3114 GILBERTSTORK AND FRANK

H. CLARKE, JR. Vol. 83

[CONTRIBUTION
FROM THE DEPARTMENT
OF CHEMISTRY,
COLUMBIA NEW Y o u 27, N. Y . ]
USIVERSITY,

Cedrol : Stereochemistry and Total Synthesis

BY GILBERTSTORK
AND FRANK
H. CLARKE,
JR.
RECEIVEDMARCH4, 1961

The total stereospecific synthesis of natural cedrol is described. This synthesis establishes the stereochemistry of cedrol
as that shown in Ib.

We have outlined elsewhere's2 the chemical and This leaves four possible relative arrangements
spectral evidence which led us to the establishment of structure I, illustrated by Ia to Id.
of structure I for cedrol and, consequently, I1 for
a-cedrene, two sesquiterpenes which occur in OH
"cedar wood oil," the essential oil of several ***OH
juniper species.

I I1
I t will be noted that the solution of the structural
problem is not completed with the establishment
of structure I since there are five asymmetric centers
IC ' Id
I
'

in the cedrol molecule, and thus there exists the Cedrol has been tiansformed into "norcedrene-
possibility of sixteen optical isomers (rather than dicarboxylic acid" (VI), with attendant destruc-
thirty-two since the carbon-carbon bonds a t 3-3a tion of ring C and of the center of asymmetry a t
and 1-7a are necessarily cis to each other). Cs. Since there are only two possible structures for
The available degradative evidence permits the norcedrenedicarboxylic acid and since such a
conclusion that the hydrogens a t position 7 and substance appeared to us an ideal intermediate
4a must be trans to each other : It has been demon- for a total synthesis of cedrol itself, we will turn
strated by Plattner, et u L . , ~ that catalytic hydro- our attention first to the synthesis of this di-
genation of the olefinic acid 111 gives a saturated carboxylic acid. We shall then return to the prob-
acid 11' d i f e r e n t from the acid V obtained by de- lem of the C3 asymmetry and of the elaboration
struction of ring B and C of the cedrol molecule. of the C ring of cedrol.
Since the acids IV and V could be equilibrated
without becoming interconverted, the non-identity
came from the different relative stereochemistry
a t positions 1 and 3 of the two cyclopentane acids.
Xs Plattner, et al.,have pointed out, the mode of
formation of IV permits only a cis arrangement a t
CH,
these centers and consequently the hydrogens a t VIb
Ci and C3 in the degradation acid V, and therefore Of the two possibilities VIa and VIb, the latter
the hydrogens a t positions 4a and 7 in cedrol itself, appeared a priori rather more likely since the trans
must be trans oriented.
fusion of the A/B system in VIa, while not impos-
sible, implies a degree of strain4 which would
be surprising on the basis of biogenetic speculations.
We therefore turned our attention initially to
the possible stereospecific synthesis of the dicar-
boxylic acid VIb. As will become apparent, the
synthesis was accomplished and VIb proved ideiiti-
cal with norcedrenedicarboxylic acid from natural
\ cedrol.
For a variety of reasons, some of which will be-
come apparent, we chose to start with ring B and
proceed with the attachment of ring A. For this
I - bCozH D C O Z H
purpose the cyclopentanone diester VI11 appeared
suitable. This had been prepared previously by
V .KZH IV XCOZH
Simonsen, et al.,5 from diethyl 2,2-dimethyl-3-
carbethoxyadipate (VII), an improved synthesis
of which is given in the Experimental section of this
(1) G Stork and R. Breslom, J . A m . Chem. Soc., 1 5 , 3291 (1953).
(2) G.Stork and R. Breslow, i b i d . , 1 5 , 3292 (1953). (41 J. W. Barrett and R. P. Linstead, J . Chcm. Soc., 436 (1935).
(3) P1. A. Plattner, A. Fiirst, A. Eschenmoser, W.Reller, H. KlBui, (5) C.S. Gibson, K,V. Hariharan and J. L. Simonsen, i b i d . , 3009
St. Meyer and M. Rosner,Hclu. Chim. A d o , 86, 1845 (1953). (1927).
July 20, 1061 STEREOCHEMISTRY SYXTE-IESIS
AND TOTAL OF CEDROL 3115

EtO2C
yc;E( "q
CO,Et

VII CH?
d
Et02C

0
COZEt

\'I11
out of the plane in the enolate ion (hindrance by
the gem-dimethyl grouping can be assumed to
be substantially identical on either side of that
plane) .6

paper. A possibility for the construction of ring

A involves attachment of a carbon chain, R, a t the
@-keto ester carbon (cf. I X ) and of another, R',
a t the carbonyl center followed by ring fprmation
EtOZC q * ' C 0 2 E t -
EtOZC
" q C 0 2 E t -
between R and R'. The degree of anticipated
Ef+(02Et
0'
R02C
H
I CH,

0 Et02C'
' C q H C o z E t ZE XI
IX
steric hindrance of the cyclopentanone carbonyl 0
in IX, however, makes i t extremely dubious that The internal aldol cyclization which we wished
addition of an external reagent to the carbonyl to use to complete the attachment of ring A now
group could be successfully carried out. I t is requires the transformation of the free carboxyl
therefore R itself which must be so constructed in X I into a methyl ketone grouping. This simple
as to permit intramolecular formation of the change is complicated here because the acid chloride
carbon-carbon bond a t the carbonyl center. XII, derived from X I , is readily transformed into
It is these considerations which led to the choice the isomeric-and unreactive-chlorolactone X I I I .
of benzyl a-bromopropionate as the alkylating Thus, although XI1 can be prepared by the re-
agent: Alkylation of VI1 was carried out by letting action of the dry sodium salt of X I and oxalyl-
its sodium salt, prepared with sodium hydride, chloride, reaction with dimethylcadmium is unsuc-
react for several days a t ca. 50' in a medium of cessful (MgX2 presumably catalyzing the XII-
benzene-dimethylformamide with the bromo ester. X I I I transformation).
The resulting alkylated substance was then puri-
fied by distillation under high vacuum and the C02Et COzEt
benzyl ester (X, 90% yield) was transformed into COzEt A+ COzEt
the corresponding free acid by hydrogenolysis in
ethyl acetate in the presence of palladium-on-
charcoal. The resulting monocarboxylic acid X I
H o 2 C0 J h . . * H -L m7{)*..H
C1O --f

was obtained in 45y0 yield as colorless rectangular XI XI1

prisms, m.p. 113-115' after recrystallization from
cyclohexane. In the acid XI, the chain which is
to be used to construct ring A has now been
differentiated from the two other carboxyl functions
and is ready for further elaboration.
I
c1
XI11
It was therefore necessary to employ a method
which would not only allow the survival of the p-
X XI keto ester system but would also avoid any acid
reagent which could catalyze the XII+XIII
Before proceeding with this, we would like to transformation: reaction of XI1 with excess diazo-
remark on the stereochemistry indicated in X I . methane' produced the diazoketone XIV, neutral-
Two new asymmetric centers have been intro- ity being maintained by the instantaneous reaction
duced with a remarkable degree of stereospecificity. of the liberated hydrogen chloride and diazometh-
One of these, a t the methyl-bearing carbon alpha ane. The crystalline diazoketone was converted
to the free carboxyl group of XI, is of no concern with anhydrous hydrogen chloride8 in ether into
to us a t this point. The stereochemical arrange- the chloromethyl ketone XV, m.p. 72-73', and the
ment a t this center will not become fixed until the latter was finally reduced with zinc dust and
final removal of the carbonyl function adjacent acetic acid in the presence of potassium iodideg
to it. What is, however, of considerable moment
a t this stage is that the cis arrangement of the two (6) The same stereochemical result would be obtained if the transi-
tion state for the alkylation looked very much like the product because
carboxyl functions a t C1 and C3 in X I , which is the bulk of the alkylating agent is greater than that of the carbethoxyl
that present in norcedrenedicarboxylic acid VIb, group present at the alkylation site and XI would be the less hindered
has now been established. of the two possible products.
The lower energy transition state for the alkyla- (7) Cf.A. L. Q'ilds and T. L. Johnston, J . A m . Chcm. SOL.,70, 1166
(1948,.
tion of VI1 must involve approach of the bulky ( 8 ) Cf.Ch. R. Engel and G. Just, ibid., 76, 4909 (1954).
benzene a-bromopropionate fragment from the side (9) Cf.A. GrUssner. J,-P. Bourquin and 0. Schnider, Helu. Chim.
away from that carbethoxyl group which remains A d o , aa,-bi? (1946).
3116 AND FRANK
GILBERTSTORK H. CLARKE,
JR. VOl. 53

to the desired methyl ketone XVI, a low melting C'O Et

solid which was further characterized as its semi-
carbazone, m.p. 191-193'.

O' XVIl XVII

I

XIV
COZEt

The remarkable efficiency of this sequence of

steps is illustrated by the 79% over-all yield
of crystalline methyl ketone XVI obtained from COaEt,?
the sodium salt of the acid XI. The methyl ke-
tone is undoubtedly a mixture of epimers a t the
center next to the side chain carbonyl. This is
reflected in the fact that XVI gave analytical
values in excellent agreement with the expected
structure, although its melting point remained a t
55-60' after several crystallizations.
We were now in position to complete the con- XVIIIa
struction of the A/B skeleton by aldol cyclization related @-lactone X X I I has been shown to form
of XVI to the important intermediate XX. easily when the hydroxy acid X X I is treated with
Attempts to effect this apparently simple step acetic anhydride. lo An important additional fac-
by treatment with p-toluenesulfonic acid in hot tor here is that the y-lactone XVIIIa not only
benzene or with aluminum t-butoxide in boiling has the rather strained bicyclo [1,2,2]heptane
toluene remained without effect. Treatment with
potassium t-butoxide in boiling t-butyl alcohol ap-
peared to produce cyclization and dehydration :
It was possible to obtain from one such experi-
ment, followed by careful chromatography on
alumina of the neutral products, a trace of crystal-
H HO
o , C ~ C o z H
XXI
omcozH XXII
line material, m.p. 65-66', which had the expected system, but additionally the equivalent of a diaxial
absorption in the ultraviolet: XEt,"," 228 m,u, log methyl/methyl interaction, and consequently y-
E 4.1. The main product of such experiments, lactone formation is here much less favorable than
however, was clearly a different cyclopentenone
system than X X : the crude material had A:"."," -
240 nip while the infrared spectrum, although dif-
usual.
A successful solution of the vexing problem of
achieving a satisfactory transformation of XVI into
ferent from that of X X nevertheless clearly re- the bicyclic enone X X was reached by noting that
vealed by its intense -61 p band coupled with the the step XVII + XVIII must be considerably
carbonyl absorption a t -5.82 p its cyclopentenone slower than the formation of the aldol XVII from
nature. The position of the ultraviolet absorp- the methyl ketone XVI: Treatment of XVI for
tion band (corresponding to one more degree of ten minutes a t room temperature with potassium
substitution than is present in XX) together with t-butoxide in t-butyl alcohol led to almost com-
the isolation from the crude material of a crystal- plete transformation into XVII as shown by in-
line semicarbazone, m.p. 174-176' which analysis frared spectra (disappearance of bands due to
showed to be that of a decarbethoxy derivative of XVI; appearance of hydroxyl absorption; no
XX, leaves no doubt that this product has the cyclopentenone peaks). The aldol XVII could
structure shown in XIX. then be dehydrated to the desired bicyclic enone
A likely reaction sequence for the surprising X X by refluxing for one hour with p-toluenesul-
formation of X I X from XVI is outlined. fonic acid in benzene solution.
The main point of interest here is the postu-
lated participation of a @-lactone XVIII rather
than the possible y-lactone XVIIIa involving the
other carbethoxyl group. This is not as unlikely
as might initially appear because the rigid system
OH
_Lt_
present in XVII removes most of the entropy dif- XX
XVII
ficulties which serve to explain the very slow for-
mation of four-rxembered compared to three- (IO) N. J. Toivonen, P. Hirsjarvi, A. Melaja, A. Kainulainen, A.
membered rings. For instance, the rather closely Halanen and E. Pulkkinen, Acto Chcm. Scand., 8, 991 (1949).
July 20, 1961 STEREOCHEMISTRY
AND TOTAL
SYNTHESIS
OF CEDROL 3117

The numerous difficulties we had encountered oily mother liquors which remained after separation
with the XVI + X X conversion made the almost of the crystalline ketodiester. This ketodiester
80% yield of crude crystalline X X a source of is correctly represented by XXIIIa and is identical
considerable satisfaction. with that from lithium-ammonia reduction: the
We are now ready to consider the establishment disparity in melting points of the dinitrophenyl-
of the cis fusion of rings A and B. Two methods hydrazones from the two sources was due to di-
come into consideration in this particular case : morphism. Recrystallization of the 1GO-16l0 form
Reduction of X X can be carried out either catalyti- from ethyl acetate-ethanol converted i t to the al-
cally or chemically. The stereochemical course of lomorph, m.p. 152-154', undepressed on admixture
the catalytic hydrogenation of a system such as X X with the product from the lithium-ammonia re-
is not easily predictable because catalyst-substrate duction.
hindrance appears considerable on either side of The demonstration that catalytic hydrogenation
the molecule. Such hindrance as exists would of X X gives the cis-bicyclic ketodiester XXIIIa
seem to favor adsorption from the side opposite substantiated our speculations on the possible
the two carbethoxyl groups and might then lead stereochemical results of this particular reaction
to a trans A/B junction, but the considerably higher and greatly simplified the preparation of XXIIIa.
energy of the trans form of the [0,3,3]bicyclo- In the latter, three of the asymmetric centers of
octane system than that of the cis isomer4 might norcedrenedicarboxylic acid have now been in-
well raise the energy of the hydrogenation transi- troduced while the fourth center-that bearing the
tion state leading to the trans system to a higher methyl group in ring A- will now be considered:
value than that leading to the cis. We will show that the more stable arrangement a t
Chemical reduction of the bicyclic enone X X the methyl-bearing carbon coincides in fact with
with lithium-liquid ammonia would, on the other that which exists in the norcedrenedicarboxylic acid
hand, be expected to lead to the desired cis A/B molecule itself, i.e., that the more stable arrange-
stereochemistry": Of the tu70 products in which ment of the methyl group is cis to the angular
the 4a-hydrogen is axial to ring A, the isomer with carboxyl group.
the cis =Z/B system is certainly the more stable. Such a conclusion is of some practical importance
It was of course necessary to carry out the reduc- since the methyl group, being adjacent to the ke-
tion without much excess lithium to prevent as tonic function in ring A, is readily equilibrated to
much as possible reduction of the ester functions. the more stable arrangement, but i t is also of
Under such conditions the saturated ketodiester considerable interest from a conformational analy-
XXIIIa was obtained and characterized as its sis point of view.
2,4-dinitrophenylhydrazone,m.p. 152-154', after We first point out that a conclusion opposite
crystallization from ethyl acetate-ethanol to that we have just mentioned above for XXIIIa
would be reached by consideration of the simple
"model" 2-methyl-3-oxocyclopentanecarboxylic
acid which is certainly more stable with the
methyl and carboxyl groups in trans relatioriship,l2
but i t is immediately evident that this situation-
1 which would be unfortunate for our purposes-
,--

need not apply in the more complex case which

concerns us here. A more closely related model is
available in the 14-iso steroid series, for instance
the pair of acids XXIV and XXV. Here again,
the known position of equilibrium which is in

XXIIIa SXIIIb
With the authentic cis-bicyclic ketone XXIIIa in
hand, i t was now possible to investigate the course
of the catalytic hydrogenation of X X : with pal- XXIV xxv
ladium-harcoal as a catalyst, in ethanol solution, favor of X X V a might seem an unfavorable omen.
a saturated ketone was obtained as a crystalline We must, however, realize that the energy dif-
solid, m.p. 33.5-35'. This was converted into ferences which will dictate which of the two isomers
a dinitrophenylhydrazone melting a t 160-161' XXIIIa and XXIIIb is the more stable are quite
after recrystallization from ethanol. The same
dinitrophenylhydrazone was obtained from the (12) Cf. A. P. Arendaruk, E. I. Budovskil, B. P. Gottikk, M. Y.
Karpekkil, L. I. Kudryashov, A. P. Skoldinov, N. V. Smirnova, A. Y.
(11) G. Stork and S. D. Darling, J . A m . Chem. Soc., 88, 1512 Khorlin and N. K. Kochetkov, Zhur. Obshchci Khim., 27, 1312 (1957).
(1980). The bracketed structures in which O* means either 0-or 0- (13) See, for instance, A. Lardon and T. Reichstein, Rrlu. Chim.
illustrate the two conformations of the transition state for protonation. Acto, 41, 904 (1958).
3118 GILBERTSTORK H. CLARKE,
AND FRANK JR. Vol. 83

small (of the order of one kilocalorie) and that con- dicarboxylic acid prepared from natural (+)-
sequently even as close a model as XXIV may be credrol.
an improper one. Specifically, we can predict
that the undesired trans arrangement of the methyl
and carboxyl groups should be considerably less
favorable in the czs five-five system than in a six-
five model such as XXIV: Contraction from a six
to a five-membered ring imposes a rotation on the XXIX VIb
starred carbon atom (cf. XXVI) such that the group
R which in XXVI is staggered between the two The conclusion that the dicarboxylic acid VIb
hydrogens on the starred carbon is now opposed W ~ Sactually dl-norcedrenedicarboxylic acid was
to one of them as in XXVII 2nd the equilibrium substantiated by the resolution of our synthetic
should now be in favor of XXVIII. acid. This proved remarkably easy: Slow evapo-

H pH$b
ration of a solution of dl-noscedrenedicarboxylic
acid and one equivalent of quinine in acetone pro-
duced a crop of crystals which after one recrystal-
lization had the same rotation, [ a I z 7 ~-123',
. .R . . .R .H as that of the quinine salt of authentic (-)-
norcedrenedicarboxylic acid and did not depress its
XXVI Rt XXVII melting point. Regeneration ,of the resolved acid
from its salt gave synthetic ( - )-norcedrenedicar-
boxylic acid, m.p. 212-213', undepressed by mixing
with authentic material.14 The rotation of the syn-
H p : thetic acid was [ c x ] ~ I ' D -38.9 *
1.5' in acetone.
The authentic (-)-acid has E rotation [a]27D-38.3
XXVIII +1.0.l4 Infrared spectra were, of course, also
identical.16
These considerations convinced us that in the We now have an extremely valuable synthetic
particular system a t hand the desirable stereo- intermediate for the further construction of the
chemical situation in which the 7-methyl group is cedrol molecule and have incidentally established
cis to the angular carbethoxyl group (XXIIIa) the position of the XXIIIa 4XXIIIb equilibrium
would probably be favored if the methyl group to be, as we had surmised, in favor of XXIIIa.
were allowed to equilibrate to the more stable Before proceeding with the elaboration of ring
arrangement. The validity of the assumption C, it is necessary to settle the remaining stereo-
was in any event soon verified by transformation chemical uncertainty. The synthetic route to
to norcedrenedicarboxylic acid. norcedrenedicarboxylic acid establishes that cedrol
Before proceeding with the removal of the ketonic is correctly represented by either Ia or I b and rules
function of XXIIIa, i t was necessary to establish out the other two possibilities which existed a
whether the methyl group in XXIIIa was already p Y i O V i .
in the more stable configuration: Prolonged re- .I decision between Ia and I b can be reached
fluxing of the ketodiester XXIIIa with 20% b-! noting thdt in l a the tertiary hydroxyl group
aqueous alcoholic potassium hydroxide led to a. is axial while i t is eyuetorial in Ib. We were able
ketodiacid which was converted into its crystalline to show that I b is the correct representation of the
anhydride, 1n.p. 171-173', on refluxing with acetic cedrol molecule in the following manner: Epoxi-
anhydride. Hydrolysis of the anhydride by re- dation of cedrene (11) can only take place from the
fluxing with aqueous dioxane and esterification side of ring C opposite the gem-dimethyl group
with diazoethane produced a ketodiester which had which effectively blocks approach to the side of
the same infrared spectrum 2 s the original diester ring C facing it. The resulting epoxide X X X
and which gave the same ~,~-dinitrophenylhy-could be reduced with lithium aluminum hydride
drazone (m.p. 159-161' from ethanol, alone or with the formation of an alcohol which was not
mixed with an authentic sample). This result oxidized by chromic a.cid and was therefore tertiary.
establishes that the C7-methyl already possesses This alcohol, which must be Ia, was not cedrol
the more stable conformation in X X I I I a itself and the latter must then be represented by Ib.
and confirms (reformation of XXIIIa from the This conclusion not only settles the stereo-
anhydride) the cis disposition of the two carbcthoxyl chemistry of the cedrol molecule, i t also has definite
groups. synthetic implications. It would not be easily
We now had to remove the carbonyl group o i feasible, for instance, to synthesize cedrol VZU
XXIIIa : Treatment of XXIIIa with ethanedithiol cedrene since direct hydration or, as we have just
in the presence of anhydrous hydrogen chloride gave seen, epoxidation followed by reduction would
the crystalline thioketal XXIX, m.p. 75-76O, which lead t o the epimeric axial alcohol. In fact the
was desulfurized by hydrogenolysis with Raney (14) Norcedreaedicarbonylic acid from natural (+)-cedrol is re-
nickel. Saponification of the diester gave a crys- ported t o have m.p. 209' and [n]'4~-39.4' (in CHCla); cf. ref. 16.
talline diacid VIb, m.p. 221-223' after recrystalli- (15) The identity of our cis fused bicyclic diacid with norcedrenedi-
zation from ether-pentane. This substance gave carboxylic acid from cedrol proves th: cis fusion of the two &membered
the analytical values expected for dl-norcedrenedi- rings in cedrol itself. T h a t equilibrium in the pair X X I I I a XXIIIb
4

carboxylic acid and its infrared spectrum was in- isPlattner,

in favor of X X I I I a then follows from the fact that the experiments of
et d , , a rigorously establish that the C ~ m e t h y must
l be c i s t o
distinguishable from a sample of ( -) -noreedrene- the 4r-hydrogen as discua4ed a t the beginning of this paper.
July 20, 1961 STEREOCHEMISTRY SYNTHESIS
AND TOTAL OF CEDROL 3119

major enol ether should be XXXV. Remarkably,

the only crystalline product isolated from the re-

H H
I1 xxx

&I H, &
:bH XXXIV XXXV XXXVI

. . 'CH? ' ' 'CH3 duction was a small (ca. 18%) yield of a saturated
CHJ CH3
alcohol, n1.p. 120-121 ', later identified (see below)
H H as the alcohol XXXVII. Its formation, while sur-
Ia Ib prising, can be rationalized from the isomeric
stereochemistry of I b strongly suggests that the enol ether XXXVa undoubtedly present in the mix-
desired result should be obtained by Grignard ture. The structure of the alcohol XXXVII follows
reaction on the ketone X X X I which thus became TOEt -.Al-
I
our next synthetic objective.

XXXVa
XXXI H
The half-methyl ester X X X I I of norcedrene-
dicarboxylic acid was prepared, essentially as de-
scribed previously by Plattner, et a1.,16 by the
partial saponification of the dimethyl ester ; the
free carboxyl group was then transformed into a
methyl ketone by either of two routes. The acid XXXVI XXXVII
chloride prepared by reaction of X X X I I with
oxalyl chloride could be treated, as mentioned in from an unambiguous synthesis: Reduction of the
an earlier paper,l with dimethylcadmium or the ketoester X X X I I I with lithium aluminum hy-
sequence acid chloride 4 diazoketone + chloro- dride followed by oxidation with chromic acid-
methyl ketone + methyl ketone which we used for pyridinela led t o the crude related ketoaldehyde.
the conversion of X I into XVI could be employed. This was cyclized with aqueous base to the cyclo-
I n spite of the number of steps, the efficiency of the hexenone XXXVI which was characterized as its
latter sequence is extremely high and the crude 2,4-dinitrophenylhydrazone,brilliant red prisms,
methyl ketone XXXIII is thus obtained in 95% melting at 164-167'. Reduction of the crude cyclo-
yield from the half-acid. The methyl ketone ester hexenone with palladium-charcoal to the cor-
was characterized by its 2,4-dinitrophenylhydra- responding saturated ketone and treatment of
zone, m.p. 140-142'. The same dinitrophenyl- the latter with lithium aluminum hydride gave the
hydrazone was obtained from the methyl ketone ciystalline alcohol XXXVII, m.p. 120-121 ' alone
made by either route. Treatment of the crude or mixed with the previously obtained sample.
ketoester X X X I I I with potassium t-butoxide in Although we had thus available the alcohol cor-
responding to the ketone which was our immediate
synthetic objective, the yields in either sequence
were low and it seemed desirable to find a better
u u route. The course of the reduction of the enol
ether of the P-diketone XXXIV suggested that the
XXXIl SXXlII latter substance itself should result in the forma-
l-butyl alcohol resulted in cyclization in high yield tion of the alcohol XXXVII or of the correspond-
to the P-dikctone XXXIV, m.p. 302-204'. ing cyclohexenol : Salt formation on addition of
I n the /3-diketone XXXIV we have built up the lithium aluniinum hydride to the 8-diketonc
skeleton of the desired ketone XXXI and it is only XXXIV should lead predominantly to the struc-
necessary to remove the extra carbonyl group. ture XXXVIII since the aluminum is thus attached
Reduction with lithium aluminum hydride of the to the less hindered oxygen (see selective formation
enol ether mixture prepared from the B-diketone of the half-ester XXXII). Further reduction
was initially considered a plausible route to the would then result in the cyclohexenone XXXVI.
cyclohexenone XXXV117on the assumption that the It might be expected that XXXVI would then be
reduced to the unsaturated alcohol, but the r educ-
(16) P1. A Plattner, G. W. Kusserow and H. Klaui, Helv. C h i n . tion product was mainly saturated alcohol: Oxida-
A d a , 26, 1345 (1942).
(17) Cf.R. L. Frank and H. K. Hall, Jr., J. Am. Chetn. Soc., 72, (18) Cf. G. I. Poos, W. F Johns a n d L. H. Sarett, tbid.. 'I?.1026
1645 (1950). (1954).
3120 GILBERTSTORKAND FRANK
H.CLARKE,
JR. Vol. 83

u-t" - XXXI Ib
XXXVII synthesis of cedrol is thus complete and its stereo-
OH OH chemistry is established as Ib.22*23
Experimental
Ethyl 2,3-dicyano-3-methylbutyrate was prepared ac-
cording to Smith and Horwitz.2' Yields of 76% to 82%
were obtained; b.p. 99-1140 (cu. 0.01 mm.). When the
product was distilled a t a hlgher pressure, b.p. 120-127"
(4 mm.) with a bath temperature of 165-170" some de-
XXXVI XXXVII composition took place and the infrared spectrum of the
product showed a double bond band a t 6.18 fi which was
lacking in the spectrum of the product obtained by distil-
tion of the alcohol mixture gave ketones, the ultra- lation a t the lower pressure.
violet spectrum of which indicated only 10--15% Ethyl 2-methylpentane-Z,3,4-tricarboxylate(VII) was
a,@-unsaturatedcarbonyl. The main constituent prepared in the following manner. Ethyl 2,3dicyan(~3-
methylbutyrate was cyanoethylated with acrylonitrile in
of the alcohol mixture was in fact the same ctystal- t-butyl alcohol using choline as a catalyst. The crude
line alcohol XXXVII, m.p. 120-121°, which we product was isolated and hydrolyzed to the triacid without
had obtained previously. Whether the saturated purification. After removal of water and ammonium chlo-
alcohol comes from 1,4-reduction of the enone ride from the hydrolysis mixture, the triacid was esterified
with alcohol and sulfuric acid to give the triester VII.
XXXVI followed by reduction of the ketone, The following details are typical of the procedure used.
liberated from its enolate during work-up, by excess Ethyl 2,3-dicyano-3-methylbutyrate(250 g.) was dis-
hydride, or whether the saturated alcohol is pro- solved in 232 g. of t-butyl alcohol, and 87 g. of acrylonitrile
duced by hydride reduction of the double bond of and 5.0 ml. of a 40% solution of choline in methanol were
added. The reaction mixture was surrounded with ice for
initially formed unsaturated alcohol is not known. 6 hours, then left a t room temperature for 18 hours and
I n either case the facile reduction of the double finally at 50" for 40 hours. The reaction mixture was then
bond implies unusual reactivity presumably be- poured into a separatory funnel and cooled until crystals of
cause of the strain involved in putting a double t-butyl alcohol began to appear in the upper layer. The top
layer was then separated and the solvent removed on the
bond in this particular position. 19,20 steam-bath under reduced pressure. The residue was com-
Oxidation of the now readily available alcohol bined with the bottom layer, ether was added and the solution
XXXVII with chromic acid gave the ketone was washed 5 times with a cool solution of sodium chloride.
XXXI, characterized as its 2,4-dinitrophenyl- The aqueous layer was washed once with ether, the ether
layers were combined and the solvent was removed on the
hydrazone, m.p. 146-147'. Addition of the ketone steam-bath under reduced pressure.
to a solution of methyllithium in ether followed by The combined residues from two batches (638 g.) were
refluxing gave, as anticipated, (+)-cedrol (Ib), dissolved in 1300 ml. of concentrated hydrochloric acid and
m.p. 8ti-87' after recrystallization from aqueous 1300 ml. of water was added. The solution was refluxed
for 48 hours and then evaporated to near dryness in two
methanol. The melting point of the mixture large evaporating dishes left on steam-baths overnight.
with natural cedrol, m.p. 86-87', was undepressed. In the morning the mixture was cooled and the ammonium
The infrared spectrum of the synthetic material chloride collected by suction filtration. The filtrate was
was identical with that of natural cedrol. Finally, taken to dryness on the steam-bath, under reduced pressure,
to give an oily solid. The ammonium chloride was washed
the rotation of the synthetic cedrol in chloroform thoroughly with absolute alcohol and the oily solid was dis-
was [ C X ] ~ ~+10.5
D h0.8' and that of natural cedrol solved in the alcohol washings and filtered from ammonium
was [ a I z 8 +9.9
~ 3~ 0.4°.21 The stereospecific chloride.
T o the filtrate (3500 ml.) was added 700 ml. of concen-
(19) The reduction of a symmetrical p-diketone with lithium alurni- trated sulfuric acid while stirring in an ice-bath. The solu-
num hydride has been reported by A. S. Drtidiug and J. A. Hart- tion was refluxed for 4.5 days. Then about 1 1. of the
man, ibid., 76, 3723 (1953), and was shown t o proceed, essentially a s alcohol was distilled off and the remainder of the solution
we have outlined here, t o the allylic alcohol. The above results sug- was cooled and poured (in three portions), into saturated
gest that the reduction of unsymmetrical p-diketones in which one of the sodium chloride solution. The ester, which separated as an
carbonyls is considerably more hindered than the other might generally oil, was washed with fresh sodium chloride solution and the
result in the removal of the more hindeted oxygen functioo. This aqueous washings were extracted with ether. The ether
could be of considerable synthetic utility but would have to be ex- extract was dried (sodium sulfate) and the solveiit removed
amined in further cases before any confidence could be placed in the
anticipated result. (22) The absolute configuration, now that the relative configuration
(20) Th spectrum of the crude enone XXXVI from the cyclization is known, can be shown to be as implied in all the structures used in
of the ketodldehyde had XZ:" 243 my (see Experimental). This this paper: Ozonization of the exocyclic double bond isomer of cedrene,
indicates considerable strain in the ground state (expected XmSx 227 obtained by pyrolysis of the acetate of natural (+).cedrol gave (-I-)-
mp) thus resulting in lower energy for the transition t o the excited norcedranone (XXXI). This showed a positive Cotton effect curve
state. It is presumably this strain which causes the easy reduction of and therefore the absolute configuration shown in XXXI (see C.
the double bond of XXXVI or the corresponding allylic alcohol during DJerassi "Optical Rotatory Dispersion," McGraw-Hill Book Co., Inc.,
the lithium aluminum hydride reaction. New York, N. Y.,1960). The same conclusion as to absolute configu-
(21) The melting point of natural cedrol is reported t o be 86-81' ration has recently been reached on a different basis by Biichi, e2 ai.
while the rotations include [-In 4-9" 3 l ' a n a +loo 31' (in chloroform). (G.Biichi, R. E. Erickson and N. Wakabayashi, J . Am. Chcm. Soc..
See E. Guenther, "The Essential Oils," Vol. 11, D. Van Nostrand 83, 927 (1961)).
Co., Inc., New York, N. Y., 1949, p. 284. The rotation reported (23) A preliminary report of this synthesis has been published:
here was obtained by us on a sample of the natural crystalline sub- G. Stork and P.H. Clarke, Jr., ibid., 77, 1072 (1855).
stance. (24) P. A. S. Smith and J. P.Horwitz, ibid., 71, 3418 (1949).
July20, 1961 STEREOCHEMISTRY
AND TOTAL OF CEDROL
SYNTHESIS 3121

to give more of the triester. Distillation of the ester gave a For the preparation of the acid chloride XI1 the dry
small forerun and a main fraction (480 g.), b.p. 134-160' sodium salt (11.6 g.) was powdered and covered with 100
(ca. 0.01 mm.), with a residue of 158 g. Redistillation of ml. of dry benzene which contained 1.0 ml. of pyridine.
the main fraction gave 410 g., b.p. 154-156' (cu. 0.01 mm.). The mixture was cooled until the benzene was partly frozen
The residues from the distillations were hydrolyzed and re- and 15 ml. of oxalyl chloride was added. The benzene
esterified to give more of the triester. melted from the heat of reaction and the mixture was left
Ethyl 3,3-dimethylcyclopentan-3-one- 1,4-dicarboxyIate a t room temperature for an hour and a half. The mixture
( V U ) was prepared according to Gibson, el ~ l . using , ~ was then cooled in ice and the oxalyl chloride and benzene
powdered sodium in benzene. The sodium was powdered were removed under vacuum a t room temperature or lower.
with fast stirring in boiling toluene. An oversize reaction Traces of oxalyl chloride were removed by adding fresh
flask was necessary because the reaction once started was benzene and removing it under reduced pressure as before.
vigorous. Usually a few drops of ethanol was added to the The residue was taken up in benzene and the precipitated
boiling benzene solution to initiate the reaction. After the salts were removed by filtration through Celite.
vigorous reaction had subsided the mixture was refluxed for T o form the diazoketone, the benzene solution of the acid
3 hours and worked-up as described.j The product was chloride was poured slowly into an excess of diazomethane
distilled through a Vigreux column, under good diffusion in ether (previously prepared from 45 g. of nitrosomethyl-
pump vacuum, a t 132-133' to give the cyclic keto-diester urea and dried for several hours a t 0" over potassium hy-
in 64% yield. droxide pellets). The solution was left in an ice-bath for
Benzyl a-bromopropionate was prepared by dissolving an hour and a half and then the diazomethane and ether were
34.0 g. of benzyl alcohol, 48.0 g. of a-bromopropionic acid removed on the steam-bath in the hood. The residual ben-
and about 0.5 g. of p-toluenesulfonic acid in 150 ml. of zene solution was taken to dryness under reduced pressure
benzene and refluxing the solution for 18 hours with a water to give an orange colored oil which crystallized on standing
separator. T o the cooled benzene solution ether was added or seeding.
and the solution was washed three times with ice-cold 5% The crude diazoketone XIV was converted to the chloro-
sodium hydroxide solution, dried over anhydrous sodium methylketone by dissolving it in 150 ml. of dry ether and
sulfate and the solvent was distilled off. The residue was passing an excess of dry hydrogen chloride into the ice-
distilled a t 139-141' ( 7 mm.) to give 55.5 g. of ester (73% cooled solution. The solution was filtered through cotton
yield). and evaporated under reduced pressure. The crude oil
Alkylation of the Cyclic Keto-diester.-The cyclic keto- which remained after removal of the ether crystallized
diester VI11 was converted to its sodium salt and alkylated exothermally on seeding. The solid chloromethylketone
with benzyl a-bromopropionate as follows. T o 6.0 g. of was dissolved in cyclohexane and most of the color was re-
sodium hydride and 100 ml. of dry benzene was added 50.0 moved with activated charcoal. Concentration of the
g. of the cyclic keto-diester VI11 in 80 ml. of dry benzene. solution and seeding gave 10.7 g. (90% yield) of the chloro-
The solution was refluxed until hydrogen evolution had methyl ketone XV as yellow crystals, m.p. 65-70'. A por-
ceased and was then decanted from unreacted sodium hy- tion of the chloromethylketone, after several recrystalliza-
dride. The latter was washed with 30 ml. of dry benzene tions from cyclohexane, melted a t 78-73'.
and the combined benzene solutions were distilled until 155 Anal. Calcd. for C~HZ,OBC~: C, 56.58; H, 6.98.
ml. of distillate had been collected. T o the residual solu- Found: C, 56.41; H, 7.08.
tion of the sodium salt 47.5 g. of benzyl a-bromopropionate Reduction of the Chloromethylketone XV.-The chloro-
and 100 nil. of dry dimethylformamide (dried by azeo- methylketone XV (19.2 g.) and 20 g. of powdered potas-
tropic distillation with 10% of benzene; b.p. 148-153"), sium iodide were dissolved in 250 ml. of glacial acetic acid.
was added and the mixture was heated a t 50' under a nitro- Zinc dust (100 9.) was then added slowly with stirring a t
gen atmosphere. A second batch of 71.1 g. of the cyclic 25" and stirring was continued for 6 hours. Water (50 ml.)
keto-diester was alkylated simultaneously using propor- was then added and stirring was continued overnight. The
tionate amounts of reagents. At the end of 1 week the solids were filtered off and washed with 80% acetic acid.
solutions were combined and poured into ether. The ether The combined washings and filtrate were taken to dryness
solution was washed three times with ice-cold 570 sodium on the steam-bath under reduced pressure and the residue
hydroxide solution and three times with water, then dried was taken up in water and ether. The ether solution was
over anhydrous sodium sulfate and the solvent distilled. A washed with water, then three times with aqueous potas-
short-path distillation of the residue using good diffusion sium carbonate solution containing some sodium thio-
pump vacuum gave 179 g. (9070 yield) of the triester X sulfate. Finally, the ether solution was again washed with
boiling at 175-192" a t a bath temperature of 200-205'. water, then dried over anhydrous potassium carbonate and
Hydrogenolysis of the Benzyl Ester.-The benzyl ester the ether was removed to give an oily residue which crystal-
X (77.0 g.) was hydrogenolyzed with 8.0 g. of 10% palla- lized on standing to give 17.1 g. (m.p. 50-57"). The solid
dium-on-charcoal catalyst in 150 ml. of ethyl acetate, using was dissolved in petroleum ether (b.p. 30-60') and the solu-
a Parr shaker. The pressure fell from 50.0 Ib. to 33.0 lb. in tion was treated with activated carbon, then concentrated
12 minutes and to 32.6 lb. in 35 minutes when the hydrogen and cooled. Colorless crystals of the methyl ketone XVI
uptake ceased. The palladium-charcoal was collected on were obtained which melted a t 55-60', yield 15.3 g. (8870).
Celite and washed with dry ether. The ether and most of A portion of the methylketone was distilled twice in high
the ethyl acetate were removed on the steam-bath under vacuum a t 105" for analysis.
reduced pressure. The residue was poured into an erlen- Anal. Calcd. for C1&$,06: C, 62.56; H, 8.03. Found:
meyer flask and boiling petroleum ether (b.p. 30-60") was C, 62.34; H, 7.92.
added until an oil precipitated out of solution. The oil
crystallized on seeding and 47.0 g. of oily crystals was col- The semicarbazone formed in poor yield either in pyri-
lected. Three recrystallizations from cyclohexane gave dineethanol or with sodium acetate in aqueous ethanol.
27.0 g. (45% yield) of the acid XI as colorless rectangular Colorless crystals, m.p. 191-193', were obtained after
prisms, m.p. 113-115". recrystallization from ethanol-water and drying for analysis
in a high vacuum a t the temperature of boiling methanol.
Anal. Calcd. for C I ~ H H O C,~ : 58.52; H, 7.37. Found: Anal. Calcd. for C18H18O~N1: C, 56.38; H, 7.62.
C, 58.47; H, 7.43. Found: C, 56.41; H, 7.91.
Preparation of the Chloromethylketone XV.-The crystal- Cyclization of the Methylketone XV1.-Repeated at-
line acid XI was transformed first to the dry sodium salt, tempts t o obtain the cyclopentenone X X directly from t h e
and then to the acid chloride with oxalyl chloride. Treat- methylketone XVI using potassium t-butoxide in refluxing
ment of the acid chloride with diazomethane gave a crystal- t-butyl alcohol gave only low yields of the desired product.
line diazoketone which was transformed directly into the A considerable quantity of the cyclopentenone monoester
chloromethyl ketone with hydrogen chloride in ether solu- X I X was formed simultaneously: To a solution of 0.7 g.
tion. of potassium in 70 ml. of dry t-butyl alcohol was added 4.3
T o prepare the sodium salt, the crystalline acid X I was g. of the crude methyl ketone XVI in 30 ml. of dry t-butyl
powdered in a mortar and mixed with a 570 excess of sodium alcohol and the mixture was refluxed for 3 hours under
bicarbonate. The mixture was placed in an oversized nitrogen. Addition to cold water, extraction with ether,
flask, water was added, and the mixture was stirred until followed by washing of the ether extracts twice with cold
the solid had all dissolved. The water was then removed 5% sulfuric acid, twice with water, six times with 5% so-
under vacuum a t room temperature or lower. dium hydroxide, drying over sodium sulfate and removal of
3122 AND FRANK
GILBERTSTORK H. CLARKE,
JR~ Vol. 83
solvent gave 1.47 g. of a neutral fraction with an ultraviolet Anal. Calcd. for C2&00&~: C, 56.32; H, 6.17.
maximum at 237 mp. Distillation under diffusion pump Found: C, 56.26; H, 6.04.
vacuum gave a mobile oil at a temperature of 100-110"
which still showed AzzH 237 mp.
A mixture of 0.4 g. of the above oil, 0.4 g. of semicar-
The 2,4-dinitrophenylhydrazone was peculiar in that
when recrystallized from ethyl acetate-ethanol it formed
matted needles which slowly changed to hard compact
bazide hydrochloride, 1.65 ml. of pyridine and 8 ml. of prisms on standing in the solution.
ethanol was refluxed for 1.5 hours and allowed to stand over- Lithium-Liquid Ammonia Reduction of the Cyclopen-
night. Removal of the solvent under reduced pressure and tenone XX.-The cyclopentenone X X (307 mg.) was dis-
trituration with ether gave an ether extract which left a solved in 2 ml. of dry tetrahydrofuran and the solution
gum after evaporation. Crystallization was effected from added to 15 ml. of anhydrous liquid ammonia. With
benzene-cyclohexane t o give the semicarbazone of the vigorous stirring 15 to 20 mg. of lithium was added and the
cyclopentanone monoester XIX. After three recrystalliza- color changed to yellow, pink, blue and finally the solution
tions from ethyl acetate-cyclohexane the semicarbazone became colorless in about 10 seconds. Ammonium chloride
melted a t 174-176". (0.6 9.) was then added and, after stirring for a short time,
Anal. Calcd. for C&&T\T3: C, 61.41; H, 7.90; X, the ammonia was allowed to evaporate. The residue was
14.33. Found: C, 61.51; H,8.12; N,14.80. taken up in ether and water and the ether solution, after
It was eventually found that if the t-butoxide reaction washing several times with water, was dried over anhydrous
was carried out for a short time a t room temperature the sodium sulfate and evaporated. The residue, 271 mg., was
intermediate aldol could be isolated in high yield and this a yellowish oil. It was mostly dissolved in petroleum ether
could easiIy be dehydrated with acid catalysis to give the (b.p. 30-60') and chromatographed on Merck alumina.
cyclopentenone X X . The first eluates, using benzene-petroleum ether (1:l ) , were
Potassium (3.0 E . ) was dissolved with refluxing in 150 yellowish and contained some enone (as judged from the
ml. of dry t-butyl akohol. The solution was cooled-to room infrared spectra). Later fractions were colorless, however,
temperature and added t o a solution of 20.1 g. of the and these were combined (115 mg.) and converted to the 2 . 3 -
methyl ketone XVI (m.p. 55-60') in 150 ml. of dry t-butyl dinitrophenylhydrazne (134 mg., m.p. 115-125O). The
alcohol. The resulting deep red solution was left a t room crystals of the 2,4-dinitrophenylhydrazonewhich were di-
temperature for 10 minutes and then poured into a mixture morphic from ethyl acetate-ethanol, melted a t 152-154'
of ether and water The aqueous layer was separated and after three recrystallizations from this solvent mixture.
washed twice with ether. The ether extracts were washed The mixture melting point with the 2,4-dinitrophenylhydra-
repeatedly with sodium chloride solution until the aqueous zone prepared from the palladium-charcoal reduced keto-
layer was colorless and neutral. The ether solution was diester (also dimorphic crystals from ethyl acetate-ethanol;
dried over anhydrous sodium sulfate and the solvent was m.p. 152-154') was not depressed. It is concluded that the
removed on the steam-bath, finally under reduced pressure. more stable cis ring junction of the five-membered rings is
The crude aldol (18.6 6.) was a mobile yellow oil, the in- present in the keto-diester X X I I I a prepared by palladium-
frared spectrum of which possessed a strong hydroxyl band charcoal reduction of the cyclopentenone XX.
a t 2.8 p. Stereochemistry of Substituents in the Reduced Keto-
The aldol was dehydrated without purification by dis- diester XXII1a.-The stereochemical relationship of the
solving it with 2.0 g. of p-toluenesulfonic acid monohydrate secondary methyl, and of the secondary ester to the tertiary
in 400 ml. of benzene and refluxing for 1 hour. After cool- ester in the reduced keto-diester XXIIIa was deduced from
ing, the benzene solution was washed three times with potas- the following series of experiments. The keto-diester
sium carbonate solution and the aqueous solutions were XXIIIa (655 mg.) was hydrolyzed by dissolving it in 10 ml.
washed twice with ether. The organic layers were com- of 807, ethanol containing 2.0 g. of potassium hydroxide
bined, dried over anhydrous potassium carbonate and the and refluxing for 36 hours under nitrogen. The solution
solvent was removed under reduced pressure. Petroleum was poured into water and the aqueous solution was washed
ether (b.p. 30-60", ca. 30 nil.) was added to the oily residue twice with ether. The aqueous solution was then acidified
and, after seeding, the mixture was left in the refrigerator with concentrated hydrochloric acid and the precipitate was
for several days. The crystalline cyclopentenone XX was removed by ether extraction. The ether snlution was dried
collected to give 15.2 g., m.p. 46-58' (80y0 yield), over anhydrous sodium sulfate and evaporated to give a
228 mp, log e 4.06. gummy residue (528 mg.) which would not crystallize.
Crystalline cyclopentenone preparations melting over the The residue was therefore converted directly to the anhy-
range of 46-60" were obtained from a total of 32.1 g. of the dride by refluxing for 3 hours in 6 ml. of acetic anhydride.
methylketone and combined to give 23.6 g. These were The solvent was removed in a stream of nitrogen, finally
recrystallized from petroleum ether (b.p. 30-60") to give a on the steam-bath under reduced pressure. The brown
first crop of 11.4 g. of large crystals, m.p. 63-66', and a residue solidified and was sublimed in high vacuum a t 160-
second crop of 8.0 g., m.p. 48-58". Recrystallization of 170" to give a colorless sublimate of oily crystals. Tritu-
the first crop using charcoal gave 10.8 g. of colorless crystals, ration with ether gave 174 mg., m.p. 154-172'. One
m.p. 65-67'. The yield of this isomer was 35yc7,.A recrystallization from ether gave 117 mg. of the anhydride
sample was evaporatively distilled in high vacuum at 90" lor ascolorlessprisms,m.p. 171-173" (infrared: 5 . 5 5 a n d 5 . 6 6 ~ ) .
analysis. The anhydride was hydrolyzed by dissolving it in 6 ml.
Anal. Calcd. for C17Hza06: C, 66.21; €1, 7.85. Found: of 50% aqueous dioxane and refluxing the solution for 20
C, 66.06; H, 7.87. hours. Removal of the solvent, first under nitrogen and
Attempts to purify a second isomer of the cyclopeiiteiiorie finally on the steam-bath under reduced pressure, gave 3
S X by recrystallization from petroleum ether were unsuc- viscous oil. The oil was dissolved in ether and added to
cessful. an ethereal solution of diazoethane (prepared from 5 ml.
Reduction of the Cyclopentenone XX.-The cyclopen- of S-riitroso-N-ethyl carbamate) and the solution 'ivi~sleft
tenone X X (m.p. 85-67", 504 mg.) was dissolved in 16 1111. at rnom temperature for 1 hour. The solvent was removed
of absolute ethanol and hy3rogenated at atmospheric pressure under reduced pressure, then toluene was added and re-
using 136 mg. of 105% palladiuin-on-c1iarco:il catalyst. moved three times on the steam-bath under reduced pres-
The uptake of hydrogen was 38.7 ml. in 20 iniiiutes a t sure. The residue was taken up in petroleum ether (b.p.
25.4' (calculated uptdke: 40.0 i d . ) . The palladium- 30-60') and the solution was filtered and evaporated.
charcoal was collected on a mat of filter-aid and washed well The residual yellow oil was evaporatively distilled a t 100-
with alcohol. The solvent was removed on the steani-bath 110' in a high vacuum. The infrared spectrum of the dis-
under reduced pressure to give 486 mg. of a colorless oil. tillate (102 mg.) was the same as that of the starting keto-
The oil crystallized on cooling in ice and the solid was col- diester XXIIIa. The 2,4-dinitrophenylhydraznne was
lected and recrystallized once from petroleurn ether (b.1,. prepared, and after four recrystallizations from ethanol the
30-60') to give 325 mg. of colorless crystals of the keto di- melting point was constant a t 159-161'. The mixture melt-
ester X X I I I a , m.p. 33.5-35.0'. The oil obtained from the ing point with the 2,4-dinitrophenylhydrazone of the start-
mother liquors was converted to the 2,4-dinitrophenyl- ing ketodiester XXIIIa (m.p. 159-161' from ethanol) was
hydrazone. After two recrystallization: from ethanol the not depressed.
melting point was constant a t 160-161 . The same 2,4- Since the methyl group alpha t o the carbonyl mas not
dinitrophenylhydrazone (same melting point and mixture isomerized by the treatment with strong base, i t must al-
melting point) was obtained from the crystalline keto. ready be in the more stable configuration in the keto-di-
diester. ester XXIIIa.
July 20, 1961 AND TOTAL
STEREOCHEMISTRY SYNTHESIS
OF CEDROL 3123
dl-Norcedrenedicarboxylic Acid (VIb) .-This diacid The latter was prepared by adding solid VIb t o a n excess of
could be obtained from the keto-diester X X I I I a by Raney diazomethane in ether solution. From 10.00 g. of NCDA
nickel desulfurization of the corresponding thioketal fol- (m.p. 212-213') a yield of 10.59 g. of the dimethyl ester
lowed by alkaline hydrolysis. To form the thioketal the (b.p. 95-132' (1 mm.)) was obtained.
keto-diester X X I I I a (m.p. 31-34', 1.03 g.) was dissolved in For the hydrolysis, 10.5 g. of NCDA dimethyl ester was
10 ml. of chloroform and 1.0 ml. of ethane dithiol was added. dissolved in 35 ml. of absolute ethanol containing 2.4 g . of
The solution was then cooled, saturated with dry hydrogen potassium hydroxide and the solution was refluxed for 2
chloride and left a t 0" overnight. The solvent was then hours. The solution was cooled, poured into water and t h e
removed on the steam-bath under reduced pressure and aqueous solution washed three times with ether t o remove
toluene was added and removed in this manner three times. unchanged diester (ca. 2 g.). After acidification with hydro-
The residue was a colorless oil which crystallized on stand- chloric acid the h'CDA monomethyl ester X X X I I was re-
ing. The solid thioketal was collected and washed with moved by ether extraction. The ether extract was dried
petroleum ether (b.p. 30-60") to give 0.86 g. of colorless over anhydrous sodium sulfate and concentrated in an erlen-
crystals, m.p. 72-75'. After several recrystallizations meyer flask. As the volume became small, petroleum ether
from pentane the pure thioketal XXIX melted a t 75-76', was added slowly to the boiling solution to displace the
Anal. Calcd. for C18H3001SZ: C, 59.05; H , 7.83. ether. When crystals appeared the mixture was cooled and
Found: C, 59.23; H , 7.77. the colorless crystals were collected to give 6.20 g. of
For the desulfurization the thioketal (1.32 g., m.p. 72- X X X I I , m.p. 129-132' (reported'6 m.p. 130-131').
75") was dissolved in 125 ml. of absolute ethanol in which The methylketone methyl e,ster X X X I I I was prepared
was suspended Raney nickel prepared according t o Pavlic from NCDA monomethyl ester X X X I I via the acid chlo-
and Adkins2j from 50 g. of Raney nickel alloy. The mix- ride, diazoketone and chloroketone by a procedure similar to
ture was refluxed with stirring for 15 hours. After cooling, that used in the preparation of the methylketone XVI.
the Raney nickel was collected on filter aid and washed well The NCDA monomethyl ester X X X I I I (8.05 g.) was
with ethanol. The solvent was removed on the steam- converted to the acid chloride by dissolving it in 20 ml. of
bath under reduced pressure. The residue was taken up in oxalyl chloride. The solution was left a t room tempera-
ether and the solution was filtered and evaporated to give ture for 1.5 hours and then excess oxalyl chloride was re-
0.86 g. of the diester of dl-norcedrenedicarboxylic acid as a moved under reduced pressure. Dry benzene was then
colorless mobile oil. added and removed under reduced pressure three times.
The diester (0.82 g.) was hydrolyzed in 10 ml. of 9570 The acid chloride was dissolved in dry ether and the solu-
ethanol containing 2.0 g. of potassium hydroxide by re- tion poured into a solution of diazomethane in ether (pre-
fluxing the solution for 24 hours. After cooling, the mix- pared from 45 g. of N-nitroso-N-methylurea and dried over
ture was poured into water and the aqueous solution was potassium hydroxide pellets). The solution was left a t
washed well with ether and acidified with 10 ml. of concen- room temperature for 1 hour and then the excess diazo-
trated hydrochloric acid. The precipitated diacid was methane and the ether were removed on the steam-bath,
removed by ether extraction. The ether solution, after finally under reduced pressure.
drying over anhydrous sodium sulfate, was evaporated t o The crude diazoketone was dissolved in 50 ml. of dry
give 0.62 g. of dl-norcedrenedicarboxylic acid (VIb) a s a ether and the solution was saturated a t 0" with dry hydro-
colorless crystalline solid, m.p. 210-215". One recrystal- gen chloride. The solvent was then removed on the steam-
lization from ether-pentane gave 454 mg., m.p. 219-221'. bath under reduced pressure.
Further recrystallizations from ether-pentane did not raise The crude chloromethyl ketone was dissolved with 9 g.
the melting point above 221-223'. The infrared spectrum of potassium iodide in 100 ml. of 80% acetic acid, and 45 g.
was the same a s that of (-)-norcedrenedicarboxylic acid of powdered zinc was added slowly with stirring and cooling
from natural cedrene. in ice. The mixture was stirred overnight a t room tem-
perature. The solids were then collected and washed well
Anal. Calcd. for C13H2004: C, 64.98; H, 8.39. Found: with 80% acetic acid. The solution was evaporated t o
C, 65.02; H, 8.36. dryness on the steam-bath under reduced pressure and the
Resolution of dl-Norcedrenedicarboxylic Acid.-To a residue taken up in water and ether. The ether layer was
hot solution of 410 mg. (1.0 equiv.) of quinine in acetone washed several times with water, then with aqueous potas-
was added a solution of 300 mg. of dl-norcedrenedicarboxylic sium carbonate solution containing some sodium thiosulfate
acid in acetone. No precipitate appeared when the stop- and finally dried over anhydrous potassium carbonate and
pered flask was allowed t o stand a t room temperature for 24 evaporated to give the crude methylketone methyl ester
hours, but when the flask was left open to permit slow evap- XXXIII (7.73 g., 96% yield).
oration of the solvent colorless crystals formed and were The 2,4-dinitro~henvlhvdrazoneof the methyl ketone
collected after one day. This first crop, m.p. 204-206", methyl ester X X A I I I -was prepared: m.p. 14O-i42O from
weighed 156 mg. Concentration of the mother liquor and cyclohexane. The same 2,4-dinitrophenylhydrazone (m .p .
seeding with the first crop gave more crystals. There was 138-139', mixture melting point 139-141"), and hence the
obtained a F t a l of 252 mg. (71y0 of theory) which melted same ketone, was obtained previously by Stork and Breslow'
above 204 . One recrystallization from chloroform-ace- by the reaction with dimethylcadmium of the acid chloride
tone gave 210 mg., m.p. 209-210°, [ a ] Z i ~-123' ( c 1.04, prepared from X X X I I .
chloroform). One more recrystallization from chloroform-
acetone did not change the melting point or rotation and Anal. Calcd. for C21H2806N4: C, 58.32; H , 6.53.
there was no depression of the melting point of the mixture Found: C, 58.46; H , 6.63.
with the quinine salt of natural norcedrenedicarboxylic Cyclization to the @-Diketone XXXIV.-The crude
acid, m.p. 209-210", [ a I z 7-122'
~ (c 1.00, chloroform). methylketone methyl ester X X X I I I (7.73 g.) was dissolved
Anal. Calcd. for C ~ J H ~ , O ~ N C,~ : 70.18; H , 7.85. in a solution of 3.0 g. of potassium in 120 ml. of dry t-butyl
Found: C, 70.13; H, 8.18. alcohol and the solution was refluxed for 3 hours. The
To obtain the free acid, the quinine salt (121 mg.) was cooled solution was poured into water and the aqueous solu-
shaken for 15 minutes with 5 ml. of 6 N hydrochloric acid tion was washed four times with ether, then acidified
and the solid was collected and washed with 6 N hydro- (ice cooling) with concentrated hydrochloric acid and the
chloric acid. The free acid weighed 46 mg. (89% yield), colorless precipitate was removed by chloroform extraction.
m.p. 212-213". One recrystallization from ether-pentane Removal of the chloroform on the steam-bath under re-
did not change the melting point and the mixture melting duced pressure gave the @-diketoneXXXIV as a solid residue
point with natural norcedrenedicarboxylic acid (m.p. 213- which was triturated with ether and collected to give 6.49
214OI4) was not depressed. The optical rotation of the g. (83.5% yield), m.p. 199-201". One recrystallization
resolved diacid in acetone solution, [ C X ] * ~ D-38.9 =J= 1.5" from dioxane gave 4.90 g., m.p. 200-202'. Further re-
(c 1.08), is the same as the rotation of the natural ( - ) - crystallization from dioxane did not raise the melting point
norcedrenedicarboxylic acid, [ a ] 2 7-38.3
~ =k 1.0" (c 1.09), above 202-204'. The crystals were sublimed in high
in acetone.14 vacuum a t 160-170° for analysis.
Monomethyl ester of ( - )-norcedrenedicarboxylic acid Anal. Calcd. for ClrHnaOn: C, 76.32; H , 9.15. Found:
(XXXII) was prepared according t o Plattner, Kusserow C, 76.66; H , 9.07.
and K l a u P by the half hydrolysis of NCDA dimethyl ester. Preparation of the Secondary Alcohol XXXVII from the
P-Diketone XXXIV.-Direct reduction of the @-diketone
(25) A. A. Pavlic and H.Adkins, J . A m . Chcm. J o c . , ~1471
~ , (1846). XXXIV with lithium aluminum hydride gave a solid mix-
3124 GILBERTSTORK
AND FRANK
H. CLARKE,
JR. Vol. 83
ture of saturated and unsaturated alcohols from which the weaker hydroxyl band and only one carbonyl band at 6.1 M.
pure saturated alcohol XXXVII could be separated by One recrystallization from pentane gave 347 mg. of colorless
chromatography. The same saturated alcohol was also needles, m.p. 120-19lo. Chromatography of the residue
isolated when the mixture of enol ethers formed from the from the mother liquor gave more solid (282 mg., m.p.
8-diketone with ethyl alcohol and a trace of acid was re- l l O - l l G o ) on elution with benzene. Two recrystallizations
duced with lithium aluminum hydride. from petroleu? ether gave 161 mg. of colorless crystals,
( a ) Reduction of the @-Diketone XXXIV.-The 8-dike- m.p. 120-121 . This compound (total 508 mg.) was
tone X X X I V (m.D. 200-202", 2.00 n.) was added with stir- identical in all respects (infrared spectra, melting point
ring to an ice-coded suspension of 3.0 g. of lithium alumi- and mixture melting point) with the alcohol XXXVII
num hydride in 50 ml. of dry ether. The mixture was re- obtained by direct reduction of the P-diketone XXXIV
fluxed with stirring for 17 hours, then cooled and the excess and also by reduction of the saturated ketone X X X I .
lithium aluminum hydride was decomposed with saturated Other fractions obtained by chromatography of the mother
aqueous sodium sulfate solution. Anhydrous sodium sul- liquors included the fraction eluted first with petroleum
fate was added and the solids were collected and washed ether (infrared spectrum lacks carbonyl bands; ultra-
well with ether. Removal of the ether gave 1.91 g. of violet spectrum in ethanol had Xmnx 225 mp, log E 3.3;
colorless crystals, m.p. 95-110". Trituration of the decolorizes bromine in carbon tetrachloride) and a fraction
crystals with boiling pentane, then cooling and filtering eluted with a 1:l mixture of petroleum ether and benzene
gave colorless needles (1.27 g.), m.p. 114-116'. Further (infrared had sharp band a t 6.1 p ; ultraviolet in ethanol
recrystallizations from cyclohexane failed to raise the melt- had, , ,A 241 mp, log E 4.0).
ing point of the product above 117-118'. However, Preparation of the Cyclohexenone XXXV1.-The methyl-
chromatography on alumina of solids obtained from the ketone methyl ester X X X I I I was reduced to the correspond-
mother liquors resulted in the separation of the saturated ing diol which was then oxidized t o the ketoaldehyde, and
alcohol on elution with 1: 1 petroleum ether-benzene. the latter on treatment with aqueous base gave the cyclo-
After several recrystallizations from petroleum ether the hexenotie XXXVI. Intermediate compounds were not
saturated alcohol XXXVII melted at 120-121' and the purified.
melting point was not depressed on admixture with the The methylketoiie methyl ester X X X I I I (5.21 9.) was
alcohol of the same melting point obtained by reduction of dissolved in dry ether and the solution was added slowly t o
the enol ether mixture from X X X I V (see below). a stirred, ice-cooled suspension of 7.77 g. of lithium alumi-
Anal. Calcd. for Cl,H%O: C, 80.71; H, 11.61. Found: num hydride in 250 ml. of anhydrous ether. The mixture
C, 80.65; H, 11.58. was stirred under reflux for 2 hours and at room tempera-
The crude alcohol, m.p. 114-116", gave a mixture of ture overnight. Excess lithium aluminum hydride was
saturated and unsaturated ketone on oxidation with chromic decomposed by adding a saturated aqueous solution of
acid-pyridine complex (for procedure, see below). The sodium sulfate to the ice-cooled mixture. Anhydrous SO-
ultraviolet absorption spectrum in 95% ethanol had diuin sulfate was added and the solids were collected and
243 mp, log c 3.22, (estimated 10-15% unsaturated ketone). washed well with ether. Removal of the ether gave a color-
The saturated ketone X X X I was separated by chromatog- less foam (4.11 g.) which had a strong hydroxyl band in the
raphy on alumina by elution with petroleum ether. The infrared spectrum but lacked carbonyl bands.
2,4-dinitrophenylhydrazone was prepared and i t was A suspension of chromic acid-pyridine complex'* was
identical (melting point and mixture melting point 146- prepared by adding 8.43 g. of chromic anhydride to 84 ml.
147') with that of the ketone prepared by reduction of the of pyridine with cooling. T o this was added a solution ol
cyclohexenone XXXVI or from the pure alcohol XXXVII the crude diol (4.11 8.) in 75 rnl. of pyridine and the mixture
(see below). was left a t room temperature overnight (17 hours). The
(b) Reduction of the Enol Ether(s) of XXX1V.-The mixture was then poured into 750 ml. of ether and the
B-diketone XXXIV (2.84 g.) was dissolved in 50 ml. of solids were collected and washed well with 250 ml. of ether.
absolute ethanol and 0.10 g. of p-toluenesulfonic acid mono- The filtrate was concentrated on the steam-bath t o about
hydrate was added.20 The mixture was distilled during 4 half its volume and was then washed three times with water,
hours through a 24-cm. column of glass helices: 53 ml. of three times with 200 ml. of 5% aqueous potassium hy-
distillate was collected. The residue was dissolved in ether droxide, once with water, twice with dilute hydrochloric
and the solution was washed twice with aqueous sodium acid, and finally twice with water. The ether solution was
bicarbonate and once with water, then dried over anhydrous then dried over anhydrous magnesium sulfate and the sol-
sodium sulfate. The oil remaining after removal of the vent was removed to give 2.81 g. of a pale yellow oil. The
solvent was dissolved in petroleum ether and filtered to re- infrared spectrum had aldehyde and ketone bands.
move a few crystals of unchanged B-diketone (m.p. 198- For the cyclization a solution of 12.5 g. of potassium
200"). Removal of the solvent from the filtrate gave 2.82 hydroxide in 500 ml. of water was first boiled under reduced
g. of a mobile, nearly colorless oil (presumably a mixture of pressure for a short time to remove oxygen, The alkaline
both possible enol ethers X X X V and XXXVa). solution was then added to 2.50 g. of the crude aldehyde-
The crude mixture of enol ethers was dissolved in 25 ml. of ketone. The mixture was stirred under reflux at 90 mm.
dry ether and added over a 10-minute period to a stirred pressure (about 50') for 14.5 hours using a mercury trap.I8
suspension of 1.0 g. of lithium aluminum hydride in 150 The mixture was cooled and washed twice with ether. The
ml. of ether at 0'. The mixture was stirred a t room tem- ether solution was washed with water, dried over anhydrous
perature for 15 minutes, then under reflux for 15 minutes magnesium sulfate and evaporated to give 2.02 g. of the
and finally left standing overnight a t room temperature. crude cycloliexenone XXXVI as a mobile yellow oil. Most
In the morning, water was added, with stirring but without of the crude product (1.99 g.) was purified by chromatog-
external cooling, followed by 3 N sulfuric acid. The ether raphy on alumina. Eluates with petroleum ether, ben-
layer was separated, washed twice with dilute sulfuric acid, zene-petroleum ether and benzene which appeared to give
three times with water, dried over anhydrous potassium only the cyclohexenone carbonyl band a t 6.1 p in the in-
carbonate and evaporated. The residue (2.45 g.) was a frared spectra were combined to give a total of 1.47g. (39%
nearly colorless crystalline solid. The infrared spectrum over-all yield) of purified cyclohexenone XXXVI. The
showed a hydroxyl band a t 2.9 p and carbonyl bands at 5.9 product was evaporatively distilled at 140-1SO" at 1 mm.
and 6.1 p . I t seemed likely that the carbonyl band at 5.9 fi ,A,(, 243 m p , log e 3.91 in 95% ethanol). The 2,4-dinitro-
and the hydroxyl band corresponded t o a @-hydroxyketone phenylhydrazone was prepared and recrystallized from
which on dehydration should give a cyclohexenone with a methanol to give brilliant red prisms, m.p. 161-167".
carbonyl band at 6.1 p . Accordingly, the crude mixture Anal. Calcd. for C20H~O,N4: C, 632.48; H, 6.29.
was dissolved in 50 ml. of benzene containing 0.20 g. of p- Found: C, 62.19; H, 6.33.
toluenesulfonic acid monohydrate and the solution was re-
fluxed for 1.5 hours. The cooled solution was washed with Preparation of the Saturated Ketone XXXI. (a) Re-
aqueous sndium bicarbonate, aqueous sodium hydroxide and duction of the Enone XXXV1.-The cyclohexenone XXXVI
then several times with water. After drying over anhydrous (327 mg.) was reduced with 108 mg. of 10% palladium-on-
potassium carbonate, the solvent was removed under re- charcoal catalyst in 10 ml. of absolute ethanol. The hy-
duced pressure. The crystalline residue (2.18 g.) had a drogen uptake was 64.3 ml. in 20 minutes at 23" and atmos-
pheric pressure, There was no further uptake of hydrogen
(26) Cf. E. D. Rergmann and J. Szmuszkovicz, J . A m . C h e w after 1 hour. The calculated uptake is 63.8 ml. The
sac.,rti, 3226 ( 1 ~ 1 5 3 ) . catalyst was collected and washed with ether. Removal of
July 20, 1961 DISVBSTITUTED-1 -ABIINO-7-I~IJNO-CYCLOI-IEPTATRIEI\'ES 3125

solvent from the filtrate gave the saturated ketone XXXI further recrystallization from petroleum ether did not raise
(524 mg.) as il colorless oil. The 2,4-dinitrophcnylhydra- the melting point. The melting point was not depressed
zone was prepared and recrystallized from tnethanol to give on admixture with the saturated alcohol XXXVII obtained
a mat of fine yellow needles which changed to stout prisms from the pdiketone XXXIV and the infrared spectra ware
on standing in the soIution; m.p. 14G-147'. identical.
Preparation of Cedrol from the Ketone XXX1.-Methyl-
Anal. Calcd. for C20H2604Nl: C, 62.16; H, 6.78. lithium was prepared by adding 6 ml. of methyl iodide t o
Found: C, 62.20; H, 6.95. 1.8 g. of lithium (in 6 pieces) in 40 ml. of dry ether, with
(b) Oxidation of the Alcohol XXXVI1.-A solution of 219 stirring and cooling in ice. When the initial vigorous reac-
mg. of the saturated alcohol XXXVII (m.p. 120-121') in tion had subsided the mixture was refluxed with stirrin.: for
2.2 ml. of pyridine was added to a suspension of the chromic 4 hours. The mixture was then cooled and the unreacted
acid-pyridine complex prepared by adding 224 mg. of lithium (0.90 g.) was removed (calculated amount of methyl-
chromic anhydride to 2.2 ml. of pyridine. The mixture was lithium: 0.06 mole).
left a t room temperature for 23 hours and dry ether was then The ketone X X X I (446 mg., 0.0021 mole) in dry ether
added. The solids were collected and washed with dry was added slowly to the solution of methyllithium and the
ether. The filtrate was washed with dilute hydrochloric mixture was refluxed with stirring for 18.5 hours. Excess
acid and water, dried over anhydrous sodium sulfate and methyllithium was decomposed by adding aqueous sodium
gave, after removal of the solvent, 190 mg. of the ketone sulfate solution containing some sodium thiosulfate to the
X X X I as a yellowish oil. The 2,4-dinitrophenylhydrazone stirred, ice-cooled solution. Water was added and the
was prepared and recrystallized from methanol; m.p. 146- ether layer was separated and washed well with water, dried
147". The melting point was not depressed on admixture over anhydrous sodium sulfate and evaporated to give 480
with the 2,4-dinitrophenylhydrazone of the ketone obtained mg. of crude cedrol as a solid residue. The crude product
by reduction of the enone XXXVI. was recrystallized from methanol-water four times to give
Reduction of the Ketone XXX1.-The saturated ketone colorless needles: 1. obtained 380 mg., m.p. 80-81"; 2.
XXXI (164 mg.) was added in ether solution to a stirred obtained 340 mg., m.p. 83-87'; 3. obtained 284 mg., m.p.
suspension of 0.40 g. of lithium aluminum hydride in ether 85-87'; 4. obtained 215 mg., m.p. 86-87'.
(total volume 30 ml.). The mixture was stirred under re- Natural cedrol was also recrystallized from methanol-
flux for 17 hours and then cooled. Excess lithium aluminum water to give colorless needles, m.p. 86-87O.m The mixture
hydride was decomposed with an aqueous sodium sulfate melting point of natural and synthetic cedrol was 88-87"
solution and anhydrous sodium sulfate was added. The and the infrared spectra were identical. The rotations
solids were collected and washed with dry ether and the were measured in chloroform solution: natural cedrol:
filtrate was evaporated to give 140 mg. of a colorless solid, [ a ] z a ~ 9.9 f 0.4' (c 5.00); synthetic cedrol: [a]'*D 10.5 i:
m.p. 118-120'. One recrystallization from petroleum ether 0.8' ( G 5.00); reportedlo values include +9' 31' and + l o "
gave 96 mg. of colorless needles, m.p. 120-121'. One 30' in chloroform solution.

[CONTRIBUTION KO. 600 FROM THE CENTRAL

RESEARCH STATION, E. I. DU POXT DE NEMOURS
DEPARTMENT, EXPERIMENTAL
AND CO.,WILMINGTON 98. DEL.]

N,N'-Disubstituted-l-amino-7-imino-~,3,5-cycloheptatrienes,
a Non-classicalAromatic
System
BY W.R.BRASEN,H. E. HOLMQUIST
AND R.E. BENSON
RECEIVED
JANUARY 26, 1961

N.N'-Disubstituted-l-amino-7-imino-1,3,5-cycloheptatrienes are stable, highly colored compounds that are readily

accessible from the tetrafluorocycloheptadjenes derived from tetrafluoroethylene and cyclopentadiene. Chemical studies of
these aminoimines have established tha'f they have appreciable aromatic character. Although structurally related t o
tropolone, the reasons for aromaticity in these two systems appear t o be fundamentally different. Whereas aromaticity in
tropolone has been attributed t o resonance forms containing the sextet, n.m.r. and infrared spectral data together with dipole
moment studies of the aminoimines clearly rule out a major contribution from sextet forms. The evidence appears more
consistent for a peripheral 10 *-electron system involving hybridization of the two non-bonding electrons on nitrogen of the
NHR substituent with the 8 x-electrons of the double bonds.

A new synthesis of tropolone recently was re- compounds that is not conveniently accessible from
ported from these laboratories, based on the hy- tropolone or its derivatives. The disubstituted
drolysis of the tetrafluorocycloheptadienes accessible aminoimines are stable, highly colored compounds
from cyclopentadiene and tetrafIuoroethy1ene.I
0 ---t

+ I, R = H, alkyl, aryl
that exhibit aromatic reactivity similar to that of
p*0 tropolone i t ~ e l f . ~
It appears, however, that the
reasons for aromaticity in these two systems are
fundanientally different. The present paper de-
scribes additional studies of the aminoimines.
The previouslyreported examples of 1-arninc-7-im-
The tetrafluorocycloheptadienes are also interme- ino- 1,3,5-cycloheptatrienes are the compounds 11-IV.
diates for the synthesis of the 1-amino-7-imino-
(3) Excellent reviews of the chemistry of tropolone include (a)
1,3,5-cycloheptatrienes (I),2 a little-studied class of P . L. Pauson, Chcm. Reus., 66, 9 (1955); (b) T. Nozoe, Forfschr.
(1) J. J. Drysdale, W. W,Gilbert, H. K. Sinclair and W. H. Sharkey, Chcm. Org. Nafursfofc,13, 232 (1955); and ( c ) T. Nozoe in "Non-
J . A m . Chem. Soc., 80, 3672 (1958). Benzenoid Aromatic Compounds," D. Ginsburg, Editor, Intrtscience
(2) W. R. Brasen, H . E. Holmquist and R. E. Benson, $bid.. 82, Publishers, In:., New York, N . Y.,1959.
995 (1960). (4) R. E. Henson, J . A m . Chem. Soc.. 83, 5948 (1960).