Studies Towards Asymmetric Catalyzed Metallo-ene Reactions

T

@

t

Vol. 3 N 3 p 6

Perganmn

@1 A r

T

A

C

1 L

S

e

reserved. Printed in Great Britain 0040-4039/97$17.00 + O.(M

PII: S0040-4039(97)01417-2

S

E

M

R

s

WolfgangOppolzer+,DavidL. Kuo, MichaelW. Hutringer,RogerIAger,Jean-O.Dumssd*,ColiISLeslie. D

A w

:T o

d C

a u

O

U

c t

P

m r

a

d G

r o t

C

w

1G @ S

s E 4 0 1 E

s

S

r

bu ts 4 L b

n

e na e

The metallo-ene reaetionl is a powerful tool for organic chemists. Intramolecular PdO1b,NiO1b,RhIIC, ZnOid
I 1 1 - Iv v Furthermore, metallo-ene type cyclizations proceed with excellent regio and stereoselectivities, allowing the preparation of enantiomerically pure intermediates or natural products”-m through efficient chirality transfer horn optically pure substrates. Howewer, the asymmetric version of the catalytic metallo-ene reaetion, where enantiospecificity is induced by chiral ligands complexed on the metal moiety, remains to be elaborated. We now report our studies towards enantioselective catalyzed metallo-ene reaction. One major diftlculty in our work is that the mechanism of the metallo-ene reaction is still not totally elucidated although a major advance towards its understanding has recently been published. It involves a zallyl metal species, where the transition state of the reaetion could have a high degree of freedom, certainly more important than in related reactions such as the Heck3 and the Pd catalyzed Alder-ene4 type cyclisations

involving a C-Pal rs-bond in the transition state. Few examples of good asymmetric inductions were reported concerning the Heck3 reaction and only one recent example of asymmetric Pd catalyzed Alder-ene type cyclisation4ahas been published, with enantioselectivities up to %~0. The ligands of Scheme 1 were examined for their effects on yield and enantioselectivity on the model substrate 1 in the Rh, Pd, Ni catalyzed metalloene reaetion. The ligands TI, T115’-d,TRAP5e-f,MOPI, MOP115g-i,were synthesized as described in the literature. The ligand TI116was prepared from enantiomerically pure7acyclohexenediaenine7bby coupling with 2-diphenylphosphinobenzoic acid using Trost’s procedure. The results are shown in Table 1.

~ // iizb —

\.

( R

t I m * a F :0

o o p :U 5 e

s C 0 d

j

o

—

d U

( S S

1 M 4.

HP u

6213

—

, (

S

1

E

B

u 3 p M -r

( (

I :R H I :R= M o c n 5

5 o

)

ee d - a

6

6214

-p N

Ts

J

A

Metal, L

/

“ ‘x

1

E

L 1 2

1 1

( (

1

$ S

4

10%(

5

10%(R)- BINAP

6

4

7

3

9 10 11

( (

-

20?%(R)-MOP1 8Y0(R,R IOYO (~ R 10%(~ R

12 IO%(@ K+TRAP a I

y

P N :X=

M

3

8

R : X=

b H

a

/

C P

Y 5

0

THF,60”C,4 h AcOH,80”C,8h

93

5

50

8

1 1 2

P N H

M T A

6 o 2

8

P

MeOH,65T, 18h THF,60W,20 h

53

10%Ni(COD)2

o

/

8%Pd(dba)2 10%Ni(COD)2 5 HRh(PPh3)4

AcOH,70”C,3 h THF,RT,24 h AcOH,80”C,8 h

50 24 26

4 12

10%Pd(dba)2

AcOH,80”C,1h

79

c

C T

O

7 Y

7 6 8

h

3 h 1 h 7h

d )

d i

0

R

i

o

e (

10%Ni(COD)2 4% ~(D10p)213

c

M

(

/ 0

R R

8

d

0

20 s

n

1

Standard ligands such as DIOP and BINAP, which are sometimes efficient for palladium-catalyzed allylic substitution 8 , were first examined. They gave an acceptable yield for the palladium-ene reaction but no me~c ~duction ww obgerv~ (enties 1,4). Haymhi’s monophosphines type ligands MOP5g,which have a reactivity different from that of bidentate phosphines for palladium-catalyzed allylic reduction 5b and subatitution5i reactions, were then studied. A lower yield was obtained using MOP I and the enrmtioselcction was still very low (entry 7). Baaed on the enhanced reactivity and selectivity observed in the nickel versus palladium-ene type cyclisationl’-b, better enantioselectivities could have been expected using nickel as catalyst. Unfortunately, if a better yield was obtained with DIOP, the enantioselectivity was low (entry 2). Furthermore, no reaction was observed with the atropoisomenc Iigands BINAP and MOP I (entries 5, 8). As the rhodium-ene reaction is known to give some major differences of selectivity in relation to Pd and Ni-ene reactions 1aC,it was worth trying to improve enantioselectivity using rhodium, but the results were still disappointing (entries 3,6). We then turned to the idea advanced by Trost. In order to bring the stereocontroling centers on the phosphines as close to the reaction centers as possible, Trost described the ligands of type T which have a large bite angle54b, suitable for inter5C’dand intramolecuhw5*bpalladiumcatalyzed allylic substitutions. The ligand TI provided a slightly better enantioselectivity than the previous ones but only in the case of the Ni-cne reaction, at room temperate. However, the chemical yield was worse (entries 9-11). To increase the bite angle to its maximum, Ito’s trans chelating Iigand TRAP5’-fwas used. TRAP ligand gave 20V0ee, in 79% yield (entry 12) when the ene-reaction was performed with palladium. Encouraged by this latter result, we decided to change our model substrate. Since there is no possibility of contmling the olefin facial selectivity of the molecule 1, we synthesized the prochiral substrate 4 witb a cyclopentadiene3* moiety suitable for metallo-ene cyclisation and olefin facial selectivity control. The synthesis is outlined as follows :

h

i

6215

”w’;Ts”PTr”Ac

4 2 3 a i) Acroleine, AcOH, H20, 987. ii) (CF3CH20)#OCH2CO~Me, KHMDS, 18-crown-6,THF, 80?/. b) i) NaBH4, MeOH,88”/. ii) TsC1,C5H5N,DMAP, 850/. c) i) DibalH, CH2C12,809/. ii) DBU,PhH, 540/.iii) AcCl, gH5N, THF, 92%

2-methyl-1,3-cyclopenanedione on acroleine9, followed by Wittig reaction under

Michael addition of

Still’s conditions’” gave the cis methyl acrylate 2. The diketone moiety was then reduced and tosylated giving 3 as a mixture of diastereomers. Reduction of the ester, elimimtion of the tosylates then acetylation of the allylic alcool furnished the desired substrate 4 in 23% overall yield. The Ni and Rh catalysts were ineffective on this substrate. However, the Pd-ene reaction led to a mixture of two diastereomers 5a” (major) and 5b (minor). The results of the Pd-ene reaction are outlined in Table 2.

J~~*” @

r 4

E

Conditions c = O.lM

Ligand

P 3

6

S

A

y)

3%

6.5%(R)-BINAP

A

y

4

2

(R)-MOP II

4 4

2

(

I

M

7

3

(

1

M

I

@ + Gb)

5

7

4

8 9 ,

1

(

M

(

S

8

Conversion(%?) ratio 5a / d 5

30

1

MeOH,70”C,2

69131

14

/

100

74/26

23

22

100

62138

2

100

4

1

100

A

7

7

9

8

5

8

A

8

3

100

8

0

0

A

7

7

100

8

32

26

100 100 100

88/ 12 93I 7 88/12

34 45 40

19 10 ““

9

87/ 13

47

15

8

8

? (

S

3

6

(

R

II

4 4 4%

I ( S 1 ( R)-TI 12‘X.(R, R)-TIII

M 7 “ 8 MeO~ 40“C,3 MeOH,72T, 2

12

4%

1

M

a G

A

e 2 -

c o t

(

R c

4

H

3

b P .

b \ @

e 5a ee f+b(~o)”

H

II

OMe

a c

f G A T

c D c

0

0

6

32

27

6 /

4

3

I

a cn c

ad 2 cy

L

2 )

R

S

A E

2

As expected, the Pd-ene reaction applied to the substrate 4 gave better enantiosclectivities than the substrate 1 using standard ligands such as DIOP and BINAP (entries 1,2). BINAP gave more then 20°/0ee in an average diastereomeric ratio. The monodentate phosphine MOP II was very effective in catalyzing the reaction, but no asymmetric induction was observed (entry 3). By contrast, 32Y0 ee was obtained using the ligand MOP I in the same conditions (entry 4). This difference of behaviour could he due to the free naphthol group of MOP I, suitable for Pd chelation, which could confer on this Iigand a bidentate nature. Lowering the temperature to 45°C improved the ee to qz~., but increased the reaction time, and the diastereomeric ratio was still low (entry 5). Ligands having a large bite angle and the trans chelating ligrmd TRAP were then studied. Surprisingly, the ligand TII which has the largest bite angle of the ligands of type T, and the TRAP ligrm~ were effective to catalyze the reaction in an acceptable diastereomeric ratio, but no asymmetric induction was obtained (entries 6, 7). By contrast, TI increased the ee to the value observed for MOP I, in a much better diaste~omeric ratio (entries 8, 9). The reaction was best performed at 45°C, in MeOH, with 2 to 3 equivalents

P

o n

6216

of ligand per Pd (entryIO). Under these conditions, the diastereomeric ratio was excellent and the ee reached 4s~0. A slightly increase of the ee was even observed with the more rigid ligand TIII under the same conditions (entry 12). In conclusion, we have opened the way to new methodologies toward enantioselective catalyzed metalloene reaction. We have shown that ligands with closely related structures can have totally different behaviors. The best inductions were obtained using the Pd-ene reaction applied to the substrate 4 derived from cyclopentadiene. The bidentate Iigands TI1 and TIII were shown to be the most effective for this substrate. Thus, new improvements can be expected in the field of asymmetric catalysis. : F

A

E N a a u

R M m d

C

s o C W t

o t cw i g M C W

b t a

S N S F k e ti N S a n W t M A P M J Se c M D K mfg o M J eD M I S M P M ai M R S af k i s s

nu

o

References and notes 1. a) G

W C O C E A oE W S rF Wp G 1 1 mb h g W P a C 1 1 a r c t c Onp W Fh A H C 9 9 1 7 2 d O W S F T L 1 33 7 e M C M pe1 c G N JF . O C 1 6 o8 t O W X J S C H C Au 1 o 6 1 h O W K 9T eH K D W , 9 g O W R C T 4 pL P L 1 3 1 i D T Y eA N S Y K T2 T J O C a t e a 1 6 2 j T M K R $ 1 3 2 k O 6 W G JM B T. M., h 0 T L 1 1G W 9B e T B G T8 L 1 1 ti 8 e p 1 m O W B H G A J A C S 1 e 1 p i n 1 2 G E C JM E AM M o G A C I E 1 c m 9a u S 1 7 3 a O T K K A M S M S M J A C a1 h a r c h o t b A A M T O L P D A C [ . 1 h c R L Hh A 9 v a J O C 1 9 9 h 7 4 a G A S M K R I Y A C W E 1 3 6 a b T BM C B T L 1 9 e 9 t e 4 5 a T BM V V DL C R 1 b T BM V V 9 D ,B C J A C r 9 K A C B R C Z J . A 3C S 1 1 d r S 1 I 9 c Trost B T BM K M R R Z G J A C S 1 1 6 e S M [r Y a 1 2 5 t S M H H S s M K R I Y O 9t a a T :A e 1 1 g U Y T A L S Y4 H T J O C 1 5 1 a h, H T o i H T I H N M M Y U Y M M Y K J A w C S 1 a 1 o K M U Y LC S C C 1 a 9 o 25=+59 (c = 1,CH2C12).IH NMR ( MHZ, CDC13): 1.8-1.9 ( 2 H 2 ( 2H 4 6. D f l T :[ i 4.2 ( 2 H), 5 ( 2 H 6 ( J=8H 2H 6 ( 2H 7 ( 2 H 7 ( 2H 3 N ( . . M C :D 9 ‘ a L J J E G Y H Y N X Z C O C 1 1 ba W D R a DP F J G J J M C 1 1 a 3 9 8. a G J S A S M S R J L T e 1 b Ku N 9 e a r e G J T A 1 1 e s 5 t 9 9 O Z H G P R O C V 7 r g a a, 10. S C G C T M 1 9 e e 8 t 11. D f 5 I 3 1 1 1 0 73 I N M 3C 76 1 ( 1 3H 1 ( 2H 1 ( 2H 2 ( 3H 2 ( J = 8 H 1 . H 1H 2 . ( 1H 5 ( ( J = 1 Hz, 1 Hz, 1 H 5 (t, J = 1 H 1H 5 d J = 5 H 1 H 1H 5 J = 1 Hz, 1.SHz, 1 H 5 ( J = 5 H 1H 5 ( 1H 1 N ( M C 1 ( 1 ( 1 ( 1 (. 1 ( 8 ( 5 ( 5 ( 4 ( 3 ( 2 ( 2 ( 2 ( M :2 ( 1 ( 1 ( 1 ( ( c 1 [ C C 1 ( 1 ( 9 ( 9 ( 7 ( 6 ( H :1 1 5 w n i I s w a f r 1a 3 s t oe s 1 P a w p al t R M lF M P M b aS 1 1 d i1 b t c r i e s o t c ( o P s t d o t e t c u f Pi s ds o x M M P F P R J O C 1o 9 r l a 9 1 F A J BR C WR h $ 1 8 e 9 u O A C 7 T

(Received in France 17 May 1997; accepted 10 July 1997)

t

e

y

1

8

8

6