Enantioselective tandem conjugate addition elimination reactions 3

3.1 Tandem CA-E reaction between linear Morita-Baylis-Hillman MBH allylic bromides and 1,3-dicarbonyl compounds 3.1.1 Synthesis of substrates Inspired by the results obtained from the

Chapter 3

Tandem Conjugate Addition-Elimination Reaction of Linear

Activated Allylic Bromides

3.1 Tandem CA-E reaction between linear Morita-Baylis-Hillman (MBH) allylic bromides and 1,3-dicarbonyl compounds

3.1.1 Synthesis of substrates

Inspired by the results obtained from the tandem CA-E reaction of cyclic MBH allylic bromides, we were keen to examine the reaction of linear substrates As shown in Scheme 3.1, various MBH allylic bromides were prepared by DABCO promoted Baylis-Hillman reactions followed by bromination with concentrated HBr and H2SO4 (1)1 or NBS together with dimethyl sulfite (2)2 A wide range of commercially available aldehydes and activated alkenes allowed us to prepare a variety of linear MBH allylic bromides Subsequently, we subjected these MBH allylic bromides to the tandem CA-E reaction

conditions S,S'-Di-tert-butyl dithiomalonate was firstly investigated as the nucleophile

for this reaction

OH

conc HBr/H2SO4

CH2Cl2, 0oC-rt

COR' Br

(Z)

OH COR' NBS, Me2 S

CH2Cl2, 0oC-rt

COR' Br

(Z)

H O

+ COR'

COR'

CH2Cl2or THF, rt

1 eq DABCO

OH COR'

R

R

(1)

(2)

Scheme 3.1 Synthesis of linear MBH allylic bromides

3.1.2 Reaction between linear MBH allylic bromides and S,S'-di-tert-butyl

dithiomalonate

1

(a) C Börner, J Gimeno, S Gladiali, J Goldsmith, D Ramazzotti and S Woodward, Chem Comm., 2000, 2433-2434

(b) L Fernandes, A J Bortoluzzi and M M Sá, Tetrahedron, 2004, 60, 9983-9989

2

H M R Hoffmann and J Rabe, J Org Chem., 1985, 50, 3849-3859

While triethylamine was proved to be inefficient for this reaction, a more nucleophilic base, DABCO was employed as the promoter The achiral tandem CA-E reaction of

various linear MBH allylic bromides 39a-j (Table 3.1) was achieved by using 2

equivalents DABCO (Scheme 3.2) Tandem CA-E product was obtained as a single product

Br

CO2Me R

COStBu

COStBu +

CH2Cl2, rt

2 eq DABCO

CO2Me R

COStBu

tBuSOC

SN2' type product

39a-j

N N

Scheme 3.2 Achiral tandem CA-E reaction between 39a-j and S,S'-di-tert-butyl

dithiomalonate

With the model reaction in hand, we started to investigate the asymmetric tandem

CA-E reaction between linear substrates 39a-j and S,S'-di-tert-butyl dithiomalonate

(Table 3.1) These reactions were generally slower than those of cyclic substrates Hence,

2 equivalents chiral promoter was used to enhance the reaction rate However, when CPS

spot was observed on TLC We have also attempted different solvent pairs to develop the TLC, but the separation of the two product spots remained a challenge The side product was assigned as SN2 type product by 1H NMR analysis However, the reason for the formation of the SN2 type product in the reaction of linear MBH allylic bromides is unknown

Table 3.1 Effect of different substitutions on the aryl group of MBH allylic bromides

CO2Me R

COStBu

COStBu +

CH3CN, rt

2 eq 11h

CO 2 Me

COStBu

tBuSOC

S N 2' type product

CO2Me R

COStBu

COStBu

S N 2 type product

39a-j

40a-j

NH MesO2S

N

Entry 39, R Product Time(hr) Yield/%a ee/%b SN2′: SN2

ratioc

a

Isolated yield of both SN2′ type and SN2 type products bDetermined by chiral HPLC analysis cDetermined by 1H NMR analysis dVery poor conversion

As shown in Table 3.1, a general observation of asymmetric tandem CA-E reaction was that the yield was typically low, often in the range of 30-60% and the ratio of SN2’ to

SN2 type products can vary from ratios of 5:1 to 2:1 When the aryl group contains no substituent (entry 1), the reaction was extremely slow; and a very small amount of product could be obtained for chiral HPLC analysis MBH allylic bromides with electron withdrawing groups were relatively more reactive and provided moderate ee values

(entries 2-5) It was also found that some allylic bromides with substituents in the meta

position of the phenyl ring tend to give better ees (entries 7,8 and 10) However, allylic

bromide 39i with NO2 group in the meta position produced same enantioselectivity as

39b When the substituent in the meta position was changed from chlorine to a bigger

group, bromine, the ee dropped from 60% (entry 7) to 47% (entry 8)

Br

CO2Me

COStBu

COStBu +

CH3CN, rt

2 eq 11h 41

no reaction

Scheme 3.3 Reaction between an alkyl substituted MBH allylic bromide and

S,S'-di-tert-butyl dithiomalonate

It was noteworthy that when alkyl MBH allylic bromide 41 was employed as the

substrate for tandem CA-E reaction, no product was observed (Scheme 3.3) Therefore,

we moved on to synthesize MBH allylic bromides from 4-nitrobenzaldehyde and different activated alkenes With a variety of commercially available acrylates and vinyl ketones, we tried to vary the activated alkene part of the substrates and test how these substituents affect the tandem CA-E reaction

As shown in Table 3.2, when a phenyl acrylate derived MBH allylic bromide 42a was employed, the enantioselectivity decreased by 23% when compared with 39b though an

excellent yield was obtained (entry 1) Other phenyl acrylate derived MBH allylic

bromides 42b and 42c were proved to be poor substrates in terms of enantioselectivity

As nitro group could form hydrogen bonding with the CPS promoter, it might unlock the substrate-promoter complex which gives high enantioselectivity It was also observed that alkyl acrylates derived MBH allylic bromides (entries 4,5 and 7) gave similar levels of enantioselectivities However, the substrate with a long alkyl chain (entry 6) or electron withdrawing group (entry 8) gave very poor ee values

In order to further explore the substrate scope of the reaction, we attempted several

MBH allylic bromides prepared from vinyl ketones (entries 9-11) It was found these reactions were generally slow and gave poor yields The best ee (57%) was obtained with

42j which was prepared from ethyl vinyl ketone

Table 3.2 Effects of different substitutions on activated alkene part of MBH allylic

bromides.a

COR COStBu

COStBu +

CH3CN, rt

2 eq 11h

COR

COStBu

tBuSOC

O2N

Entry 42, R Product Time(hr) Yield/%b ee/%c

a

The SN2′:SN2 ratio was obtained approximately 5:1 by 1H NMR analysis bIsolated yield

of both SN2′ type and SN2 type products cDetermined by chiral HPLC analysis

We have also tested the tandem CA-E reaction between 24b and a bulky

dithiomalonate (Scheme 3.4) Nonetheless, only 35% ee was obtained along with poor yield (20%) after 22 hours Therefore, we proceeded to investigate the reactions promoted by different CPS promoters

O 2 N Br

CO2Me

39b

+ COR

3 CN, rt

2 eq 11h

S R=

CO2Me

COR ROC

O2N

44

22h, 20%yield, 35%ee

Scheme 3.4 CPS 11h promoted reaction between 39b and S,S'-di-tert-octyl

dithiomalonate

Table 3.3 Effects of different CPS promotersa

CO2Me

COStBu

COStBu +

2 Me

COStBu

tBuSOC

O2N

2 eq promoter

Entry Promoter Time(hr) Yield/%b ee/%c

Bn NHTs

11a

48 47 0

RO 2 S

N

R = 2,4,6-triisopropylpheny

11i

48 60 60

MesO2S

N

OMe

11j

48 43 51

a

The SN2′: SN2 ratio was determined as 5:1 by 1H NMR analysis bIsolated yield of both

SN2′ type and SN2 type products cDetermined by chiral HPLC analysis

When CPS 11a was used to promote the reaction between 39b and S,S'-di-tert-butyl

dithiomalonate, no enantioselectivity was observed (Table 3.3, entry 1) CPS 11i and 11j could also promote this reaction and gave similar results as 11h

3.2 Other linear substrates

In addition to MBH allylic bromides, we have prepared other linear substrates and

subjected them to the tandem CA-E reaction condition Investigations of CPS 11h

promoted reactions between other linear substrates and 1,3-dicarbonyl compounds were shown in Table 3.5

We envisioned that MBH allylic iodide 453 would be an efficient substrate as a better leaving group may enhance the reaction rate of nucleophilic substitution by CPS promoter To our surprise, the reaction rate was not improved while the enantioselectivity increased by 8% when compared with its corresponding allylic bromide (entry 1) We suspected that the starting material may decompose during the reaction process, which

resulted in the moderate yield Other two substrates 46 and 47 were also synthesized from

MBH allylic alcohol and tested for the reaction with S,S'-di-tert-butyl dithiomalonate

(entries 2,3) These two substrates gave similar level of yields and SN2′: SN2 ratio as MBH allylic bromide However, the enantioselectivities of these two reactions decreased dramatically A possible explanation to this is that these two substrates may undergo double tandem CA-E process or direct nucleophilic substitution reaction to yield product

40b (Scheme 3.5) Therefore, the complexes formed from substrates and promoters might

be different from that of MBH allylic bromides

Table 3.4 11h promoted reaction of other substrates derived from Morita-Baylis-Hillman

allylic alcohol

COStBu

COStBu +

2 Me

COStBu

tBuSOC

O2N

2 eq 11h

40b

substrates

45-47

3

(a) B Das, A Majhi, J Banerjee, N Chowdhury and K Venkateswarlu, Tetrahedron Lett., 2005, 46, 7913-7915 (b) J

Li, X Wang and Y Zhang, Synlett, 2005, 6, 1039-1041

Entry Substrate Time(hr) Yield/%a ee/%b

1

CO2Me

45

24 35 62

OAc

36 33 15

OBoc

O2N

47

36 24 30

a

Isolated yield of both SN2′ type and SN2 type products bDetermined by chiral HPLC analysis

OAc

O2N

NR3

O OMe

OAc

O2N

O OMe

NR3

O2N

O OMe

NR3 AcO

COStBu

COStBu H

R3N

O2N

O OMe

NR3 AcO

tBuSOC COStBu

COStBu

tBuSOC

Scheme 3.5 Formation of 40b via double tandem CA-E process

3.2.2 Reaction of other linear substrates

We have also applied the tandem CA-E reaction condition to other activated allylic

bromides such as 48 and 51

Methyl 2-(bromomethyl)acrylate 48 is a commercially available activated allylic bromide With 2 equivalents CPS 11h, several 1,3-dicarbonyl compounds 49a-d were

used to react with substrate 48 (Table 3.5) It is interesting to note that quaternary carbons could be constructed in one step when 49b-d were used as the donor (entries 2-4)

Although these reactions could provide moderate to very high yields, no enantioselectivity was obtained, which indicates that CPS may not be a suitable promoter for this reaction

Table 3.5 11h promoted reaction of 48

49

CH2Cl2, rt

2 eq 11h

+ X Y Z

CO2Me Y

Z X

Entry Donor Product Time(hr) Yield/%a ee/%b

1 49a

CN

2

49b

CN

3 49c

O O

4 49d

O O

a

Isolated yield bDetermined by chiral HPLC analysis

In addition to commercially available substrate, we have also synthesized 51 from

dimethyl itaconate (Scheme 3.6) and subjected it to the tandem CA-E reaction condition

MeO

O

OMe

O CH2Cl2, 0

o C-rt

Br2

MeO

O

OMe O

Br

Br CH2Cl2, rt

Et3N

MeO

O

OMe O Br

51

Scheme 3.6 Synthesis of 37

When S,S′-di-tert-butyl dithiomalonate was used to react with 51, no reaction was observed even with more equivalents of CPS promoter Thus, a less hindered donor,

S,S'-diethyl dithiomalonate was employed The reaction gave tandem CA-E product 52 in only 20% yield and 20% ee, indicating that 51 is a less reactive acceptor than MBH

allylic bromides for the tandem CA-E reaction (Scheme 3.8)

MeO

O

OMe O

MeO

O

OMe O

Br

+ COSEt

COSEt CH2Cl2, rt

2 eq 11h

52

21 hrs, 20% yield, 20% ee

51

Scheme 3.7 Reaction between 51 and S,S'-diethyl dithiomalonate

In conclusion, in this chapter, we described an asymmetric tandem conjugate addition-elimination (CA-E) reaction between linear Morita-Baylis-Hillman allylic bromides and 1,3-dicarbonyl compounds promoted by chiral pyrrolidinyl sulfonamide (CPS) Generally, moderate enantioselectivities were obtained but the yields were less than satisfactory and the reactions often gave a mixture of SN2′-type and SN2 products Future work includes improving the reaction rate and enantioselectivity using more efficient promoters