11. C. CASERIO, H. E. SINMONS. JR., A. E. JOHNSON
3 102
AKLI
3. 11. IZUUERTS
Vol. s2
hydroxide a t 12 mm. Five colorless fractions, boiling of platinum oxide at room temp. After 25 hours t h e tlieolii-68', were obtained; the residue from the distillation retical quantity of hydrogen (12.8 Ib.) was absorbed consisted of a drop or so of a dark brown liquid. The T Z ~ O D On working up and distilling at 13 mm., there was obtained values for these fractions were 1.5256, 1.5261, 1.5262, 1.5262 2.89 g. of colorless amine, b.p. 64",~ Z O D 1.4234, ( Y P ~ D and 1.5265. Fractions 2 , 3 and 4 were combined and their 1.95" ( I 1 dm.). rotation determined (polarimeter tube filled in the ordinary A n d . Calcd. for C s H l r S : C, 74.32; H , 14.85; S , 10.8.1 \vva>-. i.~.,no effort made vigorously to exclude COn or Found: C,i3.99; H, 15.05; 9, 11.01, 10.96. IIzO): ( ~ 2 % i . 0 i " . Reduction of 2-Nitrooctane.--X solution of 8.5 g . of The benzamide, thrice recr) stallized from c) clohexanc, ,f -)-2-nitrooctane, b.p. 59" (1 mm.), %?OD 1.4277, C Y ~ ~ Dhad m.p. 99.5-100.5". -8.25' ( I 1 dm.), in 35 ml. of glacial acetic acid was shaken IND wit11 hydrogen ( 3 atms. pressure) in the presence of 0.45 g. LAFAYETTE,
+
2146 FROX TEIE GATESA S D CRELLIN LABORATORIES OF CHEMISTRY, TECHNOLOGY, AND THE DEPARTMENT O F CHEMISTRY, MASSACHCSETTS INSTITUTE
[COXTRIBUTION NO
CAIJFORNIA INSTII'UTE 01.' O F TECHSOLOGC]
Small-Ring Compounds. XXVI. Nucleophilic Displacement Reactions of Some Halogen-substituted Phenylcyclobutenones' BY ~ I X R J UC. R ICASERIO, E HOWARD E. SIMRZO~\'S, JR., A. EARLJOHNSON
AND JOHN
D. ROBERTS
RECEIVED NOVEMBER 6 , 1959 Displacement reactions between alkali hxlides and several cyclic allylic-type halides, namely, 2,2-dihalo-3-phenylcycli,butenones, have been found t o proceed with rearrangement to give 2,4-dihalo-3-phenylcyclobutenones.T h e 2,2-dihalocompounds also were found to rearrange t o the corresponding allylic isomers on heating with triethylamine. The allylic rearranged in reactions with halogens to form 2,2iodides, 2-fluoro-4-iodo- and 2-chloro-4-ic~do-3-plienyl-2-cyclobutenones, dihalo-3-phenylcyclobutenones Evidence is presented for the stereospecific nature of these reactions.
The interest which attends bimolecular nucleophilic displacements of allylic halides owing to the possibility of substitution with and/or without rearrangement,2 s ~ ' ) 'and s ~ ? ,respectively, prompted t h e investigation we now describe of the reactions of some halogen-substituted phenylcyclobutenones with various nucleophilic reagents. Displacements with Halide Ions.-A series of 2,2-dihalo-3-phenylcyclobutenones( I i has been found t o react with alkali halides in acetone solution t o give exclusively the product of substitution with rearrangement. Thus 2.2-dichloro-, 2-chloro-2 bromo- and 2-fluoro-2-bromo-:-l-phenylcyclobutenones (Ia, I b and IC,respectively) rearranged to their 2,4-isomers IIa, I I b and IIc on treatment with the appropriate lithium halide.
a-substituted open-chain allylic compounds. The ease of rearrangement in the series of gem-dihalides here investigated appears to decrease with decreasing size and nucleophilicity of the halogen. IVhereas bromide ion caused the 2-chloro-2-bromoketone I b to rearrange rapidly a t room temperature, the 2fluoro-2-chloroketone Id was recovered unchanged after prolonged heating with lithium chloride in acetone. However, Id appears to be attacked by chloride ion with comparative ease since optically active Id, [a]2 5 4~ 12.0°, racemized with a half-life of approximately 19 hours a t room temperature in acetone solution saturated with lithium chloride. I t seems likely t h a t racemization results from chloride exchange by an S s 2 mechanism. It was reported in earlier work* t h a t triethylamine caused the rearrangement of 2,2-dichloro-3phenylcyclobutenone (1,). Under similar conditions, ketones I b and IC have also been found to rearrange to give I I b and IIc, respectively, but again the fluorochloroketone Id did not rearrange. The resistance of Id to rearrangement by chloride IIa, XI = S n = C1 Ia, XI = Xz = C1 was not matched by iodide ion. Thus, I d reacted b, XI = C1, XZ = Br b, XI = C1, Xz = Br c, XI = F, X2 = Br c, XI = F, Xn = Br readily with sodium iodide in acetone to give the d, . d , Xi = F, Xip = '21 product of rearrangement, 2-fluoro-4-iodo-3-phenvlThe susceptibility of these allylic gem-dihalides 2-cyclobutenone (IIf'i. This result poses the yuesto rearrange by what is very probably an S N ~ ' tion as to whether IIf is formed directly by SS?' mechanism may be attributed largely to the pres- attack of iodide ion or by SNL' attack to form the ence of two a-halogen substituents on the allylic gem-fluoroiodoketone as an intermediate followed by its rapid rearrangement. \I.'hile the greater bulk I l l carbon system -Cy= C, -C,--. This has the of iodide relative to chloride ion would render attack by an S s 2 ' mechanism more favorable, t h e effect of promoting s N ~ /attack in competition with two-step mechanism is not unreasonable and no S N 2 attack by a combination of polar and steric clear choice can now bc ninde between the two alfactors, as has been amply demonstrated by de la ternatives. Mare and co-workers3 for substitution reactions of ( 3 ) P. B. 11 d e la Mare a n d C A . Vernon, J . C h e w S O L , 3 3 2 5 , ,
,
(I) Supported in p a r t b y t h e National Science Foundation. (2) F o r leading references on t h e abnormal mode of substitution ~ S u ' l ' ) +cc' R . H. DeWolfe and W . G . Young, Chew. R ~ a s . .56, 753 [ I !).>I;),
,2331, 3628 (1952), 3,555 (1953); P. B. D . d e la LIare, E . D. Hughes P C. hIerriman, I,. Pichat and C . A . Vernon, ibid..2563 (1958). f4) J n. Knherts. 0 B. Kline a n d H. E Simmons, J r . , T H I S J O U R N A L , 75, 47&j (1953).
NUCLEOPHILIC DISPLACEMENT IN
June 20, 1960
= XB = c1 b, SI= C1, X2 = Br tl, S i = F, Xp = C1
Ia, S I
IIe,X1 e, XI
f,
= = XI =
PHENYLCYCLOBUTENONES
3 103
C1 C1
F
IIa, XI = Xz = C1 b, XI = C1, X2 = Br
The 2,2-dichloroketone I a and its 2,4-isomer I I a both reacted with sodium iodide in acetone to give the same product, 2-chloro-4-iodo-3-phenyl-2-cyclobutenone (IIe). Similarly, iodide displaced bro100 200 300 400 500 minutes, mide from the isomeric pair I b and I I b to give IIe. Fig. 1.-Polarimetric rate of rearrangement of 2-chloroThus, the 2,4-dihaloketones react with iodide ion without apparent rearrangement. In addition t o 2-bromo-3-phenylcyclobutenone with lithium bromide in the obvious s N 2 formulation of the latter reaction, acetone at 25'. it is possible that I I a and I I b react by two con- min.-l (Fig. 1). The racemic product was identiHowever, if such is the case, fied as 2-chloro-4-bromo-3-phenyl-2-cyclobutenone secutive S N ~steps. ' then rearrangement of the 2,2-dihaloketones must (IIb). Further, the second-order specific decay necessarily proceed rapidly relative to the 2,4- rate of positive rotation, calculated from the expresdihaloketones since the isolable products are the e pure 2,4-isomers. Experiment supports this stipv- sion k, = k2[Br], was found to be the same (within lation for the iodide-displacement reactions with experimental error) as that determined independthe 2,2- and the 2,4-dichloroketones Ia and I I a ently for the rate of racemization of I I b by lithium since a second-order rate law was observed for both bromide in acetone (Table 11). Clearly, the transketones (Table I) but the 2,2-isomer reacted some formation Ib to I I b is stereospecific, although the 18.6 times faster than the 2,4-isomer. However, results do not describe the degree of stereospecithe behavior of the fluorochloroketone Id toward ficity nor the stereochemical relationship between chloride ion is exceptional in t h a t the position of the entering and leaving groups. However, Young, equilibrium, if equilibrium is established, must lie Webb and Goering6have suggested that the attackwell on the side of the 2,2-isomer. ing group enters cis to the leaving group in SN2' displacements. This has been elegantly demonTABLE I strated by Stork and White7 who observed only R A T E O F REACTIOX B E T W E E S S O D I U M IODIDE A S D "2- A S D CiS-sN2' attack of piperidine and malonic ester an~,~.-DICHLORO-3-PHENYLCYCLOBCTEliOXE Ili ACETOXE ion on trans-6-alkyl-2-cyclohexen-l-yl 2,6-dichloroSOLCTION AT 39.7" benzoates since the product of trans configuration Ketone I a , Ketone IIa, h7aI, k2 x 108 was formed exclusively. By analogy, therefore, .1I M .M I. mole-1 min.-' CiS-sN2' attack of bromide on I b is very probable. 0 01 IJ 1 89.6 zk 0 . 9 Since the product I I b is racemized by bromide ion, 0.04 0 1 4 . 8 0 3z 0 . 0 3 i t seemed likely that bromide exchange takes place I n order to investigate the stereochemical aspects by an sN2 mechanism. The intriguing alternative of the observed Sx2' displacements, i t was necessary of bromide exchange by two consecutive s N 2 ' to prepare optically active forms of the resolvable steps is unlikely since this should not result in race' is indeed 100yc stereospeketones. Partial resolutions of Id, IIb, IIe and IIf mization if S N ~attack were achieved by asymmetric destruction with the cific. aid of b r ~ c i n e . ~Compound I b was prepared parTABLE I1 tially active by the action of bromine on active IIe, O F 2-CHLORO-4-BROMO-3-PHESYL-"as will be described later. m'ith the exception of RATEO F RACEMIZATIOS CYCLOBCTENOSE B Y BROMIDEIONI N ACETOXEAT 25' Id, all of these substances racemized on heating, a KeKetone, LiBr. k? phenomenon which was studied in some detail in the tone .M JI X 103 10s ki, m i n . - l 1. mole-l min -1 case of optically active 2,4-dichloro-3-phenylcyclo- I I b 21 8 0.406 6 23 i 0 11 0 886 f 0 005 butenone (IIa) .5 IIb ,406 10.9 3 06 i . 1 7 281 i ,013 The reaction of active Id with sodium iodide in Ib ,348 9 01 2 . 9 0 i 03 ,293 i ,003 refluxing acetone gave inactive product IIf. However, IIf racemized under these conditions; hence A\ 287 =k 008 no conclusions could be drawn concerning the sterRearrangement of I b with [ a ]D -2.0" induced by eochemistry of the reaction of Id with iodide ion. The rearrangement of partially active 2-chloro-2- triethylamine in acetone resulted in appreciable bromo-3-phenylcyclobutenone (Ib) was more con- discoloration owing to partial destruction of the clusive. Thus, the optical rotation of active I b in ketone by the base. Solutions were consequently an acetone solution 0.001 Af in lithium bromide a t opaque to the sodium light of the polarimeter. However, after 1 5 hours a t room temperature, the 25" increased from [ a ]2 5 ~ 1.92' a t zero time to a ( 6 ) W G Young I n n ' e h b and H L Goerlng, z b t d , maximum of [ a ]2 5 +5.34' ~ before decaying to zero 79, 1076 (1951) by a pseudo first-order process, k , = 2.90 X (7) G Stork arid U' K White t b i d 7 5 , 4119 (1953), 7 8 , 160 t ( 5 ) E . F. Jenny xnd J I> I l o b c r t i ,
'CHIS J O U R N A L ,
7 8 , 2005 (1Y56).
(1956)
product I I b was isolated and shown to have a rotation [.ID +2.04". Clearly some racemization occurred. It was suggested previously4s6that rearrangement by triethylamine probably involves a reversible Menschutkin reaction. This may now be elaborated as the initial formation of a quaternary ammonium bromide followed by cis-Sx2' attack of bromide ion on excess ketone. The product is subsequently raceniized by bromide ion. Reactions with Halogens.-The allylic iodides I Ie and IIf underwent an interesting reaction with halogens to generate gem-dihaloketones. Iodine monochloride and I I e or IIf in carbon tetrachloride liberated iodine and formed ketones Ia or Id, respectively. Similarly, iodine monobromide afforded the 2fluoro-2-bromoketone ICfrom IIf, and the 2-chloro2-bronioketone I b from IIe. With excess iodine HI
Ix 7
6 5 '+ XI
Ile, XI = C1 f, Xi = F
(n.m.r.). The ultraviolet spectra of the gem-tlihaloketones showed absorption maxima a t lowcr wave lengths relative to the 2,4analogs (Table 111) TABLE
T
7
~
Compound
Ia Ib IC Id IIa
111
, . h ~w : w~r I o \~ S~P E ~C ~ R OF ~ ACYCLOBUTF\O\EI' ~ ~ * ~ ~
SI
c1
s! c1 Ar Br C1 H
2, x2 s i
A,,,*,
111~
e,,> - 73 !I IC I? I3r I1 mechanism which we accordingly illustrate below i n I ti F C1 11 73 2 :I the case of IIe. i
I
$0
IIa Cl H C1 -11 .. 1111 Br 1-I C1 -:filler, ; I . , THIS J O U R N A L , 7 9 , 4170
i1g.57) (3) T h e all>-lir iodides I I c and I i f dcc,)miioseil s I o \ v I v in s o l u t i o n Iihrratr i o d i n e N'hrn air a n d 1i::lit wcrc excliidccl, soI%i!i.iii- srmaiiieil \:al~le f i i r l i i i i ~ 1 ~ r i o c I scd tirnc
I
