- Email: [email protected]
Allyl and benzyl migrations in ortho -semibenzenes
TetrahedronLetters No. 21 pp 1691-1694, 19'15. Pergamon Press. Printed in Great Rritain.
ALLYL MD
BENZYL MIGRATIOIJS IN OPTRO-SIEIBENZEVES
By Bernard Miller* and Mohammed Resa Saidi Departmentof Chemistry,Universityof Massachusetts,Amherst, !iass. 01002
(Receivedin USA 2 February 1975; received in UK for publication4 April 1975) It was recently shown that ally1 groups in para-semibenzenes(~-sllyl-l-methylenecyclohexadienes)such as 1 readily undergo migration to the exocyclicmethylene group by free
/Q\ CHZ
fl
R
radical chain
1 processes.
Aromatic products, in principle,could also be obtained from 1 by
rearrangementvia two successiveCope migrations. The correspondingcyclohexadienones rearrange exclusivelyby Cope migrationsat 30-100°.2 However, no evidence for such migrationscould be observed in the reactionsof para-semibensenes. It was thereforeof interestto examine the thermal rearrangementsof
in which Cope rearrangementswould give aromatic
products,to see if concertedor free radical paths are favored. Reaction of methyl lithium with naphthalenonesa-c gave a single alcohol in each case. Addition of the nmr shift reagent Eu(fod)33resulted In larger shifts
in the positionsof the
allylic or benzylic methylene signals than in the peaks for the methyls at C-l, suggesting that, as expected,the lithium reagent had added mt.0
the larger group in 2.
Reaction
of methyl lithium with naphthalenonesza-c gave a mixture of isomericalcohols in each case. Again,
use of the shift reagent suggestedthat the z isomers,resultingfrom attack of methyl
lithium at the less hindered sides of the ketones,were predominant. Despite carefulmonitoringof the reactionby nmr, dehydrationof dienol ia with phosphorus
1691
1692
IJ,
No. 21
R = CH3
oxychloridein pyrldine at 3Oe, or thionyl chloride in pyridine at O", gave no evidence for formationof the desired semibenzeneza. Instead, a single aromaticproduct was obtained and was assigned structure&a. Dehydrationof dienol Lb gave (in additionto 10% of the cleavage product 1,2_dimethylnaphthalene) the correspondingrearrangementproduct gb. Dienols xa aud 3
reacted very slowly with phosphorusoxychloridein pyridine at temper-
atures below 50°. At i'O",dehydrationof Ia gave the two rearrangementproductsga and ga In
@
~
sR
3a,R=H b,R=CH3
7a,
P
(R=H)_6a+
ma
/
/
/
a
an
ao 8 to 1 ratio. Dehydrationof p
under the same conditionsgave 1,2-dimethylnaphthalene,
&b, b,and 2 in the yields shown below.
09 R3 CH3
1:;
&
(32%)
+
-8b (22%)
”
2
(8%)
The locationsof the substituentsin each of the reactionproductswere establishedfrom their nmr spectra. In each product signals for methylene or methyl groups in a-positions (deshieldedby the p&hydrogen8
) were located s.
0.2 ppm downfieldfrom sigusls for similar
substituentsln B-positions. Formationof hydrocarbons&a and gb from &a and 4-bcau only be explainedby postulating the transientformation
of
the semibenzenesza and _$b,vhich rearrangetoo rapidly for detec-
tion. The reglospecificrearrangementof !$ clearly proceedsby a [3,3] sigmatropicshift, rather than a free radical path similarto that observed in the rearrangementof &.
Hydrocarbons
1693
No. 21
ga and gb must sinilarly be formed by Cope rearrangements Formation of ga and kb from dehydration
of the appropriate semibensenes. 4
of Ia and lb can best be explained by Wagner-
Meerwein shifts of ally1 and crotyl groups in the a-naphthyl carbonium ions formed from 1. more stable B-naphthyl carbonium ions formed from the rearrangement and sb.
The
then yield semibenzenes za
The much higher yield of rearranged product from the crotyl derivative
is in accord
with this hypothesis. Hydrocarbon 2 can also arise &
a carbonium ion shift, as is shown by the fact that it is
the only rearrangement product formed, together with 1,2_dimethylnaphthalene, Lb with sulfuric acid in acetic acid or with aqueous hydrochloric 7_a similarly gives k-allyl-1,2-dimethylnaphthalene dehydration of the corresponding carbonium ion mechanism, The lack of x
(g).5
acid.
from reaction of
Under these conditions
Formation of 2 but not of 10 from
slcohols in pyridine is consistent with formation of 2 by a
in which a crotyl group is a much better migrator than an ally1 group.
formation of.10 - during dehydration of xa in pyridine suggests that 2 is formed
only via the carbonium ion path, and not from a Cope rearrangement of the semibenzene, since there seems no reason for a crotyl, but not an allyl, group to migrate to C-4 rather than the exocyclic methylene of a semibenzene. The very rapid Cope rearrangements
of ally1 groups in the semibensenes reported above
suggested that benzyl groups in m-semibenzenes
might undergo the corresponding
12 was prepared by dehydration of ic. To see if this were so, semibenzene -
[3,3] shifts.
In contrast to
ally1 substituted semibenzenes, 12 was easily isolated as a yellow oil which resisted rearrange-
2C
4c
ment at temperatures below 120°.
12
x-2
At 165" in diglyme solution, 12 rearranged to give 13,
together with a small amount of 1,2-dimethylnaphthelene. Dehydration of dienol lc gave an inseparable mixture which appeared to consist of 12 and 14 -*
Rearrangement
and 2,
of this mixture at 165' for 1G hours gave approximately equal amounts of 13
again accompanied by a small yield of 1,2-dimethylnaphthalene.
No evidence for the
Occurence of any [3,3] migration could be observed. The formation of [1,3] migration products from 12 and &presumably
proceeds via free
6
1694
No. 21
radical chains similar to those responsiblefor [1,5] benzyl migrationsin E-semibenzenes.1 Acknarledgment: We thank the donors of the PetroleumResearchFund, administeredby the American ChemicalSociety, for a grant in support of this work. References 1) B. Miller and K.-H. Iai, J. Amer. Chem. Sot., 94, 3472 (1972). 2) R. Barner, A. Boiler, J. Borgulya,E.G. Herzog, W. von Phillipsborn,C. von Plants, A. F&St, and H. Schmid, Helv. Chim.B,
9,
94 (1965).
R.E. Rondeau and R.E. Sievers, 3) Fod=6,6,7,7,8,8,8,-heptafluoro-2,2-dimethyl-3,5-octanedione: J. Amer. Chem. Sot., z,
1523 (1971).
similarlyappear to migrate v& *I) Propargylgroups in ortho-semibenzenes
[3,3] sigmatropic
shifts: P. Gilgen, J. Zsindely,and H. Schmid, -_--* Relv. Chim. Acta -9 56 1457 (1973). 5) For examples of [3,3] migrationsin cyclohexadienylcarboniumions, see H.-J. Hansen and H. S&mid, --.------* Helv. Chim. Acta .-* 51 828 (1968);B. Miller and K.-H. Lai, Ts+hedron, 21, 2221 (1972);-_ J. Org. Chem., 37, 2505 (1972). to e-allyltoluene 6) The "all carbon Claisen"rearrangementof 3-butenylbenzene
at 350' in
the presence of potassiumt_butoxidehas been reported: W. von E. Doering and R.A. Bragole, Tetrahedron,22, 385 (1966).
ALLYL MD
BENZYL MIGRATIOIJS IN OPTRO-SIEIBENZEVES
By Bernard Miller* and Mohammed Resa Saidi Departmentof Chemistry,Universityof Massachusetts,Amherst, !iass. 01002
(Receivedin USA 2 February 1975; received in UK for publication4 April 1975) It was recently shown that ally1 groups in para-semibenzenes(~-sllyl-l-methylenecyclohexadienes)such as 1 readily undergo migration to the exocyclicmethylene group by free
/Q\ CHZ
fl
R
radical chain
1 processes.
Aromatic products, in principle,could also be obtained from 1 by
rearrangementvia two successiveCope migrations. The correspondingcyclohexadienones rearrange exclusivelyby Cope migrationsat 30-100°.2 However, no evidence for such migrationscould be observed in the reactionsof para-semibensenes. It was thereforeof interestto examine the thermal rearrangementsof
in which Cope rearrangementswould give aromatic
products,to see if concertedor free radical paths are favored. Reaction of methyl lithium with naphthalenonesa-c gave a single alcohol in each case. Addition of the nmr shift reagent Eu(fod)33resulted In larger shifts
in the positionsof the
allylic or benzylic methylene signals than in the peaks for the methyls at C-l, suggesting that, as expected,the lithium reagent had added mt.0
the larger group in 2.
Reaction
of methyl lithium with naphthalenonesza-c gave a mixture of isomericalcohols in each case. Again,
use of the shift reagent suggestedthat the z isomers,resultingfrom attack of methyl
lithium at the less hindered sides of the ketones,were predominant. Despite carefulmonitoringof the reactionby nmr, dehydrationof dienol ia with phosphorus
1691
1692
IJ,
No. 21
R = CH3
oxychloridein pyrldine at 3Oe, or thionyl chloride in pyridine at O", gave no evidence for formationof the desired semibenzeneza. Instead, a single aromaticproduct was obtained and was assigned structure&a. Dehydrationof dienol Lb gave (in additionto 10% of the cleavage product 1,2_dimethylnaphthalene) the correspondingrearrangementproduct gb. Dienols xa aud 3
reacted very slowly with phosphorusoxychloridein pyridine at temper-
atures below 50°. At i'O",dehydrationof Ia gave the two rearrangementproductsga and ga In
@
~
sR
3a,R=H b,R=CH3
7a,
P
(R=H)_6a+
ma
/
/
/
a
an
ao 8 to 1 ratio. Dehydrationof p
under the same conditionsgave 1,2-dimethylnaphthalene,
&b, b,and 2 in the yields shown below.
09 R3 CH3
1:;
&
(32%)
+
-8b (22%)
”
2
(8%)
The locationsof the substituentsin each of the reactionproductswere establishedfrom their nmr spectra. In each product signals for methylene or methyl groups in a-positions (deshieldedby the p&hydrogen8
) were located s.
0.2 ppm downfieldfrom sigusls for similar
substituentsln B-positions. Formationof hydrocarbons&a and gb from &a and 4-bcau only be explainedby postulating the transientformation
of
the semibenzenesza and _$b,vhich rearrangetoo rapidly for detec-
tion. The reglospecificrearrangementof !$ clearly proceedsby a [3,3] sigmatropicshift, rather than a free radical path similarto that observed in the rearrangementof &.
Hydrocarbons
1693
No. 21
ga and gb must sinilarly be formed by Cope rearrangements Formation of ga and kb from dehydration
of the appropriate semibensenes. 4
of Ia and lb can best be explained by Wagner-
Meerwein shifts of ally1 and crotyl groups in the a-naphthyl carbonium ions formed from 1. more stable B-naphthyl carbonium ions formed from the rearrangement and sb.
The
then yield semibenzenes za
The much higher yield of rearranged product from the crotyl derivative
is in accord
with this hypothesis. Hydrocarbon 2 can also arise &
a carbonium ion shift, as is shown by the fact that it is
the only rearrangement product formed, together with 1,2_dimethylnaphthalene, Lb with sulfuric acid in acetic acid or with aqueous hydrochloric 7_a similarly gives k-allyl-1,2-dimethylnaphthalene dehydration of the corresponding carbonium ion mechanism, The lack of x
(g).5
acid.
from reaction of
Under these conditions
Formation of 2 but not of 10 from
slcohols in pyridine is consistent with formation of 2 by a
in which a crotyl group is a much better migrator than an ally1 group.
formation of.10 - during dehydration of xa in pyridine suggests that 2 is formed
only via the carbonium ion path, and not from a Cope rearrangement of the semibenzene, since there seems no reason for a crotyl, but not an allyl, group to migrate to C-4 rather than the exocyclic methylene of a semibenzene. The very rapid Cope rearrangements
of ally1 groups in the semibensenes reported above
suggested that benzyl groups in m-semibenzenes
might undergo the corresponding
12 was prepared by dehydration of ic. To see if this were so, semibenzene -
[3,3] shifts.
In contrast to
ally1 substituted semibenzenes, 12 was easily isolated as a yellow oil which resisted rearrange-
2C
4c
ment at temperatures below 120°.
12
x-2
At 165" in diglyme solution, 12 rearranged to give 13,
together with a small amount of 1,2-dimethylnaphthelene. Dehydration of dienol lc gave an inseparable mixture which appeared to consist of 12 and 14 -*
Rearrangement
and 2,
of this mixture at 165' for 1G hours gave approximately equal amounts of 13
again accompanied by a small yield of 1,2-dimethylnaphthalene.
No evidence for the
Occurence of any [3,3] migration could be observed. The formation of [1,3] migration products from 12 and &presumably
proceeds via free
6
1694
No. 21
radical chains similar to those responsiblefor [1,5] benzyl migrationsin E-semibenzenes.1 Acknarledgment: We thank the donors of the PetroleumResearchFund, administeredby the American ChemicalSociety, for a grant in support of this work. References 1) B. Miller and K.-H. Iai, J. Amer. Chem. Sot., 94, 3472 (1972). 2) R. Barner, A. Boiler, J. Borgulya,E.G. Herzog, W. von Phillipsborn,C. von Plants, A. F&St, and H. Schmid, Helv. Chim.B,
9,
94 (1965).
R.E. Rondeau and R.E. Sievers, 3) Fod=6,6,7,7,8,8,8,-heptafluoro-2,2-dimethyl-3,5-octanedione: J. Amer. Chem. Sot., z,
1523 (1971).
similarlyappear to migrate v& *I) Propargylgroups in ortho-semibenzenes
[3,3] sigmatropic
shifts: P. Gilgen, J. Zsindely,and H. Schmid, -_--* Relv. Chim. Acta -9 56 1457 (1973). 5) For examples of [3,3] migrationsin cyclohexadienylcarboniumions, see H.-J. Hansen and H. S&mid, --.------* Helv. Chim. Acta .-* 51 828 (1968);B. Miller and K.-H. Lai, Ts+hedron, 21, 2221 (1972);-_ J. Org. Chem., 37, 2505 (1972). to e-allyltoluene 6) The "all carbon Claisen"rearrangementof 3-butenylbenzene
at 350' in
the presence of potassiumt_butoxidehas been reported: W. von E. Doering and R.A. Bragole, Tetrahedron,22, 385 (1966).