Modifying Vibrational Energy Flow in Aromatic Molecules: Effects of

Jan 15, 2014 - ... determined the energies of multiple molecular vibrational modes, ... was no energy transfer from nitro to phenyl or from nitro to g...
0 downloads 0 Views 3MB Size
Article pubs.acs.org/JPCA

Modifying Vibrational Energy Flow in Aromatic Molecules: Effects of Ortho Substitution Brandt C. Pein and Dana D. Dlott* School of Chemical Sciences, University of Illinois at UrbanaChampaign, Urbana, Illinois 61801, United States S Supporting Information *

ABSTRACT: Ultrafast infrared (IR) Raman spectroscopy was used to measure vibrational energy transfer between nitrobenzene nitro and phenyl groups, in the liquid state at ambient temperature, when ortho substituents (−CH3, −F) were introduced. Quantum chemical calculations were used to assign the vibrations of these molecules to three classes, phenyl, nitro, or global. Combining transient anti-Stokes and Stokes Raman spectra determined the energies of multiple molecular vibrational modes, which were summed to determine the aggregate energies in the phenyl, nitro, or global modes. In a previous study (Pein, B. C. ; Sun, Y.; Dlott, D. D., J. Phys. Chem. A 2013, 117, 6066−6072) it was shown that, in nitrobenzene, there was no energy transfer from nitro to phenyl or from nitro to global modes, but there was some transfer from phenyl to nitro and phenyl to global. The ortho substituents activated energy flow from nitro-to-phenyl and nitro-to-global and reduced phenyl-to-nitro flow. The −CH3 substituent entirely shut down the phenyl-to-nitro pathway, presumably by efficiently directing some of the phenyl energy into methyl bending excitations. There is (inefficient) unidirectional vibrational energy flow in nitrobenzene only in the nitro-to-phenyl direction, whereas in o-nitrotoluene, vibrational energy flows only in the nitro-to-phenyl direction.

1. INTRODUCTION

Generally speaking, there are two approaches to understanding vibrational energy transport at the molecular level that are associated with either chemical physics or materials chemistry. In the chemical physics approach, the focus is on understanding how vibrational energy is transferred among different vibrational states, and in polyatomic molecules these processes become exceedingly complex,15 with a high degree of sensitivity to the details of the molecule’s vibrational potential energy surface (PES).16,17 The chemical physics approach can be studied by spectroscopic methods, such as the IR-Raman method used here,18−20 that monitor the time-dependent excitations of many individual vibrational states. The materials science approach, sometimes called “nanoscale thermal transport”,21 considers aggregate energy flow across engineered interfaces using simple extensions of familiar heat transfer concepts, such as interface thermal conductance. For instance, vibrational energy transport across long-chain molecules can be studied by attaching them to metal surfaces, using lasers to deposit energy into the metal by flash-heating, and probing the arrival of that energy in the molecules22−25 or in the surrounding medium.26,27 The chemical physics approach is useful in understanding how vibrational energy is transferred from one state to another, but because it does not typically emphasize the spatial nature of energy transfer, it can be cumbersome when it is employed to understand how vibrational energy is transported from one

Understanding vibrational energy flow in molecules is important for a number of applications, including chemical reaction dynamics,1 molecular electronics,2−4 phononics5−7 and shock initiation of energetic materials.8−10 To build useful optimized molecular devices such as current-carrying molecular wires2−4 or thermal diodes,5 it would be helpful to understand how to design molecules to control vibrational energy transport,4 in analogy to the way heat flow is managed in macroscale machinery. We envision a toolbox of molecular moieties that might be rationally combined to create the components of molecular machines and wires having desired vibrational energy transport properties.11 As an early step toward that goal, we previously investigated how vibrational energy flow was affected by different substituents on benzene.11−13 In the present study, we focus on how intramolecular vibrational energy flow between a phenyl group and a nitro substituent13 can be affected by additional substituents, specifically ortho substituents −F and −CH3 adjacent to the nitro groups. All our studies were performed on molecules in the liquid state at ambient temperature. We minimized the role of the liquid environment in our interpretation by concentrating only on shorter times, where intramolecular processes are dominant. In studies of benzene, our group previously showed that when ring CH-stretch transitions were pumped, there was essentially no energy dissipated to the bath prior to ∼5 ps, so processes observed in the 0−5 ps time range will be viewed as primarily intramolecular.14 © 2014 American Chemical Society

Received: December 9, 2013 Revised: January 11, 2014 Published: January 15, 2014 965

dx.doi.org/10.1021/jp4120546 | J. Phys. Chem. A 2014, 118, 965−973

The Journal of Physical Chemistry A

Article

symmetry phenyl (C2h). Symmetry lowering splits several of the benzene degenerate modes and might be expected to open up more channels for vibrational energy transfer by relaxing selection rules for mode-to-mode couplings.36 To investigate those effects, we studied the four aryl halides C6H5−X (X = F, Cl, Br, I).12 The lifetimes of phenyl vibrations in those molecules were slightly decreased due to changes in the substituent masses and electronegativities, but the vibrational energy pathways were affected only minimally. We then studied nitrobenzene13 (NB) and made a remarkable observation. When we pumped the phenyl group, via the ring CH-stretch (νCH at ∼3100 cm−1), there was some phenyl-to-nitro and phenyl-to-global transfer, but when we pumped the nitro group, via the νNO1 + νNO2 combination band near 2900 cm−1, there was no detectable nitro-to-phenyl or nitro-to-global transfer. This finding, that energy transfer between phenyl and a nitro substituent is unidirectional, was the motivation for the present study, where we seek to understand how vibrational energy flow in nitrobenzenes could be modified by additional substituents. We also investigated a few alkyl benzenes.11 When we looked at phenyl-to-alkyl energy transfer, we found that phenyl-to methyl transfer (in toluene) was significantly more efficient than transfer from phenyl to other alkyl groups such as isopropyl or tert-butyl.11 Unfortunately, due to the difficulty of IR-Raman measurements, we could not study as many substituted nitrobenzenes as we wished. We were limited to those that are liquids at ambient temperature that produce minimal optical emission when irradiated by intense IR and visible picosecond pulses, and that have relatively large Raman cross sections.18 We found two commercially available liquids that worked well, onitrotoluene and o-fluoronitrobenzene, where the ortho substituents were −CH3 and −F. Going forward, we will use the following abbreviations: nitrobenzene (NB), o-nitrotoluene (ONT), and o-fluoronitrobenzene (OFNB). Adding an ortho substituent could affect phenyl-to-nitro vibrational energy transfer in several ways. The ortho substituent would further lower the molecule’s symmetry to Cs, potentially accelerating vibrational energy transfer. There will be steric interactions between the two adjacent substituents. Steric interactions could affect vibrational energy flow in two ways, by altering the NB geometry, and by opening up pathways for energy transfer between the substituents. Gas phase37 and solid phase experiments,38 and theoretical37,39 studies have shown that steric interactions frequently cause the nitro group to rotate out of the phenyl plane.37,38

location to another. The materials chemistry approach is useful in understanding how aggregate vibrational energy is transported from one location to another, but it does not provide much insight into fundamental mechanisms, the nature of the excitations that transport the energy, and whether those excitations are in thermal equilibrium or not. In the IR-Raman technique,11,13,18−20,28−31 vibrational energy is input into a selected molecular vibrational transition, typically a higher-energy vibrational fundamental such as a C− H stretch, by a ∼1 ps IR pulse. Anti-Stokes Raman spectroscopy is then used to monitor the molecule’s vibrational energy in all Raman-active modes simultaneously. Complementary two-dimensional IR (2DIR) methods have also been developed by other laboratories for related measurements.32,33 Both IR-Raman and 2DIR methods are well-suited for the chemical physics viewpoint. In recent years, we have shown how to adapt IR-Raman experiments to study point-to-point vibrational energy transport.11,13,34 To accomplish this, we need to know that we are probing transitions that are primarily localized on specific parts of a molecule, and for this reason our most recent studies have focused on substituted benzenes.11−13 These molecules clearly consist of separate moieties, and their vibrational spectra have transitions that can clearly be identified as phenyl or substituent vibrations. The protocol we have devised and used previously works as follows. First we use quantum chemistry calculations to categorize the selected molecule’s vibrations (30 or more for a substituted benzene) as being predominantly phenyl, substituent, or global. As a reminder, phenyl and substituent vibrations are those with a potential energy distribution (PED) primarily contributed from atomic displacement on the ring or substituent respectively whereas global vibrations have a significant PED contributed simultaneously from the ring and substituent. We use 2D IR-Raman excitation spectroscopy13,20 to find one IR pump wavelength to selectively generate substituent excitations, and one IR wavelength to selectively generate phenyl excitations. This is accomplished by scanning the IR pump pulses while monitoring the prompt appearance at t = 1 ps, to avoid coherent artifacts that sometimes appear near t = 020 of phenyl or substituent excitations. Then we measure the time evolution of vibrational energy, with phenyl or substituent pumping, by summing the energies of many transitions, to create aggregate time-dependent phenyl, substituent or global energies. Without attempting to summarize all our work in this area, we need to mention a few prior results relevant to the present study. The Raman probe method has a selection bias, because it does not observe transitions with smaller Raman cross sections. In studies of benzene,14,35 and d6-benzene,14 our group monitored the anti-Stokes transitions associated with 10 (C6H6) or 12 (C6D6) of the 30 normal modes. Using a method we termed “ultrafast thermal calorimetry”,14,31 we also measured the time-dependent aggregate energy dissipated by benzene molecules to the bath. Combining these two measurements allowed us to infer the behavior of the vibrational energy that was not directly probed by Raman. We found that, at least in the case of benzene, the time dependence of the aggregate energy in the modes we did observe was representative of, and a reasonably good approximation for the time dependence of all vibrational energy. To study substituted benzenes, we needed to characterize how vibrational energy dynamics were changed in going from high-symmetry benzene (D6h) to the lower-

2. IR-RAMAN TECHNIQUE The IR-Raman technique used in our laboratory has been described in detail previously.14,40−42 The liquids were supplied by Sigma-Aldrich and were used without additional purification. The purity of OFNB was 99% and ONT ≥99%. The liquids, at ambient temperature, were flowed through a 50 μm diameter recirculating stainless steel capillary microjet. The IR pump and visible (532 nm) probe pulses were about 0.7 ps in duration. The IR pulse energies were 35−50 μJ, depending on the IR wavenumber, and the visible pulse energy was 50 μJ. The pump and probe pulses were focused to 60−70 and 60 μm (1/e2) diameters, respectively. Raman detection used a spectrograph with a 532 nm holographic notch filter and a 1340 × 100 CCD detector, which simultaneously acquired Stokes and anti-Stokes spectra at a ∼10 cm−1 resolution. Due to the spectral widths of 966

dx.doi.org/10.1021/jp4120546 | J. Phys. Chem. A 2014, 118, 965−973

The Journal of Physical Chemistry A

Article

cm−1 were noisy and were removed. The assignments39,43−45 of the transitions probed by IR-Raman spectroscopy are indicated in Figure 1. The full list of vibrational assignments, as well as mode vector illustrations, are given in the Supporting Information. Our assignment labeling scheme is explained as follows: νCH‑phen (∼3100 cm−1) is the phenyl C−H stretch, νCC modes (∼1600 cm−1) are phenyl C−C stretches with some in-plane phenyl C−H bend character, νC−N (∼1100 cm−1) or νC−C(Me) (∼1200 cm−1) are global modes involving displacement of the substituent-to-phenyl bonds, as well as in-plane C−H bending, double junction (2-jnct) modes (∼900 and 800 cm−1) are global modes involving the displacement of both substituents away from the ring, γCH (∼1170 cm−1) are phenyl in-plane C− H bends, βCH (∼1000 cm−1) are phenyl breathing modes, φdef (∼670 cm−1) are phenyl C−C−C deformations with some nitro scissoring, and νC−F (∼1200 cm−1) is a phenyl mode involving displacement of the fluorine atom. The methyl C−H stretches (∼2950 cm−1) are labeled νCH‑Me. The νNO modes (∼1540 and 1350 cm−1) are nitro modes with N−O stretching, the nitro and methyl rocking modes ρNO (∼500−600 cm−1) are global modes involving in-plane tilt of the substituent, as well as ring C−C−C deformation and out-of-plane C−H bending. With OFNB, we monitored 21% of the phenyl, 33% of the global, and all the nitro modes, with the exception of the ∼60 cm−1 hindered rotation,46 which we regard as a bath mode47 rather than a nitrobenzene mode. With ONT, we monitored 28% of the phenyl, 21% of the global, 33% of the methyl, and all the nitro modes, again with the exception of the hindered rotation. B. 2D Excitation Spectroscopy. 2D IR-Raman excitation spectra for ONT and OFNB are shown in Figure 2. The waterfall plots are anti-Stokes spectra at 1 ps pump−probe delay, with the ambient temperature background signals subtracted away. The excitation spectra indicate how the initial energy distribution at t = 1 ps depends on pump wavenumber.11,13,20 The peaks on the diagonal result from a combination of parent vibrational populations plus a coherent artifact due to nonlinear light scattering,33,35 so the t = 1 ps diagonal peak heights are not an accurate measure of the parent populations.42,48 However, the coherent artifact decays with the apparatus temporal response (∼1 ps), whereas the population contribution decays more slowly, with the excited state lifetime T1. Provided T1 > 1 ps, we can track the parent populations after ∼1 ps, but the parent populations at t = 0 are uncertain. The off-diagonal peaks in Figure 2 represent daughter transitions, which grow in as the parent excitations decay. The instantaneous population of each mode is proportional to its anti-Stokes intensity divided by its Stokes intensity.31 As a guide to the Stokes intensities of each mode, we plotted the Stokes spectra below each waterfall plot. In addition, we denoted the mode assignments on each Stokes spectrum. We determined the wavenumber-integrated intensities of the moreintense anti-Stokes transitions by fitting them to Voigt line shape functions, as described previously.11,13 The anti-Stokes and Stokes intensities were used to determine absolute vibrational energy densities (J cm−3) at each time delay. These energy densities are spatially averaged over the volume of sample being probed. The pump-wavenumber dependence of the energy densities of nitro stretches and the phenyl νC−C stretch, at 1 ps delay, was plotted in Figure 3. In Figure 3, the IR absorption spectrum was also plotted as a reference, showing

the picosecond 532 nm probe pulses, the overall spectral resolution for the Raman spectroscopy system was 25 cm−1, which in many cases was greater than the natural line widths. The energy density in each mode was computed, as described previously.11,13 Briefly, the energy densities are the populations, computed from the ratios of anti-Stokes and Stokes spectral intensities, scaled by the Raman-shift frequency factors, the transition energies of the modes and the number density of the liquid. IR spectra shown in Figures 1 and 3 were obtained using a commercial FTIR, with 4 cm−1 resolution. The observed vibrational line widths in the IR spectra were the natural line widths. Using the Gaussian 09 computational package, the normal modes of OFNB and ONT were calculated with MP2 perturbation theory utilizing the 6-31G basis. The resulting normal mode vectors and frequencies were used in conjunction with literature assignments39,43−45 of the vibrational spectra. The modes were classified as substituent, phenyl, or global on the basis of their potential energy distributions (PED), as explained in the Supporting Information. According to our assignments, OFNB had 19 phenyl, 15 global, and 2 nitro modes, whereas ONT had 14 phenyl, 24 global, 2 nitro, and 5 methyl modes.

3. RESULTS A. IR and Raman Spectra and Assignments. The Stokes-Raman and IR spectra of ONT and OFNB are shown in Figure 1. The lower frequency regions of the IR spectra 5 ps were not produced by intramolecular phenyl-to-nitro transfer. Instead, these were secondary excitations associated with intermolecular processes. As the pumped NB molecules lost their excess energy to the bath, ultimately leading to a bulk temperature jump of ∼30 K,13 other molecules could uptake some of this energy, which increased the excitation levels of the lower-energy phenyl and global modes. In other words, the longer-time phenyl and global excitations that appeared after nitro pumping of NB resulted from an indirect (nitro vibration → bath → phenyl vibrations on other molecules) process, rather than the direct nitro-to-phenyl process. In addition, the longer-time decay of the nitro population of NB was caused by energy dissipation to the path rather than to the phenyl group. Thus, on the basis of the t < 5 ps data, in NB nitro-to-phenyl and nitro-to-global processes were essentially absent. With phenyl pumping of NB, we must take into account that some nitro excitations were also generated by the pump pulses. Looking at the phenyl pumping transients in Figure 6a, the IRpumped nitro excitations were those that appeared instantaneously, resulting in a population that tracked the rising edge of the phenyl excitation. After this instantaneous rise of nitro 970

dx.doi.org/10.1021/jp4120546 | J. Phys. Chem. A 2014, 118, 965−973

The Journal of Physical Chemistry A

Article

between the nitro group and the ortho substituent modes or the global modes created by adding the ortho substituents. The methyl group seems to be more effective in the competition with the nitro group for the phenyl energy, and that is consistent with our previous study of toluene,11 where we found that phenyl-to-methyl transfer could be efficient.

5. SUMMARY AND CONCLUSIONS In this study, we have applied a method for interpreting IRRaman data that creates a simplified view of point-to-point molecular vibrational energy transfer. We believe this view could be particularly useful for the development of a toolkit of molecular structures to help design molecules with controlled vibrational energy transport. In our studies, the substituted benzenes played a central role, because we can readily view those molecules as consisting of individual parts: a phenyl group whose vibrational transitions were only slightly affected by the substituent, and the substituents themselves. Because phenyl, biphenyl, terphenyl and so forth groups are basic building blocks of many molecular devices, especially molecular electronic devices, this proves to be a useful artifice. So far we have made a number of useful observations. In particular, NB was found to exhibit unidirectional energy transfer, where there was some phenyl-to-nitro transfer but no nitro-to-phenyl transfer. In the present study, we investigated how additional substituents might affect this unidirectional energy transfer. We found that substituents ortho to the nitro groups, which were known to have significant steric interactions with the nitro groups, destroyed the unidirectional property of NB by opening up nitro-to-phenyl and nitro-to-global channels. We also found that adding a methyl substituent could shut down the phenyl-to-nitro process. Because we previously observed in studies of toluene that phenyl-to-methyl transfer was facile, we believe the methyl group successfully competes against the nitro group for the phenyl energy. A diode is an electronic component where current flows only in one direction. The efficiencies of the vibrational energy transfer processes we observe are nothing like what is seen with today’s highly developed electronic devices, but according to our findings, NB is a vibrational energy diode where energy flows only in the phenyl-to-nitro direction, whereas ONT is a vibrational energy diode where energy flows only in the nitroto-phenyl direction.

Figure 6. Aggregate energy densities of phenyl, nitro, and global modes after either nitro or phenyl pumping. The dashed curves denoted “parent” represent the time dependence of vibrational energies of the parent phenyl or nitro modes, but the parent amplitudes were normalized to put them on scale, so the parent energy densities are not as indicated. Nitro pumping initially produced primarily nitro excitations, but phenyl pumping was less selective, so phenyl pumping also generated some nitro excitations. With phenyl pumping the nitro populations generated by IR pumping rose instantaneously. The nitro populations created by phenyl-to-nitro transfer rose more gradually over the first ∼2 ps.

With phenyl pumping of ONT, the entire nitro population seen in Figure 6b has an instantaneous rise, so it was all generated by the IR pump pulses. There was no delayed rise in the nitro population, and therefore in ONT with phenyl pumping there is practically no phenyl-to-nitro transfer. There is phenyl-toglobal transfer in ONT, at about the same level as with NB. In the OFNB data in Figure 6c, it can again be seen from the nitro pumping data that the ortho substituent −F allows energy to flow from the nitro groups into the phenyl and global modes, with about the same efficiency as with ONT. With phenyl pumping, we observed instantaneously generated nitro excitations and a small amount of delayed nitro excitation, which caused the nitro population in Figure 6c to rise slightly slower than the phenyl population. Thus there was significantly less phenyl-to-nitro transfer in either ONT or OFNB compared to NB. From Figure 6, we conclude that ortho substituents activate the nitro-to-phenyl and nitro-to-global pathways that were absent in NB. We also conclude that ortho substituents suppressed phenyl-to-nitro transfer, compared to NB. This suppression was more efficient with the ortho methyl substituent than with the ortho fluoro substituent. At the present time, it is not possible to present compelling and conclusive quantitative explanations for these effects, but reasonable qualitative interpretations are possible, at least at the “toolkit” level. A nitro-group excitation on NB is isolated from the phenyl and global modes. But in the ortho-substituted species, there are steric interactions between the nitro and the ortho substituents, allowing the nitro excitations to transfer energy to modes involving ortho substituent displacements. These interactions activate nitro-to-phenyl and nitro-to-global transfer. The ortho substituents also tend to suppress phenylto-nitro transfer, with suppression by the ortho methyl substituent being more effective. Presumably, the suppression of nitro-to-phenyl transfer results results from a competition



ASSOCIATED CONTENT

S Supporting Information *

Details on the MP2 normal mode calculations and mode classifications. Included are the illustrations of each normal mode, as well as tables indicating computed frequencies, our mode numbering and classification, computed normalized Raman intensities and potential energy distribution percentages (PED%) used to classify each mode. This information is available free of charge via the Internet at http://pubs.acs.org.



AUTHOR INFORMATION

Corresponding Author

*D. D. Dlott: e-mail, [email protected]. Notes

The authors declare no competing financial interest. 971

dx.doi.org/10.1021/jp4120546 | J. Phys. Chem. A 2014, 118, 965−973

The Journal of Physical Chemistry A



Article

(22) Wang, Z.; Carter, J. A.; Lagutchev, A.; Koh, Y. K.; Seong, N.-H.; Cahill, D. G.; Dlott, D. D. Ultrafast Flash Thermal Conductance of Molecular Chains. Science 2007, 317, 787−790. (23) Wang, Z.; Cahill, D. G.; Carter, J. A.; Koh, Y. K.; Lagutchev, A.; Seong, N.-H.; Dlott, D. D. Ultrafast Dynamics of Heat Flow across Molecules. Chem. Phys. 2008, 350, 31−44. (24) Carter, J. A.; Wang, Z.; Dlott, D. D. Spatially Resolved Vibrational Energy Transfer in Molecular Monolayers. J. Phys. Chem. A 2008, 112, 3523−3529. (25) Carter, J. A.; Wang, Z.; Fujiwara, H.; Dlott, D. D. Ultrafast Excitation of Molecular Adsorbates on Flash-Heated Gold Surfaces. J. Phys. Chem. A 2009, 113, 12105−12114. (26) Putnam, S. A.; Cahill, D. G.; Braun, P. V. Thermal Conductivity of Nanoparticle Suspensions. J. Appl. Phys. 2006, 99, 084308. (27) Ge, Z. B.; Cahill, D. G.; Braun, P. V. Thermal Conductance of Hydrophilic and Hydrophobic Interfaces. Phys. Rev. Lett. 2006, 96, 186101. (28) Deàk, J. C.; Iwaki, L. K.; Dlott, D. D. Vibrational Energy Redistribution in Polyatomic Liquids: Ultrafast Ir-Raman Spectroscopy of Acetonitrile. J. Phys. Chem. 1998, 102, 8193−8201. (29) Iwaki, L. K.; Dlott, D. D. Three-Dimensional Spectroscopy of Vibrational Energy Relaxation in Liquid Methanol. J. Phys. Chem. A 2000, 104, 9101−9112. (30) Dlott, D. D. Vibrational Energy Redistribution in Polyatomic Liquids: 3D Infrared-Raman Spectroscopy. Chem. Phys. 2001, 266, 149−166. (31) Pein, B. C.; Dlott, D. D. Vibrational Energy and Molecular Thermometers in Liquids: Ultrafast IR-Raman Spectroscopy. In Ultrafast Infrared Vibrational Spectroscopy; Fayer, M. D., Ed.; CRC Press Taylor & Francis Group: Boca Raton, FL, 2013; pp 269−304. (32) Tesar, S. L.; Kasyanenko, V. M.; Rubtsov, I. V.; Rubtsov, G. I.; Burin, A. L. Theoretical Study of Internal Vibrational Relaxation and Energy Transport in Polyatomic Molecules. J. Phys. Chem. A 2013, 117, 315−323. (33) Chen, H. L.; Bian, H. T.; Li, J. B.; Wen, X. W.; Zheng, J. R. Ultrafast Multiple-Mode Multiple-Dimensional Vibrational Spectroscopy. Int. Rev. Phys. Chem. 2012, 31, 469−565. (34) Deàk, J. C.; Pang, Y.; Sechler, T. D.; Wang, Z.; Dlott, D. D. Vibrational Energy Transfer across a Reverse Micelle Surfactant Layer. Science 2004, 306, 473−476. (35) Iwaki, L. K.; Deàk, J. C.; Rhea, S. T.; Dlott, D. D. Vibrational Energy Redistribution in Liquid Benzene. Chem. Phys. Lett. 1999, 303, 176−182. (36) von Benten, R. S.; Liu, Y. X.; Abel, B. Dynamical Consequences of Symmetry Breaking in Benzene and Difluorobenzene. J. Chem. Phys. 2010, 133. (37) Shishkov, I. F.; Khristenko, L. V.; Vilkov, L. V.; Samdal, S.; Gundersen, S. Molecular Structure of Ortho-Fluoronitrobenzene Studied by Gas Electron Diffraction and Ab Initio MO Calculations. Struct. Chem 2003, 14, 151−157. (38) Yakovenko, A. A.; Antipin, M. Y.; Timofeeva, T. V. Molecular and Crystal Structure of Low Melting Nitrotoluene Isomers. Cryst. Growth Des. 2009, 9, 57−65. (39) Arjunan, V.; Govindaraja, S. T.; Sakiladevi, S.; Kalaivani, M.; Mohan, S. Spectroscopic, Electronic Structure and Natural Bond Orbital Analysis of o-Fluoronitrobenzene and p-Fluoronitrobenzene: A Comparative Study. Spectrochim. Acta. A 2011, 84, 196−209. (40) Shigeto, S.; Pang, Y.; Fang, Y.; Dlott, D. D. Vibrational Relaxation of Normal and Deuterated Liquid Nitromethane. J. Phys. Chem. B 2007, 112, 232−241. (41) Fang, Y.; Shigeto, S.; Seong, N. H.; Dlott, D. D. Vibrational Energy Dynamics of Glycine, N-Methylacetamide, and Benzoate Anion in Aqueous (D2O) Solution. J. Phys. Chem. A 2009, 113, 75−84. (42) Pakoulev, A.; Wang, Z. H.; Dlott, D. D. Vibrational Relaxation and Spectral Evolution Following Ultrafast OH Stretch Excitation of Water. Chem. Phys. Lett. 2003, 371, 594−600. (43) Rao, P. M.; Rao, G. R. Vibrational-Spectra and Normal Coordinate Analysis of Monohalogenated Nirtobenzenes. J. Raman Spectrosc . 1989, 20, 529−540.

ACKNOWLEDGMENTS The research described in this study is based on work supported by the Office of Naval Research under award N00014-11-1-0418, the National Science Foundation under award DMR-09-55259, and the U.S. Air Force Office of Scientific Research under award number FA9550-09-1-0163.



REFERENCES

(1) Voth, G. A.; Hochstrasser, R. M. Transition State Dynamics and Relaxation Processes in Solutions: A Frontier of Physical Chemistry. J. Phys. Chem. 1996, 100, 13034−13049. (2) Segal, D.; Nitzan, A. Heating in Current Carrying Molecular Junctions. J. Chem. Phys. 2002, 117, 3915−3927. (3) Segal, D.; Nitzan, A.; Hänggi, P. Thermal Conductance through Molecular Wires. J. Chem. Phys. 2003, 119, 6840−6855. (4) Galperin, M.; Ratner, M. A.; Nitzan, A. Molecular Transport Junctions: Vibrational Effects. J. Phys. Condens. Matter 2007, 19, 103201. (5) Segal, D.; Nitzan, A. Heat Rectification in Molecular Junctions. J. Chem. Phys. 2005, 122, 104704. (6) Segal, D.; Nitzan, A. Molecular Heat Pump. Phys. Rev. E 2006, 73, 026109. (7) Li, N.; Ren, J.; Wang, L.; Zhang, G.; Hänggi, P.; Li, B. Colloquium: Phononics: Manipulating Heat Flow with Electronic Analogs and Beyond. Rev. Mod. Phys. 2012, 84, 1045−1066. (8) Dlott, D. D.; Fayer, M. D. Shocked Molecular Solids: Vibrational up Pumping, Defect Hot Spot Formation, and the Onset of Chemistry. J. Chem. Phys. 1990, 92, 3798−3812. (9) Tokmakoff, A.; Fayer, M. D.; Dlott, D. D. Chemical Reaction Initiation and Hot-Spot Formation in Shocked Energetic Molecular Materials. J. Phys. Chem. 1993, 97, 1901−1913. (10) Tarver, C. M. Multiple Roles of Highly Vibrationally Excited Molecules in the Reaction Zones of Detonation Waves. J. Phys. Chem. A 1997, 101, 4845−4851. (11) Pein, B. C.; Sun, Y.; Dlott, D. D. Controlling Vibrational Energy Flow in Liquid Alkylbenzenes. J. Phys. Chem. B 2013, 117, 10898− 10904. (12) Pein, B. C.; Seong, N.-H.; Dlott, D. D. Vibrational Energy Relaxation of Liquid Aryl-Halides X-C6H5 (X = F, Cl, Br, I). J. Phys. Chem. A 2010, 114, 10500−10507. (13) Pein, B. C.; Sun, Y.; Dlott, D. D. Unidirectional Vibrational Energy Flow in Nitrobenzene. J. Phys. Chem. A 2013, 117, 6066−6072. (14) Seong, N.-H.; Fang, Y.; Dlott, D. D. Vibrational Energy Dynamics of Normal and Deuterated Liquid Benzene. J. Phys. Chem. A 2009, 113, 1445−1452. (15) Nitzan, A.; Jortner, J. Vibrational Relaxation of a Molecule in a Dense Medium. Mol. Phys. 1973, 25, 713−734. (16) Kenkre, V. M.; Tokmakoff, A.; Fayer, M. D. Theory of Vibrational Relaxation of Polyatomic Molecules in Liquids. J. Chem. Phys. 1994, 101, 10618−10629. (17) Gulmen, T. S.; Sibert, E. L. Vibrational Energy Relaxation of the OH Stretch in Liquid Methanol. J. Phys. Chem. A 2004, 108, 2389− 2401. (18) Deàk, J. C.; Iwaki, L. K.; Rhea, S. T.; Dlott, D. D. Ultrafast Infrared-Raman Studies of Vibrational Energy Redistribution in Polyatomic Liquids. J. Raman Spectrosc. 2000, 31, 263−274. (19) Iwaki, L. K.; Deàk, J. C.; Rhea, S. T.; Dlott, D. D. Vibrational Energy Redistribution in Polyatomic Liquids: Ultrafast IR-Raman Spectroscopy. In Ultrafast Infrared and Raman Spectroscopy; Fayer, M. D., Ed.; Marcel Dekker: New York, 2001; pp 541−592. (20) Sun, Y.; Pein, B. C.; Dlott, D. D. Three-Dimensional Spectroscopy of Vibrational Energy in Liquids: Nitromethane and Acetonitrile. J. Phys. Chem. B 2013, 117, 15444−15451. (21) Cahill, D. G.; Ford, W. K.; Goodson, K. E.; Mahan, G. D.; Majumdar, A.; Maris, H. J.; Merlin, R.; Phillpot, S. R. Nanoscale Thermal Transport. J. Appl. Phys. 2003, 93, 793−818. 972

dx.doi.org/10.1021/jp4120546 | J. Phys. Chem. A 2014, 118, 965−973

The Journal of Physical Chemistry A

Article

(44) Qayyum, M.; Reddy, B. V.; Rao, G. R. Vibrational Analysis of Mononitro Substituted Benzamides, Benzaldehydes and Toluenes Part I. Vibrational Spectra, Normal Coordinate Analysis and Transferability of Force Constants of Nitrobenzamides, Nitrobenzaldehydes and Nitrotoluenes. Spectrochim. Acta A 2004, 60, 279−290. (45) Varsányi, G. R.; Láng, L. Assignments for Vibrational Spectra of Seven Hundred Benzene Derivatives; Wiley: New York, 1974. (46) Rice, B. M.; Trevino, S. F. An Intermolecular H - O Potential for Methyl Rotations in Solid Nitromethane. J. Chem. Phys. 1991, 94, 7478−7485. (47) Deàk, J. C.; Iwaki, L. K.; Dlott, D. D. Vibrational Energy Redistribution in Polyatomic Liquids: Ultrafast IR-Raman Spectroscopy of Nitromethane. J. Phys. Chem. A 1999, 103, 971−979. (48) Pakoulev, A.; Wang, Z. H.; Pang, Y. S.; Dlott, D. D. Vibrational Energy Relaxation Pathways of Water. Chem. Phys. Lett. 2003, 380, 404−410. (49) Graener, H.; Laubereau, A. Ultrafast Overtone Excitation for the Study of Vibrational Population Decay in Liquids. Chem. Phys. Lett. 1983, 102, 100−104. (50) Deàk, J. C.; Iwaki, L. K.; Dlott, D. D. When Vibrations Interact: Ultrafast Energy Relaxation of Vibrational Pairs in Polyatomic Liquids. Chem. Phys. Lett. 1998, 293, 405−411.

973

dx.doi.org/10.1021/jp4120546 | J. Phys. Chem. A 2014, 118, 965−973