On the Mechanism of Solvation Dynamics in Imidazolium-Based Ionic

Jul 2, 2013 - Experimental studies of solvation dynamics in imidazolium-based ionic liquids (ILs) have revealed complex kinetics over a broad range of...
0 downloads 0 Views 1010KB Size
Article pubs.acs.org/JPCB

On the Mechanism of Solvation Dynamics in Imidazolium-Based Ionic Liquids Z. L. Terranova and S. A. Corcelli* Department of Chemistry and Biochemistry, University of Notre Dame, Notre Dame, Indiana 46556, United States S Supporting Information *

ABSTRACT: Experimental studies of solvation dynamics in imidazolium-based ionic liquids (ILs) have revealed complex kinetics over a broad range of time scales from femtoseconds to tens of nanoseconds. Microsecond-length molecular dynamics (MD) simulations of coumarin 153 (C153) in 1-ethyl-3-methyl imidazolium tetrafluoroborate, [emim][BF4], were performed to reveal the molecular-level mechanism for solvation dynamics in imidazolium-based ILs over the full range of time scales accessed in the experiments. The solvation response of C153 in [emim][BF4] compared favorably with experiment. An analysis of the structure of the IL in the vicinity of the C153 dye revealed preferential solvation by the [emim] cations. Despite this observation, decomposition of the solvation response into components from the anions and cations and also from translational and rotational motions shows that translations of the [BF4] anions are the dominant contributor to solvation dynamics. The kinetics for the translation of the [BF4] anions into and out of the first solvation shell of the dye were found to mimic the kinetic profile of the solvation dynamics response. This mechanism for solvation dynamics contrasts dramatically with conventional polar liquids in which solvent rotations are generally responsible for the response.

I. INTRODUCTION

detailed microscopic mechanisms and time scales of solvation dynamics in ILs is of pressing relevance. Solvation dynamics are measured in time-dependent fluorescence Stokes shift (TDSS) experiments. TDSS experiments employ fluorescent probe molecules to measure the response of the environment to an instantaneous perturbation of the charge distribution of the probe molecule.27−30 In brief, the experiment begins with an initial laser pulse, ν(0), that electronically excites the probe molecule. This excitation occurs effectively instantaneously, such that the molecular geometry of the probe molecule is unchanged, but its charge distribution is significantly altered. After the initial excitation, the solvent environment begins to reorganize to accommodate the new charge distribution. The reorganization stabilizes the excited electronic state of the probe relative to its ground electronic state, and subsequent fluorescence, ν(t), is red-shifted with respect to the initial fluorescence, ν(0). The time scales of the environment reorganization are typically characterized by a solvation response function

Ionic liquids (ILs) are attracting considerable attention because of their attractive properties as environmentally friendly alternatives to volatile organic solvents1,2 and their applications involving the production, storage, and efficient utilization of energy3 while also demonstrating tremendous promise in a variety of liquid separation and extraction strategies.4 ILs exhibit unique physical properties relative to conventional liquids in terms of vapor pressure, viscosity, electrical and thermal conductivity, solubility of polar and nonpolar molecules, and melting point.5 Moreover, these properties can be tuned to specific applications by chemically modifying the molecules that comprise the liquid, and/or by forming multicomponent mixtures of anions and cations, making ILs highly adaptable for a variety of tasks. In general, dynamics play an important role in determining mass and heat transport properties of ILs, which are crucial to many energy related applications. There have been numerous experimental measurements of solvation dynamics in ILs.6−25 Solvation dynamics refers to molecular reorganizations in response to a perturbation in the geometric or electronic structure of a solute and are particularly important for chargetransfer reactions whose kinetic rates are determined almost exclusively by solvent reorganization. Solvent reorganization also influences the kinetic rates and mechanisms of other classes of chemical reactions that involve polar transition states.26 Many proposed and actual applications of ILs involve charge-transfer reactions (e.g., dye-sensitized solar cells, batteries, and many catalytic reactions), so understanding the © 2013 American Chemical Society

S( t ) =

v (t ) − v (∞ ) v(0) − v(∞)

(1)

Special Issue: Michael D. Fayer Festschrift Received: June 28, 2013 Revised: July 2, 2013 Published: July 2, 2013 15659

dx.doi.org/10.1021/jp406419y | J. Phys. Chem. B 2013, 117, 15659−15666

The Journal of Physical Chemistry B

Article

electronically polarizable force field, and their results for the solvation response functions agree well with experiment, although the long time decay was slightly too slow. Once again, translational motions of the molecules were implicated as being particularly important to the solvation response. By decomposing the total response in terms of auto- and crosscorrelation functions for the anions and cations, Schmollngruber, Schröder, and Steinhauser found that the anion and cation autocorrelation functions were similar, and the crosscorrelation functions were highly anticorrelated. In this paper, we present calculations of the solvation response over the full range of time scales (from 10 fs to 10 ns) accessed in measurements of C153 in 1-ethyl-3-methyl imidazolium ([emim]) and tetrafluoroborate ([BF4]). These calculations utilized microsecond-length molecular dynamics (MD) simulations with the C153 and IL molecules modeled in full atomistic detail. The calculated response is compared directly with experiment, and analysis of the simulations with decomposition strategies developed previously to understand solvation dynamics in complex biological systems49−52 provide insights into the mechanism of solvation dynamics in this imidazolium-based IL.

where ν(∞) is the emission frequency after the environment has completely responded to the excited state charge distribution of the dye. The majority of the solvation dynamics measurements on ILs employ the time-correlated single photon counting (TCSPC) technique to measure the TDSS following optical excitation of the probe molecule.11,12,16−25 TCSPC has a time resolution of approximately 20 ps; therefore, any solvation dynamics occurring faster than 20 ps are left largely unresolved. Since the TDSS often persists to ∼10 ns in ILs, TCSPC is an ideal technique to capture these relatively long time scales. In practice, though, TCSPC will miss about half of the total solvation response.22 Several investigators have utilized fluorescence upconversion spectroscopy with sub-100 fs time resolution31,32 to measure faster portions of the TDSS in ILs.9,10,13−15 By combining the two measurement techniques, the full solvation response from ∼10 fs to ∼10 ns is elucidated.9,10,14,15 Recently, Maroncelli, Ernsting, and coworkers have measured the TDSS of coumarin 153 (C153) in a series of ILs from 50 fs to 20 ns.6,7 Immediately apparent in their results is a complex, highly nonexponential kinetic profile with a noticeable plateau in the region between 1 and 10 ps, where the TDSS slows appreciably. Trends with respect to varying the identity of the cations and anions are not immediately apparent, although the slowest time scales do correlate well with viscosity.8,11,20 Additionally, there have been a number of previous theoretical and simulation studies of solvation dynamics in ILs.33−45 Kim and co-workers have examined important aspects of IL dynamics, including solvation dynamics, with both united and all-atom fully flexible models.36−38,43 For the united atom studies, solvation responses were computed to time scales as long as 1 ns, whereas shorter time scales of 10 ps were considered for the all-atom simulations. In these studies, the solute was modeled either as a diatomic or with a benzene-like structure. Kobrak and co-workers performed simulations of solvation dynamics that closely mimic experiment in so much as the all-atom, fully flexible liquids contain a realistically modeled C153 solute.39,40 The solvation response was computed to 10 ps, and a number of decomposition strategies were employed to understand the factors responsible for the onset of solvation dynamics in ILs. In particular, Kobrak emphasizes the importance of the translations of the ions to the early solvation response.40 This is in stark contrast to conventional polar liquids, whose short-time solvation dynamics are dominated by rotations.29 Kobrak encountered some difficulties in applying certain decomposition strategies to his calculated solvation response functions because of the nonpairwise-additive nature of the Ewald summation technique.46−48 Roy and Maroncelli performed extensive equilibrium and nonequilibrium simulations of a variety of model solutes, as well as a fully atomistic C153 probe, in a coarse grained IL model.34 For the C153 solute, the computed solvation response was in generally good agreement with experiment with the amplitude of the short time response being slightly overestimated. Roy and Maroncelli also noted the importance of translational motions of the IL molecules to the inertial solvation response, as well as reasonable agreement between the equilibrium and nonequilibrium solvation response functions implying the validity of the linear response approximation. Most recently, Schmollngruber, Schröder, and Steinhauser investigated the solvation dynamics of C153 in three imidazolium-based ILs.33 Their simulations utilized an

II. THEORETICAL METHODOLOGY A. Solvation Response Calculations. The solvation response of C153 in [emim][BF4] was computed using a methodology that has been utilized and validated extensively in previous MD studies of solvation dynamics.53−55 The central quantity in this approach is ΔE(t) = Ee(t) − Eg(t), which represents the difference in the interaction energy of the probe molecule with its environment for its excited (Ee) and ground (Eg) electronic states. The electronic states are modeled classically as two different distributions of atomic-centered charges, which have been derived and validated previously from density functional theory calculations.56 This allows for the calculation of the equilibrium time correlation function for the fluctuations in the solvation energy differences within a MD simulation C(t ) =

⟨δ ΔE(0)δ ΔE(t )⟩ ⟨(δ ΔE)2 ⟩

(2)

where δΔE(t) = ΔE(t) − ⟨ΔE⟩ and ⟨...⟩ is an ensemble average in the ground electronic state of the probe. Within linear response theory, which previous studies have generally found to be applicable to solvation dynamics in ILs at early times in the response,41 C(t) is equal to S(t), thus allowing direct comparisons with experiment.55,57 During calculations of ΔE(t), the long-ranged electrostatic interactions between C153 and the IL molecules were computed with the damped shifted force (DSF) method.58 The DSF method is used as a viable alternative to the traditional Ewald summation technique,46−48 and facilitates decomposition of the solvation response (see section II.B) because it is based explicitly on a pairwise sum. B. Decomposition of the Solvation Response by Solvent Component. The total solvation response can be decomposed into contributions from the different constituents present in the liquid. Such decompositions are made possible because of the additive nature of the solute interaction energy, ΔE(t) = ∑α ΔEα(t), where ΔEα(t) is the solute interaction energy with solvent component α, in this case the anions, cations, and intramolecular electrostatic interactions within the 15660

dx.doi.org/10.1021/jp406419y | J. Phys. Chem. B 2013, 117, 15659−15666

The Journal of Physical Chemistry B

Article

Since imidazolium-based ILs are known to exhibit slow (nanosecond) relaxation dynamics, we employed a rigorous equilibration protocol before performing the production-run simulations from which the solvation response function was calculated. Following an initial period of minimization, the temperature was slowly raised from 0 to 300 K over a period of 200 ps. This was followed by a 4 ns simulation in the NPT ensemble (300 K and 1 atm) using a Nosé−Hoover thermostat and barostat.67 The size of the simulation box was isotropically scaled to reflect the average density and was then simulated in an NVT ensemble for 1 ns at 300 K; the temperature was raised to 600 K over 1 ns to destroy any pseudostable ionic cages that may have formed. Next, the temperature was reduced back to 300 K in 1 ns, followed by an NVT simulation for 1 ns at a constant temperature of 300 K. The final velocities were scaled to 300 K, and a preproduction NVE simulation was performed for 11 ns. Production runs (comprising a total of 5.015 μs) were performed in the NVE ensemble with a 2 fs integration time step and a collection resolution of 10 fs. A 100 ns control simulation of pure [emim][BF4] was performed to validate the IL force field. The average density of the liquid was found to be 1.19 g/cm3, which compares reasonably well to experiment (1.28 g/cm3).68 It is important to note that the partial charges for [emim] and [BF4] were empirically scaled by a factor of 0.80, guided by DFT calculations,69 to achieve better agreement with experiments where dynamic properties are of interest. Charge scaling has become a fairly common practice in IL MD simulations and a valid means of increasing diffusion.70 The scaling factor can be regarded as an additional empirical force-field parameter whose physical role is to account, in an approximate and empirical fashion, for the effects of electronic polarizability. This adjustment decreases the calculated density from 1.27 g/cm3 in a simulation with full charges64 to 1.19 g/cm3 with scaled charges. However, the computed self-diffusion constants, 8.1 × 1011 m2/s for [emim] and 5.5 × 1011 m2/s for [BF4], compare more favorably with experiment (5.0 × 1011 and 4.2 × 1011 m2/ s) than diffusion constants reported for simulations with full charges (1.1 × 1011 and 0.9 × 1011 m2/s).64

probe itself. Unfortunately, there is not a unique decomposition of eq 1 simply based on expressing ΔE(t) as a sum. However, one decomposition strategy is formally compatible with linear response theory42,50,59 C α(t ) =

⟨δ ΔE α(t )δ ΔE(0)⟩ ⟨(δ ΔE)2 ⟩

(3)

where the superscript α signifies the solvent component of interest. Previous investigations have confirmed empirically that, when linear response theory holds, Cα(t) corresponds to the contributions to S(t) from the solvent components (note that S(t) does uniquely decompose due to the additive property ΔE(t)).59,60 Kobrak and Znamenskiy42 were the first to employ eq 3 in the context of solvation dynamics. They decomposed the solvation response of the solute betaine-30 in 1-butyl-3methyl imidazolium ([bmim]) and hexafluorophosphate ([PF6]) over the range from 1 to 5 ps and found that the anion dominates the response on this time scale. C. Decomposition of the Solvation Response by Translations and Rovibrations. The solvation response functions for the anions and cations of the IL can be further decomposed into contributions from their respective translational and rovibrational motions. The translational contribution, ΔEαtrans, to ΔEα for the cations and anions is computed by regarding each relevant IL molecule as a single point charge located at its center of charge. The rovibrational contribution, ΔEαrovib, to ΔEα is then just obtained as the difference, ΔEαrovib = ΔEα − ΔEαtrans. With these definitions, the correlation functions for the cations and anions decompose into translational and rovibrational contributions α α C α(t ) = C trans (t ) + Crovib (t )

(4)

where the functions Cαtrans(t) and Cαrovib(t) are computed with eq 3 using Cαtrans(t) and Cαrovib(t) in place of ΔEα(t). D. Molecular Dynamics Simulations. The MD simulations were performed using the large-scale atomic/molecular massively parallel simulator (LAMMPS)61 program with periodic boundary conditions and a cubic simulation box containing 256 [emim] cations, 256 [BF4] anions, and one C153 molecule. All of the molecules were modeled as fully flexible, except for covalent bonds containing hydrogen which were fixed at equilibrium lengths using the SHAKE algorithm.62 For the C153 and [emim] molecules, force-field parameters for the bonds, bends, dihedrals, and atomic-centered LennardJones sites were adopted from the generalized Amber force field (GAFF).63 Parameters for [BF4] are not available in the GAFF, so these parameters were obtained from Liu et al.64 Atomiccentered partial charges for the IL molecules and the ground state of C153 were calculated via the Merz−Singh−Kollman65 analysis of the electron density of the optimized geometry of the molecules obtained with density functional theory (DFT) with a B3LYP functional and the aug-cc-pVDZ basis set. The changes in the partial charges for C153 upon electronic excitation were calculated and validated previously by Cinacchi et al.;56 thus, the excited state partial charges for C153 are obtained by simply adding these changes to the computed ground state charges. In the MD simulations, the long-ranged electrostatic interactions were computed with the particle-mesh Ewald summation method with a 15 Å real-space cutoff.46,66 This same cutoff distance was utilized when computing interactions between Lennard-Jones sites.

III. RESULTS AND DISCUSSION The calculated solvation response for C153 in [emim][BF4] is shown in Figure 1 along with the experimental measurement6,7

Figure 1. Calculated (black) and experimental6,7 (green) solvation response functions in the range from 50 fs to 1 ns. 15661

dx.doi.org/10.1021/jp406419y | J. Phys. Chem. B 2013, 117, 15659−15666

The Journal of Physical Chemistry B

Article

for direct comparison. The simulated solvation response is in good overall agreement with experiment, and it reproduces the complex nonexponential decay, including the plateau in the ∼1−10 ps regime reminiscent of dynamics in supercooled liquids. The pronounced oscillations in the calculated response at times less than 10 ps are due to internal motions of the C153 probe molecule and will be discussed in more detail below. While the calculated response is qualitatively similar to experiment over all time scales, it decays too quickly in the regime beyond 10 ps. To quantify the differences in the longtime decay between experiment and theory, both solvation response curves were fit with a stretched exponential function, A exp(−(t/τ)β), in the range from 10 ps to 1 ns (Table 1 and Table 1. Parameters for a Stretched Exponential, A exp(−(t/ τ)β), Fit to the Calculated and Experimental Solvation Response Functions in the Range from 10 ps to 1 ns A τ (ps) β

experiment

theory

0.38 115.7 0.55

0.32 81.4 0.74

Figure 2. The calculated total simulated solvation response (black) is decomposed using the methodology described in section II.B into contributions from the anions (red), the cations (blue), and the internal motions of the C153 solute (purple).

correlation functions presented most recently by Schmollngruber, Schröder, and Steinhauser,33 where it is not clear that the anions are the dominant contributor to the solvation response over all time scales. The results shown in Figure 2 suggest that the [BF4] anions play a particularly important role in determining the solvation response. To begin to understand the role of the anions in the solvation response, we first focused on the structure of the IL in the vicinity of the C153 probe. Shown in Figure 3 are radial

Figure S1, Supporting Information). The time constant, τ, is shorter in the theoretical result (81.4 ps) than in the experiment (115.7 ps). It has been empirically established that the long-time decay of the solvation response in imidazolium-based ILs is viscosity dependent.8,11,20 Thus, the slightly too large self-diffusion constants for the molecules in the neat [emim][BF4] simulation (section II.D) are a likely culprit for the slightly too fast long-time decay of the solvation response. While it is beyond the scope of the present study, it might be possible to utilize the experimental solvation response data to further refine the dynamical properties of the IL force field via the empirical charge scaling factor discussed in section II.D. While modest discrepancies in the long-time decay exist, the calculated solvation response captures the complex kinetic behavior of the experimental measurement, thus validating the simulations directly with experiment, and we will now proceed to further analyze the simulations to uncover the molecular mechanisms responsible for solvation dynamics in [emim][BF4]. Figure 2 shows the decomposition of the total calculated solvation response of C153 in [emim][BF4] into contributions from the anions, cations, and internal motions of the C153 solute. It is immediately apparent that the anions dominate the solvation response. Not only is the anion contribution substantially larger than that of the cations or C153 for all time scales, it also mimics closely the characteristic shape of the total response curve. The internal motions of the C153 solute do not contribute to the decay of the solvation response (i.e., the C153 component of the response is generally flat). However, these motions do engender pronounced oscillations in the total response function. Such oscillations in the solvation response have been observed in high temporal resolution measurements in solution71 and for a dye molecule (Hoechst 33258) bound to DNA.72 For H33258 bound to DNA, an analysis of the oscillations computed with theoretical methodology similar to that employed in the present work yielded reasonably good agreement with experiment.51 The decomposition in Figure 2, which is formally consistent with the linear response approximation, results in a less ambiguous interpretation than the analysis in terms of auto- and cross-

Figure 3. The radial distribution function, g(r), for C153−[emim] and C153−[BF4] pairs, where r is defined as the distance between the center-of-mass positions of the relevant pair of molecules.

distribution functions, g(r), for the center of mass of the cations and anions relative to the center of mass of the C153 solute. The most striking feature is the dramatic enhancement of [emim] cations in the vicinity of the C153 molecule, represented by the pronounced peak near 4 Å, and the corresponding depletion of [BF4] anions close to C153. The preferential solvation of C153 by [emim] is perhaps not surprising, since the molecules can align their flat surfaces and take advantage of favorable π−π stacking interactions. While the radial distribution function shown in Figure 3 does not give a full characterization of the three-dimensional solvent structure around the nonspherical C153 solute, it does indicate that the 15662

dx.doi.org/10.1021/jp406419y | J. Phys. Chem. B 2013, 117, 15659−15666

The Journal of Physical Chemistry B

Article

response (but still significantly less than the anion translations); however, these contributions are largely offset by the cation rotations, which are uniformly negative in sign on all time scales. Negative values of the cation rotational contribution to the total solvation response indicate that these motions are actually counterproductive, on average, to stabilizing the excited state charge distribution of the C153 solute. The final unresolved issue for obtaining a complete view of the mechanism of solvation dynamics in the [emim][BF4] IL is to determine which specific translational motions of the [BF4] anion are relevant. For this, we computed the residence-time correlation function (RTCF) for the anions and cations in the first solvation shell of the C153 solute, ⟨n(0)n(t)⟩, where n(t) = 1 if the molecule of interest is in the first solvation shell and n(t) = 0 if it is not (Figure 5). For the purposes of the RTCF

[emim] cations are, on average, more proximal to the C153 solute than the [BF4] anions. This observation is consistent with previous studies of the three-dimensional solvent structure around aromatic compounds in imidazolium ILs.73,74 The observation of preferential solvation creates a conundrum for understanding solvation dynamics in [emim][BF4]. Conventionally, solvation dynamics is dominated by molecules close to the fluorescent dye molecule. In this case, however, the anions, which have been implicated as the dominant player in the response dynamics (Figure 2), are generally further away from the C153 dye molecule than the cations. Before we resolve this issue, it is worth noting that the observation of preferential solvation has broader implications. In most experiments that employ an exogenous spectroscopic probe molecule, the general assumption is that the presence of the solute does not greatly alter the native structure and dynamics of interest. In this case, the C153 molecule is apparently inducing structure in the IL. This implies that there is some ambiguity as to whether the experiment is really measuring dynamics that are native to the neat IL. Preferential solvation effects could also be relevant to the interpretations of other experiments on IL structure and dynamics. To determine which specific motions of the [emim] and [BF4] molecules are most relevant to solvation dynamics, we further decomposed the anion and cation solvation responses into contributions from translations and rovibrations (Figure 4). Focusing first on the [BF4] anions, the results are

Figure 5. The residence-time correlation function (RTCF) for the anions and cations in the first solvation shell of the C153 solute. The first solvation shell was defined as any molecule whose center of mass lies within 5.6 Å of the center of mass of the C153 probe.

calculation, the first solvation shell was defined as any molecule whose center of mass lies within 5.6 Å of the center of mass of the C153 probe. 5.6 Å was chosen because it is the first minimum of the C153−[emim] radial distribution function. The RTCF describes the time scales for molecules, either anions or cations, to enter or leave the first solvation shell of the solute. The RTCF for the anion is generally similar to the solvation response. It appears to exhibit the same unusual kinetic profile as the solvation response, which is suggestive that these anionic translational motions are relevant for the response. In contrast, the RTCF for the cations is significantly slower than the anions and its shape is qualitatively different from the solvation response. The time scales for the decay of the anion RTCF are similar, although clearly slower, than the solvation response. However, the difference in the times scales is likely due to the somewhat arbitrary definition of the solvation shell. Changing the cutoff distance for the solvation shell alters the time scales for the RTCF decay of both anions and cations, but the qualitative profiles are unchanged (data not shown). The results in Figures 3−5 support a solvation dynamics mechanism in [emim][BF4] where the translational motion of the anions into and out of the first solvation shell of the C153 probe molecule potentially plays an important role.

Figure 4. The solvation responses of [emim] (blue) and [BF4] (red) decomposed into contributions from translational (solid) and rovibrational (dashed) motions. For reference, the total solvation response minus the contribution from the internal motions of the C153 solute is also shown (black).

unambiguous: translational motions of the anions dominate the total response, whereas [BF4] rovibrational motions are negligible. Note that the shape of the anion translational decomposition nearly perfectly mimics the total solvation response. Since the [BF4] molecule is symmetric and possesses no permanent dipole moment, it is not surprising that its rovibrational motions do not affect the solvation dynamics. However, since solvation dynamics in conventional polar liquids is governed mostly by rotational motions of the solvent, the central importance of the translational motions of [BF4] to the solvation response of C153 in [emim][BF4] is unusual and warrants further investigation. The cations also exhibit unusual and interesting behavior in Figure 4. The translations of the cations are also a significant contributor to the total solvation

IV. SUMMARY Extensive MD simulations have revealed the detailed molecular motions responsible for solvation dynamics of C153 in the 15663

dx.doi.org/10.1021/jp406419y | J. Phys. Chem. B 2013, 117, 15659−15666

The Journal of Physical Chemistry B

Article

visualization expertise. The authors are also grateful for helpful discussions with Professor Mark Maroncelli at The Pennsylvania State University and for generously providing his experimental data for the solvation dynamics of C153 in [emim][BF4].

imidazolium-based IL [emim][BF4] (Figure 6). Solvation dynamics in liquids generally involves collective motions of



(1) Welton, T. Room-Temperature Ionic Liquids. Solvents for Synthesis and Catalysis. Chem. Rev. 1999, 99, 2071−2083. (2) Sheldon, R. Catalytic Reactions in Ionic Liquids. Chem. Commun. 2001, 2399−2407. (3) Wishart, J. F. Energy Applications of Ionic Liquids. Energy Environ. Sci. 2009, 2, 956−961. (4) Huddleston, J. G.; Willauer, H. D.; Swatloski, R. P.; Visser, A. E.; Rogers, R. D. Room Temperature Ionic Liquids as Novel Media for “Clean” Liquid−Liquid Extraction. Chem. Commun. 1998, 1765−1766. (5) Wishart, J. F.; Castner, E. W. The Physical Chemistry of Ionic Liquids. J. Phys. Chem. B 2007, 111, 4639−4640. (6) Zhang, X.-X.; Liang, M.; Ernsting, N. P.; Maroncelli, M. Complete Solvation Response of Coumarin 153 in Ionic Liquids. J. Phys. Chem. B 2013, 117, 4291−4304. (7) Maroncelli, M.; Zhang, X.-X.; Liang, M.; Roy, D.; Ernsting, N. P. Measurements of the Complete Solvation Response of Coumarin 153 in Ionic Liquids and the Accuracy of Simple Dielectric Continuum Predictions. Faraday Discuss. 2012, 154, 409−424. (8) Samanta, A. Solvation Dynamics in Ionic Liquids: What We Have Learned from the Dynamic Fluorescence Stokes Shift Studies. J. Phys. Chem. Lett. 2010, 1, 1557−1562. (9) Kimura, Y.; Fukuda, M.; Suda, K.; Terazima, M. Excited State Intramolecular Proton Transfer Reaction of 4′-N,N-Diethylamino-3hydroxyflavone and Solvation Dynamics in Room Temperature Ionic Liquids Studied by Optical Kerr Gate Fluorescence Measurement. J. Phys. Chem. B 2010, 114, 11847−11858. (10) Arzhantsev, S.; Jin, H.; Baker, G. A.; Maroncelli, M. Measurements of the Complete Solvation Response in Ionic Liquids. J. Phys. Chem. B 2007, 111, 4978−4989. (11) Jin, H.; Baker, G. A.; Arzhantsev, S.; Dong, J.; Maroncelli, M. Solvation and Rotational Dynamics of Coumarin 153 in Ionic Liquids: Comparisons to Conventional Solvents. J. Phys. Chem. B 2007, 111, 7291−7302. (12) Paul, A.; Samanta, A. Solute Rotation and Solvation Dynamics in an Alcohol-Functionalized Room Temperature Ionic Liquid. J. Phys. Chem. B 2007, 111, 4724−4731. (13) Halder, M.; Headley, L. S.; Mukherjee, P.; Song, X.; Petrich, J. W. Experimental and Theoretical Investigations of Solvation Dynamics of Ionic Fluids: Appropriateness of Dielectric Theory and the Role of DC Conductivity. J. Phys. Chem. A 2006, 110, 8623−8626. (14) Lang, B.; Angulo, G.; Vauthey, E. Ultrafast Solvation Dynamics of Coumarin 153 in Imidazolium-Based Ionic Liquids. J. Phys. Chem. A 2006, 110, 7028−7034. (15) Arzhantsev, S.; Jin, H.; Ito, N.; Maroncelli, M. Observing the Complete Solvation Response of DCS in Imidazolium. Ionic Liquids, From the Femtosecond to Nanosecond Regimes. Chem. Phys. Lett. 2006, 417, 524−529. (16) Mandal, P. K.; Samanta, A. Fluorescence Studies in a Pyrrolidinium Ionic Liquid: Polarity of the Medium and Solvation Dynamics. J. Phys. Chem. B 2005, 109, 15172−15177. (17) Saha, S.; Mandal, P. K.; Samanta, A. Solvation Dynamics of Nile Red in a Room Temperature Ionic Liquid Using Streak Camera. Phys. Chem. Chem. Phys. 2004, 6, 3106−3110. (18) Ito, N.; Arzhantsev, S.; Heitz, M.; Maroncelli, M. Solvation Dynamics and Rotation of Coumarin 153 in Alkylphosphonium Ionic Liquids. J. Phys. Chem. B 2004, 108, 5771−5777. (19) Ito, N.; Arzhantsev, S.; Maroncelli, M. The Probe Dependence of Solvation Dynamics and Rotation in the Ionic Liquid 1-Butyl-3methyl-imidazolium Hexafluorophosphate. Chem. Phys. Lett. 2004, 396, 83−91.

Figure 6. A schematic depiction of the mechanism of solvation dynamics in C153/[emim][BF4].

the solvent environment in the vicinity of the fluorescence probe molecule. In this case, however, we have identified a particular molecular motion that appears to be especially relevant to solvation dynamics in [emim][BF4], namely, the translational motion of the [BF4] anion into and out of the first solvation shell of C153. This proposed mechanism is consistent with a number of previous simulation studies that have implicated translational motions and/or motions of the anions as being especially relevant to solvation dynamics in imidazolium-based ILs.34,38,40−42 The MD simulations have also revealed that the IL forms a specialized structure around the C153 solute, in which the first solvation shell is enriched in [emim] while the second solvation shell is enriched in [BF4]. An important question remains for future studies: how general is this anion-translation mechanism? It seems likely that other imidazolium-based ILs, whose cations can preferentially solvate dyes with fused ring structures, will exhibit this mechanism. However, as one deviates from the structural motifs of the C153/[emim][BF4] system, new solvation dynamics mechanisms could emerge. This presents both challenges and opportunities for the theoretical and experimental communities interested in designing ILs with properties selectively tuned for specific applications.



ASSOCIATED CONTENT



AUTHOR INFORMATION

REFERENCES

S Supporting Information *

Details regarding the stretched-exponential fits of the calculated and experimental solvation response functions. This material is available free of charge via the Internet at http://pubs.acs.org. Corresponding Author

*E-mail: [email protected]. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS This work was supported by the American Chemical Society Petroleum Research Fund (52648-ND6) and the Sustainable Energy Initiative at the University of Notre Dame. The authors are thankful for high performance computing resources and support from the Center for Research Computing at the University of Notre Dame, and to Kristina E. Davis for her 15664

dx.doi.org/10.1021/jp406419y | J. Phys. Chem. B 2013, 117, 15659−15666

The Journal of Physical Chemistry B

Article

(20) Arzhantsev, S.; Ito, N.; Heitz, M.; Maroncelli, M. Solvation Dynamics of Coumarin 153 in Several Classes of Ionic Liquids: Cation Dependence of the Ultrafast Component. Chem. Phys. Lett. 2003, 381, 278−286. (21) Chakrabarty, D.; Harza, P.; Chakraborty, A.; Seth, D.; Sarkar, N. Dynamics of Solvent Relaxation in Room Temperature Ionic Liquids. Chem. Phys. Lett. 2003, 381, 697−704. (22) Ingram, J. A.; Moog, R. S.; Ito, N.; Biswas, R.; Maroncelli, M. Solute Rotation and Solvation Dynamics in a Room-Temperature Ionic Liquid. J. Phys. Chem. B 2003, 107, 5926−5932. (23) Karmakar, R.; Samanta, A. Dynamics of Solvation of the Fluorescent State of Some Electron Donor-Acceptor Molecules in Room Temperature Ionic Liquids, [BMIM][(CF3SO2)2N] and [EMIM][(CF3SO2)2N]. J. Phys. Chem. A 2003, 107, 7340−7346. (24) Karmakar, R.; Samanta, A. Steady-State and Time-Resolved Fluorescence Behavior of C153 and PRODAN in Room-Temperature Ionic Liquids. J. Phys. Chem. A 2002, 106, 6670−6675. (25) Karmakar, R.; Samanta, A. Solvation Dynamics of Coumarin153 in a Room-Temperature Ionic Liquid. J. Phys. Chem. A 2002, 106, 4447−4452. (26) Sumi, H.; Marcus, R. A. Dielectric Relaxation and Intramolecular Electron Transfers. J. Chem. Phys. 1986, 84, 4272. (27) Fleming, G. R.; Cho, M. H. Chromophore-Solvent Dynamics. Annu. Rev. Phys. Chem. 1996, 47, 109−134. (28) Jimenez, R.; Fleming, G.; Kumar, P. V.; Maroncelli, M. Femtosecond Solvation Dynamics of Water. Nature 1994, 369, 471− 473. (29) Stratt, R. M.; Maroncelli, M. Nonreactive Dynamics in Solution: The Emerging Molecular View of Solvation Dynamics and Vibrational Relaxation. J. Phys. Chem. 1996, 100, 12981−12996. (30) Maroncelli, M. The Dynamics of Solvation in Polar Liquids. J. Mol. Liq. 1993, 57, 1−37. (31) Zhang, X. X.; Wurth, C.; Zhao, L.; Resch-Genger, U.; Ernsting, N. P.; Sajadi, M. Femtosecond Broadband Fluorescence Upconversion Spectroscopy: Improved Setup and Photometric Correction. Rev. Sci. Instrum. 2011, 82, 063108. (32) Zhao, L. J.; Lustres, J. L. P.; Farztdinov, V.; Ernsting, N. P. Femtosecond Fluorescence Spectroscopy by Upconversion with Tilted Gate Pulses. Phys. Chem. Chem. Phys. 2005, 7, 1716−1725. (33) Schmollngruber, M.; Schröder, C.; Steinhauser, O. Polarization Effects on the Solvation Dynamics of Coumarin C153 in Ionic Liquids: Components and Their Cross-Correlations. J. Chem. Phys. 2013, 138, 204504. (34) Roy, D.; Maroncelli, M. Simulations of Solvation and Solvation Dynamics in an Idealized Ionic Liquid Model. J. Phys. Chem. B 2012, 116, 5951−5970. (35) Daschakraborty, S.; Biswas, R. Ultrafast Solvation Response in Room Temperature Ionic Liquids: Possible Origin and Importance of the Collective and the Nearest Neighbour Solvent Modes. J. Chem. Phys. 2012, 137, 114501. (36) Shim, Y.; Kim, H. J. MD Study of Solvation in the Mixture of a Room-Temperature Ionic Liquid and CO(2). J. Phys. Chem. B 2010, 114, 10160−10170. (37) Jeong, D.; Choi, M. Y.; Jung, Y.; Kim, H. J. 1/f Spectrum and Memory Function Analysis of Solvation Dynamics in a RoomTemperature Ionic Liquid. J. Chem. Phys. 2008, 128, 174504. (38) Shim, Y.; Kim, H. J. Dielectric Relaxation, Ion Conductivity, Solvent Rotation, and Solvation Dynamics in a Room-Temperature Ionic Liquid. J. Phys. Chem. B 2008, 112, 11028−11038. (39) Kobrak, M. N. A Comparative Study of Solvation Dynamics in Room-Temperature Ionic Liquids. J. Chem. Phys. 2007, 127, 184507. (40) Kobrak, M. N. Characterization of the Solvation Dynamics of an Ionic Liquid via Molecular Dynamics Simulation. J. Chem. Phys. 2006, 125, 064502. (41) Shim, Y.; Choi, M. Y.; Kim, H. J. A Molecular Dynamics Computer Simulation Study of Room-Temperature Ionic Liquids. II. Equilibrium and Nonequilibrium Solvation Dynamics. J. Chem. Phys. 2005, 122, 044511.

(42) Kobrak, M. N.; Znamenskiy, V. Solvation Dynamics of RoomTemperature Ionic Liquids: Evidence for Collective Solvent Motion on Sub-Picosecond Timescales. Chem. Phys. Lett. 2004, 395, 127−132. (43) Shim, Y.; Duan, J.; Choi, M. Y.; Kim, H. J. Solvation in Molecular Ionic Liquids. J. Chem. Phys. 2003, 119, 6411−6414. (44) Kashyap, H. K.; Biswas, R. Dipolar Solvation Dynamics in Room Temperature Ionic Liquids: an Effective Medium Calculation Using Dielectric Relaxation Data. J. Phys. Chem. B 2008, 112, 12431−12438. (45) Song, X. Solvation Dynamics in Ionic Fluids: an Extended Debye−Hückel Dielectric Continuum Model. J. Chem. Phys. 2009, 131, 044503−044503−8. (46) Darden, T.; York, D.; Pedersen, L. Particle Mesh Ewald: an N Log(N) Method for Ewald Sums in Large Systems. J. Chem. Phys. 1993, 98, 10089−10092. (47) Onsager, L. Electric Moments of Molecules in Liquids. J. Am. Chem. Soc. 1936, 58, 1486−1493. (48) Ewald, P. The Calculation of Optical and Electrostatic Grid Potential. Ann. Phys. 1921, 64, 253−287. (49) Furse, K. E.; Corcelli, S. A. Dynamical Signature of Abasic Damage in DNA. J. Am. Chem. Soc. 2011, 133, 720−723. (50) Furse, K. E.; Corcelli, S. A. Molecular Dynamics Simulations of DNA Solvation Dynamics. J. Phys. Chem. Lett. 2010, 1, 1813−1820. (51) Furse, K. E.; Corcelli, S. A. The Dynamics of Water at DNA Interfaces: Computational Studies of Hoechst 33258 Bound to DNA. J. Am. Chem. Soc. 2008, 130, 13103−13109. (52) Furse, K. E.; Lindquist, B. A.; Corcelli, S. A. Solvation Dynamics of Hoechst 33258 in Water: An Equilibrium and Nonequilibrium Molecular Dynamics Study. J. Phys. Chem. B 2008, 112, 3231−3239. (53) Ingrosso, F.; Ladanyi, B. M.; Mennucci, B.; Elola, M. D.; Tomasi, J. Solvation Dynamics in Acetonitrile: A Study Incorporating Solute Electronic Response and Nuclear Relaxation. J. Phys. Chem. B 2005, 109, 3553−3564. (54) Kumar, P. V.; Maroncelli, M. Polar Solvation Dynamics of Polyatomic Solutes: Simulation Studies in Acetonitrile and Methanol. J. Chem. Phys. 1995, 103, 3038. (55) Carter, E. A.; Hynes, J. T. Solvation Dynamics for an Ion Pair in a Polar Solvent: Time-Dependent Fluorescence and Photochemical Charge Transfer. J. Chem. Phys. 1991, 94, 5961. (56) Cinacchi, G.; Ingrosso, F.; Tani, A. Solvation Dynamics by Computer Simulation: Coumarin C153 in 1,4-Dioxane. J. Phys. Chem. B 2006, 110, 13633−13641. (57) Laird, B. B.; Thompson, W. H. On the Connection Between Gaussian Statistics and Excited-State Linear Response for TimeDependent Fluorescence. J. Chem. Phys. 2007, 126, 211104. (58) Fennell, C. J.; Gezelter, J. D. Is the Ewald Summation Still Necessary? Pairwise Alternatives to the Accepted Standard for LongRange Electrostatics. J. Chem. Phys. 2006, 124, 234104. (59) Nilsson, L.; Halle, B. Molecular Origin of Time-Dependent Fluorescence Shifts in Proteins. Proc. Natl. Acad. Sci. U.S.A. 2005, 102, 13867. (60) Golosov, A. A.; Karplus, M. Probing Polar Solvation Dynamics in Proteins: A Molecular Dynamics Simulation Analysis. J. Phys. Chem. B 2007, 111, 1482−1490. (61) Plimpton, S. Fast Parallel Algorithms for Short-Range Molecular Dynamics. J. Comput. Phys. 1995, 117, 1−19. (62) Forester, T. R.; Smith, W. SHAKE, Rattle, and Roll: Efficient Constraint Algorithms for Linked Rigid Bodies. J. Comput. Chem. 1998, 19, 102−111. (63) Wang, J. M.; Wolf, R. M.; Caldwell, J. W.; Kollman, P. A.; Case, D. A. Development and Testing of a General Amber Force Field. J. Comput. Chem. 2005, 25, 1157−1174. (64) Liu, Z.; Huang, S.; Wang, W. A Refined Force Field for Molecular Simulation of Imidazolium-Based Ionic Liquids. J. Phys. Chem. B 2004, 108, 12978−12989. (65) Singh, U. C.; Kollman, P. A. An Approach to Computing Electrostatic Charges for Molecules. J. Comput. Chem. 1984, 5, 129− 145. 15665

dx.doi.org/10.1021/jp406419y | J. Phys. Chem. B 2013, 117, 15659−15666

The Journal of Physical Chemistry B

Article

(66) Essmann, U.; Perera, L.; Berkowitz, M. L.; Darden, T.; Lee, H.; Pedersen, L. G. A Smooth Particle Mesh Ewald Method. J. Chem. Phys. 1995, 103, 8577−8593. (67) Shinoda, W.; Shiga, M.; Mikami, M. Rapid Estimation of Elastic Constants by Molecular Dynamics Simulation Under Constant Stress. Phys. Rev. B 2004, 69, 134103. (68) Nishida, T.; Tashiro, Y.; Yamamoto, M. Physical and Electrochemical Properties of 1-Alkyl-3-Methylimidazolium Tetrafluoroborate for Electrolyte. J. Fluorine Chem. 2003, 120, 135−141. (69) Chaban, V. V.; Voroshylova, I. V.; Kalugin, O. N. A New Force Field Model for the Simulation of Transport Properties of Imidazolium-Based Ionic Liquids. Phys. Chem. Chem. Phys. 2011, 13, 7910−7920. (70) Schröder, C. Comparing Reduced Partial Charge Models with Polarizable Simulations of Ionic Liquids. Phys. Chem. Chem. Phys. 2012, 14, 3089. (71) Pérez Lustres, J. L.; Kovalenko, S. A.; Mosquera, M.; Senyushkina, T.; Flasche, W.; Ernsting, N. P. Ultrafast Solvation of N-Methyl-6-quinolone Probes Local IR Spectrum. Angew. Chem., Int. Ed. 2005, 44, 5635−5639. (72) Sajadi, M.; Furse, K. E.; Zhang, X.-X.; Dehmel, L.; Kovalenko, S. A.; Corcelli, S. A.; Ernsting, N. P. Detection of DNA-Ligand Binding Oscillations by Stokes-Shift Measurements. Angew. Chem., Int. Ed. 2011, 50, 9501−9505. (73) Harper, J. B.; Lynden-Bell, R. M. Macroscopic and Microscopic Properties of Solutions of Aromatic Compounds in an Ionic Liquid. Mol. Phys. 2004, 102, 85−94. (74) Hanke, C. G.; Johansson, A.; Harper, J. B.; Lynden-Bell, R. M. Why Are Aromatic Compounds More Soluble Than Aliphatic Compounds in Dimethylimidazolium Ionic Liquids? A Simulation Study. Chem. Phys. Lett. 2003, 374, 85−90.

15666

dx.doi.org/10.1021/jp406419y | J. Phys. Chem. B 2013, 117, 15659−15666