Abstract
Fluorescence emission from the first, S1, and second, S2, excited π-electronic singlet states to the singlet ground state, S0, has been observed for a number of aromatic molecules in dilute, deoxygenated n-heptane solutions. Detailed fluorescence spectra from S1 and S2 for 3,4-benzpyrene and 1,12-benzperylene over the temperature range -90 to + 90°C are reported. The S2 fluorescence is very strongly dependent on temperature and it results mostly from thermal repopulation of S2 from S1. The ratio, R, of the integrated fluorescence quantum intensity from S2 to that from S1 is 0.09 at -90°C and 0.8 at +90°C for 1,12-benzperylene. For 3,4-benzpyrene, R = 0.06 at -90°C and R = 0.4 at +90°C. These studies show that fluorescence emission from the second excited π-singlet state of aromatic molecules in solution is not uncommon and that the internal conversion process cannot be regarded as being completely irreversible in these systems.
| Original language | English |
|---|---|
| Pages (from-to) | 471-483 |
| Number of pages | 13 |
| Journal | Journal of the Chemical Society, Faraday Transactions 2: Molecular and Chemical Physics |
| Volume | 69 |
| DOIs | |
| State | Published - 1973 |
Funding
C. E. Easterly L. G. Christophorou J. G. Carter Fluorescence emission from the first, S 1 , and second, S 2 , excited π -electronic singlet states to the singlet ground state, S 0 , has been observed for a number of aromatic molecules in dilute, deoxygenated n-heptane solutions. Detailed fluorescence spectra from S 1 and S 2 for 3,4-benzpyrene and 1,12-benzperylene over the temperature range –90 to + 90°C are reported. The S 2 fluorescence is very strongly dependent on temperature and it results mostly from thermal repopulation of S 2 from S 1 . The ratio, R , of the integrated fluorescence quantum intensity from S 2 to that from S 1 is 0.09 at –90°C and 0.8 at +90°C for 1,12-benzperylene. For 3,4-benzpyrene, R = 0.06 at –90°C and R = 0.4 at +90°C. These studies show that fluorescence emission from the second excited π -singlet state of aromatic molecules in solution is not uncommon and that the internal conversion process cannot be regarded as being completely irreversible in these systems. Fluorescence from the Second Excited dinglet State of Aromatic Hydrocarbons in Solutiont BY C. E. EASTERLY§, AND J. G. CARTERL. G. CHRISTOPHOROU* Health Physics Division, Oak Ridge National Laboratory, Oak Ridge, Tennessee 37830, U.S.A. Received 20th October, 1972 Fluorescence emission from the first, Si, and second, Sz, excited r-electronic singlet states to the singlet ground state, So, has been observed for a number of aromatic molecules in dilute, deoxygenated n-heptane solutions. Detailed fluorescence spectra from Si and S2 for 3,4-benzpyrene and 1,12-benzperylene over the temperature range -90 to +9VC are reported. The S2fluorescence is very strongly dependent on temperature and it results mostly from thermal repopulation of SZ from S1. The ratio, R, of the integrated fluorescence quantum intensity from S2 to that from Si is 0.09 at -90°C and 0.8 at +90°C for 1,lZbenzperylene. For 3,4-benzpyreneY R = 0.06 at -90°C and R = 0.4 at +9O"C. These studies show that fluorescence emission from the second excited T-singlet state of aromatic molecules in solution is not uncommon and that the internal conversion process cannot be regarded as being completely irreversible in these systems. As a result of the fast relaxation processes which operate among the higher excited z-electronic states of organic molecules, optical emission from organic mole- cules (with the exception of low pressure vapours) has been observed, quite generally, only from the lowest vibrational level, 0, of the first excited n-singlet (Sl,o)or n-triplet (Tl,o) state down to the various vibrational levels, n, of the singlet ground state (S0,J. For organic molecules in solution, azulene and some of its derivatives have been a distinct exception to the above generalization. For these molecules fluorescence was observed from the second excited n-singlet state S2,0 to So,,. Kobyshev and Terenin suggested the possibility of S2,0-+S0,n emission in chlorophyll and its phthalocyanine analogues in solution, and Dawson and Kropp observed a temperature dependent anti-Stokes fluorescence from 1,12-benzperylene in plastic solution which they analyzed in terms of S2,0+So,n fluorescence. Distinct examples of S2.0-4O.n fluorescence emission have been reported by the authors for 1,2- benzanthracene and 3,4-benzpyrene in n-heptane solutions. Recently Kropp and Stanley have reported a completely temperature dependent anti-Stokes fluorescence from ovalene in toluene and in plastic solutions. In organic vapours S2,0+S0,n fluorescence has been reported for pyrene,6-8 3,4-ben~pyrene,~p~and naphthalene.1° Since our original report on 1,2-benzanthracene and 3,4-benzpyrene we have observed S2,0+S0,n emission also from 3-methylpyrene, 2'-methyl-lY2-benzanthracene, 7-methyl- lY2-benzanthracene, 1,12-benzperylene and 3,4-benzotetraphene, all in n-heptane solutions. In this paper we report the results of a detailed investigation ofthese new emissions for 1,lZbenzperylene and 3,4-benzpyrene in n-heptane solutions ?Research sponsored by the U.S. Atomic Energy Commission under contract with Union Carbide Corporation. also Department of Physics, University of Tennessee, Knoxville, Tennessee. J Oak Ridge Graduate Fellow, University of Tennessee, Knoxville, Tennessee. 471 AROMATIC HYDROCARBONS IN SOLUTION over the temperature range -90 to +90°C. In the present study a more elaborate deoxygenation procedure and a more accurate method for determining the integrated fluorescence quantum intensities for S2 and S1 have been adopted, as compared to our original report on 1,2-benzanthracene and 3,4-benzpyrene. EXPERIMENTAL The apparatus used for measurement of the S2 fluorescence (Sz,o+So,o+ hv2) is con- structed around two Bausch and Lomb 500 mm grating monochromators (see fig. 1) with LiGHT SELECTIkG MONOCHROMATOR DEWAR LlOLilD -NITROGEN PYREX TUBE -INSULATION SAMPLE CHAMBER -SAMPLE THERMOCOUPLE -MERCURY WELL FIG.1.-Schematic diagram of the experimental FIG.2.-Sample cell deoxygenation apparatus used for measurement of the fluores- assembly. cence spectra. gratings blazed at 3000A. The light source is a lo00 W high pressure mercury-xenon lamp. The exciting light after passing through a mechanical chopper is focused on the entrance slit of the selecting monochromator. The resulting monochromatic light is focused on a sealed Suprasil cuvette fixed inside an evacuated chamber at the end of a heat exchange rod. Between the cuvette and the selecting monochromator, a Suprasil plate is mounted at 45" to the exciting light beam so that part of the exciting light is reflected onto an intensity monitoring vacuum photodiode. The sample chamber is mounted on an x--y table and the cuvette can be rotated for either front surface or right angle observa- tion. Continuously variable cooling of the curvette is obtained by a commercial Joule- Thomson refrigeration device using nitrogen gas ;heating is done by applying a well regulated voltage through heater coils wrapped around the heat exchange rod of the refrigeration unit. The temperature stability is 1.0"C and is measured by a thermocouple fixed in a Suprasil tube which is sealed in the center of the cuvette (see fig. 2). The emission from the sample passes through the second monochromator to a cooled EM1 62568 photomultiplier tube whose signal is processed by a lock-in amplifier which derives the y-axis of the recorder. An automatic wavelength drive provides the voltage for the x-axis. The absorption spectra were recorded with a Cary 14R spectrophotometer. The compounds studied (J. Hinton, Philadelphia) were zone-refined and were used without further purification. Chromatographic tests by the Analytical Chemistry Division of the Oak Ridge National Laboratory have shown them to be better than 99.9 % pure. The solvent, n-heptane, was spectroquality grade (Columbia Organic Chemicals, Columbia) C. E. EASTERLY, L. G. CHRISTOPHOROU, J. G. CARTER and was not purified further. n-Heptane was selected as the solvent because it is a non-polar liquid which has a low freezing (-91OC) and a high boiling (+98"C) point; this permits a study over a wide temperature range in the liquid phase. Further, the absorption spectrum of n-heptane lies l1 beyond 4.18 eV which corresponds to the most energetic (2967A) exciting light used. Thus the solute was excited directly rather than by energy transfer from the solvent. No n-heptane emission was observed for the wavelength region covered in this study. Deoxygenation of the solutions was accomplished by the conventional freeze-pump-thaw method. In addition to the normal precautions taken in order to avoid sample contamina- tion, a molecular sieve cold trap was placed as close to the sample as was physically possible. During the freeze portion of the procedure a layer of insulation and a Pyrex tube protect the cuvette from direct contact with liquid nitrogen (see fig. 2). This precaution was found necessary because of the poor strength characteristics of the internal edges of most cuvettes. The initial concentration of the solution for each compound studied was 5x M. However, we estimate an increase of 5 to 10 % of this concentration as a result of solvent loss during deoxygenation. At a maximum concentration of 5.5 x M there is virtually no possibility for interaction between excited and unexcited solute molecules, thus eliminating the possibility of molecular complex formation. The low concentration also limits self- absorption losses to a maximum of -3 %. All data have been corrected for the spectral response of the system as determined with a spectrally calibrated light source. In addition, the n2 (refractive index squared) geometrical correction to intensity measurements, as a function of wavelength and temperature, has been applied. Also, the measured relative quantum intensities have been corrected for changes in concentration with changing temperature. RESULTS DETERMINATION OF THE RELATIVE FLUORESCENCE YIELDS FROM s1 AND s2 In order to determine the relative fluorescence quantum intensities * F(9 from the first and second excited z-singlet states, it is necessary to separate the S2 and S1 emissions in the region of overlap. The general shape of an S1,o+So,n emission spectrum of an aromatic hydrocarbon in solution is normally one of sharp and intense peaks near the high energy onset with gradually decreasing sharpness and intensity at lower energies (longer wavelengths) until, finally, there are no distinct peaks and the intensity falls off nearly exponentially with decreasing energy. This is a consequence of the spacing of the vibrational levels in the ground state, So, and the Franck-Condon overlap factors. For both 1,12-benzperylene and 3,4-benzpyrene the high energy region of the S2,0+S0,n emission, namely that between Es,,, and Es,,ohas very distinct transitions with the same energy spacings present in the Sl,o-+ So,, emission spectrum (see following section). Since the S2,0+ So,, transitions must, moreover, end at the same upper vibrational levels of So as do the Sl,o+So,n transitions, we assume that the lower energy part of the S2,0+So,n emission spectrum resembles the general shape of the low energy part of the SI,O+SO,n emissionspectrum. In order to separate the two spectra, the wavelength scale is first changed to an energy scale and a point, A, (see fig. 3) is chosen on the S2,0+S0.n spectrum as near as possible to the Sl,o+So,o transition where there is yet no contribution to the fluores- cence signal from the S1 emission. The energy difference between this point, A, and the S2,0+So,o peak is then subtracted from the energy of the Sl,o+So,o peak giving the position, C, on the S1spectrum, i.e., a point which is energy wise the same * The function F(G) is defined as the number of emitted photons with wave numbers between 6and G+d3 and can be normalized by the relation q = where q is the fluorescence quantum ef5ciency. AROMATIC HYDROCARBONS IN SOLUTION distance from the 0-4transition in the S1band as is the point A from the 0-0 transition in the S2 band. The shape of C-D (i.e., the remaining part of the S1 spectrum) is used for the low energy tail, A-B of the S2 emission spectrum by normalizing the intensity at C to that at A. Subtracting the constructed part of the Sz emission spectrum from the total emission yields a slightly different intensity against energy spectrum which is used to construct a new A-B intensity distribution for the S2 emission spectrum. The process is repeated until no change results within 5 %. This was done by a computer program and normally no more than one iteration was necessary. FIG.3.-Relative fluorescence quantum intensity (right angle illumination) as a function of wave number, s, for a 5 x M solution of 1,lZbenzperylene in n-heptane at +5°C. -, overall experi- mental spectrum ;I-, S1fluorescence spectrum ;---, Szfluorescence spectrum (see text). FLUORESCENCE SPECTRA FROM THE FIRST AND SECOND EXCITED Z-SINGLET STATES OF 1,12-BENZPERYLENE AND 3,4-BENZPYRENE In fig. 4 we present the fluorescence quantum intensity as a function of wavelength for 1,12-benzperyiene for three different temperatures (-90, +5, and +85"C). The spectra were taken under identical conditions except for temperature; the differences seen are the result of only the differences in temperature. Below the peak at 4060& which corresponds to the Sl,o-+So,otransition, the spectra are seen to depend strongly on temperature. From the position of the S0,0+S2.0 transition shown in table 1, the emission peak at 3850 A can be identified as resulting from the TABLE1.--bOSITIONS OF THE so,o+s~,o AND s~,o-'sz,o TRANSlTIONSFOR 1,12-BENZPERYLENE AND 3,4BENZPYRENEIN Il-HEPTANE AIA so.o+sl,o_molecule vlcrn-1 AIA so.o-+~z.o i;/cm-1 1,12-benzperylenea 4055 24660 3835 26080 3,4-benzpyrene 4025 24850 3833 26 090 0 The absorption spectrum for 1,12-benzperylene was also taken in benzene solution and is in good agreement with that of E. Clar, PoZycycIic Hydrocarbons (Academic Press, New York, 1964). S2,0+So,o transition. Hence we attribute the fluorescence emission between 3850 and 4060 to the direct radiative transitions S2,03S0,n of 1,12-benzperylene. Consistent with this interpretation is the information presented in fig. 5 where the C. E. EASTERLY, L. G. CHRISTOPHOROU, J. G. CARTER fluorescence spectrum from S1 and the new emission attributed to the S2,0-+S0,n transitions are shown. The S2 fluorescence spectrum has been constructed (see fig. 3) from the known part (27 000-24880 cm-l) plus that (V<24 880 cm-') which was unfolded from the overall experjmental spectrum using the procedure outlined in the previous section and shifted in energy so that the peaks due to the S2.0-+S0.0 and S1,o+So,o transitions coincided. The two fluorescence spectra were normalized at their most intense peaks. The similarity in the positions of the vibrational peaks in the two spectra in fig. 5 for the energy range (3~24800 crn-l) between Es,,, and Es,,, clearly shows that both emissions terminate at the same vibrational levels of the ground state So. The S2 emission is seen to be quite intense, extending down to the 0-0 absorption peak S2,0, of the second excited n-singlet state. This observa- tion tends to rule out the possibility of hot-band emission from S1. A/nm FIG.4.-Relative fluorescence quantum intensity (right angle illumination) as a function of wave- length, A, for a 5 x 1W6 M solution of 1,12-benzperylene in n-heptane at -90, +5, and +85"C. Exciting wavelength is the 2967 A Hg line. From the spectra in fig. 4 and 5 it is evident that the most intense peak in both the S1 and S2 fluorescence bands is not that due to the Sl~o(S2,0)+So,otransition but that due to the transition S, ,0(S2,0)-+S0,nwhere n= 5. This implies a difference in the equilibrium internuclear separations for both the S, and S2 excited states, compared to those for the ground state So. The fluorescence spectra of 3,4-benzpyrene for three temperatures (-89, -1, and +86°C)are given in fig. 6. The fluorescence quantum intensities at each tempera- ture are related to each other absolutely. The S2 fluorescence is clearly seen to AROMATIC HYDROCARBONS IN SOLUTION xT/cm-' FIG.5.-Relative fluorescence quantum intensity (right angle illumination) as a function of wave number, Y, for a 5 x M solution of 1,12-benzperylene in n-heptane at +85°C. Exciting wave- length is the 2967A Hg line. The S1and S2 spectra have been separated as described in the text and have been normalized to their most intense peaks. The Szspectrum has been shifted so that its 0-0 transition coincides with the O--O transition of S1. -, Sl,o+So emission ; ---,shifted S2,0--f So,n emission. 0 FIG.6.-Relative fluorescence quantum intensity (right angle illumination) as a function of wave-length, A, for a 5x M solution of 3,4-benzpyrene in n-heptane at -89, -1, and +86"C. Exciting wavelength is the 2%7 A Hg line. C. E. EASTERLY, L. G. CHRISTOHPOROU, J. G. CARTER be quite temperature dependent. The S, fluorescence is very much more dependent on temperature for 3,4-benzpyrene than for 1,12-benzperylene (see fig. 4 and 6). The general features of the fluorescence spectra for 3,4-benzpyrene are the same as those for 1,12-benzperylene except that the most intense peak of the S, fluorescence corresponds to the 0-0 transition. At a slightly lower energy than the S2,0+S0,0 transition for each molecule, a weak fluorescence emission of undetermined origin is detected which is slightly less dependent on temperature than the corresponding S2 fluorescence. It should be noted that for both 1,lZbenzperylene and 3,4-benz- pyrene, the sharpness in the S2spectra decreases, as that in S1,with increasing tempera- ture. The relative fluorescence quantum intensities integrated over energy (in wave numbers, f) for the first excited z-singlet state, F, = F(V) dij,1 SlrO-+SO,n and for the second excited n-singlet state, F2 = F($)dis,1 S2,O +Sam have been determined as a function of temperature for 1,12-benzperylene and 3,4- benzpyrene. These are presented in fig. 7-and 8, respectively. The ratios F2/Fl temperaturer C FIG.7.4ntegrated fluorescence quantum intensities of S1 (upper curve) and S2 (lower curve) in relative units for 5 x M solutions of 1,lZbenzperylene in n-heptane. The different symbolsrefer to data taken with different samples. as a function of temperature have also been determined for both molecules and are given in fig. 9. It is seen that for 1,12-benzperylene, F2/F1is 0.09 at -90°C and 0.8 at +90"C. For 3,4-benzpyrene, F2/Fl is 0.06 at -90°C and 0.4 at +90°C. Although the functional dependence of the ratio F2/F1for 3,4-benzpyrene does not differ appreciably from that we reported earlier its absolute magnitude does by a factor of -6. The differences between the present values of the F2/F1ratios for AROMATIC HYDROCARBONS IN SOLUTION 3,4-benzpyrene and the earlier ones can be ascribed mainly to two factors : (i) the more accuratedetermination of F2 and Fl in the present study and (ii) the more elaborate deoxygenation procedure employed in the present work. The fluorescence quantum intensity for both the S1 and the S2 emissions were found to be proportional to the exciting light intensity over a wide range of light intensities. This rules out a biphotonic origin of the S2emission. It is finally noted that the ratio F2/Fl for 3,4-benzpyrene at 23°C has been determined for excitation wavelengths between 2650 and 3650A and it has been found to be constant within experimental error. temperat ure/"C FIG.8.-Integrated fluorescence quantum intensities of S1(upper curve) and S2 (lower curve) in relative units for 5 x lo-" M solutions of 3,4-benzpyrene in n-heptane. The different symbols refer to data taken with different samples. temperature/"C FIG.9.-Ratios of the second to the first excited state integrated fluorescence quantum intensities as a function of temperature for 5 x lo-" M solutions of 1,lZbenzperylene (upper dotted curve) and 3,4-benzpyrene (lower dotted curve) in n-heptane. C. E. EASTERLY, L. G. CHRISTOPHOROU, J. G. CARTER DISCUSSION In fig. 10 the positions of the first, second, and third excited z-singlet states for 1,lZbenzperylene are shown along with the photon excitation energy, hvexo and a simplified de-excitation scheme with the associated rate constants. The radiative rate constants from the Sl,o and S2,0levels to the ground state So are, respectively,kF1and kF2;kI1and k12are the rate constants for internal quenching of fluorescence from S1 and S2 respectively, and include both direct quenching to the ground state and intersystem crossing to the triplet manifold;* k12and kzr are, respectively, the 1,lZ-benzperylene 3,4benzpyrene 34 070 I 33.730 II I iIt m 2 15-to-t r5-r FIG. 10.-Energy level diagram for 1,12-benzperylene and 3,4benzpyrene. Energies are from absorption spectra taken in n-heptane at +23"C. Note that huexccoincides with the S3,0 of 1,12-benzperylene, but it lies slightly below the S~,Oof 3,4-benzpyrene. rate constants for the radiationless population of S1 from S2 (i.e., for the process S2,0+Sl,n) and radiationless thermal repopulation of S2 from S1.n (i.e., for the process Sl,n-+S2,0) where n20. Competing radiationless transitions from S2will be neglected, i.e., the rate constant k12is taken to be very small compared to the spin allowed kI2. Nothing explicit can be said about the temperature dependence of the radiative rate constant kF2. Dawson and Kropp reported kF1to be temperature * We assume that intersystem crossing from S, or Sz to the triplet manifold is irreversible, i.e., once the cross-over takes place, it is followed by strong internal conversion within the triplet mani-fold. AROMATIC HYDROCARBONS IN SOLUTION independent. From relevant studies on other systems l2 kF1was found not to be a strong function of temperature and one may expect kF2 to behave similarly. The rate constant k,, may contain a temperature dependent component as well as a temperature independent component. The rate constant kP1will be assumed to have the form kzl = kil exp (-W2,/kT). Let No be the number of molecules per second which arrive at the S2,0level via light excitation, and [M*Is2 and [M*Is,, respectively, the concentrations of excited molecules in S2 and S1at time t. Then we may write -d[M*]s2/dt = No -kF2[M*]S2 -k12[M*Isz +k21[M*Isl (1) and -d[M*]S,/dt = k12[M*]Sz-kF1[M*]S, -kll[M*]Si -k2l[M*ISl. (2) It we now let kl = kF1+krl+kzl and k, = kF2+k12,we have, under photostationary conditions, [M*]s2 = No/(k2 --(3)k1:21) and Hence, with ~LY,~,>&~,~,the quantum efficiencies for S, and S2 emission can be expressed as We shall substitute for q1 and q2 the measured quantities F,(T) and F,(T) which have been made to agree with Dawson and Kropp by taking for Fl +F2 at +23"C the value of 0.36. [The data in ref. (3) were taken in poly(methy1 methacrylate) solution.] The rate constant kFl(T)is then obtained from Using the present data for Fl(T)and for zF1(T)the results of Dawson and Kr~pp,~ we have found kF1to be temperature independent at 1.3 x lo6s-l. The internal conversion rate constant kIl can be found from The rate constant kI1is found to be essentially temperature independent at 3.4 x lo6 s-l. Additionally, it is seen from that a plot of In F2(T)/F1(T)against l/kT should yield the activation energy Wzl, C. E. EASTERLY, L. G. CHRISTOPHOROU, J. G. CARTER 481 the pre-exponential factor and the constant term. A plot of In F2(T)/F1(T)against l/kT is given in fig. 11. The data points from -80 to +4WC have been fitted by a computer least squares procedure to determine through eqn (9) W,1, k;1, and kF2/kl2. This fitting was performed using either all of the data points between -80 and +4O"C or any set of points in excess of 5 in this temperature range. From this analysis it was found that F2(23/F1(T) (0.021+0.005)+(13+4) exp [-(690+60 cm-l)/kT]. (10)= Eqn (10) is represented in fig. 11 by the broken line. Hence, kF2/k12 = (0.006+ 0.001) and kzl= (2.8k0.5)x lo9 exp [-(690+60 cm-l)/kT] s-l. For +20"C, kzl is equal to 9.5 x lo7 S-l. If we now take for kF2the value 2.2 x lo8 s-l l3 we find that k12 = (3.7k0.6)~1O1O s-l. It is noted also that from eqn (5) and (6) one obtains for the fluorescence lifetimes zF1 and 2F2, respectively, the values of 190 ns and 0.5 ns at +20°C. L-0.6--0.4-0.3-0.2 -0.1y 008-0 -The fluorescence quantum efficiency for the first excited state of 3,4-benzpyrene in alcohol at room temperature is 0.42,14 and the lifetime in cyclohexane at room temperature is 49 ns.15 It is unfortunate that these data were taken in solvents different from the one used in the present study. In the absence of any other informa- tion we will use these data to analyze our results on 3,4-benzpyrene (fig. 6, 8 and 9)" in a way similar to that employed for 1,lZbenzperylene. The data in fig. 8 were thus adjusted so that Fl+F, = 0.42 at +23"C. The radiative rate constant for the first excited state is expected to be essentially temperature independent l2 and kF1 is 8.6~was so assumed. Hence, from eqn (9, lo6 s-l and from eqn (8) the internal conversion rate constant krl varies from 4.6 x lo6 s-l at -90°C to 2.7 x lo7 s-l at +9O"C and is 1.5 x lo7 s-l at +23"C. The data points from -90 to -10°C * If future data shows this to be an unreasonable assumption, the data presented in fig. 8 can be re-evaluated using the revised values for the quantum yield and/or the fluorescence lifetime. It must be noted, however, that the calculated activation energy, WZ1,is insensitiveto the magnitude of these two values. AROMATIC HYDROCARBONS IN SOLUTION were computer fitted to eqn (9) yielding kF2/k12= (2.0k0.27) x and k,, = (1.4 & 0.39) x lo9 exp [ -(629 33 cm-')/kT] s-l. If we now take for kF2the value 2.0~10' s-l as obtained through the Strickler-Berg equation,16 the value of k12is found to be (1.0+0.15)x 1O1O s-l. Data points above -10°C have been excluded from the computer fitting because they would cause the standard deviations to become significantly larger and/or kF2/kl to become slightly negative. Although the higher temperature data for both 1,12-benzperylene and 3,4-benzpyrene deviate from the model proposed, a very good fit for the entire temperature range -90 to + 90°C is found for 1,Zbenzanthracene and a number of other m01ecules.l~ From the data and the analysis presented, it is reasonable to conclude the following. (i) The fluorescence quantum yield from S, is primarily due to an indirect fluorescence originating from thermal repopulation of S2from S1. (ii) The activation energies, Wzl,as determined in the present analysis are roughly one-half the actual energy separation E~,,o-E~l,obetween the first and second excited states (see table 1). This may indicate that S2,0 in addition to being repopulated from S1,o is also re- populated from vibrationally excited S, (i.e., S2,0-+Sl,n-+S2.0, n> 0) or possibly from an optically forbidden state located between S1,o and S2.0. Polarization measurements of 3,4-benzotetraphene and other five-ringed aromatic hydrocarbons have shown new electronic transitions, which are not observed in absorption spectra, between the second and third excited states in each rnolecule.18 (iii) The internal conversion process cannot be considered as being completely irreversible. In view of the present work it can be generally concluded that fluorescence from the second excited n-singlet state of aromatic molecules in solution does not appear to be as uncommon as it has been thought in the past. Such emissions, depending on the temperature, affect strongly the overall spectral characteristics of organic molecules. These conclusions are supported by additional data from 1,2-benzan- thracene and several other aromatic hydrocarbons in solution which will be presented later. The dependence of the S2 emission on temperature and on the energy separation of S1,oand S2,* is under further study. We gratefully acknowledge stimulating discussions with Dr. J. B. Birks of the University of Manchester, Manchester, England. We also wish to thank J. T. Bell of the Chemical Technology Division of the Oak Ridge National Laboratory for assistance in taking the absorption spectra and R. D. Birkhoff of the Health Physics Division of the Oak Ridge National Laboratory for constructive criticism. M. Beer and H. C. Longuet-Higgins,J. Chem.Phys., 1955,23,1390 ;G. Binsch, E. Heilbronner, R. Jankow and S. Schmidt, Chem. Phys. Letters, 1967, 1, 135. *G. I. Kobyshev and A. N. Terenin, in Proc. Int. Conference on Lunzinescence (Akadhiai Kiadii, Budapest, 1968), p. 520. W. R. Dawson and J. L. Qopp, J. Phys. Chem., 1969,73, 1752. C. E. Easterly, L. G. Christophorou, R. P. Blaunstein and J. G. Carter, C/zem.Phys. Letters, 1970, 6, 579. J. L. Kropp and C. C. Stanley, Chern. Phys. Letters, 1971, 9, 534. P. A. Geldof, R. P. H. Rettschnick and G. J. Hoytink, Chem. Phys. Letters, 1969, 4, 59. 'A. Nakajima and H. Baba, Bull. Chem. SOC.Jupun, 1970,43,967* H. Baba, A. Nakajima, M. Aoi and K. Chihara, J. Chem. Phys, 1971, 55, 2433. P. Wannier, P. M. Rentzepis and J. Jortner, Cizem. Phys. Letters, 1971, 10, 102. lo P. Wannier, P. M. Rentzepis and J. Jortner, Chem. Phys. Letters, 1971, 10, 193. l1 This can be inferred from work (E. N. Lassettre, A. Skerbele and M. A. Dillon, J. Chern. Phys., 1968, 49, 2382) on ethane, methane, propane, butane, and perdeuteroethane for which the first energy loss peak was found to occur at energies >, 8.5 eV, and from work (F. Hirayamaand S. Lipsky, J. Chern.Phys., 1969,51,3616) reporting that the emission from heptane peaked at 207 nm using 14708,wavelength for excitation. C. E. EASTERLY, L. G. CHRISTOPHOROU, J. G. CARTER l2 J. B. Birks, Photophysics of Aromatic Molecules (Wiley-Interscience, New York, 1970). l3 The value of kFz was calculated3 from the absorption spectrum using the StricMer-Berg equation (S. J. Strickler and R. A. Berg, J. Chern. Phys., 1962, 37, 814), assuming that the fluorescence yield from S2is unity and that the mirror image relationship applies. l4 C. A. Parker, C. G. Hatchard and T. A. Joyce, J. Mol. Spectr., 1964,14, 311. l5 D. J. Dyson, Ph.D. Thesis (University of Manchester, 1963). l6 S. J. Strickler and R. A. Berg, J. Chem. Phys., 1962, 37, 814. C. E. Easterly and L. G. Christophorou, to be published. l8 R. Kiessling, G. Hohlneicher and F. Dorr, 2.Nuturforsch., 1967, 22a, 1097.