COMXUNICATIONS TO T H E EDITOR
504s
a singlet of the type of form 11, referred to6 as the "charge-transfer" band. Addition of hydrogen chloride to the 2-propanol solution leads to no change in the 249-inp band, but t o marked decrease of the band a t 352 mp. The pk', of I)MAl13P.estimated from the decrease in this absorption with addition of hydrogen HC1 the long chloride, is 1 . 7 in 2-propanol. At 2 wave length band is a t 314 mp and of low intensity (log 6 2 . 2 8 ) , comparable t o the corresponding band of benzophenone a t 334 mp (log e 2.18); i t appears to arise from an n~-r*transition which leads to a chernically reactive state, presumably the triplet.8 Photoreduction of 0.1 -If DMXBP in 2-propanol in Pyrex by a C,. E. 851V h3 ultraviolet lanip proceeds very slowly, about n.00:3 the rate of benzophenone. The rate however rises regularly with addition of hydrogen hydrochloride, while the absorption a t long wave length decreases, leveling a t ca. 0.7 S HC1 a t a rate about 0.2 t h a t of benzophenone, indicating a quantum yield of about 0.4.The rate of photoreduction of benzophenone in 2-propanol is unaffected by 0.5 HCl. The rate of photoreduction of DMXBP in 2-propanol appears to rise more rapidly than the "charge-transfer" band is suppressed with addition of hydrogen chloride, indicating that the ground state may be a weaker base than is the n-r* excited triplet. In 1 : 1 2-propanol-water the rate of photoreduction of DMXBP is negligible and is again greatly increased by addition of hydrogen chloride. Higher concentrations of the acid are required to reach maximum rate of photoreduction than in the absence of water, reflecting the lower protonating power of the aqueous system. IYater reduces the maximum rate of photoreduction of DM;\BP and also that of benzophenone itself, effects which will be described in a subsequent report. The product of photoreduction of 0.1 M DMABP in 0.5 LV HC1 in %propanol, formed quantitatively, is 4,1'-bis(dimethylaminobenzpinacol) dihydrochloride, m.p. 150-151° dec. This was converted quantitatively to the free base, m.p. and m.m.p. 181-183', identical with an authentic sample prepared by reduction of I) \I.XHP by magnesium-magnesium iodide. :lnul. Calcd. for CJuHa??j202: C , 79.7; H , 7.OX; N , G.19. Found: C, SO.0; H , 7.10; N.6.21. The pinacol was readily distinguished from p-dimethylaminobenzhydrol, 1n.p (Xio, prepared by reduction of the ketone by sodium borohydride. .1 nul. Calcd. for ClsH,7r\;O: C. 7!).:3: H , 7.40; Ti, R . l f i , mol. wt. 227. Found: C.79.5; H , 7 . 2 4 : N, li.19; mol. wt. 230. Preliminary experiments have shown t h a t p - and o-aniinobenzophenorie are also reduced readily in 2propanol in the presence of acid. X 0.01 -21 solution o f the para compound was reduced less than 770 in the absence of hydrogen chloride after irradiation for 21 h r . . and was conipletely reduced in less than this time in the presence of 0.5 2' HC1. o-Aminobenzophenone, 0 . 0 1 .\I. showed no reduction after irradiation for 17 hr. in the absence of acid and was almost completely reduced after irradiation for 1 hr. in the presence of O . , j -VHC1. The o r t / compound ~~ is of particular interest since its reduction might normally be prevented C;, S
Iiammontl a t i d 11' 51
l l < , o r e , .I
C'hcw
.io(.. 81, 6 M 4
Vol X ( i
both because of an unreactive low triplet of structure related to that of form I1 and because of intramolecular hydrogen abstraction related to photoenolization.I f ) The observed photoreduction in acid medium indicates that the o-KHs- group. unlike -CH3, does not transfer hydrogen rapidly to the oxygen of the n-r* tripletcarbonyl group, and that the latter is relatively positive and its activity in hydrogen atom abstraction is electrophilic in character. Since this work is part of a study of photoreduction and its inhibition by sulfur compounds" in aqueous media, we have prepared the quaternary methochloride from DMABP, 4-benzoylphenyltrimethylammonium chloride, n1.p. 187-189' dec. ilnal. Calcd. for C160
H&OCl: C, 69.3; H , 6.53; N, 5.08; C1, 13.0. Found: C, 69.3; H , 6.42; N , 5 . 1 7 ; C1, 13.1. It showed Amax 340 mp (log E 2.20), presumably due to an n--K* transition, similar to that of benzophenone, and no high intensity long wave length absorption characteristic of the unquaternized and unprotonated aminoketones. This quaternary compound was also readily photoreduced in 1 : I 2-propanol-water, a t a rate about 0.35 that of benzophenone in this solvent pair, with estimated quantum yield ca. 0.25. (10) N . C . Y a n g and C. R i v a s , ibid., 85, 2213 (1961). (11) S. G. C o h e n , S. O r m a n , a n d D. 6. L a u f e r . ibid., 84, 3905 (1962)
DEPARTMEKT OF CHEMISTRY BRAXDEIS UNIVERSITY WALTHAM, MASSACHVSETTS02154
SAULG . COHEN M . NASIMSIDDIQUI
RECEIVED AUGUST21, 1964
Homolytic Aromatic Substitution. Phenylation of Phenanthrene
V.'
Sir : Electrophilic and homolytic aromatic substitutions of phenanthrene have been investigated to evaluate predictions, based on molecular orbital theory, that the reactivities of the five positions in this arene follow the sequence 9 > 1 > 4 > 3 > 2 . ? For example, nitration has been found to give the order 9 > l > 3 > 2 > 4,3 and homolytic phenylation with two sources (benzoyl peroxide and diazoaminobenzene) has been reported to correspond to the sequence 9 > 1 > 3 > 2 . 4 The latter studies failed to reveal 4-phenylphenanthrene among products from either source, and steric hindrance has been suggested by way of explanatian. We now report that the reactions of phenanthrene with three sources of phenyl radicals, including diazoaminobenzene, produce the 4-isomer in substantial amounts and that a t least one source yields 4-phenylphenanthrene in amounts t h a t are indicative of steric (1) P a p e r I V . S C. Ilickerrnan. M. Klein, a n d G B. V e r m o n t , J Orp C h ~ m, in press, paper 111: S C IXckerman, .4.5 4 . Felix, and I,. B. 1,evy. ibiii., 29, 26 (1i)fil) ( 2 ) 4 . Streitwieser. J r . , "Molecular Orbital T h e o r y for Organic Chemists," 1961. John R'iley and Sons, I n c . , New Y o r k , K.Y., (L3) AT J . S. I l e w a r , T. Mole, a n d I2 \I' T. Warford, J . C h r m . Sor., 3,511
I' 4) A . L. J . Beckwith and &I. J T h o m p s o n . ibtd , 7 3 (1961).
COMMUNICATIONS TO THE EDITOR
S o v . 20, 1964
acceleration rather than retardation. Other findings are relevant to problems concerning mechanisms of homolytic aromatic substitution. Authentic samples of the five phenylphenanthrenes were prepared by the photochemical method of Mallory and co-workers5 and were used to establish that g.1.c. would suffice for individual analysis of the 4-, 3-, and 2-isomers and for mixtures of 9- and l-phenylphenanthrenes. Meerwein, N-nitrosoacetanilide, and diazoaminobenzene phenylations of phenanthrene were studied and gave the data presented in Table I. The identity and homogeneity of that reaction product 'which has the same retention time as authentic 4phenylphenanthrene were confirmed by ultraviolet and infrared analysis of samples collected from the gas chromatograph. Infrared analysis also served to establish the presence of both the 9- and 1-isomers in the single peak exhibited by these compounds Examination of the data in Table I reveals t h a t all three sources of phenyl radicals produce 4-phenylphenanthrene in amounts up to a third of the total
5049
nitrosoacetanilide phenylations of phenanthrene, in aqueous acetone and a t 30°, do not yield identical mixtures of phenylphenanthrenes. Differences of this kind may arise through the intervention of side reactions which selectively alter the composition of i s ~ m e r s . ~ , ~Such O side reactions have not been detected in Meerwein phenylations of anthracene, although other radical sources invariably yield dimers, etc.8 The mechanism of Meerwein arylation of arenes has been discussed,8 and the absence of these types of side reactions has been attributed to the presence of an internal oxidizing agent, cupric chloride, which efficiently aromatizes intermediate radicals, e.g., I. On the other hand, the mechanism of N-nitrosoacetanilide phenylation is not well understood. Recent
I TABLEI HOMOLYTIC PHEXYLATION Temp, Sourceb
'c.
21
NAA NAA NAA NAA DAB
30 30 30 4 60 156
DAB
152
OF PHENANTHRENE4
--Isomers, ~o~vent9 I 4
+
Acetonee Acetone' Benzene Benzene Benzene Bromobenzene Sonel
Con- Recovversion,e e r y d
yo-3
2
%
7 4
5.8 8 2
4 3 0.34 3.5
53
31 24 21 15 2.5 23
4.1
28
56 56 56 61 R7
11 13 12 10
7 6
15
10
16
12
11 12
7~ 97 82 101
a ilnalyses by g.1.c. with a probable error of 4-5W per 100. M (Meerwein), S X A (S-nitrosoacetanilide), DAB (diazoaminobenzene). ' Conversion to phenylphenanthrenes based on phenanthrene. Recovery of phenanthrene as phenanthrene and phenylphenanthrenes. e Plus 20 vol. YG water and 2.8 vol. yG benzene. Essentially a duplicate of the reaction conditions employed by Beckwith and Thompson.?
'
In fact, Meerwein phenylation must correspond to a reactivity sequence 9 > 4 > 1 > 3 > 2 or 4 > 9 > 1 > 3 > 2 . Such discrepancy with theory is similar to that observed in studies of rates of deuteriodeprotonation and basicities of phenanthrene and phenanthrenelike hydrocarbons. Steric acceleration has been suggested to account for these results and we infer t h a t relief of nonbonded interaction is the most reasonable explanation for the enhanced reactivity of the 4,5positions of phenanthrene in homolytic phenylation. Although steric acceleration has not been observed previously in reactions of radicals with aromatic hydrocarbons, i t is not entirely unexpected in view of earlier work which has indicated t h a t addition of aryl radicals must be directed essentially perpendicular to the nodal plane.7s8 The data in Table I provide a direct comparison of radical sources and reveal t h a t Meerwein and N(5) F . B . M a l l o r y . C . S Wood. a n d J . T . G o r d o n , J . A m . Chem. S o c . , 86, 3094 (1964). (6) G . I)allinga, P . J. S m i t , a n d E . L . M a c k o r i n "Steric Effects in Conjugated S y s t e m s , " G . W. G r a y , E d . , B u t t e r w o r t h s a n d C o . , L o n d o n , 1958. C h a p t e r 13, see also ref. 2 , p. 344. ( 7 ) H. W e i n g a r t e n , J . Ovp. C h e m . ,!46, 730 (1961). ( 8 ) S C D i c k e r m a n a n d G . B . V e r m o n t . J . A m Chem. Sac., 84, 4150 11962)
developments may be summarized as follows: (a) the cage mechanism has been eliminated from consideration, l 1 s I 2 (b) evidence of induced decomposition has been given," and (c) an analogy has been drawn to the diazoanhydride mechanism of the Gornberg-Bachmann r e a ~ t i 0 n . I ~A noncage mechanism for N-nitrosoacetanilide phenylation, accompanied by side reactions, offers an explanation for the observed differences in isomer distribution of phenylphenanthrenes between this and Meerwein phenylation (Table I ) . Isolation of 1,4dihydrobiphenyl from decompositions of N-nitrosoacetanilide in benzene a t GOo confirms the occurrence of such reactions, a t least a t this temperature, and provides additional evidence against the cage mechanism,l 4 The effect of temperature on the reactions of Nnitrosoacetanilide with phenanthrene in benzene was investigated and it was found that an increase in temperature favored formation of 4-phenylphenanthrene largely a t the expense of the 3 - and 2-isomers (Table I ) . These results are unusual in that an increase in reaction temperature is normally accompanied by an increase in those isomers whose formation requires the largest energies of activation. Present speculations concerning the origin of the temperature effect include the more favorable entropy of activation associated with formation of radical I relative to the isomeric intermediates and/or to changes in the selectivity of the side reactions. (9) T h e s e side reactions involve dimerization a n d disproparti,,natiljn oi i n t e r m e d i a t e radicals a n d were discovered in peroxide a r y l a t i ~ r n of s benzene, D. F . D e T a r a n d K A. Long, i b z d . . 80, 17-12 (1958) (IO) W h e t h e r s u c h side reactions selectively a l t e r t h e " t r u e " cumposition o f isomeric biaryls h a s been a subject of considerable c o n t r o v e r s y . T h e original presumption of general selectivity, which was baqed on a j w i o ~ z reasoning, h a s been shown t o be incorrect thrnuFh a comprehensive s t u d y of peroxide a r y l a t i o n s of derivatives of benzene: I 68 kcal., i e . , ETI 3 68 kcal. Moreover, I can hr f o r m e d jrom TI], where ETIr< 50 kcal. since T I 1 must 50 kcal. = be less energetic than singlet I1 (Es,, 570 m p ) . j On prolonged unsensitized irradiation (Xi5 mp) of benzene or ethanol solutions of I there were formed the (15) T h i s investigation was s u p p o r t e d b y a Public H e a l t h Predoctoral Fellowship (5F1 - G R I - 2 2 , 577-02) f r o m t h e S a t i o n a l I n s t i t u t e of General dimer V,la 3,4-diphenylisocoumarin (VI) and, in Medica! Sciences. ethanol, the stereoisomeric ethers VII. The isocouDEPARTMENT OF CHEMISTRY S. CARLTON DICKERMAN marin VI was found to arise exclusively from 11. NEW YORKUNIVERSITY ISAAK ZIMMERMAS'' Thus on irradiation of a thermally equilibrated mixture N E W YORK,XELVYORK 10153
12 kcal.), such a nonvertical energy transfer process to give the lower energy T I I ~might be expected. The absence of this abnormal
-
