JOURNAL O F T H E AMERICAN CHEMICAL SOCIETY Registered in U.S. Patent Office.
@ Copyright, 1976, by the American Chemical Society
VOLUME 98, NUMBER8
APRIL 14, 1976
Thermal, Photochemical, and Photophysical Processes in Cyclopropanone Vapor Hector J. Rodriguez,2 Jung-Ching Chang,* and Timothy F. Thomas* Contribution from the Department of Chemistry, Unioersity of Missouri-Kansas City, Kansas City, Missouri 641 10. Received June 26, I975
-
Abstract: The photochemical quantum yields of cyclopropanone vapor have been measured from 365 to 291 nm. For decom~ from 0.60 to -I .O as the irradiation wavelength position of the parent molecule 6 1.0 a t all wavelengths, while @ c ~ Hrises and Tf(calcd) 2 is decreased. The fluorescence quantum yield was measurable only at ,A, 365 nm, where & = 6.8 X X s. A detailed mechanism is proposed involving predissociation of the ‘(n,a*) state leading to two biradical intermediates. The thermal reaction of cyclopropanone, yielding polymer, was found to be surface catalyzed, with an activation energy of -2 kcal mol-’ in a Pyrex ir cell between 20 and 50 O C . An experimental value of $3.8 kcal mol-’ was obtained for the heat of formation of cyclopropanone and used to predict an activation energy of -30.5 kcal mol-’ for its homogenous decomposition via the oxyallyl biradical.
Cyclopropanone and its derivatives were proposed as intermediates in organic reactions several decades3-’ before the first spectroscopic e v i d e n ~ e for ~ , ~the existence of the molecule was obtained. Following the first reports of a synthetic method yielding workable quantities of the unsubstituted compound,l0>” there have been several studies of its reactions with a wide variety of acids, bases, and other adducts in solution, generally at subzero temperatures because of cyclopropanone’s high r e a ~ t i v i t y . ’ , ’ ~Numerous -~~ theoretical calculations to various levels of approximation have been made regarding the structure,I5 heat of formation,I6 and potential energy surfacesl7-I9 of cyclopropanone. But experimental measurements on the isolated molecule have been limited to determinations of its structure by microwave spectroscopy20 and electron diffraction,21 plus our preliminary report on its photochemistry in the gas phase.22 W e present here a detailed report of the experimental properties of gaseous cyclopropanone, including its photochemical and fluorescence quantum yields as a function of irradiation wavelength, its thermal decomposition rate constants (experimental vs. “theoretical”), and a detailed mechanism for its reactions via the ‘ ( n , r * ) state.
Experimental Seetion Materials. Ketene was synthesized by passing acetic anhydride, carried in a stream of dry nitrogen, over a heated tungsten filament, removing by-products by traps at -22 and -63 OC, then collecting the desired product at -132 OC. Subsequent trap-to-trap distillation on a high vacuum line yielded material with -0.80 Torr vapor pressure at -132 O C , which was shown to be 99+% pure ketene by GC and ir analysis. Cyclopropanone was synthesized by the reaction between ketene and diazomethane,’Ox’ with the modification that diazomethane was generated in a flow system (dry
-
argon carrier) and bubbled through pure liquid ketene at -132 O C . The diazomethane was produced by the reaction between sodium hydroxide and N-nitrosomethylurea in ethylene glycol, taking pains to thoroughly dry the resulting diazomethane-argon stream. Product cyclopropanone was purified by: (i) removing the ketene solvent at -78 O C ; (ii) distilling the residue at -45 O C ; (iii) distilling the distillate from (ii) at -63 OC and keeping the middle fraction. The resulting material was shown to have a purity of -9896, the main impurities being cyclobutanone, -1.5%, and ethylene, -0.5%. Its vapor pressure at -45 OC was -0.8 Torr and its melting point lay between -78 and -63 OC. Yields of cyclopropanone were low (56 were detected. Absorption maxima in the uv spectra of the vapor were found at 312 ( h l ) nm, e 17 M-’ cm-I; and at 201 nm, t I .O X lo3 M-’ cm-’ (see ref 22 for complete spectrum). I r peaks (gas phase) include: 3086 (vw), 3020 (w), 3005 (w), 2990 (w), 1980 (m), 1905 ( s ) , 1868 (m), 1818 (vs), 1050 (w), 1028 (w), 980 (m), 949 (m), 920 (m), 91 1 (m) cm-I. Interference by peaks due to polycyclopropanone formed on the KBr windows occurred in the 1500-900-cm-’ region, making identification of the weaker monomer peaks in that region somewhat tentative. Preliminary results with cyclopropanone-dz and -dq indicate the multiplicity of peaks in the carbonyl region is probably due to Fermi resonance with overtones or combination bands of vibrations whose fundamentals lie in the 9 0 0 - 1 0 5 0 - ~ m - ~region, rather than the co-existence of several isomeric forms of ~yclopropanone.~3 The added gases used in the photochemical studies were all research grade: COz, from Air Products; 0 2 , from Matheson Gas Products; 1,3-butadiene and cyclopentane from Phillips Petroleum I s z i
2027
2028 Co. Compounds used in the cyclopropanone synthesis had the following sources and purity: acetic anhydride, M C / B reagent grade; ethylene glycol, M C / B chromatographic quality; N-nitrosomethylurea, K & K Laboratories, containing 3% acetic acid. Cyclobutanone was obtained from Aldrich Chemical and showed >99% purity by GC, after trap-to-trap distillation. Azomethane was synthesized from sym-dimethylhydrazine dihydrochloride by the method of Renaud and L e i t ~ h then , ~ ~ purified by trap-to-trap distillation. Quinine sulfate, used as the emission standard in the fluorescence yield determination, was from K & K Laboratories and was used without further purification. Procedures. Rates of the thermal reaction of cyclopropanone as a function of temperature were measured using a water-jacketed 10-cm Pyrex ir gas cell and recording the transmittance at 1905 cm-’ as a function of time on a Perkin-Elmer 621 ir spectrophotometer operating with X5 scale expansion. The 1980- and 1818cm-’ peaks were occasionally monitored and showed the same decay rate as the 1905-cm-I peak, which was preferred because it is free from interference by traces of cyclobutanone which absorbs strongly at 1816 cm-l. Temperatures were read using a calibrated thermister placed in a well just above the ir beam passing through the cell. For the surface-effect studies a stainless steel beam-conforming ir cell and a quartz photolysis cell (see below), in one case containing 0.88 g of thoroughly cleaned 200 mesh Pyrex powder, were also used. For these latter two cells the Cary 1501 spectrophotometer was used to follow the cyclopropanone disappearance. Samples for photolysis were prepared on a mercury-free highvacuum line (routinely maintained at Torr) and contained in 10-cm long, 1.9 cm o.d., fused quartz optical cells with MartinKontes Kel-F high-vacuum stopcocks. A Bausch and Lomb SP200 mercury arc was used as the radiation source; 313- and 365nm mercury lines were selected by interference filters having a half-band width of 12 nm, while 334 and -291 nm were obtained using a Bausch and Lomb High Intensity Monochromator (AAl12 5 nm) preceded by water and bromine filters. Gaseous azomethane at 2-3 Torr .was used as an actinometer to calibrate the RCA No. 935 vacuum photodiode used to monitor light intensities. Actinometry was done about once for every two cyclopropanone photolyses, and was based on the assumption that @ c ~ H=~ 1.0. Extinction coefficients used in the quantum yield calculations were obtained on both Cary 14 and 1501 spectrophotometers for azomethane and cyclopropanone; in the case of cyclopropanone, c was found to be constant to & 5 6 % between 0.7 and 33 Torr. The effective extinction coefficient of cyclobutanone was determined directly from transmittance of a 45-Torr gas sample in the photolysis system, because of its sharply structured absorption spectrum. Product analyses were carried out by means of gas chromatography (Varian 1200, hydrogen flame detector), mass spectrometry (Varian MAT CH4B), and ir (Perkin Elmer 621) and uv (Cary 1501) spectrophotometry. The G C was used to measure the amount of C2H4 produced in photolysis of cyclopropanone and C2Hs from azomethane, with cyclopentane as the internal standard in all samples. Columns used included 20% SF-96 on Chromosorb 102 at 100 OC (for C2H4 and cyclopentane) and 10% OV-17 on Chromosorb G at 120 OC (searching for other products). The following photoproducts were looked for but not found by G C (detection limits): acrolein ( < 2 p ) , acetone ( < 2 p ) , cyclohexane-I ,4-dione, a possible dimer of cyclopropanone ( < 3 p ) , cyclobutanone ( < 3 p ) , and cyclopropanone-butadiene adducts such as 3-vinyl~yclopentanone~~ ( < 3 p ; 2-octanone was used to set an upper limit to f R and detection limit for the adduct). The mass spectrometer was used to confirm that the ratio of ethylene and carbon monoxide formed by photolysis of cyclopropanone was unity, within experimental error of f < 3 % . In addition, a mass spectrum of cyclopropanone photolyzed to 93% completion showed no peaks greater than 1% of the base peak (28) for m/e >70. Attempts to monitor cyclopropanone itself by G C were unsuccessful, even using an all glass inlet system and a glass column packedwith Kel-F beads. Instead the absorbance at 48 600 cm-I, where t is 763 M-’ cm-l for cyclopropanone but the absorption by cyclobutanone impurity and CzH4 product was negligible, was used to measure the rates of both photo- and thermal decomposition of cyclopropanone. The uv absorption was also used to search for another possible photoproduct, ketene. This compound has emax 4.17 X l o 3 M-’ cm-‘ at 47 080 cm-I which permitted 2-5 p to be detected. Several experiments were done in which a sample of
Journal of the American Chemical Society
/
98:8
/
cyclopropanone was photolyzed while in the sample beam of the ir spectrophotometer, using a gas cell with two quartz and two KBr windows, in hopes of detecting any transient photoproducts. Scans of the ir spectrum during and immediately after photolysis of a 16-Torr cyclopropanone sample by A,,, 313 nm revealed only C2H4, C O , and polymer. The polymer formed in both thermal and photochemical reactions was characterized by its N M R spectrum, obtained on a Varian T-60 N M R spectrometer, as well as by uv and ir absorption spectra (see Results section). The procedures used to measure fluorescence quantum yields have been described in detail in a recent publication.26 They were improved over ones used to obtain an earlier value22 for brof cyclopropanone vapor in several significant respects: a Wood’s horn light trap was added opposite the viewing window of the fluorescence cell; the photomultiplier tube was cooled to -25 OC; emitted light intensity was measured on an S S R No. 1 1 I O photon counter instead of a lock-in amplifier; and a quinine sulfate solution was used as the fluorescence standard at all wavelengths rather than biacetyl at A,, 365 nm and hexafluoroacetone at shorter hex’s.The first three changes reduced the scattered light contribution and the random scatter of the data; the last improved the absolute accuracy of the resultant 4;s. Because the continuous formation of polymer on the windows increased the scattered light contribution during the course of a measurement, the scattered light correction was determined initially on the evacuated fluorescence cell, then again, after measuring emission of the sample, with the sample frozen down by liquid nitrogen. An attempt was also made to measure the fluorescence lifetime of a 1.8-Torr sample of cyclopropanone, using the 337.1-nm line of a nitrogen laser,26 but no difference in the intensity or time profile of oscilloscope traces could be discerned between the filled and the evacuated fluorescence cell. The fluorescence spectrum was scanned using a Perkin-Elmer “linear energy” spectrofluorimeter with the excitation monochromator system replaced by a mercury arc-interference filter combination. Because of the very weak emission and the relatively large contribution of scattered light (the spectrum of which suggested emission from cell windows and/or polymer deposit on same), the resulting spectrum had S / N 5 2:1, permitting only the conclusion that the emission extends roughly from 410 to 610 nm, with a maximum in the vicinity of 480 nm. The appearance potential of the CrH4+ ion formed from cyclopropanone by electron impact was determined using a Fox-type RPD appearance potential accessory mounted on a Nuclide 1290-G mass ~pectrometer.~’ The retarding voltage was modulated at 26 Hz with an amplitude of 100 mV permitting “lock-in” amplification of only those ions produced by a 100-mV slice out of the energy distribution of the ionizing electrons. High-frequency pulsing of the ion repeller electrode caused ionization to occur in a field-free region; the ionizing voltage was scanned digitally in steps of 10 mV. Further details will be given in a subsequent paper reporting appearance potential data for a series of cyclic and bicyclic ketones.
Results The Thermal Reaction. Because t h e t h e r m a l reaction o f t h e g a s e o u s c y c l o p r o p a n o n e proceeded a t a significant rate at room t e m p e r a t u r e , a c t u a l l y a t a slightly faster r a t e t h a n the photolyses a t 2 9 1 nm and a b o u t t h e same r a t e as t h e photolysis a t 3 3 4 n m , it w a s necessary t o investigate t h e t h e r m a l reaction before o b t a i n i n g q u a n t i t a t i v e p h o t o c h e m i cal data. T h e a n a l y t i c a l t e c h n i q u e s described in t h e Experim e n t a l S e c t i o n w e r e used t o s e a r c h for p r o d u c t s of t h e t h e r mal reaction, b u t n o volatile p r o d u c t s w e r e ever f o u n d . Inf r a r e d s p e c t r a r u n repetitively on gaseous c y c l o p r o p a n o n e s a m p l e s showed t h a t t h e decay of p e a k s in t h e 1 8 0 0 to 2000 cm-’ region w a s a c c o m p a n i e d by a c o r r e s p o n d i n g rise in p e a k s a t 1450, 1 3 1 0 , 1125, 1 0 0 5 , 975, a n d 945 c m - I , plus several w e a k e r p e a k s in t h e s a m e region. T h e s e p e a k s rem a i n e d a f t e r e v a c u a t i o n of t h e ir cell a n d t h e y c o r r e s p o n d satisfactorily t o t h e absorption s p e c t r u m previously rep o r t e d for polycyclopropanone.” An N M R s p e c t r u m of t h e p o l y m e r formed in t h e t h e r m a l reaction, dissolved in CDC13 with 1% TMS, showed a single p e a k a t 6 1.0 p p m , in good
April 14, 1976
2029
1
1.6
1.3
4
0 a
-
1.1 -
1.2
0
A
1.0
-
.7
1
‘60
10
20
30
40
50
60
70
80
0.5
t (Min)
Figure 1. Thermal reaction rate of cyclopropanone: A = absorbance at 1905 cm-’ in a IO-cm Pyrex ir cell; P c p = 8 1 0 ~ T; = 49.5 O C .
agreement with Schaafsma, et al.’s value of 1 .O for polycyclopropanone.] The kinetic order of the thermal reaction was estimated by continuously recording the transmittance a t one of the “carbonyl” peaks in the water-jacketed gas ir cell described earlier. As shown in Figure 1 the decay of the absorbance due to cyclopropanone fits a first-order plot well for nearly 2 half-lives. Thus, the thermal reaction of cyclopropanone seems well summarized by
1
0.40.3I
I
Figure 2. Temperature dependence of the thermal rate constant: P C P N 8 2 0 ~ . Table 1. Effect of Surface on Thermal Rate Constant“
Vessel
S/V, cm-‘
Quartz photolysis cell (seasoned) Pyrex ir cell (KBr windows) Stainless steel ir cell (KBr windows) Quartz photolysis cell (+ pyrex powder)
r a t e = k,[CP]
The indicated structure of the polymer is due to Schaafsma, et al., whose conclusion was based on identification of characteristic cyclopropane ring frequencies in the near-ir spectrum of the polymer, as well as on its N M R spectrum. The first-order rate constant, k T , was measured in the water-jacketed gas ir cell a t temperatures ranging from 21.8 to 49.5 OC. The results, shown in Figure 2, lead to a n Arrhenius activation energy of 11.8 f 1.7 kcal mol-’. The large uncertainty in E , as well as the apparent curvature in the plot may be partally a consequence of temperature gradients of the order of 2-3 OC along the length of the cell and 8-9 O C on the surface of the windows, but could conceivably indicate a change in reaction mechanism a t higher temperatures. The thermal reaction rate was also measured in a stainless steel ir cell and in a pair of quartz uv cells, a “seasoned” one and a clean one with added Pyrex powder. The results given in Table I show a strong dependence of kr on both the surface-to-volume ratio and the nature of the surface in the container used. Our conclusion is that the thermal reaction of cyclopropanone, forming polycyclopropanone, is partly, if not entirely, a surface-catalyzed process. Thus, the value of the activation energy obtained from Figure 2 is only a lower limit to the activation energy for any homogeneous thermal reaction that may occur. The Photochemical Reaction. Using the analytical methods described in the Experimental Section the only photoproducts detected were ethylene and carbon monoxide (in a 1 to 1 ratio) and polymer. As detailed in our previous report,22 the identification of polymer as a photoproduct was based on the nearly twice as rapid increase of absorbance in
I
I
a
T = 23 O
Pcp, Torr
kT,
min-‘
2.20
0.75
8X
1.58
0.80
4 x 10-3
3.13
0.70
2X
13.03
0.75
7x
C
the -1 90-220-nm region for the evacuated photolysis cell as for the matched dark reaction cell. After completion of all the photochemical experiments reported here, the photolysis cell was washed out with CDC13 containing 1% TMS. The resulting solution gave an N M R spectrum with a single peak a t 6 1.0 ppm and an ir spectrum with showing peaks a t 1450, 1310, 1125, 1010, 975, and 945 cm-I, both essentially identical with spectra for the polycyclopropanone formed in the thermal reaction alone. In calculating the photochemical quantum yields for loss of cyclopropanone and formation of ethylene (the amount of polymer formed per run being too small to be measured accurately) the simultaneous loss of cyclopropanone via the thermal reaction was accounted for as shown in the following equations. Since in no case case did the absorbance of a sample at the wavelength of irradiation exceed 0.04, the photolysis followed first-order kinetics: D-0
kP
-+
products
r a t e = k,,[CP]
where k p = 2.303$cptlIo. Here $ c p is the quantum yield of cyclopropanone disappearance; I , the cell path length; t, the molar extinction coefficient; and IO,the incident light intensity. Since simultaneous steps 1 and 2 are both first order,
Rodriguez, Chang. Thomas
/ Thermal and Photo Processes in Cyclopropanone Vapor
2030 the change in cyclopropanone concentration during photolysis time t is given by
In [CP],/[CP]o = -(A+
+ h,)t
(3)
The ratio of concentrations is now replaced by the ratio of absorbances a t 48 600 cm-’ and the equation solved for k p , giving finally:
Table 11. Extinction Coefficients (M-’ cm-l) Compd ~
E365
6734
6313
E291
2.7
11.8 4.78 n.d.
16.8 2.89 5.61
8.9 0.77 n.d.
~~
Cyclopropanone Azomethane Cyclobutanone
3.55 n.d
lyzed, using the same methods of sample preparation, photolysis, analysis, and actinometry. For A,,, 3 13 nm the quanSimilarly, tum yields obtained were: propylene, 0.05; cyclopropane, 0.23; ethylene, 0.65. At 4.5 Torr, 40 OC, and 313 nm b o C,H, CO (5) McGee28 reports the 4’s: propylene, 0.02; cyclopropane, rate = L,[CP], 0.28; ethylene, 0.70, Allowing for th known pressure effect on the propy1ene:cyclopropane ratio,29 the agreement seems where kE = 2.303~C2H,€l~0 and [CP], can be obtained from satisfactory. eq 3. The integrated rate equation from ( 5 ) is Fluorescence Yields and Lifetimes. Table IV shows the fluorescence quantum yields measured for cyclopropanone vapor, with previously reported values for cyclobutanoneZb included for comparison. The 4f listed for cyclopropanone which yields, on using eq 3 and solving for kE: a t A,, 365 nm replaces an earlier, less accurate value of -5 X (see Experimental Section for improvements made in the method of measurement).22 The earlier value was probably too high because of insufficiently accurate correcwhere ~ c is ~the Hquantum ~ yield of C2H4 formation. tion for the relatively large contribution by “scattered Because of the heterogeneous nature of the concurrent light” from polymer being continuously deposited on the thermal reaction, it seemed advisable to determine the k T cell windows. Estimated “probable errors” are: f 9 X used in eq 4 for each photolysis. This was routinely done by a t Aey. 365 nm for cyclopropanone; f 3 X a t A,, 334 measuring the decrease of the far-uv absorption band for a nm, and f 5 X a t the other wavelengths for cyclobuta1 -h time period after the photolysis but before G C analysis none. of products, Values of k T so determined varied erratically As stated earlier, the attempt to measure the lifetime of between the extremes of 5.3 X and 2.3 X min-]; the fluorescing state of cyclopropanone was not successful. the k T appearing in the first line of Table I represents a However, the natural radiative lifetime, TO,of the l(n,r*) mean value. The value of A[CzH4] appearing in eq 7 was state was calculated by the method of Strickler and Berg30 determined by the GC analysis, with correction for a small to be 1.2 f 0.2 X s, assuming = 1.1 X 10-13 amount of C2H4 present in the blank photolysis cell The inicm3 (corresponding approximately to a mean fluorescence tial concentration of cyclopropanone, [CPIo, was obtained wavelength of 480 nm). The Strickler-Berg formula has from the absorbance a t 48 600 cm-I before starting the been found to be inaccurate for the n r* transition i n kephotolysis. l o was based on the average current of the vacutones, however, giving values for TO that are too long by a um photodiode, calibrated as described in the Experimental factor of 4 in the case of c y c l ~ b u t a n o n eand ~ ~ by a factor of Section. The experimental values of the extinction coeffi8 in acetone.32The smaller discrepancy in the cyclic comcients used are given in Table 11. pound may be a consequence of a more constrained geomeThe experimental values obtained for 4cp and ~ . J c * H try ~ . near the carbonyl c h r ~ m o p h o r e ;therefore, ~~ a correcfrom eq 4 and 7, respectively, are shown in Table 111. The tion factor of 4 will be used in the case of cyclopropanone, indicated uncertainties in the average 4’s shown for each yielding the estimate: TO = 3.0 X s. The fluorescence wavelength are standard deviations. The estimated experilifetimes, or their limits, shown in Table IV were calculated mental errors in individual quantum yields ranged from from this value and the experimental results for 4f via the f0.02 a t 3 13 nm to f0.05 a t 29 1 nm, for ~ c ~andHfrom ~ , formula f 0 . 0 7 a t 313 nm to f 0 . 1 5 a t 291 nm, for dcp. Examination Tf = (8) of the results shows that, even a t pressures above one atmos.~’ The To used for cyclobutanone was 2.46 X sphere, none of the added gases had an effect on the two Heat of Formation. A total of nine determinations of the quantum yields in excess of the estimated experimental appearance potential of the C2H4+ ion from cyclopropaerror. In addition, no new products were found when the ponone were made on three different samples, two containing tentially reactive gases 0 2 and 1,3-butadiene were added to xenon and one containing benzene to calibrate the energy the samples to be photolyzed. In particular, no cyclopropascale. The two samples containing xenon yielded AP(28) = none-butadiene adduct could be detected.25There is, how9.67 f 0.03 and 9.66 f 0.03 eV; the sample containing benever, a clear effect of the wavelength of irradiation on zene gave AP(28) = 9.73 f 0.03 eV. Making a small (-0.2 @c~H the~ quantum , yield increasing with decreasing Ai,, until the limiting value of 1.0 is attained by 291 nm. In conkcal/mol) correction from the source temperature of 50 to 25 OC and using literature values for34 AHfO’s of C O and trast, 4cp appears to be essentially unity at all irradiation CzH4+ gives the following heats of formation for cycloprowavelengths. Values for 4cp could not be obtained in the panone: AHf0(298 K) = +4.2, +4.5, and +2.8 kcal/mol. runs in which cyclopentane or butadiene were added gases, We have a slight preference for the latter value, since the because of strong absorption by these compounds a t 48 600 shape of the ionization efficiency curve for C&j+ near cm-l. In these cases [CP]o was determined from the meaonset more closely resembles that of C2H4’ from cycloprosured initial pressure of cyclopropanone, corrected for the panone than does the Xe+ curve, possibly leading to some presence of small amounts of C2H4 and cyclobutanone imsmall cancellation of errors. But the differences are so close purities. to the estimated experimental precision of f 0 . 7 kcal mol-‘ As a check on the experimental procedures used for cyclopropanone, a 902-c( sample of cyclobutanone was photothat it only seems justified to take the mean value, f 3 . 8
2.303
+
-
@fro
Journal of the American Chemical Society
/
98:8
/
April 14, I976
203 1 Table 111. Photochemical Quantum Yields"
P0,CP. @
Padd,
Torr
10-15 x io, photons S-I
t , min
(A0/Aob
APCzHo /J
4CP
bC2H4
Xi, 365 nm
783 815 774 78 1 783 796 745 772 718 519 780
0 0 0
197 C02 398 c02 792 C02 1268 C02 327 c - C ~ H I ~ 100 0 2
420 0 2 65 1 (1,3-C4Hs)
7.69 7.57 6.86 7.03 7.54 7.69 6.86 6.17 6.56 6.20 14.00
60 60 65 65 60 60 65 70 65 72 45
38.9 45.9 40.3 44.5 43.3 43.1 41.0 37.5 34.2 24.2 65.0
1.14
35.2
1.15
59.6 126.0 685.0 68.5 89.0 126.0 56.1 125.0
1.14 1.30 1.29 1.49 1.32 1.38 1.14 n.d.
28.2 28.5 28.4 24.4
I .30 1.21 1.33 I .39
0.55 0.62 0.59 0.63 0.62 0.63 0.63 0.58 0.57 0.57 0.65 Av 0.60 f 0.03
1.11
1.14 1.13 1.14 1.15 1.14 n.d. 1.16 1.28 n.d.
0.94 0.79 0.97 0.89 1.02 0.87 0.88 n.d. 1.oo
0.95 n.d. 0.92 f 0.07
Xi, 334 nm
762
0
1.40
60
0.64
0.97
A,,, 3 I3 nm
687 759 4340 348 614 754 694 763
0 0 0
0
364 C02 1307 C02 11302 65 1 (1,3-C4H8)
14.0 13.1 12.5 12.8 13.4 13.7 13.4 12.8
7 15 15 20 15 15
7 15
0.72 0.74 0.73 0.70 0.64 0.72 0.72 0.72 Av 0.71 f 0.03
0.97 I .oo 0.98 1.12 I .01 1.12 1.01
n.d. I .03 f 0.06
Xi, 291 nm
669 769 73 I 610
0
0 1 150 COZ 389 0 2
0.724 0.703 0.739 0.748
90 90 90 90
I .05 0.93 0.97 1.01 Av 0.99 f 0.05
0.92 1.05
0.92
1 1 .oo
0.97 f 0.06
7'= 23 "C, photolysis cell V = 29.7 cm3. A = absorbance at 205.7 nm
kcal mol-', as the enthalpy of formation of cyclopropanone a t 298 O K , and increase the range of uncertainty to f l kcal mol-'. Dewar et a1.I6 have previously calculated the heat of formation of cyclopropanone to be -34.1 kcal/mol using the M I N D 0 / 2 method which admittedly underestimated the strain energies of cyclopropane rings. If the group additivity method of B e n ~ o nis~used ~ to calculate the heat of formation, using the specified 27.6 kcal mol-' correction for strain energy in the cyclopropane ring, a value of -13.9 kcal mol-' is obtained. For the group additivity method to reproduce our experimental value, the ring strain energy correction in cyclopropanone must be set equal to 45.3 kcal mol-'. Reasonably enough, this value for the strain energy lies between the 39.3 kcal mol-' calculated for the methylene cyclopropane structure36 and the 53.7 kcal mol-' given for cy~lopropene.~~ Discussion The photochemical results and calculated fluorescence lifetimes show that essentially all cyclopropanone molecules produced in the '(n,r*) state by absorption a t any of the wavelengths used undergo chemical reaction very rapidly, in 5 2 X s. Considerable previous work on the photochemistry of cyclic ketones, in both gaseous and liquid phases, suggests that the initial chemical step is scission of the bond between the carbonyl and a-carbon atoms (ClC2).37-39The intermediacy of the 3(n,r*) state in this process has been directly established in the case of acetone in the vapor phase40 and cyclic ketones in various solvent^,^' but only indirectly indicated in the case of cyclic ketones in the vapor phase.42 Rate constants for intersystem crossing
Table IV. Fluorescence Yields and Lifetimes"
4f ,A,
nm
365 334 313 302.5 29 I
CP
rr, s
CBd
6.8 X
