[COKTRIBUTION FROM
TIIE
DEPARTMENT O F CHEMISTRY, THEUNIVERSITY
OF
TEXAS]
Preparations and Reactions with Amines of Some 1,2-Dibenzoylalkened PHILIP S. BAILEY, SHEAFFERS S. BATH, WILLIAM F. THOMSEN, HERBERT H. NELSON, AND ELIAS E. KAWAS
Received Xovember 7, 1955 The reactions of several dibenzoylalkenes with amines are compared with the reaction of 1,2-dibenzoylpropene with amines to give l-amino-2,3-dibenzoylpropanes. The Wolff-Kishner reduction of 3-benzoyl-2,5-diphenylfuran gave 3,5diphenyl-4-(P-phenylethyl)pyrazole as a by-product. CH3 I n earlier papers the reactions of 1,2-dibenzoylpropene, 1,2-dibenzoyl-l-butene, and 2,3-dibenH--C---C-A=CH, zoyl-1-(4-morpholiny1)-2-butene4with morpholine /I /I CsH5-C C-C& to give 1,2-dibenzoyl-3-(4-morpholinyl)propaneand butane and 2,3-dibenzoyl-l,4-di(4-morpholinyl)butane, respectively, and of 1,2-dibenzoylpropene I11 (a) R = H; R’ = isopropyl IV (b) R = H; R’ = benzyl with hydrogen halides to give 1,2-dibenzoyl-3(c) R = R ‘ CHE halopropanes or 3-halomethy1-2,5-diphenylfuran~~~~(d) R = R’ = H (e) R = R’ = CHzBr have been reported. I n the present paper the syntheses and reactions with amines of three similar 0 compounds, cis-l,2-dibenzoyl-3-methyl-l-butene I1 (VIa), cis-2,3-dibenzoyl-2-butene (VIc), and cis- C6Hs-C-CH2-C-C-C6Hj /I I/ 1,Z-dibenzoyl-3-phenylpropene (VIb) are described. 0 R’-C-R The synthesis of cis-l,2-dibenzoyl-3-niethyl-l- V (a) R = R’ = CH3 (b) R = H ; R’ = phenyl butene (VIa) started with 3-acetyl-2,5-diphenylfuran (Ia) and involved as the first step reaction 0 C&,--C-C--K II with methylmagnesium iodide to give 3-(2-hydroxy-2-propyl)-2,5-diphenylfuran (IIa) which /I C6H5-C-C-R’ spontaneously dehydrated to 3-isopropenyl-2,5/I diphenylfuran (IV). Catalytic reduction of IV 0 gave 3-isopropyl-2,5-diphenylfuran (IIIa) which VI (a) R = H, R’ = isopropyl underwent the characteristic nitric acid oxidative (b) R = H, R‘ = benzyl (c) R = R’ = CH3 cleavage7 to the desired dibenzoylisopropylethylene (d) R = CH3, R’ = CH~SCIHRO (VIa). Attempts to produce I I I a by a direct The synthesis of cis-1,2-dibenzoyl-3-phenylproFriedel-Crafts reaction between 2,5-diphenylfuran pene (VIb) involved the Wolff-Kishner reduction of (IIId) and isopropyl chloride failed. 3-benzoyl-2,5-diphenylfuran(Ib) to 3-benzyl-2,50 G diphenylfuran (IIIb) followed by nitric acid oxidaII H-C--C-C!-R H-C-C(CHa)2 tion of I I I b to VIb. The cis configuration is assigned I/ I/ I1 to both VIa and VIb by analogy to the nitric acid CsH6-C C-CsHa C81-Is-b C-C6H5 oxidations of similar furans.’ The reductive cycliza\/ \/ 0 0 tion of each back to the corresponding furans I1 (a) G = OH I (a) R = CH3 (IIIa and b) is evidence that no unexpected struc(b) G = C1 (b) R = phenyl tural changes occurred during the oxidations. cis-2,3-Dibenzoyl-2-butene (VIc) was obtained (1) Taken in part from the M.A. Thesis of W. F. T. as shown previously,* by the nitric acid oxidation (January 1950) and H. H. N. (January 1951) and from a of 3,4-dimethyl-2,5-diphenylfuran (IIIc). The latter portion of the Ph.D. dissertation of S. S. B. (soon to be was synthesized by a new method, however, which presented). This work was supported in part by grants-inaid from the Research Corporation, New York, N. Y., and involved the dibromomethylation of 2,s-diphenylfuran (IIId to IIIe) followed by reductive dehaloThe University of Texas Research Institute. (2) Lutz and Bailey, J . Am. Chem. Soc., 67, 2229 (1945). geiiation. (3) Bailey and Hakki, J . Am. Chem. SOC.,71, 2886 The dibenzoylisopropylethylene (VIa) failed to (1949). undergo additi‘on with dimethylamine, morpholine, (4) Bailey and Nowlin, J . Am. Chem. Soc., 71, 732 or piperidine. Instead, whether in ether solution (1949). (5) Bailey, Nowlin, Pomerantz, Waggoner, and Kawas, at room temperature or with the amine direct, even J . Am. Chem. Soc., 73, 5560 (1951). a t goo, it isomerized to 1,2-dibenzoyl-3-methyl-2-
‘d
b
(6) Bailey and Pomerantz, J . Am. Chem. Soc., 75, 281 (1953). (7) Lutr and Wilder, J. Am. Chem. SOC.,56, 978 (1934).
297
(8) Lutz and Kibler, J . Am. Chem. Soc., 62, 1520 (1940).
291
BAILEY, BATH, THOMSEN, NELSON, AND KARVAS
VOL.
21
0 butene (Va). Under more drastic conditions it gave 1 2 /I a tar. The structure of Va was proven by reductive CsH6-C-CH=C--C-C8Hj cyclization to IIIa and by ozonolysis, whereby II I 0 CH, +CH, acetone was produced. In contrast, VIa gave no a 4 acetone upon ozonolysis. VI1 In an earlier paper3 it was suggested that the R’ CHRM cause for the difficulty with which 1,2-dibenzoyl-l‘ I butene undern-ent the amine addition in compariC~HS-C-CH-CH-C-C~H~ /I I/ son to 1.2-dibenzoylpropene might be the inductive 0 0 effect of the additional methyl group (carbon 4 in VI11 M = morpholinyl structure C‘II) which could conceivably decrease (a) R’ = H ; R = phenyl (b) R = H, R‘ = CHI the ease with n-hich the hydrogens on carbon 3 (of VU) could undergo the 1,3-shift necessary to G bring about the reaction.?^^ However the results 0 H-C--C-&HCsHa with the isopropyl compound (T’Ia) show that this /I /I Il is not so, since the 1,3-hydrogen shift occurs readily CcHb-CH-C-C-CaHs CeHs-C C-CsHs but addition to the product (T‘a) does not occur. CHZ /I Instead, it appears much more likely that steric IX XI (a) G = Br hindrance in the product of the 1,3 hydrogen shift (b) G = morpholinyl is the reason why homologs and structurally similar (c) G = OH derivatives of 1,2-dibenzoylpropene undergo the C~HS-C-CH?-C-C-C~HS novel amine addition less readily than does 1,2/I I/ 0 0 dibenzoylpropene. Indeed, a Fisher-Hirschfelder x model of T a shows that it would be extremely difficult for addition with an amine as large as PLs with 2,3-dibenzoyl-l-(4-morpholinyl)-2-bumorpholine or piperidine to occur. tene (VId),4 it was necessary to heat a morpholine The benzyldibenzoylethylene (T’Ib) also reacted solution of 2,3-dibenzoyl-2-butene (VIc) for 21 in accordance with this picture. In a dilute solution hours in order to bring about reaction. The product of morpholine in ether, VIb simply isomerized to was the known 2,3-dibenzoyl-l-(4-morpholinyl)2,3-dibenzoyl-l-phenylpropene(Trb). When treated butane (VIIIb).4 Since not even isomerization with a concentrated morpholine-ether solution or occurred with VIc under milder conditions it is with morpholine alone, however, addition occurred reasonable t o suppose that in the case of VI- and to give 1 2-ilibenzoyl-3- (4-morpholinyl)-3-phenylT’Id not only is steric hindrance a factor, but also propane (T’IIIa). It was under the same conditions the extent to which the product of the 1,a-shift of that 1,2-dibenzogl-I-butene reacted with morphohydrogen occurs in the equilibrium mixture proline togive 1.2-dibenzoyl-3-(l;-morpholinyI)-butane. duced by treating the original olefin with morphoFischer-Hirsvhfelder models of 2,3-dibenzoyl-1line. For example, IX (the product of the 1,3-shift phenylpropene (T%) and of 1,2-dibenzoyl-2-butene of hydrogen with VIc) should be much less stable (the product of the 1,3-hydrogen shift with 1,2than VIc, and therefore present t o a much lesser dibenzoyl-1-butene) show that whereas these extent in the equilibrium mixture, because it not molecules :Ire sterically hindered, they are not only possesses a shorter conjugated system than as sterically hindered as T‘a and morpholine addition T’Ic, but also is stabilized through the hyperconshould occur. jugation of only one hydrogen atom whereas T‘Ic The structural proof of Vb consisted of its ozonolis stabilized through the hyperconjugation of six ysis to the enol of l14-diphenyl-1,2,4-butanetrione hydrogen atoms. With the other systems discussed. (XI9 which is a known compound. The odor of hyperconjugation either makes n o difference, or benzaldehyde was also apparent, but no attempt else stabilizes the product of the 1,3-shift to a was made to isolate the material. greater extent than the original olefin, thus making The structure of VIIIa was proven by cyclizathe equilibrium more favorable for the prodwt of tion to XIh which i n turn was synthesized from the 1,3-shift than in the case of TIC. 3-benzyl-2 5-diphcnylfuran (IIIb) by bromination During the work leading to the preparation of with S-broiiiosucciiiimid~ to XIa followed by IIIa and I I I b the corresponding his-furans (XIIa aminolysis of X I a to XIb. The bromo compound and b) were obtained. The his-isopropylfuran (XIa) had lire\-iously been made another way by (XIIa) was obtained, instead of IIIa, by reduction Iiohler and aJones.lo of IT’ with hydriodic acid or stannous chloride and
1“’
b
‘d
ii
(9) Lutz, Wilder, and Parrish, J . Ani. CAern. SOC.,5 6 , 1980 (1931); J a t z and St,uart. J . Am. Chem. Soc., 5 8 , 1885 (1936); L u t z and King, J . Org. Chem., 17, 1519 (1952). (10) Iiohler and Jones, J . Am. Chem. SOC.,41, 1249 (1919).
hydrochloric acid. Stannous chloride-hydrochloric acid reductions of T’Ia and T’a likewise gave XIIa. It was also produced by treatment of IV, T’a or T’Ia with a hydrochloric acid-acetic acid mixture followed by coupling with copper bronze. Evi-
1956
MARCH
RE.4CTIONS OF AMINES WITH
dently in these last reactions, at least, halogen derivatives are intermediates, I I b from IT', and either or both I I b and the product of a conjugate addition of hydrogen chloride to Va in the cases of T'a and Y1a.l1 A Fisher-Hirschfelder model of Va shows that hydrogen chloride addition could much more easily occur than the morpholine addition discussed earlier. Similarly. XI1t) was produced by Ponndorf reduction" of the benzoylfiiraii (Ib) to the corresponding alcohol (XIc) followed by stannous rhloride-hydrochloric avid reduction of XIc. It jvas also obtained from the coupling of the bromobenzylfuran ( M a ) hy means of copper bronze. R
R
,
I
II-C--(~---C--C--C--C--€i ll 1 ' I c6IIj-c C R' R'
I'
\pC J I ~
299
droxide solution. Thege reactions are entirely typical only of the pyrazole (XIII). The material was shown to be the pyrazole by an unambiguous synthesis, the principal step of which was the conversion of the corresponding 1,3-diketone (XT'I) to the pyrazole by the general method of treating with hydrazine. -4 possible course by whirh the benzoylfuran (Ib) was converted to the pyrazole (XIII) involves first the formation of the hydrazone (XYII). The hydrazone could then be converted to the benzylfuran (IIIb) on the one hand and on the other undergo an internal simultaneous aminolytic cleavage of the furan ring and formation of the pyrazole ring to forin XT'III which would ketonize and undergo Wolff -Kishner reduction to XI11 H-C-C-
I1
d
1,%DIBENZOYLALKENES
C-CsHj
C-CJIj
A/
GH~ CH?
XI1 (a) R = R' = (b) R' =: H : R = phenyl
SI11
H
SI-I11
CH2CsHs
I (\ H--G C-CEH5 I/ CbHS--C k
II-C-C-CH2C6H5 C6Hj-C
11
C
'I
c'-CsH,
'
\/
s I
\/
n.
XHz XIV
An alternative route to XT'III (or its keto form) involves first the aminolytic cleavage of the furan ring of I b by hydrazine to give XIX, followed by the intramolecular interaction of the substituted hydrazine group and the carbonyl group of X I X
XV 0
0
iI
11
C~H~,C-CH-C--C~HS
I
CHZCH~C~H~ XVI
The \Yolff-I
