Time-Resolved in-Situ Experiments on the Crystallization of Natural

---

6412

J. Phys. Chem. 1996, 100, 6412-6414

Time-Resolved in-Situ Experiments on the Crystallization of Natural Gas Hydrates Carolyn A. Koh* Department of Chemistry, King’s College London, Strand, London WC2R 2LS, U.K.

Jeffrey L. Savidge Basic Research Department, Gas Research Institute, Chicago, Illinois 60631

Chiu C. Tang Daresbury Laboratory, Warrington, WA4, U.K. ReceiVed: December 22, 1995X

Time-resolved X-ray synchrotron techniques have been used, for the first time, to capture the crystalline structural dynamics of common hydrate structures during formation. Measurements were performed on carbon dioxide and propane hydrates at 273.8 K and 3.29-0.42 MPa. Our results indicate that the in-situ crystal lattice structure is in agreement with previous low-temperature single-crystal data, although the in-situ unit cell parameter is lower than those based on low-temperature, single-crystal measurements. We show the effects of two kinetic inhibitors, polyvinylpyrrolidone and tyrosine, on the energy dispersive spectra obtained for the inhibited crystal hydrate.

Introduction Gas hydrates are solid inclusion compounds. Their existence has been reported in the literature since the early nineteenth century.1-3 Previous hydrate work in the literature has concentrated on macroscopic measurements to determine the equilibrium aspects of hydrate formation,4,5 perform fundamental model calculations,6-8 and obtain spectroscopic information at low temperatures.9-11 Our early infrared studies12,13 and complementary molecular simulations14 suggested the presence of hydrogen-bonded water rings, similar to those suggested by Benson.15 Time-resolved in-situ structural data have not been obtained to date for these systems. X-ray synchrotron measurements provide fast diffraction patterns which have enabled us to extend single-crystal X-ray studies on gas hydrates to dynamic in-situ measurements at temperatures and pressures close to industry operating conditions. Our new results indicate this is important in determining kinetic mechanisms of crystallization control. Experimental Section Energy-dispersive diffraction was used in this study, where the intensity peaks were obtained as each set of lattice planes of interplanar spacing (d) from the crystalline sample diffracts X-ray of energy (E), given by E (keV) ) 6.199/(d sin θ). The incident X-ray beam was reduced by a series of slits and finally defined by a 1 mm collimator, and the diffracted X-rays were detected by an Eg & G ORTEC germanium solid state detector with an energy resolution of ∆E/E ≈ 0.007 at 50 keV. The detector was set at a fixed angle (2θ) of 5.0402 ( 0.0005°, and the collected photons were electronically separated by the instrument multichannel analyzer. A radioactive dial source was used to calibrate the energy scale of the energy-dispersive detector system. Energies were at the CuKR (8.1125 keV), CuKβ (8.9682 keV), RbKR (13.4500 keV), RbKβ (15.0332), MoKR (17.5093), AgKR (22.1693), and BaKR (36.3988) lines. 2θ was calibrated using Si(1,1,1; 2,2,2; 2,2,0) and NaCl (1,1,1; 2,0,0; 2,2,0; 3,1,1) and refined to be 5.0402(5)°. X

Abstract published in AdVance ACS Abstracts, March 15, 1996.

0022-3654/96/20100-6412$12.00/0

The gas hydrates were prepared by introducing water from a reservoir into a high-pressure, low-temperature stirred reactor (see Figure 1). CO2 or C3H8 gas was then injected into the chamber from a high-pressure gas cylinder. The pressure and temperature of the water/gas mixture were then fixed to the appropriate hydrate formation conditions. Time-resolved energy dispersive spectra were recorded for both carbon dioxide and propane hydrates during growth over the temperature range 300-250 K and pressure range from atmospheric to 3.5 MPa. Spectra were recorded every 200 s throughout the experiment. Results and Discussion Our diffraction spectra for carbon dioxide hydrate prepared at 273.8 K show that six major reflections are present after crystal growth occurred, which we indexed in terms of a cubic lattice with unit cell parameter a ) 1.192(1) nm (Figure 2). These results were compared with reported values for known hydrate structures (see Table 1). We found that although the lattice type agrees with previous determinations of structure I gas hydrates, the a values obtained were significantly smaller than the value of 1.207 nm previously reported16 at 77 K. We made similar measurements for propane hydrate. Nine major reflections were observed after crystallization and indexed in terms of a diamond lattice with unit cell parameter a ) 1.7196(2) nm (Figure 3). This agreed with the lattice type for structure II gas hydrates; however, the unit cell parameter measured in solution was found to be slightly smaller than the previous value obtained at low temperature for propane of a ) 1.74 nm (ref 16). To our knowledge, there have been no previous in-situ X-ray measurements on gas hydrates at these pressures, and therefore comparisons with literature values are only to confirm the structure of the hydrates. Our carbon dioxide hydrate crystal growth data reveal a dynamic intermediate region (Figure 4), which occurs between gas dissolution and catastrophic growth of the hydrate crystal. On further uptake of gas by the system, the final stable hydrate crystal structure is obtained. Similar observations were made during the formation of propane hydrate. This behavior was observed during the formation of both structure I and structure © 1996 American Chemical Society

Crystallization of Natural Gas Hydrates

J. Phys. Chem., Vol. 100, No. 16, 1996 6413

Figure 3. In-situ energy dispersive X-ray diffraction pattern after crystallization of propane hydrate; T ) 273.8 K, P ) 0.42 MPa.

Figure 1. Schematic of the in-situ high-pressure, low-temperature reaction cell used to synthesize carbon dioxide and propane gas hydrates.

Figure 4. In-situ time-resolved energy dispersive X-ray diffraction patterns during growth of carbon dioxide hydrate.

Figure 2. In-situ energy dispersive X-ray diffraction pattern after crystallization of carbon dioxide hydrate; T ) 273.8 K, P ) 3.29 MPa.

TABLE 1: Crystal Hydrate Unit Cell Parameters at Different Temperatures and Pressures carbon dioxide hydrate

propane hydrate

T (K), P (MPa)

unit cell parameter, a (nm)

T (K), P (MPa)

unit cell parameter, a (nm)

273.8, 3.29 77, atm pressure

1.192 1.207 (ref 16)

273.8, 0.42 77, atm pressure

1.7196 1.74 (ref 16)

II hydrates. The presence of this dynamic region suggests that the mechanism of hydrate formation is not as simple a process as previous hypothetical models have suggested.4

We considered a number of hydrate inhibitor mechanisms: disruption to the prenucleation hydrogen bond environment, adsorption onto the crystal surface, and incorporation into the hydrate cage. NMR and FT-infrared spectroscopies were used to probe the effect of the inhibitor in the bulk liquid phase. Micro-FT-infrared was used to investigate functional group interactions at the vapor-liquid and vapor-solid interfaces. X-ray diffraction measurements were made to examine the effect of inhibitors on the hydrate crystal spectra. Our X-ray measurements for polyvinylpyrrolidone (PVP) and tyrosine were to determine their effect on the hydrate structure. Our data for PVP show the index reflections (222), (410), and (520) for carbon dioxide hydrate have been significantly suppressed in the presence of PVP (Figure 5a). One explanation may be the preferred orientation occurring in the presence of this polymer, although the combination of (222), (410), and (520) does not seem too likely; alternatively, there may be poor sampling of the “Debye-Scherrer powder diffraction cone” during synthesis. The explanation of this phenomenon, however, is not yet clear. Application of complementary techniques, such as micro-FT-infrared spectroscopy,17 strongly suggests that

6414 J. Phys. Chem., Vol. 100, No. 16, 1996

Koh et al. Conclusions Our in-situ time-resolved X-ray synchrotron data show there is a dependence between the crystal hydrate unit cell parameter, a, and the formation temperature and pressure conditions. Evidence of dynamic structural changes has been observed during the formation and decomposition of structure I and II gas hydrates. The presence of inhibitor molecules such as polyvinylpyrrolidone and tyrosine was found to alter the energy dispersive spectra of the crystal hydrate. Acknowledgment. We gratefully acknowledge the financial support of the Gas Research Institute Basic Research Program. Thanks are also due to R. E. Motie, R. A. Nooney, and A. Lewis for their help in performing the X-ray measurements, to A. Neild for the cell constructions, and to R. Westacott for the graphics input. Lastly, we thank S. C. Nyburg for fruitful discussions of these experiments.

Figure 5. In-situ energy dispersive X-ray diffraction patterns of carbon dioxide hydrate at T ) 273.8 K, P ) 3.29 MPa with (a) polyvinylpyrrolidone (the labeled peaks have been suppressed) and (b) tyrosine (peaks with asterisks have shifted with respect to carbon dioxide hydrate without tyrosine).

PVP adsorption can be explained in terms of a hydrogenbonding interaction between the carbonyl functional groups of PVP and an OHfree group on specific hydrate crystal planes. This is consistent with adsorption isotherm measurements made on PVP on an ice surface using second harmonic generation.17 Our X-ray work on the inhibition mechanism of tyrosine revealed that tyrosine induces a change in the energy dispersive spectrum (Figure 5b), which is different than our observations on PVP. Again, this may be attributed to the preferred orientation in the presence of tyrosine molecules during hydrate formation, or poor sampling of the “Debye-Scherrer powder diffraction cone” during synthesis. The action of tyrosine on the hydrate growth process was suggested from previous Raman measurements17 to involve the incorporation of a tyrosine molecule into the hydrogen-bonded water cavities.

References and Notes (1) Davy, H. Philos. Trans. R. Soc. London 1811, 101, 1. (2) Faraday, M. Quant. J. Sci. Lit. Arts 1823, 15, 71. (3) Lowig, C. Ann. Chem. Phys. Sci. 1829, 42, 113. (4) Ann. N.Y. Acad. Sci. 1994, 715, 561. (5) Bishnoi, P. R.; Dholabhai, P. D. Fluid Phase Equilibria 1993, 83, 455. (6) Van der Waals, J. H.; Platteeuw, J. C. AdV. Chem. Phys. 1959, 2, 1. (7) Tanaka, H.; Kiyohara, K. J. Chem. Phys. 1993, 98, 4098. (8) Rodger, P. M. Mol. Simulation 1990, 5, 315. (9) Davidson, D. W. In Water: A ComprehensiVe Treatise; Franks, F., Ed.; Plenum: New York, 1972; Vol. 2, p 115. (10) Jeffrey, G. A.; McMullan, R. K. Prog. Inorg. Chem. 1967, 8, 43. (11) van Stackelberg, M.; Muller, H. R. J. Chem. Phys. 1951, 19, 1319. (12) Koh, C. A.; Muller, E.; Zollweg, J. A.; Gubbins, K. E.; Savidge, J. L. Ann. N.Y. Acad. Sci. 1994, 715, 561. (13) GRI Final Report, 1993, GRI-94/0272. (14) Baez, L.; Clancy, P. Ann. N.Y. Acad. Sci. 1994, 715, 177. (15) Benson, S. W.; Siebert, E. D. J. Am. Chem. Soc. 1992, 114, 4269. (16) Von Stackelburg, M.; Jahns, W. Elektrochem. 1954, 58, 162. (17) Koh, C. A.; Klenerman, D.; Motie, R. E.; Tanaka, H.; Savidge, J. L. Proc. Annu. Meet. Gas. Hydrates 1994, (Ottawa).

JP960094S