Raman spectroscopy of optically levitated supercooled water droplet

Hidenori Suzuki, Yoshiki Matsuzaki, Azusa Muraoka, and Maki Tachikawa Citation: J. Chem. Phys. 136, 234508 (2012); doi: 10.1063/1.4729476 View online: http://dx.doi.org/10.1063/1.4729476 View Table of Contents: http://jcp.aip.org/resource/1/JCPSA6/v136/i23 Published by the American Institute of Physics.

Additional information on J. Chem. Phys.

Journal Homepage: http://jcp.aip.org/ Journal Information: http://jcp.aip.org/about/about_the_journal Top downloads: http://jcp.aip.org/features/most_downloaded Information for Authors: http://jcp.aip.org/authors

Downloaded 09 Sep 2012 to 193.48.219.8. Redistribution subject to AIP license or copyright; see http://jcp.aip.org/about/rights_and_permissions

THE JOURNAL OF CHEMICAL PHYSICS 136, 234508 (2012)

Raman spectroscopy of optically levitated supercooled water droplet

Hidenori Suzuki, Yoshiki Matsuzaki, Azusa Muraoka, and Maki Tachikawaa)
Department of Physics, Meiji University, 1-1-1 Higashimita, Tama-ku, Kawasaki, Kanagawa 214-8571, Japan

(Received 5 April 2012; accepted 31 May 2012; published online 21 June 2012) By use of an optical trap, we can levitate micrometer-sized drops of puried water and cool them below the melting point free from contact freezing. Raman spectra of the OH stretching band were obtained from those supercooled water droplets at temperatures down to 35 C. According to the two-state model, an enthalpy change due to hydrogen-bond breaking is derived from temperature dependence of the spectral prole. The isobaric heat capacity calculated from the enthalpy data shows a sharp increase as the temperature is lowered below 20 C in good agreement with conventional thermodynamic measurements. 2012 American Institute of Physics. [http://dx.doi.org/10.1063/1.4729476]
I. INTRODUCTION

Although water is one of the most common materials on Earth, it is distinguished from other simple liquids for unique physical properties such as a negative thermal expansion below 4 C and a thermodynamic singularity around 45 C.16 It is generally accepted that strong hydrogen bonding between water molecules is responsible for these characteristics. However, their detailed mechanism and microscopic picture are still controversial. There is a hypothesis that predicts the existence of a critical point associated with two phases of liquid water; one with a low density and the other with a high density.7 This idea of water polyamorphism was inspired by the discovery of low- and high-density amorphous ice and the rst-order phase transition between the two solid phases.810 According to the hypothesis, those solid phases are connected to corresponding liquid phases at higher temperatures, and the observed anomalies of liquid water are interpreted as a manifestation of the critical phenomenon near the liquidliquid critical point, so-called the second critical point. Thermodynamic analysis and numerical simulations1116 predict that the second critical point lies deep in the supercooled region well below the homogeneous nucleation temperature TH ,6, 17 which is about 38 C at 1 bar. Even if the critical point exists, this makes its direct observation extremely difcult. There are other thermodynamic models called the singularity-free scenario,18, 19 the stability-limit scenario,20 and the critical-point free scenario.21 Each of the three models explains anomalies of liquid water without assuming the second critical point. According to Stokely et al.,22 difference of the four scenarios comes from the estimation of the strength and cooperativity of hydrogen bonds. Careful comparison with experimental data of supercooled water is important in testing these models. Since interesting anomalies of liquid water become pronounced at temperatures far below the melting point, special care should be taken in experiments to avoid freezing of the
a) Author to whom correspondence should be addressed. Electronic mail:

tachikaw@isc.meiji.ac.jp. 0021-9606/2012/136(23)/234508/6/$30.00

aqueous sample. Contact freezing on a solid surface is inevitable for bulk water contained in a vessel, and the liquid water temperature normally accessible is higher than 25 C. Isolation of water samples in ne glass capillaries23 or emulsion droplets24 effectively reduces the nucleation probability and enabled the measurements of heat capacity,1, 2, 5 isothermal compressibility,3, 5, 6 and the coefcient of thermal expansion46 at temperatures down to 38 C. Those quantities tend to diverge as the temperature approaches 45 C, which is the temperature of the expected Widom line25 at 1 bar. They well t to a scaled parametric equation of state derived on the assumption that the liquidliquid critical point does exists.26, 27 Water conned in a nanometer-sized pore is free from freezing even at temperatures below 200 K.28 The calorimetric measurements on the nanopore water revealed that the heat capacity had a maximum around 233 K in favor of the second critical point scenario.29 It should be noted, however, that nanometer-scale connement generally changes thermodynamic properties of the sample. As a method of isolating supercooled bulk water from solid boundary, particle levitation is a promising option.30 Compared to other levitation techniques such as electrodynamic balance31 and acoustic levitation,32 optical trapping is versatile in manipulation of the position and motion of the trapped particles.33, 34 Motivated by atmospheric science, optical levitation has been employed for thermodynamic measurements and observation of the freezing process of low-temperature aqueous solutions.3537 Trapping of a micrometer-sized ice crystal and imaging of crystal growth to a snow ake were also demonstrated.37, 38 Optical trapping enables spectroscopic investigation of water medium even in highly supercooled state. Raman spectroscopy of liquid water provides information on intermolecular motions in the hydrogen-bonded network as well as the perturbation to the normal vibrational modes of a single water molecule.3956 So far, experiments using the Raman tweezers spectroscopy mostly aimed to probe physical and chemical processes in aqueous aerosol droplets containing salt or acid as a solvent.35, 37, 5759 On the other hand, Raman spectroscopy has been applied to detect density uctuations near the liquid vapor critical point in several materials.60, 61
2012 American Institute of Physics

136, 234508-1

Downloaded 09 Sep 2012 to 193.48.219.8. Redistribution subject to AIP license or copyright; see http://jcp.aip.org/about/rights_and_permissions

234508-2

Suzuki et al.

J. Chem. Phys. 136, 234508 (2012)

In this paper, we analyze Raman spectra from a puried water droplet levitated in air by a dual-beam optical trap.33 It is our major objective to clarify how thermodynamic anomalies appear in spectral proles of molecular vibrations, and to establish a methodology of spectroscopic investigation of highly supercooled bulk water. We focus on the OH stretching band in the frequency range of 27003900 cm1 . Raman spectrum observed between 90 C and 35 C is decomposed into Gaussian components and their origin and temperature dependence are discussed. Based on the so-called two-state model,62, 63 an enthalpy change due to hydrogen-bond breaking is obtained from the temperature dependence of the spectral proles. The isobaric heat capacity calculated from the enthalpy data is compared with results of the conventional measurements.2
II. EXPERIMENTAL

A schematic of our experimental setup is shown in Fig. 1(a). A detailed description of the trapping optics and the cooling system is given in Ref. 38. In brief, the optical trap33 uses counter-propagating Gaussian beams from a 532 nm Nd:YVO4 laser (Coherent, Verdi-V10). The two

beams of equal power are focused by a lens pair in such a way that their beam waists are positioned in front of each other. The trap connes a dielectric particle three-dimensionally in the center of the two focal points. An axial displacement from the equilibrium position causes an imbalance of the radiation pressure forces from the opposing lights, which produces a net restoring force along the optical axis. In radial directions, the optical gradient force works as a restoring force. The dual-beam trap is installed in an open cold chamber assembled with a copper cylinder and a coolant jacket. Temperature distribution inside the chamber is controlled via the coolant temperature which is adjustable from 25 C to 50 C with a refrigerated circulator (Julabo, F83). The ambient temperature around the trap region is measured by a thermocouple that is positioned 10 mm aside from the trap center not to intercept the laser beams. Temperature difference across the offset is less than 0.1 K. Ultrapure water with a resistivity of 18.2 M cm is turned into a ne mist and guided to the cold chamber by an ultrasonic nebulizer (Omron, NE-U07). Water droplets that have entered the cold chamber are rapidly thermalized to the ambient temperature and become supercooled. The optical trap captures a water droplet that has drifted into the laser beams, as is shown in Fig. 1(b). The size of the droplet can be accurately determined from the interference fringes due to Mie scattering of the trapping beams. The diameter of the water droplets ranges from a few micrometer to several tens micrometer. The trapped water droplet slightly absorbs the 532 nm trapping radiation. The resultant temperature rise estimated from the absorption coefcient is less than 0.1 K. Stokes and anti-Stokes photons are scattered from water molecules excited by the 532 nm radiation. Raman spectra are observed in the X(Y, X + Y)Z conguration. The pump radiation is horizontally polarized, and vertically scattered light is focused on the input end of an optical ber feed for the CCD spectrograph (Acton Research, SpectraPro 2300i) after passing a depolarizer. The spectral sensitivity of the detection system was calibrated with a NIST-calibrated tungsten halogen lamp (Ocean Optics, LS-1). Wavelength calibration of the spectrograph was performed using Ne lines as references. Figure 2 compares Raman spectrum of the OH stretching band observed at different temperatures from 90.0 C to 34.6 C. For temperatures above 5 C, spectra from an ultrapure water sample contained in a Pyrex glass cell were

FIG. 1. (a) Schematic of the experimental apparatus. (b) Image of a single water droplet in the dual beam optical trap and its Mie scattering fringes.

FIG. 2. Raman spectra of the OH stretching band of water at different temperatures.

Downloaded 09 Sep 2012 to 193.48.219.8. Redistribution subject to AIP license or copyright; see http://jcp.aip.org/about/rights_and_permissions

234508-3

Suzuki et al.

J. Chem. Phys. 136, 234508 (2012)

FIG. 3. Decomposition of the Raman spectrum, observed at T = 23.3 C and 27.7 C, into Gaussian components. Sum of the Gaussian components is represented by a red solid line.

FIG. 4. Temperature dependence of the integrated intensity of each Gaussian component. I1 , I2 , I3 , I4 , and I5 represent intensities of the 3050 cm1 , 3200 cm1 , 3400 cm1 , 3500 cm1 , and 3650 cm1 components, respectively. Solid lines are just a guide to the eye.

recorded. The intensity of the Raman signal depends on the size of the water droplet being trapped. For the sake of comparison, a spectral prole was normalized so that its integrated intensity becomes equal. A typical Raman spectrum of this band has two major peaks. As the temperature is lowered, the low-frequency peak around 3200 cm1 becomes more prominent than the high-frequency peak around 3400 cm1 , and the whole spectrum is shifted to lower frequencies. It is also noticeable that a shoulder observed around 3600 cm1 at room temperatures disappears at temperatures below 15 C. Following the conventional analysis of the OH stretching band of H2 O,42, 47, 56 the spectrum is decomposed into ve Gaussian components located around 3050, 3200, 3400, 3500, and 3650 cm1 , as shown in Fig. 3. The linear combination of these components is tted to the observed spectral proles by adjusting the center frequency, the width, and the height of the Gauss functions. The ratio of each components integrated intensity to the total intensity is plotted as a function of the temperature in Fig. 4. As the temperature rises, relative integrated intensities of the 3200 cm1 and 3400 cm1 components (I2 and I3 ) decrease and that of the 3500 cm1 component (I4 ) increases. The intensities of the 3050 cm1 and 3650 cm1 components (I1 and I5 ) are little dependent on the temperature. According to the assignment in literature,42, 56 the 3200 cm1 and 3400 cm1 components are due to the in-phase and out-of-phase stretching modes of a fully hydrogen-bonded water molecule in tetrahedral grouping of ve molecules. Such a molecule forms four hydrogen bonds, two with protons and two with lone electron pairs. The origin of the 3050 cm1 component is thought to be the Fermi resonance between the rst overtone of the bending vibration and the in-phase stretching mode. Breakage of the proton-donating bonds yields a water molecule with a

freely vibrating OH group. The 3500 cm1 and 3650 cm1 components are due, respectively, to vibrations of the hydrogen-bonded OH group and the dangling OH group in such a molecule. We have conrmed the assignments for the collective vibrations and the vibration of the dangling bond thorough ab initio MO calculation of vibrational modes for a tetrahedral water pentamer which is a component part of ice Ih structure.56 Geometry optimization and Raman shift analysis were made by the Hartree-Fock approach using 6-31+G* basis set in GAUSSIAN 09 program package.64

III. ANALYSIS AND DISCUSSION

To describe the structure and thermodynamic properties of liquid water, the so-called two-state model has been proposed,62, 63 which assumes water is a mixture of two different states of intermolecular bonding. One is an icelike state where water molecules are coupled by hydrogen bonding to form tetrahedral network. The other is a state where water molecules are more densely packed and hydrogen bonds are disrupted by distortion. The assumption of two distinct domains in liquid is sometimes criticized from the viewpoint that presumes continuous distribution of molecular coordinates.39, 65 Meanwhile, recent data of x-ray spectroscopy66, 67 support the multi-state model, and a new theoretical approach has been presented based on the twostate model, successfully describing the emergence of the second critical point.68, 69 Here, we stand on the concept of the two-state model and divide water molecules into two categories; a molecule with two OH groups both hydrogen-bonded to the tetrahedral network (hydrogen-bonded molecules), and a molecule with one or two dangling OH groups whose hydrogen bond to nearest neighbors is broken or weakened by severe distortion

Downloaded 09 Sep 2012 to 193.48.219.8. Redistribution subject to AIP license or copyright; see http://jcp.aip.org/about/rights_and_permissions

234508-4

Suzuki et al.

J. Chem. Phys. 136, 234508 (2012)

of the network (non-hydrogen-bonded molecules). Based on the assignments, the former contributes to the Raman spectral components I1 , I2 , and I3 , and the latter to the components I4 and I5 . This is consistent with the observed temperature dependence of the integrated intensity since thermal motion of molecules disrupts more hydrogen bonds with an increase in temperature. +I2 +I The ratio I1I4 +I5 3 is proportional to the equilibrium constant K, K= I1 + I2 + I3 fH B =A , fNH B I4 + I5 (1)

droplet is size-dependent and higher than the external pressure because of the surface tension. The internal pressure of a 5 m droplet, for example, is about 1.6 bar. However, the enthalpy of liquid water is little dependent on the pressure. For the internal pressure range of the trapped droplets, it is reasonably assumed that the enthalpy change derived from the droplet data is equal to that of bulk water under the standard pressure.5 In the two-state model, the standard heat capacity of liquid water is given by,43
H Cp = Cp B + (fH B 1)

where fHB is the fraction of hydrogen-bonded H2 O molecules, fNHB = 1 fHB the fraction of non-hydrogen-bonded molecules, and A is a constant independent of the temperature. According to the vant Hoff equation, ln K is related to the temperature T as,43, 47, 56 d ln K H . = d T 1 R (2)

H fH B (1 fH B ) + T R

H T

, (3)

Here R is the gas constant, and H denotes the standard enthalpy change of formation when a hydrogen-bonded molecule is produced from a non-bonded molecule at 1 bar of pressure. +I2 +I In Fig. 5, ln I1I4 +I5 3 is plotted as a function of the inverse of the temperature. Errors on the vertical scale were estimated by considering tolerated shifts of the tted parameters. The error bars indicate the range of ln K in which residuals between the observed spectral prole and the tted function always fall within the noise level of the Raman signal. The standard enthalpy change H , obtained from the slope of the curve, is almost constant over the temperature range from 5 C90 C, but gradually increases in magnitude as the temperature is lowered below 0 C. A test function g(x) = ax + b + cexp (x) is tted to the experimental data, and the best-tted one is shown by the solid curve in Fig. 5. This function was chosen since it well describes both the linear part below x = T1 = 0.0036 K1 and the monotonically increasing convex part above 0.0037 K1 . The value of H is calculated to be 2.0 kcal/mol at temperatures between 5 C90 C, in good agreement with previous measurements.42, 43, 47, 51 Droplet samples differ in size, and due to evaporation, even a single water drop gradually shrinks during the observation. It should be noted that the pressure inside a water

H where Cp B is the standard heat capacity of hydrogen-bonded water molecules. The isobaric heat capacity Cp can be derived as a function of the temperature using experimentally determined fHB and H . Absolute values for fHB can be obtained from g(x) and Eq. (1) if the value of A is known. The enthalpy H change H can be replaced by R dg(x) . Two quantities Cp B dx and A are treated as adjustable parameters. Figure 6 compares the heat capacity Cp thus derived and data of the thermodynamic measurements so far reported.2 Data at temperatures below 3 C are those of emulsion H droplets. For the curve in the gure, Cp B is xed to 8R, and A is set to 0.59. The value of A reasonably agrees to the Raman spectral intensity ratio obtained from our ab initio MO calculation. The empirical curve based on our Raman spectroscopic measurements reproduces overall temperature-dependence of the heat capacity of liquid water as well as its absolute values. The heat capacity is almost constant above 0 C but exhibits a sharp increase as the temperature is lowered below 20 C. The curve predicts that the heat capacity of liquid water takes a maximum value around T = 210 K. This unimodal feature is intrinsic to the two-state system, but the curvature of the emulsion data cannot be tted satisfactorily without introducing the experimentally determined temperature dependence of H . According to MD simulations,53, 54 the molecular coordinates in liquid water are continuously distributed, and the line shape of the OH stretching band is attributed to inhomogeneous broadening due to local electric eld variations.

FIG. 5. vant Hoff plot derived from the integrated Raman intensities.

FIG. 6. Temperature dependence of the isobaric heat capacity of liquid water calculated from the experimentally determined parameters according to Eq. (3). Data from emulsion measurements2 are shown as crosses.

Downloaded 09 Sep 2012 to 193.48.219.8. Redistribution subject to AIP license or copyright; see http://jcp.aip.org/about/rights_and_permissions

234508-5

Suzuki et al.
6 C.

J. Chem. Phys. 136, 234508 (2012) A. Angell, Annu. Rev. Phys. Chem. 34, 593 (1983), and references therein. 7 P. H. Poole, F. Sciortino, U. Essmann, and H. E. Stanley, Nature (London) 360, 324 (1992). 8 O. Mishima, L. D. Calvert, and E. Whalley, Nature (London) 310, 393 (1984). 9 O. Mishima, L. D. Calvert, and E. Whalley, Nature (London) 314, 76 (1985). 10 O. Mishima, J. Chem. Phys. 100, 5910 (1994). 11 P. H. Poole, U. Essmann, F. Sciortino, and H. E. Stanley, Phys. Rev. E 48, 4605 (1993). 12 H. E. Stanley, C. A. Angell, U. Essmann, M. Hemmati, P. H. Poole, and F. Sciortino, Physica A 205, 122 (1994). 13 O. Mishima and H. E. Stanley, Nature (London) 392, 164 (1998). 14 O. Mishima and H. E. Stanley, Nature (London) 396, 329 (1998). 15 R. C. Dougherty, Chem. Phys. 298, 307 (2004). 16 H. Kanno and K. Miyata, Chem. Phys. Lett. 422, 507 (2006). 17 H. Kanno, R. J. Speedy, and C. A. Angell, Science 189, 880 (1975). 18 S. Sastry, P. G. Debenedetti, F. Sciortino, and H. E. Stanley, Phys. Rev. E 53, 6144 (1996). 19 L. P. N. Rebelo, P. G. Debenedetti, and S. Sastry, J. Chem. Phys. 109, 626 (1998). 20 R. J. Speedy, J. Phys. Chem. 86, 3002 (1982). 21 C. A. Angell, Science 319, 582 (2008). 22 K. Stokely, M. G. Mazza, H. E. Stanley, and G. Franzese, Proc. Natl. Acad. Sci. U.S.A. 107, 1301 (2010). 23 S. C. Mossop, Proc. Phys. Soc. London, Sect. B 68, 193 (1955). 24 D. H. Rasmussen and A. P. MacKenzie, J. Chem. Phys. 59, 5003 (1973). 25 J. L. F. Abascal and C. Vega, J. Chem. Phys. 133, 234502 (2010). 26 D. A. Fuentevilla and M. A. Anisimov, Phys. Rev. Lett. 97, 195702 (2006); 98, 149904 (2007). 27 C. E. Bertrand and M. A. Anisimov, J. Phys. Chem. B 115, 14099 (2011). 28 F. Mallamace, C. Corsaro, M. Broccio, C. Branca, N. Gobzlez-Segredo, J. Spooren, S.-H. Chen, and H. E. Stanley, Proc. Natl. Acad. Sci. U.S.A. 105, 12725 (2008). 29 A. Nagoe, Y. Kanke, M. Oguni, and S. Namba, J. Phys. Chem. B 114, 13940 (2010). 30 E. H. Brandt, Science 243, 349 (1989). 31 A. Pluchino, J. Opt. Soc. Am. A 4, 614 (1987). 32 Y.-J. L, W.-J. Xie, and B.-B. Wei, Chin. Phys. Lett. 19, 1543 (2002). 33 A. Ashkin, Phys. Rev. Lett. 24, 156 (1970). 34 A. Ashkin, J. M. Dziedzic, J. E. Bjorkholm, and S. Chu, Opt. Lett. 11, 288 (1986). 35 C. Mund and R. Zellner, J. Mol. Struct. 661, 491 (2003). 36 K. Anders, N. Roth, and A. Frohn, J. Geophys. Res. 101, 19223, doi:10.1029/95JD03227 (1996). 37 S. Ishizaka, T. Wada, and N. Kitamura, Chem. Phys. Lett. 506, 117 (2011). 38 K. Taji, M. Tachikawa, and K. Nagashima, Appl. Phys. Lett. 88, 141111 (2006). 39 T. T. Wall and D. F. Hornig, J. Chem. Phys. 43, 2079 (1965). 40 G. E. Walrafen, J. Chem. Phys. 47, 114 (1967). 41 G. DArrigo, G. Maisano, F. Mallamace, P. Migliardo, and F. Wanderlingh, J. Chem. Phys. 75, 4264 (1981). 42 W. B. Monosmith and G. E. Walrafen, J. Chem. Phys. 81, 669 (1984). 43 G. E. Walrafen, M. R. Fisher, M. S. Hokmabadi, and W.-H. Yang, J. Chem. Phys. 85, 6970 (1986). 44 J. L. Green, A. R. Lacey, and M. G. Sceats, J. Phys. Chem. 90, 3958 (1986). 45 G. E. Walrafen, J. Phys. Chem. 94, 2237 (1990). 46 D. E. Hare and C. M. Sorensen, J. Chem. Phys. 96, 13 (1992). 47 D. M. Carey and G. M. Korenowski, J. Chem. Phys. 108, 2669 (1998). 48 C. P. Lawrence and J. L. Skinner, J. Chem. Phys. 117, 8847 (2002); 118, 264 (2003). 49 S. A. Corcelli and J. L. Skinner, J. Phys. Chem. A 109, 6154 (2005). 50 P. L. Geissler, J. Am. Chem. Phys. 127, 14930 (2005). 51 J. D. Smith, C. D. Cappa, K. R. Wilson, R. C. Cohen, P. L Geissler, and R. J. Saykally, Proc. Natl. Acad. Sci. U.S.A. 102, 14171 (2005). 52 R. Kumar, J. R. Schmidt, and J. L. Skinner, J. Chem. Phys. 126, 204107 (2007). 53 B. Auer, R. Kumar, J. R. Schmidt, and J. L. Skinner, Proc. Natl. Acad. Sci. U.S.A. 104, 14215 (2007). 54 B. M. Auer and J. L. Skinner, J. Chem. Phys. 128, 224511 (2008). 55 Y. Sekine and T. Ikeda-Fukazawa, J. Chem. Phys. 130, 034501 (2009). 56 G. E. Walrafen, in Water: A Comprehensive Treatise, edited by F. Franks (Plenum, New York, 1972), Vol. 1, Chap. 5.

Raman spectral analysis by Smith et al.51 showed that the rise of temperature gradually distorts intermolecular geometries and releases more OH oscillators from restriction of hydrogen bonding. Though still inconclusive, the continuum picture is supported by recent numerical analyses and seems more natural as a real water structure. In describing thermodynamic properties, however, a continuous system may reasonably be modeled by combination of multiple states with continuously variable fractions. Our approach provides a simple method of decomposing the Raman spectrum into two parts with opposite temperature dependence to t into the two-state model. Raman spectra at much lower temperatures down to 50 C are necessary to test for the existence of the predicted second critical point. The accessible temperature in the present spectroscopic measurements was limited by freezing of water droplets, but it could be further lowered by the following strategy. Since homogeneous nucleation of ice is a stochastic process and the probability of freezing is volumeand surface-dependent,70 the lifetime of supercooled state is largely prolonged for smaller samples. Water of micrometer or submicrometer size is still large enough to have thermodynamic properties of a bulk sample. Selective trapping of a submicrometer droplet can be achieved by rst trapping a larger droplet and heating it with an infrared radiation until the droplet shrinks to an appropriate size through evaporation. When the heating radiation is turned off, the drop instantaneously thermalizes to the ambient temperature, and will remain liquid for a certain amount of time.

IV. CONCLUSION

We have demonstrated that combination of optical trapping and Raman spectroscopy is an effective tool to probe thermodynamic anomalies of liquid water in highly supercooled state. Detailed analysis of the OH stretching Raman components, based on the two-state model, yielded temperature dependences of the standard enthalpy change and the standard heat capacity. Anomalous increase of the heat capacity was observed at temperatures below 20 C. Although accessible temperature in the present experiment was limited to 35 C, rapid cooling and continuous in situ observation of a submicrometer water drop may push the limit to much lower temperatures. This versatility of manipulating a single droplet is a major advantage of the optical tweezers spectroscopy, which may be a key to the no-mans land.14

ACKNOWLEDGMENTS

The authors thank Dr. Tomoko Fukazawa for fruitful discussions on Raman spectroscopy. The computations were performed using Research Center for Computational Science, Okazaki, Japan. This work was supported by JSPS Grant-inAid for Scientic Research (Grant No. 19340112).
1 C. A. Angell, J. Shuppert, and J. C. Tucker, J. Phys. Chem. 77, 3092 (1973). 2 C.

A. Angell, M. Oguni, and W. J. Sichina, J. Phys. Chem. 86, 998 (1982). J. Speedy and C. A. Angell, J. Chem. Phys. 65, 851 (1976). 4 D. E. Hare and C. M. Sorensen, J. Chem. Phys. 87, 4840 (1987). 5 R. J. Speedy, J. Phys. Chem. 91, 3354 (1987), and references therein.
3 R.

Downloaded 09 Sep 2012 to 193.48.219.8. Redistribution subject to AIP license or copyright; see http://jcp.aip.org/about/rights_and_permissions

234508-6
57 C.

Suzuki et al.
65 J.

J. Chem. Phys. 136, 234508 (2012) A. Pople, Proc. R. Soc. London, Ser. A 205, 163 (1951). Wernet, D. Nordlund, U. Bergmann, M. Cavalleri, M. Odelius, H. Ogasawara, L. . Nslund, T. K. Hirsch, L. Ojame, P. Glatzel, L. G. M. Pettersson, and A. Nilsson, Science 304, 995 (2004). 67 T. Tokushima, Y. Harada, O. Takahashi, Y. Senba, H. Ohashi, L. G. M. Pettersson, A. Nilsson, and S. Shin, Chem. Phys. Lett. 460, 387 (2008). 68 E. G. Ponyatovsky, V. V. Sinitsyn, and T. A. Pozdnyakova, J. Chem. Phys. 109, 2413 (1998). 69 C. T. Moynihan, Mater. Res. Soc. Symp. Proc. 455, 411 (1997). 70 T. Kuhn, M. E. Earle, A. F. Khalizov, and J. J. Sloan, Atmos. Chem. Phys. 11, 2853 (2011).
66 Ph.

Esen, T. Kaiser, and G. Schweiger, Appl. Spectrosc. 50, 823 (1996). D. King, K. C. Thompson, and A. D. Ward, J. Am. Chem. Soc. 126, 16710 (2004). 59 J. Buajarern, L. Mitchem, and J. P. Reid, J. Phys. Chem. A 111, 13038 (2007). 60 M. J. Clouter, H. Kiefte, and N. Ali, Phys. Rev. Lett. 40, 1170 (1978). 61 M. Musso, F. Matthai, D. Keutel, and K.-L. Oehme, J. Chem. Phys. 116, 8015 (2002). 62 C. M. Davis,Jr and T. A. Litovitz, J. Chem. Phys. 42, 2563 (1965). 63 M. Vedamuthu, S. Singh, and G. W. Robinson, J. Phys. Chem. 98, 2222 (1994). 64 M. J. Frisch, G. W. Trucks, H. B. Schlegel et al., GAUSSIAN 09, Revision A.1, Gaussian, Inc., Wallingford, CT, 2009.
58 M.

Downloaded 09 Sep 2012 to 193.48.219.8. Redistribution subject to AIP license or copyright; see http://jcp.aip.org/about/rights_and_permissions