Oct.

20, 1952 : RLDOXIMES

RECKMANNREARRANGEMENTS 5 15 1
ethyl ketone gave a yellow product, m.p. 264-266'; analysis Hydrolysis and Cyclization of VI.-A suspension of 2.0 g.
(calcd. for Cl~H1106:C, 62.06; H , 4.86. Found: C!, 62.10; of VI in 30 ml. of 5% sodium hydroxide solution was heated
H , 5.00) indicated that this material of unknown structure (95-100') for 20 minutes; the resulting clear solution waq
had the same composition as the usual product. diluted and acidified strongly with concd. hydrochloric acid.
Ozonolysis of VI.-A solution of 0.30 g. of VI in 150 ml. After standing a t room temperature for one hour, a small
of chloroform was treated with a stream of ozone for 30 amount of gum was removed, and after further dilution to a
minutes. The solution was stirred vigorously with 75 ml. total volume of 100 ml. the solution was allowed to stand for
of 5% sodium hydroxide solution containing 24 ml. of per- five days. The crystalline product was separated and re-
hydro1 (30%) for 15 minutes, and after separation of the crystallized from aqueous ethanol to yield nearly colorless
chloroform the organic solution was washed with 5% sodium plates, m.p. 248-247" (with shrinking and discoloration), of
hydroxide solution, with dilute acetic acid, with 5y0sodium material whose analysis and ultraviolet spectrum (Fig. 3)
bicarbonate solution, and with water. After drying and indiEated that cyclization had occurred; the product was
evaporation of the chloroform there was obtained 0.1 g. of evidently l-methyl-6,7-dimethoxy-3-naphthoic acid.
veratraldehyde, identified through the 2,4-dinitrophenylhy- Anal. Calcd. for Cl4H1404: C , 68.28; H , 5.73; neut.
drazone (m.p. 260-262', undepressed on mixture with an equiv., 246.3. Found: C, 68.71; H , 5.87; neut. equiv., 246.
authentic sample).
Reduction of VI.-When VI was subjected to catalytic Acylation Experiments.-In attempts to define the scope
reduction in acetic acid a t 70' with a 5% palladium-carbon of the acylation reaction leading to V I , reactions were
catalyst, sufficient hydrogen t o reduce the olefinic linkage carried out with the corresponding diester (obtained by com-
was absorbed. It was not possible to isolate crystalline plete esterification of I ) rather than the diacid or anhydride,
material after the reduction; cyclization studies on the re- with benzylsuccinic anhydride, and with benzoic anhydride
duced material were inconclusive: and, while a reaction in place of acetic anhydride in the usual acylation. No
apparently occurred with 2,4-dinitrophenylhydrazine, the products corresponding to V I could be found.
product could not be purified. BETHESDA,
MARYLAND

[CONTRIBUTION FROM THE NATIOXAL NATIONAL

HEARTINSTITUTE, U. S. PUBLIC
INSTITUTESOP HEALTH, SERVICE
HEALTH
FEDERALSECURITY AGENCY]

Beckmann Rearrangements. Aldoximes

BY E. C. HORNING
AND V. L. STROMBERG
RECEIVED
MARCH1, 1952

The Beckmann rearrangement has been extended to aldoximes through use of polyphosphoric acid as a reagentfor the
rearrangement. For the first time this permits a direct study of the configuration of an aldoxime through examindim of the
products of rearrangement. The present assignment of syn and anti structures to the two known benzaldoximeu'has been
confirmed.

A few limited observations may be found in the provided that the configuration of the oxime is not
on reactions involving the con- affected during the reaction, or that both oximes
older l i t e r a t ~ r e l - ~
version of an aldoxime to an amide v i a a Beckmann can be isolated and are found to give different
rearrangement. In general, however, the usual products. In the case of ketoximes, it is some-
conditions of Beckmann rearrangements result in times possible t o isolate two stereochemically
partial or complete dehydration of aldoximes to different oximes whose structure can be determined
give the corresponding nitriles, and as a result in this way. Some oximes, such as acetophenone
special or indirect methods have been employed oxime, are known to exist in only one form, and in
where rearrangement is desired. N-Alkyl ethers4s5 these cases only one rearrangement product has
and acetyl derivatives6 have been used in such been observed. Other ketoximes, including methyl
studies. n-propyl ketoxime, are apparently homogeneous,
We have recently found that polyphosphoric acid but give two amides on rearrangement, indicating
is an excellent agent for effecting Beckmann re- that both forms of the oxime are present during
arrangements of ketoximes,' and the study has now rearrangement.
been extended to aldoximes. As an example of an The assignment of configuration to aldoximes has
aliphatic aldoxime, we investigated the rearrange- of necessity been based upon indirect evidence.
ment of n-heptaldoxime. Under the conditions Hantzsch8 first recognized the possibility of isom-
described in the experimental section, a 92% yield erism of aldoximes, and i t has since become
of n-heptamide was obtained. This result is com- standard practice t o treat the acetyl6 or benzoylg
parable to the yields obtained with k e t o x i m e ~ , ~derivatives of aldoximes with alkali in order t o
and i t indicates that the Beckmann rearrangement determine the structure. The acylated anti-
is indeed a general reaction of both ketoximes and oxime yields a nitrile, while the syn compound does
aldoximes. not react or regenerates the original oxime. The
Since the Beckmann rearrangement is stereo- rearrangement of an aldoxime with polyphosphoric
specific, the structure of the starting oxime may be acid provides a means of studying aldoxime struc-
determined from the structure of the product, tures in a direct way, and we have therefore ex-
(1) W. R. Dunstan and T. S. Dymond, J . C h e w . Soc.. 66, 206 amined the best-known case of aldoxime isomerism,
(1894). that of benzaldoxime, lo with this method. When
(2) W.Comstock, Am. Chcm. J . , 19, 485 (1897). the syn-oxime (liquid) was treated with poly-
(3) A. Hantzsch and A. Lucas, Bey., 28, 744 (1895).
(4) E. Beckmann, ibid., 26, 2272 (1893). (8) A. Hantzsch. Rer., 24, 21 (1891).
(5) E. Beckmann, ibid., 37,4136 (1904). (9) G.Vermilion and C.R. Hauser, THISJOURNAL, 62, 2939 (1940)
(6) C. R. Hauser and E. Jordan, THIS JOURNAL, 67, 2450 (1935). C. R. Hauser and G . Vermilion, i b i d . , 63, 1224 (1941).
(7) E.C. Horning and V. L. Stromberg, i b i d . , 74, 2680 (1952). (10)E. Beckmann, Be?.. 22, 1531 (1889).
phosphoric acid a t 90*, an RC; yield of formanilide the stable hydrochloride of benzaldoxime is that
dIld a lOT{ vield of benLamide resulted; 13;' of the derived from the anfi forni, this salt should yield
osime was recovered unchanged. .It 1:iO t h e bc.nzaniide 011 rearrangement. IVhen the anti-
osiiiie hydrochloride was treated with polyphos-
ayn C611 -c 11 phoric acid a t 130' for five minutes, an SO(;/, .yield
of benzamide was obtained. The configuration of
the oxime in the anti form was evidently retained
in the salt. 'This rnodification of the usual rear-
mngenient may he useful where stable salts can be
I)rel)ared.
Acknowledgment.-b:e are indebted to hlrs. Iris
Simwrs for t h e infrared spectrophotoinetric data.
Experimental
yields of formanilide and benzamide were 9 and Rearrangement of n-Heptaldoxime.-A 2.00-g. sample of
25'70, respectively, and a t this temperature no iz-heptaldoxime was heated in polyphosphoric acid at 130",
oxime was recovered. It is apparent that both
forms of the oxime are present during rearrange-
according t o the previously published Procedure A.7
yield was 1.83g. (92%) of crude n-heptamide, m.p. 78-81 .TF
Recrystallization from pentane-benzene gave colorless rods,
ment in this case. 111.p. 9 - 9 2 " . Theeidentity of the crude product and the
K'hen anti-benzaldoxime (solid, m.p. 12;') purified material was established by comparison of the in-
was subjected to rear gement with polyphos- frared absorption spectra with that of an authentic sample.
phoric acid at %lo, a ,- yield of benzonitrile re- A niised melting point determiriation with an authentic
sample showed 110 depressioii.
sulted, with a 30% rec ry of oxime. At 130°, Rearrangement of syn-Benza1doxime.-The rearrange-
a 409; yield of benzamide was obtained. No form- iiieiit of 20.0 g. of syn-benzaldoxime (liquid form) was
anilide was found. ,Ilthough these yields are not carried out i n polyphosphoric acid a t 90' according to Pro-
high, it may be concluded that the syn and anfi cedure R.' The result was 7.4 g. of light yellow oil which
\\-as separated into its components by chromatography on
configurations are indeed correctly assigned, but alumina. The chief fractions, with the solvents employed,
that direct rearrangement of the syn compound is w r e its follows: ( a ) 1.55 g. (8%) of formanilide, m.p. 40-
accompanied by partial isomerization to the anti 43', with 3:1 pentane-benzene; ( b ) 2.66 g. (13%) of re-
form, with subsequent formation of both possible covered oxime (identified by infrared spectra), with 1 : 1
benzene-ethyl acetate; and (c) 1.94 g. (10%) of benzamide,
amides. m.p. 126-128', with 9: 1 ethyl acetate-ethanol. The
The hydrochloride of the anti-oxime may be ob- products were identified by mixed melting points and by
tained by treatment of the syn-oxime with hydrogen comparison OF t h e iiifrned spectra with thosc of authentic
chloride. The anti-oxime, a solid, is obtained from s:1111plrs.
-4 siniilar rearrangement was carried out, on a smaller
this salt, and is reported to be unstable with respect wale (11.3 g. of oxime), at 130", according t o Procedure B.'
to the syn-oxime, a liquid. The conversion has Two products were isolated by chromatography on alumina
been described as occurring on heating or on con- of the crude product (5.2 g.): 1.25 g. (9%) of formanilide
tact with acids. Our experiments indicate that and 3.51 g. (25%) of benzamide, identified as before.
Rearrangement of anti-Benza1doxh.e.-The rearrange-
the anti-oxime is not isomerized by heating in poly- ment was carried out as described for the syn-oxime, and the
phosphoric acid, but that these conditions do result products were separated in pure form by chromatography
in partial isomerization of the syn form. This is on alumina. U'ith a temperature of 90", the yield of benz-
the reverse of what might be expected from the amide was 5y0,with 30% of recovered oxime. At 130", n o
oxime was recovered, and a 41yo yield of benzamide was
nature of the reaction conditions, if the assumption obtained. Identifications were made as before.
that the y n form is the most stable form is ac- Rearrangement of anti-Benzaldoxime Hydrochloride.-A
cepted. mixture of 2.0 g. of anti-benzaldoxime hydrochloride and
Although Beckmann rearrangements have been 3 1 g. of polyphosphoric acid was heated to 130' and maiti-
tained at 130-135° for five minutes. Hydrogen chloride
demonstrated to occur through the influence of was evolved. The mixture was treated with cold water,
heat alorie on certain acyl derivatives of oximes, arid the clear aqueous solution was neutralized with potas-
it is likely that acid catalysis is a part of the usual sium hydroxide a t a temperature below 10". The aqueous
Beckmann procedure. Polyphosphoric acid is not solution was extracted with ether, and the ether solution
was dried and evaporated with the aid of an air jet. A
a strong acid, but by using the oxime hydrochloride yield of 1.2 g. (8Oyo)of colorless benzamide, m.p. 118-122',
rather than the free oxime, it is possible t o provide was obtained. The crude a n d :i purifietl saniple were iden-
more effective conditions of acid catalysis. Since tified b y infrared spectra.
( 1 1 ) h l Xiillsir 4 Ile,ri f rJll \, I J w p l ' , I' 1 l o - , 14141 f3l?TiiI?SDA, x I A R Y I . A S I )