1167
NOTES addition of vinyl chloride. They indicate that 82 is strongly negative, extremely composition and pressure dependent near the critical point, and that no discontinuity exists a t the phase boundary. The pronounced minima in the 52 vs. composition isobars near the critical point (accompanied by maxima in 81) arc also consistent with the data (Figure 3 ) . It would therefore seem that strong ordering must occur over a rather broad region in solutions of this type both in the one-phase and in the two-phase region, accompanied by the formation of molecular clusters whose structure is highly composition and pressure dependent near the critical point. 3 m
by F. Sheldon Wettack' Department of Chemistry, The Un'niversity of Texas, Austin, Texas (Received October 1 0 , 1 9 6 8 )
The absolute fluorescent efficiency, Qt,of 2-pentanone as a function of excitation wavelength and the effect of pressure and wavelength on this emission have been determined. The value of Qf (number of quanta emitted by excited singlet state/number of quanta absorbed) obtained at 3130A agrees well with the relative (to biacetyl) value reported by Ausloos and Murad,2 and the pressure dependence can be interpreted as confirming the wavelength effect noted3for the gas-phase Norrish type I1 reaction. The method used for calculating Qr is similar to method B of Noyes, Mulac, and Harter4 for benzene except that integrations were performed over phototube sensitivity and filter transmission for the entire emission spe~trum.~ Figure 1 shows the gas-phase emission spectrum6 and Table I shows the measured values of Qr. The overlap of the emission spectrum with the gas-phase absorption spectrum5 indicates that the excited singlet energy is about 3.75 elr (3300 A). The phosphorescence spectrum of 2-pentanone in EPA a t 77°K has also been Table I: Fluorescence Efficiency of 2-Pentanonea
A
3130 2925 2804 2730 2654
Uayelength (A)
5600
Figure 1. The gas-phase emission spectrum of 2-pentanone recorded with a Jarrell-Ash 3/4-m Czerny-Turner scanning spectrometer.
Fluorescence of 2-Pentanone
A,
4000
91
x
108
2.67 f 0.29* 2.01 1.24 1.34 1.49
a 25 Torr a t 25". Deviation from the mean for six measurements made with varying slit conditions. Values a t X # 3130 h; are relative to the 3130-h; value but have been corrected for any variation in phototube sensitivity.
determined and the features are essentially the same as those reported by Borkman and Kearns' for acetone. Figures 2 and 3 show the effect of added l13-butadiene and n-butane on Qf. Both gases tend to increase Qr, which indicates that they are removing vibrational and not electronic energy.
Discussion The values obtained for Qr confirm that fluorescence is a minor process from the excited singlet state, and seem to substantiate that both fluorescence and part of the Norrish type I1 reaction occur from the same electronic state. As Qr decreases by a factor of 2 in going from 3130 to 2700 A, the quantum yield of type I1 from the singlet state, IdII, increases by a factor3 of 2.2 over the same wavelength range. The increase in '411 is based on the fraction of the type I1 yield which is unaffected by a high pressure of butadiene. Since the diolefin does not quench the ketone fluorescence in the pressure range studied it is assumed that the inhibited fraction of type I1 occurs from a triplet state while the uninhibited fraction has a singlet state precursor. This assumption is valid if fluorescence and type I1 from the singlet state are competitive. The present data do not prove this to be true, but the relationship between the decrease in Qfand the increase in ~C#J[Iin going from 3130 to 2700 would seem to indicate that as type I1 becomes a more important singlet state process the singlet emission becomes less important. (1) Department of Chemistry, Hope College, Holland, Mich. 49423. (2) P. Ausloos and E. Murad, J . Phys. Chem.. 6 5 , 1519 (1961). (3) F. S. Wettack and W. A. Noyes, Jr., J. Amer. Chem. Soc., go, 3901 (1968). (4) W. A. Soyes, Jr., W. -4. Mulac, and D. A. Harter, J . Chem. Phys., 44, 2100 (1966). ( 5 ) For experimental details, see F. 9. Wettack, Ph.D. Dissertation, The University of Texas a t Austin, 1968. (6) The author wishes t o thank Dr. H. D. Mettee for help in obtaining this spectrum. (7) R . F. Borkman and D. R. Kearns, J. Chem. Phys., 44,945 (1966). Volume 79,Number 4 April 1960
1168
NOTES
1
‘a0
I
0,
0
1.0 0
50
2800 A E x e f t a t t o n 150
IO0
‘Butadiene (torr’
Figure 2. Effect of 1,3-butadiene on the fluorescence efficiency of 2-pentanone.
Thus one is tempted to conclude that each excited singlet molecule is capable of either fluorescing or undergoing the type 11, and, hence, the two processes are competitive. However, such a conclusion is strictly valid only if both processes occur from the same vibronic states. That this condition is met in the 2-pentanone system cannot be rigidly established by the present data. The question of competitiveness was raised recentlya since the total type I1 quantum yield, 411, appears essentially independent of both wavelength and pressure but Qf depends markedly on these parameters. However, 411 may be a function of competition between formation of type I1 products and formation of the parent ketone from some intermediate other than the singlet formed by light absorption, while ‘411 may be determined by competition within the singlet state between fluorescence and abstraction of a y-hydrogen to form the intermediate. The pressure dependence of Qf substantiates the assumption made previously3 regarding the loss of vibrational energy at long and short wavelengths. Furthermore, the similarity between the effect of 1,3-
4-01
butadiene and n-butane provides evidence that the inhibition of the type I1 reaction by diolefins in the gas phase is due t o an electronic energy transfer from the triplet state of the ketone as was assumed.3 The electronic energy for the upper singlet state of diolefins is uncertain,s and a singlet transfer might be possible if fluorescence and singlet state type I1 are not competitive. This possibility seems to be negated by the present results. The pressure dependence allows one to calculate5 an effective cross section for removal of vibrational energy in a two-step mechanism at the various wavelengths. The resulting cross sections are 0.81 Az (3130A) and 0.20fi2 (2800A). More than two steps are undoubtedly involved in the deactivation and so these values should represent maximum cross sections. At 3130 and 100 Torr such a cross section indicates that vibrational relaxation will occur in 2.9 X lov8 sec, which is comparable to the mean lifetime of the singlet state, 2.8 X lo-* sec, and that of the triplet state, 1.6 X sec, as calculated from the rate constants reported in ref 3. These values show that vibrational relaxation should be complete at total pressures of about 100 Torr. At short wavelengths the vibrational energy would be equilibrated in 1.0 X lo-’ sec, but the singlet and triplet mean lifetimes are 1.0 X lovs sec-’ and 3.6 X 10-lo sec-l, respectively. Thus high vibronic states will most certainly be involved in the reaction processes initiated with wavelengths below 2800 It should be noted that the lifetimes mentioned are dependent on the assumptions regarding diolefin quenching and cannot be considered as absolute values, particularly at the shorter wavelengths. In light of recent observations concerning the 0-0 band for the upper singlet state of diolefins, an interesting aspect of this work is that no quenching of Qr by butadiene is observed. Quenching of singlet emission from benzeneeJOand the xylenes’O by diolefins has been observed, and indicates that the diolefin 0-0 band is at a wavelength longer than about 2650A (-4.7 eV). This conclusion assumes, of course, that the fluorescence quenching observed is due to an exothermic electronic energy transfer. If singlet vibronic energy of the ketone is transferrable to the olefin as electronic excitation one might expect such a fluorescence quenching to occur when the The lack of ketone is excited in the region of 2800 quenching indicates either (1) that the transfer does not occur or (2) that fluorescence and the transfer occur from two different excited states. If the latter is true and if the type I1 reaction attributed to the singlet state also occurs from a nonfluorescent state, the agreement between the 1+11 increase and the Qf decrease
A.
A.
(Pi
1.0
I 0
100
200
3m
4 ;:,torr
o
3130 E x c i t e t l o n
A
2925 E x c i t a t i o n
0
2800 E x c i t a t i o n
400
500
600
Total Pressure (torr)
Figure 3. Effect of total pressure on the fluorescence efficiency of 2-pentanone. Added gas, n-butane. The Journal of Physical Chemistry
For example, see R. Srinivasan, Adoan. Pholochem., 4, 113 (1966). (9) A. Morikawa and R. J. Cvetanovic, J. Chem. P h y s . , 49, 1214 (1968). (10) W. A. Koyes, Jr., private communication. (E)
NOTES
1169
is fortuitous. In this case the diolefin experiments a t short wavelengths would have little meaning so far as triplet-state involvement in the type I1 process is concerned. If option (1) is correct, one is led to the conclusion that the electronic energy of the first excited singlet state of butadiene is probably above 4.4 eV (2800 A). If the energy is much less than 4.4 eV, one should observe energy transfer when the collision rate approaches the rate of dissipation of vibronic energy by other routes. The rates compared above indicate that vibrational relaxation is slow compared to reaction processes a t 2800 A. However, the concentration range investigated would seem sufficient to show the electronic energy transfer since Qr does increase with increasing butadiene pressure in this range. This is presumably due to a relaxation of vibrational energy of the excited ketone through a collisional interaction. At 150 Torr of butadiene measurable vibrational relaxation has occurred but &f is still significantly lower than that found with 3130-A excitation. Hence the collisional interaction is probably between a pentanone singlet possessing about 4.4 eV of vibronic energy and a ground-state diolefin. Thus the effect of butadiene on Qt and the relative changes in and Qr with decreasing wavelength indicate that collisions between 2-pentanone singlets and butadiene do occur but that energy transfer between the two does not occur following excitation at 2800 A. This conclusion, coupled with the quenching observationseJOmentioned above, would Feem to place the 0-0 band of butadiene between 2650A and approximately 2800 A. Acknowledgment. The author wishes to thank Professor W. A. Noyes, Jr., for guidance during the course of this work, which was supported by NSF Grant G P 6713.
The cis-trans Effect in the H-Atom Addition to Olefins by Richard D. Kelley, Ralph Klein, and Milton D. Scheer National Bureau of Standards, Washington, D . C . (Received October 14, lQ68)
were obtained for both cis and trans-2-butene by comparison of the rate of addition to the olefin with the rate of abstraction from n-butane. The ratio of the rates for cis-2-butene and n-butane and for trans-2butene and n-butane were compared, the faster rate of addition to trans with respect to the cis being demonstrated. This result is rather surprising for not only are the two molecules identical with respect to electron density around the ?r bond, but the free energy of formation of the cis-2-butene, 67.588 kJ/mol (16.154 kcal/mol) is more positive than that of the trans2-butene, 64.534 kJ/mol (15.424 kcal/mol) Because of the small difference of the rates and the absence of a simultaneous comparison between the two olefins, further confirmation of a cis-trans effect was sought. For this purpose, we have utilized the low-temperature technique4 and employed the compound cis, trans-2,6octadiene. Since both functional groups are on the same molecule, certain possible extraneous effects, such as local concentration differences, are avoided and relative rates can be directly determined. It has been demonstrated416that under the present experimental conditions of high dilution in propane a t 90°K the only significant radical reactions are the disproportionation and combination reactions of two alkyl radicals. Combination of large alkyl radicals is small a t low temperatures4p6 and may be neglected. Thus the observed products result from the disproportionation of the CSradicals formed by H-atom addition. A mixture of the three isomers of 2,6-octadiene was separated by gas chromatography, with each component being identified by ir absorption.' A 6-ft 20% glycerol (saturated with AgN03) column a t 50" with a helium flow of 50 cc/min gave excellent resolution of the three components. The cis, trans-2,6-octadiene purified in this manner showed less than 0.1% of the trans, trans isomer and no detectable cis, cis isomer. The reaction products of the H-atom addition to the 2,G-octadiene were observed by gas chromatography using a 20-ft 30% p,p'-oxydipropionitrile at 30" and a helium flow of 60 cc/min. No quantitative analysis was made of the product diolefins other than the 2,6-octadiene isomers. The octadiene, diluted with propane in a 1 to 500 ratio at 90"K, was exposed to gas-generated hydrogen (1) K. R. Jennings and R. J. Cvetanovic, J . Chem. Phys., 3 5 , 1233
(1961).
80334
Gas-phase experiments a t room temperature showed that hydrogen atoms add to trans-2-butene 20% faster than to cis-2-butene,' although earlier work using a molybdenum trioxide detector technique showed the reverse effect.2 The room temperature gas-phase data
(2) P. E. M.Allen, H. W. Melville, and J. 0. Robb, Proc. R o y . SOC., AZ18, 311 (1953). '(3) F. D. Rossini, K. 5. Pitzer, W. J. Taylor, J. P. Ebert, J. E. Kilpatrick. 0.W. Beckett. M. G . Williams, and H. G. Werner, "Selected Values of Properties of Hydrocarbons," National Bureau of Standards Circular C4G1, U. S. Government Printing Offlce. Washington, D. C., 1947. (4) R. Klein, M. D. Scheer, and R. Kelley, J . Phys. Chem., 6 8 , 598 (1964). (5) R. D. Kelley, R. Klein, and M. D. Scheer, {bid., 69, 905 (1965). (6) R. Klein and M. D. Scheer, ibtd., 6 7 , 1874 (1963). (7) W. von E. Doering and W. R. Roth, Tetrahedron, 18, 67 (19G2).
Volume 79, Number 4 April 1969
