Water Interface Revealed by

Apr 27, 2016 - Department of Chemistry, Graduate School of Science, Tohoku University, Sendai 980-8578, Japan. # Elements Strategy Initiative for Cata...
0 downloads 21 Views 1003KB Size
Download PDF
Letter pubs.acs.org/JPCL

Efficient Spectral Diffusion at the Air/Water Interface Revealed by Femtosecond Time-Resolved Heterodyne-Detected Vibrational Sum Frequency Generation Spectroscopy Ken-ichi Inoue,† Tatsuya Ishiyama,‡ Satoshi Nihonyanagi,†,§ Shoichi Yamaguchi,†,∥ Akihiro Morita,⊥,# and Tahei Tahara*,†,§ †

Molecular Spectroscopy Laboratory, RIKEN, 2-1 Hirosawa, Wako, Saitama 351-0198, Japan Department of Applied Chemistry, Graduate School of Science and Engineering, University of Toyama, Toyama 930-8555, Japan § Ultrafast Spectroscopy Research Team, RIKEN Center for Advanced Photonics (RAP), 2-1 Hirosawa, Wako, Saitama 351-0198, Japan ∥ Department of Applied Chemistry, Graduate School of Science and Engineering, Saitama University, 255 Shimo-Okubo, Sakura, Saitama 338-8570, Japan ⊥ Department of Chemistry, Graduate School of Science, Tohoku University, Sendai 980-8578, Japan # Elements Strategy Initiative for Catalysts and Batteries (ESICB), Kyoto University, Kyoto 615-8520, Japan ‡

S Supporting Information *

ABSTRACT: Femtosecond vibrational dynamics at the air/water interface is investigated by time-resolved heterodyne-detected vibrational sum frequency generation (TR-HD-VSFG) spectroscopy and molecular dynamics (MD) simulation. The low- and high-frequency sides of the hydrogen-bonded (HB) OH stretch band at the interface are selectively excited with special attention to the bandwidth and energy of the pump pulses. Narrow bleach is observed immediately after excitation of the high-frequency side of the HB OH band at ∼3500 cm−1, compared to the broad bleach observed with excitation of the low-frequency side at ∼3300 cm−1. However, the time-resolved spectra observed with the two different excitations become very similar at 0.5 ps and almost indistinguishable by 1.0 ps. This reveals that efficient spectral diffusion occurs regardless of the difference of the pump frequency. The experimental observations are well-reproduced by complementary MD simulation. There is no experimental and theoretical evidence that supports extraordinary slow dynamics in the high-frequency side of the HB OH band, which was reported before.

W

a recent combined theoretical and experimental study has suggested that even hydrogen bonds of the HB OH of the water molecule that has a free OH on the other side is only slightly weaker than the hydrogen bonds in the bulk.6 Although VSFG spectra indicate that the HB OH is similar to that in the bulk from the steady-state viewpoint, dynamical information obtainable with time-resolved measurements can provide much more insight, such as inhomogeneity, spectral diffusion, and vibrational relaxation at the water interface.7 Therefore, it is of essential importance to study the vibrational dynamics of the HB OH at the water interface to elucidate the property of interfacial water. The vibrational dynamics at the aqueous interfaces has been studied by time-resolved (TR-) VSFG, in which femtosecond infrared excitation (pump) and VSFG measurements (probe) are combined.8−10 Conventional TR-VSFG measures the pump-induced intensity change of the SFG light (ΔI), and it

ater is the most important liquid, and it has been extensively studied in a variety of contexts of science and technology. In particular, the water interface has attracted much attention because it provides unique environments that are important for atmospheric chemistry, electrochemistry, organic chemistry, and so on.1,2 At the interface, the threedimensional hydrogen-bonded (HB) network of water is truncated on the atomistic scale, and it is of great interest to elucidate how this truncation affects the property of interfacial water. Vibrational sum frequency generation (VSFG) spectroscopy has been playing crucial roles in investigating the structure of water interfaces. For instance, VSFG spectra at the air/water interface show a sharp OH stretch band at ∼3700 cm−1, which is attributed to a non-HB OH (free OH) that exists only at the interface.3 Observation of this free OH manifests the powerfulness of VSFG spectroscopy for studying the water interface. VSFG spectra at the air/water interface also show a broad HB OH stretch band peaked at ∼3450 cm−1. Its frequency and bandwidth (>200 cm−1)4,5 are very similar to those of the HB OH band in bulk, indicating that HB OH at the interface realizes an almost bulk-like hydrogen bond. In fact, © XXXX American Chemical Society

Received: March 29, 2016 Accepted: April 27, 2016

1811

DOI: 10.1021/acs.jpclett.6b00701 J. Phys. Chem. Lett. 2016, 7, 1811−1815

Letter

The Journal of Physical Chemistry Letters

neither long-lasting inhomogeneity nor extraordinarily slow dynamics was observed. The steady-state χ(2) spectrum of the air/water interface is shown in Figure 1. This χ(2) spectrum is essentially the same as

provides information on the change of the square of secondorder nonlinear susceptibility (χ(2)) (2) ΔI ∝ Δ |χ (2) |2 ≅ 2Re[χsteady * Δχ (2) ]

Here, the asterisk denotes the complex conjugate, and χ(2) steady and Δχ(2) are the steady-state signal of χ(2) and its pumpinduced change, respectively. Obviously, this quantity does not have a simple physical meaning,11 and it is difficult to interpret the data. Time-resolved VSFG using heterodyne detection12,13 (TR-HD-VSFG) can solve this problem7,14−16 because it directly provides the pump-induced change of Im χ(2) spectra (ΔIm χ(2)). Because ΔIm χ(2) is interpreted in the same manner as the time-resolved IR spectrum which corresponds to ΔIm χ(1) (χ(1): linear susceptibility), we can readily interpret the time-resolved spectra obtained with TR-HD-VSFG. Furthermore, TR-HD-VSFG has been extended to 2D spectroscopy (2D HD-VSFG) to obtain complete information on ultrafast vibrational dynamics of interfacial water.7,11,17,18 At present, TR-HD-VSFG and 2D HD-VSFG are the most powerful methods to investigate vibrational dynamics of water at interfaces. The first 2D HD-VSFG spectra of the OH stretch region of the air/water interface were reported by our group in 2013.17 In that study, we observed that the peak frequency of the bleach of the HB OH stretch showed pump frequency dependence, revealing inhomogeneity of the HB OH at the interface for the first time. It was also found that the bleach generated by the different excitations from 3200 to 3600 cm−1 became indistinguishable within a few hundred femtoseconds owing to efficient spectral diffusion at the water interface. Later, Bonn and co-workers also carried out 2D HD-VSFG experiments for the air/water interface using pump pulses having a narrower bandwidth (100 cm−1) under a stronger excitation condition.18 Thanks to the narrower pump bandwidth, they observed pump frequency dependence of the bleach more clearly immediately after photoexcitation. In addition, very surprisingly, they reported that spectral diffusion did not complete even at 1.5 ps when the water was excited at 3500 cm−1. They concluded that HB OH at around 3500 cm−1 was isolated from other HB OH and the interfacial water is very inhomogeneous. Furthermore, in another recent study, they reported that the vibrational relaxation time (T1) of the OH stretch level measured with 3500 cm−1 excitation was ∼0.75 ps, which was more than twice larger than the T1 time obtained with 3300 cm−1 excitation (∼0.35 ps).19 Because such a long-lasting inhomogeneity at the air/water interface was not observed in our first study, these time-resolved data look incompatible. Because the existence of the extraordinarily slow dynamics is a critical issue for understanding the property of the air/water interface, it is highly desirable to solve the discrepancy. In the present study, we carried out TR-HD-VSFG measurements for the air/water interface with selective excitation of high- and low-frequency sides of the HB OH band. We paid special attention to the bandwidth and pulse energy of the pump pulses, which are the possible reasons for the discrepancy. We performed experiments making the bandwidth of the infrared pump pulses as narrow as 120 cm−1, keeping its energy as low as 15 μJ. Under this condition, the narrower bleach of HB OH was observed with excitation of the 3500 cm−1 region of the HB OH band immediately after excitation. However, spectral diffusion takes place rapidly, and

Figure 1. (Black) Real and (red) imaginary parts of χ(2) spectra of the air/water interface.

those reported recently.20,21 As the sign of Im χ(2) gives information about the net orientation of the transition dipole moments,22 the sharp positive band at ∼3700 cm−1 is assigned to the “H-up” oriented free OH, whereas the broad negative band peaked at ∼3450 cm−1 is attributed to the “H-down” oriented HB OH which we discuss in this study. Figure 2a shows the time-resolved ΔIm χ(2) spectra measured with excitation of the low-frequency side (3300 cm−1) of the

Figure 2. Time-resolved ΔIm χ(2) spectra with (a) 3300 and (b) 3500 cm−1 excitations. The steady-state Im χ(2) spectrum and the pump spectrum are shown at the top of each figure.

HB OH stretch band. At −1.0 ps without pump influence, no transient change is observed, and the spectrum indicates the noise level of the present experiment. Immediately after excitation at 0.0 ps, a positive band centered at ∼3400 cm−1 appears, which is attributed to the bleach of the negative HB OH stretch (ν = 0 → 1 transition). Although this bleach is broader than the bandwidth of the pump pulse, it is narrower than the HB OH stretch band in the steady-state spectrum. This implies that a spectral hole is created, reflecting the inhomogeneity of the HB OH. This positive bleach signal is accompanied by a negative band in the lower-frequency region, which is assigned to the hot band (ν = 1 → 2) of the HB OH stretch. The frequency of the hot band (ν = 1 → 2) is lower than the bleach frequency (ν = 0 → 1) due to the 1812

DOI: 10.1021/acs.jpclett.6b00701 J. Phys. Chem. Lett. 2016, 7, 1811−1815

Letter

The Journal of Physical Chemistry Letters

ΔIm χ(2) spectra obtained with two excitations become very similar to each other at 0.5 ps, and they are almost indistinguishable at 1.0 ps. It is evident that, although the ΔIm χ(2) spectra immediately after excitation show substantial difference, efficient spectral diffusion takes place regardless of the difference of pump frequency and the noticeable difference does not remain at 1.0 ps. We also evaluated the vibrational relaxation (T1) time, following analysis in the previous study of the Bonn group.19 In this analysis, the ΔIm χ(2) signal at around 3500 cm−1 where the signal becomes zero at 1.5 ps is chosen to be plotted against the delay time. Because the ΔIm χ(2) signal of this region is considered free from the contribution of the thermalized spectrum, it is expected that this signal represents the intensity of the “pure” bleach signal and that the vibrational relaxation time can be extracted. As shown in Figure 3b, the ΔIm χ(2) signal decays with a time constant of 0.39 ± 0.04 ps for the 3300 cm−1 excitation and 0.35 ± 0.03 ps for 3500 cm−1 excitation. They are indistinguishable within the error. This is counterevidence of the marked difference in the vibrational relaxation time claimed in the previous study.19 The vibrational relaxation times of interfacial water obtained in the present study look slightly longer than the reported values for bulk water (0.26−0.35 ps),19,24,25 which are evaluated with excitation of the 3200−3500 cm−1 range. Nevertheless, the difference is small even if it exists, indicating that HB OH at the interface realizes almost a bulk-like hydrogen bond. In the present experiment, the extraordinarily slow dynamics with 3500 cm−1 excitation was not observed, even though we used pump pulses having a narrow bandwidth. Then, the possible reason for the discrepancy is the different pump energy between the present measurement (15 μJ) and the experiments of the Bonn group (40−100 μJ).18,19 To examine this possibility, we measured ΔIm χ(2) spectra with changing pump energy (Figure S1). It was found that the ΔIm χ(2) spectrum at 1.5 ps shows blue shifts as the pump energy is increased from 15 μJ to 35 μJ. More importantly, the ΔIm χ(2) spectrum obtained with 3500 cm−1 excitation is more blue shifted than that with 3300 cm−1 excitation under the stronger excitation condition, which makes the two spectra look noticeably different. It seems that this difference is the same as the spectral difference that provided a basis for claiming the long-lasting spectral inhomogeneity.18 This energy-dependent blue shift is readily explained by the difference in the thermalized spectra reflecting different local temperatures, which becomes noticeable under high-energy excitation conditions because the difference in the local temperature can be large under strong excitation conditions. Such a difference in the thermalized spectrum can affect the shape of the 2D spectrum as well as the vibrational relaxation time deduced from the decays of the ΔIm χ(2) signal at ∼3500 cm−1. To rationalize the experimental observation in the present study, we performed MD simulation (see the SI for details of the calculation). The basis of the argument of long-lasting inhomogeneity at the water interface is the existence of two types of HB OH at the air/water interface18 (Figure 4a); one is HB OH in doubly donating water molecules (2D OH), and the other is HB OH of water molecules that have free OH on the opposite side (1D-BOH). We calculated the velocity−velocity autocorrelation function (VVAF) of hydrogens of interfacial water and its temporal change. VVAF allows for straightforward decomposition into these types of OH vibrations and investigation of their spectral diffusion dynamics separately.

anharmonicity. The transient signal is also seen in the free OH region at 0.0 ps. The appearance of this transient is explained by the anharmonic coupling between HB OH and free OH.17,23 At 0.1−0.3 ps, the positive bleach band becomes broader due to the spectral diffusion. Then, at a delay time later than 0.5 ps, the transient spectrum exhibits a positive signal at 3500 cm−1, which reflects the frequency upshift of the HB OH band due to thermalization; the pump energy absorbed by the water finally converts to thermal energy, which increases the temperature of interfacial water. This weakens the hydrogen bond and causes a blue shift of the HB OH band.14 Figure 2b shows the time-resolved ΔIm χ(2) spectra measured with excitation of the high-frequency side (3500 cm−1) of the HB OH band. Basically, the ΔIm χ(2) spectrum obtained with 3500 cm−1 excitation exhibits temporal evolution similar to that observed with 3300 cm−1 excitation; the positive bleach and negative hot band appear at 0.0 ps, the bleach becomes broader due to spectral diffusion at 0.0−0.3 ps, and the spectral feature due to thermalization appears after 0.5 ps. However, the time-resolved spectrum in the early delay time region is significantly different from that observed with 3300 cm−1 excitation, particularly at 0.0 ps. The difference is more clearly recognized in Figure 3a, where the time-resolved spectra measured with 3300 and 3500 cm−1

Figure 3. (a) Comparison of the time-resolved ΔIm χ(2) spectra with 3300 (black) and 3500 cm−1 (red) excitations. The two spectra at each time delay are scaled at the maximum intensities. (b) Time evolution of the area intensity of the ΔIm χ(2) spectra at around the zero crossing frequency at 1.5 ps (near 3500 cm−1). Black and red markers represent the data points for 3300 and 3500 cm−1 excitations, respectively. The solid lines are the fitting results, corresponding to the singleexponential decays with time constants of 0.39 ± 0.04 ps for 3300 cm−1 excitation and 0.35 ± 0.03 ps for 3500 cm−1 excitation. The instrumental response is taken into account by a 0.2 ps Gaussian function in the fitting.

excitations are overlapped at each delay time after intensity normalization. At 0.0 ps, the bleach band obtained with 3300 cm−1 excitation exhibits a peak at 3410 cm−1 with the bandwidth of 200 cm−1 (fwhm). On the other hand, the bleach observed with 3500 cm−1 excitation shows the peak at 3470 cm−1 with a much narrower bandwidth of 120 cm−1. The bleach bands shift according to the change of the pump frequency, but their bandwidths are distinctly different from each other. This pump frequency dependence of the ΔIm χ(2) spectrum is observed very clearly in the present study thanks to the narrow pump bandwidth employed. However, Figure 3a also shows that this spectral difference vanishes quickly. The 1813

DOI: 10.1021/acs.jpclett.6b00701 J. Phys. Chem. Lett. 2016, 7, 1811−1815

Letter

The Journal of Physical Chemistry Letters

(Figure 4c). The time resolution of the present experiment is ∼200 fs, and hence, our experiments cannot resolve the dynamics that occurs within this time resolution. Therefore, it is possible that the broad bleach observed at 0.0 ps with 3300 cm−1 excitation arises, at least in part, from the faster spectral diffusion that occurs within the time resolution of the measurement. In addition to the different efficiency of the spectral diffusion, the different bandwidth of the bleaches observed at 0.0 ps may also arise from the frequency dependence of Fermi resonance with the bend overtone.11,28−30 Because the H2O bend overtone frequency is located at ∼3300 cm−1,31,32 the Fermi resonance is expected to be most significant for the OH stretch at around 3300 cm−1. Therefore, the bleach generated with excitation of this frequency region can be instantaneously broadened with Fermi resonance because the depletion of the ground state affects both of the Fermi split bands. On the other hand, the bleach with 3500 cm−1 excitation is less affected by the Fermi resonance because of its larger frequency mismatch with the bend overtone, resulting in the narrower bleach at 0.0 ps. We note that the calculated bandwidths with 3300 and 3500 cm−1 excitation at 0.0 ps shown in Figure 4c are similar to each other because our classical MD simulation does not completely treat the effect of the Fermi resonance.33 In summary, we carried out TR-HD-VSFG experiments and MD simulation of the air/water interface to clarify the difference in the vibrational dynamics of the low- and highfrequency sides of the HB OH band. At 0.0 ps, the narrow bleach band in the ΔIm χ(2) spectrum is observed with the high-frequency 3500 cm−1 excitation, compared to the lowfrequency 3300 cm−1 excitation. This difference is rationalized by the slow spectral diffusion as well as the absence of Fermi resonance at around 3500 cm−1. However, the ΔIm χ(2) spectra with 3300 and 3500 cm−1 excitations become very similar at 0.5 ps and almost indistinguishable by 1.0 ps due to efficient spectral diffusion. These experimental observations are wellreproduced by complementary MD simulation. The vibrational relaxation time deduced from the signal at around 3500 cm−1 is about 0.4 ps, which is not significantly different from that of bulk water. The present study shows that there is no evidence that supports extraordinarily slow dynamics at ∼3500 cm−1 of the air/water interface and indicates that the effect of the truncation of the hydrogen-bond network is screened very efficiently at the water interface.

Figure 4. (a) Schematic of the air/water interface. (Green) HB OH in doubly donating water molecules (2D OH). (Blue) HB OH of water molecules that have free OH on the opposite end (1D-BOH). (Black) Free OH. (b) Calculated VVAF of 2D OH (green) and 1D-BOH (blue) at the air/H2O interface. (c) Calculated transient changes of VVAF (ΔVVAF) with 3300 (black) and 3500 cm−1 (red) excitations.

Here we define the 2D OH and 1D-BOH by classifying each H as bonded H when the H···O distance is less than 2.5 Å or otherwise as nonbonded H. The interfacial region is defined as the region in 3 Å depth from the Gibbs dividing surface of water toward liquid. As shown in Figure 4b, the calculated VVAFs show that the bands of 2D OH (green) and 1D-BOH (blue) mostly overlap each other, although the peak of the 1DBOH band (∼3500 cm−1) is slightly higher than that of 2D OH (∼3450 cm−1).26 Therefore, neither the pump pulses at 3300 cm−1 nor those at 3500 cm−1 can selectively excite the 2D OH and 1D-BOH components. We confirmed that these VVAFs are almost independent of the details of the hydrogen bond definition (Figure S3). Then, we calculated the temporal change of VVAF (ΔVVAF) after exciting the 3300 and 3500 cm−1 vibrations of interfacial water (Figure 4c) by employing the transient kinetic energy method (see the SI for details).27 The ΔVVAF represents temporal evolution of the vibrational frequency, which is observed as spectral diffusion in the TR-HD-VSFG experiment. At 0.0 ps, each ΔVVAF exhibits a clearly distinct peak having a similarly narrow bandwidth. At 0.2 and 0.4 ps, substantial broadening of ΔVVAF is seen, and it is recognized that the broadening after 3300 cm−1 excitation proceeds much faster than that after 3500 cm−1 excitation. Nevertheless, the two ΔVVAFs almost overlap with each other at 0.6 ps, indicating that the spectral diffusion almost completes within a time as short as 0.6 ps regardless of the difference in the pump frequency. The calculated ΔVVAF reproduces the efficient spectral diffusion, and it rationalizes the experimental observation that ΔIm χ(2) spectra observed with 3300 and 3500 cm−1 excitations become very similar at 0.5 ps. We note that ΔVVAF calculated in the present study is also consistent with the simulated 2D HD-VSFG spectra reported recently.23 The calculated ΔVVAF also provides insight into the different bleach bandwidth observed immediately after excitation in the TR-HD-VSFG experiments. As clearly shown in Figure 2 (as well as Figure 3a), the bleach band observed with 3300 cm−1 excitation is much broader than that observed with 3500 cm−1 excitation. On the other hand, the bandwidths of ΔVVAF are the same at 0.0 ps, and that with 3300 cm−1 excitation gets rapidly broadened in 200 fs, whereas that with 3500 cm−1 excitation remains relatively narrow

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.jpclett.6b00701. Experimental details, pump energy dependence, computational procedures, and hydrogen-bond definition dependence of VVAF (PDF)

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was supported by a Grant-in-Aid for Scientific Research on Innovative Area (No. 25104005) from the 1814

DOI: 10.1021/acs.jpclett.6b00701 J. Phys. Chem. Lett. 2016, 7, 1811−1815

Letter

The Journal of Physical Chemistry Letters

Vibrational Relaxation in Bulk and Surface Water Reveals SubPicosecond Structural Heterogeneity. Nat. Commun. 2015, 6, 8384. (20) Nihonyanagi, S.; Kusaka, R.; Inoue, K.; Adhikari, A.; Yamaguchi, S.; Tahara, T. Accurate Determination of Complex χ(2) Spectrum of the Air/Water Interface. J. Chem. Phys. 2015, 143, 124707. (21) Yamaguchi, S. Development of Single-Channel HeterodyneDetected Sum Frequency Generation Spectroscopy and Its Application to the Water/Vapor Interface. J. Chem. Phys. 2015, 143, 034202. (22) Nihonyanagi, S.; Yamaguchi, S.; Tahara, T. Direct Evidence for Orientational Flip-Flop of Water Molecules at Charged Interfaces: A Heterodyne-Detected Vibrational Sum Frequency Generation Study. J. Chem. Phys. 2009, 130, 204704. (23) Ishiyama, T.; Morita, A.; Tahara, T. Molecular Dynamics Study of Two-Dimensional Sum Frequency Generation Spectra at Vapor/ Water Interface. J. Chem. Phys. 2015, 142, 212407. (24) Lock, A. J.; Bakker, H. J. Temperature Dependence of Vibrational Relaxation in Liquid H2O. J. Chem. Phys. 2002, 117, 1708−1713. (25) Ramasesha, K.; De Marco, L.; Mandal, A.; Tokmakoff, A. Water Vibrations Have Strongly Mixed Intra- and Intermolecular Character. Nat. Chem. 2013, 5, 935−940. (26) Gan, W.; Wu, D.; Zhang, Z.; Feng, R.-r.; Wang, H.-f. Polarization and Experimental Configuration Analyses of Sum Frequency Generation Vibrational Spectra, Structure, and Orientational Motion of the Air/Water Interface. J. Chem. Phys. 2006, 124, 114705. (27) Yagasaki, T.; Saito, S. A Novel Method for Analyzing Energy Relaxation in Condensed Phases Using Nonequilibrium Molecular Dynamics Simulations: Application to the Energy Relaxation of Intermolecular Motions in Liquid Water. J. Chem. Phys. 2011, 134, 184503. (28) Sovago, M.; Campen, R. K.; Wurpel, G. W. H.; Müller, M.; Bakker, H. J.; Bonn, M. Vibrational Response of Hydrogen-Bonded Interfacial Water Is Dominated by Intramolecular Coupling. Phys. Rev. Lett. 2008, 100, 173901. (29) Nihonyanagi, S.; Yamaguchi, S.; Tahara, T. Water Hydrogen Bond Structure near Highly Charged Interfaces Is Not Like Ice. J. Am. Chem. Soc. 2010, 132, 6867−6869. (30) Singh, P. C.; Nihonyanagi, S.; Yamaguchi, S.; Tahara, T. Interfacial Water in the Vicinity of a Positively Charged Interface Studied by Steady-State and Time-Resolved Heterodyne-Detected Vibrational Sum Frequency Generation. J. Chem. Phys. 2014, 141, 18C527. (31) Max, J.-J.; Chapados, C. Isotope Effects in Liquid Water by Infrared Spectroscopy. III. H2O and D2O Spectra from 6000 to 0 cm−1. J. Chem. Phys. 2009, 131, 184505. (32) Freda, M.; Piluso, A.; Santucci, A.; Sassi, P. Transmittance Fourier Transform Infrared Spectra of Liquid Water in the Whole Mid-Infrared Region: Temperature Dependence and Structural Analysis. Appl. Spectrosc. 2005, 59, 1155−1159. (33) Nihonyanagi, S.; Ishiyama, T.; Lee, T.-k.; Yamaguchi, S.; Bonn, M.; Morita, A.; Tahara, T. Unified Molecular View of the Air/Water Interface Based on Experimental and Theoretical χ(2) Spectra of an Isotopically Diluted Water Surface. J. Am. Chem. Soc. 2011, 133, 16875−16880.

Ministry of Education, Culture Sports, Science and Technology (MEXT). K.I. acknowledges the Special Postdoctoral Researchers (SPDR) program of RIKEN.

■

REFERENCES

(1) Ravishankara, A. R. Heterogeneous and Multiphase Chemistry in the Troposphere. Science 1997, 276, 1058−1065. (2) Jung, Y.; Marcus, R. A. On the Theory of Organic Catalysis “on Water. J. Am. Chem. Soc. 2007, 129, 5492−5502. (3) Du, Q.; Superfine, R.; Freysz, E.; Shen, Y. R. Vibrational Spectroscopy of Water at the Vapor/Water Interface. Phys. Rev. Lett. 1993, 70, 2313−2316. (4) Gopalakrishnan, S.; Liu, D.; Allen, H. C.; Kuo, M.; Shultz, M. J. Vibrational Spectroscopic Studies of Aqueous Interfaces: Salts, Acids, Bases, and Nanodrops. Chem. Rev. 2006, 106, 1155−1175. (5) Feng, R.-r.; Guo, Y.; Lü, R.; Velarde, L.; Wang, H.-f. Consistency in the Sum Frequency Generation Intensity and Phase Vibrational Spectra of the Air/Neat Water Interface. J. Phys. Chem. A 2011, 115, 6015−6027. (6) Stiopkin, I. V.; Weeraman, C.; Pieniazek, P. A.; Shalhout, F. Y.; Skinner, J. L.; Benderskii, A. V. Hydrogen Bonding at the Water Surface Revealed by Isotopic Dilution Spectroscopy. Nature 2011, 474, 192−195. (7) Singh, P. C.; Nihonyanagi, S.; Yamaguchi, S.; Tahara, T. Ultrafast Vibrational Dynamics of Water at a Charged Interface Revealed by Two-Dimensional Heterodyne-Detected Vibrational Sum Frequency Generation. J. Chem. Phys. 2012, 137, 094706. (8) McGuire, J. A.; Shen, Y. R. Ultrafast Vibrational Dynamics at Water Interfaces. Science 2006, 313, 1945−1948. (9) Smits, M.; Ghosh, A.; Sterrer, M.; Müller, M.; Bonn, M. Ultrafast Vibrational Energy Transfer between Surface and Bulk Water at the Air-Water Interface. Phys. Rev. Lett. 2007, 98, 098302. (10) Eftekhari-Bafrooei, A.; Borguet, E. Effect of Surface Charge on the Vibrational Dynamics of Interfacial Water. J. Am. Chem. Soc. 2009, 131, 12034−12035. (11) Inoue, K.; Nihonyanagi, S.; Singh, P. C.; Yamaguchi, S.; Tahara, T. 2D Heterodyne-Detected Sum Frequency Generation Study on the Ultrafast Vibrational Dynamics of H2O and HOD Water at Charged Interfaces. J. Chem. Phys. 2015, 142, 212431. (12) Shen, Y. R. Phase-Sensitive Sum-Frequency Spectroscopy. Annu. Rev. Phys. Chem. 2013, 64, 129−150. (13) Nihonyanagi, S.; Mondal, J. A.; Yamaguchi, S.; Tahara, T. Structure and Dynamics of Interfacial Water Studied by HeterodyneDetected Vibrational Sum-Frequency Generation. Annu. Rev. Phys. Chem. 2013, 64, 579−603. (14) Nihonyanagi, S.; Singh, P. C.; Yamaguchi, S.; Tahara, T. Ultrafast Vibrational Dynamics of a Charged Aqueous Interface by Femtosecond Time-Resolved Heterodyne-Detected Vibrational Sum Frequency Generation. Bull. Chem. Soc. Jpn. 2012, 85, 758−760. (15) Watanabe, K.; Inoue, K.; Nakai, I. F.; Matsumoto, Y. Nonadiabatic Coupling between C-O Stretching and Pt Substrate Electrons Enhanced by Frustrated Mode Excitations. Phys. Rev. B: Condens. Matter Mater. Phys. 2010, 81, 241408. (16) Xiong, W.; Laaser, J. E.; Mehlenbacher, R. D.; Zanni, M. T. Adding a Dimension to the Infrared Spectra of Interfaces Using Heterodyne Detected 2D Sum-Frequency Generation (HD 2D SFG) Spectroscopy. Proc. Natl. Acad. Sci. U. S. A. 2011, 108, 20902−20907. (17) Singh, P. C.; Nihonyanagi, S.; Yamaguchi, S.; Tahara, T. Communication: Ultrafast Vibrational Dynamics of Hydrogen Bond Network Terminated at the Air/Water Interface: A Two-Dimensional Heterodyne-Detected Vibrational Sum Frequency Generation Study. J. Chem. Phys. 2013, 139, 161101. (18) Hsieh, C.-S.; Okuno, M.; Hunger, J.; Backus, E. H. G.; Nagata, Y.; Bonn, M. Aqueous Heterogeneity at the Air/Water Interface Revealed by 2D-HD-SFG Spectroscopy. Angew. Chem., Int. Ed. 2014, 53, 8146−8149. (19) van der Post, S. T.; Hsieh, C. S.; Okuno, M.; Nagata, Y.; Bakker, H. J.; Bonn, M.; Hunger, J. Strong Frequency Dependence of 1815

DOI: 10.1021/acs.jpclett.6b00701 J. Phys. Chem. Lett. 2016, 7, 1811−1815