The first synthesis of optically active Δ2-isoxazolines

The first synthesis of optically active Δ2-isoxazolines

TetrahedronLetters No. 34, pp. 2993-2996,1970. PergamonPress. Printed in Great Britain. THE FIRST SYNTHESIS OF OPTICALLY ACTIVE R. J. MacConaill A2-...

212KB Sizes 2 Downloads 27 Views

TetrahedronLetters No. 34, pp. 2993-2996,1970. PergamonPress. Printed in Great Britain.

THE FIRST SYNTHESIS OF OPTICALLY ACTIVE R. J. MacConaill

A2-ISON?GOLINES

and F. L. Scott,

Chemistry Department,

University

College,

Cork, Ireland.

(Received in UK 29 May 1970; acceptedfor publication18 June 1970) We

have established

cyclisation

using a variety of methods' that the base-induced

of Mannich base oxime methiodides

to yield A'-isoxazolines

involves oximate anion as a powerful neighbouring of this process

should then involve preservation

at the reaction

terminus.

of oxime anchimerism

species which have not been previously

of overall optical activity

R R' Ph-C-&H-&&R" II N

R 7' Ph-C-CH-CH-NR"

AH

R' = Me, NR"

= piperidino.

b. R = H,

R' = Me, NR"

= dimethylamino.

c. R = H,

R' = Ph, I&“

= dimethylamino.

d. R = Me, R' = H,

NR"

= dimethylamino.

e. R = Ph, R' = H,

NR"

= dimethylamino.

usual oximation

(Ia), m.p. 153-154',

[lit.,' m-p. 147-143',

was prepared by the method of Andrisano via the (-) dibenzoyl

[a] -18.5'

gave the Mannich base oxime

(c = 3.7 in DMF).

in ethanol)].

This salt when sub.ject to our

which with methyl iodide gave the (-) methiodide [a] g&I -6.6'

(c=5

[a] Egg -23'

et a1.,2 resolution being achieved

tartarate derivative.

procedure3

.MeI

III

a. R = H,

The (-) Mannich base hydrochloride in water),

-

6H II

I

consequence

active A2-isoxazolines

reported.

!I

(c=2.8

The stereochemistry

We have now used this stereochemical

to prepare several optically

R R' Ph-CO-&H-&H-NR".HCl

group.

Treatment

2993

(IIa)4, m.p. 171-173O

(IIIa), m.p. 200-200.5',

of this methiodide

with sodium

No.34

2994

ethoxide

(1.5 eq.) in ethanol for 3 hr. at 50' gave, after dry column

chromatography

on alumina

methyl-3-phenyl

(activity III) with methylene

isoxazoline

(IVa) (85$), m-p. 58.5',

chloride,

(+)-5-

[a] gE9 + 125.5'

(c = 1 in CHCls). Similarly

from the workup of the diastereoisomeric

the (+) Mannich base hydrochloride

(c = 5 in water) was obtained and from it on oximation 175-176'.

This gave the (+) methiodide

in CHCls). one oxime

isoxazoline

[a] ggg + 5.40

sodium ethoxide gave (-)Eg9 -1160

(c = 7.0

product -anti-1-phenylbutan-2-en-l-

(Vap (5-10s) m-p. 88.5' was obtained

from the cyclisation

reaction.

Ph-C-CR-Cm'

Ph-C-CHR

1

!

the oxime (IIa),m-p.

(IVa) (88%) m-p. 57.5',[u]

In both cases the elimination

tartarate

[a] gg9 + 16.0'

(IIIa), m-p. 193-194'.

(c = 1.7 in DME'), which when reacted with ethanolic 5-methyl-3-phenyl

dibenzoyl

(Ia), m.p. 164-165',

,

$

\/*

;H V

IV

a. R = H,

R' = Me.

b. R = H,

R' = Ph.

c. R = Me, R' = H. d. R .= Ph, R' = H.

These stereochemical alterations.

transformations

[as in (Ia)] to dimethylamino corresponding tartaric acid.

were quite sensitive to structural

Thus on changing the Mannich base amino moiety from piperidino

oxime

(Ib) neither the (Ib) free base nor the

(IIb) could be successfully

However,

resolved using (-) dibenzoyl

the racemic oxime methiodide

base to give the A* -isoxazoline

(IV) in 74% yield.

(IIIb) did cyclise in

Again when the J-methyl

function in (Ia) was replaced by a 3-phenyl group as in (Ic) attempted resolution

of the corresponding

lead to elimination (86%).

free base with

of dimethylamine

Again, the racemic methiodide

(-) dibenzoqrl tartaric acid

with formation of benzylideneacetophenone (111~) could be cyclised

to give a high

2995

No.34

yield of 3,5-diphenyl-A2-isoxazoline

(IVb)(> 91%).

We next examined a number of g-substituted Mannich base hydrochloride

(Id),prepared by literature

our usual oximation procedure a liquid.

Mannich bases.

a low yield

This formed the methiodide

methods',

(IIId), m-p. 223-224',

(c = 9.0 in DMF), which when reacted with sodium ethoxide,

isoxazoline

(+)

gave by

(13.5%) of the oxime (IId) as

ethanol for 20 hr. at 50°, gave on chromatographic phenyl

me

(IVc) (26$), m.p. 71-S-72',

workup

[a] $gg +5.15'

(10 eq.) in (+)-4-methyl-3-

[a] Egg + 100.00 (c = 2.2 in

CHCls) and eyn-2-methyl-l-phenylprop-2-en-l-one

oxime (Vc) (67%). m.p.

112-113'.

In contrast with the 3-phenyl compound when the inactive 2-phenyl

methiodide

(IIIe) was subject to our cyclisation

conditions

isolrazoline, forming only eyn-1,2-diphenylprop-2-en-l-one

it formed no

oxime (Vd) (97$),

m-p. 133-134'. The predominance of the 2-substituted in an unfavourable

of elimination compounds

ment of the oximes

whereas

(IIId and e) arose because

(~JIJ to 1-phenyl) configuration

Beckmann rearrangements,

phenyl migration,

substituted propionanilides

similar reactions with

of the type (CeHsCOWHCHsCHsWRs)

Thus the oximes ion and the oxime its methiodide

corresponding

configuration.

to give the isoxazoline

undergo a ralid E-anti

and anti isomers which we

(IVc) and the s

ketoxime

base catalysed

isomer undergoes

(Vc). production

isomerization

Such an isomerization

isomerization

of some of

followed by

has been reported p-amino aldoximes

in alkali.

sequence involving the production

of optically active isoxazoline

configurat-

The oxime (IId) and

by Ginsburg and Wilson7 who found that certain quaternary

The following

(IId) products

(25$).

While it is most likely that the anti- isomer

to give the m-vinyl

cannot be ruled out.

to

of alkyl migration

505%. and with

(IIId) appear to be a mixture of s

the isoxazoline by a w-anti cyclisation

cyclisation.

Such rearrange-

(IIa), (IIb) and (11~) have the --anti-1-phenyl

have not been able to separate.

only elimination

this.

(25%) and alkyl migration

(IIe) the syn-1-phenyl

cyclises exclusively

the oxime group was

in ca. 70% yields,

(IIe) gave the product

derived from both phenyl migration

in the case

for anchimeric

using PC15 in ether, confirmed

(IIa), (IIb) and (11~) gave products namely

namely benzamides

reactions over cyclisation

from a E-precursor

of a small quantity is thus explicable

anti-isomer either on the basis of a small a;nount (< 5%) of --

being present

2996

No.44

or of a small amount of E-anti cyclisation

procedure.

isomerism taking place during the

Thus the E

Mannich base oxime (IId) prepared

by the above method may be resolved with

(+)-dibenzoyltartaric

give the (-) dibenzoyl tartarate m.p. 147.5-149' z9

-51.05'

(from methanol),

[z]

(c = 10.5 in DMF), the free base of the oxime being released

with dilute ammonia and the oxime methylated 232-232.5'

acid to

(from ethanol),

of this methiodide

[a] Eg9 -1.140

(c = 8.3 in Dm).

43O, and reduced activity

m-p.

Treatment

(with sodium ethoxide in a similar manner

gave on workup the isoxazoline

Satisfactory

to give a methiodide

to IIId)

(IVc) both in small yield (3.0$), m.p. 41.5-

[a] gg9 + 39.2' (c L= 2.1 in CHCls).

microanalytical,

compounds have been obtained.

i.r., and n.m.r. data for the above

Rotations

were measured on a Perkin-Elmer

and we wish to thank Dr. V. Chester of A. I. Guinness

141 polarimeter

Sons (St. James Gate, Dublin)

and

for the use of this instrument.

References 1.

(a) F. L. Scott, J. C. Riordan and A. F. Hegarty, Tetrahedron

Letters,

537 (1963): (b) F. L. Scott and R. J. MacConaill,

Letters,

Tetrahedron

3685 (1967). 2.

R. Andrisano,

A. S. Angeloni,

P. de,Baria and M. Tramontini,

J. Chem.

Sot., (C), 2307 (1967). 3.

F. L. Scott, R. J. MacConaill

and J. C. Riordan, J. Chem. Sot.,

(Cl,

44 (1967). 4.

The Mannich base oxime hydrochloride racemic.

The material remaining

deposited

in solution

as the free oxime with dilute sodium hydroxide 5.

The configurations

and anti to describe

the position

active.

mentioned have

and by n.m.r. studies.

relative to the l-phenyl group in the compounds 6.

and was optically

of this and other vinyl ketoximes

been deduced by Beckmann rearrangement the terms u

from solution was

(19%) was liberated

We use

of the oxime -OH (II-V).

A. Pohland, L. R. Peters and H. R. Sullivan, J. Orq. Chem., 22, 2483 (1963).

7.

S. Ginsburg and I. B. Wilson, J. Amer. Chem. Sot., 22,

481 (1957).