Proton transfer reactions of 4'-chloro substituted 3-hydroxyflavone in solvents and aqueous micelle solutions.

PCCP View Article Online

Published on 10 March 2014. Downloaded by Northern Illinois University on 03/09/2014 23:06:18.

PAPER

Cite this: Phys. Chem. Chem. Phys., 2014, 16, 8594

View Journal | View Issue

Proton transfer reactions of 4 0 -chloro substituted 3-hydroxyflavone in solvents and aqueous micelle solutions Deborin Ghosh,a Anup Kumar Pradhan,a Samiran Mondal,b N. A. Begumb and Debabrata Mandal*a Flavonol 4 0 -chloro,3-hydroxyflavone (Cl-3HF) has been investigated in solvents of varying polarity and hydrogen-bonding capacity as well as in aqueous micelle solutions. Quantum chemical calculations indicate that although the Cl-atom at the 4 0 -position of the 2-phenyl ring weakly perturbs the electron distribution of the parent 3-hydroxyflavone, the nuclear framework remains largely intact, and excited state intra-molecular proton-transfer (ESIPT) is feasible. The ESIPT process in both polar solvents and micelles was found to be fast and irreversible, with remarkably long time-constants of several tens of picoseconds. This dramatic inhibition of the ESIPT rate (which is intrinsically a sub-picosecond event) could be rationalized in terms of the emergence of complexes between the solvent and the enol form of Cl-3HF, whose dynamics is coupled to the relatively slow dynamics of inter-molecular hydrogen bonds. In the micelle solutions, spectroscopic data establish that the guest Cl-3HF molecules localized

Received 25th May 2013, Accepted 5th March 2014

almost exclusively at the polar exterior shell, where they experienced a nearly uniform local environment similar to that in moderately polar solvents. Thus, the Cl-3HF molecules tend to avoid the non-polar

DOI: 10.1039/c3cp52209a

core of the micelles, in spite of being strongly hydrophobic themselves. This apparently unusual observation is explained by the formation of inter-molecularly hydrogen-bonded complexes between

www.rsc.org/pccp

the guest Cl-3HF and the water molecules tethered to the polar shells of the micelles.

1. Introduction 3-Hydroxyflavone (3HF) has long been regarded as a model compound for the study of excited state intra-molecular proton transfer reactions (ESIPT). The proximity of the ring carbonyl and the 3-hydroxy group in the enol-like 3HF facilitates formation of an intra-molecular H-bond, which in turn is believed to promote proton-transfer in the excited state, creating a tautomer (Fig. 1a). Usually both the excited enol E* and excited tautomer T* are fluorescent, leading to dual fluorescence of 3HF and its derivatives. However, in a H-bonding solvent, the situation may become immensely complicated. Such a solvent may be represented by the general molecular formula H–J, where a H atom is directly bonded to a sufficiently electronegative atom in the J moiety. Thus H–J can operate either as a H-bond donating acid through the H atom (Fig. 1b) or as a H-bond accepting base by virtue of one or more electronegative

a

Department of Chemistry, University College of Science & Technology, University of Calcutta, 92, A.P.C. Road, Kolkata 700 009, India. E-mail: [email protected] b Bio-Organic Chemistry Lab, Department of Chemistry, Visva-Bharati University, Santiniketan 731 235, India

8594 | Phys. Chem. Chem. Phys., 2014, 16, 8594--8607

atoms in the J moiety (Fig. 1c) or as a simultaneous H-bond donor and acceptor (Fig. 1d). In the case of protic solvents like alcohol and water, J = OR and OH, respectively, while for nonprotics like DMF and DMSO which are incapable of H-bond donation, J = the entire molecule. As a result of inter-molecular H-bond formation between 3HF enol and H-bonding solvents, enol–solvent complexes with a variety of structures may materialize, some of which are illustrated in Fig. 1.1–3 In some of these solvated complexes, e.g., [E-solv]00 , [E-solv] 0 0 0 and [E-solv] 0 0 0 0 , the 3-OH group engages in H-bonding, which may cause it to be polarized to the point of dissociation, generating the 3HF anion. These putative structures offer valuable clues for elucidating the 3HF photophysics in solvents where inter-molecular H-bonding interferes with ESIPT.1–6 Since the proton-transfer event is associated with a substantial realignment of electron density, the nature and position of substituents on the 3HF skeleton might significantly modulate the photophysics of 3HF derivatives. So far, the vast majority of the work on 3HF derivatives has focussed mainly on those with an electron-donating substituent at the 4 0 -position of the 2-phenyl ring, especially a dialkylamino substituent (Fig. 2a). Here, intra-molecular charge transfer was found to interfere with the proton-transfer event.6–12 Keeping these in mind,

This journal is © the Owner Societies 2014

View Article Online

PCCP


Paper

Fig. 1 ESIPT and other possible proton transfer pathways of 3-hydroxyflavone in: (a) intra-molecularly hydrogen-bonded enol, (b–d) several different types of solvated complexes in hydrogen-bonding solvents.

we undertook to probe the effect of a halogen atom as a substituent in the 4 0 -position of the 2-phenyl ring. Here we present our findings on 4 0 -chloro,3-hydroxyflavone (Fig. 2b), which we will henceforth refer to as Cl-3HF in this article. The first part of our work consists of quantum chemical calculations. In order to understand the effect of the 4 0 -Cl atom on the photophysics, we identified the enol, the tautomer and


the transition state involved in proton transfer in both the ground and excited state energy surfaces and optimized their energy levels and molecular geometries. The molecular orbitals and electronic transition were also determined. These calculations were done in vacuum, i.e., in the isolated molecule state. Next, we extended our calculations to methanol, which is a typically good H-bonding solvent. Here, we identified the most

Phys. Chem. Chem. Phys., 2014, 16, 8594--8607 | 8595

View Article Online


PCCP

Paper

aggregation of surfactant molecules in aqueous solution, is usually spherical in shape, with an average diameter within few nanometers. It consists of a non-polar interior core enveloped within a polar exterior shell, and thus embodies heterogeneity in the nm length-scale. In many ways, it resembles lipid bilayers,13 which have a non-polar interior enclosed within a polar outer surface (Fig. 3). Last but not least, ESIPT-based organic systems have been intensively studied for potential technological applications in advanced optoelectronic devices including photoprotectors, photomemories, photochromic products, photosensors, etc.14,15 Keeping this in mind, our present study attempts to focus on some basic aspects of ESIPT.

2. Experimental 2.1.

Fig. 2 Structures of 3-hydroxyflavone derivatives: (a) 4 0 -N,N-dimethylamino,3-hydroxyflavone, and (b) 4 0 -chloro,3-hydroxyflavone.

stable 1 : 1 complexes between methanol and the enol, tautomer and anionic forms of Cl-3HF. Spectral transitions for these complexes were then calculated and compared with real experimental values in an attempt to determine the most probable structures of solvated Cl-3HF in methanol. We also tried to calculate the spectral transitions of Cl-3HF residing in aqueous micelles, following the clues provided by our experimental work, described below. The second part of our work consists of spectroscopic experiments on Cl-3HF in a series of solvents with varying H-bonding capacity and polarity to test the effect of the properties of solvents on the photophysics. The experiments were further extended to include aqueous micelles as media. A micelle, formed by the

Fig. 3 Schematic representation of micelle formation from a surfactant monomer.


Quantum chemical calculations

All theoretical calculations were performed using the Gaussian 09 program.16 Geometries of the enol (E) and tautomeric (T) forms of the Cl-3HF molecule in the ground state (S0) were optimised under vacuum by the density functional theory (DFT) using the hybrid functional B3LYP with a 6-31G(d,p) basis set. The calculation of transition energies and oscillator strengths as well as the geometry optimisation in the first singlet excited state (S1) were done by the time-dependent density functional theory (TDDFT) using the same functional with the same basis set. The transition states for the proton transfer process which connect both species in S0 as well as S1 states were detected and optimized by using the synchronous transit-guided quasiNewton QST2 method using an empirical estimate of Hessian and suitable starting structures. This method was chosen since it is known to converge efficiently to the actual transition state structure.17,18 Highest occupied molecular orbital (HOMO) and lowest unoccupied molecular orbital (LUMO) diagrams were also constructed on the optimized structures. Ground state potential energy surfaces for proton transfer were subsequently generated using the intrinsic reaction co-ordinate (IRC) method, where the optimized geometry of the S0 transition state between the E and T states (obtained by the QST2 method described above) was specified as the initial input. Computational studies were then carried out on Cl-3HF in the polar protic solvent methanol. Both un-solvated and solvated structures were considered. We first calculated the electron density distribution for the un-solvated Cl-3HF molecule immersed in methanol taken as a uniform dielectric medium. From these results, we formulated the solvated structures, i.e., 1 : 1 H-bonded complexes between Cl-3HF and methanol. All the possible structures were optimised by using the same functional (B3LYP) with the same basis set (6-31G(d,p)). Next the absorption and emission wavelengths were calculated by using the TDDFT method on each optimized ground state geometry. The effect of the solvent was introduced through the use of the polarisable continuum model (PCM).17,19,20 In our calculations on Cl-3HF in aqueous micelles, we assumed the micellar environment to be a uniform dielectric. The refractive index and dielectric constant of micelle–water


View Article Online

Paper

PCCP

interfaces reported in the literature21,22 were used to characterize this environment. This choice is justified by the fact that our experimental data strongly indicate that Cl-3HF molecules occupy mainly the hydrophilic external shell of the micelle, as discussed elaborately in Section 3.3 below.


2.2.

Chemicals and spectroscopy

3-Hydroxyflavone was purchased from Sigma-Aldrich. The derivative, 4 0 -chloro,3-hydroxyflavone, was prepared by the Alger–Flynn– Oyamada reaction23 using 2-hydroxyacetophenone and 4-chlorobenzaldehyde. The surfactants cetyltrimethylammonium bromide (CTAB), sodium dodecylsulfate (SDS), Triton X-100 (TX-100) as well as the amphiphilic block copolymer Pluronic F-68 were all purchased from Sigma-Aldrich. All solvents were of

UV-spectroscopy grade and distilled freshly by a standard procedure before use. Absorption and fluorescence spectra were measured using a HITACHI UV spectrophotometer (U-3501) and Perkin Elmer (LS 55) spectrofluorimeter, respectively. Picosecond fluorescence dynamic studies of the fluorophore solutions were performed using a time correlated single photon counting (TCSPC) system by employing a picosecond diode laser operating at lex = 375 nm, and a pulsewidth of B70 ps.

3. Results and discussion 3.1.

Quantum chemical calculations

3.1.1. Molecular geometries. The optimized geometries of Cl-3HF in the ground-state (S0) and the first excited singlet state

Fig. 4 Optimized molecular structures of enol, tautomer and proton-transfer transition states of Cl-3HF in S0 and S1 electronic states. Numerical figures indicate bond lengths in Å unit.



View Article Online


PCCP

(S1) are shown in Fig. 4. The bond lengths are found to maintain a close similarity with the corresponding bond lengths reported for the original 3HF molecule.17,24 Thus, the 4 0 -Cl atom does not appear to strongly perturb the molecular geometries. In particular, a close examination of the O H O bond distances spanning the 3-OH and ring carbonyl groups reveals a competition between the O atoms of these two groups to hold on to the hydroxyl proton. The O H distance for the 3-OH group increases B1.8 times while that for the ring carbonyl group decreases B1.9 times on moving from S0(E) to S0(T). In the S1 surface, the former increases B1.95 times, while the latter decreases B1.73 times. Moreover, on moving from S0(E) to S0(T.S.) the former increases B1.4 times and the latter decreases B1.7 times, while on moving from S1(E*) to S1(T.S.*) the former increases B1.13 times and the latter decreases B1.26 times. Obviously, S0(T.S.) displays a much more advanced stage of proton transfer than S1(T.S.*). In other words, the S0 transition state resembles S0(T), while the S1 transition state resembles S1(E).

Paper

3.1.2. Spectroscopic transitions and orbital electron distributions in vacuum. The above point is further stressed in the spectral transitions depicted in Fig. 5. In terms of both energy and position along the proton-transfer coordinate, S0(T.S.) is closer to the S0(T), while the S1(T.S.*) is closer to the S1(E*). In fact, activation energy required for enol - tautomer conversion is nearly B10 fold lesser in the excited state than in the ground state. It is also B10 fold lesser than that required for tautomer enol back proton transfer in the excited state. Therefore we can safely conclude that: (i) enol - tautomer ESIPT is highly irreversible and (ii) ESIPT is much more facile than ground state intramolecular proton transfer. From Fig. 5, we also note that the initial S1 0 (E*) excited state generated by Franck–Condon vertical transition is located 0.165 eV above S1(T.S.*), which is sufficient to render the ESIPT movement practically barrierless. The HOMO–LUMO diagrams for both enol and tautomer forms display a considerable rearrangement of electron densities following the S0 - S1 transition. By contrast, the electron distribution in the enol and tautomer forms within the same

Fig. 5 Relative energy levels (in eV units, not in scale) of enol, tautomer and proton-transfer transition states of Cl-3HF in S0 and S1 electronic states. Wavelengths (in nm units) corresponding to a given electronic transition are shown within parantheses. HOMO and LUMO of enol and tautomer forms as well as the ground-state potential energy surface are also shown.



View Article Online

Paper Table 1

PCCP Calculated spectral parameters for electronic transitions of enol and tautomer forms of Cl-3HF in vacuum

Calculated Absorption

Emission

Form of Cl-3HF

Orbital transition

Absorption lmax (nm)

Oscillator strength

Orbital transition

Emission lmax (nm)

Oscillator strength

Enol

H-L H-1 - L

349 295

0.44 0.14

L-H

399

0.44

L-H

553

0.31


Tautomer

energy surface (S0 or S1) is not very different. This implies that the intramolecular proton transfer event – either in the ground or excited state – is highly localized within the ring carbonyl and 3-hydroxy groups; it does not upset the overall molecular electronic distribution to any great extent. Incidentally, the HOMO–LUMO diagrams of the Cl-3HF enol are nearly the same as those of 3HF enol,17 except that a sizeable electron density tends to reside at the electronegative 4 0 -Cl atom of Cl-3HF. The calculated main orbital transitions in vacuum, listed in Table 1, are found to be dominated by HOMO–LUMO transitions in both the enol and tautomer forms. It is instructive to compare these results – obtained by precluding any external perturbation on the Cl-3HF molecule – with its spectral data recorded in purely non-polar solvents like n-heptane, where solute–solvent interactions like H-bonding or dispersive forces are either extremely weak or absent. Thus, we found that the absorption spectrum of Cl-3HF in n-heptane peaks at 342 nm (see below), which is quite close to the calculated absorption lmax of 349 nm. However, while calculation produces two prominent emissions at lmax = 399 nm and 553 nm pertaining to the LUMO - HOMO transitions of enol and tautomer forms, respectively, the emission spectrum in n-heptane features only a single peak at 525 nm. Hence, in reality, E* emission has almost zero intensity although calculations predict a non-negligible oscillator strength for it. We believe that since ESIPT is highly irreversible, E* and T* do not co-exist, i.e., E* - T* conversion occurs before E* has the chance to emit fluorescence. 3.1.3. Spectroscopic transitions in protic solvent. Since our experiments were performed in polar solvents and micelles, a more realistic insight into the system is gained by computing the electronic transition in the protic H-bonding solvent – methanol. As mentioned above, the Cl-3HF molecule likely forms H-bonded complexes in such a solvent. The calculated electronic charge densities for different states of the Cl-3HF molecule in methanol, displayed in Table 2, reveal a large

Table 2

accumulation of negative charge on the O atoms of ring carbonyl and 3-hydroxy groups. Thus, these centres are particularly susceptible to H-bonding with methanol, which is an excellent H-bond donor and a moderate H-bond acceptor with a = 0.93 and b = 0.62.25 Based on this clue, several possible 1 : 1 complexes between Cl-3HF and CH3OH were constructed and tested for their energies, of which the most stable structures were identified, as shown in Fig. 6. Note that the structures E-I, E-II and E-IV closely resemble the complexes [E-solv] 0 , [E-solv]00 and [E-solv]0 0 0 0 , respectively, but no structure resembling [E-solv]0 0 0 counts among the most stable forms. Since a 4 b for methanol, complexes where methanol acts only as H-bond acceptor base are not favored. The computed relative energies of the 3 most stable enol–solvent complexes lie in the sequence: [E-IV] (0.00 eV) o [E-I] (0.07 eV) o [E-II] (0.20 eV). The higher stability of E-IV derives from the fact that it comprises of 2 H-bonds, while E-I and E-II comprise of only 1H-bond each. Furthermore, Table 2 demonstrates that electronic charge density of the enols is higher on the ring carbonyl O atom than on the 3-hydroxy O atom. Hence, the former is a better H-bond acceptor than the latter, causing E-I to be more stable than E-II. For all the complexes short-listed in Fig. 6, the spectral transitions in methanol were computed and listed in Table 3. Like in vacuum, HOMO–LUMO was the dominant transition in methanol. For the enol form, the calculated absorption lmax lie between 353 and 362 nm, and thus coincide well with the wavelength range between the 347 nm peak and 361 nm shoulder of the Cl-3HF absorption spectrum in methanol, as displayed in Fig. 7. Calculations further yield emission lmax between 418 nm and 440 nm for the enols and between 556 nm and 580 nm for the tautomers, respectively. By comparison, the experimental emission spectrum of Cl-3HF in methanol at excitation wavelength l = 370 nm consists of broad bands around 410 nm and 532 nm, attributed to the enol and tautomer forms, respectively, as shown in Fig. 8 and discussed elaborately

Calculated electronic charge densities for different ground and excited state forms of the Cl-3HF molecule in methanol

Enol (E) & tautomer (T) forms of Cl-3HF

Ring A + C Ring B Ring carbonyl O 3-Hydroxy O 4 0 -Cl a

Anion of Cl-3HF (A)

S0 (E)

S1 0 a (E*)

S1b (E*)

S1b (T*)

S0 (A)

S1 0 a (A*)

S1b (A*)

0.77 0.09 0.59 0.24 0.03

0.66 0.12 0.63 0.15 0.00

0.67 0.12 0.65 0.14 0.00

0.73 0.06 0.19 0.58 0.02

0.45 0.09 0.59 0.69 0.08

0.31 0.06 0.62 0.57 0.06

0.31 0.06 0.62 0.57 0.06

Franck–Condon excited S1 state.

b

Relaxed S1 state.



View Article Online


PCCP

Fig. 6

Paper

Optimized structures of possible stable H-bonded complexes of enol (E), tautomer (T) and anion (A) forms of Cl-3HF with methanol.

below in Section 3.2. Thus, the calculated lmax for all un-solvated and solvated Cl-3HF fall within the wavelength range spanned by the experimental emission bands. However, due to the large broadness of the enol emission band, it is impossible to attribute the spectral transitions to any one form of Cl-3HF enol specifically. Therefore we summarize the situation by noting that Cl-3HF enol in methanol exists in a variety of solvation states. In the case of the Cl-3HF anion, the stable H-bonded complexes include both 1 : 1 and 1 : 2 stoichiometries, as shown in Fig. 6 (A-I and A-II, respectively). But A-II is eliminated from consideration because calculations predict an extremely low emission oscillator strength for it, rendering it non-fluorescent. For A-I, calculated absorption and emission lmax are 450 nm and 560 nm, respectively. However, experimental absorption and emission peaks for the Cl-3HF anion in methanol are 425 nm and 460 nm, respectively, as discussed below in Section 3.2. Although the results qualitatively capture the redshift of electronic transitions of the anion compared to the enol, they lack quantitative accuracy. In our method, we have neglected electron correlation to minimize computational cost. But this becomes an important factor for the anion,


which contains an extra electron. Moreover, presence of this extra electronic charge in the anions causes their electronic orbitals to be more expanded than the neutral molecules. Hence, for a better description of the situation, more diffuse functions need to be incorporated into the applied basis set. We are currently exploring this problem. 3.2.

Spectroscopy in polar solvents

3.2.1. Steady-state spectra. Absorption and emission spectra of Cl-3HF were recorded in the solvents n-heptane, 1,4-dioxane, acetonitrile, N,N-dimethylformamide (DMF) and methanol. The hydrogen bond donor acidity (a) and hydrogen bond acceptor basicity (b) values of these solvents25 are listed in Table 4. The absorption spectra in Fig. 7 show a slight but persistent red-shift as the polarity of the solvent is increased. An additional peak at l 4 400 nm appears prominently in DMF and methanol, less prominently in acetonitrile and dioxane, but is totally absent in n-heptane. Such a l 4 400 nm peak has been observed previously in 3HF and its other derivatives, where it has been assigned to the anion (Fig. 1c).1–3,6 As discussed above in Section 3.1, our computational results are also in qualitative agreement with this assignment.


View Article Online

Paper

PCCP

Table 3 Calculated spectral parameters for electronic transitions of various forms of Cl-3HF in methanol and aqueous micelles. Experimentally observed transition wavelengths are also shown for comparison

Calculated


Absorption Medium

Form of Cl-3HF

Methanol (e = 32.7)

Enol

a

CTAB-water (e = 29) SDS-water (e = 48)a

a

TX-100-water (e = 32) a

State of complexation

Un-complexed E-I E-II E-IV Tautomer Un-complexed T-I T-II T-IV Anion Un-complexed A-I Enol Un-complexed Tautomer 00 00 Enol Tautomer 00 00 Enol Tautomer 00

Experimental

Emission

Emission Oscillator Absorption Emission Orbital Absorption Oscillator Orbital transition lmax (nm) strength transition lmax (nm) strength lmax (nm) lmax (nm) H H H H

-

L L L L

353.52 358.97 354.32 361.92

0.53 0.58 0.68 0.53

H-L H-L H-L

478.78 453.91 367.19

0.36 0.38 0.72

H-L

367.62

0.73

H-L

367.26

0.72

L L L L L L L L L L L L L L L L

-

H H H H H H H H H H H H H H H H

422.17 425.89 418.15 440.42 579.35 580.36 556.56 567.32 566.00 568.54 422.02 579.17 422.71 580.02 421.96 579.31

0.78 0.83 0.80 0.76 0.60 0.61 0.64 0.60 0.50 0.52 0.78 0.60 0.79 0.61 0.78 0.60

B350

B400

—

B530

B420

B460

B350 — B350 — B350 —

B410 B550 B410 B550 B410 B550

According to ref. 22.

Fig. 7 Absorption spectra of Cl-3HF in various solvents. Inset shows the expanded view of the absorption spectra from 380 nm to 480 nm.

Keeping in mind the existence of various un-solvated and solvated Cl-3HF species in polar solvents, we recorded fluorescence spectra using several different excitation wavelengths across the absorption window: 300 nm, 340 nm, 370 nm and 400 nm. The results are shown in Fig. 8. Using the first three wavelengths, the spectra of Cl-3HF in dioxane and DMF clearly exhibit dual fluorescence involving two peaks: the one at the shorter emission wavelength is attributed to the excited enol (E*), while the other at the longer emission wavelength to the excited tautomer (T*) generated by ESIPT. Dual fluorescence is also discernible in acetonitrile, although the E* intensity in this solvent is much subdued. In the polar protic solvent, methanol, instead of dual fluorescence, we have three prominent peaks at 410 nm, 532 nm, and 460 nm. On the other hand, when an excitation wavelength of 400 nm is used, a dramatic change is observed in these polar


solvents: the Cl-3HF fluorescence becomes dominated by a completely new peak at 460–480 nm. The excitation spectra of Cl-3HF in methanol, shown in Fig. 9, reveal that the fluorescence around lem = 460 nm derives maximum contribution from a species absorbing at l 4 400 nm, which we had identified above as the anion of Cl-3HF. The other polar solvents give the same result. Therefore, the fluorescence at 460–480 nm likely arises from the excited anion A*. However, as pointed out above, our computational results at this point are not reliable enough to predict the correct state of complexation of the anion. We note that the peak positions of A* and T* fluorescence tend to localize around 460–480 nm and 525–530 nm, respectively, regardless of the solvent under study. In contrast, the E* fluorescence appears to depend quite strongly on the nature of the solvent. Thus, the E* peak of Cl-3HF is completely absent in n-heptane (spectra not shown), barely apparent in acetonitrile, and prominent in the other solvents under study. Even among the latter, the E* peak in low polar dioxane (e = 2.25) is markedly blue-shifted than in high polar DMF (e = 36.7) or methanol (e = 32.7). 3.2.2. Time-resolved fluorescence. The time-resolved fluorescence of Cl-3HF in dioxane and methanol at different emission wavelengths is displayed in Fig. 10. In dioxane, the emission at lem = 430 nm (dominated by E* emission) undergoes an ultra-fast decay, resulting in a time-profile that is nearly indistinguishable from that of the instrumental response function of the experiment. In comparison, the emission at lem = 540 nm (dominated by T* emission) decays much more slowly. Moreover, as the inset of Fig. 10a reveals, while the E* emission is a pure decay, the T* emission consists of an initial ultrafast growth feature followed by decay. In methanol on the other hand, the behavior of the E* emission (lem = 410 nm) is strikingly similar to that in dioxane. Also, here too, the T* emission


View Article Online

PCCP

Paper Table 4

Solvent

a

b

Acetonitrile 1,4-Dioxane DMF Methanol

0.19 0.00 0.00 0.93

0.31 0.37 0.69 0.62

a


Solvent H-bonding propertiesa

According to ref. 25.

Fig. 9 Comparison of the absorption spectrum of Cl-3HF in methanol and its fluorescence excitation spectra at different emission wavelengths.

T* emissions: they consist of a very long decay component, longer than the T* decay component. The time-profiles were all fitted with a polyexponential function of the general form: FðtÞ ¼

X i

Fig. 8 Steady-state fluorescence emission spectra of Cl-3HF in various solvents. The excitation wavelengths used are: 300 nm (solid circle), 340 nm (open circle), 370 nm (solid curve), 400 nm (dotted curve).

(lem = 530 nm) consists of an initial ultrafast growth feature followed by decay which is much longer than that of the E* decay. However, the emission time-profiles of Cl-3HF in methanol recorded at intermediate emission wavelengths (lem = 470 nm and 500 nm) exhibit a marked difference from either E* or


t ai exp ti

(1)

The results are listed in Table 5. E* emission of Cl-3HF in both dioxane and methanol undergoes ultrafast decays within several tens of picoseconds. However, since our experimental time-resolution is 70 ps, we are unable to capture this component reliably. Hence we have designated it as ‘‘o0.07 ns’’ in Table 5. Incidentally, this component also appears in the T* emission for both solvents, where it carries a negative preexponential factor, implying an initial ultra-fast rise in the timeprofile. Note that the change-over from ultrafast decay to ultrafast initial rise occurs as the emission wavelength is increased, as can be seen from the data for methanol (lem = 410 nm - 470 nm - 500 nm - 530 nm). The very fact that the ultrafast component occurs both in the decay of E* emission as well as in the initial rise of T* emission clearly indicates that it represents the ESIPT dynamics connecting the two species. Once the T* species has been generated from the E* via the ultrafast ESIPT, it decays with a rather slow time-constant: 420 ps in methanol and 1.25 ns in dioxane. In methanol, a B3 ns component appears in the timeprofiles at all emission wavelengths, carrying the highest relative weightage at lem = 500 nm. Since, according to Fig. 8, the emission in this region draws a heavy contribution from the


View Article Online


Paper

PCCP

Fig. 10 Picosecond fluorescence time-profiles of Cl-3HF at different emission wavelengths in (a) dioxane, and (b) in methanol. Insets show the expanded view of the time-profiles at the initial stages.

Cl-3HF anion, this component is assigned to the decay of the excited A* species. That a very minor B3 ns component also appears at other emission wavelengths is probably because of the spectral overlap of E* or T* fluorescence with A* fluorescence,

Table 5

which causes the former to be somewhat contaminated by the latter. The ESIPT reaction of 3HF essentially implies a proton jump between two neighbouring electronegative O atoms separated over an extremely short distance. Thus, in the gas phase or in a non H-bonding apolar solvent, where the 3HF enol remains intra-molecularly H-bonded and free from solvent interference, the ESIPT time-constant is inherently sub-picosecond,26,27 limited only by the vibrational period of the intra-molecular hydrogen bonded O–H bond stretch. As a result, the E* emission is completely extinguished, and only the T* emission is observed. In Cl-3HF too, the fluorescence spectra in n-heptane (not shown) comprises of only the T* peak. In contrast, E* fluorescence of Cl-3HF is definitely detectable in polar solvents, indicating that intra-molecular H-bonding and associated sub-picosecond ESIPT of the Cl-3HF enol is disrupted by competitive inter-molecular H-bonding with the solvent. In other words, the enol in polar solvents is largely trapped in [E-solv]-type complexes, where H-bonding with the solvent inhibits ESIPT. It is particularly inhibited in complexes like [E-solv] 0 and [E-solv] 0 0 0 0 , where the ring carbonyl O atom of Cl-3HF remains H-bonded with methanol. Incidentally, our computational results presented above demonstrate an obvious preference for [E-solv] 0 0 0 0 and [E-solv] 0 -type complexes. Upon excitation, some of these complex-bound enols extricate from intermolecular H-bonding and engage in ESIPT (Fig. 1). Thus, in this mechanism, the overall ESIPT process is coupled with the destruction of the initial enol–solvent complexes, so that its rate is controlled by the relatively slow dynamics of the inter-molecular H-bonds. This explains why the ESIPT time-constants in polar solvents extend to several tens of picoseconds. Although [E-solv] 0 0 0 type complexes are insignificant in methanol, they may play a decisive role in solvents where a = 0 or a { b. According to Fig. 1c, tighter is the (enol J) H-bond in the [E-solv] 0 0 0 complex, more efficiently does it hinder the intra-molecular H-bond within the enol. Thus, E* emission is enhanced with the b-value: it is extremely weak in acetonitrile but prominent in 1,4-dioxane and DMF. Finally, visual comparison as well as fitting analysis confirms that the long-time decay behaviour of the emission time-profiles of E* and T* are dissimilar for both dioxane and methanol. This indicates that the ESIPT of Cl-3HF in polar solvents is fast and irreversible.28

Fitting parameters for fluorescence time-profiles of Cl-3HF in different solvents

(i) Dioxane lem = 430 nm (E*)

lem = 540 nm (T*)

t1 (a1)

t2 (a2)

t1 (a1)

t2 (a2)

o0.07 ns (96%)

3.00 ns (04%)

o0.07 ns (55%)

1.25 ns (45%)

(ii) Methanol lem = 410 nm (E*)

lem = 470 nm

t1 (a1)

t2 (a2)

t1 (a1)

t2 (a2)

t3 (a3)

t1 (a1)

t2 (a2)

t3 (a3)

t1 (a1)

t2 (a2)

t3 (a3)

o0.07 ns (97%)

3.10 ns (03%)

o0.07 ns (89%)

3.00 ns (06%)

0.41 ns (05%)

o0.07 ns (11%)

3.00 ns (78%)

0.43 ns (11%)

o0.07 ns (47%)

3.00 ns (01%)

0.42 ns (52%)


lem = 500 nm

lem = 530 nm (T*)


View Article Online


PCCP

Paper

3.2.3. Tautomer emission decays. We performed time-resolved fluorescence experiments on Cl-3HF in several other solvents – n-heptane, acetonitrile and DMF. In the case of the polar solvents, E* time-profiles had almost identical behavior to those in dioxane and methanol. However, for T* emission, the decay following the initial ultrafast growth has a remarkable dependence on the nature of the solvent. The T* time-profiles plotted in Fig. 11 show that the decay rate becomes markedly slower in the sequence methanol, DMF - acetonitrile - dioxane - n-heptane, i.e., in the order of lower polarity. The decay rates were single-exponential with time-constants listed in Table 6. 3.3.

Spectroscopy in micelles

3.3.1. Steady-state spectra. Since the Cl-3HF molecule is nearly insoluble in water, in a micellar solution it is expected to reside almost entirely within the micelle. However, within the host micelle are regions of very different micro-environment, and the guest Cl-3HF may partition among these regions. The effect of its specific environment must be reflected in the photophysics of Cl-3HF, which we already found to be extremely sensitive to the polarity of its immediate surroundings. The spectra of Cl-3HF in different micelles are displayed in Fig. 12. We have selected three different well-known surfactants: cationic CTAB, anionic SDS, and neutral TX-100, as well as a

Fig. 11 Picosecond fluorescence time-profiles of Cl-3HF in tautomer form in (a) various solvents, and (b) in various solvents together with micelles (latter represented as dotted curves).


Table 6 Decay time-constants for T* emission of Cl-3HF in different solvents

Solvent

Time-constant (ns)

n-Heptane Dioxane Acetonitrile DMF Methanol

3.55 1.25 1.00 0.43 (major) 0.42 (major)

neutral amphiphilic block copolymer Pluronic F68, as the components of our micelles. The fluorescence spectra in all micelles exhibit dual emission when excited at 340 nm or 370 nm. Now, had all the Cl-3HF molecules been buried deep down the hydrophobic core of the micelle where the environment resembles that of a non-polar hydrocarbon like n-heptane, the E* emission would have been totally extinguished, as explained above. Hence, the dual emission behaviour strongly suggests that at least a sizeable part of Cl-3HF population occupies the only other region available to it: the polar shell of the micelles. Spectral transitions calculated for Cl-3HF in different micelle–water interfaces yield absorption and emission lmax that tally well with experimental spectral bands of Cl-3HF in aqueous micelles, as seen from Table 3 and Fig. 12. Hence, compared to the hydrophobic core, the microenvironment at the hydrophilic shell resembles a micelle–water interface, consisting of water molecules solvating the polar head groups of the surfactants. Cl-3HF molecules residing at the shell region have direct access to these water molecules whereby intermolecular H-bond can form between them, generating the same type of [E-solv] complexes as in polar solvents. 3.3.3. Time-resolved fluorescence. The fluorescence timeprofiles of Cl-3HF in the micelles are shown in Fig. 13. For a given micelle, the emission wavelengths were chosen to coincide with the E* and T* fluorescence peaks of Cl-3HF in that micelle. As in the case of polar solvents, the responses of the two fluorescence time-profiles in micelles are categorically distinct: E* displays only fast decay, while T* shows initial ultrafast growth, followed by decay. Fitting analysis was performed with a polyexponential function as before, and the results are listed in Table 7. In all the micelles, an ultrafast E* decay time-constant appears just as in the polar solvents. Since this decay is too short for our experimental time-resolution of 70 ps, we designate it as the ‘‘o0.07 ns’’ component, as before. Moreover, since this ultrafast component also appears as the growth timeconstant of the corresponding T* emission, it may safely be attributed to the ESIPT process occurring from the destruction of [E-solv]-like complexes, where solv = water molecules of the micelle shell. A minor component of B3 ns is also detected in some cases, similar to the A* emission of Cl-3HF in polar solvents. Although several authors have ruled out deprotonation of 3HF in pure water,1,3 there is evidence that 3HF gives rise to anions in micellar media.29 Therefore, here too, the B3 ns component is assigned to the A* generated from [E-solv] 0 0 0 0 -like complexes.


View Article Online


Paper

PCCP

Fig. 13 Picosecond fluorescence time-profiles of (a) E* form, and (b) T* form of Cl-3HF in various miceller media. Inset in (a) highlights the long time-scale behavior of the T* emission time-profiles.

Fig. 12 Steady-state absorption spectra (a), and fluorescence emission spectra at (b) lex = 340 nm and at (c) lex = 370 nm, of Cl-3HF in various miceller media.

Turning to the T* emission, we find that in all micelles except TX-100, the T* emission of Cl-3HF register single-exponential decay, following the initial ultra-fast growth phase. The single decay time-constant behaviour strongly suggests that the T* molecules in a given micelle all experience an absolutely uniform environment, i.e., they are localized in the same region of the micelle. To interpret the results, we emphasize that T* emission of Cl-3HF can possibly originate from only two types of species:


(i) Enols that are totally intra-molecularly H-bonded (Fig. 1a). These reside in the micelle core composed of the long hydrocarbon chain-like tails of the surfactants. Here, free from solvent interference, the Cl-3HF enol is expected to undergo sub-picosecond ESIPT. The corresponding T* emission timeprofile must then resemble that in a pure hydrocarbon like n-heptane. (ii) Enols engaged in [E-solv]-like complexes (Fig. 1b–d), which must be localized at the polar shell of the micelle. Any other enol or enol-derived species will not generate T*. The T* decay time-constants in micelles are around 800 ps to 1.5 ns, similar to those in moderately polar solvents (see Tables 5 and 7). Even in the case of TX-100 micelles, where the T* decay requires a bi-exponential fit, both the two decay timeconstants (800 ps and 1.5 ns) are considerably shorter than that in a pure hydrocarbon like n-heptane (tdecay B 3.55 ns, Table 6). In fact, careful inspection of Fig. 11b reveals that the T* emission profiles in all the micelles tend to cluster around


View Article Online

PCCP Table 7

Paper Fitting parameters for fluorescence time-profiles of Cl-3HF in different micelles


E* Micelle

t1 (a1)

CTAB SDS TX-100 Plu-F68

o0.07 o0.07 o0.07 o0.07

T*

ns ns ns ns

(94%) (99.5%) (85%) (69%)

t2 (a2)

t3 (a3)

t1 (a1)

3.3 ns (06%) 3.0 ns (0.5%) 0.25 ns (14%) 0.25 ns (29%)

— — 3.0 ns (01%) 3.00 ns (02%)

o0.07 o0.07 o0.07 o0.07

those of acetonitrile and dioxane, i.e., they register decays much faster than the T* emission in n-heptane. Therefore, the possibility (i) cited above must be rejected, while possibility (ii) is established: that the Cl-3HF molecules reside overwhelmingly in the shell region of the micelle, where the environment resembles that in a polar solvent. This observation is apparently surprising, since Cl-3HF itself is strongly hydrophobic, and almost insoluble in water. At this point, we recall that the water molecules inhabiting the micelle polar shell are quite different from those in bulk, homogeneous liquid water, on account of their confinement. It is an established fact that these two types of water molecules exhibit distinct response in many physico-chemical phenomena, like polarity and solvation dynamics.30 Since these water molecules are essentially tethered to the surfactant headgroups, the Cl-3HF molecule H-bonding with them also loses their mobility significantly. Thus, the majority of Cl-3HF population get restricted within the polar shell of the micelle, only a few may stray elsewhere.

4. Conclusion In this work, we have explored the photophysics of the Cl-3HF molecule in a variety of H-bonding environments: polar solvents, protic solvents and aqueous micelles. Computational results emphasize that Cl-3HF undergoes ESIPT just like the parent 3-HF molecule, and that the ESIPT process is intrinsically almost barrier-less. Computed spectral transitions show close agreement with experimental spectral data and can also qualitatively reproduce the red-shift in spectral transitions of the Cl-3HF anion compared to Cl-3HF enol. Combination of experimental and computational data further highlight the existence of a variety of H-bonded complexes of Cl-3HF enol in protic solvents like methanol or interfacial water in aqueous micelle solutions. Since inter-molecular H-bonding opposes ESIPT, it can proceed only after such complexes are destroyed or weakened in the excited state. Thus, the overall ESIPT process is coupled intimately with the destruction of the initial enol–solvent complexes, so that its rate is controlled by the relatively slow dynamics of the intermolecular H-bonds. As a consequence, the ESIPT in polar solvents and micelles is retarded to a time-scale of several tens of picoseconds, although it could not be accurately detected by our experimental time-resolved fluorescence setup, which has a timeresolution of B70 ps. With a better time-resolution, it may even be possible to distinguish the ESIPT time-constants emanating from various different enol–solvent H-bonded complexes. Our results conclusively show that, although Cl-3HF is strongly hydrophobic, in aqueous micelles it avoids the hydrophobic


t2 (a2) ns ns ns ns

(37%) (50%) (35%) (31%)

1.00 1.50 0.80 0.80

ns ns ns ns

t3 (a3) (63%) (50%) (23%) (68%)

— — 1.50 ns (42%) 3.00 ns (01%)

micelle core and resides exclusively at the hydrophilic micelle shell. This is ostensibly because water molecules occupying the micelle-water interface are known to have remarkably different physico-chemical properties from water molecules in the bulk state. H-bonding between these water molecules and Cl-3HF reduces the mobility of the latter and forces them to remain localized within the micelle shell region.

Acknowledgements We sincerely thank Soumya Ganguly Neogi and Dr P. Chaudhury, Department of Chemistry, University of Calcutta, India, for assistance with quantum chemical calculations using GAUSSIAN 09. We also gratefully acknowledge the Council of Scientific & Industrial Research (CSIR), India, for financial support (under Project No. 01(2390)/10/EMR – II). D.G. and A.K.P. thank CSIR for awarding research fellowships.

References 1 P. K. Mandal and A. Samanta, J. Phys. Chem. A, 2003, 107, 6334–6339. 2 A. Douhal, M. Sanz, L. Tormo and J. A. Organero, ChemPhysChem, 2005, 6, 419–423. 3 S. Protti, A. Mezzetti, J. Cornard, C. Lapouge and M. Fagnoni, Chem. Phys. Lett., 2008, 467, 88–93. 4 R. Das, A. S. Klymchenko, G. Duportail and Y. Mely, J. Phys. Chem. B, 2008, 112, 11929–11935. 5 J. A. Organero, L. Tormo, M. Sanz, A. Roshal and A. Douhal, J. Photochem. Photobiol., A, 2007, 188, 74–82. 6 M. Sanz, J. A. Organero and A. Douhal, Chem. Phys., 2007, 338, 135–142. 7 C. A. Kenfack, A. S. Klymchenko, G. Duportail, A. Burger and Y. Mely, Phys. Chem. Chem. Phys., 2012, 14, 8910–8918. 8 A. D. Roshal, J. A. Organero and A. Douhal, Chem. Phys. Lett., 2003, 379, 53–59. 9 P. T. Chou, S. C. Pu, Y. M. Cheng, W. S. Yu, Y. C. Yu, F. T. Hung and W. P. Hu, J. Phys. Chem. A, 2005, 109, 3777–3787. 10 W. Zhang, B. Shi and J. Shi, THEOCHEM, 2005, 731, 219–224. 11 A. Douhal, M. Sanz, M. A. Carranza, J. A. Organero and L. Santos, Chem. Phys. Lett., 2004, 394, 54–60. 12 S. Ameer-Beg, S. M. Ormson, X. Poteau, R. G. Brown, P. Foggi, L. Bussotti and F. V. R. Neuwahl, J. Phys. Chem. A, 2004, 108, 6938–6943. 13 D. L. Nelson, A. L. Lehninger and M. M. Cox, Lehninger Principles of Biochemistry, W.H. Freeman, 5th edn, 2008.


View Article Online


Paper

14 J. E. Kwon and S. Young, Adv. Mater., 2011, 23, 3615–3642. 15 J. Wu, W. Liu, J. Ge, H. Zhang and P. Wang, Chem. Soc. Rev., 2011, 40, 3483–3495. 16 M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B. Mennucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li, H. P. Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Sonnenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven, J. A. Montgomery, Jr., J. E. Peralta, F. Ogliaro, M. Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov, T. Keith, R. Kobayashi, J. Normand, K. Raghavachari, A. Rendell, J. C. Burant, S. S. Iyengar, J. Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene, J. E. Knox, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, R. L. Martin, K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador, J. J. Dannenberg, S. Dapprich, A. D. Daniels, O. Farkas, J. B. Foresman, J. V. Ortiz, J. Cioslowski and D. J. Fox, Gaussian 09, Revision B.01, Gaussian, Inc., Wallingford, CT, 2010. ´s, O. Vendrell, M. Moreno, J. M. Lluch and 17 R. Casadesu K. Morokuma, Chem. Phys., 2006, 325, 243–250.


PCCP

18 C. Peng and H. B. Schlegel, Isr. J. Chem., 1993, 33, 449. 19 B. Mennucci, J. Tomasi, R. Cammi, J. R. Cheeseman, M. J. Frisch, F. J. Devlin, S. Gabriel and P. J. Stephens, J. Phys. Chem. A, 2002, 106, 6102–6113. 20 C. A. Kenfack, A. S. Klymchenko, G. Duportail, A. Burgerc and Y. Mely, Phys. Chem. Chem. Phys., 2012, 14, 8910–8918. 21 C. H. Tan, Z. J. Huang and X. G. Huang, Anal. Biochem., 2010, 401, 144–147. 22 G. G. Warr and D. F. Evans, Langmuir, 1988, 4, 217–224. 23 M. A. Smith, R. M. Neumann and R. A. Webb, J. Heterocycl. Chem., 1968, 5, 425–426. 24 C. Lapouge and J. P. Cornard, J. Phys. Chem. A, 2005, 109, 6752–6761. 25 M. J. Kamlet, J.-L. M. Abboud, M. H. Abraham and R. W. Taft, J. Org. Chem., 1983, 48, 2877–2887. 26 D. McMorrow and M. Kasha, J. Am. Chem. Soc., 1983, 105, 5133–5134. 27 C. Chudoba, E. Riedle, M. Pfeiffer and T. Elsaesser, Chem. Phys. Lett., 1996, 263, 622–628. 28 V. I. Tomin, S. Oncul, G. Smolarczyk and A. P. Demchenko, Chem. Phys., 2007, 342, 126–134. 29 S. Mondal, S. Basu and D. Mandal, Chem. Phys. Lett., 2009, 479, 218–223. 30 K. Bhattacharyya, Chem. Commun., 2008, 2848–2857.


E.s.r. of free radicals in aqueous solutions of substituted pyrimidines.

Proton Transfer Dynamics of 4'-N,N-Dimethylamino-3-hydroxyflavone Observed in Hydrogen-Bonding Solvents and Aqueous Micelles.

Proton transfer reactions of methylglyoxal synthase.

Aqueous proton transfer across single-layer graphene.

Photodegradation of 2-mercaptobenzothiazole and 1,2,3-benzotriazole corrosion inhibitors in aqueous solutions and organic solvents.

Proton transfer reactions and hydrogen-bond networks in protein environments.

Studies of the conformation of apomyoglobin in aqueous solutions and denaturing organic solvents.

Assessing the importance of proton transfer reactions in DNA.

The reactions of alpha-chymotrypsin and related proteins with ester substrates in non-aqueous solvents.

Exploring the metric of excited state proton transfer reactions.

Infrared-activated proton transfer in aqueous nafion proton-exchange-membrane nanochannels.

The proton transfer reactions catalyzed by yeast pyruvate kinase.

Photoelectric conversion based on proton-coupled electron transfer reactions.

The solubilization of some para-substituted acetanilides in aqueous solutions of polyoxyethylene-polyoxpropylene block copolymers [proceedings].

Sonochemical redox reactions of Pu(III) and Pu(IV) in aqueous nitric solutions.

Solvent proton relaxation of aqueous solutions of the serum proteins alpha 2-macroglobulin, fibrinogen, and albumin.

Photodynamics of a Proton-Transfer Dye in Solutions and Confined Within NaX and NaY Zeolites.

Multicomponent diffusion in solute-containing micelle and microemulsion solutions.

Predicting proton titration in cationic micelle and bilayer environments.

Competition and interplay of various intermolecular interactions in ultrafast excited-state proton and electron transfer reactions.

The influence of substituted phenols on the sol:gel transition of hydroxypropyl methylcellulose (HPMC) aqueous solutions.

Optimizing the transfer of [¹⁸F]fluoride from aqueous to organic solvents by electrodeposition using carbon electrodes.

A modified two-state empirical valence bond model for proton transport in aqueous solutions.

Photochemical redox reactions of copper(II)-alanine complexes in aqueous solutions.