You are on page 1of 7

Ind. Eng. Chem. Res.

2005, 44, 7373-7379


NMR and Viscosity Investigation of Clathrate Hydrate Formation and Dissociation
Shuqiang Gao,† Walter G. Chapman,*,† and Waylon House*,‡
Chemical Engineering Department, Rice University, Houston, Texas 77251, and Petroleum Engineering Department, Texas Tech University, Lubbock, Texas 79406

To better understand clathrate hydrate mechanisms, nuclear magnetic resonance (NMR) and viscosity measurements were employed to investigate tetrahydrofuran (THF) hydrate formation and dissociation processes. In NMR experiments, the proton spin lattice relaxation time (T1) of THF in deuterium oxide (D2O) was measured as the sample was cooled from room temperature down to the hydrate formation region. The D2O structural change around THF during this process was examined by monitoring the rotational activation energy of THF, which can be obtained from the slope of ln(1/T1) vs 1/T. No evidence of hydrate precursor formation in the hydrate region was found. T1 measurements of THF under constant subcooling temperature indicate that THF hydration shells do not undergo much structural rearrangement during induction. The T1 of THF was also measured as the sample was warmed back to room temperature after hydrate dissociation. T1 values of THF after hydrate dissociation were consistently smaller than those before hydrate formation and never returned to original values. It was proposed that this difference in T1 after hydrate dissociation indicates that the THF-D2O solution is more microscopically homogeneous than before hydrate formation. In viscosity experiments, a Champion Technologies hydrate rocking cell (CTHRC) was used to probe the residual viscosity phenomenon after Green Canyon (GC) gas hydrate as well as THF hydrate dissociation. The residual viscosity reported in the literature was observed after GC hydrate dissociation but not after THF hydrate dissociation. Because GC hydrate behavior involves significant amounts of gas mass transfer while THF hydrate does not, one might conclude that the residual viscosity observed after GC hydrate dissociation was likely caused by the supersaturated gas concentration and its general effect on solvent viscosity, not necessarily by a clathrate water structure lingering from the solid.
1. Introduction Gas hydrates, also known as clathrate hydrates, are nonstoichiometric crystalline compounds composed of cooperative hydrogen-bonded water molecules forming nanoscale cage-like structures, which accommodate smaller guest molecules in nearly spherical cavities. Depending on the size and the composition of the guest molecules, gas hydrates usually form in at least three types of crystal structures, I, II, and H.1 It has been widely recognized that gas hydrates are significant factors in flow assurance,2 global warming,3 and seafloor stability4 and are a potential energy supply.5 Gas hydrates have also been proposed for use in gas separation,6 transportation, and storage.7 Since the first discovery that water forms clathrate hydrate with chloride in 1810 by Sir Humphrey Davy, about 200 years of research has been devoted to understanding hydrate phenomenon. This continuous effort has well-established the thermodynamic, physical, and structural properties of gas hydrates and discovered a rich collection of hydrate formers. However, clathrate hydrate formation and decomposition mechanisms and kinetics are still far from clear due to a lack of appropri* To whom correspondence should be addressed. W.G.C: phone, (713) 348-4900; fax, (713) 348-5478; e-mail, wgchap@ W.H.: phone, (806) 742-3573; fax, (806) 742-3502; e-mail, † Rice University. ‡ Texas Tech University.

ate experimental techniques capable of probing dynamic structural information on the molecular level. The actual formation and dissociation mechanisms have important impacts on all hydrate applications. Clarification of the formation mechanism is especially important for designing kinetic inhibitors or antiagglomerant hydrate inhibitors that are expected to intervene with the hydrate formation process on the molecular scale. Current hypotheses1 about hydrate formation and dissociation involve hydrate precursor and postcursor structures. The hypothesis states that, prior to hydrate formation, water molecules form individual clathrate cages (hydrate precursors) around the dissolved guest molecules with one guest molecule inside each cage. Under favorable conditions, these cages will agglomerate and form solid clathrate hydrate. The hypothesis also suggests that, upon melting, the solid hydrate phase dissolves into hydrate cage clusters. Those clusters, also known as residual structures or postcursors, persist in the liquid phase over a long period of time provided the temperature stays within certain limits.8 To understand hydrate formation and dissociation mechanisms, many researchers have studied hydrates with a particular emphasis on the precursor and postcursor hypotheses. Molecular simulations generated voluminous numerical data on the elemental molecular detail of hydrophobic hydration processes, which is important for evaluating the precursor hypothesis. A vigorous debate continues on whether hydrophobic solutes structure the water molecules around them. One

10.1021/ie050464b CCC: $30.25 © 2005 American Chemical Society Published on Web 08/09/2005


Ind. Eng. Chem. Res., Vol. 44, No. 19, 2005

view argues that nonpolar solutes enhance the order of the hydrogen-bonding network and create structured hydration shells around them.9-20 On the other hand, some simulation results showed no indication of enhanced structure around hydrophobic solutes.21,22 Even though computer simulation studies have reached conflicting conclusions on whether nonpolar solutes structure water network around them, they all seem to agree that long-lived clathrate cages in solution do not exist. The orientation of hydrogen bonds in the hydration shell is qualitatively the same as in hydrate clathrate cages, but the organization of the hydrogen bonds is less ordered and the number of water molecules involved in each hydration shell is also different from that in a clathrate hydrate cage. Experimental work has also been carried out to probe the water structures around dissolved nonpolar solutes. Data seem to agree with those from computer simulation; that is, only loosely organized hydration cages form around hydrophobic solutes, with no evidence of regular clathrate cages in solution.23-27 The fact that no regular clathrate cages have been experimentally confirmed in the water phase before hydrate formation indicates that the actual hydrate formation process may be much more complicated than the hypothesis suggests. Studies of molecular hydrate formation mechanisms present great experimental challenges due to the low equilibrium concentrations of gases in water and the high pressures required. Experiments usually require specially designed sample cells and high sensitivity instruments capable of providing structural information on a molecular scale. NMR T1 measurement has been shown to be a powerful technique in studying microdynamic behavior of liquids28 and providing local molecular-level structural information surrounding the NMR responding guest molecules. THF and water form structure II hydrate at ∼4.5 °C under ambient pressure. It is a good model system to study clathrate hydrate mechanisms because its formation conditions are mild and THF is miscible with water at the conditions of interest. In this work, proton T1 of THF in D2O was measured as the sample was cooled from room temperature to hydrate formation conditions. This allowed us to monitor the structural change of the hydration shells around THF molecules as a function of temperature and provided an opportunity to examine whether hydrate precursors form under hydrate conditions. The residual structure dissociation hypothesis is not as widely studied as the hydrate precursor hypothesis. It has been observed that re-formation of gas hydrates from this “hydrate melt” requires less subcooling and less induction time. This phenomenon is called the memory effect. Vysniauskas and Bishnoi29 proposed that it is the existence of residual clathrate structure in water after hydrate dissociation that is responsible for the memory effect. More recently, Sloan et al.30 discovered that the viscosity of water after methane hydrate dissociation is initially higher than that before hydrate formation under the same pressure and temperature. The viscosity returns to normal only after a long period of time. This macroscopic residual viscosity phenomenon is argued as evidence of residual structure, i.e., clathrate aggregates remaining in water after gas molecules are released.31 However, it is well-known that the diffusion rate of methane in water is very small. After hydrate

Figure 1. Schematic of the Champion Technologies hydrate rocking cell apparatus.

dissociation, it is very likely there is still a large amount of methane dissolved in water solution due to mass transfer rate limitations. Since dissolved methane structures the hydrogen-bonding network in water,32 water supersaturated with methane will have a higher viscosity than water simply saturated with methane under the same pressure and temperature. Therefore, it is not clear whether the observed residual viscosity is caused by remnant clathrate structures or an excess amount of dissolved methane. To eliminate the complication of mass transfer phenomenon on the experimental results mentioned above and obtain a better understanding of the residual hydrate structure hypothesis, in this work we chose to study the possibility of residual viscosity after THF hydrate dissociation. THF is miscible with water at the concentration of solid THF hydrate;33,34 thus, experiments can be performed at constant concentration before, during, and after hydrate formation. Stangret and Gampe35 recently demonstrated that the hydrophobic hydration of THF definitely dominates the hydrophilic one at low THF concentrations. This is the case at hydrate composition THF:H2O ) 1:17 (molar), so water-THF hydrogen bonds are not very important for the fluid structure of dilute solutions. Therefore, in terms of water structure change, THF is not far from other nonpolar solutes. Viscosity of the sample was monitored before hydrate formation and after dissociation using a Champion Technologies hydrate rocking cell (CTHRC) (Figure 1) to detect the reported residual viscosity phenomenon.30 Since THF concentration in water is constant, if the residual viscosity phenomenon is observed, it would be a strong demonstration of the existence of residual clathrate hydrate structures, i.e., postcursors. Fleyfel et al.36 combined macroscopic hydrate experiments (visual rocking cell) with microscopic hydrate experiments (NMR) to investigate clathrate residual structure at the point where hydrate particles become invisible. They found that some clathrate cages might still exist in solution after hydrates visually disappear. In this work, to investigate the molecular evidence of residual structure for THF hydrate, we measured T1 values of THF as a function of temperature after THF hydrate dissociation in the NMR experiment mentioned earlier. Results were compared with that before hydrate

Ind. Eng. Chem. Res., Vol. 44, No. 19, 2005 7375

Figure 2. Schematic of the NMR experimental setup.

formation to detect any difference caused by the history of hydrate formation. 2. Experimental Section 2.1. NMR Microscopic Measurements. 2.1.1. Theoretical Background. Spin lattice relaxation mechanisms for THF are composed of two parts, intermolecular dipole-dipole and intramolecular dipole-dipole interactions. When the THF to D2O molar ratio is 1:17 (the same composition as in THF hydrate), most of the THF molecules are completely surrounded by D2O hydration shells, and intra-dipole-dipole is the main spin relaxation mechanism. Orientational structures of D2Os in hydration shells have direct impact on the rotational motion of enclosed THF molecules. On the basis of NMR theory,37 T1 is inversely proportional to the rotational correlation time τc. τc is usually believed to follow the Arrhenius behavior

1 ∝ τc T1 τc ) τoeEa/RT

(1) (2)

where τo is a constant preexponential factor; Ea is the rotational activation energy, a manifestation of hydration cage structure around THF and not sensitive to temperature; T is the temperature in kelvin; and R is the molar gas constant. Combining eqs 1 and 2 yields eq 3.

d ln(1/T1) d(1/T)


Ea R


The slope of ln(1/T1) versus (1/T) gives the rotational activation energy Ea. Therefore, by tracking the changes in the slope of ln(1/T1) versus (1/T), we can measure Ea and evaluate the changes in hydration shell structure around dissolved THF molecules as a function of temperature. 2.1.2. Experimental Details. T1 measurements of THF (Aldrich, 99+%) in D2O (Cambridge Isotope Laboratories, D 99.9%) at various temperatures were performed on an 85-MHz Oxford horizontal 31 cm wide bore NMR with imaging capability, using a LITZ RF Volume Coil (with 14 cm internal diameter) from Doty Scientific, Inc. (Figure 2). Data were acquired and processed using Varian VNMR software and INOVA hardware systems. An Air-Jet temperature controller blew dry and cold air to control the sample temperature. It is capable of controlling temperature from -40 to 100 °C with (0.1 °C stability. A glass bottle with a cap that has a Teflon

liner was used to contain the deoxygenated THF:D2O ) 1:17 (molar ratio) mixture. A LUXTRON fluoroptic thermometer was mounted into the glass container through the cap to monitor system temperature. Its output reading resolution is 0.1 °C. Since trace amounts of oxygen may alter the T1 of THF significantly, we deoxygenated pure D2O and THF liquids separately in Teflon containers by periodically flushing the gas above the liquid phase with pure nitrogen gas while periodically shaking the sample containers to accelerate the diffusion of oxygen from the liquid samples. After flushing six or seven times over about 12 h, THF and D2O were mixed on the molar basis of 1:17. All these processes were operated within a closed glovebox with nitrogen environment. After allowing the sample to stabilize for half a day, we put the sample into the NMR probe for T1 measurements. (The success of deoxygenation was demonstrated by that fact that, after deoxygenation, T1 increased about 2 s compared to the sample without deoxygenation). T1 values were measured using the inversion recovery technique. All samples’ 90° and 180° pulses were calibrated before every measurement. VNMR software, given inputs of possible minimum and maximum T1 values, automatically generates standard 180°-τ-90° pulse sequences with various values of τ. It took 4-6 min to take a T1 data point. The sample was cooled from room temperature (∼25 °C) to subzero temperatures with an average cooling rate of ∼0.5 C°/h until hydrate formation occurred. Hydrate formation was indicated by a jump in the sample temperature. Oyama et al.38 proposed that, during the induction period, water around dissolved CO2 takes time to rearrange into individual clathrate cages before hydrate formation. To test this hypothesis, we measured T1 as a function of time while the temperature was kept constant in subcooling state. If the hydration shell structures around THF experience rearrangements, the T1 of THF would vary. After hydrate formation, temperature was raised to completely dissociate the hydrate. T1 was also measured as the sample was heated back to about 25 °C. 2.2. Viscosity Macroscopic Measurements. The viscosity experiments for the THF (99+%, Aldrich)deionized (DI) water system were performed on a stateof-the-art CTHRC, which is an apparatus with highpressure sapphire tubes placed on a rack. It is capable of handling pressures up to 5000 psi. The rack is immersed in a temperature-controlled thermal bath and it rocks (45° from horizontal position through a computer-controlled step motor. As the rack rocks, a ball inside the tube rolls from one end to the other, and each tube functions such as a rolling ball viscometer. The motion of the ball is detected through two sensors mounted on both ends of each tube. The time that the ball takes to travel from one sensor to the other is recorded as an indication of the fluid viscosity inside the tube. Changes of the fluid viscosity in the tubes can be sensitively detected by measuring the ball travel speeds. Hydrate formation is indicated by an increase in fluid apparent viscosity while system temperature and pressure were kept constant. The ball was eventually immobilized by the hydrate blockage. Hydrate onset can also be detected by visual observation of hydrate crystal deposit on the tube’s inside wall. Ball travel time, temperature, and pressure data are collected every minute into a terminal computer using LabView.


Ind. Eng. Chem. Res., Vol. 44, No. 19, 2005

Figure 4. T1 of THF in D2O during induction time at -1.7 °C. Figure 3. T1 of THF in D2O before hydrate formation and after hydrate dissociation, plotted as ln(1/T1/s) vs 1/T/K. The activation energies (Ea) were calculated from the slopes.

Green Canyon (GC) gas hydrate experiments were first performed on CTHRC to duplicate the residual viscosity phenomenon reported in the literature30 in order to ensure that the rocking cell we use is sensitive enough. GC gas is composed of 87.2% (mol) CH4, 7.6% C2H6, 3.1% C3H8, 0.8% n-C4H10, 0.5% i-C4H10, 0.2% n-C5H12, 0.2% i-C5H12, and 0.4% N2. Every tube was first charged with 9 mL of DI water. After the system was purged three times with 50 psi of GC gas, 1000 psi of GC gas was introduced into the tubes. The pressure was kept constant at 1000 psi during the experiment. Once charged, the cell temperature was initially raised to 30 °C to eliminate any possible preexisting extra structure in water. The system was then lowered to 20 °C and equilibrated for ∼2 h to establish a viscosity baseline at this temperature. The temperature was then further cooled to 15 °C to initiate hydrate formation. The predicted hydrate equilibrium temperature by CSMHYD1 is 17.8 °C for this pressure and gas composition. To ensure complete hydrate formation, the temperature was lowered to 5 °C before it was raised to dissociate the hydrate. Then the system was slowly warmed back to 20 °C until complete hydrate dissociation occurred. The system was equilibrated to establish the viscosity baseline at this temperature again. Viscosities of the THF-water system were measured in a similar fashion. DI water and THF were added in a H2O:THF ) 17:1 molar ratio, the composition in THF hydrate. The tubes were then sealed and heated to 30 °C to eliminate any preexisting extra structure in the water. The system temperature was then lowered to hydrate formation conditions. During this cooling process, the system was allowed to equilibrate at 10 and 6 °C for about 300 min in order to establish the viscosity baselines at these temperatures. After complete hydrate formation, the system was warmed to dissociate the hydrate. The system was equilibrated again at 10 and 6 °C to establish the viscosity lines. 3. Results and Discussion 3.1. NMR Results. As shown in Figure 3, the ln(1/ T1) vs 1/T slope does not change dramatically as the sample is cooled from room temperature into the hydrate region. This indicates that the hydration cage structure around THF in the hydrate region before the phase transition is not very different from that at room temperature. The slope gives the average rotational activation energy of 21.18 kJ/mol for THF in water solution, which is very different from that in the solid hydrate phase, 4.14 kJ/mol.39 Since the rotational

activation energy reflects local water structure around THF, the result indicates that THF hydration shells in the aqueous phase are dissimilar in nature from the regular clathrate cages in the hydrate phase, even under hydrate formation conditions. The suggested hydrate precursors were not found and the proposed hydrate formation hypothesis1 is not supported. To compare with the time-dependent results reported by Oyama et al.,38 we measured T1 as a function of time while the temperature was kept constant under hydrate conditions. As shown in Figure 4, T1 does not change much with respect to time during the induction period. This implies that the hydration shells around THF have no dramatic structural change and they do not rearrange themselves into regular clathrate cages during induction time. One interesting feature is that even though the slope of ln(1/T1) vs 1/T does not change dramatically as temperature goes from ∼25 °C to hydrate formation regions, it does show a slight increase at around 8-9 °C. In this experiment, the activation energy changes from 19.97 to 23.83 kJ/mol at ∼8.5 °C. This transition in the slope is reproducible and it is under further investigation. After hydrate dissociation, even though the slope is similar to that before hydrate formation, there is a consistent shift of T1, which never returned to the original value. This variation of T1 implies a structural change of water molecules around THF due to hydrate formation and dissociation. However, this structural change is not caused by residual clathrate structure, because the effect still exists above 25 °C. There are possible microscopic heterogeneities40 in fresh THFD2O solution, even though THF and D2O are miscible at all concentrations under conditions of interest. Since THF becomes perfectly distributed as a result of hydrate formation, subsequent dissociation would result in a more microscopically homogeneous THF-D2O solution. Therefore, the T1 shift is probably caused by a more uniform distribution of THF in D2O after THF hydrate dissociation than before hydrate formation. It was also observed that before hydrate formation the THF-D2O solution had a slight milky appearance and after hydrate dissociation the solution became crystal clear. Without external disturbance, the milky appearance can last over 2 weeks at room temperature. 3.2. Viscosity Results. In the GC gas hydrate viscosity experiment, shown in Figure 5, the viscosity baseline after hydrate dissociation is obviously higher than that before hydrate formation at temperature 20 °C and pressure 1000 psi. The residual viscosity phenomenon reported in the literature30 was successfully reproduced.

Ind. Eng. Chem. Res., Vol. 44, No. 19, 2005 7377

Figure 5. Residual viscosity phenomenon observed at 20 °C after Green Canyon gas hydrate dissociation. The pressure was kept constant at 1000 psi.

Figure 6. No residual viscosity is observed at 6 and 10 °C after THF hydrate dissociation.

However, results from THF-water hydrate experiments (Figure 6) do not demonstrate this residual viscosity effect. Viscosities baselines established at 6 and 10 °C before hydrate formation are the same as those after hydrate dissociation. Macroscopic evidence of the residual structure is absent in this experiment. The different results from THF and GC experiments may be explained as follows. In GC hydrate experiment, after hydrate dissociation, a large amount of hydrocarbons have to get out of the liquid phase because the concentrations of hydrocarbon gases in the hydrate phase are much higher than their equilibrium concentrations in liquid water under the same condition. The hydrocarbon gases resulting from hydrate dissociation can come out of liquid phase by diffusion or by forming bubbles. Since the diffusion coefficients of hydrocarbons in water are very small, it is much more efficient for gases to escape as bubbles. However, from visual observation, few bubbles formed during GC hydrate dissociation under constant pressure. Therefore, gases come out of water mainly by diffusion, which makes this process very time consuming. During the time frame of viscosity measurements, liquid water could be still highly supersaturated with hydrocarbon gases. This would cause higher than normal apparent water viscosity.41 THF does not have the above problem because its concentration in the liquid phase is the same as that in the hydrate phase

and it stays constant under all conditions. Since residual viscosity was absent after THF hydrate dissociation, the observed residual viscosity phenomenon in GC hydrate experiment was probably caused by supersaturated hydrocarbons in water after hydrate dissociation, not necessarily by remaining clathrate structure in the aqueous phase. The fact that residual viscosity was not observed after THF dissociation but an overall structural change around THF was indicated by NMR results may be caused by two reasons. First, strong mixing by the steel ball inside the sample tube may have already made THF-D2O microscopically homogeneous even before hydrate formation and destroyed any possible residual structure as soon as the hydrate melted. Second, even though CTHRC can reproduce the residual viscosity phenomenon for the GC hydrate experiment, it may not be sensitive enough to detect the structural change after THF hydrate dissociation. In any case, rocking-cell types of apparatus30 may not be appropriate tools to measure residual structure. 4. Conclusions NMR and CTHRC techniques were used to investigate the formation and dissociation mechanisms of THF hydrate. In NMR experiments, T1 of THF in D2O was measured as the sample was cooled from room temper-


Ind. Eng. Chem. Res., Vol. 44, No. 19, 2005
(11) Geiger, A.; Rahman, A.; Stillinger, F. H. Molecular dynamics study of the hydration of Lennard-Jones solutes. J. Chem. Phys. 1979, 70, 263. (12) Guillot, B.; Guissani, Y. A computer simulation study of the temperature dependence of the hydrophobic hydration. J. Chem. Phys. 1993, 99, 8075. (13) Chau, P. L.; Mancera, R. L. Computer simulation of structural effect of pressure on the hydrophobic hydration of methane. Mol. Phys. 1999, 96, 109. (14) Long, J.; Sloan, E. D. Quantized water cluster around apolar molecules. Mol. Simul. 1993, 11, 145. (15) Skipper, N. T. Computer simulation of methane-water solutions. Evidence for a temperature-dependent hydrophobic attraction. Chem. Phys. Lett. 1993, 207, 424. (16) Mancera, R. L.; Buckingham, A. D.; Skipper, N. T. The aggregation of methane in aqueous solution. J. Chem. Soc., Faraday. Trans. 1997, 93, 2263. (17) Silverstein, K. A. T.; Haymet, A. D. J.; Dill, K. A. Molecular model of hydrophobic solvation. J. Chem. Phys. 1999, 111, 8000. (18) Silverstein, K. A. T.; Haymet, A. D. J.; Dill, K. A. A simple model of water and the hydrophobic effect. J. Am. Chem. Soc. 1998, 120, 3166. (19) Mancera, R. L. Computer simulation of effect of salt on the hydrophobic effect. J. Chem. Soc., Faraday Trans. 1998, 94, 3549. (20) Lazaridis, T. Solvent reorganization energy and entropy in hydrophobic hydration. J. Phys. Chem. B. 2000, 104, 4964. (21) Meng, E. C.; Kollman, P. A. Molecular dynamics studies of the properties of water around simple organic solutes. J. Phys. Chem. 1996, 100, 11460. (22) Hernandez-Cobos, J.; Mackie, A. D.; Vega, L. F. The hydrophobic hydration of methane as a function of temperature from histogram reweighting Monte Carlo simulations. J. Chem. Phys. 2001, 114, 7527. (23) Broadbent, R. D.; Neilson, G. W. The interatomic structure of argon in water. J. Chem. Phys. 1994, 100, 7543. (24) De Jong, P. H. K.; Wilson, J. E.; Neilson, G. W.; Buckingham, A. D. Hydrophobic hydration of methane. Mol. Phys. 1997, 91, 99. (25) Koh, C. A.; Wisbey, R. P.; Wu, X.; Westacott, R. E.; Scoper, A. K. Water ordering around methane during hydrate formation. J. Chem. Phys. 2000, 113, 6390. (26) Filipponi, A.; Bowron, D. T.; Lobban, C.; Finney, J. L. Structural determination of hydrophobic hydration shell of Kr. Phys. Rev. Lett. 1997, 79, 1293. (27) Bowron, D. T.; Filipponi, A.; Roberts, M. A.; Finney, J. L. Hydrophobic hydration and the formation of a clathrate hydrate. Phys. Rev. Lett. 1998, 81, 4164. (28) Hertz, H. G. Microdynamic behavior of liquids as studied by NMR relaxation times. Prog. Nucl. Magn. Reson. Spectrosc. 1967, 3, 159. (29) Vysniauskas, A.; Bishnoi, P. R. A kinetic study of methane hydrate formation. Chem. Eng. Sci. 1983, 38, 1061. (30) Sloan, E. D.; Subramanian, S.; Matthews, P. N.; Lederhos, J. P.; Khokhar, A. A. Quantifying Hydrate Formation and Kinetic Inhibition. Ind. Eng. Chem. Res. 1998, 37, 3124. (31) Makogon, Y. F. Hydrates of Natural Gas; PennWell Books: Tulsa, OK, 1981 (32) Frank, H. S.; Wen, W. Y. Structural aspects of ion-solvent interaction in aqueous solutions: A suggested picture of water structure. Discuss. Faraday. Soc. 1957, 24, 133. (33) Matous, J.; Hrncirik, J.; Novak, J. P.; Sobr, J. Liquidliquid equilibrium in the system water-tetrahydrofuran. Collect. Czech. Chem. Commun. 1970, 35, 1904. (34) Lejcek, P.; Matous, J.; Novak, P.; Pick, J. Phase equilibria and excess molar volumes of tetrahydrofuran (1) + deuterium (2). J. Chem. Thermodyn. 1975, 7, 927. (35) Stangret, J.; Gampe, T. Hydration of tetrahydrofuran derived from FTIR spectroscopy. J. Mol. Struct. 2005, 734, 183 (36) Fleyfel, F.; Song, K. Y.; Kook, A.; Martin, R.; Kobayashi, R. Interpretation of C-13 NMR of methane propane hydrates in the metastable nonequilibrium region. J. Phys. Chem. 1993, 97, 6722. (37) Fukushima, E.; Roeder, S. B. W. Experimental Pulse NMRsA Nuts and Bolts Approach; Westview Press: Boulder, CO, 1981.

ature to hydrate formation conditions. By calculating the rotational activation energy of THF in D2O, it was shown that the THF hydration shell structure in solution changes little as temperature goes from room temperature to subzero. No indication of hydrate precursor formation was observed. During the induction period, T1 stayed about the same as a function of time at constant temperature, which implied that water molecules in THF hydration shells do not undergo much rearrangement during the induction period. These results do not support the hydrate precursor hypothesis in THF. After hydrate dissociation, T1s were consistently smaller than that before hydrate formation, which suggests a structural change in the solvent after hydrate dissociation. The failure of this change to disappear even at higher temperatures suggests it is probably caused by a more uniform distribution of THF in D2O after hydrate dissociation. Using the CTHRC apparatus, a residual viscosity effect reported in the literature was reproduced in a GC gas hydrate experiment but not in a THF hydrate system. After GC gas hydrate dissociation under constant pressure, it would take a long time for hydrocarbon gases to reach their equilibrium concentrations because of their small diffusion constants in water. Combined with results from the THF viscosity experiment, it is concluded that the residual viscosity phenomenon after GC gas hydrate dissociation is probably due more to gas supersaturation than residual clathrate structure. Acknowledgment We gratefully acknowledge the financial support of the Robert A. Welch Foundation. S.G. expresses his gratitude to Champion Technologies Inc for giving him access to their state-of-the-art Champion Technologies hydrate rocking cell. The Livermore Chair and MRIPAC at TTU provided support for the NMR experiments. Literature Cited
(1) Sloan, E. D., Jr. Clathrate Hydrates of Natural Gases; Marcel Dekker: New York, 1998. (2) Sloan, E. D., Jr. Hydrate Engineering; Society of Petroleum Engineers: Richardson, TX, 2000. (3) Dickens, G. R.; O’Neil, J. R.; Rea, D. K.; Owen, R. M. Dissociation of oceanic methane hydrate as a cause of carbon isotope excursion at the end of the Paleocene. Paleoceanograhy 1995, 10, 965. (4) Kvenvolden, K. A. Gas hydratessGeological perspective and global change. Rev. Geophys. 1993, 31, 173. (5) Kerr, R. A. EnergysGas hydrate resource: Smaller but sooner. Science 2004, 303, 946. (6) Kang, S. P.; Lee, H. Recovery of CO2 from flue gas using gas hydrate: Thermodynamic verification through phase equilibrium measurements. Environ. Sci. Technol. 2000, 34, 4397. (7) Yevi, G. Y.; Rogers, R. E. Storage of fuel I hydrates for natural gas vehicles (NGVs). J. Energy Resour. Technol.sASME. 1996, 118, 209. (8) Lederhos, J. P.; Long, J. P.; Sum, A.; Christiansen, R. L.; Sloan, E. D. Effective kinetic inhibitors for natural gas hydrates. Chem. Eng. Sci. 1996, 51, 1221 (9) Owicki, J. C.; Scheraga, H. A. Monte Carlo calculation in isothermal-isobaric ensemble. 2. Dilute aqueous solution of methane. J. Am. Chem. Soc. 1977, 99, 7413. (10) Swaminathan, S.; Harrison, S. W.; Beveridge, D. L. Monte Carlo studies on the structure of dilute aqueous solution of methane. J. Am. Chem. Soc. 1978, 100, 5705.

Ind. Eng. Chem. Res., Vol. 44, No. 19, 2005 7379
(38) Oyama, H.; Ebinuma, T.; Shimada, W.; Takeya, S.; Nagao, J.; Uchida, T.; Narita, H. Viscosity increase before nucleation of CO2 gas hydrate. Ann. N. Y. Acad. Sci. 2000, 912, 561. (39) Albayrak, C.; Zeidler, M. D. NMR relaxation of tetrahydrofuran in clathrate hydrate. Ber. Bunsen-Ges. Phys. Chem. 1989, 93, 1119. (40) Katz, E. D.; Ogan, K.; Scott, R. P. W. Distribution of a solute between 2 phasessThe basic theory and its application to the prediction of chromatographic retention. J. Chromatogr. 1986, 352, 67. (41) Uchida, T.; Ohmura, R.; Nagao, J.; Takeya, S.; Ebinuma, T.; Narita, H. Viscosity of aqueous CO2 solutions measured by dynamic light scattering. J. Chem. Eng. Data. 2003, 48, 1225.

Received for review April 18, 2005 Revised manuscript received July 7, 2005 Accepted July 12, 2005 IE050464B