ENERGY TRANSFER IN FLUORESCENT PLASTIC SOLUTlONS BY J. B. BIRKS AND K. N. KUCHELA The Physical Laboratories, The University, Manchester Received 24th March, 1959 Solvent-solute energy transfer in tetraphenyl-1 : 3-butadiene + polystyrene solutions has been studied by observing the fluorescence excitation spectra from 220-350 mp. At low con- centrations the transfer is purely radiative, but at c > 10-4 M non-radiative transfer occurs. The transfer is independent of excitation wavelength from 270-240 mp, but decreases at shorter wavelengths, probably due to the reduced excitation depth. There is some evidence for a change in the polystyrene emission spectrum when excited by 220 mp radiation. The results are compared with data on similar solutions, excited by ionizing radiations.Studies 1-4 of the fluorescence of polystyrene solutions of organic fluors, excited by ionizing radiations, have shown that efficient solvent-solute energy transfer occurs in these systems. If ionizing radiation is used for the stimulation of the fluorescence, the energy transfer is only one of several alternative or suc- cessive processes by which the initial ionization and excitation energy of the solvent molecules is dissipated, degraded, or ultimately converted into the observed solute fluorescence. If ultra-violet radiation is used for excitation, the energy transfer can be studied directly by observing the fluorescence excitation spectrum. In the present experiments this method, which has been previously used for liquid solutions,L 6 has been applied to a typical plastic solution system, that of 1 : 1’ : 4 : 4’-tetraphenyl-1 : 3-butadiene (T.P.B.) in polystyrene.The behaviour of this system for P-particle and a-particle excitation has been investigated by Swank and Buck.3 EXPERIMENTAL AND RESULTS The solutions, ranging in concentration c from 10-6 M to 10-2 M, were supplied by Messrs. Nash and Thompson, Ltd., and were prepared in the standard manner3 by the thermal polymerization of solutions of T.P.B. in styrene monomer. Cylindrical specimens, 2.2 cm diam. by 1 cm long, were machined from each solid solution. The ultra-violet absorption spectra of the solvent and solute (fig. 1) were measured with a Unicam SP 500 spectrophotometer. For polystyrene the extinction coefficient K was derived (i) from measurements on solutions in chloroform, extrapolated to 100 % concentration, and (ii) from measurements on thin polystyrene films.For T.P.B. the molar extinction coefficient E was derived from measurements (i) on solutions in poly- styrene at wavelengths down to 310 mp, and (ii) on solutions in cyclohexane. The rela- tive values of K and E determine small correction factors in the subsequent analysis. They can be estimated adequately for this purpose, despite the differences in the absorption in the liquid and solid phase. The fluorescence excitation spectra of the solution specimens were measured by a method similar to that of Birks and Cameron.6 The specimen was placed on a Chance OY 10 filter, adjacent to the photo-cathode of an E.M.I. 6097 B photomultiplier mounted vertically.The upper face of the specimen was illuminated with monochromatic radiation from a Unicam SP 500 spectrophotometer with a hydrogen lamp source, the horizontal exit beam being deflected vertically downwards by a polished aluminium mirror. The filter is transparent to the T.P.B. fluorescence, but it is opaque to the incident radiation and to most of the polystyrene fluorescence emission. The photomultiplier output 5758 ENERGY TRANSFER I N FLUORESCENT PLASTIC SOLUTIONS current was corrected for the small dark current component. Except at low concentra- tions, where there is a small contribution due to polystyrene fluorescence, the current is proportional to the intensity of the T.P.B. fluorescence. 4 10 3 10 \ \ \ \ A (mt4 FIG. 1.-Absorption spectra of (1) polystyrene (a) in chloroform, (b) solid, and (2) T.P.B.(a) in polystyrene, (6) in cyclohexane. 9-1 Z A (m) FIG. 2.-Fluorescence excitation spectra. Relative fluorescence quantum intensity I against wavelength X for specimens of concentration c. The relative quantum intensity of the incident radiation at different wavelengths was determined by replacing the specimen by a standard calibrating solution of 1 cm thick- ness, contained in a cylindrical fused silica vessel of similar dimensions to the specimen. 2 x 10-2 M solutions of 1-dimethylaminonaphthalene 7-sodium sulphonate in water, andJ. B . BIRKS AND K . N. KUCHELA 59 1-dimethylaminonaphthalene 5-sulphonic acid in N sodium hydroxide solution, which have constant fluorescence quantum efficiencies 7 at wavelengths from 210-400 mp, were used for calibration.Both solutions gave consistent and reproducible results. The I I I I 2 7 0 O l 2 2 0 2 3 0 240 2 5 0 2 6 0 (w) FIG. 3.-Energy transfer coefficient f against wavelength A, for specimens of concentration c. S C FIG. 4.-(a) Energy transfer coefficient f against concentration c (A = 240-270 mp). (6) Scintillation pulse height S against c (Swank and Buck 3). fluorescence intensity I of each specimen, normalized to a constant flux of incident quanta, was measured from 350 mp to 220 mp wavelength. Typical fluorescence excitation spectra are plotted in fig. 2. The energy transfer coefficient f, defined as the fraction of quanta initially absorbed by the solvent which are transferred to the solute, has been evaluated from the excitation60 ENERGY TRANSFER I N FLUORESCENT PLASTIC SOLUTIONS spectra.6 At X > 300m,u where the polystyrene is almost transparent, the T.P.B.fluor- escence is excited directly. Its intensity attains a limiting value l o in the higher con- centration specimens, where the incident radiation is completely absorbed by T.P.B. At lower concentrations Ill0 is approximately equal to the fraction of incident radiation absorbed by T.P.B. in the specimen, as calculated from E and K (fig. 1). At X < 270mp, practically all the incident radiation is absorbed by polystyrene in the lower concentration specimens, and assuming the fluorescence quantum efficiency of T.P.B. to be independent of wavelength, f = I/Zo. At higher concentrations a correction must be applied for the component of the T.P.B.fluorescence directly excited by the incident radiation,6 so that The values off as a function of excitation wavelength X for specimens of different c are plotted in fig. 3. f i s found to be independent of X for X = 240-270 mp, i.e. for excitation within the first absorption band of polystyrene. The variation off with c in this spectral region is plotted in fig. 4(a). FIG. 5.-Excitation at X = 270 m p : (a) direct ; I against concentration c ; (b) indirect (by polystyrene emission) ; I' against concentration c. The radiative component of the energy transfer, due to absorption of polystyrene fluorescence by T.P.B., was studied by a method similar to that of Swank and Buck.3 A thin disc of pure polystyrene, 0.05 cm thick, was placed on top of the specimen with a thin intermediate film of glycerine to reduce internal reflection at the interface. The polystyrene disc completely absorbs the incident radiation at X < 270mp and converts it into polystyrene emission.The intensity of the specimen fluorescence I', due to this indirect excitation by the polystyrene emission, was compared with the intensity I due to direct excitation, at X < 270 mp. The curves of Ip against concentration at X = 240-270 mp are identical within the experimental error. The variation of Ip and I with c at X = 270 mp is shown in fig. 5. DISCUSSION The energy transfer is independent of wavelength in the region X = 240-270 m,u, which corresponds to excitation into the 1st excited state of polystyrene.The polystyrene fluorescence emission spectrum 3 ~ 4 extends from X = 285-360 mp.J . B . BIRKS AND K . N. KUCHELA 61 This emission is efficiently absorbed by T.P.B. at relatively low concentrations. The indirect excitation curves of Ip against c (fig. 5 ) show that 90 % absorption of the polystyrene emission occurs at about c = 10-4 M, corresponding to an effective molar extinction coefficient E - 104 for a 1-cm specimen. This agrees with the value of E for T.P.B. (fig. 1) averaged over the polystyrene emission spec- trum. The direct excitation intensity 1 is proportional to Ip (allowing for the small component due to unabsorbed polystyrene emission) up to c - 10-4 M, showing that the energy transfer at these concentrations is purely radiative. Similar conclusions have been reached 4s 5 for excitation by ionizing radiations. The radiative transfer coefficient f R = Aqo, where qo is the fluorescence quantum efficiency of polystyrene, and A is the fraction of the emission absorbed by T.P.B., which can be estimated from the Ip against c curve.I 0 -3 0 -2 i il M C FIG. 6.-Energy transfer coefficient f against concentration c (A = 240-270 mp). (a) radiative f R , (b) non-radiative f N R , (c) total f, from (2) and (3) ; I, experimental. indicates that an additional transfer process, which is non-radiative, becomes operative. If the alternative processes competing for the solvent excitation energy are emission, self-quenching, quenching by the solute, and non-radiative transfer to the solute, and their relative probabilities are Pe, pq, psc and ptc respectively, then the non-radiative transfer coefficient will be given by PtC fNR = Pe 3- Pq + (ps + p t ) ~ 1 + (TIC a (TC =- Since the non-radiative transfer competes with the solvent emission, f R becomes A90 1 + (T'c' f R = ~ (3) Fig.6 shows a comparison between the experimental data and the values of f N R , f R and f( = f R + f n R ) from (2) and (3, taking qo = 0.1, (T = (T' = 30.62 Similar equations to (2) and (3), with a = a’ and A = 1 (for c > 10-4 M) have been derived by Swank and Buck.3 Their experimental data on the relative scintillation pulse height S of polystyrene + T.P.B. solutions for p-particle ex- citation are plotted in fig. 4(6). Within the experimental error S is proportional to f over the range of concentrations studied, indicating that when the system is excited by ionizing radiation the energy transfer occurs from the 1st excited state of the solvent. A decrease is observed in the energy transfer coefficient f at h < 240mp, c > 10-4 M.This spectral region corresponds to excitation in the 2nd absorption band of polystyrene. One possible explanation is that additional quenching processes (e.g. singlet-triplet state conversion) compete with the internal conversion of the polystyrene excitation energy from the 2nd to the 1st excited singkt state, and thus reduce f. An alternative, and more probable, explanation is that the ENERGY TRANSFER I N FLUORESCENT PLASTIC SOLUTIONS C FIG. 7.-Excitation at A = 220 and 240 mp : (a) direct ; I against c ; (b) indirect ; I p against c.effect is associated with the reduced depth of penetration x of the incident radiation due to the increase in K. For K = 104, 90 % of the radiation is absorbed within 1 p of the surface. A similar decrease is observed in the scintillation efficiency of organic crystals and solutions,ss 9 excited by ionizing particles which penetrate less than a few p below the surface. This effect 10 is attributed to the surface escape or quenching of the excitation energy. It is found that the scintillation efficiency decreases with x towards a limiting value at the surface, which is 0-5 of that in the interior of the material. In the present experiments, for c > 5 x 10-4 M, f decreases with A, i.e. with decrease in x , from A = 240 mp to 220 mp to about 0-5 of its maximum value.This similarity of behaviour for excitation by short- range ionizing and ultra-violet radiations indicates that the decrease in f is due to a similar surface effect. The shape of the I against c curve at h = 220 mp differs from that at /I = 230-270 mp, at c < 5 x 10-4 M (fig. 7). It appears that, while the non-radiative transfer f n R is unchanged apart from the surface effect, the radiative transfer f R which operates at lower c is modified. The results suggest a change in the spectrum of the polystyrene emission, since it is completely absorbed by lower concentrations of T.P.B. Further experiments are required to elucidate thisJ . B. BIRKS AND K . N. KUCHELA 63 apparent change, which might be due to emission from a triplet state, from the 2nd excited singlet state,ll or from surface impurities. It is interesting to note that at h = 220mp there is a clear discrimination between the alternative energy transfer processes. 1 Koski, Physic. Rev., 1951, 82, 230. 2 Pichat, Pestail and Clement, J . Chim. Physique, 1953, 50, 26. 3 Swank and Buck, Physic. Rev., 1953,91, 927. * 4 Krenz, Trans. Faraday SOC., 1955, 51, 172. 5 Cohen and Weinreb, Proc. Physic. SOC. B, 1956, 69, 593. 6 Birks and Cameron, Proc. Physic. SOC., 1958, 72, 53. 7 Weber and Teale, Trans. Faraday SOC., 1958, 54, 640. 8 King and Birks, Physic. Rev., 1952, 86, 568. 9 Birks and Brooks, Proc. Physic. SOC. B, 1956, 69, 721. IoBirks, Physic. Rev., 1952, 85, 569; 1953, 90, 1131. 11 Birks, Physic. Rev., 1954, 94, 1567.