J. Phys. Chem. B 2007, 111, 11209-11213
11209
Hydration of NaCl on Glassy, Supercooled-Liquid, and Crystalline Water Ryutaro Souda* Nanoscale Materials Center, National Institute for Materials Science, 1-1 Namiki, Tsukuba, Ibaraki 305-0044, Japan ReceiVed: April 2, 2007; In Final Form: June 29, 2007
Interactions of sodium chloride with amorphous and crystalline water films, leading to the possible formation of a dilute NaCl solution, were investigated using time-of-flight secondary ion mass spectrometry as a function of temperature. A monolayer of NaCl tends to remain on the surface or in subsurface sites of thick amorphous solid water films (200 monolayers); the Na+ ion is hydrated preferentially, whereas the Cl- ion is segregated at the surface. The hydration structure of NaCl is fundamentally unchanged for viscous liquid water that appears at temperatures higher than 136 K. The solubility of NaCl increases abruptly at 160 K because of the evolution of supercooled liquid water, which can hydrate the Cl- ion efficiently. However, the diffusion of the ions toward the bulk of supercooled liquid water is interrupted by crystallization; therefore, the dilute NaCl solution that is characterized by completely separated Na+-Cl- pairs may not be formed. When NaCl is deposited on the crystalline ice film, hydration of NaCl is enhanced above 160 K as well, indicating that a liquidlike phase coexists with crystals.
1. Introduction The surface composition and structure of aqueous sodium chloride solutions have attracted considerable attention in terms of the heterogeneous chemistry of sea salt aerosols.1,2 It is believed that no atomic ions exist at the air/solution interface because electrostatic forces repel ions from the surface into the aqueous bulk.3 This traditional view has been challenged recently by molecular dynamics (MD) simulations,2,4-6 which predict that heavier, polarizable halide ions such as iodide, bromide, and chloride tend to be segregated at the surface because of their imperfect hydration relative to the unpolarizable cations. To date, several groups have used surface-sensitive experimental techniques to study surface ion concentrations in salt solutions. Nonlinear optical experiments have revealed that the hydrogen bond strength of the water molecules in the NaBr and NaI solutions is weakened, which has been ascribed to higher concentrations of the Br- and I- ions in the surface region.7,8 On the other hand, conflicting results have been obtained from measurements of high-pressure photoelectron spectroscopy (PES); the valence-band photoelectron spectra from a liquid jet of alkali iodide solutions suggest that ions are depleted at the surface,9 whereas a considerable enhancement of the halide ion concentration in the surface region was inferred for KBr and KI single-crystal samples in contact with water vapor (several Torr).10 The inconsistency of these experiments might arise from the inherent difficulty in obtaining detailed molecular information for liquid surfaces that are equilibrated with the vapor phase. Recently, the surface composition of alkali halide adsorbed on thin films of amorphous solid water (ASW) has been investigated under ultrahigh vacuum (UHV) conditions. Using a low-energy sputtering technique, Kim et al.11 revealed that Na+ ion migrates from the surface into the interior of the ASW films at temperatures higher than 120 K, in contrast to the * To whom correspondence should be addressed. E-mail: SOUDA.
[email protected].
surface residence of the Cl- ion. Hofft et al.12 found that surface segregation of iodine occurs above 130-135 K for the CsI/ ASW system. The results of these UHV experiments, which are more sensitive to the outermost surface layer than photoelectron spectroscopy, show agreement with the MD simulations for aqueous alkali halide solutions. However, the assumption that ASW is a thermodynamic extension of liquid water11,12 is questionable because the local structure of glassy water resembles that of crystalline ice rather than liquid water.13 A liquid phase might occur by annealing ASW above the glass-transition temperature, which has been assigned to Tg ) 136 K,14 but supercooled water might not be accessible because crystallization occurs at 150 K.15 For this study, we investigate the interaction of NaCl with the ASW film on the basis of time-of-flight secondary ion mass spectrometry (TOF-SIMS). In precedent studies, TOF-SIMS has been used for analyses of the glass-liquid transition of water at the molecular level;16,17 those studies demonstrated that the self-diffusion of water molecules commences at 136 K, and then the fluidity evolves at 160-165 K, leading to film dewetting. Furthermore, a concentrated LiCl solution is formed at around 160-165 K when the ASW film is deposited on a polycrystalline LiCl film.18,19 The present study examines hydration of one monolayer (1 ML) of NaCl deposited on a thick ASW film (200 ML) and discusses the possible formation of a dilute NaCl solution in the deeply supercooled region. Distribution of the hydrated NaCl species in the water film is analyzed using TOFSIMS as a function of temperature, and the hydration structures of NaCl are discussed from comparison with the results of the MD simulations. 2. Experimental The experiment was performed in a UHV chamber with base pressure of less than 1 × 10-8 Pa. TOF-SIMS measurements were made by incidence of a pulsed He+ beam (2.0 keV) extracted from an electron-impact-type ion source onto a sample surface that was floated with a bias voltage of (500 V. A
10.1021/jp0725580 CCC: $37.00 © 2007 American Chemical Society Published on Web 08/31/2007
11210 J. Phys. Chem. B, Vol. 111, No. 38, 2007 grounded stainless steel mesh was placed immediately in front of the sample surface (ca. 4 mm) to avoid leakage of the electric field. The positive and negative ions were subsequently extracted perpendicularly from the surface through the mesh, detected by a channel electron multiplier after traveling through a fieldfree linear TOF tube (ca. 65 cm), and then were pulse-counted using a multichannel scaler (LN-6500R; Laboratory Equipment Inc.). For measurements of negative ions, small ferrite magnets were placed at the entrance of the TOF tube to reduce the yield of secondary electrons emitted from the surface. A Ni(111) surface that was spot-welded to a Ta holder using Ni strips was used as a substrate. The sample was inserted in the UHV chamber via a load-lock system. Cooling of the sample was achieved by thermal contact between mirror-finished faces of the Ta holder and a sapphire plate that was attached to a copper rod extended from a closed-cycle helium refrigerator. Heating was achieved by electron bombardment from behind through a square hole opened in the sapphire plate. The surface was heated to ca. 1200 K for cleaning and then was cooled to 10 K. The sample temperature was monitored at two points of the Cu rod close to the sample using Au-Fe chromel thermocouples and was controlled using a digital temperature controller (model 9650; Scientific Instruments Inc.). The H2O molecules were admitted into the UHV chamber through high-precision leak valves; then, the ASW film was deposited on the Ni(111) surface by backfilling the UHV chamber. The coverage of the water molecules was determined from the evolution curves of sputtered ion intensities as a function of exposure. One monolayer of the water molecules was attained by exposure of around 2.5 langmuir (1 L ) 1 × 10-6 Torr s). The NaCl molecule was evaporated thermally from a resistively heated Ta basket that was placed in front of the surface. TOF-SIMS spectra were recorded continually every 30 s at a ramping speed of 5 K min-1.
Souda
Figure 1. Evolutions of (a) Na+ and (b) Na+(NaCl) ion intensities in TOF-SIMS as a function of the NaCl evaporation time. Three substrates are compared: the Ni(111) surface at room temperature and the ASW films (200 ML) at 100 and 135 K.
3. Experimental Results The TOF-SIMS spectra of cations sputtered from the H2O film consist of bare protons and a series of hydrated protons, H+(H2O)n, whereas the Na+ ion is sputtered predominantly from the NaCl film, together with a much smaller amount of Na+(NaCl) ions. Upon adsorption of NaCl on the H2O film, the Na+(H2O)n ions are sputtered in addition to these species. On the other hand, the H-, O-, and OH- ions are the main species in the anion TOF-SIMS spectra from the H2O film, and the Cl- ion is sputtered from the adsorbed NaCl species. Figure 1 compares the intensities of the (a) Na+ and (b) Na+(NaCl) ions as a function of the deposition time of NaCl on the clean Ni(111) surface and the ASW film. The ASW film was grown by deposition of the 200 ML H2O molecules on the Ni(111) substrate at 100 K. The Na+ and Na+(NaCl) ions evolve more steeply from the Ni(111) surface than those from the ASW film. Although no saturation behaviors are observed clearly, we can assign the deposition time required for the 1 ML NaCl formation to 4 min from these evolution curves. The ion intensities increase more gradually when NaCl is deposited on the ASW film. The smaller Na+ intensity from the ASW film may be caused by additional formation of the Na+(H2O)n ion because of interaction with the H2O molecules, but this cannot completely explain the loss of the Na+ intensity. The NaCl molecules are likely to be incorporated in the ASW film as inferred from a more gradual increase in the ion intensities at 135 K relative to that at 100 K. Very few Na+(NaCl) ions are sputtered from the ASW film at submonolayer coverage of NaCl. Because Na+(NaCl) ion comes from the NaCl islands, its absence indicates that the adsorbed NaCl molecules have been hydrated efficiently.
Figure 2. Evolutions of Na+, Cl-, and Na+(NaCl) ions sputtered from 1 ML NaCl adsorbed on the Ni(111) substrate as a function of the coverage of the H2O molecules. Measurements were made at 100 K.
Figure 2 shows intensities of the Na+, Cl-, and Na+(NaCl) ions sputtered from NaCl (1 ML) adsorbed on the Ni(111) surface as a function of the coverage of the H2O molecules. The experiment was performed at 100 K, and the intensities were normalized relative to the values before H2O adsorption. The Na+(NaCl) ion disappears almost completely when the surface is covered with a monolayer of the H2O molecules. The Cl- ion is emitted from a shallow region of the water film, whereas the Na+ ion can pass through a considerably thick water film. Thus, the mean free path of the ion is found to differ greatly between Na+ and Cl-. The smaller the size of the ion, the longer the mean free path of the ion appears to become, as inferred from comparison with the Li+ ion from LiCl (not shown).
Hydration of NaCl on Glassy and Crystalline Water
J. Phys. Chem. B, Vol. 111, No. 38, 2007 11211
Figure 4. Intensities of typical cations sputtered from the NaCl adsorbed crystalline-ice film as a function of temperature. The crystalline film was formed by heating the ASW film (200 ML) to 165 K, and then 1 ML of NaCl was adsorbed on it at 100 K.
Figure 3. Intensities of (a) cations and (b) anions sputtered from the NaCl adsorbed ASW film as a function of temperature. The ASW film was formed at 100 K by deposition of the 200 ML H2O molecules, and 1 ML of NaCl was adsorbed on it at 10 K. The temperature was increased at a rate of 5 K min-1.
Evolutions of the typical secondary ions as a function of temperature are shown in Figure 3. The ASW film was prepared by deposition of the 200 ML H2O molecules at 100 K; then, the 1 ML NaCl molecules were adsorbed on it at 10 K. The Na+ intensity decreases gradually by increasing the temperature to 100 K; it then exhibits a steep drop at around 160 K. The H2O film evaporates at around 185 K, where the intensities of the Na+ and Ni+ ions increase. The H3O+ ion intensity is almost unchanged until the evaporation of the H2O film. Regarding the anions, the Cl- ion shows similar behavior to that of the Na+ ion, although the intensity tends to increase at temperatures higher than 80 K following the initial decrease. Dissolution of NaCl in the bulk of the water film is responsible for the sharp drop in intensities of both Na+ and Cl- ions at 160 K. The intensities of the O- and OH- ions from the H2O film increase at this temperature, which is observed for the pure H2O film as well. The H+ intensity from the pure H2O film is known to exhibit a dip at this temperature because of film dewetting.17,18 The dewetting is caused by the surface tension of the fluidized film and is observed for the films that are thinner than 80 ML. The morphology of the sufficiently thick ASW film used in the present study may not change greatly at 160 K. The enhanced solubility of NaCl in water of temperatures higher than 160 K is consistent with the previous observation that the concentrated aqueous LiCl solution is formed when the wateradsorbed polycrystalline LiCl film is heated at around this temperature.18,19 These behaviors strongly suggest that some phase transition of water occurs at 160 K. Regarding the phase transition, glassy water is known to crystallize at around 150-160 K as revealed by measurements of differential scanning calorimetry,14 Raman spectroscopy,20 thermal desorption spectroscopy,21 and infrared absorption
spectroscopy.22 However, the dissolution of NaCl at temperatures higher than 160 K cannot be explained by crystallization. The interactions of this “crystalline” ice film with the adsorbed NaCl molecules are investigated to gain more insight into the origin of the phase transition and the nature of the crystalline film. Figure 4 shows the experimental results. The crystalline film was prepared by heating the ASW film (200 ML) at 165 K; then, the NaCl molecules (1 ML) were deposited on it at 100 K. Although the Na+ intensity decreases rather gradually compared to that from the ASW film shown in Figure 3a, NaCl tends to dissolve even in the crystalline ice film at temperatures higher than 160 K. The eutectic temperature of the water/NaCl system is 251.8 K. Therefore, the aqueous NaCl solution should not be formed at the interface of the crystalline water and NaCl. This behavior strongly suggests that the water film after the phase transition cannot be regarded as comprising ordinary crystals. A liquidlike phase is thought to coexist with the crystals. The Na+ intensity in Figure 3a appears to decay in two steps: a steep decay at 160-162 K is followed by a more gradual decay at 162-175 K. The former is thought to be caused by the formation of supercooled liquid water, whose properties resemble those of normal water, whereas the latter fits to the decay curve of Na+ after crystallization occurs, as shown in Figure 4. 4. Discussion It remains an open question how amorphous water at low temperature is connected thermodynamically with normal liquid water under atmospheric pressure. Smith and Kay claimed that supercooled liquid water is formed in the temperature range of 150-158 K before crystallization occurs,21 but we denied this assumption on the basis of the infrared absorption study.19 Kim et al.11 suggested that the thermodynamic state of ASW at temperatures higher than 120 K can be regarded as an extrapolation from liquid water because of the occurrence of the complete dissociation of NaCl into the Na+ and Cl- ions. Regarding glassy states of water, two distinct phases are known to exist in the deeply supercooled region.23 Therefore, two corresponding liquid phases might exist above Tg. According to this conjecture, a distinct liquid phase, termed low-density liquid (LDL), should appear from ASW at temperatures higher than Tg ) 136 K: LDL is a tetrahedrally structured liquid that is characterized by its ultraviscous nature, and hydrophobic species can be incorporated in its bulk.17,18,24 On the other hand, the high solubility of electrolytes, leading to the formation of
11212 J. Phys. Chem. B, Vol. 111, No. 38, 2007
Figure 5. Escape depths of Na+ and Cl- ions from NaCl (1 ML) adsorbed on the ASW film (200 ML) as a function of temperature, which were converted from the TOF-SIMS intensities (Figure 3a) on the basis of the intensity-to-coverage relation (Figure 2).
uniform electrolyte solutions, is expected to occur for normal or supercooled liquid water. On the basis of the previous observations of dehydration of hydrophobes,17 formation of the LiCl solution,18,19 and dewetting,17 we assigned the phase transition at around 160 K (165 K for D2O) to the transformation of LDL into supercooled liquid water (liquid-liquid phase transition). Crystallization is thought to occur as a result of spontaneous nucleation in supercooled liquid water that is unstable below 235 K. The crystallized water film is known to exhibit a viscous nature at temperatures of 140-210 K.25 Our results have shown that the liquidlike phase coexists with crystals on the basis of combined infrared absorption and TOFSIMS measurements.19 It is important to clarify the hydration structures of the ionic species in ASW and LDL because they are thought to be distinct phases with properties that differ largely from those of liquid water. Figure 5 shows the mean escape depths of the Na+ and Cl- ions sputtered from NaCl adsorbed on the ASW film as a function of temperature. They are converted from the TOFSIMS intensities of the Na+ and Cl- ions (Figure 3) using their decay curves (Figure 2): The attenuation of the ion intensity (Figure 2) is caused by the loss of kinetic energies during collisions with the water molecules, so that the normalized ion intensities correlate with the mean free paths of the respective ions if no mixing occurs at the interface. In the present case, NaCl is neither permeated by the water molecules nor segregated to the surface because the water molecules are deposited on the NaCl monolayer at a temperature well below Tg (100 K). As seen in Figure 3, the lower ion intensities deposited on the ASW film than those after evaporation of the water film (T > 185 K) can be ascribed to the incorporation of NaCl into the bulk or subsurface sites of the film; the mean escape depths of the Na+ and Cl- ions are estimated using the results in Figure 2. Figure 5 shows that both Na+ and Cl- ions are located close to the surface immediately after deposition of the NaCl molecules. Upon heating, the Na+ ions tend to be incorporated
Souda in deeper sites of the film, whereas the surface segregation appears to occur for the Cl- ion. Consequently, the difference in the mean escape depths between the Na+ and Cl- ions finally becomes as large as ca. 1.5 ML of H2O, but the NaCl molecules fundamentally remain in the subsurface regions of ASW (T < 136 K) and LDL (136 K < T < 160 K) without diffusion into the bulk. At 160 K, the NaCl molecules can be transported to deeper film sites because of the formation of supercooled liquid water. Actually, the Na+ and Cl- ion distributions are much shallower than the film thickness (200 ML). The incomplete diffusion of the NaCl molecules into the bulk is attributable to the occurrence of crystallization, although NaCl diffuses gradually, even after the crystallization, because of the coexisting liquidlike phase, as shown in Figure 4. Density profiles that were simulated for the concentrated NaCl solutions (6.1 M)4 show the formation of a double-layer structure at the airsolution interface with the ion pairs separated by about 3 A. The separation between the Na+ and Cl- ions in Figure 5 can be estimated as 4.1 A using the O-O distance of hexagonal ice (2.75 A). The experimental result shows only qualitative agreement with those of the MD simulations. Large errors ((50%) exist in the estimated escape depths, arising from the conversion of exposure to film thickness. Moreover, because the sputtering yield of NaCl from the ASW film is thought to be lower than that from the heavier Ni(111) substrate, the ion escape depths shown in Figure 5 might be overestimated. The present result for T < 160 K is fundamentally consistent with that of a low-energy Cs+ sputtering experiment by Kim et al.;11 they revealed that the Na+ ion migrates from the surface into the interior of the ASW films at temperatures higher than 120 K. From those observations, they claimed the complete dissociation of NaCl into Na+ and Cl- ions and the resemblance of ASW to liquid water. However, the poor solubility of NaCl below 160 K is not expected for liquid water. Moreover, the concentrated solution that might be formed only at the surface region should not yield the completely separated Na+-Cl- ion pairs.4 The theoretical calculations of NaCl(H2O)6 clusters predicted the presence of two nearly isoenergetic isomers;5 one is a solvent-separated ion pair and the other is a contact ion pair, and both are surrounded by the water hexamer. Nevertheless, their hydrogen bond structures are largely different. In the contact ion pair configuration, the Na+ ion is almost completely embedded in the water cluster, whereas Cl- remains unhydrated at the surface. The hydrogen bonds of the water molecules surrounding the Na+ ion are shown to be almost intact. On the other hand, the water molecules surrounding the separated Na+Cl- ion pair are aligned along the electric field between the ions with very little residual hydrogen bonding between the water molecules. The strong hydration of the Na+ ion for the contact ion pair configuration is consistent with the present observation seen in Figure 5. The absence of the sputtered Na+ ions by low-energy (35 eV) Cs+ ions11 might also be explained using the formation of the hydrated Na+ ions without assuming the complete dissociation into Na+ and Cl- ions. The contact ion pair is expected to be formed for ASW and LDL that have local-order similarities to crystalline ice.13 On the other hand, the hydrogen bond of liquid water is thought to be distorted from a perfectly tetrahedral configuration, that is, a mixture of straight and strong hydrogen bonds and weak and bent hydrogen bonds, leading to the higher density or coordination number of liquid water than crystalline water. Therefore, the solventseparated ion pair is likely to be created in (supercooled) liquid water.
Hydration of NaCl on Glassy and Crystalline Water The residence of the NaCl molecules in the near surface region of ASW and LDL is explicable by the surfactant effect of the contact ion pairs having hydrophobic (Cl-) and hydrophilic (Na+) moieties; the Na+ ion is hydrated preferentially by the hydrogen-bonded water hexamer. The hydration of the large Cl- ion requires water molecules with more distorted hydrogen bonds,4 so that NaCl can be incorporated in the bulk after liquid (or supercooled liquid) water evolves. According to the Debye-Huckel theory, the ions are practically unpaired in the case of infinite dilution. The NaCl solution formed at temperatures higher than 160 K might possess completely separated Na+-Cl- pairs; they have much less surfactant ability than the contact ion pairs, thereby enabling ion diffusion into the bulk. In reality, however, the dilute NaCl solution should undergo phase separation into pure crystal ice Ic and more concentrated solutions as inferred from the results of hyperquenched LiCl solutions.26,27 The MD simulation revealed that the ion pairing occurs as the dominant process in the saturated solutions.4 Therefore, the solvent-separated ion pairs are formed only transiently during the liquid-liquid phase transition at 160 K and become minor species after the crystallization occurs. The enhanced solvation of ionic salts at temperatures around 160 K has not been observed in the previous studies for NaCl (0.8 ML) on D2O (4 ML)11 and CsI on H2O (10 ML).12 At least six water molecules are necessary for bulk hydration of a NaCl molecule.5 Therefore, the thin ASW films used in these studies might not engender the solvent-separated ion pairs, in contrast to the present observation. 5. Conclusion The NaCl species tend to remain at the surface or in subsurface sites of the ASW film. The same is true for the viscous liquid (LDL) phase, which appears above Tg ) 136 K, indicating that the bulk hydration of NaCl is not induced simply by the self-diffusion of the water molecules. According to the MD simulation, stronger segregation (hydration) of the Cl(Na+) ion appears to occur for the contact ion pair configuration than for the solvent-separated ion pairs.5 Hydrogen bonds of the water molecules are much less disrupted in the former, so that the contact ion pair is more likely to be formed on the surface of ASW and LDL, which have local structure similarities to the crystalline ice. The surface segregation of NaCl can be attributed to the surfactant effect of the contact ion pair. The properties of water change drastically at around 160 K, where supercooled liquid water is formed (the liquid-liquid phase transition). The NaCl species tend to be incorporated in the bulk of supercooled liquid water because the Cl- ion can be hydrated
J. Phys. Chem. B, Vol. 111, No. 38, 2007 11213 after the formation of the solvent-separated ion pair. Because of the occurrence of spontaneous nucleation, however, the diluted NaCl solution, if any, would undergo phase separation into crystalline ice Ic and concentrated solutions immediately. The ASW cannot be regarded as a simple analogue of liquid water, and supercooled liquid water is formed only temporarily at around 160 K. Consequently, the completely separated Na+Cl- ion pair might not be a dominant species at cryogenic temperatures. The hydration behavior of NaCl on the crystalline ice film resembles that on the ASW film because crystallization to ice Ic is not complete; viscous liquid water coexists with crystals and plays a role in the gradual incorporation of NaCl in the film at temperatures higher than 160 K. References and Notes (1) Oum, K. W.; Lakin, M. J.; DeHaan, D. O.; Brauers, T.; FinlaysonPitts, B. J. Science 1998, 279, 74. (2) Knipping, E. M.; Lakin, M. J.; Foster, K. L.; Jungwirth, P.; Tobias, D. J.; Gerber, R. B.; Dabdub, D.; Finlayson-Pitts, B. J. Science 2000, 288, 301. (3) Onsager, L.; Samaras, N. N. T. J. Chem. Phys. 1934, 2, 528. (4) Jungwirth, P.; Tobias, D. J. J. Phys. Chem. B 2000, 104, 7702. (5) Jungwirth, P.; Tobias, D. J. J. Phys. Chem. B 2002, 106, 6361. (6) Garret, B. C. Science 2004, 303, 1146. (7) Raymond, E. A.; Richmond, G. L. J. Phys. Chem. B 2004, 108, 5051. (8) Peterson, P. B.; Johnson, J. C.; Knutsen, K. P.; Saykalley, R. J. Chem. Phys. Lett. 2004, 397, 46. (9) Weber, R.; Winter, B.; Schmidt, P. M.; Widdra, W.; Hertel, I. V.; Dittmar, M.; Faubel, M. J. Phys. Chem. B 2004, 108, 4729. (10) Ghosal, S.; Hemminger, J. C.; Bluhm, H.; Mun, B. S.; Hebenstreit, E. L. D.; Ketteler, G.; Ogletree, D. F.; Requejo, F. G.; Salmeron, M. Science 2005, 307, 563. (11) Kim, J. H.; Shin, T.; Jung, K. H.; Kang, H. Chem. Phys. Chem. 2005, 6, 440. (12) Hofft, O.; Borodin, A.; Kahnert, U.; Kempter, V.; Dang, L. X.; Jungwirth, P. J. Phys. Chem. B 2006, 110, 11971. (13) Finney, J. L.; Hallbrucker, A.; Kohl, I.; Soper, A. K.; Bowron, D. T. Phys. ReV. Lett. 2002, 88, 225503 (14) Hallbrucker, A.; Mayer, E.; Johari, G. P. J. Phys. Chem. 1989, 93, 4986. (15) Fisher, M.; Devlin, J. P. J. Phys. Chem. 1995, 99, 11584. (16) Souda, R. Phys. ReV. Lett. 2004, 93, 235502. (17) Souda, R. J. Chem. Phys. 2004, 121, 8676. (18) Souda, R. J. Chem. Phys. 2006, 125, 181103. (19) Souda, R. J. Phys. Chem. B 2007, 111, 6528. (20) Tulk, C. A.; Klug, D. D.; Branderhorst, R.; Sharpe, P.; Ripmeester, J. A. J. Chem. Phys. 1998, 109, 8478. (21) Smith, R. S.; Kay, B. D. Nature (London) 1999, 398, 788. (22) Ostblom, M.; Ekeroth, J.; Konradsson, P.; Liedberg, B. J. Phys. Chem. B 2006, 110, 1695. (23) Mishima, O.; Stanley, H. E. Nature (London) 1998, 396, 329. (24) Paschek, D. Phys. ReV. Lett. 2005, 94, 217802. (25) Jenniskens, P.; Banham, S. F.; Blake, D. F.; McCoustra, M. R. J. Chem. Phys. 1997, 107, 1232. (26) Angell, C. A.; Sare, E. J. J. Chem. Phys. 1968, 49, 4713. (27) Suzuki, Y.; Mishima, O. Phys. ReV. Lett. 2000, 85, 1322.