VISCOSITY AND TEMPERATURE EFFECTS IN FLUORESCENCE BY E. J. BOWEN Physical Chemistry Laboratory, Oxford University Received 30th September, 1958 Fluorescence measurements have been the means of examining the nature of the transfer of electronic energy from one molecuie to another and from one part of a molecule to another part. In this paper the possibility is explored of using fluorescence measurements to investigate the change over from one electronic state of a molecule to another. Fluorescence is generally enhanced by lowering the temperature. The measurement of this effect is not easy because attention must be paid to changes of concentration, refractive indices, extinction coefficients, and band shapes, while oxygen quenching, concentration quenching and effects such as dimerization must be eliminated.Mr. J. Sahu has recently measured the quantum yields of fluorescence of solutions of substituted anthracenes in a number of organic solvents from - 70" to 70°C with due care about the above factors.1 The effect of temperature on the yield F may be interpreted by the scheme: A f hv -+ A*, (1) A* + A + hv', (2) A* --f A. (3) If process 2 is independent of temperature and process 3 needs activation energy E it follows that 2 ( l / F ) - 1 = k exp (- E/RT). All the 9-substituted anthracene derivatives studied showed approximate agreement with this equation, which implies that F tends to unity at T = 0°K. On the other hand, the yields for side-substituted anthracene derivatives appeared to tend to values less than unity at T = O"K, which could be explained by supposing that process 3 takes two courses, a temperature-dependent and a temperature-independent one.The above equation then becomes ( l / F ) - K = k exp (- E/RT), where K is a constant greater than unity. Unfortunately one is not able here to evaluate K, k, and E separately with any accuracy, as this would require more precise values of F than it has been possible to obtain. Where there are only two constants, as for the 9-substituted derivatives, this difficulty does not arise. If this mode of treatment of the data is correct the heat of activation E may be interpreted in terms of a point (or more probably an ill-defined region) where the potential-energy surfaces of the excited singlet and the ground states come together, and the temperature-dependent effect is associated with a direct energy degradation.The temperature-independent effect may then be related with the change over from the excited singlet state to the triplet without the need for appreciable activation. There may therefore be a means here of separating two types of energy degradation. For the 9-substituted anthracenes, where the number of constants is advantage- ously reduced to two, the k and E values obtained showed for organic solvents of several types a much clearer dependence on the anthracene than on the solvent. In viscous solvents, however, both constants appeared to be markedly higher for each solute. 40E . J . BOWEN 41 This was further investigated by using as solvents a series of paraffinic mixtures of different viscosities. Table 1 gives for these solvents the quantity T/q (absolute temperature/viscosity in poise) x 10-4.temp. "C - 70 - 50 - 30 - 10 10 30 50 70 A TABLE 1 B solvent C D 1.03 1-8 2.88 1.07 0.28 1 4.3 1 -64 0.484 0.171 6.02 2-40 0.807 0.350 8.23 3-31 1 *20 0.626 10.68 4-42 1.75 1 *09 5.68 2.42 1.80 In interpreting the results, the problem is to separate the effect of temperature on the solvent from that on the solute in respect of process 3 above. It was found that the empirical equation : (l/F) - 1 = k(T/q)1/4 exp (- E/RT) gave a good fit for all the results, with very little variation of k or E with solvent. This is shown in the tables 2, 3 and 4 below giving experimental and calculated values of the fluorescence yields F for three compounds.Constants for three other anthracenes are given in table 5. TABLE 2.-9-METHYL ANTHRACENE solvent A B C k= 6.7, k= 6-12, k = 6.05, E= 2350, E=2350. E=2350, temp. O C expt. calc. expt. calc. expt. calc. - 70 0.82 0.81 - 50 0.68 0.68 - 30 0.56 0.58 0.67 0.67 0.75 0.74 - 10 0.46 0.46 0.56 0.56 0-64 063 10 0.38 0.37 0.46 0.46 0.53 0.53 30 0.30 0.30 0.37 0.37 0.43 0.43 50 0.24 0.24 0.29 0.30 0.36 0-35 70 0.24 0.25 0.29 0.29 TABLE 3.-9-METHOXY ANTHRACENE solvent A B C k=516, k = 506. k= 545, E= 5000, E= 5000, E=5000, temp. "C expt. calc. expt. calc. expt. calc. - 70 0.97 0.98 - 50 0.91 0.92 - 30 0.79 0.80 0.89 0.85 - 10 0.63 0.64 0.72 0.70 0.76 0.75 10 0.46 0.46 053 0.52 0.58 0.58 30 0.31 0.31 036 0.37 0.40 0-40 50 0.20 0.20 0.23 0.24 0.26 0.27 70 0.15 016 0.16 0.18 D k= 5.26' E=2350.expt. calc. 0.73 0.73 0.62 0.62 0.51 0.51 0.42 0.42 0.34 0-34 D k= 553. E= 5000. expt. calc. 0.81 0.79 0.62 0.62 0.44 0.44 0.29 0.29 0.18 0.19 These tables show that irrespective of solvent a single value of E can be found for each solute, and that the introduction of the term (T/q)1/4 makes the remaining constant k vary rather little with change of solvent.42 VISCOSITY AND TEMPERATURE EFFECTS IN FLUORESCENCE temp. OC - 50 - 30 - 10 10 30 50 70 solute 9-ethyl 9-phen yl p-chloro A k= 18.1, E= 3500, expt. calc. 0-94 092 0.85 0.85 0.75 0.74 0.63 0.63 0.51 0.52 0.40 0.42 TABLE 4.-9 : 10-DICHLORO ANTHRACENE solvent B C k=20*3, k= 19.5, E= 3500, E= 3500, expt. calc. expt. calc. 0-80 0.79 0.83 0.82 0.68 0.68 0.73 0.73 0.56 0.56 0.60 0-60 0.46 0.45 0-50 0.50 0.35 0.35 0-40 0.40 TABLE 5 solvent A B C k E k E k E 8.42 2500 6.92 2500 7-07 2500 1.15 1800 1-17 1800 1-36 1800 52.5 2900 49.7 2900 44.0 2900 D k = 18-4.E= 3500. expt. calc. 0.85 0.86 0-76 0.77 0.64 0.64 0.53 0.52 0.43 0.42 D k E 6-48 2500 1.28 1800 A possible interpretation of these results may be found along the following (a) A* 3 A + hv', (b) A* 3 A', (4 A' --f A*, (4 A' + A 4- hv', (4 (f 1 Here A* represents the excited molecule in its lowest vibrational level, becoming A' on receiving activation energy E to reach a change-over point on to the ground- state potential-energy surface. It is assumed that both A* and A' can radiate with a rate constant kh that the rate of process c is given by kl exp (- E/RT), d by k2, and f by k3 (TI$. This last assumption implies that diffusion-type movements, less frequent in viscous solvents, permit process f to take place. On this hypothesis (1IF)- 1 = (ki/kf) eXP ( - E/RT) (T/q)/{T/T -k k2fk3 -k (kl/k3) eXP ( - E/RT) -k kf/k3). If kl and kf are relatively small the expression reduces to lines. A + hv -+ A*, A' + solv. -+ A. (kllkf) exp ( - E/RT) (T/q)/(T/q -k k2/k3). The empirical function (T/7)1/4 is fairly closely reproducible by the function k(T/$/(T/q + C) over the range of values here used with k 12-20 and C 0.2-1.4. If the thread of these arguments is sound it indicates that for a singlet-excited state molecule to degrade directly to the singlet ground state, thermal activation energy is necessary but not sufficient ; in addition, the requisite freedom to reach certain amplitudes of vibrational movement against the solvent viscosity is also needed. 1 Bowen and Sahu, J. Physic Chem., 1959,63,4. 2 Terenin, Acfu physicochim., 1948,18,210.