Chiral poisoning in the kinetic resolution of allylic alcohols

Tetrahedron Letters, Vol. 34, No. 46. pp. 1359-1362,1993 Printed in Great Britain

Chid

Poisoning

GWI-4039193 $6.00 + .Ml

Pcrgamon Press Lid

in the Kinetic Resolution

of Allylic

Alcohols

J.W. Faller* and Makotu Tokunaga

Department of Chemistry,

Abstract:

Recently,

pure phosphines, These

alcohols. deactivating racemic

hydrogenations

of allylic alcohols

such as (R)-BINAP kinetic

resolutions

one enantiomer

were reported

using ruthenium

catalysts

with enantiomerically

to be useful for the kinetic resolution

can also be effected

of the catalyst

(BINAP)-Ru~l~(dm~~

Yale University, New Haven, CT 06511 USA

using

racemic

with an enantiomerically

with (lR,Z~)-ephedrine

provided

BINAP

pure chiral

of cyclic allylic

and preferentially

poison.

(R)-Z-cyclohexenol

Poisoning

of

in 93% ee at 72%

conversion.

Noyori and his coworkers’ with enantiomerically cyclic allylic alcohols.

have shown that hydrogenation

pure phosphines,

of allylic alcohols using ruthenium catalysts

such as (R)-BINAP have proven effective

Since enantiomerically

for kinetic resolution

pure BINAP, 2,2’-bis(diphenylphosphino)-1-l’-binapthyl,

of is

OH

9H H2

(R)-Ru(BINAP)

I

+

-0

4

obtained via a classical resolution,’ the separation is time consuming and results in a rather high cost for a pure en~tiomer.

We have adopted a different aproach in which a chit&

a racemic catalyst and so give high enantioselectivity

catalyst costs in situations where the cost of the ligands is high. of in situ resolutions,4’

where the deactivation

our findings that ephedrine,

poison

in asymmetric

can deactivate one en~tiomer

reduction.3

There have also been several other reports

of one hand of a catalyst has been proposed.

a relatively inexpensive

of

This sharply reduces the

and readily available enantiomerically

can effectively

function to poison a ruthenium catalyst containing racemic BINAP.

complemental

to Sharpless epoxidation.

We report here

pure amino alcohol,

This provides a method

The Sharpless procedure is effective for flexible acyclic substrates,

but does not always work well with cis olefins, particularly cyclic allylic alcohols.‘j’* Using a racemic BINAP-ruthenium to prepare (R)-2-cyclohexenol loading and have observed

catalyst with (-)-( lR,2S)-ephedrine

in >95% ee. 7040%

conversion

For ease in comparison in one hour.

over moderate times.

7359

as a poison we have been able

of results we have used 0.3% catalyst

There appears to be no degradation

in catalyst

7360

The original studies of kinetic resolutions [(S)-BINAP]Ru(acetate),.’

In situ formation

popular since it eliminates

the need to synthesize

are summarized

determined

by conversion

and handle the air-sensitive

1 complex’2,‘4 for convenience

racemic (BINAP)-RuCl,(dmf),, conditions

of cyclic allylic alcohols by asymmetric

of (R)- or (S)-BINAP-ruthenium

in Table

1.15 The enantiomeric

complexes,

hydrogenation

used

has become more

acetate. 9-‘4 We have used the

in handling.

Our results for a number of

purity of the remaining

cycloalkenol

was

to the MTPA ester.16

OH

HZ racemic Ru(BINAP)

OH

(-)-(1 R,SS)-ephedrine

Table 1.

Kinetic

Resolution

Poison

none, (R)-BINAPRub,” none, (*)BINAP

of 2-cyclohexenol

by Chill

Poisoning

of a Racemic

Ru(BJNAP)

rxn time

convn

ee

min

%

%

60 60

>95 0

Eq poison

H* atm

0 0

11 10

15 3

config

Catalyst

k&”

(S)

>15 ___

(lR,2S)-Ephedrine

10

2

720

66

53

(R)

2.8

(lR,2S)-Ephedrine

10

10

10

28

15

(R)

2.6

(lR,2S)-Ephedrine

10

10

40

57

58

(R)

4.4

(lR,2S)-Ephedrine

10

10

55

72

93

(R)

6.4

(lR,2S)-Ephedrine

10

10

60

77

>95

(R)

>6.4

(lS,2R)-Ephedrine

10

10

60

79

>95

0)

>6.4

( lR,2S)-Ephedrined

30

10

120

40

28

(R)

3.2

(lR,2S)-Ephedrine

10

100

7

57

44

(R)

3.0

(lR,2S)-N-methylephedrine

30

10

3

50

38

(R)

3.1

5

10

130

44

7

(S)

1.3

10

10

60

49

20

(S)

1.8

Quinine

5

10

3

68

8

(R)

1.2

Quinidine

5

10

3

73

5

(R)

1.1

(lS,2R)-Norephedrine (lS,2S)-Pseudoephedrine

The solvent ruthenium

was catalyst

2:

I

CH,CI,/hkOH

with a substrate

two hours prior to pressurizing

b)

The value of k/k,,

c)

First entry is for 100% ee BINAP-Ru

d)

The’substrate

if the reactions

to catalyst

to catalyst

were first order in substrate. catalyst;

ratio of 300.

with hydrogen.

the other entries

The amine was mixed with the solution

The reactions

were conducted

of

at 21 f 2%.

This is not the case for certain conditions. used 0% ee BINAP-Ru

catalyst.

ratio for this run was 900.

The first entry in Table 1 shows that the analogue of 1 prepared from pure (R)-BINAP is quite effective the kinetic resolution of 2-cyclohexenol.

Since the remaining (S)-2cyclohexenol

for

is obtained in >95% ee, this

7361

indicates

that the (R)-BINAP-Ru

hexenol.

When the reaction was carried out with racemic 1, the observed

mutual kinetic resolution;

catalyst

hydrogenates

(R)-2-cyclohexenol

i.e., in this case both enantiomers

much faster than (S)-2-cyclorate was much faster owing to

of the catalyst have the favored enantiomer

of the

substrate available to them. Table 1 summarizes

the effect of the number of equivalents

on the enantiomeric

purity of the remaining

Addition

noticeably

of amine

deactivates catalyst.

both the

slows

2-cyclohexenol

the reaction.

(R)- and (S)-BINAP-Ru

deactivating

the (R)-BINAP-Ru

is obtained component

initial rate with the enantiomerically faster than that poisoned

of the racemic catalyst.

of the ratio of the rate constants simple equation relates k,J/$,,,

with and partially with one hand of the

was the most effective.

the poison

This was confirmed

is selectively

by noting that the

catalyst poisoned with (lS,2R)-ephedrine

is 1.9 times

studies are normally analyzes by assuming that the reactions

allows the determination

of the substrate.

A relatively

purity. These ratios are now generally reported in kinetic resolution

values increase

of the kinetic resolution. with % conversion

for the studies with 10 eq of (lR,2S)-

and 10 atm H,, which indicates that the reaction is not first order in substrate. Preliminary

measurements

of the (Rj-BINAP complex poisoned with (lS,2R)-ephedrine

first order in substrate, but fractional or zero order. for estimation effectiveness

of % ee as a function of the kinetic resolution

more complete

are first

to % ee and % conversion. 6 From this ratio one can predict the % conversion

studies and give a measure of the effectiveness &&,,,,

of % ee at a particular % conversion

for the faster and slower reacting enantiomers

required to achieve a desired enantiomeric The calculated

to cyclohexanol.

with (lR,2S)-ephedrine.”

order in substrate. In these cases, a measurement

Thus the conventional

of % conversion. if one extrapolates

That equation

confirmed

kinetics

that the reaction is not

equation6 does not hold in this case leads to an underestimate

from % ee at low conversions.‘8

of the

We will report an

kinetic analysis in a future paper.

Preliminary 61% conversion

results with 2-pentenol using similar conditions after 30 minutes and a remaining

(R)-2-pentenol

of 66% ee was obtained.

resolution of other allylic alcohols. (lR,2S)-ephedrine

($0.50/g).

motivation

Acknowledgement:

References

interacts

in high ee with (lR,2S)-ephedrine,

pure (R)-BINAP-Ru

The results of kinetic resolution

economic

The amine presumably

catalysts, but it does so more selectively

A survey of a number of amines and amino alcohols showed that ephedrine

Since the (R)-2-cyclohexenol

ephedrine

of amine added per mole of Ru complex

as a function of % conversion

(10 atm H,) and (lR,2S)-ephedrine

It appears that this approach

(R)-BINAP is much more expensive

may be applicable

to the kinetic

($320/g, Aldrich Chemical Co.) than

Since racemic BINAP should be much less expensive,

for using our chiral poisoning

showed

substrate of 20% ee. After 120 min and 89% conversion,

there is a significant

strategy.

We thank the National Science Foundation

for support of this research.

and Notes:

1.

M. Kitamura, 1. Kasahara, K. Manabe, R. Noyori, H. Takaya,

J. Org. Chem.,

2.

H. Takaya, K. Mashima, K. Koyano, M. Yagi, H. Kumobayashi,

H. Taketomi, S. Akutagawa, R. Noyori,

J. Org.

Chem., 1986, 51, 629.

1988 53, 710.

7362

3.

J.W. Faller , J. Parr, J. Am. Chem. Sot.,

4.

N.W. Alcock, J.M. Brown, P.J. Maddox, J. Chem. Sot., Chem Commun., Maddox, Ch~rali~,

1991, 3,

1993, 115, 804. 1532; J.M. Brown, P.

1986,

345.

5.

K. Maruoka, H. Yamamoto,

.I. Am. Chem. Sot., 1989, 111, 789.

6.

VS. Martin, S.S. Woodard,

T. Katsuki, Y. Yamada, M. Ikeda, K.B. Sharpless,

J. Am. Chem.

SOL,

1981, 103, 6237. 7.

Y. Gao, R.M. Hanson, J.M. KIunder, S.Y. Ko, H. Masamune, K.B. Sharpless, J. Am. Chem.

1987,

Sac.,

109, 5765. 8.

S.L. Schreiber,

9.

B. Heiser, E.A. Broger, Y. Crameri, Tetrahedron:

10.

D.F. Taber, L.J. Silverberg,

11.

T. Manimaran,

M.T. Goulet, G. Schulte, J. Am. Chem. Sot.,

Tetrahedron

1987,

Asymmetry,

109,

4718.

1991, 2, 51.

Left., 1991 57, 4053.

T.-C. Wu, W. Dirk Klobucar, C.H. Kolich, G. P. Stahly, F.R. Fronczek,

Organometallics,

1993,

SE. Watkins,

12 1467.

12.

M. Kitamura, M. Tokunaga, T. Ohkuma, R. Noyori, Tetrahedron

13.

M. Kitamura, M. Tokunaga, R. Noyori,

14.

M. Kitamura, M. Tokunaga, T. Ohkuma, R. Noyori,

15.

A typical reaction procedure follows.

Lett.,

1991, 32, 4163.

J. Org. Chem., 1992,57,4053. Organic

Syntheses,

1992,

71, 1.

A dry 20 mL Schlenk tube was charged with (lR,2S)-ephedrine

(16.5 mg, 0.1 mmol) and 1 mL of CH,CI,, degassed by three freeze-thaw cycles, and then filled with N,. 1, (9.0 mg, 0.01 mmol calculated with x = 1.5) was added to this solution under

RuCl*[(~)-binap](dm~~, N, and the mixture

allowed

2_cyclohexenol(300

to stand for 2 h.

A dry 80 mL Schlenk

tube was charged

with

mg, 3.06 mmmol), CH,CI, (1 mL) and MeOH (1 mL) and degassed by three freeze-

thaw cycles and then filled with N2. This solution was added to the solution of catalyst via cannula using N, pressure.

The reaction mixture was quickly transferred

to a N,-filled

reaction vessel of a 75

mL Parr high pressure reactor under a stream of N,. The reactor was flushed with 30 atm H, two times with 10 atm H, . After the desired time had passed, the H, pressure was released

and then pressurized

and the reaction mixture concentrated NMR . Conversion

analysis of enantiomeric 16.

Enantiomeric

on a rotary evaporator. Percent conversion

of the mixture of 2-cyclohexenol

and cyclohexanol

was determined

by ‘H

to the (R)-MTPA esters allowed

purity.16

purity was determined

by ‘H NMR after conversion

to the (R)-MTPA ester.

the olefinic protons are well separated and appear as multiplets in the diastereomers

One set of

at 6 5.74 (R,S) and

6 5.83 (R ,R), (CDCl,, 300 MHz). 17.

Rates were determined

18.

The conventional

by following changes in H, pressure.

equations6 can take different form if the reactions are zero or fractional order.

certain circumstances,

as a function of % conversion

can be complicated,

first order behavior.

We are currently

appropriate

for the kinetic resolution.

equations

Under

such as the approach to saturation kinetics, the ability to accurately predict %ee but can lead to better results than expected on simple

examining

(Received in USA 27 August 1993; accepted 16 September

a broader profile of conditions

1993)

to determine

the