21 November 1997

CHEMICAL PHYSICS LETTERS ELSEVIER

Chemical Physics Letters 279 (1997) 303-308

Determination of the nature of the lowest triplet state of the intramolecular charge-transfer probes DMABN and DMABA by laser-induced optoacoustic spectroscopy Nitin Chattopadhyay ~, Mark Van der Auweraer, Frans C. De Schryver Lat~oratoO" ~!]'Molecular Dynamics and Spectroscopy, Department ~ Chemistry, Katholieke Unil,ersiteit Leu~,en, 200 F Cele.~tijnenlaan, B-3001 Hererlee, Belgium Received 6 March 1997: in final form 15 September 1997

Abstract

The possible electrostatic volume constriction associated with the excited state photophysics of the triplet states of the intramolecular charge transfer probes p-N,N-dimethylaminobenzonitrile and p-N,N-dimethylaminobenzaldehyde has been studied in apolar media by laser-induced optoacoustic spectroscopy, using a series of normal alkanes ranging fiom pentane to hexadecane as solvents. A nanosecond laser was used for excitation and the optoacoustic signal was determined by a piezoelectric transducer. Hardly any volume change is observed, suggesting that, in these solvents, the lowest triplet states of both p-N,N-dimethylaminobenzonitrile and p-N,N-dimethylaminobenzaldehyde do not have appreciable charge transfer character. © 1997 Elsevier Science B.V.

1. Introduction

The intramolecular charge transfer (ICT) fluorophores, since their discovery by Lippert [1], attracted the attention of photochemists for their wide variety of applications [2]. The mechanism of the excited state charge transfer process is still controversial [3,4]. For molecules such as p-N,N-dimethylaminobenzonitri[e (DMABN), it is often found that upon increasing the solvent polarity, a low-energy emission corresponding to the [CT state grows at the cost of the high-energy emission of the locally excited

Permanent address: Department of Chemistry, Jadavpur Universily, Calcuna 700032, India.

state (LE). However, the overall fluorescence quantum yield (q'~.), decreases. This decrease in the radiative yield has been explained by an enhancement of the non-radiative decay of the excited fluorophore to the ground state through the stabilized ICT state [5-7]. In a recent laser-induced optoacoustic spectroscopy (LIOAS) study [8] on D M A B N and pN,N-dimethylaminobenzaldehyde (DMABA) indications were found that while klc, the rate constant for the internal conversion (IC) process, becomes more important, kls c, the rate constant for intersystem crossing (ISC), decreases as one moves towards more polar solvents. However, their sum does not change much, indicating that the decrease in @r is mainly due to a decrease in k~, the fluorescence rate

0009 2 6 1 4 / 9 7 / $ 1 7 . 0 0 © 1997 Elsevier Science B.V. All rights reserved. Pll S 0 0 0 9 - 2 6 1 4 ( 9 7 ) 0 1 0 7 1 - 3

304

N. Chattopadhyay et al. / Chemical Physics Letters 279 (1997) 303-308

constant. This study also revealed that for both molecules, ISC is the principal deactivation channel irrespective of the solvent polarity (up to acetonitrile). This emphasizes the participation of the triplet state in the photophysics of those compounds. Involvement of the higher n - v * triplet state has also been proposed for DMABA, particularly in apolar solvents [9]. Although time-resolved infrared spectroscopy indicates that, for DMABN, the amount of the ICT triplet formed is proportional to the amount of the ICT singlet [10], nanosecond transient studies indicate that the lowest triplet does not have a strong CT character [11]. Drickamer et al. have shown that the lowest triplet state of DMABA is of n-'rr * character in a poly(ethyl methacrylate) film at atmospheric pressure. Under pressure, the rr,rr* character increases, changing its emitting properties [12]. From phosphorescence studies in condensed phases, it is known that the nature of the emitting triplet state may change from n-'rr * to ~-'rr * on increasing the solvent polarity [9]. Thus, the nature of the lowest excited triplet states for the two fluorophores mentioned above is yet to be established, particularly in fluid solution, where phosphorescence spectra cannot be easily obtained. Photothermal techniques, in particular laser-induced optoacoustic spectroscopy (LIOAS), can yield information about the volume changes associated with an excited state photoprocess [13-15]. There are two possible contributions to the volume change (AV) in a photoinduced reaction in solution: (a) the change in intrinsic reaction volume (AVe), which is determined by the difference in molar volume between the products and the reactants; and (b) the expansion of the medium upon the release of heat (AVth). AVr is independent of the thermoelastic properties of the environment, at least to the first approximation. Only AV~h depends on the thermoelastic parameters of the solution (Ce, the specific heat capacity at constant pressure; fl = ( O V / 3 T ) p ( I / V ) , the volume expansion coefficient; and p, the solution density) [13]. Optoacoustic measurements as a function of these parameters allow the separation of both terms [13,14] and reflect the nature of the lowest triplet states. The present Letter reports our studies relating to the solvent-dependent LIOAS of DMABN and DMABA, which indicates that the lowest excited

triplets of the fluorophores do not have any appreciable CT character in apolar solvents.

2. Experimental The compounds DMABN, DMABA and 2-hydroxybenzophenone (HBP) (all from Aldrich) were recrystallized twice from methanol. The purity of the compounds was checked by absorption and fluorescence spectroscopy as well as with TLC. The normal alkanes, pentane (PEN), hexane (HEX), heptane (HEP), octane (OCT), nonane (NON), decane (DEC), dodecane (DODE), and hexadecane (HEXD), were all high-purity solvents from Fluka and were used as received. HBP was used as the reference compound [c~ (fraction of the total excitation energy liberated as heat) = 1.0] for the photoacoustic measurements. In each solvent, three solutions of the probes and HBP, with optical densities 0.08, 0.10 and 0.12 were prepared. The solutions were degassed by bubbling with argon for 15 rain. The H / H R values (see below) were determined for the three solutions in a particular solvent and the mean value was used. The fluorescence quantum yields of DMABN, were determined using quinine sulphate solution in 0.05 M H2SO 4 as a reference (q~f = 0.54) [16]. The absorption and fluorescence spectra were recorded on a Perkin-Elmer Lambda 6 U V / V i s spectrophotometer and a Spex Fluorolog fluorimeter, respectively. The LIOAS apparatus has been described previously [17-19]. The 320 nm excitation pulse was provided by an Nd:YAG/rhodamine 6 G laser system with 10 Hz repetition rate. The laser beam diameter was set with a pinhole of 1 mm diameter, so that the effective acoustic transit time (% = d/va; d = diameter of the beam, Va = sound velocity) was < 1 I.zs for all solutions. 5% of the energy of the laser beam was reflected to the pyroelectric energy meter (Molelectron Rj 7100) by a beam splitter placed before the point where the beam entered the cell. The energy fluency of the laser pulses was varied by a variable neutral density filter. The range of laser energy used spanned from 0 to 30 IxJ. The sound wave was detected with a PZT (Panametrics) detector attached to the wall of the 1 cm path length

N. Chattopadhyay et al. / Chemical Physics Letters 279 (1997) 303-308

quartz cuvette. The average of 100 signals was stored in a transient recorder (Gould 4072 oscilloscope).

305

first put forward by Callis [22], the relevant equations will be discussed. The thermally produced volume change, AVth can be related to the volume expansion coefficient, /3, by

3. Results and discussion

±v,h= /3wxr= The fluorescence spectrum of DMABN in all alkane solvents consists of a single band whose maximum and features are within the experimental error independent of the chain length of the alkane. The fluorescence quantum yields (qbf,DMAm~), do not change strongly (Table 1), although they show a slight increase upon moving towards the higher alkanes. DMABA, on the other hand, is almost non fluorescent in alkanes [8]. Hence, the non-radiative processes are the principal channel for the energy dissipation from both photoexcited fluorophores. Assuming AV, = 0, a quantitative estimation of the individual non-radiative channels IC and ISC) yields the values of qb~c = 0.14 and qblSc = 0.73 for DMABN in HEP, while qblsc is ~ 1.0 for DMABA [8] in this solvent. The two terms constituting the volume changes associated with a photoinduced chemical reaction, i.e. AV, and AVth, can be separated by photohermal measurements as a function of the thermoelastic parameters of the medium [ 13,14]. Thus, the intrinsic volume change associated with the given reaction can be estimated. Before exploiting this principle,

Table 1 Fluorometric, LIOAS, and thermoelastic data for DMABN and DMABA in different alkanes

ISCAPP R Solvent (/)LDMABN~DMABN Cp p / j~a H//HDMABN H~ HRMABA ( J / m ~) PEN HEX HEP OCT NON DEC DODE HEXD

0.13 0.12 0.13 0.13 0.15 0.15 0.15 0.16

0.64 0.68 0.6g 0.62 0.70 0.59 0.61 0.63

853.62 0.39 1101.96 0.37 1 2 2 6 . 5 7 0,36 1 3 4 9 . 7 9 0,41 1 4 3 3 . 5 0.32 1566.12 0.41 1720.59 0.39 1 9 2 4 . 6 1 0.37

0.26 0.29 0.27 0.28 0.26 0.27 0.27 0.26

ISCAPP is the apparent quantum yield of intersystem crossing, @t~MaUN assuming that the optoacoustic signal is completely due to the thermal effect. ~'Refs. [20,21].

I3/( c p) ,

(1)

where c~ is the fraction of the absorbed energy evolved as 'prompt' heat, E~ is the absorbed energy and c t, is the specific heat capacity of the solution at constant pressure. The intrinsic volume change AVr, originating from the molecular photoprocess occurring with a quantum yield q~r, which is to be added to AVth, is, to the approximation, independent of the temperature change. Hence. the equation describing the behavior of the optoacoustic signal of a sample solution, including also this volume change, should be given by

n = k[

r,) + a'r Vr],

(2)

where k reflects the geometrical and instrumental parameters. For dilute solutions, with low absorbances, E~ is proportional to the laser energy ( E o) and the absorbance (A) of the solution ( E,, = E~( I 10 a)) [13]. For Eq. (2), it is assumed that no multiphoton effects or ground state depletion occur. This can be, and was verified by the linear dependence of the optoacoustic signal on the energy of the laser pulse. For the reference solution, where AVr = 0 and oe = 1.0, Eq. (2) turns into Eq. (3):

H"=

,.,,p).

(3)

When the sample solution has the same absorbance as the reference solution, one obtains from the ratio of Eqs. (2) and (3),

.R =

+

E--7--,

(4)

From Eq. (4), it is clear that a plot of the fluencynormalized optoacoustic signal versus the thermoelastic parameter of the solvent, given by cpo//3 should be linear, the intercept giving the value of a. From the slope one can determine the volume change associated with the photoprocess.

N. Chattopadhyay et al. / Chemical Physics Letters 279 (1997) 303-308

306

The values of H / H R for D M A B N and D M A B A in different alkane media were estimated using HBP as reference. Table l gives the experimental LIOAS data and the thermoelastic parameters of the alkane solvents. The plots of the H / H R values against the thermal expansivity ( c e p / f l ) show a straight line parallel to the axis representing the solvent expansivity parameter (Fig. 1). This indicates that there is no substantial volume change associated with the formation of the transient state. The fact that qbf of D M A B N varies slightly with the alkanes makes, in principle, the value of H / H R also a function of the solvent through the solvent dependence of ce. However, the small variation of qbt of between 12% and 16% (leading to an even smaller variation of ce) does not basically change o~ upon changing the alkane solvents. Nevertheless, this effect was considered by modifying Eq. (4) to Eq. (5) and using the values of [ H / H R + ~.(Ef/Eo)] determined for different alkanes: . .

Plotting [ H / H R + (be( Ef/Eo)] versus Cp p/t9 yields a straight line parallel to the abscissa. This suggests that either q~,AV,/E~ equals zero, or that the dependence of qbls c Ew/E o on cv p / ~ is within the experimental error, equal and opposite to that of (ClgrAVJE,)(cpp/~). For an excited state dipole moment of 20 debye (which is much larger than the dipole moment expected for the excited triplet state)

0.6

C] DMAFIN

--

/h DMABA

. . . . . . Linear regression (DMABA)

Linear regression (DMABN)

05

0.4

f~

u

[]

[]

~9

~9

u

[] 03

/4/h~

~" . . . . . . . . . . . . . . . . & . . . . . . ~X . . . . . . k ,

"lX. . . . . . . ~, . . . . . . . . . ~ . . . . . . . . . . .

A

0.2

01 0 800

1000

1200

cVp/[5o f

1400 alkanes

1600

1800

2000

(J/ml)

Fig. 1. Plot of H/H ~ against cpp/~ for DMABN and DMABA in various alkanes. Points from left to right correspond to solvents PEN, HEX, HEP, OCT, NON, DEC, DODE and HEXD, respectively. The error bars are omitted for the clarity of the figure.

and assuming a cavity radius of 4.0 × 10 ~°m, clglscE.r/E o would change by ~ 3.9 × 10 -3, which is clearly within the experimental uncertainty. For DMABA, the latter possibility becomes even less likely, as for all solvents, characterized by different values of Cv O/fi, the values of @isc ET/Eo should be equal to those of (clgrAVJE,)(cpp/fi) using an intersystem quantum yield of 1.0 + 0.05 in all solvents. Hence, for both probes it is unlikely that the independence of the apparent value of a on the ratio Cpp/fl, is due to an accidental cancellation of the changes of the thermal and the volume contributions to H / H R. Starting from S o , all molecules have relaxed to S 0, l ~s after excitation, except for the molecules intersystem crossing to T~, the lowest triplet state. Hence, the value of qOIscAVr/E a should only correspon_d to qblsc(VT, -- ~Zs¢)/NAht~ or cIglscA~/NAhP, the V terms representing the molar volumes in the corresponding states indicated by the subscripts, and change of the molar volume. NA is the Avogadro constant and h v is the energy of the absorbed photon. The experimental error in the determination of the measured parameter ( H / H R ) , as well as the experimentally accessible range of Cp P/B, determines the minimum values of the molar volume change that can be detected. In the present case, the minimum change in H / H R detected over the series of alkane solvents would be 0.045 and 0.016 for D M A B N and DMABA, respectively. This corresponds to a change in the molar volumes of 1 . 4 × 1 0 5 and 5 . 0 × 1 0 - 6 m 3 mol-~ for D M A B N and DMABA, respectively. The values of AVr include all conformational changes in the excited molecule due to the rearrangement of the solvent around the fluorophore (AV t) and intramolecular changes in bond lengths or bond angles (AV 2). One could expect an appreciable electrostriction of the solvent around the fluorophore, if a highly polar triplet were be formed, resulting in a negative value of AP t . For a molecule with an excited state dipole moment /z E in a spherical cavity with radius a, AV I can be estimated using Eq. (6) [23,24]:

. . . . . . ~--~), 4,rr~0a3(2Er + 1) 2 (dEr

(6)

N. Chattopadhyay et al. / Chemical Physics Letters 279 ( 1997; 303-308

where /*a and /*E represent the dipole moment (C m) of the ground and excited states, respectively, Er is the relative dielectric constant of the solution and % is the permittivity of vacuum (8.85 × 10 -~2 C 2 J ~ m t). As the values of d q / d P are not readily available, it is useful to transform Eq. (6) into following expression [25], using the ClausiusMosotti equation [26,27]:

4~re0a-~(2q + 112 ×(dirtY /

i d P Jr'

d,nV) T

17)

:0d+--

2

pep

(8)

where ~,~,is the sound velocity in the solvent. Assuming a cavity radius of 4 . 0 × 10 -~° m for both D M A B N and D M A B A and a ground state dipole moment of 6 debye, it is possible to estimate AV 2 for different values of the excited state dipole moment. The results of this estimation are given in Table 2 for nonane, which has a central position in the series of alkanes used. The values in Table 2 suggest that, when the excited state dipole moment is larger than 10 debye and assuming a negligible value of AVe, the electrostriction would lead to a dependence of H / H R on c e p / ~ exceeding the experimental error.

Table 2 Estimated volume contraction m nonane /*~ (D)

q~1sc AVlsc (m ~ mol - i )

0 9 12 15 18

6 . 5 × I11-6 - 8 . 1 × 10 -6 l . g x 10 -5 -3.4>110 s - 5 . 2 × 10 5

The volume contraction is calculated for different values of the excited state dipole moment using Eq, (7) and assuming a cavity radius of 4.0 X l0 ~o m. The isothermal compressibility of nonane amounts to 9 . 8 1 X 1 0 i¢~ m 3 / j as calculated from the sound propagation velocity [20,28,29], density [20], thermal expansion coefficient, and specific heat. A value of 1.97 was used for the relative dielectric constant of nonane [30,31].

3117

For large conjugated molecules without a large dipole moment in the ground or excited states, e.g. porphyrines, triplet formation in an aqueous medium is accompanied by a volume contraction of 7 × 10 ~' to 1.1 × 10 -s m 3 mol --~ due to strengthening of the hydrogen bonds [32,33]. For xanthene dyes, where this phenomenon is less important, volume changes between - I . 0 × 10 -~ and 1.5X 10 ~ m 3 tool L are reported [32]. Since DMABN and DMABA are conjugated molecules, one could expect a value of [AVll below 5 X 10 -~' m -~ tool ~, related to the formation or decay of the Testate in the absence of large configurational rearrangements, e.g. rotation of the dimethylamino moiety. It is not clear to what extent twisted intramolecular charge transfer (TICT) formation, although unlikely for a triplet state in an apolar solvent, would affect [AV~I. Intuitively a r o b ume increase is expected. However, a value of - 2 . 9 × 10 -5 m s mol ~ is observed lk3r AV,, lk~r the S L state of tetraphenylethylene dissolved in alkanes [34], where there is some evidence of TICT formation. Taking into account that a dipole moment of only 6 debye was determined for the TICT-state of tetraphenylethylene, the experimental value of AV,. is four times larger than expected. The latter observation suggests that the formation of a twisted excited state is accompanied by an important volume contraction, not related to the electrostriction effect. A similar "intrinsic" volume contraction ('XVe) was observed upon intramolecular photoinduced electron transfer and consecutive folding of a bridged donor-acceptor compound [25] or for intermolecular exciplex formation [25,35]. If this observation is general, AV~ would strengthen rather than cancel the effect of AV~ when a polar TICT state is li)rmed. This completely contradicts our observations. =..

4. Conclusions

The solvent-dependent laser-induced optoacoustic spectroscopic study establishes that the lowest triplet states of D M A B N and D M A B A do not have pronounced charge transfer character in apolar solvents. Under the assumption that rotation of the dimethylamino moiety would lead to an important volume contraction, this process can also be eliminated for the systems studied.

308

N. Chattopadhyay et al, / Chemical Physics Letters 279 (1997) 303-308

Acknowledgements NC thanks the Research Council, KU Leuven for a 'Junior Fellowship'. MVdA is an 'Onderzoeksdirecteur' of the F.W.O. (Fonds voor Wetenschappelijk Onderzoek Vlaanderen). The continuing support of the F.W.O., de 'Nationale Loterij' and DWTC through IUAP IV-II is gratefully acknowledged.

References [1] E. Lippert, W. Liider, H. Boos, in: A. Mangini (Ed.), Advances in Molecular Spectroscopy, Pergamon, Oxford, 1962, p.443. [2] E. Lippert, W. Rettig, V. Bona6ic-Kouteck~), F. Heisel, J.A. Mieh~, Adv. Chem. Phys. 68 (1987) 1. [3] K. Rotkiewicz, K.H. Grellmann, Z.R. Grabowski, Chem. Phys. Lett. 19 (1973) 315. [4] W. Scbuddeboom, S.A. Jonker, J.M. Warman, U. Leinhos, W. Ktihnle, K. Zachariasse, J. Phys. Chem. 96 (1992) 10809. [5] K. Bhattacharyya, M. Chowdhury, Chem. Rev. 93 (1993) 507. [6] S. D~ihne, W. Freyer, K. Teuchner, J. Dobkowski, Z.R. Grabowski, J. Luminesc. 22 (1980) 37. [7] J. Dobkowski, E. Kirkor-Kaminska, J. Koput, A. Siemiarczuk, J. Phys. Chem. 27 (1982) 339. [8] N. Chattopadhyay, J. Rommens, M. Van der Auweraer, F.C. De Schryver, Chem. Phys. Lett. 364 (1997) 265. [9] H. Giirner, H.J. Kuhn, J. Phys. Chem. 90 (1986) 5946. [I0] M. Hashimoto, H. Hamaguchi, J. Phys. Chem. 99 (1995) 7875. [11] G. Ktihler, G. Grabner, K. Rotkiewicz, Chem. Phys. 173 (1993) 275. [12] Z,A. Dreger, J.M. Lang, H.G. Drickamer, J. Phys. Chem. 100 (1996) 4646. [13] S.E. Braslavsky, G.E. Heibel, Chem. Rev. 92 (1992) 1381, [14] M.S. Churio, K.P. Angermund, S.E. Braslavsky, J. Phys. Chem. 98 (1994) 1776.

[15] J.-L. Habib Jiwan, A.K. Chibisov, S.E. Braslavsky, J. Phys. Chem. 99 (1995) 10246. [16] A. Onkelinx, F.C. De Schryver, L. Viaene, M. Van der Auweraer, K. Iwai, M. Yamamoto, M. Ichikawa, H. Masuhara, M. Maus. W. Rettig, J. Am. Chem. Soc. 118 (1996) 2892. [17] P. Van Haver, L. Viaene, M. Van der Auweraer, F.C. De Schryver, J. Photo-chem. Photobiol. A 63 (1992) 265. [18] M. Van der Auweraer, L. Viaene, P. Van Haver, F.C. De Schryver, J. Phys. Chem. 97 (1993) 7178. [19] P. Van Haver, N. Helsen, S. Depaemelaere, M. Van der Auweraer, F.C. De Schryver, J. Am. Chem. Soc. 113 (1991) 6849. [20] D.R. Lide (Ed.), CRC Handbook in Chemistry and Physics, 73rd ed., CRC Press, London, 1992-1993. [21] J. Timmermans (Ed.), Physico Chemical Constants of Pure Organic Compounds, Elsevier, London, 1950. [22] J.B. Callis, W.W. Parson, M. Gouterman, Biochim. Biophys. Acta 267 (1972) 348. [23] P. Drude, W. Nernst, Z. Phys. Chem. 15 (1894) 79. [24] E.J. Whalley, J. Chem. Phys. 38 (1963) 1400. [25] B. Wegewijs, J.W. Verhoeven, S.E. Braslavsky, J. Phys. Chem. 100 (1994) 8890. [26] R.C. Mufioz, R.A. Holyrod, K. Itoh, K. Nakagawa, M. Nishikawa, K. Fueki, J. Phys. Chem. 91 (1987) 4639. [27] R.C. Mufioz, R.A. Holyrod, K. Itoh, K. Nakagawa, M. Nishikawa, K. Fueki, J. Phys. Chem. 97 (1993) 4639. [28] S. Braslavsky, K. Heihoff, in: J,C. Scaiano (Ed.), Handbook of Organic Photochemistry, vol. 1, CRC Press, Boca Raton, FL, 1989, p.327. [29] W. Schaafs, in: K.H. Hellwege, A.M. Hellwege (Eds.), Landolt B~Srnstein N.S., Group II: Atomic and Molecular Physics, vol. 5, Springer, Berlin, 1967, ch. 2. [30] A.J. Gordon, R.A. Ford (Eds.), The Chemist's Companion, Wiley, New York, 1972, pp. 1-13. [31] R.J. Shaw, J. Chem. Eng. Data 14 (1979)431. [32] T. Gensch, S. Braslavsky, J. Phys. Chem. B 101 (1997) 101. [33] J. Feitelson, D. Mauzerall, J. Phys. Chem. 100 (1996) 7698. [34] J. Morals, J, Ma, M. Zimmt, J. Phys. Chem. 95 (1991) 3885. [35] P. Pollmann, D. Rehm, A. Weller, Ber. Bunsenges. Phys. Chem. 79 (1975) 692.