Picosecond Solvation Dynamics in Nanoconfinement: Role of Water

Article Cite This: J. Phys. Chem. B 2018, 122, 3996−4005

pubs.acs.org/JPCB

Picosecond Solvation Dynamics in Nanoconfinement: Role of Water and Host−Guest Complexation Suman Biswas,† Santanu Santra,‡ Semen Yesylevskyy,§ Jyotirmay Maiti,† Madhurima Jana,*,‡ and Ranjan Das*,† †

Department of Chemistry, West Bengal State University, Barasat, Kolkata 700126, India Molecular Simulation Laboratory, Department of Chemistry, National Institute of Technology, Rourkela 769008, Orissa, India § Institute of Physics, National Academy of Sciences of Ukraine, 03028 Kyiv, Ukraine J. Phys. Chem. B 2018.122:3996-4005. Downloaded from pubs.acs.org by ST FRANCIS XAVIER UNIV on 09/02/18. For personal use only.

‡

S Supporting Information *

ABSTRACT: The dynamics of solvation of an excited chromophore, 5-(4″dimethylaminophenyl)-2-(4′-sulfophenyl)oxazole, sodium salt (DMO), has been explored in confined nanoscopic environments of β-cyclodextrin (βCD) and heptakis(2,6-di-O-methyl)-β-cyclodextrin (DIMEB). Solvation occurs on a distinctly slower time scale (τS3 ∼ 47 ps, τS4 ∼ 517 ps) in the host cavity of DIMEB than in that of βCD (τS3 ∼ 20 ps, τS4 ∼ 174 ps). The calculated equilibrium solvation response of DMO was characterized by four relaxation components (τS1 ∼ 0.46−0.48 ps, τS2 ∼ 3.2−3.4 ps, τS3 ∼ 32.3−37.7 ps, and τS4 ∼ 232−485 ps), of which the longer ones (τS3, τS4) are well-consistent with experiments, whereas the ultrafast components (τS1, τS2) are unresolved. The observed time constant (τS3) within the ∼20−47 ps range arises from slow water molecules in the primary hydration layers of the host CDs and is slower for DIMEB than for βCD presumably due to longer-lived and stronger hydrogen bonds that water forms with the former host relative to the latter. Decomposition of the calculated solvation response of DMO has revealed that conformational fluctuations of the macrocyclic hosts give rise to the observed long-time relaxation component (τS4), which is much slower for the inclusion complexes with DIMEB than for those with βCD because of slower conformational dynamics of the former host than that of the latter.

1. INTRODUCTION In nanoscale confinement, water plays an essential role in a multitude of chemical and biological processes, and its unique properties stem from an extended hydrogen bonding network,1 which evolves continuously on a picosecond time scale, facilitating processes ranging from proton diffusion to protein folding.2,3 It was experimentally demonstrated that the structural and dynamical features of confined water are different from those exhibited by bulk water.4−11 Cyclodextrins (CDs) are used as model host systems for studying water in confined nanoscopic environments.12−14 Fleming et al.12 studied the dynamics of water in restricted environments by comparing the spectral dynamics of an optically excited probe molecule, coumarin 480, embedded in γ-cyclodextrin (γCD) with the spectral dynamics of the same probe molecule in bulk water. The spectral dynamics reflects the collective rearrangement of the solvating water and was found to be much slower within the cavity of γCD than in bulk water. Bhattacharyya et al.13 studied the solvation dynamics of coumarin 153 (C153) encapsulated within two β-cyclodextrins (βCD) of similar sizes, albeit different functionalization, and observed ultraslow relaxation components of few nanoseconds. Recently, extensive molecular dynamics (MD) simulations14 were carried out to explore the origin of these nanosecond relaxation components13 in the © 2018 American Chemical Society

solvation response of C153 entrapped within the hydrophobic cavities of β-cyclodextrins (βCD) in water. In this study Laria et al.14 were unable to detect any slow dynamic modes of water in the solvation response and ascribed the ultraslow dynamics of solvation of C153 to the gauche−trans interconversions in the primary hydroxyl chains of the βCDs, which is not directly connected to the optical excitation of the probe. This intramolecular motion in the βCD as the new source of ultraslow solvation dynamics has never been contemplated in the previous analyses.12,13 Recently, Corcelli et al.15,16 used MD simulations to calculate the equilibrium and nonequilibrium solvation responses to excitation of Hoechst 33258 bound to DNA and showed that DNA motion was responsible for the long-time relaxation component (∼20 ps) in the solvation dynamics. Therefore, study of solvation dynamics in host− guest inclusion complexes provides a particularly intriguing avenue of exploration of the origin of the slower relaxation component of few hundreds of picoseconds to few nanoseconds. Received: October 19, 2017 Revised: March 9, 2018 Published: March 12, 2018 3996

DOI: 10.1021/acs.jpcb.7b10376 J. Phys. Chem. B 2018, 122, 3996−4005

Article

The Journal of Physical Chemistry B

2. EXPERIMENTAL SECTION 2.1. Materials. 5-(4″-Dimethylaminophenyl)-2-(4′sulfophenyl)oxazole, sodium salt, was purchased from Molecular Probes Inc. Heptakis(2,6-di-O-methyl)-β-cyclodextrin (DIMEB) and β-cyclodextrin (βCD) were obtained from Sigma-Aldrich and used as received. The stock solutions of CDs were prepared in deionized water (from Millipore Milli-Q nanopure water system). For the measurements of UV−vis, steady-state, and time-resolved fluorescence decays, the final concentration of DMO was maintained at 5 μM and the concentrations of βCD and DIMEB were maintained at 750 and 250 μM, respectively. Quantum yields (Φ) of the dye were determined with respect to its solution in ethanol (Φ = 0.72) as a reference.23 All of the spectroscopic measurements were performed at 25 °C in cuvettes of 1 cm optical path. Steadystate absorption and fluorescence spectra were recorded on a PerkinElmer Lambda 35 UV−vis spectrometer and a PerkinElmer LS 55 spectrofluorimeter, respectively. Time-resolved fluorescence measurements were recorded with a commercial time-correlated single-photon counting (TCSPC) setup from Edinburgh Instruments (LifeSpec-ps) described elsewhere.23 The system is equipped with a 375 nm diode laser as the excitation source (PicoQuant LDH-P-C-375, 80 ps full width at half-maximum (FWHM)) and a microchannel plate photomultiplier (Hamamatsu R3809U-50) as the detector. The decays were analyzed using FAST software of Edinburgh Instruments. The time resolution of the equipment is ∼16 ps after instrument response function reconvolution. Time-resolved fluorescence intensity decays were fitted with the following multiexponential function

In the present study, we have used picosecond-resolved emission spectroscopy in conjunction with MD simulations to probe the solvation response of optically excited fluorophore 5(4″-dimethylaminophenyl)-2-(4′-sulfophenyl)oxazole, sodium salt (DMO), in the cavities of pure β-cyclodextrin (βCD) and heptakis(2,6-di-O-methyl)-β-cyclodextrin (DIMEB), (Scheme 1) with a limited number of co-included water Scheme 1. Molecular Structures of DMO, (a) βCD, and (b) DIMEB

n

I (t ) =

⎛

t⎞ ⎟ ⎝ τi ⎠

∑ ai exp⎜− i=1

(1)

where ai’s are amplitudes of the decay components with time constants of τi. The average excited-state fluorescence lifetime is given by the following equation: τ = ∑ni=1aiτi, where ∑ni=1ai = 1. The time-resolved fluorescence anisotropy decays (r(t)) were determined by measuring the parallel-polarized (IVV(t)) and perpendicularly polarized (IVH(t)) fluorescence transients using eq 2 as shown below. r (t ) =

IVV(t ) − GIVH(t ) IVV(t ) + 2GIVH(t )

(2)

The magnitude of G, the grating factor of the emission monochromator of the TCSPC system, is obtained using a long-tail matching technique.24 Time-resolved anisotropy decays, r(t), were fitted reasonably well with a monoexponential decay function (3) both in water and βCDs.

molecules. The CDs are toroidal-shaped molecules containing seven glucopyranose units (Scheme 1) with a hydrophobic cavity surrounded by a hydrophilic exterior comprising primary and secondary hydroxyl (OH) groups.17−19 These hydroxyl groups on the rims of a CD cavity may strongly perturb the structure and dynamics of water molecules surrounding the inclusion complex.20−22 The goals of the present study are to investigate the effects of increased confinement on the dynamics of solvation to the optical excitation of a chromophore in the nanoscopic domains of the CD cavities and to explore the origin of the slower relaxation, which have remained elusive, so far.

⎛ t ⎞ r(t ) = r0·exp⎜ − ⎟ ⎝ θ1 ⎠

(3)

where r0 is the initial anisotropy and θ1 is the rotational correlation time. Time-resolved emission spectra (TRES) were generated from a set of emission decays (at least 16 wavelengths) recorded at 10 nm intervals spanning the fluorescence spectrum using the “spectral reconstruction” method as described elsewhere.25 The time evolution of the peak wavenumbers, ν(t), in the TRES was fitted using a lognormal line-shape function 3997


Article


Four systems were simulated: βCD/DMO in ground and excited states and DIMEB/DMO in ground and excited states. All systems were initially equilibrated for 20 ns. After that, the production simulations of 200 ns were performed on each system. The trajectories were saved each 20 fs. After that, the electrostatic interaction energies between DMO and the rest of the system were computed for each trajectory frame using ground- and excited-state topologies of DMO. Solvation response functions C0(t) and C1(t) are defined below and were computed using the equilibrium time correlation function of fluctuations from the solvation energy differences as described elsewhere.15

I(ν , t ) = A exp{−ln 2(ln[1 + 2bi(ν − νi)/Δi ]/bi)2 } for α > 1 = 0 for α ≤ 1, where α = 2bi(ν − νi)/Δi

(4)

where A, νi, bi, and Δi are the peak height, peak wavenumber, asymmetry parameter, and width parameter, respectively. The normalized spectral shift correlation function or the solvent response function, C(t), was calculated according to C(t ) =

ν(t ) − ν(∞) ν(0) − ν(∞)

(5)

where ν(0), ν(t), and ν(∞) are the wavenumbers of the emission maxima at time 0, t, and ∞, respectively. The decay of C(t) is satisfactorily fitted by the following biexponential function C(t ) = a3 exp( −t /τS3) + a4 exp(−t /τS 4)

C0,1(t ) =

⟨δ ΔE(0)δ ΔE(t )⟩0,1 ⟨|δ ΔE|2 ⟩0,1

where δΔE(t) = ΔE(t) − ⟨ΔE⟩0,1. ΔE(t) = Eexcited(t) − Eground(t) is the difference of electrostatic energies of the dye in excited and ground states for the same system configuration observed at time t in the MD trajectory. The angular brackets represent an ensemble average over the equilibrium MD trajectory computed in either the ground (subscript 0) or the excited (subscript 1) state of the dye. GROMACS tools were used to compute the autocorrelation functions. The calculated solvation response functions exhibit multiexponential behavior, and the number of their exponential components was not known in advance. One of the most successful techniques of decomposing such complex signals into exponential components is the maximum entropy method (MEM).34,35 In this work, we used MEMfit software, which was previously used with great success in the analysis of multiexponential relaxation in nonequilibrium MD simulations.36 It allows finding the spectrum of exponential components of the signal automatically in completely modelfree manner. Root-mean-square deviations (RMSDs) of βCD and DIMEB were computed over first 50 ns of production trajectories using all atoms including hydrogens. Cyclodextrin molecules were structurally aligned to the conformation from the first frame, and the RMSD was computed using the Gromacs “msd” tool. Autocorrelation of RMSD was computed using the standard Gromacs analysis tool over the time window of 20 ns. Although RMSD is not the most precise method of comparing conformational mobility, it is known to work well in most cases where a rough estimate of conformational changes is sufficient.37 In our case, conformational changes are not too large to produce redundant results in RMSD measurements. We use this method only to visualize an existence of conformational changes in βCD and DIMEB in the nanosecond time scale.

(6)

where τS3 and τS4 and a3 and a4 are the solvent relaxation components and their corresponding amplitudes, respectively.

3. MOLECULAR DYNAMICS SIMULATIONS The initial configurations of the 1:1 inclusion complexes of βCD/DMO (complex-1) and DIMEB/DMO (complex-2) in aqueous medium were prepared as follows. The coordinates of the βCD molecule were taken from the corresponding crystal structure.26 The coordinates of DIMEB were obtained by replacing the hydroxyl (OH) groups of βCD by methoxy (OCH3) groups. The initial coordinates of DMO and the topologies of DMO, βCD, and DIMEB were generated by CHARMM-GUI.27 The structures of the DMO probe in ground and the first excited states were optimized using Gaussian 09 at the B3LYP/ 6-31G level of theory.28 After that, the electrostatic potential atomic charges were computed on optimized structures using the same level of theory. The charges of structurally equivalent atoms were averaged. Obtained charges were assigned to the draft topology of DMO, leading to two different topologies for ground and excited states of the probe. This approach where the topologies for ground and excited states differ only by atomic charges, whereas all bonded and van der Waals parameters remain the same, is commonly employed for solvation response simulations.15 The initial configurations of the complexes were prepared by fitting the center of the dye at the center of the CD molecules. The two complexes were then solvated in the cubic simulation boxes containing ∼1300 water molecules. A single Na+ ion was inserted randomly into the simulation box to balance the negative charge of DMO. System preparation was facilitated by Pteros molecular modeling library.29,30 All simulations were performed in the Gromacs 2016.1 package31 using the CHARMM36 force field. The TIP3P water model was used. All simulations were performed in NPT conditions with 300 K temperature and 1 bar isotropic pressure maintained by a v-rescale thermostat and a Berendsen barostat, respectively. The time step of 1 fs was used. The parameters for cutoffs recommended for the CHARMM36 force field were used as suggested in the CHARMM-GUI output. The minimum image convention32 was employed to calculate the short-range Lennard-Jones interactions using a spherical cutoff distance of 12 Å with a switch distance of 10 Å. The long-range electrostatic interactions were calculated using the particle-mesh-Ewald method.33

4. RESULTS AND DISCUSSION 4.1. Steady-State Emission Spectra. DMO is known to display strong polarity-sensitive fluorescence. For example, the steady-state emission peak maximum of the dye is significantly red-shifted from 420 nm in n-hexane to 582 nm in water, which is attributable to a highly polar excited state originating from an intramolecular charge transfer process.38 Addition of βCD and DIMEB to an aqueous solution of DMO causes blue shifts of the emission peak maximum by 25 and 38 nm, respectively (Figure 1 and Table 1) along with the significant enhancement of the fluorescence quantum yield in comparison to that for water. These observations are in good agreement with those in 3998


Article


less polar and more hydrophobic nature of the cavity of DIMEB relative to that of βCD. 4.2. Formation of Dye/CD Inclusion Complexes. The encapsulation of DMO into the cavities of βCD and DIMEB leads to the formation of host (CD)/guest (dye) inclusion complexes. Because of the lower concentrations of CD used (≤0.75 mM) in the present study, it is reasonable to assume the formation of 1:1 CD/dye complexes.40,41 This is well corroborated from the excellent fitting (correlation coefficients R2 ≥ 0.996) of the experimental data of emission intensity at different CD concentrations (Figure 2) by the following equation based on the formation of 1:1 CD/dye complexes.42 I=

Table 1. Steady-State Photophysical Parameters and Parameters for dye/βCD Complexesa βCD DIMEB water

λem./nm

Φ

K1/M−1

353 354 340

557 544 582

0.18 0.53 0.04

5923 21 330

(7)

where I0 and I1 denote the fluorescence intensity in pure water and in the complex, respectively, and I is the fluorescence intensity at a given CD concentration. K1 is the association constant for the formation of the 1:1 CD/dye complex. From the nonlinear regression analysis of the experimental data, the association constants for the formation of the 1:1 βCD/DMO and DIMEB/DMO complexes are found to be 5923 and 21 330 M−1, respectively, indicating much stronger host−guest binding interaction for the latter than for the former. Given the large values of the association constants, for a CD/dye ratio of 150:1 or 50:1, all of the DMO molecules are present in the 1:1 complexes with βCD and DIMEB, respectively. 4.3. Time-Resolved Fluorescence Anisotropy Studies. Time-resolved fluorescence anisotropy decays of DMO/βCD and DMO/DIMEB systems (Figure 3) could be fitted reasonably well with a monoexponential function (eq 3), yielding rotational correlation times (θ1) of 496 and 632 ps, respectively. Because the rotational correlation time (θ1) is directly proportional to the hydrodynamic volume of the fluorophore,24,43 nearly 3-fold increase in the value of θ1 for the dye in the presence of the host CDs from that in water23 (θ1 ∼ 167 ps) confirms the formation of inclusion complexes,44 which is consistent with the steady-state spectral data. 4.4. Time-Resolved Emission Spectra: Solvation Dynamics in Inclusion Complexes of CDs. Figure 4

Figure 1. Steady-state fluorescence spectra of DMO in water and host cyclodextrins. The excitation wavelength is 370 nm. The concentrations of DMO, βCD, and DIMEB are 5, 750, and 250 μM, respectively.

λabs./nm

I0 + I1K1[CD] 1 + K1[CD]

λabs., λem, and Φ are the absorption peak maximum, emission peak maximum, and fluorescence quantum yield, respectively. K1 is the association constant for the formation of 1:1 dye/CD complexes. a

the bile salt aggregates39 and confirm the encapsulation of the probe molecules into the cavities of βCD and DIMEB, where the local environments of DMO are less polar and more hydrophobic than those in water. Furthermore, the larger blue shift and higher quantum yield in DIMEB are indicative of a

Figure 2. Plot of fluorescence intensity I vs varying concentrations of DIMEB and βCD, for DMO. Open circles are the experimental data points, and full lines (blue and red) are the nonlinear regression fits to the experimental data points following eq 7. 3999


Article


Figure 4. Picosecond-resolved emission transients of DMO in βCD and DIMEB. Solid lines (colored) represent fits to raw fluorescence decay data (shown in circles) following eq 1.

Figure 3. Time-resolved anisotropy decays of DMO in CDs. Solid lines (red line, blue line) represent fits to raw anisotropy decay data (shown in circles) following eq 3. Solid lines in green (green line) display reference lines corresponding to anisotropy value zero.

displays wavelength-dependent emission transients of DMO in βCD and DIMEB at three characteristic wavelengths, from the blue edge to the red edge of the steady-state emission spectrum. All transients exhibit a general trend of decay at the blue side and initial rise at the red side, which is typical of solvation dynamics45 due to reorganization of the surrounding water molecules around the excited-state dipole of the fluorophore. As the chromophore excited state progresses to lower energy due to solvation, a time-dependent red shift of the fluorescence spectra (Figure 5) in both the CDs is observed. The decays of the reconstructed solvent response functions, C(t), corresponding to the time scale of this spectral relaxation process are displayed in Figure 6. The solvation correlation functions obtained are the response of local environments around DMO to its sudden change of dipole moment upon excitation. Under this perturbation, the response, in principle, can result from both the surrounding water molecules and the conformational flexibility of host CDs.15,16 The role of the water dipolar reorientation and relaxation of the dye environment in the spectral relaxation is elucidated from time-resolved anisotropy decay. For the host−guest complexes, the anisotropy decays to 0 on a much longer time scale (6 ns, Figure 3) compared to the time scale of solvent relaxation (0.8−1.8 ns, Figure 6). These results indicate that the orientational motion of probe DMO is

Figure 5. Time-resolved emission spectra (TRES) of DMO in host cyclodextrins. Solid lines (colored) represent log-normal fits to TRES (shown by colored circles) following eq 4.

4000


Article


complexes along with their initial structures are displayed in Figure 7. We calculated the ground- and excited-state equilibrium correlation functions, C0(t) (Figure 8) and C1(t), respectively,50 for simulations of DMO in water and in complexes with host CDs and extracted the time scales of solvation response by fitting C0,1(t) to a multiexponential function. The results of our calculations along with the experimental results obtained from time-resolved fluorescence measurements are shown in Table 2. The relaxation components associated with the calculated equilibrium solvation correlation functions were extracted using the maximum entropy method (MEM),34,35 and the standard errors of relaxation times are estimated from the full width at half-maximum (FWHM) of the peaks generated by MEM. In the case of the experimental solvent response function, C(t), the solvent relaxation components were extracted by fitting the data with a biexponential function (eq 6) in a nonlinear leastsquares curve fitting routine based on the Levenberg− Marquardt51,52 algorithm implemented in Origin. The standard errors to the experimentally obtained relaxation times were generated from the nonlinear least-squares curve fitting routine and are displayed in Table 2. The solvation response function computed for DMO in aqueous solution is characterized by an ultrafast inertial relaxation component (∼10 fs) due to librational motions53 of water molecules along with two slower relaxation components (∼0.4 and 1.3 ps), which is typical of solvation of a molecular probe in bulk water.54 The ultrafast inertial response is responsible for nearly 50% of the total solvation response in water on a time scale of ∼20 fs.53,55 In inclusion complexes with host CDs, the ultrafast inertial response is absent (Table 2) and the calculated solvation response of DMO is characterized by four relaxation components. Among them, the relaxation components on the time scale of sub-picosecond to few picoseconds (τS1 ∼ 0.46− 0.48 ps and τS2 ∼ 3.2−3.6 ps) correspond to diffusive motions in water53−55 and account for nearly 60% of the total solvation response in the dye/CD complexes. This is quite reasonable given that nearly half of the chromophore (DMO) is protruded out of the CD cavities (Figure 7) and into their hydration layers. As a result, relatively mobile waters15,16,54,56 in the hydration layer or water molecules from the bulk in the vicinity of the chromophore and the host CDs make significant contribution to the collective solvation response observed for DMO in the host−guest inclusion complexes. These relaxation components (τS1 and τS2) are unresolved because of the limited time resolution of our experimental setup (∼16 ps), although similar time constants have been resolved earlier13,52 with femtosecond

Figure 6. Decay of the solvation correlation function, C(t), with time. Solid lines (colored) represent fits to the decay of C(t) (represented by colored circles) following eq 6.

significantly hindered in the cavities of βCD and DIMEB, indicating that the probe remains mostly restricted in motion during the time scale of solvation. The bimodality of the observed solvation response of DMO in its inclusion complexes with the host CDs (Table 2) reflects two types of relaxation of the dye environment. On the basis of recent MD simulations of Bagchi et al.,46,47 the faster component (τS3 ∼ 20−47 ps) may be attributed to slow orientational and translational motion of water molecules in the primary hydration layers of βCD and DIMEB. This component is slower for DIMEB than for βCD presumably due to longlived and stronger hydrogen bonds that water forms with the former host relative to the latter in the primary hydration layer.47 The methoxy groups of DIMEB likely form stronger Hbonds with water molecules than the hydroxyl groups of βCD, which leads to a slowdown in the reorientational dynamics of water molecules due to the transition state H-bond effect.48,49 In the case of the significantly longer relaxation component of few hundreds of picoseconds (τS4 ∼ 174−517 ps, Table 2), conformational fluctuation of macrocyclic hosts may be involved given the works of Corcelli et al.15,16 To explore the origin of this slow relaxation on the time scale of few hundreds of picoseconds, we have carried out MD simulations to calculate directly the equilibrium solvation response of DMO incorporated in the host cavities of βCD and DIMEB. 4.5. Total Solvation Response from MD Simulations. The simulated configurations of the 1:1 host−guest inclusion

Table 2. Solvation Correlation Times for DMO in Water and in Complexes with host CDs from Experimental Solvation Responseb and Calculated Equilibrium Correlation Functions, C0,1(t)a water βCD

DIMEB

C0(t) expt. C0(t) C1(t) expt. C0(t) C1(t)

τS1/ps

a1

τS2/ps

a2

0.010 ± (0.005)

0.51

0.41 ± (0.13)

0.20

0.47± (0.2) 0.48 ± (0.22)

0.39 0.35

3.4 ± (1.5) 3.6 ± (1.6)

0.21 0.32

0.46 ± (0.12) 0.48 ± (0.15)

0.35 0.41

3.2 ± (1.9) 3.2 ± (1.8)

0.20 0.18

τS3/ps 1.33 19.8 37.7 45.0 46.9 32.3 22.6

± ± ± ± ± ± ±

(0.80) (0.7) (6.8) (7.1) (1.3) (4.1) (5.2)

a3 0.29 0.89 0.23 0.23 0.81 0.21 0.17

τS4/ps

a4

± ± ± ± ± ±

0.11 0.17 0.10 0.19 0.24 0.24

174.0 232.0 254.3 517.4 484.8 569.7

(40.0) (46.3) (37.2) (44.0) (85.7) (92.0)

C(t)/C0,1 = ∑ni=1ai e−t/τns ; ∑ni=1ai = 1, τSn and an are correlation times and their amplitudes, respectively, obtained from multiexponential fitting. Statistical errors of the measured time scales are shown within parentheses. bSolvent response function, C(t), constructed from time-resolved data. a

4001


Article


Figure 7. Initial (left column) and representative simulated configurations (right column) of the 1:1 inclusion complexes of βCD/DMO (upper panel) and DIMEB/DMO (lower panel).

component (τS4) of few hundreds of picoseconds becomes evident from the component-based decomposition of the total calculated response into contributions from the individual water and sodium ions and macrocyclic hosts. Decomposition of C0(t) reveals that the component due to water and sodium ions decays rapidly (Figure 8), whereas the decay profile of the component due to βCD or DIMEB is almost identical to the total solvation response of the inclusion complex with βCD or DIMEB. This demonstrates that the long-time relaxation constant (τS4) is associated exclusively with the conformational dynamics of the macrocyclic hosts. This derives further support from decomposition of C1(t) (Figure S1) into individual response functions.50 To visualize the existence of conformational dynamics of CDs, we have also computed the RMSDs of the host−guest inclusion complexes (Figure 9) and the time evolution of their autocorrelations (Figure 10). The fluctuations of the structures of CDs that occur on the time scale of hundreds of picoseconds and nanoseconds are evident. The autocorrelation of RMSD fluctuations converges to zero (Figure 10) on a time scale of few nanoseconds, confirming the existence of long-time dynamics of both βCD and DIMEB. Moreover, the decay of the autocorrelation is much slower for DIMEB relative to that for βCD, which is consistent with the slower dynamics of the conformational fluctuation of the former host than that of the latter. The slower dynamics of the conformational fluctuation of DIMEB relative to that of βCD gives rise to a much slower collective solvation response of DMO in its inclusion complex with the former than that in the latter.

Figure 8. Ground-state equilibrium correlation functions, C0(t), calculated for fluorescent probe DMO encapsulated in inclusion complexes with βCD (black), DIMEB (gray), and free in water (green). Decomposition of C0(t) into individual contributions from water and Na+ ion for both complexes (blue line, light blue line), βCD (red line), DIMEB (pink line).

resolution. The other two calculated relaxation components (τS3 ∼ 32−37 ps and τS4 ∼ 232−485 ps) are well-consistent with the experimental results. Whereas the relaxation dynamics on the time scale of few tens of picoseconds (τS3) is due to slow water molecules in the primary hydration layers46,47 of βCD and DIMEB, the molecular origin of the much slower 4002


Article


evolution of the autocorrelation of RMSD fluctuations displays slower dynamics of conformational fluctuations of DIMEB than that of βCD, which is consistent with the much slower relaxation component (τS3 ∼ 517 ps) observed for the inclusion complexes with DIMEB than that for those with βCD (τS3 ∼ 174 ps). In the present study, host CDs effectively “solvate” the probe along with water and exhibit a distinctly different time dependence of reorganization compared to that for water.

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.jpcb.7b10376. Profile of the calculated excited-state equilibrium correlation functions, C1(t), for chromophore DMO in water and inclusion complexes with βCD and DIMEB; decomposition of C1(t) into individual contributions from water, Na+ ion, and host CDs (Figure S1) (PDF)

Figure 9. Time evolution of the RMSDs for all of the atoms of the complex with βCD and DIMEB with respect to their initial structures.

■

AUTHOR INFORMATION

Corresponding Author

*Tel: 91 0 98361 54202. ORCID

Madhurima Jana: 0000-0002-6567-4490 Ranjan Das: 0000-0003-4171-0941 Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS Generous financial support to R.D. from CSIR, Government of India, vide CSIR project 01(2445)/10/EMR-II is gratefully acknowledged. M.J. acknowledges the Department of Science and Technology, Government of India (SB/FT/CS-065/ 2012), for grant support for creating high-performance computing facility. S.S. thanks NIT, Rourkela, for providing a scholarship. Sincere thanks are to Priya for her help with TCSPC measurements. S.Y. was supported by the NATO Science for Peace and Security programme under the project SPS 985291.

Figure 10. Time evolution of autocorrelation of RMSDs for the complexes with βCD and DIMEB.

5. CONCLUSIONS Time-resolved Stokes shift experiments have revealed that the collective solvation response to excitation of fluorescent probe DMO is bimodal and remarkably slower in inclusion complexes with DIMEB (τS3 ∼ 47 ps, τS4 ∼ 517 ps) than in those with βCD (τS3 ∼ 20 ps, τS4 ∼ 174 ps). These observations are wellconsistent with the equilibrium solvation response calculated from MD simulations. However, because of the limited time resolution, ultrafast relaxation components (τS1 ∼ 0.46 ps, τS2 ∼ 3.2 ps) could not be resolved, as predicted from the equilibrium solvation response calculations. These time constants likely originate from relatively mobile waters that solvate the probe itself as well as the host CDs.15,16 The relaxation component of few tens of picoseconds (τS3 ∼ 20−47 ps) is ascribed to slow orientational and translational motion of water molecules in the primary hydration layers of βCD and DIMEB on the basis of recent MD simulations.46,47 This component is slower for DIMEB than for βCD presumably due to long-lived and stronger hydrogen bonds that water forms with the former host relative to the latter in the primary hydration layer.47 The equilibrium solvation response calculations clearly demonstrate that the long-time relaxation dynamics (τS4 ∼ 174−517 ps) on the time scale of hundreds of picoseconds is associated with the conformational dynamics of the host CDs. Furthermore, time

■

REFERENCES

(1) Steinel, T.; Asbury, J. B.; Zheng, J.; Fayer, M. D. Watching hydrogen bonds break: a transient absorption study of water. J. Phys. Chem. A 2004, 108, 10957−10964. (2) Laage, D.; Hynes, J. T. A molecular jump mechanism of water reorientation. Science 2006, 311, 832−835. (3) Laage, D.; Hynes, J. T. On the molecular mechanism of water reorientation. J. Phys. Chem. B 2008, 112, 14230−14242. (4) Bellissent-Funel, M. C. Status of experiments probing the dynamics of water in confinement. Eur. Phys. J. E: Soft Matter Biol. Phys. 2003, 12, 83−92. (5) Tsukahara, T.; Hibara, A.; Ikeda, Y.; Kitamori, T. NMR Study of Water Molecules Confined in Extended Nanospaces. Angew. Chem., Int. Ed. 2007, 46, 1180−1183. (6) Pal, S. K.; Zhao, L.; Zewail, A. H. Water at DNA surfaces: Ultrafast dynamics in minor groove recognition. Proc. Natl. Acad. Sci. U.S.A. 2003, 100, 8113−8118. (7) Moilanen, D. E.; Levinger, N. E.; Spry, D. B.; Fayer, M. D. Confinement or the Nature of the Interface? Dynamics of Nanoscopic Water. J. Am. Chem. Soc. 2007, 129, 14311−14318. (8) Moilanen, D. E.; Piletic, I. R.; Fayer, M. D. Water Dynamics in Nafion Fuel Cell Membranes: The Effects of Confinement and 4003


Article

The Journal of Physical Chemistry B Structural Changes on the Hydrogen Bond Network. J. Phys. Chem. C 2007, 111, 8884−8891. (9) Moilanen, D. E.; Spry, D. B.; Fayer, M. D. Water Dynamics and Proton Transfer in Nafion Fuel Cell Membranes. Langmuir 2008, 24, 3690−3698. (10) Moilanen, D. E.; Piletic, I. R.; Fayer, M. D. Tracking Water’s Response to Structural Changes in Nafion Membranes. J. Phys. Chem. A 2006, 110, 9084−9088. (11) Dokter, A. M.; Woutersen, S.; Bakker, H. J. Inhomogeneous dynamics in confined water nanodroplets. Proc. Natl. Acad. Sci. U.S.A. 2006, 103, 15355−15358. (12) Vajda, Š.; Jimenez, R.; Rosenthal, S. J.; Fidler, V.; Flemimg, G. R.; Castner, E. W., Jr. Femtosecond to Nanosecond Solvation Dynamics in Pure Water and inside the γ-Cyclodextrin Cavity. J. Chem. Soc., Faraday Trans. 1995, 91, 867−873. (13) Sen, P.; Roy, D.; Mondal, S. K.; Sahu, K.; Ghosh, S.; Bhattacharyya, K. Fluorescence Anisotropy Decay and Solvation Dynamics in a Nanocavity: Coumarin 153 in Methyl β-Cyclodextrins. J. Phys. Chem. A 2005, 109, 9716−9722. (14) Rodriguez, J.; Marti, J.; Guardia, E.; Laria, D. Exploring the Picosecond Time Domain of the Solvation Dynamics of Coumarin 153 within β-Cyclodextrins. J. Phys. Chem. B 2008, 112, 8990−8998. (15) 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. (16) Furse, K. E.; Corcelli, S. A. Molecular Dynamics Simulations of DNA Solvation Dynamics. J. Phys. Chem. Lett. 2010, 1, 1813−1820. (17) Uekama, K.; Hirayama, F.; Irie, T. Cyclodextrin Drug Carrier Systems. Chem. Rev. 1998, 98, 2045−2076. (18) Rekharsky, M. V.; Inoue, Y. Complexation Thermodynamics of Cyclodextrins. Chem. Rev. 1998, 98, 1875−1918. (19) Szejtli, J. Introduction and General Overview of Cyclodextrin Chemistry. Chem. Rev. 1998, 98, 1743−1754. (20) Starikov, E. B.; Brasicke, K.; Knapp, E. W.; Saenger, W. Negative solubility coefficient of methylated cyclodextrins in water: A theoretical study. Chem. Phys. Lett. 2001, 336, 504−510. (21) Heine, T.; Dos Santos, H. F.; Patchkovskii, S.; Duarte, H. A. Structure and Dynamics of β-Cyclodextrin in Aqueous Solution at the Density-Functional Tight Binding Level. J. Phys. Chem. A 2007, 111, 5648−5654. (22) Hansen, J. E.; Pines, E.; Fleming, G. R. Excited-state proton transfer of protonated 1-Aminopyrene complexed with β-cyclodextrin. J. Phys. Chem. 1992, 96, 6904−6910. (23) Maiti, J.; Sarkar, Y.; Parui, P. P.; Chakraborty, S.; Biswas, S.; Das, R. Photophysical study of a charge transfer oxazole dye in micelles: Role of surfactant headgroups. J. Lumin. 2015, 163, 21−27. (24) Lakowicz, J. R. Principles of Fluorescence Spectroscopy; Kluwer Academic/Plenum: New York, 2006. (25) Horng, M. L.; Gardecki, J. A.; Papazyan, A.; Maroncelli, M. Subpicosecond Measurements of Polar Solvation Dynamics: Coumarin 153 Revisited. J. Phys. Chem. 1995, 99, 17311−17337. (26) Sharff, A. J.; Rodseth, L. E.; Quiocho, F. A. Refined 1.8-Å Structure Reveals the Mode of Binding of β-Cyclodextrin to the Maltodextrin Binding Protein. Biochemistry 1993, 32, 10553−10559. (27) Jo, S.; Kim, T.; Iyer, V.; Im, W. CHARMM-GUI: A web-based graphical user interface for CHARMM. J. Comput. Chem. 2008, 29, 1859−1865. (28) Frisch, M. J. Gaussian 09; Gaussian Inc.: Wallingford, CT, 2009. (29) Yesylevskyy, S. O. Pteros: Fast and easy to use open-source C++ library for molecular analysis. J. Comput. Chem. 2012, 33, 1632−1636. (30) Yesylevskyy, S. O. Pteros 2.0: Evolution of the fast parallel molecular analysis library for C++ and python. J. Comput. Chem. 2015, 36, 1480−1488. (31) Abraham, M. J.; Murtola, T.; Schulz, R.; Szilard, P.; Smith, J. C.; Hess, B.; Lindahl, E. GROMACS: High performance molecular simulations through multi-level parallelism from laptops to supercomputers. SoftwareX 2015, 1−2, 19−25. (32) Allen, M. P.; Tildesley, D. J. Computer Simulations of Liquids; Clarrendon Press: Oxford, 1987.

(33) 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. (34) Skilling, J. Classic Maximum Entropy; Kluwer Academic: Norwell, MA, 1989. (35) Steinbach, P. J.; Ansari, A.; Berendzen, J.; Braunstein, D.; Chu, K.; Cowen, B. R.; Ehrenstein, D.; Frauenfelder, H.; Johnson, J. B.; et al. Ligand binding to heme proteins: connection between dynamics and function. Biochemistry 1991, 30, 3988−4001. (36) Yesylevskyy, S. O.; Hushcha, T. O. Conformational relaxations of human serum albumin studied by molecular dynamics simulations with pressure jumps. Biopolym. Cell 2012, 28, 486−492. (37) Kufareva, I.; Abagyan, R. Methods of protein structure comparison. Methods Mol. Biol. 2012, 857, 231−257. (38) Pal, K.; Chandra, F.; Mallick, S.; Koner, A. L. Effect of solvents and cyclodextrin complexation on acid-base and photophysical properties of dapoxyl dye. J. Photochem. Photobiol., A 2015, 306, 47−54. (39) Maiti, J.; Kalyani, V.; Biswas, S.; Rodriguez-Prieto, F.; Mosquera, M.; Das, R. Slow solvation dynamics in supramolecular systems based on bile salts: Role of structural rigidity of bile salt aggregates. J. Photochem. Photobiol., A 2017, 346, 17−23. (40) Kusumoto, Y. A. Spectrofluorimetric method for determining the association constants of pyrene with cyclodextrins based on polarity variation. Chem. Phys. Lett. 1987, 136, 535−538. (41) Roberts, E. L.; Chou, P. T.; Alexander, T. A.; Agbaria, R. A.; Warner, I. M. Effects of Organized Media on the Excited-State Intramolecular Proton Transfer of 10-Hydroxybenzo[h]quinoline. J. Phys. Chem. 1995, 99, 5431−5437. (42) Nigam, S.; Durocher, G. Spectral and Photophysical Studies of Inclusion Complexes of Some Neutral 3H-Indoles and Their Cations and Anions with β-Cyclodextrin. J. Phys. Chem. 1996, 100, 7135−7142. (43) O’Connor, D. V.; Phillips, D. Time Correlated Single Photon Counting; Academic Press: New York, 1984. (44) Shaikh, M.; Mohanty, J.; Sundararajan, M.; Bhasikuttan, A. C.; Pal, H. Supramolecular Host-Guest Interactions of Oxazine-1 Dye with β- and γ-Cyclodextrins: A Photophysical and Quantum Chemical Study. J. Phys. Chem. B 2012, 116, 12450−12459. (45) Singh, P.; Choudhury, S.; Singha, S.; Jun, Y.; Chakraborty, S.; Sengupta, J.; Das, R.; Ahn, K.-H.; Pal, S. K. A sensitive fluorescent probe for the polar solvation dynamics at protein-surfactant interfaces. Phys. Chem. Chem. Phys. 2017, 19, 12237−12245. (46) Mondal, S.; Mukherjee, S.; Bagchi, B. Origin of diverse time scales in the protein hydration layer solvation dynamics: A simulation study. J. Chem. Phys. 2017, 147, 154901−154911. (47) Mukherjee, S.; Mondal, S.; Bagchi, B. Distiguishing dynamical features of water inside protein hydration layer: Distribution reveals what is hidden behind the average. J. Chem. Phys. 2017, 147, 024901− 024912. (48) Fogarty, A. C.; Laage, D. Water dynamics in protein hydration shells: The molecular origin of the dynamical perturbation. J. Phys. Chem. B 2014, 118, 7715−7729. (49) Sterpone, F.; Stirnemann, G.; Hynes, J. T.; Laage, D. Water hydrogen-bond dynamics around amino acids: The key role of hydrophilic hydrogen-bond acceptor groups. J. Phys. Chem. B 2010, 114, 2083−2089. (50) Supporting Information file. (51) Levenberg, K. A Method for the Solution of Certain Non-Linear Problems in Least Squares. Q. Appl. Math. 1944, 2, 164−168. (52) Marquardt, D. An Algorithm for Least-Squares Estimation of Nonlinear Parameters. J. Soc. Ind. Appl. Math. 1963, 11, 431−441. (53) Jimenez, R.; Fleming, G. R.; Kumar, P. V.; Maroncelli, M. Femtosecond solvation dynamics of water. Nature 1994, 369, 471− 473. (54) Pal, S. K.; Peon, J.; Zewail, A. H. Biological water at the protein surface: Dynamical solvation probed directly with femtosecond resolution. Proc. Natl. Acad. Sci. U.S.A. 2002, 99, 1763−1768. 4004


Article

The Journal of Physical Chemistry B (55) Kumar, P. V.; Maroncelli, M. Polar solvation dynamics of polyatomic solutes: Simulation studies in acetonitrile and methanol. J. Chem. Phys. 1995, 103, 3038−3060. (56) Peon, J.; Pal, S. K.; Zewail, A. H. Hydration at the surface of the protein Monellin: Dynamics with femtosecond resolution. Proc. Natl. Acad. Sci. U.S.A. 2002, 99, 10964−10969.

4005


Picosecond Solvation Dynamics in Nanoconfinement: Role of Water

Recommend Documents