spectra show any measurable change.
those previously reported by Kolb and co-workers,6 who identify a narrow-band emission spectrum in N2/Ne matrices (quite similar to that reported here) with the Pd(N2) molecule and a narrow and intense band a t 827 nm with Pd(N2)2. Using the existing theoretical information on the M(N2) and the isoelectronic M ( C 0 ) molecules, we identify two stable electronic states of the Pd(N2) molecule coisolated in Ar matrices. The 'E+ Pd(N2) molecule corresponds to that identified by IR and Raman spectroscopy, whereas the 3Z+ Pd(N,) molecule displays an electronic specttum quite similar to that previously associated with the 4d95s1, 3D3 metastable state of the isolated Pd(4d10,'So) atoms in an Ar matrixes Finally, in neat Ar matrices the Pd(4d95s', 3D3)atom has been identified as a stable species coexisting with Pd(4dIo, 'So) atoms. There still remains the question as to whether the stabilization of Pd(3D3) atoms is the result of the presence of coisolated N2 impurities in the Ar matrix, or rather a manifestation of the interaction with the matrix cage in a special trapping site.
Conclusion From the IR and Raman spectroscopic indentification of the Pd(N2), Pd(N2)2,and Pd(N2)3molecules isolated in N2/Ar and N 2 matrices, it is found that the optical absorption spectra of these species occur in the same spectral region, making their individual assignment difficult. The corresponding emission spectra consist of extremely weak and broad bands (bandwidth 400, 450, and 1950 cm-I). The latter observation contrasts substantially with
Acknowledgment. The generous financial support of the Natural Sciences and Engineering Research Council of Canada is greatly appreciated. A 3M research award to J.G.P. is acknowledged with gratitude. The technical assistance of Dr. Mark Baker with the laser-induced emission experiment proved t o be most valuable. Registry No. Pd, 7440-05-3; Pd(N2), 40810-26-2; Pd(N2)2, 4081027-3; Pd(N2)3,41 133-52-2; Ar, 7440-37-1; N2,7727-37-9.
Low-Temperature Cross-Polarizatlon/Magic Angle Spinning 13C NMR of Solid Methane Hydrates: Structure, Cage Occupancy, and Hydration Numbert J. A. Ripmeester* and C. I. Ratcliffe Division of Chemistry, National Research Council of Canada, Ottawa, Ontario K I A OR9, Canada (Received: May 26, 1987; In Final Form: July 31. 1987)
13CNMR spectra of methane trapped in a solid type I hydrate and in a mixed methane/propane type I1 hydrate were studied at -80 OC under cross-polarization and magic angle spinning conditions. Type I could be distinguished from type I1 structures due to the chemical shift pattern of methane trapped in the large and small cages. In agreement with the statistical theory of clathrate hydrate formation, methane was found to occupy both large and small cages in each structure to a significant extent. In the case of type I methane hydrate, the distribution of methane over the two sites was obtained, and together with a previous determination of the ~1 potential of the empty lattice with respect to ice, a hydration number of 6.05 & 0.06 was determined.
Introduction Gas hydrates, ice-like lattices with guest molecules trapped in molecule-sized voids, have proved to be difficult subjects of study over the years, not in the least due to their nonstoichiometric nature.'J Once structures were established by X-ray diffraction t e c h n i q ~ e s , and ~ - ~ the theory of localized surface adsorption was adapted to describe the thermodynamic proper tie^,^^^ one of the problems has been to develop techniques to accurately measure properties which can be used to test the clathrate solid solution theory. Considerable progress has been made in measuring properties such as compositions,'0." hydration and relative site o c ~ u p a n c i e s , ' ~in~ 'addition ~ to the traditional decomposition pressure m e a ~ u r e m e n t s . ' ~ J ~ Of special interest are the gas hydrates with hydrocarbon guests, as it has been suggested that large quantities of natural gas are trapped in hydrate deposits distributed in sediment below the sea b o t t ~ m . I ~ *A' ~recent publication from this laboratory19 reported tNRCC No. 27829.
0022-365418812092-0337$01.50/0
results obtained from the application of a number of laboratory techniques, such as 'H and 13CNMR, dielectric relaxation, heat ~
~~~
~
~
~
~
(1) Davidson, D. W. In Warer. A Comprehensive Treatise; Franks, F., Ed.; Plenum: New York, 1972; Vol. 2. (2) Cady, G. H. J . Chem. Educ. 1983, 60,1915. (3) Clausen, W. F. J. Chem. Phys. 1951, 19, 259. (4) Von Stackelberg, M.; Muller, H. R. J. Chem. Phys. 1951, 19, 1319. (5) Clausen, W. F. J . Chem. Phys. 1951, 19, 1425. (6) Mfiller, H. R.; von Stackelberg, M. Natunvissenschaften 1952,39, 20. (7) Pauling, L.; Marsh, R. E. Proc. Natl. Acad. Sci. U.S.A. 1952, 38, 112. (8) Barrer, R. M.; Stuart, W. L. Proc. R. SOC.London, A 1957,243, 172. (9) Van der Waals, J. H.; Platteeuw, J. C. Adu. Chem. Phys. 1959, 2. 1 . (10) Handa, Y.P. J . Chem. Thermodyn. 1986, 18, 915. (11) Dharmawardhana, P. B.; Parrish, W. R.; Sloan, E. D. Ind. Eng. Chem. Fundam. 1980, 19, 410. (12) Cady, G. H. J . Phys. Chem. 1983, 87, 4437. (13) Ripmeester, J. A.; Davidson, D. W. Bull. Magn. Reson. 1980, 2, 139. (14) Davidson, D. W.; Handa, Y. P.; Ripmeester, J. A. J . Phys. Chem. 1986, 90,6549. (15) Barakhov, S. P.; Sawin, A. Z . ; Tsarev, V. P. Zh. Fir. Khim. 1985, 59, 1039. (16) Holder, G. D.; Gorbin, G.; Papadopoulos, K. D. Ind. Eng. Chem. Fundam. 1980, 19, 282.
Published 1988 by the American Chemical Society
338 The Journal of Physical Chemistry, Vol. 92, No. 2, I988 capacity, and X-ray powder diffraction on a natural gas hydrate sample obtained from sediment in the Gulf of Mexico. The gas present as guest material in this hydrate is thought to be mainly of thermogenic origin?O the main components being methane and propane, with smaller proportions of isobutane, ethane, and carbon dioxide. X-ray powder diffraction showed the hydrate to have ~ * ~a lattice parameter von Stackelberg's type I1 cubic s t r ~ c t u r ewith of 1.724 f 0.007 nm at 100 K. Other natural gas hydrates can ~ ~recently discovered be expected to have the trpe I ~ t r u c t u r eor~the type H structure.21 The 13CN M R spectrum of the Gulf of Mexico hydrate sample, obtained under low-temperature magic angle spinning conditions, showed the main methane resonance line to occur at a chemical shift of -2.8 ppm. A second weak line at -6.4 ppm was attributed to methane in the large type I1 cages. The presence of a cage-dependent chemical shift for enclathrated methane therefore suggests that 13CN M R should provide a direct, nondestructive technique for the determination of natural gas hydrate structures. There are indeed few documented properties that can be used to identify hydrate structures. X-ray powder diffraction, the only technique in common use, does not always give unequivocal results for complex samples. Broad-line N M R , although it can be used to identify clathrate hydrates,32 including those of natural gas,33 does not give information on structure type. In this contribution, we present results obtained with CP/MAS I3C N M R techniques on model type I and type I1 hydrate structures and confirm that the cage-dependent I3C chemical shift of enclathrated methane can be used to determine the hydrate structure. Quantitative determination of the gas distribution over the different cage sites, together with the previously determined difference in p potential between ice and the empty type I hydrate lattice,14 was then used to determine the hydration number of methane hydrate.
Experimental Section Samples were prepared by sealing in Pyrex tubes excess finely powdered ice and appropriately 13C-labeled hydrocarbon gases (MSD Isotopes) and conditioning the samples for -3 months at -40 "C and then for 1 week at -13 OC. The hydrate sample meant to simulate type I1 behavior was made with a gas mixture consisting of 30% [2-'3C]propane and 70% [I3C]methane. After conditioning, the sample tubes were cooled slowly to 77 K and opened, and the powdered material was transferred to Delrin sample spinners. These were then placed into the stator assembly of a Chemagnetics probe previously cooled to - 4 0 OC. 13C N M R spectra were recorded on a Bruker CXP-180 spectrometer at a frequency of 45.28 MHz. Single-shot cross-polarization (CP) sequences were used with contact times of 1-40 ms. The radio-frequency field strengths were 40 kHz, and the magic angle spinning (MAS) rate was 2-3 kHz. Chemical shifts were measured taking the Delrin peak to be This is a room-temperature value; however, the at 89.5 relative shifts are more important than the absolute shifts for the purposes of this paper. Results and Discussion The 13CCP/MAS spectrum of [13C]methane hydrate at -80 "C is shown in Figure 1 (bottom). There are resonance lines at -2.8 and -5.2 ppm, and as the large cages outnumber the small cages by a factor of 3 in type I hydrate, the more intense upfield line can be assigned to methane in the large 14-hedral cage. This parallels the assignment of the Iz9XeN M R lines observed for Xe (17) Trofimuk, A. A.; Cherskii, N. V.; Tsarev, V. P. Priroda (Moscow) 1979, 1, 18. (18) Kvenvolden, K. A.; McMenamin, M. A. US.Geol. Sum., Circ. 1980, 825. (19) Davidson, D. W.; Garg, S. K.; Gough, S. R.; Handa, Y . P.; Ratcliffe, C.I.; Ripmeester, J. A.; Tse, J. S.;Lawson, W. F. Geochim. Cosmmhim. Acra 1986. 50. 619. (20) Brooks, J. M.; Kennicutt, M. C.; Fay, R. R.; McDonald, T. J.; Sassen, S . Science 1984, 225, 409. (21) Ripmeester, J. A.; Tse, J. S.; Ratcliffe, C. I.; Powell, B. M. Nature (London) 1987, 325, 135.
Ripmeester and Ratcliffe
Structure U hydrate 70%methane, 3o%propane
I3C Enriched methane and propane
I
I
I
I
40
20
0
-20
ppm
Figure 1. I3C CP/MAS NMR spectra at -80 OC: bottom, type I methane hydrate; top, type I1 hydrate made from 30:70 propane/methane mixture.
TABLE I: I3C Chemical Shift Summary "CH4 type I large cage type I small cage type I1 large cage type I1 small cage gas phase
-5.21 -2.84 -6.27 -2.73 -2.3' -7.0b
type I1 large cage solution
18.5 15.9"
solution
CH3"CH2CH3
'Reference 3 1. Reference 36. hydrate, where the large cage resonance is upfield from the small cage resonance by 90 ppm.13s14 The spectrum of the type I1 propane/methane hydrate is shown in Figure 1 (top). The line at 18.5 ppm is due to the labeled propane C2 carbon in the large 16-hedral cage. The propane methyl carbon lines are too weak to be seen. The large-amplitude line at -2.7 ppm can be assigned to methane in the small type I1 cage, and the weak line at -6.3 ppm therefore represents methane in some of the large type I1 cages. Again this assignment parallels that for the 129Xeshifts in mixed Xe-containing type I1 hydrates, where the large cage resonance occurs some 150 ppm upfield from the small cage resonance.22 The 13C chemical shifts are summarized in Table I. The small-cage methane chemical shifts are nearly identical, as both small cages are pentagonal dodecahedra of almost identical dim e n s i o n ~ .The ~ ~ ~14~ and ~ 16-hedral cages, which occur in type I and type I1 hydrate, respectively, are of different size and shape and give rise to quite different chemical shifts for trapped methane. The methane chemical shift pattern therefore is a unique indicator of the structure type. The spectrum shown in Figure 1 (top) is very much like the one reported previously for the natural gas hydrate sample from the Gulf of Mexico, confirming the tentative methane line assignment proposed at that time.Ig The fact that solid-state polarization transfer techniques are used in obtaining the spectral data ensures that no signals arise from pockets of trapped liquid or gas. The cross-polarization technique depends on "solidlike" behavior; that is, there must be some static nuclear dipolar coupling between IH and I3C nuclei. The efficiency of the polarization transfer depends, in part, on (22) Ripmeester, J. A.; Ratcliffe, C. I., unpublished results. (23) McMullan, R. K.; Jeffrey, G . A. J . Chem. Phys. 1965, 42, 2725. (24) Mak, T. C.W.; McMullan, R. K. J . Chem. Phys. 1965, 42, 2732.
Solid Methane Hydrates
The Journal of Physical Chemistry, Vol. 92, No. 2, 1988 339
+A
I I
I
2c!
1
I
I
20 tcp/msec
10
0
I
I
30
40
Figure 2. 13Cline intensity variation as a function of cross-polarization time I,, for the mixed propane/methane hydrate at -80 O C .
the strength of this coupling. In "rigid" systems, such as pure solid hydrocarbons, polarization transfer can be complete in times on the order of 1 ms or less.25 In the presence of molecular motion, the polarization transfer can take a significantly longer time. In highly mobile systems such as encaged guest molecules, intramolecular dipolar couplings may be completely averaged. This is certainly the case for encaged methar~e.'~,'~ The polarization transfer therefore depends to a good extent on intermolecular dipolar couplings, that is, between a I3C nucleus in one cage on the one hand and cage water protons, or guest protons, in neighboring cages on the other. The true integrated signal intensities can be used to give information on the guest distribution in the different cages. The signal intensities of propane/methane hydrate as a function of cross-polarization time are shown in Figure 2. The 13Cpolarization reaches a maximum value around 7-8 ms, before a slow decay sets in. Between 15 and 40 ms the ratio between the two larger line intensities is essentially constant and can be taken to reflect the ratio of the population in the two sites. After allowance has been made for the fact that there are twice as many small cages as large ones, occupancy ratios can be calculated. The following occupancy ratios can be determined from the data in Figure 2. BL(CH4)/Bs(CH4) = 0.14 and BL(propane)/ Bs(CH4) = 1.14, where BL and Os are large- and small-cage occupancies. These values are indeed very close to values extrapolated from calculated results reported by van der Waals and Platteeuwg for a hydrate in equilibrium with a propane/methane gas mixture under a pressure of 15 atm at -3 OC: BL(CH4)/Bs(CH,) = 0.13 and BL(propane)/Bs(CH,) = 1.10. Our results present the first direct proof that methane does indeed occupy the large structure I1 cage to a measurable extent, in good agreement with solid solution theory predictions. In the case of methane hydrate, knowledge of the occupancy ratio of the large and small cages can lead to a value for the hydration number, a quantity difficult to measure by direct met hods.
-
-
1v2~10
(25) Alemany, L. B.; Grant, D. M.; Pugmire, R. J.; Alger, T.D.; Zilm, K. W. J. A m . Chem. SOC.1983, 105, 2133.
In the absence of guest-guest interactions and host-lattice distortions, the chemical potential of the water molecules in a structure I hydrate is given by',8*9 RT p,(h) - pw(ho) = -[3 In (1 - 6,) - In e,] 23 where p,(ho) is the chemical potential of water molecules in the lattice with empty cavities. When the hydrate is in equilibrium with ice, the left side of eq 1 becomes p,(ice) - pw(ho)= -Apwo, where A&' is the chemical potential of the empty hydrate lattice relative to ice. Recently, A h o was determined to be 1297 f 110 J/mol for a xenon hydrate sample prepared under equilibrium conditions at 0 OC,I4 in good agreement with values determined by more empirical methods. The integrated line intensities (Figure 1, bottom) again allow the relative populations of the two sites to be determined. After allowing for the fact that there are 3 times as many large cages as small ones in type I hydrate, the occupancy ratio Bs/BL = 0.916. This, together with the above Apwo value, gives absolute occupancies OL = 0.97 and Bs = 0.89. The hydration number n can then be found from n = 23/(38, + Os), giving a value of 6.05 f 0.06. The error limits reflect the uncertainty in A&', which is thought to be more important than the fact that the sample was conditioned at -13 OC rather than at 0 OC. Other recent values for n, determined by using direct as well as indirect methods, are 5.77 f 0.12,26 7,277.18,286.00 f 0.16,297.4,306.3,30and 6.00 f O.O1.lo Our value, in agreement with other values lying between -5.8 and 6.3, suggests an incomplete filling of both 12-hedral and 14-hedral cages. These results have illustrated that, in addition to providing fundamental information on gas hydrate composition and the general correctness of solid solution theory as applied to gas hydrates, I3C N M R techniques can be used as an analytical tool to assess natural gas hydrate structure. Although 13C-enriched materials were used in this study, our previous work19 has shown that the technique can be applied to natural samples as well. One desirable experimental improvement would be the use of a larger volume sample spinner for natural samples. Due to extensive molecular motion of the guest molecules, only modest spinning rates and decoupling field strengths should be required to give well-resolved, spinning-sideband-free spectra. Registry No.
Methane hydrate, 14476-19-8; propane hydrate,
14602-87-0. (26) Glew, D. N. J. Phys. Chem. 1962, 66, 605. (27) Roberts, 0. L.; Brownscombe, E. R.; Have, L. S.;Ramser, H. Pet. Eng. 1941, 12, 86. (28) Frost, E. M.; Deaton, W. M. Oil Gas J . 1946, 45, 170. (29) Gallaway, T.J.; Ruska, W.; Chappelear, P. S.;Kobayashi, R. Ind. Eng. Chem. Fundam. 1970, 9 , 237. (30) De Roo,J. L.; Peters, C. J.; Lichtenthaler, R. N.;Diepen, G. A. M. AIChE J . 1983, 29, 651. (31) Breitmaier, E.; Voelter, W. "C N M R Spectroscopy; Verlag Chemie: Weinheim, 1978. (32) Davidson, D. W.; Ripmeester, J. A. In Inclusion Compounds; Atwood, J. L., Davies, J. E. D., MacNicol, D. D., Eds.; Academic: London, 1984; Vol. 3. (33) Davidson, D. W.; Garg, S. K.; Gough, S. R.; Hawkins, R. E.; Ripmeester, J. A. Can. J. Chem. 1977, 55, 3641. (34) Zilm, K. W.; Alderman, D. W.; Grant, D. M. J. Mugn. Reson. 1978, 30, 563. (35) Garg, S. K.; Gough, S . R.; Davidson, D. W . J . Chem. Phys. 1975, 63, 1646. (36) Jameson, A. K.; Jameson, C. J. Chem. Phys. Lett. 1987, 134, 461.