Reduced fluorescence quenching of coumarin 102 at higher phenol mole fractions in cyclohexane–phenol and anisole–phenol solvent mixtures: role of competitive hydrogen bonding

Abstract

Recently, we demonstrated that the fluorescence of coumarin 102 (C102) drastically quenches upon hydrogen bonding (H-bonding) with phenol in a non-polar solvent cyclohexane due to H-bond driven photoinduced electron transfer (PET) (J. Phys. Chem. A, 2013, 117, 3945–3953). However, the work was limited to low concentrations of phenol only, where predominately a 1 : 1 C102–phenol complex exists. Herein, we report an unusual fluorescence modulation of C102 over mole fractions of the H-bond donor (phenol) in the cyclohexane–phenol mixtures covering a higher range of mole fractions. We found that the fluorescence quantum yield (ϕ_f) and fluorescence lifetime (τ_f) of C102 first diminish steadily with increase in the concentration of phenol, but after a particular mole fraction (X_PhOH = 0.013) both the quantities (ϕ_f and τ_f) increase upon further increase in the phenol content. The results can be attributed to the competitive nature of the C102–phenol and the phenol–phenol H-bonding. Since the PET quenching operates through the H-bond between the C=O group of C102 and the HO– group of phenol, the H-bond may fail to attain an optimum geometry when the phenol is linked with other phenols via the H-bond. The variation of the C102–phenol H-bonding nature with the mole fraction of phenol was supplemented by FT-IR measurements. Similar unusual variation of the C102 fluorescence is observed in the phenol–anisole mixtures to a lesser extent, but is completely absent in the anisole–cyclohexane mixtures, where H-bonding is practically impossible.

---

1 Introduction

Competitive H-bonding may be considered as an important factor for governing chemical, photo-physical and photochemical responses in complex H-bonding environments. For example, in an enzyme pocket, a substrate forming a H-bond with a particular functional group at an active site, may also simultaneously make a H-bond with other amino acids.^1,2 These additional H-bonds may influence the biological activity and substrate turnover rate depending on the direction and the nature of H-bonding.^3,4 H-bonding has important consequences in controlling many photophysical processes e.g. photoisomerization,⁵ intramolecular charge transfer (ICT),⁶ proton transfer (PT),^7–9 photo-induced electron transfer (PET),^10,11 proton-coupled electron transfer (PCET)^12,13 and excited state intramolecular proton transfer (ESIPT).^14,15

Continuous variation of H-bonding environment in a solvent mixture may shed new light on the understanding of the competitive H-bonding impact on various photophysical processes. However, this aspect remains relatively unexplored in literature. Very recently, Saini et al. reported an anomalous fluorescence modulation of curcumin in toluene–methanol mixture; the emission intensity and lifetime gradually reach to a maximum at a methanol mole fraction of 0.14 and thereafter decrease.^14,15 The excited-state intermolecular H-bonding between the pigment and the H-bond donor solvent was proposed to reduce the non-radiative decay by breaking the intramolecular H-bonding between the keto and the enolic –OH group of curcumin.¹⁶ Interestingly, the intermolecular H-bonding interaction was most effective at some intermediate mole fraction rather than in neat methanol. For a doubly H-bonded proton transfer probe 2,2′-pyridine-3,3′-diol, Mandal et al. observed that the quantum yield varies abnormally with the mole fraction in a water–DMSO mixture displaying a maximum at X_DMSO = 0.12.¹⁷ Very recently, we reported that the H-bond assisted photoinduced electron transfer (PET) between coumarin 102 (C102) and aniline exhibits an unusual dependence on the mole fraction of aniline in aniline–cyclohexane and aniline–toluene mixtures.¹⁸ The electron donor (aniline) can form H-bond with the C=O group of the acceptor (C102) through the –NH₂ group. The aniline may also simultaneously associate with other aniline molecules forming H-bond at high concentrations in the mixture or in neat aniline. Although, the non-interacting component (cyclohexane or toluene) can not directly engage either in H-bonding or in PET, may alter polarity of the medium or may perturb the aniline–aniline H-bonding by its physical presence.¹⁸ Since C102 can involve in H-bond assisted PET quenching with phenol¹⁹ and as phenol can form self-associated H-bonding network at high concentrations,²⁰ we may anticipate to observe the similar unusual pattern showing a minimum in the variation of quantum yield or lifetime with the mole fraction of phenol. We will show that the prediction was indeed correct indicating the generality of our model of competitive H-bonding.¹⁸

The C102–phenol H-bonded system has been the subject of interest as a prototype of an ideal H-bonded complex to study H-bonding both in the ground and in the electronically excited state.^21–26 The C=O group of the coumarin dye is the primary site for intermolecular H-bonding with the phenolic –OH group. The C=O stretching frequency of C102 red-shifts from 1735 cm⁻¹ to 1695 cm⁻¹ on the addition of phenol in C₂Cl₄ indicating H-bond formation between C102 and phenol.²⁴ Using ultrafast UV-IR pump–probe spectroscopy, the pioneering work of Nibbering and co-workers showed that in the excited state, the C=O stretch of the H-bond complex reverts back to 1740 cm⁻¹ similar to the original stretching frequency of unbound C102.²³ In addition, they have also observed concomitant appearance of the –OH frequency of free phenol.²³ Hence, they ascribed the observation to the excited state H-bond breaking at ∼200 fs time scales. Later, Palit and co-workers have observed similar H-bonding dynamics of C102 with aniline, where they proposed that the intermolecular H-bond in the electronic excited state may initially break in the femtosecond time scales but recombine after few picoseconds of delay.²⁷ However, Han and coworkers subsequently demonstrated that the ultrafast vibrational spectral blue-shift of the C=O stretching band should be accounted to the electronic redistribution of the C102 rather than intermolecular H-bonding interactions.²⁵ In addition, they have also established for the first time that the H-bond between C102 and phenol is significantly strengthened, instead of being cleaved in the electronically excited state by computing the C=O and O–H stretching frequencies of the hydrogen-bonded complex.²⁵ Furthermore, Liu et al. reconsidered the electronic excited-state H-bonding dynamics of the photoexcited C102 in aniline and supported the excited-state intermolecular H-bond strengthening mechanism.²⁸ Nibbering and coworkers also acknowledged the excited-state intermolecular H-bond strengthening mechanism and utilized the mechanism to explain the PET dynamics between 9-fluorenone and amine solvents.²⁹ The very important role and effect of the electronic excited-state H-bond strengthening on various photophysical process such as photoinduced electron transfer (PET),³⁰ intramolecular charge transfer (ICT),³¹ metal-to-ligand charge transfer (MLCT),³² etc. has been highlighted in a series of works by the Han group.³³ They also predicted that H-bonded C102–phenol should undergo PET in the excited state.^25,34 We have verified the H-bond assisted PET of the C102–phenol complex in cyclohexane recently.¹⁹ The fluorescence of C102 was found to quench drastically on the addition of phenol in cyclohexane. The quenching was even more prominent for p-Cl-phenol for which stronger H-bonding is expected but was absent for anisole which can not form H-bond with C102.¹⁹ Although the works was an important demonstration of H-bond assisted PET but was limited to very low concentrations of phenol. Here, we have extended the work up to a much larger mole fractions of phenol to capture complete modulation of fluorescence with the H-bonding environment.

It is certain that H-bonding situation will modulate in a mixture of H-bonding and non-H-bonding constituents. However, in a solvent mixture, not only the H-bonding environment modulates with mole fractions, other parameters like polarity, π-stacking interactions may also vary. For the cyclohexane–phenol mixture, the polarity will increase with increase in the mole fraction of phenol as the polarity of phenol is much higher than cyclohexane (Table 1). Hence, the observed fluorescence modulation may be a consequence of the polarity change also. To check the proposition, we have performed the study in phenol–anisole mixture as well. As the chemical structures of phenol and anisole differ only by a methyl substitution, both have similar polarity (π* values are 0.72 and 0.73, respectively for phenol and aniline) but anisole does not possess H-bonding ability (α = 0) (Table 1). We may assume that both anisole and phenol may have comparable propensity of π–π interaction because of the common phenyl moiety, but the methyl groups may turn anisole less susceptible to π-stacking interaction with C102. Hence, fluorescence study of C102 in anisole–phenol and cyclohexane–anisole mixture will be a good indicator to monitor the effect of π-stacking interaction and polarity on fluorescence modulation.

Table 1 Kamlet–Taft parameters, α (hydrogen bond donor), β (hydrogen bond acceptor), and π* (polarity) of the solvents used³⁵

Solvent	α	β	π*
Cyclohexane	0	0	0
Phenol	1.65	0.30	0.72
Anisole	0.00	0.32	0.73

---

2 Experimental

Coumarin 102 (C102) and phenol were purchased from Sigma-Aldrich and used as received. Anisole and cyclohexane were supplied by Merk Chemicals.

FT-IR spectra were recorded by Perkin-Elmer Spectrum Two FTIR spectrophotometer. The UV-Vis spectra were recorded in a Perkin-Elmer Lamda-750 spectrophotometer. Emission spectra were collected on Jobin-Yvon FluoroMax4 spectrofluorometer. Time-resolved fluorescence measurements were performed on a time-correlated single photon counting (TCSPC) setup LifeSpec2 (Edinburgh Instruments). For excitation at 375 nm, picosecond laser diode EPL-375 (Edinburgh Instruments) was used. Typical FWHM of the set up was ∼60 ps. The fluorescence decays were fitted using FAST software. All measurements were performed at room temperature (25 °C).

3 Results

3.1 Steady state measurements

Fig. 1 displays the absorption spectra of C102 in three different mixtures viz. cyclohexane–phenol, anisole–phenol and cyclohexane–anisole at various mole fractions of phenol or anisole. Absorption maxima of C102 in neat cyclohexane and anisole were found to be at 362 nm and 384 nm, respectively. In cyclohexane, with increase in the mole fraction of phenol from 0.0 (i.e. neat cyclohexane) to 0.30, the absorption maximum of C102 gradually shifts towards the longer wavelength with the development of an additional band at ∼395 nm. Similar observation was also noticed for the C102–phenol in C₂Cl₄.³⁶ The new band developed upon gradual addition of phenol becomes more prominent at higher concentration of phenol with the subsequent reduction of the absorption band at 362 nm. This may be a manifestation of varying polarity and H-bonding ability of the medium. Another important feature is that up to 0.030 mole fraction of phenol a clear isosbestic point is observed at 382 nm. The isosbestic point may be an indication of the equilibrium between 1 : 1 C102–phenol complex and free C102. However, upon further enrichment of phenol the isosbestic region becomes defocused. This may be due to the formation of higher order C102–(phenol)_n complex (n = 2, 3, etc.). Likewise, the absorption spectrum of C102 in anisole modulates gradually in a manner that absorption of the higher wavelength side increases with a concomitant decrease in the lower wavelength side maintaining an isosbestic point at 392 nm up to a high mole fraction (0.10) of phenol. The progressive red-shift of absorption maximum may be due to the H-bond formation between C102 and phenol because polarity of the mixture may not be very different at various compositions. On the other hand, absorption spectra of C102 in cyclohexane–anisole mixture exhibit red-shift with increase in the mole fraction of anisole without any isosbestic point. Since anisole lacks the H-bond formation ability with the coumarin molecule, this red shift of absorption spectra may solely due to increase in the average polarity of the medium with increase in the mole fraction of anisole.

Fig. 1 Absorption spectra of C102 in (a) cyclohexane–phenol, (b) anisole–phenol, and (c) cyclohexane–anisole mixtures at different mole fractions of phenol (a) and (b) or anisole (c).

The emission maximum (λ^max_em) of C102 markedly depends on the polarity and H-bonding nature of the medium.^37,38 The λ^max_em of C102 in neat cyclohexane and anisole were at 408 nm and 435 nm, respectively (Fig. 2). In cyclohexane–phenol and anisole–phenol mixtures the emission spectra of C102 show gradual red-shift with increase in the mole fraction of phenol. However, the extent (i.e. slope) of the spectral shift is much higher at lower mole fraction of phenol compared to the higher mole fraction of the same. The variation of the emission maxima in frequency ν^max_em (in cm⁻¹) against mole fraction (x) can be conveniently fitted by the equation

ν^max_em (x) = ν^max_em (x = 1) − Δνe^−kx (1)

where ν^max_em (x = 1) is the emission frequency in neat phenol (or anisole) and Δν is the total shift of the emission maxima and k is a constant. The equation can be easily obtained if one assumes that the extent (i.e. slope) of energy lowering is proportional to the difference in energy of the mixture from that of the neat H-bonded or the polar solvent (i.e. at x = 1). The details of the fit parameters are given in the ESI.† It is clear that when the H-bonding is involved (in cyclohexane–phenol and phenol–anisole mixtures), the frequency shift levels off at very low mole fraction (Fig. 3). Conversely, when only polarity (in cyclohexane–anisole) is involved the frequency shift almost varies linearly with mole fraction (Fig. 2c and 3).

Fig. 2 Emission spectra of C102 in (a) cyclohexane–phenol (λ_ex = 370 nm), (b) anisole–phenol (λ_ex = 375 nm) and (c) cyclohexane–anisole (λ_ex = 375 nm). In (a) and (b) fluorescence intensity of C102 decreases with increase in the mole fraction of phenol upto a certain mole fraction, thereafter increases with further enrichment of phenol. In (c) fluorescence intensity monotonically decreases with increase in the mole fraction of anisole.

Fig. 3 Variation of the emission maxima, ν^max_em (cm⁻¹), of C102 in the three different solvent mixtures against the mole fraction of phenol, X_PhOH (or anisole, X_anisole).

With increase in the mole fraction of phenol, the fluorescence intensity shows anomalous trend first decreases up to a certain mole fraction of phenol and thereafter increases with increase in the mole fraction of phenol (Fig. 4). For the cyclohexane–phenol and anisole–phenol mixtures, the critical mole fractions of phenol were found at 0.013 and 0.027, respectively. However, in the cyclohexane–anisole mixture, the fluorescence intensity shows a slight linear decrease with increase in the mole fraction of anisole (Fig. 4). This regular trend of fluorescence intensity change may be due to change in the polarity of the medium with increase in the mole fraction of anisole. Since anisole lacks the H-bonding ability, the absence of anomalous fluorescence variation in cyclohexane–anisole mixture hints at the possible link of the H-bonding modulation in phenol with the observed anomaly of the fluorescence quenching of C102 in cyclohexane–phenol and anisole–phenol mixtures.

Fig. 4 Variation of the quantum yield (QY) of C102 in different mixtures with mole fraction of phenol or anisole.

3.2 Time resolved measurements

The anomalous trend of steady state fluorescence variation was also evident from the lifetime measurements. Fig. 5 and 6 display the fluorescence decays of C102 in the three mixtures – cyclohexane–phenol, anisole–phenol and cyclohexane–anisole measured at the respective emission maxima at different mole fractions of phenol (or anisole). The fluorescence decays of C102 were found to be single exponential in neat solvents – cyclohexane and anisole with the lifetimes of 2.72 ns and 3.3 ns, respectively.

Fig. 5 Fluorescence decays of C102 in cyclohexane–phenol mixture at λ_ex = 375 nm at different mole fractions of phenol. Left panel represents the fluorescence decay in lower mole fraction (0.0 to 0.013), whereas the right panel represents the decay in higher mole fraction (0.013 to 0.21). Fluorescence decays become faster at low concentration region but the trend reverses in the higher mole fraction region.

Fig. 6 Fluorescence decays of C102 in anisole–phenol (left panel) and cyclohexane–anisole (right panel) mixtures at λ_ex = 375 nm at different mole fraction of phenol and anisole, respectively.

However, the fluorescence decays of C102 in the presence of phenol were found to be bi-exponential with two distinct time components in both the solvents – cyclohexane and anisole (Tables 2 and 3). The average decay time of C102 gradually decreases up to the same (X_PhOH = 0.013) mole fraction at which quantum yield was minimum (Fig. 7). The amplitude of the faster component (370–860 ps) gradually increases up to this critical mole fraction in the case of cyclohexane–phenol mixture and thereafter decreases with increase in the concentration of phenol (Table 2). For anisole–phenol mixture faster component follows the similar trend (Table 3). The fast component may be due the PET of the H-bonded complex whereas the slower component may be due to either an inappropriately oriented H-bonded complex or a non-H-bonded complex. However, in the case of cyclohexane–anisole mixture there is no significant change of the average fluorescence lifetime of C102 with increasing the mole fraction of anisole (Table 4).

Table 2 Emission maxima (λ^max_em), quantum yield (ϕ) and lifetimes (τ) of C102 in the cyclohexane–phenol mixture at different mole fraction of phenol. Excitation wavelength was at 370 nm. Lifetime measurements were done at corresponding emission maxima (λ^max_em)

XPhOH	λ^maxem (nm)	QY (ϕ)	Lifetimes (τ/ns)		〈τ〉 (ns)
			τ1 (a1)	τ2 (a2)
0.0	408	1.0	—	2.72 (1.0)	2.72
0.0005	409	0.88	0.38 (0.24)	2.66 (0.76)	2.18
0.0012	409	0.81	0.37 (0.33)	2.5 (0.63)	1.83
0.0024	409	0.66	0.46 (0.43)	2.3 (0.57)	1.52
0.004	409	0.58	0.45 (0.51)	2.2 (0.49)	1.27
0.0054	410	0.47	0.42 (0.61)	1.9 (0.39)	1.00
0.009	411	0.38	0.40 (0.68)	1.6 (0.32)	0.79
0.013	423	0.31	0.44 (0.74)	1.5 (0.26)	0.72
0.021	432	0.35	0.77 (0.87)	2.5 (0.13)	0.98
0.03	436	0.39	0.83 (0.84)	2.9 (0.16)	1.16
0.055	449	0.47	0.83 (0.65)	3.8 (0.35)	1.88
0.080	454	0.53	0.86 (0.57)	3.9 (0.43)	2.18
0.10	459	0.58	0.88 (0.53)	4.0 (0.47)	2.36
0.21	460	0.68	0.58 (0.49)	4.9 (0.51)	2.77
0.30	472	0.77	0.58 (0.48)	5.1 (0.52)	2.87

---

Table 3 Emission maxima (λ^max_em), quantum yield (ϕ) and lifetimes (τ) of C102 in the anisole–phenol mixture at different mole fraction of phenol. Excitation wavelength was at 375 nm. Lifetime measurements were done at corresponding emission maxima (λ^max_em)

XPhOH	λ^maxem (nm)	QY (ϕ)	Lifetimes (τ/ns)		〈τ〉 (ns)
			τ1 (a1)	τ2 (a2)
0.0	435	0.84	—	3.3 (1.0)	3.3
0.0007	436	0.82	0.97 (0.05)	3.2 (0.95)	3.16
0.004	439	0.81	0.86 (0.09)	3.1 (0.91)	2.97
0.009	442	0.79	1.70 (0.14)	3.2 (0.86)	2.96
0.013	446	0.76	1.90 (0.28)	3.6 (0.72)	2.94
0.027	452	0.69	1.10 (0.32)	3.7 (0.68)	2.84
0.052	455	0.75	0.85 (0.30)	3.8 (0.70)	2.92
0.10	462	0.84	0.74 (0.18)	4.0 (0.82)	3.42
0.21	466	0.85	0.54 (0.18)	4.3 (0.82)	3.65

---

Fig. 7 Variation of the average fluorescence lifetime, (〈τ〉) of C102 in different mixtures with mole fraction of phenol or anisole.

Table 4 Emission maxima (λ^max_em), quantum yield (ϕ), and lifetimes (τ) of C102 in the cyclohexane–anisole mixture at different mole fraction of anisole. Excitation wavelength was at 375 nm. Lifetime measurements were done at corresponding emission maxima (λ^max_em)

Xanisole	λ^maxem (nm)	QY (ϕ)	Lifetimes (τ/ns)		〈τ〉 (ns)
			τ1 (a1)	τ2 (a2)
0.0	408	1.0	—	2.72 (1.0)	2.72
0.1	415	0.97	0.29 (0.20)	3.0 (0.80)	2.60
0.2	418	0.96	0.23 (0.18)	3.1 (0.82)	2.54
0.3	420	0.95	0.31 (0.18)	3.1 (0.82)	2.60
0.4	422	0.94	0.37 (0.20)	3.2 (0.80)	2.60
0.5	425	0.92	0.34 (0.19)	3.2 (0.79)	2.64
0.6	429	0.90	0.34 (0.23)	3.2 (0.77)	2.55
0.7	431	0.89	0.39 (0.23)	3.2 (0.77)	2.58
0.8	432	0.88	0.31 (0.18)	3.3 (0.82)	2.70
0.9	433	0.86	0.34 (0.17)	3.3 (0.83)	2.80
1.0	435	0.84		3.3 (1.0)	3.3

---

3.3 FT-IR measurements

To monitor the ground state H-bonding nature of C102 with phenol in cyclohexane, we have measured FT-IR of C102 in the C=O stretching frequency region at several mole fractions of phenol (Fig. 8).

Fig. 8 FTIR spectra of C102 in the presence of different mole fractions of phenol in cyclohexane. The C=O stretching frequencies were 1739 cm⁻¹, 1697 cm⁻¹, 1702 cm⁻¹, 1702 cm⁻¹ and 1704 cm⁻¹ at 2 mM of C102, 0.0055, 0.013, 0.030 and 0.30 mole fraction of phenol, respectively.

In the non-interacting cyclohexane, the C=O stretching frequency of C102 was found to be at 1739 cm⁻¹ which is the characteristics of unbound C=O group.^24,27 On the addition of 50 mM (mole fraction 0.0055) of phenol, the C=O stretching frequency undergoes red-shift to 1697 cm⁻¹. The marked red-shift by 42 cm⁻¹ may indicate H-bond formation with phenol. Because of the low phenol concentration, we may assume predominantly 1 : 1 C102–phenol formation. However on further addition of phenol, the C=O, the stretching frequency shift reduces by 4–5 cm⁻¹. As the shift of carbonyl frequency is related to the H-bonding strength, we may attribute to the lowering of the shift to the weakening of the H-bond strength at higher mole fractions. However, due to change in the spectral shape and possible effect of π-stacking interaction on the carbonyl stretching frequency, we can not do any quantitative estimate of H-bonding energy or binding constant.

4 Discussion

We have shown earlier that phenol induces a non-radiative relaxation pathway for C102 deactivation via H-bond assisted PET, thus acts as a potential PET quencher to C102 fluorescence.¹⁹ Generally, addition of more quencher leads to enhanced quenching, however, we observed here reduced fluorescence quenching of C102 fluorescence at phenol-rich mole fractions in cyclohexane–phenol (or anisole–phenol) solvent mixture. To explain the observation, we need to first recall the mechanism of the H-bond controlled relaxation pathways of the C102–phenol H-bonded complex.

As predicted by the TDDFT calculations of Zhao et al.²⁵ and also supported by our earlier study,¹⁹ the PET within the H-bonded complex occurs via a transition from a locally excited (LE, S₂) state to a low-lying charge transfer (CT, S₁) state. The CT state is only supported in the H-bonded complex and the gap between the LE and the CT state depends on the strength of H-bonding. Similar H-bond assisted PET was also proposed for C102–aniline complex by Liu et al.²⁸ which was experimentally verified by us.¹⁸ Note that anisole and N,N-dimethylaniline (DMA) are analogues of phenol and aniline respectively, lacking the H-bonding ability and thus, both are unable to quench the C102 fluorescence.

At the lower concentration region, one may assume that C102 exist either as 1 : 1 H-bonded complex or as free. Since only the H-bonded complex undergoes PET, with increase of phenol, the fraction of the C102–phenol increases resulting in higher quenching. We have shown that for initial few mole fractions, the quenching behaviour follows a linear Stern–Volmer trend.¹⁹ The slope of the plot was found to depend on the strength of intermolecular H-bonding indicating the importance of H-boning in activating the PET. Here, we found that the linear quenching trend is only valid over initial few mole fractions. At higher mole fractions, the quenching markedly reduced. Several possible explanations may be discussed to account for the observation.

It is reasonable to assume that at higher concentration, a phenol donor that is linked with the C102, additionally coordinate to other phenols forming a H-bonded cluster, C102–(phenol)_n≥2. These phenol–phenol H-bonding may weaken or hinder the main C102–phenol H-bonding to attain optimum condition required for PET. Hence, the key C102–phenol H-bonding may be less effective in guiding electron transfer. Thus, the competitive nature of the C102–phenol and phenol–phenol H-bonding may reduce the PET quenching. The non-H-bonding component may assist to perturb or weaken the phenol–phenol H-bond. At the critical mole fraction where we observed the strongest fluorescence quenching the C102–phenol, H-bond may be the most effective to undergo PET. Similar competitive H-bonding model has been introduced by us to explain the quenching anomaly of C102 in cyclohexane–aniline¹⁸ and aniline–DMA³⁹ mixture.

Another important concern is possible modulation of polarity in a mixture of phenol–cyclohexane because phenol is more polar than cyclohexane. Polarity may have important consequence over the PET energy and on the C102–phenol H-bonding. It is generally considered that H-bonding in a non-polar medium is much stronger than in polar solvents due to solvation.⁸ The polarity of the phenol-rich mixture may be higher and hence the H-bonding between coumarin and phenol may be weaker. Stronger H-bonding in a less polar medium compared to the high polar medium is consistent with the higher emission shift observed at lower mole fraction. Additionally, the relative stability of the LE and CT excited states may also depends on polarity. The CT state may be stabilized in polar medium to a much higher degree compared to the LE state due to larger dipole moment of the CT state. At lower mole fraction region i.e. in low polarity medium, H-bonding is strong but the CT state may not be energetically favourable and hence, fluorescence from the LE state is dominant. At higher mole fraction, due to higher polarity, the CT state is stabilized but the H-bonding is weak and hence, PET may occur mostly via a non-H-bonding pathway, if allowed. At a critical mole fraction, both the factors may have an optimum influence and the PET rate is the highest.

To discriminate between the two possibilities (competitive H-bonding vs. polarity induced H-bonding modulation), we have chosen anisole–phenol solvent mixture. Due to similar polarity of anisole and phenol, the polarity of the mixture remains nearly invariant throughout all the compositions. Hence, the PET modulation due to change in the polarity of the mixture can be neglected. Since anisole lacks H-bond donating ability, it can not compete with the C102–phenol H-bond but it may form H-bond with phenol and thus, reduce phenol–phenol H-bonding networking when present in large quantities. We observe similar reduced quenching as higher mole fraction for phenol–anisole mixture. Both the fluorescence quantum yield and average fluorescence lifetime gradually decrease up to the critical mole fraction and thereafter increase. As polarity remains invariant, we may conclude that competitive H-bonding has an important role in the observed anomalous quenching behaviour. Very recently, we have also observed such unusual fluorescence quenching modulation of C102 in similar polarity aniline–DMA mixture.³⁹

To check the effect of polarity only and to exclude H-bonding possibility, we have considered another solvent mixture – cyclohexane–anisole. Here quantum yield of C102 decreases linearly with increase in the mole fraction of anisole. The slight linear decrease may be due to polarity change of the medium. From this we can say that H-bond assisted PET has a pronounced effect on the anomalous fluorescence modulation of C102 in the two solvent mixtures – cyclohexane–phenol and anisole–phenol.

Now an important question is whether the fluorescence modulation is due to the H-bonding modulation in the ground state or in the excited state? The FT-IR spectroscopic investigation shows that H-bonding between C102–phenol is indeed modulated differentially at low-phenol and high-phenol regions in the ground state. At low mole fraction (0.0055) of phenol in cyclohexane, the C=O stretching frequency of C102 exhibits a marked red shift of 42 cm⁻¹ indicating a strong H-bond formation with phenol. However at higher phenol concentrations, the red shift reduces by 4–5 cm⁻¹. Thus, it may be inferred that the C102–phenol H-bond becomes weaker at higher donor concentration because of self association of several donors. This result clearly illustrates that H-bonding strength is indeed depend on the mole fraction of phenol. However, for fluorescence modulation, we need to consider the H-bonding in the photoexcited state. We have shown that PET in the non-H-bonded complex is negligible and hence, significant PET in phenol containing mixtures implies that H-bonding should retain in the excited complex. Thus, we can indirectly infer that H-bonding between C102 and phenol may strengthen in the excited state, which is consistent with the prediction by Han and co-workers.²⁵ Another minor possibility is that the H-bond may initially break within few hundreds of femtoseconds of excitation but subsequently, may recombine on a picosecond time scale to induce the PET process. Similar model of H-bond dynamics was proposed by Palit and co-workers for the case of C102–aniline H-bond dynamics.²⁷ Overall, our results show the dominant role of H-bonding between C102 and phenol for controlling the anomalous fluorescence quenching within the C102–phenol H-bonded complex.

5 Conclusions

The H-bond mediated PET between C102–phenol is found to be more activated at a lower mole fraction of the H-bond donor rather than at higher mole fraction of the donor (phenol). It is proposed that at higher mole fraction, C102 may form higher order C102–(phenol)_n≥2 complex where, the phenol may associate with more phenol molecules through phenol–phenol H-bond. The phenol–phenol H-bond may perturb the H-bond between the C=O group of C102 and the HO– group of phenol to attain optimum geometry for PET. This competitive nature of H-bonding that reduces the H-bonding strength between C102 and phenol makes the complex less effective to undergo PET. This was further complemented by the FTIR measurement, which shows higher red-shift of the C=O frequency at low phenol concentration compared to the higher phenol concentrations.

Acknowledgements

This work is supported by Indian Institute of Technology Guwahati. The TCSPC measurements were carried out in the Central Instrument Facility (CIF). We also thank the Department of Science and Technology (DST), India for the project EMR/2014/000011.

Notes and references

1. J. P. Schwans, F. Sunden, A. Gonzalez, Y. Tsai and D. Herschlag, Biochemistry, 2013, 52, 7840–7855.
2. I. Dance, Dalton Trans., 2012, 41, 7647–7659.
3. Q. Cui and M. Karplus, J. Phys. Chem. B, 2002, 106, 1768–1798.
4. A. Shokri, Y. Wang, G. A. O'Doherty, X.-B. Wang and S. R. Kass, J. Am. Chem. Soc., 2013, 135, 17919–17924.
5. M. Boggio-Pasqua, M. A. Robb and G. Groenhof, J. Am. Chem. Soc., 2009, 131, 13580–13581.
6. F. A. S. Chipem, A. Mishra and G. Krishnamoorthy, Phys. Chem. Chem. Phys., 2012, 14, 8775–8790.
7. S. Chai, G.-J. Zhao, P. Song, S.-Q. Yang, J.-Y. Liu and K.-L. Han, Phys. Chem. Chem. Phys., 2009, 11, 4385–4390.
8. N. Ikeda, H. Miyasaka, T. Okada and N. Mataga, J. Am. Chem. Soc., 1983, 105, 5206–5211.
9. Y. Nagasawa, A. P. Yartsev, K. Tominaga, P. B. Bisht, A. E. Johnson and K. Yoshihara, J. Phys. Chem., 1995, 99, 653–662.
10. H. Pal, Y. Nagasawa, K. Tominaga and K. Yoshihara, J. Phys. Chem., 1996, 100, 11964–11974.
11. D. Yang, Y. Liu, D. Shi and J. Sun, Comput. Theor. Chem., 2012, 984, 76–84.
12. C. Costentin, M. Robert, J.-M. Saveant and C. Tard, Phys. Chem. Chem. Phys., 2011, 13, 5353–5358.
13. M. Sjödin, T. Irebo, J. E. Utas, J. Lind, G. Merényi, B. Åkermark and L. Hammarström, J. Am. Chem. Soc., 2006, 128, 13076–13083.
14. R. K. Saini and K. Das, J. Lumin., 2013, 144, 169–175.
15. R. K. Saini and K. Das, J. Lumin., 2014, 145, 832–837.
16. R. K. Saini and K. Das, J. Phys. Chem. B, 2012, 116, 10357–10363.
17. S. Mandal, S. Ghosh, C. Banerjee, J. Kuchlyan and N. Sarkar, J. Phys. Chem. B, 2013, 117, 12212–12223.
18. N. Barman, D. Singha and K. Sahu, Phys. Chem. Chem. Phys., 2014, 16, 6159–6166.
19. N. Barman, D. Singha and K. Sahu, J. Phys. Chem. A, 2013, 117, 3945–3953.
20. V. Poterya, L. Sistik, P. Slavicek and M. Farnik, Phys. Chem. Chem. Phys., 2012, 14, 8936–8944.
21. C. Chudoba, E. T. J. Nibbering and T. Elsaesser, J. Phys. Chem. A, 1999, 103, 5625–5628.
22. C. Chudoba, E. T. J. Nibbering and T. Elsaesser, Phys. Rev. Lett., 1998, 81, 3010–3013.
23. E. T. J. Nibbering, F. Tschirschwitz, C. Chudoba and T. Elsaesser, J. Phys. Chem. A, 2000, 104, 4236–4246.
24. E. T. J. Nibbering, C. Chudoba and T. Elsaesser, Isr. J. Chem., 1999, 39, 333–346.
25. G.-J. Zhao and K.-L. Han, J. Phys. Chem. A, 2007, 111, 2469–2474.
26. G.-J. Zhao and K.-L. Han, Acc. Chem. Res., 2011, 45, 404–413.
27. D. K. Palit, T. Zhang, S. Kumazaki and K. Yoshihara, J. Phys. Chem. A, 2003, 107, 10798–10804.
28. Y. Liu, J. Ding, D. Shi and J. Sun, J. Phys. Chem. A, 2008, 112, 6244–6248.
29. H. N. Ghosh, K. Adamczyk, S. Verma, J. Dreyer and E. T. J. Nibbering, Chem.–Eur. J., 2012, 18, 4930–4937.
30. G.-J. Zhao, J.-Y. Liu, L.-C. Zhou and K.-L. Han, J. Phys. Chem. B, 2007, 111, 8940–8945.
31. G.-J. Zhao and K.-L. Han, Biophys. J., 2008, 94, 38–46.
32. G.-J. Zhao, B. H. Northrop, K.-L. Han and P. J. Stang, J. Phys. Chem. A, 2010, 114, 9007–9013.
33. G.-J. Zhao and K.-L. Han, Acc. Chem. Res., 2011, 45, 404–413.
34. G.-J. Zhao, J.-Y. Liu, L.-C. Zhou and K.-L. Han, J. Phys. Chem. B, 2007, 111, 8940–8945.
35. R. Stenutz, Kamlet–Taft solvent parameters – Home page of R. Stenutz, http://www.stenutz.eu/chem/solv26.php.
36. F. Tschirschwitz and E. T. J. Nibbering, Chem. Phys. Lett., 1999, 312, 169–177.
37. R. S. Moog, W. W. Davis, S. G. Ostrowski and G. L. Wilson, Chem. Phys. Lett., 1999, 299, 265–271.
38. G. Jones, W. R. Jackson, C. Y. Choi and W. R. Bergmark, J. Phys. Chem., 1985, 89, 294–300.
39. N. Barman and K. Sahu, Phys. Chem. Chem. Phys., 2014,  10.1039/c4cp04441j.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c4ra11251b

---