STANLEY J. CRISTOLAND DENNISD. TANNER
3122
constants are concerned, i t is evident that the fulvenes are better treated as dienes than as aromatic systems. The C13-H spin coupling results are informative regarding the molecular orbital predictions that the exocyclic carbon-6 in the fulvene structure should bear a positive charge. Juan and Gutowsky have derived relationships between J c H , the s-character of the hybrid orbitals of the C-H bond ( u H ? ) , and a term A x which measures the “affinity” of a substituent X for the carbon 2s-character. They have shown that the value of Ax increases with the electronegativity of the substituent and t h a t the effects of multiple substituents on U H ~are additive. Using their relationships and the for the methyl group in dimethylfulvalue of the JCH vene one calculates a value of A, of 0.0168 for the fulvene residue. Using their additivity relationship and their value for a phenyl group gives a calculated value for the methylene group in dibenzylfulvene of a H 2 of
[COXTRIBUTION FROM
THE
Vol. 86
0.257 which agrees well with the experimental value of 0.256. A similar set of calculations was carried out using J c H for the methyl group in isobutylene (126 C . P . S . )and ~~ our values for allylbenzene. The value of A x for the olefin substituent was found to be 0.008. Calculated and observed values for the allylbenzene methylene were uH20.254 and 0 . 2 5 5 , respectively. The values of Ax of 0.0168 for the fulvene residue and 0.008 for the olefin residue point up the greater electronegativity of the former group, a result in keeping with the enhanced positive character of C-6. Acknowledgment.-U’e wish to acknowledge the generous financial aid of the Robert A. \i5:elch Foundation in support of this work. The helpful suggestions of D r . K. B. LViberg, D r . Wallace Brey, Jr.> and Dr. J. N. Shoolery are also gratefully acknowledged. (27) N . Muller and D E . Pritchard, J . A m . Chem. Soc , 31, 1471 (1961).
DEPARTMEKT OF CHEMISTRY, USIVERSITYOF COLORADO, BOULDER, COLO.]
Bridged Polycyclic Compounds. XXVI. The Solvolysis of Some 4-Chlorodibenzobicyclo [3.2.1 Ioctadienesl BY STANLEY J. CRISTOLA N D DENNISD . TANNER RECEIVED FEBRUARY 24, 1964 The solvolyses of ero- (11) and endo-4-syn-S-dichlorodibenzobicyclo[3.2.l]octadiene (111) and of exo-4-chlorodibenzobicyclo[3.2.l]octadiene ( V ) have been studied. .ketolysis of I11 and T‘ proceeds with good first-order kinetics to give the em-acetate IT’ and its dichloro analog, respectively. The acetolysis of I1 t o IT’ is accompanied by epimerization to 111, involving ion-pair intermediates. Solvolysis of I1 in pivalic acid gives the endo-pivalate. Various carbonium ions are considered as intermediates. and it is concluded t h a t a classical benzylic carbonium ion can accommodate the results.
Introduction In a recent study2 of the ionic chlorination of 9,lOdihydro-9,lO-ethenoanthracene(I) the sole products of the reaction were found to be exo- and endo-4-syn-8dichlorodibenzobicyclo[ 3 . 2 .lloctadiene (11 and 111).
I
I1
111
The major isomer, compound 11,produced in 88% yield, was shown to be the less stable of the two dichlorides. In view of our interest in such compounds, we have investigated the rates and products of solvolysis of the two isomers. Results Solvolysis rates were followed in dry acetic acid, with added potassium acetate. A plot of the experimental data for the acetolysis of the endo isomer I11 a t 74.74’ gave a constant first-order rate which showed no observable trends. Only one product was obtained from solvolysis of 111; this product was identified as syn-8-chlorodibenzobicyclo [3.2.1Ioctadien-exo-2-yl ace( 1 ) Previous paper in series: S J. Cristol and D. I Davies, J Ovf Chetn., 29, 1282 (1964). T h i s work was reported a t t h e 1 8 t h National Organic Symposium of t h e American Chemical Society, C o l u m b u s , Ohio. J u n e , 1969. ( 2 ) S J. Cristol, R . P Arganbright, and D D T a n n e r , J O r g C h r m , 28, 1371 (1963)
c1
,
IV
V
VI
tate (IV).2 The specific rate constant obtained graphically was 1.68 X 10F6 sec.-l. An average of three kinetic runs was used to obtain this value. When the exo isomer I1 was solvolyzed, a first-order plot of the experimental data showed a considerable downward drift. The steadily decreasing rate was shown to be caused by the concomitant isomerization of the exo to the endo isomer during the course of the solvolysis. Thus, when a sample of the exo isomer was allowed to react to about 5070 completion, the remaining (43%) dichloride which was recovered was the endo isomer 111. The acetolysis product appeared to be entirely syn-8-chlorodibenzobicyclo [3.2.1Ioctadien-exo2-yl acetate (IV), the same product obtained from the solvolysis of 111. This behavior is analogous to t h a t observed in the acetolysis of a , a-dimethylallyl ‘chloride, and the same kinetic procedure was used in this work to calculate the specific rate of solvolysis ‘for I1 and the rate of isomerization of I1 to 111, except that an analytical procedure, rather than a graphical one, was used to calculate the specific reaction rate constant^.^ ( 3 ) W G Young, S Winstein. and H I. Goering, J A m ChPm .SOL, 1 3 , 1958 (1951) ( 4 ) Details of this procedure are given in t h e P h D Thesis of D D T a n ner, University of Colorado. 1961
Aug. 5 , 1964
S O L V O L Y S I S OF 4 - C H L O R O D I B E N Z O B I C Y C L O
TABLE I RATE O F S O L V O L Y S I S O F exU-4-syn-8-DICHLORODIBEC ZOBICYCLO[3 2 11OCTADIENE AT 74 74 f 0 05' I N GLACIAL ACETIC XCID Sample data for a typical kinetic run" Time sec
--[RCI]o Obsd
-
0
0 4920 7860 11400 16200 19500 22900 27700 36000 44300 55200
0 00168 002 17 00294 00387 00468 00549 00591 00650 00715 00806 00934
87100
[ R C l l t moles,l Calcd
-
0 0 00148 00225 00309 00406 00464 00517 00580 00666 00730 OOi93 00936
Instantaneous r a t e c o n s t a n t at 1 106kl sec - 1
15 i 14 4 13 6 1% 6 11 2 10 3 9 36 8 11 6 29 4 97 4 02 3 35
Alkyl a Initial concentration: 0.0207 X 11, 0.0342 M KOAc. halide consuIned, as computed by the procedure ; standard error of estimate = 0.00019.
[3.2.1I O C T A D I E N E
rearrangement andlor return prior to titrimetrically measurable dissociation. 3 , 5 - i T h e acetolysis rate of exo-4-chlorodibenzobicyclo[3.2.l]octadiene (V) was also investigated; V showed steady first-order kinetics, k = 58 X sec.-l, when the initial concentration of the alkyl chloride samples was corrected for solvolytically active chloride, calculated from its infinity titer.8 The acetolysis of both I1 and I11 yields as the sole isolable reaction product syn-8-chlorodibenzobicyclo[3.2.1]octadien-exo-2-yl acetate (IV). The stereochemistry of IV was assigned for the following reasons. Acetolysis of I1 and I11 gives the same chloroacetate IV as was obtained by the ionic addition of t-butyl hypochlorite in acetic acid to I ; since the structure of I V had been established as either syn-8-chlorodibenzobicyclo [3.2.lloctadien-exo- or endo-2-yl acetate, only the configuration of the acetoxyl group a t the %position remained to be established.2 , 4 Evidence suggesting a strong diaxial interaction has been found for the syn-cxo-dichloride I I . * If it is as-
TABLEI1 COMPARISON O F RATEO F SOLVOLYSIS O F exU-4-syn-8-DICHLORODIBENZO[32 11OCTADIEUE IN ACETICACIDAT 74 74"
T i m e , sec
Reaction, 7" [KOAclo = 0 034 M [RXlo = 0 020 J4
3000
5
6800 11500 16100 22500 39500 69000
10 15 20 25 33 42
WITH A N D \L ITHOUT >ADDED
Reaction, 7" [KOAcIo = 0 034 A4 [RXIo = 0 020 .M
Reaction, R [KClJo = 0 0 0 9 1 M [KOAcIo = 0 0 2 5 M [RXlo = 0020 M
5 10 15 20 25 35 45
5 10 14 18 23 34 44
Table I gives the data for a sample kinetic run for 11. T h e calculated values of the instantaneous rate constants for solvolysis evaluated from the slope of the synthesized curve decrease from 16 X 1 0 P to about 3.3 X 1 0 P set.-' (approaching the rate of the solvolysis for the endo isomer) by the time the reaction is approximately 45yo complete. The specific reaction rate constant for isomerization was found to be approximately 24 X 10F set.-'. T h e solvolysis of the exo-chloride I1 was examined for its dependence on added chloride ion a t constant ionic strength. An accurate value for the rate of solvolysis with added chloride could not be readily obtained owing to the interference of appreciable concentrations of chloride ion with the observation of the titrimetric end point.3 However, comparison of the values for per cent reaction vs. time, shown in Table 11, indicates t h a t the rates of solvolysis and ionization are quite insensitive to initially added chloride ion within the limits of experimental error. Thus, there is no detectable evidence of either bimolecular isomerization of em-chloride or reaction of a carbonium ion with external chloride ion. The reaction then clearly represents another example of an internal epimerization of a type previously reported b y S. Winstein and co-workers, where p-chlorobenzhydryl chloride undergoes intramolecular racemization in conjunction with its solvolysis in glacial acetic acid.5 Numerous studies have established the occurrence of ion-pair 13) S "instein, J S Gall, 51. Hojo, a n d S S m i t h , 3 A m . C h r i n S O L , 8 8 , 1010 i 1 9 6 0 ) , t h e t r p s f o r m a t i o n of P S O t o eriilo isomer, in o u r s y s t e m , is stereochemically similar t o racemization in their case.
3123
CHLORIDE
IOU
Reaction, 70 [KClIo = 00168 M [KOAclo = 0 0 1 7 M IRXIo = 00164 M
5 9 14 17 21 30 39
sumed t h a t chloroacetate I V is the syn-chloro-exoacetate and arises from attack of acetic acid or acetate anion from the exo side, then a very large anion, larger than chloride or acetate, might attack the benzyl position from the endo side of the carbonium ion and result in a product related to the epimeric alcohol &Then I1 was treated with silver pivalate in pivalic acid and the resulting ester was hydrolyzed, a chlorohydrin, VII, epimeric to the one derived from the acetolysis product, was obtained. In the case of I1 and 111, the endo isomer 111 was shown to be the thermodynamically more stable product ; likewise, when the chloroacetate IV was treated with perchloric acid in acetic acid, the isomeric chloroacetate V I I I , derived also from VII, was obtained. U'hen VI1 was subjected to permanganate oxidation, the chloroketone, IX, that was isolated from the oxidation of synS-chlorodibenzobicyclo [?.2.1]octadien-cwo-%ol (X),? was produced. For these reasons, X has been designated as the syn-S-8x0-2-chlorohydrin and VI1 as the syn-S-endo-%chlorohydrin. (6) S Winstein, el a i , ibiri , 76, 2597 (19.54). 7 8 , :328. 2 7 6 3 2 7 6 7 . 2781 i I % j 6 ) ; 80, 169, 439 (19.781, C h r m I,rii ( I m n d o n ) . 664 (l!)j4) H t . 1 ~C h i m Acla, 41, 807 (19.58) (7) G K o h n s t a m a n d I3 Shillaker. J C'hriii S o c , 11115 ( I R 5 9 ) ( 8 ) T h e sample of V utilized for t h e r a t e determination was prepared by t h e reaction of 1 l-dibenzohicycla[2 1 S]octadienol with thi