Enantioselective tandem conjugate addition elimination reactions 2

Based on the results obtained from the model reaction between 3 and dimethyl malonate, we are keen to develop an asymmetric tandem conjugate addition-elimination CA-E reaction using chi

Chapter 2

Tandem Conjugate Addition-Elimination Reaction of Cyclic

Activated Allylic Bromides

2.1 Reaction between 2-(bromomethyl)cyclopent-2-enone and dimethyl malonate

2.1.1 Preliminary studies

As a starting point, a MBH allylic bromide 3 was prepared by an imidazole promoted

Baylis-Hillman reaction1 between 2-cyclopentenone and formaldehyde followed by bromination2 using PBr3 (Scheme 2.1) We subsequently chose this allylic bromide as the target of initial studies, which focused on standardizing reaction conditions To the best of

our knowledge, this MBH allylic bromide 3 has never been investigated as an

electrophile for any nucleophilic substitution or conjugate addition reaction before

O

H N (1.0 eq) THF-H2O (1:1, v/v), rt

PBr3ether, 0oC

3 2

1

Scheme 2.1 Synthesis of cyclic MBH allylic bromide 3

We found that with 2 equivalents of triethylamine, the reaction between

2-(bromomethyl)cyclopent-2-enone (3) and dimethyl malonate completed in 24 hrs at

room temperature in CH2Cl2 Stoichiometric base was needed for this reaction as HBr could be generated during the nucleophilic reaction process It is worth noting that SN2’ product was obtained in an isolated yield of 85% while SN2 type product was not observed (Scheme 2.2) Other dialkyl malonates such as diethyl malonate and

di-isopropyl malonate were also subjected to the reaction condition However, the SN2′ type products were obtained only in moderate yields Thus we focused on the reaction

between 3 and dimethyl malonate, which could be used as a model reaction for

1 S Luo, B Zhang, J He, A Janczuk, P G Wang and J-P Cheng, Tetrahedron Lett., 2002, 43, 7369-7371

2 H.-K Yim, Y Liao and H N C Wong, Tetrahedron, 2003, 59, 1877-1884

Scheme 2.2 Reaction between 3 and dimethyl malonate

To better understand how the SN2′ type product was formed, we particularly looked

at the mechanistic aspects of this reaction Based on the results obtained and Lee’s proposed mechanism3, we postulated that the reaction may proceed through a tandem conjugate addition-elimination (C-AE) process (Scheme 2.3)

3

CO2Me MeO2C

Scheme 2.3 Proposed mechanism of tandem CA-E process

In Scheme 2.3, an initial nucleophilic substitution by triethylamine on the bromide of

3 resulted in the formation of an ionic intermediate Deprotonation of dimethyl malonate

by triethylamine generated the carbon nucleophile Subsequently, nucleophilic attack of dimethyl malonate anion onto the β-carbon of the α,β-unsaturated double bond led to the

formation of the enolate 5 Finally, the elimination of the promoter would give rise to the

SN2′ type product 4 In contrast, the formation of the SN2 type product would be via the

direct nucleophilic substitution on the carbon adjacent to the leaving group

Based on the results obtained from the model reaction between 3 and dimethyl

malonate, we are keen to develop an asymmetric tandem conjugate addition-elimination (CA-E) reaction using chiral leaving group strategy (Scheme 2.3) Using a chiral tertiary amine as the promoter, enantioselectivity could be achieved through a tandem CA-E fashion

2.1.2 Cinchona alkaloids promoted tandem CA-E reactions

Nowadays, Cinchona alkaloid and its derivatives have attracted lots of chemists’

attention and have been widely used as organocatalysts for various kinds of reactions such as Michael reaction4, Henry reaction5, Mannich reaction6, Friedel-Crafts7 reaction, Diels-Alder reaction8 and so forth

With various kinds of commercially available Cinchona alkaloids (Figure 2.1), we

embarked on the study of asymmetric tandem CA-E reaction between 3 and dimethyl

malonate (Table 2.1) When 2 equivalents alkaloids such as quinidine and quinine were employed, the reaction could reach 100% conversion according to TLC9 after several

4 Selected examples of Michael reactions catalyzed by cinchona alkaloids: a) H Li, Y Wang, L Tang, F Wu, X Liu, C

Guo, B M Foxman and L Deng, Angew Chem Int Ed., 2005, 44, 105-108 b) M Bella and K A Jørgenson, J Am

Chem Soc., 2004, 126, 5672-5673 c) B Vakulya, S Varga, A Csámpai and T Soós, Org Lett., 2005, 7, 1967-1969

5 H Li, B Wang and L Deng, J Am Chem Soc., 2006, 128, 732-733

6 a) J Song, Y Wang and L Deng, J Am Chem Soc., 2006, 128, 6048-6049 b) A Ting, S Lou and S E Schaus, Org

Lett., 2006, 8, 2003-2006

7 Y.-Q Wang, J Song, R Hong, H Li and L Deng, J Am Chem Soc., 2006, 128, 8156-8157

8 Y Wang, H Li, Y.-Q Wang, Y Liu, B M Foxman and L Deng, J Am Chem Soc., 2007, 129, 6364-6365

9 When the substrate spot on TLC disappeared on TLC, it was considered as 100% conversion However, a big spot

was observed on the baseline (not corresponds to alkaloid), which was suspected to be the salt intermediate 6

hours at room temperature (entries 1 and 2) However, reaction yields were very poor along with moderate enantioselectivities It was deduced that the majority of substrate

remained as an ionic intermediate (Scheme 2.3, intermediate 6) after the initial

nucleophilic substitution of chiral promoter to the substrate In addition, it was observed that cinchonidine and cinchonine only promoted the reaction at very slow reaction rates

Table 2.1 Cinchona alkaloid and its derivatives promoted tandem CA-E reactions

aIsolated yield bDetermined by chiral HPLC analysis

Subsequently, the effects of several other commercially available alkaloids

hydroquinine and A-F (entries 3-9) were explored As expected, these alkaloids did not

provide decent enantioselectivities although there was an improvement on the yield of the reaction The best results (99% yield and 39% ee) were obtained with a quinidine

derivative A as the promoter It is interesting to note that when promoter G was employed,

no product was observed

N H

N

HO MeO

CH 3

N

H3C

N N

O

O Cl

O

O Cl

N

HO MeO

hydroquinine

N H

NH OMe

S HN

F3C CF3

G

Figure 2.1 Cinchona alkaloid and its derivatives

Furthermore, we surveyed the reaction between substrate 3 and dimethyl malonate

under a phase-transfer condition (Scheme 2.4) With 10 mol% N-benzylcinchonidinium

chloride as the phase-transfer catalyst (PTC) and 1M NaHCO3 aqueous solution as the

stoichiometric base, 3 reacted smoothly with dimethyl malonate in a very good yield but

with almost no enantioselectivity

N H

N

HO

Ph Cl

PTC

Scheme 2.4 Tandem CA-E reaction under PTC condition

2.1.3 Other chiral tertiary amines as promoters

In order to find a suitable promoter for this tandem CA-E reaction, we also attempted other chiral tertiary amines (Table 2.2), which are either commercially available or synthesized molecules However, chiral imidazoline (entry 1)10 and a (-)-pseudoepherin derivative (entry 2)11 were proved to be ineffective promoters Other two chiral tertiary amines (entries 3, 4) could only promote this reaction at very slow reaction rate with very poor enantioselectivies

Table 2.2 Several chiral tertiary amines promoted tandem CA-E reaction

10 J Xu, Y Guan, S Yang, Y Ng, G Peh, C.-H Tan, Chemistry: An Asian Journal2006, 1, 724-729

11 Compound was prepared by the following procedures:

CH2Cl2, Et3N

0oC-rt

Ph

N OH

O

Ph LiAlH4THF, reflux

Ph

N OH

Ph

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

1

N N Ph

- conversion Poor -

N OH

aIsolated yield bDetermined by chiral HPLC analysis

2.2 Chiral pyrrolidinyl sulfonamide (CPS)

2.2.1 Introduction

From the above results, we concluded that an efficient and highly selective promoter was desirable for this tandem CA-E reaction Therefore, we designed a class of bi-functional promoters, chiral pyrrolindinyl sulfonamide (CPS) (Figure 2.2) This promoter contains a tertiary amine which can undergo nucleophilic substitution with

MBH allylic bromide to form the salt intermediate 6 In addition, the acidic -NH group

might activate the carbonyl group of the substrate via hydrogen-bonding

Nucleophilic amine

Hydrogen-bonding donor

Figure 2.2 Chiral pyrrolidinyl sulfonamide (CPS)

As organocatalysts, simple and versatile chiral sulfonamides available for various catalytic asymmetric reactions have been developed recently Wang and co-workers

reported that a chiral sulfonamide 7 (Figure 2.3), which resembles L-proline, could be

utilized as an organocatalyst for enantioselective Michael addition reaction of aldehydes and ketones towards nitroolefins12 7 also serves as an effective organocatalyst for

promoting direct, highly enantioselective Aldol reactions of α,α-dialkylaldehydes with aromatic aldehydes13 In addition, 7 can catalyze α-selenenylation and α-sulfenylation

reactions in which L-proline shows poor catalytic activity14 The enhanced catalytic

activity and enantioselectivity for these reactions promoted by 7 are due to the acidic and

sterically bulky properties of the trifluoro-methanesulfonamide group11b

Another bifunctional chiral sulfonamide 8 (Figure 2.3) was firstly reported by Nagao

to achieve a highly enantioselective thiolysis of prochiral cyclic dicarboxylic anhydride15

Impressively, only 5 mol% catalyst 8 was required for the reaction between cyclic

anhydride and 1.2 equivalents benzyl mercapten (BnSH)

H N N

Figure 2.3 Various sulfonamides as organocatalysts

Ishihara also disclosed that an L-histidine derived chiral sulphonamide 9 acted as an

12 a) W Wang, J Wang and H Li, Angew Chem Int Ed., 2005, 44, 1369-1371 b) J Wang, H Li, B Lou, L Zu, H Guo and W Wang, Chem Eur J., 2006, 12, 4321-4332

13 W Wang, H Li and J Wang, Tetrahedron Lett., 2005, 46, 5077-5079

14 J Wang, H Li, Y Mei, B Lou, D Xu, D Xie, H Guo and W.Wang, J Org Chem., 2005, 70, 5678-5687

artificial acylase for the kinetic resolution of racemic alcohols16 Polymer-bound catalyst was also prepared and reused more than 6 cycles without loss of activity This could be a practical method to prepare chiral diols or chiral amino alcohols

An attractive direct Mannich reaction catalyzed by an axially chiral amino

sulfonamide 10 was recently reported by Maruoka for the synthesis of anti-β-amino

aldehyde17 Excellent enantioselectivities (>99%) with high anti-selectivies could be

achieved by employing only 2 mol% chiral catalyst Sulfonamide 10 has also been found

to catalyze the syn-selective direct asymmetric cross-aldol reactions between aldehydes

Similarly, with the use of 5 mol% of catalyst 5, excellent levels of enantioselectivities (92

to 99%) and syn/anti ratios (up to >20/1) were obtained for the rare example of a

syn-selective direct cross-aldol reaction via an enamine intermediate18 The use of catalyst loadings as low as 5 mol% or lower makes organocatalysts almost as competent

as the traditional organometallic catalysts, which are well known for their low catalysts loading

Inspired by these chiral sulfonamides catalyzed highly enantioselective reactions; we were keen to utilize our designed chiral pyrrolidinyl sulphonamide (CPS) on the tandem CA-E reaction

2.2.2 Synthesis of chiral pyrrolidinyl sulfonamide (CPS)

The CPS promoters could be prepared via two different routes as presented in Scheme 2.5 and 2.6 In Scheme 2.5, N-sulfonyl aziridines were readily prepared from

their corresponding commercially available chiral amino alcohols19 The regioselective

16 K Ishihara, Y Kosugi and M Akakura, J Am Chem Soc., 2004, 126, 12212-12213

17 T Kano, Y Yamaguchi, O Tokuda and K Maruoka, J Am Chem Soc., 2005, 127, 16408-16409

18 T Kano, Y Yamaguchi, Y Tanaka and K Maruoka, Angew Chem Int Ed., 2007, 46, 1738-1740

19 a) W Ye, D Leow, S L M Goh, C.-T Tan, C.-H Chian and C.-H Tan, Tetrahedron Lett., 2006, 47, 1007-1010 b)

B M Kim, S M So, H J Choi, Org Lett., 2002, 4, 949-952

ring opening reaction using pyrrolidine or piperidine afforded the desired CPS promoters.

In 3 steps, the CPSs can be obtained in multigram quantities and high yields In Scheme 2.6,

the amino group of L-tert-leucinol was protected using Boc anhydride and 2M NaOH

solution Boc-L-tert-leucinol 12 was then coupled with pyrrolidine or piperidine using

HOBt and DCC The intermediate 13 was then subjected to deprotection condition followed by reduction to afford the chiral diamine 15 The final step involved a protection

step using p-TsCl or other kinds of sulfonyl chlorides

OH

NH2

N NHR2

R1N

Scheme 2.5 Synthesis of CPS promoters from chiral amino alcohols Reagents and

conditions: (i) R1 = Bn: p-TsCl or other sulfonyl chlorides, Et3N, CH3CN; R1 = tBu:

p-TsCl or other sulfonyl chlorides, Et3N, CH3CN, MS(4A), 0oC then MsCl, DMAP, Et3N,

CH2Cl2, rt; (ii) pyrrolidine or piperidine, CH3CN, reflux

OH NHBoc

ii

NH2

R1O N OH

Scheme 2.6 Synthesis of CPS promoters from chiral amino acids Reagents and

conditions: (i) (Boc)2O, 2M NaOH, 0oC to rt; (ii) pyrrolidine or piperidine, HOBt, DCC, THF, 0oC to rt; (iii) 3M HCl in EtOAc; (iv) LiAlH4, THF, rt to reflux; (v) p-TsCl or other

sulfonyl chlorides, DMAP, Et3N, CH2Cl2, 0oC to rt

Comparing two synthetic pathways, the first one makes use of chiral amino alcohols whereas the second involves chiral amino acids As the number of steps in Scheme 2.6 is more than twice than that of Scheme 2.5, the synthetic route employing chiral amino

acids would be more time-consuming with a lower overall yield However, this may provide an easy way to install various sulfonyl groups onto the amino group to afford the CPS promoters containing same R1 group Considering the efficiency, the first route (Scheme 2.5) was the preferred synthetic pathway and it was used for the synthesis of most of the promoters used in the following research

2.3 Chiral Pyrrolidinyl sulphonamide (CPS) promoted tandem CA-E reactions

2.3.1 Reaction between 2-(bromomethyl)cyclopent-2-enone and 1,3-dicarbonyl compounds

It was found that with 2 equivalents CPS 11a, the tandem CA-E product 4 was

obtained in 41% isolated yield and 78% ee (Scheme 2.7) However, the reaction was too

slow to be useful We have recently found that S,S'-dialkyl dithiomalonates are effective

donors for chiral bicyclic guanidine catalysed Michael reactions20 due to the high

α-proton acidity Thus, we envisaged that S,S'-dialkyl dithiomalonate could be an enol

equivalent with high reactivity

NHTs

4

4days, 41%yield,78%ee

Scheme 2.7 CPS 11a promoted reaction between 3 and dimethyl malonate

Firstly, we examined the reaction between 3 and S,S'-di-n-propyl dithiomalonate 16a

by employing a variety of promoters listed in Figure 2.4 As a reliable starting point, these reactions were conducted in dichloromethane at room temperature (Table 2.3)

20 W Ye, Z Jiang, Y Zhao, S L M Goh, D Leow, Y.-T Soh and C.-H Tan, Adv Synth Catal., 2007, 349, 2454-2458

Bn

NHTs

N NHTs

N NHTs

11a

NH MesO2S

N

N Bn

NH BnO2S

11b

N Bn

NHNs

Ns = p-nitrobenzenesulfonyl

11c

NH MesO2S

N Bn

N

OMe

11j

Figure 2.4 Various CPSs

As expected, reaction with S,S'-di-n-propyl dithiomalonate 16a as the donor could

complete within 1 day and provided satisfactory results (Table 2.3, entry 1) There was no

improvement of enantioselectivity when the Ts group of CPS 11a was replaced by BnSO2

(entry 2) or Nosyl group (entry 3) When we installed a bulky 2,4,6-trimethylphenyl group onto the promoter, the enantioselectivity increased by 10% along with very good

yield (entry 4) CPS 11e with a pipridine ring was proven to be ineffective (less than 20% conversion after 24 hours) so 11g was not investigated in this part Moreover, the results

obtained with promoter 11f revealed that a bulkier tert-butyl group would be essential to

the enantioselectivity of the reaction Finally, we observed the best results with CPS 11h (entry 7) and it was used in the subsequent studies It is interesting to note that CPS 11i,

which contains a bulkier protecting group, produced almost the same ee value as that of

11h When we installed an extra chiral center on the CPS 11i, the enantioselectivity

slightly decreased (entry 9) We also attempted combining chiral pyrrolidine with

thiourea21, which was known as an activator of carbonyl group by double hydrogen bonding interactions Although 66% yield was obtained after 22 hours (entry 10), the

selectivity dropped by 11% comparing with that of 11a

+ COSnPr COSnPr

aTwo equivalents of the 16a were used to react with one equivalent of 3. bIsolated yield

cDetermined by chiral HPLC analysis The absolute configuration was determined by single crystal X-ray analysis, see Experimental chapter dSlow reaction e

11k was prepared by mixing chiral diamine 15 with corresponding isothiocyanate

When 2 equivalents CPS 11h were employed, various S,S'-dialkyl dithiomalonates

were observed to participate in the reaction with 3, giving the tandem CA-E products 17b-g in moderate to good yields (Table 2.4) Most of dithiomalonates gave similar levels

of yield and ee except S,S'-dibenzyl dithiomalonate, which is a very unreactive donor for

21 P M Pihko,Angew Chem Int Ed 2004, 43, 2062-2064

this reaction Since S,S'-di-tert-butyl dithiomalonate (16f) produced the best results, it

was used for the optimization of the reaction conditions It is noteworthy that a less

hindered donor, S,S'-diethyl dithiomalonate, afforded the tandem CA-E product with 84%

ee (entry 1) However, lowering the temperature to 0oC and -20oC slowed the reaction rate considerably without any improvement in enantioselectivity

Table 2.4 CPS 11h promoted reaction between 3 and various S,S'-dialkyl

dithiomalonates

+ COSR COSR

equivalents CPS 11a (Table 2.5, entries 1-6) It was found that non-polar solvents such as

toluene resulted in a low yield of 15% even after 48 hours of reaction time (entry 1) The low yield could be attributed to the low solvating power of toluene, as it could not fully

dissolve all the starting materials as well as promoter 11a Reactions in chlorinated

solvents such as CH2Cl2 (entry 2) and CHCl3 (entry 3) were much faster with 54% and 46% ee respectively When THF (entry 4) was used, there was a slight decrease of ee

The best results were obtained with a polar solvent, CH3CN (88% isolated yield, 59% ee)

Thus, we believed that polar solvents would be preferred in the chiral tandem CA-E

reaction by solvating the starting materials and promoters well enough, and increasing the

nucleophilicity of the promoters to achieve reasonably high yields Nonetheless, when the

reaction was conducted in another polar solvent, DMSO (entry 6), the ee decreased to

aIsolated yield bDetermined by chiral HPLC analysis cVery slow reaction

Other CPSs (11f-h) were also investigated with CH3CN as the solvent (entries 7-9)

Optimization based on making structural changes to CPS revealed that the tert-butyl group

increases the enantioselectivity of the reaction (entry 7) When the pyrrolidine ring was

replaced by a piperidine, there was a slight decrease of both yield and ee (entry 8) As

expected, the most hindered CPS 11h promoted the reaction smoothly with the best yield

and ee value (entry 9) The loading of CPS promoter could be reduced to 1.5 equivalents while the yield and enantioselectivity were maintained at a satisfactory level (entry 10) We

also examine the reaction with only 1.1 equivalents CPS 11h, but the reaction did not go to

completion even after 28 hours

Table 2.6 1.5 equivalents 11h promoted tandem CA-E reaction between 3 and 16f under

aIsolated yield bDetermined by chiral HPLC analysis

It is known that in many cases of asymmetric catalysis, lowering temperature could

be beneficial to the selectivity However, the tandem CA-E reactions were generally slow

at low temperatures Therefore, we attempted several other conditions to reduce the reaction time and while maintaining the enantioselectivity To our surprise, the tandem CA-E reaction could proceed even at reflux condition but the ee only dropped 4% (Table 2.6, entry 1) Under microwave condition, 77% ee could be obtained in only 15 minutes However, the results in H2O were quite disappointing as no enantioselectivity was obtained though the reaction was fast (entry 3) A possible explanation for the low ee could be that H2O might have acted as a H-bond acceptor to form H-bonding with the NH

group of the promoter, thereby weakening any H-bonds formed between the CPS and 3

As a consequence, the ionic intermediate formed between the CPS and 3 would not be

locked in a particular conformation to favor the attack of the nucleophile from a

particular face

Table 2.7 11h promoted reaction between 3 and various diketones or ketoesters

+ COR2COR1

Entry 18[R1, R2] Product Time(hr) Yield/%a ee/%b

aIsolated yield bDetermined by chiral HPLC analysis cReaction at 40oC with 2 equiv

CPS 11h dReaction with 2 equiv CPS 11h

In the mean time, we were delighted to find that other kinds of 1,3-dicarbonyl

compounds 18a-h could be suitable donors for the reaction (Table 2.7) In entries 2,3,5 and 7, two equivalents CPS 11h were required as these reactions were generally slow In Table 2.7, tandem CA-E products 19a-h were obtained in moderate to good yields,

though the ee values were higher than that of S,S'-dialkyl dithiomalonates Those

reactions employing ketoesters (entries 4-10) afforded the products with diastereomeric ratios of approximately 1:1 In order to improve the yield of these reactions, we also

examined conditions at 40oC with 2 equivalents 11h (entry 3) Nevertheless, the reaction

still could not go to completion though the enantioselectivity was maintained Other 1,3-dicarbonyl compounds listed in Figure 2.5 were proved to be ineffective donors

Therefore, we focused on searching for other reactive donors such as bulky S,S'-dialkyl

dithiomalonates and keto-thioesters

Figure 2.5 Other unreactive donors for tandem CA-E reaction with 3

With the optimized conditions, a series of 1,3-dicarbonyl compounds were explored

as nucleophiles towards 3 (Table 2.8) Since S,S'-di-tert-butyl dithiomalonate could

provide very good enantioselectivity, we tested other bulkier S,S'-dialkyl dithiomalonates

such as 20a (entry 1) and 20b (entry 4) Both two donors reacted smoothly with high

yield and ee values However, the reaction of S,S'-di-adamantyl dithiomalonate 20b was

carried out in CH2Cl2 as the starting material was not soluble in CH3CN

We were concerned about the high amounts of CPS required for this reaction

Fortunately, the CPS promoter 11h could be recovered with 90% yield, using a simple

acid-base workup It can also be reused without further purification, for two further

cycles without loss of yield and enantioselectivity of the product 21a (entries 2,3) This

could be a milestone as less amounts of CPS are needed for developing a number of tandem CA-E reactions

Table 2.8 11h promoted tandem CA-E reaction between 3 and 1,3-dicarbonyl compounds

under optimized conditions

O Br

+ COR2COR1

aIsolated yield bDetermined by chiral HPLC analysis cRecovered 11h, 2nd cycle

dRecovered 11h, 3rd cycle eReaction in CH2Cl2

Other 1,3-dicarbonyl compounds such as keto-thioesters were proved to be effective donors for this reaction (entries 5-7) As discussed in Table 2.5, while the enantioselectivities obtained from ketoesters were generally high, the reactions always suffered from slow reaction rates and low yields We replaced ketoester with ketothioester,

which revealed the same concept using S,S'-dialkyl dithiomalonates The ketothioesters were prepared by heating 1,3-dioxin-4-ones with tert-butyl thiol22 When S-tert-butyl

ketothioesters such as 20c-e were employed, tandem CA-E reactions were significantly faster and products 21c-e were obtained in high yields and ees, with diastereomeric ratios

22 J.-i Sakaki, S Kobayashi, M Sato and C Kaneko, Chem Pham Bull., 1990, 38, 2262-2264