Spectral Signatures of Four-Coordinated Sites in Water Clusters

Dec 23, 2010 - Spectral Signatures of Four-Coordinated Sites in Water Clusters: Infrared Spectroscopy of Phenolr(H2O)n (∼20 e n e ∼50). Toru Hamas...
0 downloads 0 Views 2MB Size
Download PDF
ARTICLE pubs.acs.org/JPCA

Spectral Signatures of Four-Coordinated Sites in Water Clusters: Infrared Spectroscopy of Phenol-(H2O)n (∼20 e n e ∼50) Toru Hamashima, Kenta Mizuse, and Asuka Fujii* Department of Chemistry, Graduate School of Science, Tohoku University, Sendai 980-8578, Japan ABSTRACT: We report infrared spectra of phenol-(H2O)n (∼20 e n e ∼50) in the OH stretching vibrational region. Phenol-(H2O)n forms essentially the same hydrogen bond (H-bond) network as that of the neat water cluster, (H2O)nþ1. The phenyl group enables us to apply the scheme of infrared-ultraviolet double resonance spectroscopy combined with mass spectrometry, achieving the moderate size selectivity (0 e Δn e ∼6). The observed spectra show clear decrease of the free OH stretch band intensity relative to that of the H-bonded OH band with increasing cluster size n. This indicates increase of the relative weight of four-coordinated water sites, which have no free OH. Corresponding to the suppression of the free OH band, the absorption peak of the H-bonded OH stretch band rises at ∼3350 cm-1. This spectral change is interpreted in terms of a signature of four-coordinated water sites in the clusters.

’ INTRODUCTION A fully hydrated water molecule is four-coordinated, and its spectral signature is expected in the hydrogen-bonded (H-bonded) OH stretch region of the infrared (IR) spectra. While the OH stretch band of ice is uniquely attributed to the infinite H-bond network only of four-coordinated waters,1,2 it has been controversial to decompose the OH stretch band of liquid water to different coordination components.3-9 Despite many proposals to determine the coordination numbers of liquid water, it still remains an open question. Identification of four-coordinated water bands in gas phase water clusters, (H2O)n, is then highly important because spectroscopic investigations of size-selected clusters combined with theoretical calculations would enable us to precisely examine the correlation between H-bonded OH stretch frequencies and finite H-bond networks involving fourcoordinated waters. IR spectra of size-selected (H2O)n have been reported up to n ∼ 10.10-14 For n g 6, H-bond networks in the minimum energy structures become three-dimensional cages. In these sizes of clusters, however, all the molecules locate at the cluster surface. Their maximum coordination number is still three, and fourcoordinated sites cannot exist in low energy stable structures. A larger number of molecules is obviously needed to accommodate a few molecules inside clusters. Such “centered cage” structures with interior water(s) have been predicted by theoretical calculations to exist only for n g 17.15-31 Each interior water molecule is fourcoordinated, and surrounding four water molecules, most of which are on the cluster surface, also become four-coordinated. In the size range of a few tens, “multiple fused cages” constructed by cube and pentaprism building blocks are alternative structures competitive to r 2010 American Chemical Society

“centered cages”. Though “multiple fused cages” are all surface type structures, four-coordinated sites are generated at shared edges between building blocks. Therefore, whichever type of structures may be preferred, spectral signatures of four-coordinated water should appear in the size region of a few tens and larger. Structural studies of (H2O)n in the size region of n = 10-50 have been led by theoretical calculations,15-31 and experimental studies have been very scarce.22,32-37 IR spectra of (H2O)n under the jet expansion22,32-34,36,37 and cold gas cell condensation22,32,35 have been reported, but only the average size could be controlled in both the cases. The reported spectra are a mixture of contributions from clusters of the wide size range. Therefore, though the spectra are consistent with the network development trend predicted by theoretical calculations, size dependence of the spectra could not be precisely discussed. Size-selective IR spectroscopy has been performed for Na(H2O)ne60,38 Hþ(H2O)ne200,39-44 Ca2þ(H2O)ne69,45 and SO42-(H2O)ne8046 in the OH stretch region. However, solvation of the electron released from the Na atom significantly contaminates the expected four-coordinated water band region in the spectra of Na(H2O)n. Also for the other clusters, solvation of the charged site would change the H-bond network structures from those in neat (H2O)n. Very recently, we have demonstrated size-selective IR spectroscopy of phenol-(H2O)n (∼10 e n e ∼50) to probe neutral H-bond networks consisting of tens of water molecules.47 A phenol molecule is compatible with a water molecule on the surface of water H-bond networks, and network structures of phenol-(H2O)n will be equivalent with those of neat (H2O)nþ1, Received: December 6, 2010 Published: December 23, 2010 620

dx.doi.org/10.1021/jp111586p | J. Phys. Chem. A 2011, 115, 620–625

The Journal of Physical Chemistry A

ARTICLE

as has been demonstrated for the relatively small sized clusters (n = 1-8).13,48-51 The phenol molecule is introduced as an inert chromophore, and the infrared-ultraviolet (IR-UV) double resonance scheme13,14 is applied for this size region. The sizeselection is performed by resonance-enhanced two-photon ionization with the broadened S1-S0 electronic transition excitation and following mass spectrometry. In consideration for the dissociation after the ionization, the moderate size selectivity (0 e Δn e ∼6) is achieved in this method.47 We have reported the spectra in the free (dangling) OH stretch region of the clusters as a sensitive probe of the H-bond network distortion and have shown the development of the network motifs from the dominance of the highly strained four-membered rings in n ∼ 10 to more relaxed five- and six-membered rings with increasing n. In the previous study, we have focused only on the free OH stretch frequency behavior, and four-coordinated sites have not been discussed at all, though the sizes of the observed clusters are large enough to have four-coordinated sites. In the present paper, we extend the measurement to the whole OH stretch region of these phenol-(H2O)n clusters. The intensity behavior of the free OH stretch band relative to the H-bonded OH stretch band is examined. Moreover, the H-bonded OH stretch band also shows size-dependent changes. With a help of theoretical simulations, we discuss spectroscopic signatures of four-coordinated sites in the finite water H-bond networks.

’ EXPERIMENTAL SECTION IR spectra of size-selected phenol-(H2O)n (n = 19-49) were measured by the modified IR-UV double resonance scheme. Details of the experiments are reported in ref 47. Briefly, phenol(H2O)n clusters were produced by a pulsed supersonic jet expansion, and the jet was skimmed to form a molecular beam. [Phenol(H2O)n]þ ions produced by the two-photon ionization of neutral clusters were mass selected and detected as a measure of the ground state population of phenol-(H2O)nþΔn. Here, Δn is the number of water molecules evaporated upon the ionization and is estimated to be 0 e Δn e ∼6 by the ionization energetics calculations. The UV frequency was tuned to 36254 cm-1, which is near the maximum of the broad absorption of large-sized clusters. An IR laser pulse was introduced prior to the UV laser pulse, and its frequency was scanned. When an IR transition occurs in phenol(H2O)nþΔn, the monitored [phenol-(H2O)n]þ signal decreases because of the vibrational predissociation of the clusters. Then IR spectra of the size-selected clusters were measured as dip spectra by monitoring the ion intensities while scanning the IR frequency. The wavelength of the IR light was calibrated by recording an ambient water vapor spectrum. The presented IR spectra were normalized by the IR power. The IR light was generated by difference frequency mixing between the second harmonics of a Nd:YAG laser and an output of a YAG laser pumped dye laser (DCM) with a LiNbO3 crystal. A Wiley-McLaren type time-of-flight mass spectrometer was used for mass spectrometry. In the present measurement, a part of the experimental conditions is modified to improve the quality of observed spectra. The pulsed jet nozzle and the pulsed UV laser were operated at 10 Hz while the pulsed IR laser was at 5 Hz. Then the active baseline subtraction technique was applied to reduce artifacts by fluctuation of the signal baseline. Stable structures of clusters were calculated by the density functional theory. All calculations in this study were carried out

Figure 1. Infrared (IR) spectra of moderately size-selected phenol(H2O)nþΔn in the OH stretching vibrational region. Each spectrum is obtained by monitoring the [phenol-(H2O)n]þ cluster ion intensity. The size uncertainty (Δn) is estimated to be 0 e Δn e ∼6. The depletion depth of the ion intensity is plotted. The gap in the spectra around 3500 cm-1 is due to the depletion of the output of the IR light source (see text).

using the Gaussian 03 program,52 and the cluster structures were visualized with the MOLEKEL program.53

’ RESULTS AND DISCUSSION Figure 1 presents the observed IR spectra of the size-selected phenol-(H2O)nþΔn (n = 19-49) in the whole OH stretching vibrational region. The size selectivity (Δn) is estimated to be 0 e Δn e ∼6.47 The gap around 3500 cm-1 is due to the IR absorption by water impurity in the LiNbO3 crystal to generate the IR light. These spectra show a sharp band near 3700 cm-1, which is assigned to the free OH stretching vibrations of dominantly three-coordinated water molecules, and the broadened absorption below 3600 cm-1 is attributed to the H-bonded OH stretching vibrations. As for the free OH band, we have already reported its slight but definite frequency shift with the cluster size and its implication on the H-bond network structures.47 In the present study, we note only the intensity of the free OH band relative to the H-bonded OH band. This is new information obtained by the measurement of the whole OH stretch region. The relative intensity of the free OH band gradually decreases with increasing n. In the observed size region, the clusters will form three-dimensional structures, and the observed free OH band is mainly attributed to threecoordinated waters on the cluster surface. On the other hand, four-coordinated waters have no free OH. Because increase of the relative abundance of four-coordinated waters results in decrease of the ratio of three-coordinated waters, the observed decrease of the free OH band intensity is regarded as an 621

dx.doi.org/10.1021/jp111586p |J. Phys. Chem. A 2011, 115, 620–625

The Journal of Physical Chemistry A

ARTICLE

indication of increase of four-coordinated waters. The similar trend of the free OH band has been reported for the average sizecontrolled neat water clusters produced by cold gas cell condensation.22,36 However, the reported spectra suffer from wide size distribution and/or influence of N2 adsorption on the cluster surface. On the other hand, the present spectra are measured with much better size selectivity and are free from perturbations of extra species. Therefore, the size dependence of the present spectra would be more reliable. The IR spectra of average-sized controlled neat water clusters have been also studied under the molecular beam condition, but no clear suppression of the free OH intensity has been seen with the change of the average size from nav = 40 to 1960.34 In this study, the authors have noted that their detection method (dissociation fragment detection by electron bombardment ionization) is sensitive to the cluster surface. Corresponding to the indication of four-coordinated water sites by the free OH band intensity, direct spectral signatures of the four-coordinated sites are expected in the H-bonded OH stretch region. In the spectrum of n = 19, the H-bonded OH band is very broad, and it shows almost flat intensity distribution in the range of 3000-3600 cm-1. With increasing cluster size, the absorption at around 3200-3500 cm-1 gradually rises, while the intensities in both the edge regions of the H-bonded OH band are suppressed. In the spectrum of n = 49, the band finally forms a small peak at around 3350 cm-1. The absorption peak at ∼3350 cm-1 has been observed also in the average-size controlled water clusters,22,34,36 but the peak formation process was unclear in the previous studies probably due to the wide size distribution. Because this spectral change in the H-bonded OH band correlates with the suppression of the free OH band, it is reasonable to attribute the rise of the absorption centered at ∼3350 cm-1 to the increase of four-coordinated sites. To examine the origin of the spectral change of the H-bonded OH stretch band, we performed theoretical calculations. Prior to description on the theoretical calculations, it may be helpful to confirm the character of the spectral carrier of the observed spectra. Because of the fragmentation of water molecules on the ionization, the present spectroscopic method is not rigorously size selective.47 Contribution of larger-sized clusters (0 e Δn e ∼6) may be involved. In addition, even in each cluster size, numerous isomers would contribute to the observed spectra. Therefore, extensive sampling of possible isomers by molecular dynamics or Monte Carlo calculations and weighted summation of all the isomer contributions at the finite temperature are requested for full analyses of the observed spectra. Such sampling is, however, still challenging even for smaller sizes,15-31,54 and this approach is beyond the focus of the present study. Then as a practical choice at the present stage, we decide to pick up the minimum energy isomers reported by the previous theoretical studies,19,21,30 for qualitative characterization of the observed spectral features. Since we produce the clusters by a supersonic jet expansion, the observed clusters would be vibrationally cold (typically below ∼100 K,55 but the broadened electronic transition prevents us from evaluating the actual vibrational temperature by the hot band intensity). Therefore, the minimum energy isomer is a reasonable choice to compare with the observed spectra. Moreover, the clusters would be “frozen”, and dynamical effects discussed in warm ionic clusters may not be important.42,56 We employed DFT calculations at the B3LYP/ 6-31þG(d) level to compare with the observed spectra.

Figure 2. Comparison between (a) observed IR spectrum of phenol(H2O)29 and (b-d) simulated spectra of (H2O)30. The simulated spectra are based on the global minimum structures at the TIP3P, TIP4P, and TTM2-F potentials, respectively, which are reoptimized at the B3LYP/6-31þG(d) level (see text). In the simulations, the black trace represents the IR spectrum, and the green, blue, and red traces are decomposition of the spectrum to the two-, three-, and four-coordinated water components, respectively.

Figure 2 displays the comparison between the observed and simulated IR spectra of phenol-(H2O)29. Spectrum a is a reproduction of the observed IR spectrum. Spectra b-d are the harmonic vibrational simulations of (H2O)30, which is supposed to have the same H-bond network structure as phenol(H2O)29 (in our previous work,47 we have confirmed by the B3LYP level calculations that the replacement of one water molecule on the cluster surface to a phenol molecule results in essentially the same vibrational spectrum at the cluster size of n = 20). These simulations are based on the stable structures shown in the insets. These three structures are obtained by reoptimization of the global minimum of (H2O)30 under the TIP3P and TIP4P potentials reported by Takeuchi,30 and under the TTM2-F potential by Bandow and Hartke.19 In the simulations, the black trace indicates the simulated IR spectrum and the blue and red traces show the decomposition of the simulated spectrum to the three- and four-coordinated water components, respectively. Only the TIP3P structure has minor contribution of a two-coordinated water, which is shown as the green trace. In this decomposition, we first determine the coordination number of each water molecule in the cluster. Then we analyze each normal mode and calculate relative weight of each local OH oscillator. The relative weight of this normal mode into the coordination number component is distributed according to the sum of the relative weights of the local OH oscillators, which belong to this coordination number. All the calculated frequencies 622

dx.doi.org/10.1021/jp111586p |J. Phys. Chem. A 2011, 115, 620–625

The Journal of Physical Chemistry A

ARTICLE

were scaled by a single factor of 0.9736. This scaling factor has been used to reproduce the free OH frequency of the clusters.47 The calculated stick spectra were transformed into the continuous IR spectra by using Lorentzian functions with 10 and 50 cm-1 full widths at half-maximum for each free OH mode and H-bonded OH mode, respectively. These global minimum structures of (H2O)30 are “centered cage” types including four water molecules inside the cage. The H-bond networks are essentially identical between the TIP4P and TTM2-F potentials, but that by the TIP3P potential is different from the others. According to the H-bond structures, the simulated spectra c and d are almost the same while spectrum b is different from the others. The decomposed spectra of the three- and four-coordinated water components are also different between TIP3P and TIP4P (or TTM2-F), but each component shows common trends. The decomposed spectra of three-coordinated waters have the wide and flat absorption in the range of 3000-3600 cm-1, and a weak peak and a shoulder are seen at ∼3100 and ∼3500 cm-1, respectively. These two features are attributed to double acceptor single donor (AAD) sites and single acceptor double donor (ADD) sites, respectively. These features of three-coordinated sites have been actually observed in the small-sized clusters (7 e n e 10).12,14 On the other hand, the absorption of four-coordinated waters is concentrated in the narrower range of 3200-3500 cm-1 and a peak is seen at 3300-3400 cm-1. These band decompositions well agree with the previous calculations for water clusters of the similar sizes.5,22,34 From these simulations, it is expected that the sum of the contributions of the three- and four-coordinate waters results in relatively wide and flat absorption when four-coordinated waters are minor, and increase of the contribution of fourcoordinated sites causes the rise of the absorption maximum centered at around 3300-3400 cm-1. As seen in Figure 2a, the observed spectrum of phenol(H2O)29 shows the wide and flat absorption in the range 3000-3600 cm-1, and reproduction by the simulations is rather poor. It should be, however, noted that the aim of the present simulation is to show the trend of the correlation between the H-bond coordination number and the H-bonded OH stretch band position, and a single isomer intrinsically cannot reproduce the observed spectrum, which is due to an ensemble of numerous isomers. In the minimum energy structures, the number of fourcoordinated sites is almost equal to that of three-coordinated sites: 16 molecules are on four-coordniated sites and 14 water molecules locate on three-coodinated sites in the TIP4P (and TTM2-F) structure. The poor reproduction suggests that the minimum energy structures overestimate the contribution of the four-coordinated site in comparison with the observed ensemble of isomers. For the size range of n = 30-34, the minimum energy structures of (H2O)n at the TTM2-F potential have been reported by Bandow and Hartke.19 The reported structures are reoptimized at the B3LYP/6-31þG(d) level, and their IR spectra are calculated at the same level. Figure 3 shows the simulated IR spectra of n = 30-34 and their decompositions into the three- and fourcoorrdinated water components. Only n = 33 has the contributions of exceptional two- and five-coordinated waters. The calculated band shapes remarkably change with the cluster size. This suggests that the IR spectrum is still sensitive to the H-bond structure in this size range, though an ensemble of isomers, which is actually observed, may easily wash out details of single isomer spectra. We also note that the decomposed spectra show the same trends as

Figure 3. Simulated IR spectra of (H2O)n in the OH stretch region. (a) n = 30, (b) n = 31, (c) n = 32, (d) n = 33, and (e) n = 34. The simulated spectra are based on the global minimum structures at the TTM2-F potential, which are reoptimized at the B3LYP/6-31þG(d) level (see text). In the simulations, the black trace represents the IR spectrum, and the blue and red traces are decomposition of the spectrum to the threeand four-coordinated water components, respectively. Spectrum d has the minor contributions of two- and five-coordinated sites, which are shown as the green and violet traces, respectively.

those found for n = 30: the three-coordinated water component has wide and rather flat absorption, while absorption by the fourcoordinated component is centered at ∼3300-3400 cm-1. These simulations imply that increase of four-coordinated sites results in rise of the absorption at ∼3300-3400 cm-1. To make further confirmation on the contribution of fourcoordinated waters, a similar comparison is carried out among the observed IR spectrum of phenol-(H2O)49 and simulated ones of (H2O)48, as shown in Figure 4. The global minimum structure of (H2O)48 under the TIP4P potential has been reported by Kazimirski et al.21 We reoptimized this structure at B3LYP, and the resultant spectral simulation is shown in Figure 4c. To highlight the feature of three-coordinated waters, we also construct a model structure consisting only of three-coodinated waters. This structure becomes a hollow-cage type and is shown with its simulated spectrum in Figure 4b. As is expected by the simulations on (H2O)30-34, spectrum b shows the broad absorption in 3000-3600 cm-1 without a single prominent maximum. This reveals that the peak at ∼3350 cm-1 in the observed spectrum (spectrum a) cannot be rationalized only with three-coordinated waters. The free OH band in spectrum b is also too intense relative to the H-bonded OH band, in comparison with spectrum a. On the other hand, spectrum c shows qualitative agreement with spectrum a; a peak appears at around 3300 cm-1, and the relative 623

dx.doi.org/10.1021/jp111586p |J. Phys. Chem. A 2011, 115, 620–625

The Journal of Physical Chemistry A

ARTICLE

’ AUTHOR INFORMATION Corresponding Author

*E-mail: [email protected].

’ ACKNOWLEDGMENT We thank Professor Naohiko Mikami in Tohoku University for his kind encouragement. We also thank Dr. Toshihiko Maeyama and Dr. Yoshiyuki Matsuda in Tohoku University, and Dr. Jer-Lai Kuo in Institute of Atomic and Molecular Sciences for their valuable discussion. This study was supported by the Grant-in-Aid for Scientific Research (Project No. 19056001 from MEXT Japan, and Nos. 20 3 5015 and 2235001 from JSPS). K.M. is supported by JSPS Research Fellowships for Young Scientists. Most of the calculations were performed using supercomputing resources at Cyberscience Center, Tohoku University. ’ REFERENCES (1) Eisenberg, D.; Kauzmann, W. J. The structure and properties of water; Oxford at the Clarendon Press: London, 1969. (2) Marechal, Y. The Hydrogen Bond and the Water Molecule; Elsevier: Amsterdam, 2007. (3) Brubach, J.-B.; Mermet, A.; Filabozzi, A.; Gerschel, A.; Roy, P. J. Chem. Phys. 2005, 122, 184509. (4) Ohno, K.; Okimura, M.; Akai, N.; Katsumoto, Y. Phys. Chem. Chem. Phys. 2005, 7, 3005. (5) Lenz, A.; Ojam€ae, L. J. Phys. Chem. A 2006, 110, 13388. (6) Schmidt, D. A.; Miki, K. J. Phys. Chem. A 2007, 111, 10119. (7) Leetmaa, M.; Wikfeldt, K. T.; Ljungberg, M. P.; Odelius, M.; Swenson, J.; Nilsson, A.; Pettersson, L. G. M. J. Chem. Phys. 2008, 129, 084502. (8) Paesani, F.; Voth, G. A. J. Phys. Chem. B 2009, 113, 5702. (9) Zhang, C.; Donadio, D.; Galli, G. J. Phys. Chem. Lett. 2010, 1, 1398. (10) Vernon, M. F.; Krajnovich, D. J.; Kwok, H. S.; Lisy, J. M.; Shen, Y. R.; Lee, Y. T. J. Chem. Phys. 1982, 77, 47. (11) Liu, K.; Cruzan, J. D.; Saykally, R. J. Science 1996, 271, 929. (12) Buck, U.; Huisken, F. Chem. Rev. 2000, 100, 3863. (13) Watanabe, T.; Ebata, T.; Tanabe, S.; Mikami, N. J. Chem. Phys. 1996, 105, 408. (14) Gruenloh, C. J.; Carney, J. R.; Arrington, C. A.; Zwier, T. S.; Fredericks, S. Y.; Jordan, K. D. Science 1997, 276, 1678. (15) Ludwig, R. Angew. Chem., Int. Ed. 2001, 40, 1808. (16) Khan, A. J. Chem. Phys. 1997, 106, 5537. (17) Wales, D. J.; Hodges, M. W. Chem. Phys. Lett. 1998, 286, 65. (18) Hartke, B. Phys. Chem. Chem. Phys. 2003, 5, 275. (19) Bandow, B.; Hartke, B. J. Phys. Chem. A 2006, 110, 5809. (20) Kabrede, H.; Hentschke, R. J. Phys. Chem. B 2003, 107, 3914. (21) Kazimirski, J. K.; Buch, V. J. Phys. Chem. A 2003, 107, 9762. (22) Buch, V.; Bauerecker, S.; Devlin, J. P.; Buck, U.; Kazimirski, J. K. Int. Rev. Phys. Chem. 2004, 23, 375. (23) Lenz, A.; Ojam€ae, L. Phys. Chem. Chem. Phys. 2005, 7, 1905. (24) Fanourgakis, G. S.; Apra, E.; de Jong, W. A.; Xantheas, S. S. J. Chem. Phys. 2004, 121, 2655. (25) Fanourgakis, G. S.; Apra, E.; Xantheas, S. S. J. Chem. Phys. 2005, 122, 13304. (26) Lagutschenkov, A.; Fanourgakis, G. S.; Niedner-Schatterburg, G.; Xantheas, S. S. J. Chem. Phys. 2005, 122, 194310. (27) Yoo, S.; Apra, E.; Zeng, X. C.; Xantheas, S. S. J. Phys. Chem. Lett. 2010, 1, 3122. (28) James, T.; Wales, D. J.; Hernandez-Rojas, J. Chem. Phys. Lett. 2005, 415, 302. (29) Kabrede, H. Chem. Phys. Lett. 2006, 430, 336. (30) Takeuchi, H. J. Chem. Inf. Model. 2008, 48, 2226.

Figure 4. Comparison between (a) observed IR spectrum of phenol(H2O)49 and (b, c) simulated spectra of (H2O)48. Simulated spectrum b is based on all three-coordinated water structure shown in the inset. Spectrum c is given by the global minimum structure at the TIP4P potential, which is taken from ref 21. Both the structures are (re)optimized at the B3LYP/6-31þG(d) level. In the simulations, the black trace represents the IR spectrum, and the green, blue, and red traces are decomposition of the spectrum to the two-, three-, and fourcoordinated water components, respectively.

intensity of the free OH stretch band is similar to the observed one. The global minimum structure of (H2O)48 is constructed by 27, 20, and 1 water molecules in the four-, three-, and twocoordinated sites, respectively.21 The contribution of the fourcoordinated waters is dominant in the simulated spectrum, and it causes the apparent maximum in the H-bonded OH region. These results show that the absorption maximum centered at ∼3350 cm-1 is reasonably interpreted as a spectral signature of four-coordinated water sites in the neutral water clusters of n e ∼50. Formation of four-coordinated sites is an important step in the development of H-bond networks from water clusters to bulk water, and the present study demonstrates its very early stage. The H-bonded OH stretch frequency of ice (Ih) is 3220 cm-1,1,2 and it is much lower than the present observation. This means that the OH frequency of four-coordinated sites in the clusters will show further lowering with the expansion of the network among fourcoordinated waters. Such a shift has been observed in much largersized clusters/ice nanoparticles.22,34 In conclusion, IR spectroscopy of moderately size-selected phenol-(H 2O)n in the whole OH stretching region demonstrated the infrared spectral signatures of four-coordinated waters in the water clusters. The free OH band shows the suppression of its relative intensity with increase of n, representing the increase of the abundance of four-coordinated waters. The rise of the absorption maximum of the H-bonded OH stretch band at ∼3350 cm-1 is also attributed to four-coordinated water sites. This is consistent with the previous theoretical studies and the average-size controlled experiments.5,22,34,36 624

dx.doi.org/10.1021/jp111586p |J. Phys. Chem. A 2011, 115, 620–625

The Journal of Physical Chemistry A

ARTICLE

(31) Bandyopadhyay, P. Chem. Phys. Lett. 2010, 487, 133. (32) Buch, V.; Devlin, J. P. (Eds.) Water in confining geometries; Springer-Verlag: Berlin, 2003. (33) Page, R. H.; Vernon, M. F.; Shen, Y. R.; Lee, Y. T. Chem. Phys. Lett. 1987, 141, 1. (34) Steinbach, C.; Andersson, P.; Kazimirski, J. K.; Buck, U.; Buch, V.; Beu, T. A. J. Phys. Chem. A 2004, 108, 6165. (35) Devlin, J. P.; Sadlej, J.; Buch, V. J. Phys. Chem. A 2001, 105, 974. (36) Paul, J. B.; Collier, C. P.; Saykally, R. J.; Scherer, J. J.; O'Keef, A. J. Phys. Chem. A 1997, 101, 5211. (37) Goss, L. M.; Sharpe, S. W.; Blake, T. A.; Vaida, V.; Brault, J. W. J. Phys. Chem. A 1999, 103, 8260. (38) Steinbach, C.; Buck, U. J. Phys. Chem. A 2006, 110, 3128. (39) Miyazaki, M.; Fujii, A.; Ebata, T.; Mikami, N. Science 2004, 304, 1134. (40) Shin, J. W.; Hammer, N. I.; Diken, E. G.; Johnson, M. A.; Walters, R. S.; Jaeger, T. D.; Duncan, M. A.; Christie, R. A.; Jordan, K. D. Science 2004, 304, 1137. (41) Wu, C.-C.; Lin, C.-K.; Chang, H.-C.; Jiang, J.-C.; Kuo, J.-L.; Klein, M. L. J. Chem. Phys. 2005, 122, 074315. (42) Douberly, G. E.; Ricks, A. M.; Duncan, M. A. J. Phys. Chem. A 2009, 113, 8449. (43) Mizuse, K.; Fujii, A.; Mikami, N. J. Chem. Phys. 2007, 126, 231101. (44) Mizuse, K.; Mikami, N.; Fujii Angew. Chem., Int. Ed. 2010, 49, 10119. (45) Bush, M. F.; Saykally, R. J.; Williams, E. R. J. Am. Chem. Soc. 2008, 130, 15482. (46) O'Brien, J. T.; Prell, J. S.; Bush, M. F.; Williams, E. R. J. Am. Chem. Soc. 2010, 132, 8248. (47) Mizuse, K.; Hamashima, T.; Fujii J. Phys. Chem. A 2009, 113, 12134. (48) Watanabe, H.; Iwata, S. J. Chem. Phys. 1996, 105, 420. (49) Roth, W.; Scmitt, M.; Jacoby, Ch.; Spangenberg, D.; Janzen, Ch.; Kleinermanns, K. Chem. Phys. 1998, 239, 1. (50) Janzen, Ch.; Spangenber, D.; Roth, W.; Kleinermanns, K. J. Chem. Phys. 1999, 110, 9898. (51) Kryachko, E.; Nakatsuji, H. J. Phys. Chem. A 2002, 106, 731. (52) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Montgomery, J. A.Jr.; Vreven, T.; Kudin, K. N.; Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone, V.; Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.; Petersson, G. A.; Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Klene, M.; Li, X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.; Morokuma, K.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich, S.; Daniels, A. D.; Strain, M. C.; Farkas, O.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.; Clifford, S.; Cioslowski, J.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.; Johnson, B.; Chen, W.; Wong, M. W.; Gonzalez, C.; Pople, J. A. Gaussian 03, Revision E.01; Gaussian, Inc.: Wallingford, CT, 2004. (53) Flukiger, P.; Luthi, H. P. ; Portmann, S. ; Weber, J. MOLEKEL 4.3; Swiss Center for Scientific Computing: Manno, Switzerland, 2000-2002. (54) Nguyen, Q. C.; Ong, Y.- S.; Kuo, J. -L. J. Chem. Theory Comput. 2009, 5, 2629. (55) Mikami, N.; Hiraya, A.; Fujiwara, I.; Ito, M. Chem. Phys. Lettt. 1980, 74, 531. (56) Iyengar, S.; Peterson, M. K.; Day, T. J. F.; Burnham, C. J.; Teige, V. E.; Voth, G. A. J. Chem. Phys. 2005, 123, 084309.

625

dx.doi.org/10.1021/jp111586p |J. Phys. Chem. A 2011, 115, 620–625