TRIPLET ~~ET'HYLENE RADICAL REACTION WITH Cis-BUTENE-2
191
Triplet Methylene Radical Reaction with cis-B~tene-2~. by D. F. Ring'b and B. S. Rabinovitch Department of Chemistry, University of Washington, Seattle, Washington 08106
(Received June 26, 1967)
A study of the pressure dependence of the Cj products of the reaction between cis-butene-2 and triplet methylene radicals is reported. Triplet methylene was produced by collisional deactivation (in the presence of a large excess of nitrogen) of the methylene produced by the photolysis of diazomethane with 4358-A radiation at 23", and with unfiltered light at 23 and 56". Reaction pressures were varied from 0.3 to 3.5 atm; a constant ratio of cis-butene-2 : diazomethane :nitrogen was maintained at 20: 1 :30,000. The observed pressure dependence of the product proportions indicates that alkene products arise, in part, by isomerization of the triplet dimethylcyclopropanes or diradicals, rather than exclusively by a concomitant noninterceptible formation reaction. There is a pressure dependence of the ratio of trans- :cis-dimethylcyclopropane which, together with the pressure variability of products such as 3-methylbutene-1, dictates caution in the interpretation of the proportions of T H 2 :1CH2 from product proportions in photolysis systems and provides an alternative interpretation of a previous finding of the authors.
Introduction It has been reported that the ratios of products,2a 'CH2:CH2, and of secondary products that arise on photolysis of diazomethane and ketene in the presence of butene-2, may vary with pressure, up to several atmospheres or more.2b I n this connection, no information has been available on the proportions of the products that arise in pure 3CH2-butene-2 systems at the higher pressures of interest, although Cvetanovii: and Duncan have reported on systematic studies at lower pressures. I n the present investigation, diazomethane was photolyzed at 4358 A, and with unfiltered light, in the presence of cis-butene-2 and excess nitrogen. The pressure of the system was systematically varied up to 3.5 atm, while constancy of composition of the reaction mixture was maintained. Singlet methylene, which is formed upon photolysis of diazomethane, may be collisionally deactivated in the presence of excess inert gas to ground-state This singlet-triplet deactivation technique can be used to produce 100% of 3CH2 in a reaction system, provided that the system is sufficiently dilute in substrate so that the initially formed 'CH2 molecules undergo a sufficient number of prior collisions with the inert gas.
Experimental Section Materials. Diazomethane (DM) was prepared by the reaction of 12 A l KOH and N-nitrosomethylurea; it was purified by trap to trap distillation and stored in dibutyl phthalate at -196"; the samples were shielded from light at all times. Upon warming of the butyl phthalate solution, the initial-pressure fraction
was discarded in order to remove any ethane or ethylene impurities. Phillips research grade cis-butene-2 (99.89%) was used without further purification; a trace of transbutene-2 was the only impurity. Airco dry nitrogen was purified by passing it through a hot copper trap, to remove oxygen, and then through two packed traps at - 196". Apparatus and Procedure. Gas handling of the cisbutene-2 and nitrogen was performed on a greaseless section of the vacuum system. Diazomethane was measured out in a well-seasoned section of the vacuum system, reserved for diazomethane handling. The only reactor used was a 1000-cc Pyrex bulb, which was pumped to mm before loading with reactants; it was well seasoned with diazomethane. The light source was a G.E. AH-6 high-pressure mercury arc lamp. Dow-Corning no. 5543 and no. 3389 filters were housed in a Pyrex water-cooled jacket; these filters and light source provided a median wavelength at 4358 A ; experiments were also performed with the unfiltered radiation of the AH-6 lamp. Reaction was carried out at 25 and 56" with mixtures of constant
(1) (a) This work was supported by the Office of Naval Research; (b) abstracted in part from the Ph.D. thesis of D. F. Ring. (2) (a) J. W. Simons and B. S . Rabinovitch, J . Phys. Chem., 68, 1322 (1964); (b) B. S. Rabinovitch, K. W. Watkins, and D . F. Ring, J. Am. Chem. SOC.,87, 4960 (1965). (3) F. J. Duncan and R. J. Cvetanovii, ibid., 84,3593 (1962). (4) (a) G. Herzberg, Proc. Roy. SOC.(London), A262, 291 (1961); Can. J . Phys., 39, 1511 (1961); (b) F. A. L. Anet, R. F. W. Bader, and A. M. Van der Auwera, J . Am. Chent. SOC.,82, 3217 (1960); (c) H.M. Frey, {bid., 82, 6947 (1960); (d) H. M. Frey, Progr. Reaction Kinetics, 2,131 (1964); (e) R. F.W. Bader and J . I. Generosa, Can. J. Chem., 43, 1631 (1965). Volume 72,Number 1 January 1968
D. F. RINGAND B. S. RABINOVITCH
192 composition of cis-butene-2 : DM :Kz = 20 : 1:30,000. The extent of photolysis of DIVI was 60-95%. Analysis. Products were separated from the nitrogen, upon completion of photolysis, by pumping the reaction mixture through packed traps at - 196". Gas chromatographic analysis was used. The analytical column was 26 f t of' 15% dinonyl phthalate on 60SO mesh Chromosorb P. Both flame and thermal detectors were used on different occasions. All products were identified with authentic samples. +
Results Nitrogen Dilution. Frey4d has noted that in the identification4" of the spectrum of aCH2,Nz:D M = 500 : 1 was employed; he proposed that about 50 collisions are required to bring lCH2 to its ground v i b w tional state and the remaining 400 or more collisions take 'CH2 to the triplet ground state. He4chas worked with dilutions of argon:&-butene-2 up to 1600:l and found that 3CHz production increased with dilution. Bell5 has estimated that ten times as many collisions are required to transport methylene to its ground electronic state as to deactivate vibrationally the singlet state. Bader and G e n e r o ~ ahave ~ ~ also substantiated the increase of 3CHzproduction with increased dilution in cis-butene-2 :inert-gas systems in which ratios up to 950: 1 were employed. Experiments were performed here at nearly constant pressure in which the ratio of Nz:cis-butene-2 was varied from 500: 1 to 2400: 1. The relative yield of products did not appear to vary with dilution between 1200:l and 2400:l. Thus the dilution of 1500:1, which was maintained in the present work, appears sufficient to ensure that collisional deactivation of 'CH, to 3CHzoccurred. Product Yields. The major C5 products from the reaction of aCHz with cis-butene-2 are trans-1,Zdimethylcyclopropane (TDMC) , cis-1,2-dimethylcyclopropane (CDMC), 2-methylbutene-1 (2MBl), 2methylbutene-2 (2bIB2) 3-methylbutene-1 (3MB1), trans-pentene-2 (TP2), and cis-pentene-2 (CP2). I n addition to these products from the photolysis reaction, the DRII-cis-butene-2 system was found to undergo a, dark reaction which produced CDMC, CP2, and 2RfB2. From blank dark runs performed at several pressures, it was determined that the corrections to be applied for the dark products were virtually independent of pressure. At 23", they amounted to 4% CDRIIC, 0.6% CP2, and 2% 2MB2 for the reaction with 4358-A radiation, and 470 CDMC, 1% CP2, and 2% 2MB2 with unfiltered radiation; at 56", they were 3% CDMC, O.S% CP2, and 2% 2XB2 with unfiltered radiation. The experimental data are summarized in Figures 1-8. I n view of the large scatter sometimes evident, subjective interpretation about trends in the data may be misleading. For this reason, we have fitted the data The Journal of Physical Chemistry
c
""I-+ - -e-
040-
a ? . "
,
I
I
I
I
I
I
8 2
I
P
3
( a t m.)
Figure 1. Variation of product composition with total pressure in the DM-cis-butene-2-nitrogen system a t 23' for photolysis with 4358-A light (dashed line and filled points) and unfiltered AH-6 radiation (solid line and open points): total 1,2-dimethylcyclopropane, C D M C TDMC, V ; cis-1,2-dimethylcyclopropane,CDMC, 0; trans-l,2-dimethylcyclopropane,T D M C , 0 .
+
I
2 P (atm.)
3
Figure 2. The change of CP2, V, TP2, A, 3MB1, 0,and 2MB2, 0, with total pressure a t 23' for 4358-A light (dashed line and filled points) and unfiltered AH-6 radiation (solid line and open points).
( 5 ) J. A. Bell, Progr. Phys. Org. Chem., 2, 1 (1964).
TRIPLET ;\/IETIIYLENE RADICAL REACTION WITH CZ'S-BUTENE-8
193
IO1
I
I
I
I
I
I
I
I
3
2
P (atm) Figure 5 . Change in product composition with total pressure in the DM-cis-butene-2-nitrogen system a t 56" for photolysis with unfiltered AH-6 radiation: CDMC TDMC, V ; CDMC, 0; TDMC, 0.
+
Figure 3. Variation of the ratios T D M C : CDMC, 0, TP2: CDMC, A, 3 N B 1 : CDMC, 0 , and 2MB2: CDMC, with total pressure a t 23" for 4358-A light (dashed line and filled points) and unfiltered AH-6 radiation (solid line and open points).
I O5 I
\
I
I
I
1
I
I
I
I
I
I
n
5
n
Q
15
3.0-
2.0N
///
a
0
-
-
\
.-
/.__l___rY 0
5-
0
I
I _
+
by least squares to a three-parameter curve, X = a bp cp2. The curves given in the figure are these least-square fits. Solid curves associated with the open points are fits to the data from photolysis runs with unfiltered radiation; the dashed curves and filled points are for data from photolysis with 4358-A light. The variation of the ratio (CDMC TDMC) :(total C,) with pressure is shown in Figure 1 for the reaction at 23" ; this ratio increases with pressure to a limiting value of 78y0for 4358-8 light and to 80% for unfiltered radiation. Figure 1 also presents the individual yields for CDMC and TDMC. Data for unfiltered and
+
Y
0.0
I
I
2 P (atrn)
3
Figure 4. Variation of the ratios T D M C : CP2, 0, TP2:CP2, A, 3 M B l : C P 2 , 0, and 2MB2: CP2, 0, with total pressure a t 23" for 4358-A light (dashed line and ffilled points) and unfiltered AH-6 radiation (solid line and open points).
I
+
Volume 71,Number 1 January 1968
D. F. RINGAND B. S. RABINOVITCH
194
difference between the increase of dimethylcycloproI panes with pressure as contrasted with the decrease of
I
3
2 P (atrn.1
Figure 7. Pressure dependence of the ratios T D M C : CDMC, 0,TP2: CDMC, A, 3MB1: CDMC, 0,and 2MB2: CDMC, 0, a t 56' for photolysis with unfiltered AH-6 radiation.
(\I
20-
a V \
.-
e
c 0
a
= IDL
a
v
-
(
I
2 P (atrn.)
n
I
n
T
?
3
Figure 8. Variation of the ratios T D M C : CP2, 0, TP2: CP2, A, 3MBl: CP2, 0, and 2MB2: CP2, 0, with change in pressure a t 56" for photolysis with unfiltered AH-6 radiation.
alkene products; the contrast extends to both "singlet" (CDRIC us. CP2, 2RIB2) and "triplet" (TCRIC us. TP2, 3XB1) products. Figure 3 gives the pressure dependence of the ratio TD3IC:CDNC between 0.3 and 3.5 atm. A maximum of 0.43 at 2.2 atm is noted in the 4358A data, although the ratio with unfiltered light increases monotonically with pressure to 0.35 at 3.5 atm. These values are lower than those reported previously found for the 3CH2-cis-butene-2 system : TDMC: C D X C to be 0.96 at dilutions similar to those employed here; Bader and C e n e r o ~ agave ~ ~ a TDMC : CDMC ratio of 0.82; and Duncan and C ~ e t a n o v i ' ~ reported a TDMC: CDMC value of 1.29 for mercuryphotosensitized production of %HZ from ketene; the data of Simons and Rabinovitch2" tended to agree with these. This apparent discrepancy is discussed later ; it led us to examine early results carefully and to repeat a number of experiments; the D3I:butene ratio is lower here than that employed in other work. By contrast with the TDRIC values, the ratios TP2: CDRIC and 3MB1: CDMC, given in Figure 3, decrease with increasing pressure for both light sources. Three ratios of products appear in Figure 4: TDMC :CP2, TP2: CP2, and 3JIB1: CP2. The former reaches a plateau for 4358-A data and increases monotonically with pressure to nearly the same value for unfiltered light. Both TP2:CP2 and 3?tIBl:CP2 are approximately constant for data from both light sources, corresponding to nearly parallel pressure dependence of all alkene products. Figures 5-8 illustrate data obtained with unfiltered radiation for the DhI-cis-butene3 system at 56". With little exception, the data are qualitatively similar to those in Figures 1-4.
Discussion filtered light show a rise to -60% CDMC, while TDMC increases with pressure to 20% at 3.5 atm. Figure 2 shows the yields of TP2, CP2, 2R4B2, and 3MB1. Over the entire pressure range, 2RIB2 is nearly constant at 2.0% for both 4358-A and unfiltered radiation. The compounds TP2, 3RIB1, and CP2 decrease with pressure, to a plateau or minimum for the latter. The decrease of 3MB1 cannot readily be compared with wherein they obthe work of Bader and Cenero~a,~e served an increase of 3MB1 with increased pressure, but with concurrent increased inert gas and hence increased dilution; also, they worked at ratios of cis-butene-2 : DhI, which were very low (2.5: 1) and which invited complicating reactions of DM. The comparison of data for the filtered and unfiltered light sources suggests some possible (but minor) real differences that may be distinguishable from experimental variability. It is worthy of note here that there is a systematic The Journal of Physical Chemistry
Reaction Scheme. Two initial reactions can occur after the photolysis of DR'I DM
+ hv
excess Ka
%H2
+ K2
One of these is addition to the double bond 3CHz
+ cis-CHSCH=CHCHa
+
CH3CH-CHCH3
I
.CH2
+*
(1)
The initially formed triplet diradicala" intermediate is vibrationally excited (*) but probably not electronically excited (t),though some uncertainty exists concerning the electronic ground stage of this diradical.6bio The other reaction of aCH2is H abstraction
TRIPLET AIETHYLENE RADICAL REACTION WITH C~S-BUTENE-~
+ cis-C:HaCH=CHCH3 + *CH3+ CH3CH=CHCH2* + ‘CHI + CH&H=CCH3
195
aCH2
C-+C-C* (2) (3)
Reaction 2 should preponderate over reaction 3. A CH insertion reaction by 3CH2was previously reported’ to occur to an extent in alkane systems, but the presence of the double bond should render such reaction negligible here. The vibrationally excited diradical formed in reaction 1 may cyclize with inversion of spin CHaCH--CHCHat* --+TDMC*,CDMC*
(CP2,TP2)*
c-+c-c*
I
7 kg
and a 1,2-methyl migration by C4 to C5 C-+C-C*
TDMC* -% TDMC
(5)
2MBlj2MB2
4
I
*
CH2 5
the following H- and methyl-shift reactions of the diradical species may be described $0 a 1,Bhydrogen shift from C3 t o C2 2MB1
(6)
+2MB2* -% 2MB2
(7)
--+
2MB1*
I
*C a 1,Zhydrogen shift from C3 to C5
.c a 1,4-hydrogen shift from C1 to C5
I
I
-+CP2*,TP2* -% CP2,TP2 (11)
.C
+ CH,CH=CHCH2.
*CH3
--+
--+
+ CH&H=CCH3
+
2h4B2* -% 2MB2
1 2 3 CHa-QH-CH-CH3
C-C-C-C*
3 n 1 ~ i *-% ~ M B I (io)
CP2*,TP2* -% CP2,TP2 (12)
By numbering the carbon atoms
C-C-C-C*
--+
*C
-% CP2,TP29
(2RIB1,2MB2)*
C-C-C-C*
(9)
a 1,2-methyl migration by C4 to C2
aCH3
a CDhIC*
-% E C
Reactions 2 and 3 may lead to recombination of the initially formed radicals
CH2
Further reactions of TDMC* and CDMC* may be summarized*
CDNC
EC*
.C
(4)
I
*
---f
3MB1* -% 3MB1
(8)
.C a 1,3-hydrogen shift from C4 to C2 with ring closure (ethylcyclopropane = EC)
(13)
Although the cis configuration tends to be preserved in the methallyl radical,” TP2 will also be formed by (6) (a) The existence of a triplet diradical intermediate has been suggested by other workers: see ref 4a-c; also K. R. Kopecky, G. S. Hammond, and P. A. Leermakers, J . Am. Chem. Soc., 8 4 , 1015 (1962); R. W. Carr and G. B. Kistiakowsky, J . Phys. Chem., 70, 118 (1966); C. McKnight and F. S. Rowland, J . Am. Chem. Soc., 8 8 , 3179 (1966); R. J. CvetanoviE, H. E. Avery, and R. S. Irwin, J . Chem. Phys., 46, 1993 (1967). (b) R. Hoffman (151st National Meeting of the American Chemical Society, Pittsburgh, Pa., March 1966, paper 109K) has reported calculations from LCAO theory, using a basis set of 2s and 2p carbon and 1s hydrogen AO’s which indicate that the trimethylene diradical has a singlet ground state in which the three C’s and four terminal H’s are coplanar and the CCC angle is approximately 125O; the two electrons on the diradical are reported to occupy a 1,3 antibonding antisymmetric orbital: this ostensibly results in a tendency to add 1,2 to olefins and to close to cyclopropane in a conrotatory manner. The ground-state diradical has barriers to rotation of the terminal CHZgroups, while the first excited singlet and triplet states are calculated to possess nearly free rotation for these groups. (c) H. E. Simmons, private communication to A. Mishra, J . Am. Chem. Soc., 88, 3963 (1966), carried out a Pariser-ParrPople calculation for trimethylene and found that the ground state is predicted to be triplet. (7) D. F. Ring and B. S. Rabinovitch, &id., 88, 4285 (1966). (8) H. M. Frey, Proc. Roy. SOC. (London), A251, 575 (1959). (9) J. A. Bell6 has suggested that linear pentenes may be explained if the diradical introconverts between a linear conformation (C-C-C-C-C) and the branched configuration formed in reaction 1, but this suggestion was questioned by G. Z. Whitten and B. S. Rabinovitch ( J . Phys. Chem., 69, 4348 (1965)) on the basis of earlier results obtained by Placzek (D. W. Placnek and B. 5. Rabinovitch, Can. J . Chem., 43, 820 (1965)). McKnight and Rowlandaa have further indicated that such a rearrangement is not a major source of pentene from dimethylcyclopropane. Very recently, Cvetanovii, et aZ.,es have presented further strong evidence against the proposed introconversion. (10) These reactions of the diradical have been suggested previously. Duncan and CvetanoviC8 proposed the relative importance of the following reactions: 10 and 11 > 8 > 7 > 6. Reaction 9 waa suggested by D. W. Setser and B. 5. Rabinovitch (Can. J . Chem., 40, 1425 (1962)). Carr and Kistiakowsky& have included reaction 8 in a reaction scheme, while McKnight and Rowlandea believe reactions 6, 7, 8, 10, and 11 may occur. (11) R. F. Kubin, B. S. Rabinovitch, and R. E. Harrington, J . Chem. Phys., 37, 937 (1962).
Volume 72, Number 1 January 1968
D. F. RINGAND B. S. RABINOVITCH
196 (12). Allylic resonance can lead to 3MB1 in the following manner CH&H=CHCH2. CH3CHCH=CH2
CHaGHCH=CH2
(14)
+ *CH3+ 3MB1* A 3MB1 (15)
In addition, other disproportionation and combination reactions of the radicals can occur t o give methane, ethane, trans-butene-2, cis-butene-2, l,Zbutadiene, 3methyl-2,5-heptadiene1 3,4-dimethyl-2,4-hexadiene, and 2,6-octadiene. A whole host of products, many in trace or minor amounts, actually are found in this system. We have defined and limited our objective to a clarification of the major Cs products, which conventionally have drawn the most attention and which are most closely related to the initial processes. While an exhaustive study of all of the products would assist the understanding of secondary and later radical reactions that occur, these may not contribute to our primary objective, and, since our interest does not extend to them per se, we have chosen to restrict our study to major products from Cz through Cs. Wavelength and Temperature Efects. The data in Figures 1-4 exhibit possible systematic differences for the two light sources used. Unfiltered (more energetic) radiation produced more CDRIC, while enhanced relative yields of TDMC, TP2, and 31IB1 were found with 4358-A light. The differences in product yields for the two light sources tended t o lessen as the pressure was increased. It might be expected that data for both light sources would be identical, since any wavelength effect in the production of 3CHz should be eliminated by the subsequent collisions with the inert deactivating gas. The observed difference, if real, could be explained if excited singlet D X t were capable of adding to cisbutene-2 in a predominantly stereospecific manner, and if the probability of this (minor) reaction were increased relative to decomposition at higher vibrational levels of DMt. Such reaction could, in any case, be only a minor component of the total reaction. Comparison of the data for unfiltered radiation in Figures 1-4 with that in Figures 5-8 indicates that the yields of (CDNC TDAIC) and of CDMC are lower at 56 than at 23”, but that the yields of TDAlC, CP2, TP2, and 3MB1 are higher at 56 than at 23”. These changes in yields could be rationalized by postulating that increase of temperature tends to enhance reaction of D M t with cis-butene, but that it also favors the formation of TDRIC in reaction 4, if an appreciable activation energy exists for rotation about the single (C2-C3) bond in the diradical, as has been proposed.lZ Previously, Setser and Rabinovitchlo deduced that 1CH2carries approximately 22-kcal excess energy from
+
The Journal of Physical Chemistry
its genesis, on the basis of AHr0(’CH2) = 90 kcal mole-’. Allowance for a possible difference in heats of formation of ‘CH2 and 3CH2of 5-10 kcal mole-‘ could reduce the exothermicity of reaction 1 to s(110 29.5) 5 80.5 kcal mole-’. Depending upon the energy splitting, z, between singlet and triplet trimethylene, the excess internal energy of the triplet diradical might be 5 (80.5 - 62 - z) = 5 (18.5 - z) kcal. Higher temperature would also favor alkene production reactions 6-11. The above is a qualitative speculative explanation of effects whose magnitudes are not large and may even be in doubt owing to experimental scatter. Product Trends with Pressure. The most important result of the present study concerns the consistent pressure trends of product formation found under all of the experimental conditions. The percentage of both TDMC and CDMC in the Cs product increased with rise in pressure. By contrast, the percentage of alkenes decreased as the reaction pressure was raised; the yields of CP2, TP2, 3iliIB1, and 2MB2 show similar pressure dependence; the three former compounds were generated in comparable amounts, while the latter arose to a lesser extent (a few per cent) a t all pressures; in addition, 2MB1 and E C were found, but only in trace amounts; each comprised less than 0.3% of the total Cs product. The observed pressure trends are in contrast to those observed over a similar pressure range from the T H 2 cis-butene-2 reaction system4din which only the major product, CDRIC, increased with rise of pressure while other major products, CP2 and 2MB2, remained nearly constant; TDMC, TP2, 3nfB1, and 2MB2 were minor products and decreased with increase of pressure. The 2MB2 product is relatively the most reduced on going from a “singlet” to a “triplet” system and, of all the products, thus appears most characteristically “singlet” in nature; conventionally, CDMC and (to a lesser extent) CP2 have been throught of as characteristic “singlet” products of reaction of methylene with cisbutene. The plots of ratios of the various products us. CDMC indicate that TDMC :CDnIC increased (to a plateau or shallow maximum) with rise of pressure, while TP2: CDMC and 3RfB1: CDMC decreased and showed very similar pressure dependence (Figures 3 and 7). If the products are plotted us. CP2, instead, TDRIC: CP2 increases dramatically with rise in pressure while
-
2&740
(12) R. J. Crawford and A. Mishra (J.A m . Chem. Soc., 88, 3963 (1966)) have estimated that the linear CS diradical, formed upon pyrolysis of l-pyrazolines, exhibits a bonding energy of about 8-12 kcal between the radical sites. Thus, if the diradical formed in reaction 1 is not free to rotate about the central C-C bond, the original cis configuration would tend to be conserved. Inasmuch as the evidence discussed in ref 9 appears to rule out a cyclised intermediate, the experimentally436observed tendency for conservation of the original butene-:! configuration in the product cyclopropanes formed by SCHZaddition strongly supports the existence of a rotational barrier in the unsymmetrical diradical.
TRIPLET n‘f ETHYLENE RADICALREACTIONWITH Cis-BUTENE-2 TP2: CP2 and 3RIIB1: CP2 are roughly constant or exhibit possible broad shallow maxima (Figures 4 and 8). Thus the pressure trends of the two sets of ratios of products, constructed relative to the two conventional “singlet” products, are dissimilar. Interpretation of Pressure Trends. A plausible explanation of the data, which accounts for the observed trends in product yields, as well as for the contrasting trends of the two series of product ratios, is that a large fraction of products arises from isomerization reactions of the initial diradical (or triplet cyclopropane) which is formed in reaction 1 and that there is an increasing tendency for reaction 4 to be favored over reactions 6-11 as the hot diradical is collisionally dea~tivated.’~ I n ref 10, it is pointed out that Cvetanovib, et aZ., had proposed that the relative importance of the reactions of the diradical to produce alkenes should be 10 and 11 > 8 > 7 > 6 ; this proposed hierarchy is supported experimentally here. Only part of the alkene products are formed by concomitant noninterceptible initial reactions. This part includes H abstraction by 3CHz (reactions 2 and 3 followed by 12-15) ; extrapolation of the alkene proportions at highest pressures plausibly corresponds to these. It is possible that at higher pressures the cis configuration in reaction 4 is favored by more rapid collisional stabilization of the initially formed diradical because of the possible existence of an appreciable activation energy for rotation around the central single bond, referred to earlier in another connection. This could help explain a leveling off or decline in the TDMC :CDMC ratio at highest pressure. At the other end, the decline of TDMC :CDMC at lower pressures prompts the following suggestion. Stereospecific CDMC formation by a reaction of D M t with cisbutene-2 may occur; a minor contribution of this kind would be more significant at lower pressures, where it provides a larger component of the total dimethylcyclopropanes because of the enhanced importance of reactions 6-11, and might account qualitatively for the observed lower pressure behavior. Additional mention should be made of the discrepancy which exists between the low values of TDMC: CDMC found here and previous values which have appeared in the l i t e r a t ~ r e . ~ b lAn ~ , ~important difference between the experimental conditions employed in this work and previous studies lies in the high ratio of cis-butene-2 :D M used here which was 20 :1 (TDMC:CDMC 0.4). Other workers have used 1: (TDMC: CDMC 6.7) through 2.5: 14e (TDMC:CDMC 0.8) to 4:14c (TDMC:CDMC 1). I n an attempt to test this possible relation between high cis-butene-2 :D M reactant ratios and high TDMC :CDMC product ratios, experiments were performed in which the ratio of cis-butene-2: D M was decreased from 20: l to l :2; the TDMC: CDMC ratio increased to ~ ~ 0 . which 6, is a t least qualitatively con-
- -
-
197
sistent.14 An explanation of the effect could again involve a reaction between DMt and cis-butene-2. The relative concentration of cis-butene-2 is decreased in going from 20: 1 to 1:2 a t constant pressure, and other processes of DMt are enhanced relative to its reaction with cis-butene-2. Although we are attracted to the suggested reaction between excited singlet D M t and cis-butene-2, in that such reaction helps rationalize three aspects of the data, a quantitative consistent interpretation on this basis does not seem at hand and this reaction must be considered speculative. TDMC:CDMC as a Measure of 3CH2 Proportion. It has been shown above that the experimental TDMC: CDMC ratio is a function of pressure even in this all-triplet system. This ratio has been employed by usBbearlier as an indicator of the relative amount of 3CH2to lCH2reaction in ordinary ketene and Dnil photolysis systems. Our present results indicate that restrictions on the use of this (and similar) diagnostic
(13) Analysis was also made in this work for C2H4, C&, C3H6, C&8, n-C4HlO, and trans-C~Hs-Z. The data are not included here, but details may be found in ref lb. In general, a t both 23 and 58 O , for both filtered and unfiltered light, the proportions of these lower products also decreased with increase of pressure. (The proportions of the total C1-Cb products a t lowest and highest pressures were, respectively: ethylene, 43.4, 31.5%; ethane, 9.3, 4.7%; propene f propane, 5.3, 1.8%; n-butane, 3.4, 3.1%; and trans-butene-2, 6.3%, 4.1%.) This suggests that, a t least in part, they too arise from decomposition reactions of excited Cs species. All of these products (excluding trans-butene-2) were also found, to some extent, in blank runs that were made without substrate. Relative amounts of compounds which were formed in blank runs were: ethylene, 67,5y0; ethane, ll.O’%; propene, 3.4%; propane, 2.8% ; cyclopropane, 1.1%; isobutane, 0.2%; n-butane, 0.6%; butene-1, 9.2%; 1,4pentadiene, 1.1%; isopentane, 0.1%; neopentane, 0.1% ; n-pentane, 0.2%; pentenel, 1.4%; 3MB1, 0.1%; 2MB2, 1.1%; and 4-methylpentene-1, 0.1%. Excluding the ethylene product (which arises in considerable amount from the reaction of triplet methylene with DM), the C6 products constituted approximately 56% of all C2-Ca products a t low pressure and 80% a t high pressure. (14) A referee has suggested that the discrepancy between our data and that of others may be due, for example, to (our) analytical error. This may be so. However, we are willing to state categorically our belief that the chemistry and products of 3CHz systems are more complex than for 1CHz systems and that high proportions of DM in triplet systems invite serious complications. (15) NOTE ADDEDIN PROOF: Experiments with trans-butene-2 reactant yielded: 3MB1, 4.4%; TDAMC,76.0%; TP2, 7.4%; CP2, 3.1%; CDMC, 8.2oJ,; and 2MB2, 0.9%. These results for cis- and trans-butene-2 differ from F. S. Rowland (private communication) in regard to relative amounts of CDMC and TDMC products. Experiments on cis-butene-2-Ne with an excess of added CO which removes ~CHZ [B. A. DeGraff and B. G. Kistiakowsky, J. Phys. Chem., 71, 1553 (1967)l did not alter the proportions of CDMC and T D M C although the total c6 yield was reduced; it seems that ‘CHZ was not a major reaction component. A liquid ~CHZ-cis-butene-2system was produced by photolysis of D M at 23O in 200-fold excess of C3F8. The yields were: 3LMB1, 6.1%; TDMC, 13.3%; 2MB1,1.9%; TP2,7.1%; CP2,9.3%; CDMC, 60.4’%; and 2MB2, 1.9%. These yields, which are characteristically “triplet,” may be compared with the characteristically “singlet” products formed on photolysis of DM at 23’ in neat liquid cisbutene-2: 3MB1, 0.2%; TDMC, 0.470; 2MB1, 0.3%; TP2, 0.0%; CP2,39.1%; CDMC, 47.5’%; and 2MB2, 12.6%. These latter results also contrast with gas-phase results (in the absence of Nz) where some “triplet” product does arise.% Evidently, in the presence of a liquid-phase interceptor, no intersystem crossing of excited singlet DM to a triplet surface occurs. The triplet products formed in liquid perfluoropropane can result from intersystem crossing by DM and obviously also by collision deactivation of GHz. Volume 72,Number 1 Januaru 1988
STANFORD L. SMITHAND RICHARD H. Cox
198 ratios exist, namely, that the value of TDMC: CDMC depends on both the pressure of the system and the ratio of cis-butene-2: DM. This cautionary remark applies as well to much of the recent work in the literature on ordinary photolysis systems in which %H2 has
been found. Our previous finding of an increase with pressure of the ratio TDMC: CDMC, up to a plateau or maximum, now may have an explanation other than the previously alleged variation of the C H 2 :‘CH2 proportion with pressure.16
Solvent-Dependent H-H Couplings in Hexachlorobicyclo[2.2,1] heptenes’ by Stanford L. Smith and Richard H. Cox Depurtment of Chemistry, University of Kentucky, Lexington, Kentucky
40606 (Received June 89, 1967)
The solvent dependence of 2 J and ~ cis ~ and trans 3 J is ~investigated ~ in 1,2,3,4,7,7hexachlorobicyclo[2.2.1 Iheptenes substituted a t the 5 position with chloro, cyano, phenyl, hydroxy, carboxyl, or acetate groups. In all six compounds, ‘ J H Hdecreased (in the absolute sense) in solvents of increasing dielectric constant. The magnitude of the change ~ for all compounds except the hydroxyl- and carboxyl-substituted cases in 2 J is ~similar where strong association effects are evident. Changes in ’JHHwere 0.25 Hz or less in all cases. The general solvent effect (in the absence of specific association) is attributed to the solvent electric field which, in turn, is controlled by the net electric dipole of the solute molecule.
Introduction The solvent dependence of spin-spin coupling constants in many relatively rigid molecules is now an established phenomenon whose causes and characteristics are rapidly being e l u ~ i d a t e d . ~Previous ,~ investigations have focused on geminal H-H and H-F couplings in sp2 hybridized systems and on various single-bond c~uplings.~JSeveral examples of solventdependent vicinal couplings are also known, again in sp2 hybridized sy~tems.~*~+’OThe solvent dependence of J,,, across C2 in 4-methyl-l,3-dioxolane2 and the solvent dependence of J,,,,, in dl-dibromosuccinic anhydride” are the only known examples of solvent-dependent couplings in “rigid” spa hybridized systems. I n neither case are the results completely unambiguous. A number of investigators have noted ~ an apparent solvent dependence of 2 J in~ flexible systems such as 1,2-dichloro- and dibromopropanes,l2 various 1,1,2-trisubstituted ethanes,13 and 3,3-dimethylbutyl iodide.14 Unfortunately, it is not possible to determine whether these observed changes arise directly from some intrinsic solvent effect or indirectly from solvent-induced changes in conformer populations. In the latter event, changes in orientation of substituents with respect to the methylene group would be expected’s to produce small changes in the geminal H-H coupling. The existence of these examples The Journal of Physical Chemistry
raises potentially serious questions concerning the utilization of H -H (and H-F) couplings in the study of concentration and solvent effects on conformationally mobile systems. We therefore decided to examine and Hcis and trans 3 J ~ ~ the solvent dependence of ~ J H in a rigid spa hybridized system. (1) A summary of this work was presented at the Southeastern Regional Meeting of the American Chemical Society, Louisville, Ky., Oct 1966. (2) S. L. Smith and R. H. Cox, J . Chem. Phys., 45, 2848 (1966), and references therein. (3) 5. L. Smith and A. M. Ihrig, ibid., 46, 1181 (1967),and references therein. (4) V. S. Watts and J. H. Goldstein, J . Phys. Chem., 70, 3887 (1966). (5) H. M. Hutton, E. Bock, and T. Schaefer, Can. J . Chem., 44, 2772 (1966). (6) R. H. Cox and S. L. Smith, J . Mol. Spectry, 21, 232 (1966). (7) S. L. Smith and A. M. Ihrig, ibid., 22, 241 (1967). (8) H.M. Hutton and T. Schaefer, Can. J . Chem., 43, 3116 (1965). (9) H.M.Hutton and T. Schaefer, ibid., 45, 1111 (1967). (10) P. Laszlo and H. J. T. Bos, Tetrahedron Letters, 1325 (1965). (11) L. E.Erickson, J . A m . Chem. SOC.,87, 1867 (1965). (12) H. Finegold, J. Chem. Phys., 41, 1808 (1964). (13) E.I. Snyder, J . A m . Chem. Soc., 88, 1155 (1966). (14) G. M. Whitesides, J. P. Sevenair, and R. W. Goetz, ibid., 89, 1135 (1967). (15) A. A. Bothner-By in “Advances in Magnetic Resonance,” Vol. 1, J. S. Waugh, Ed., Academic Press Inc., New York, N. Y., 1965,p 197.
