Download PDF
Synlett 2012(3): 453-357  
DOI: 10.1055/s-0031-1290316
LETTER
© Georg Thieme Verlag Stuttgart ˙ New York

Stereoselective Synthesis of vic-Halohydrins via l-tert-Leucine-Catalyzed syn-Selective Aldol Reaction

Atsushi Umehara, Takuya Kanemitsu, Kazuhiro Nagata, Takashi Itoh*
School of Pharmacy, Showa University, 1-5-8 Hatanodai, Shinagawa-ku, Tokyo 142-8555, Japan
Fax: +81(3)37845982; e-Mail: itoh-t@pharm.showa-u.ac.jp;

Further Information

Publication History

Received 9 November 2011
Publication Date:
19 January 2012 (online)

 PDF Download Permissions and Reprints All articles of this category

Abstract

l-tert-Leucine was found to be an effective organocatalyst for the asymmetric aldol reaction of chloroacetone. The stereoselective synthesis of vic-halohydrins was accomplished with excellent regioselectivity (>99%) to generate α-chloro-β-hydroxy ketones with high syn selectivity (syn/anti = 16:1) and enantioselectivity (up to 95% ee).

Key words

aldol reaction - amino acids - asymmetric synthesis - chloroacetone - organocatalysis

Optically active vic-halohydrins are versatile building blocks and key intermediates for the synthesis of biologically active compounds and natural products such as β-­agonists, [¹] substituted pyrrolidines, [²] a polychlorosulfolipid, [³] and insect sex pheromones. [4] Thus, the development of effective synthetic methods for the production of these compounds has attracted much attention. The most common procedure for the preparation of optically active vic-halohydrins is the ring opening of enantiomerically pure epoxides, [5] but this approach results in a mixture of regioisomers. On the other hand, asymmetric reductions of prochiral α-halo ketones by chiral oxazaborolidine catalysts with borane [¹b] [c] [²] [6] and asymmetric hydrogenation using chiral Ru [¹a] [7] or Rh [7c] [8] catalysts have shown excellent enantioselectivity to afford a variety of vic-halohydrins. Moreover, organocatalyzed direct asymmetric aldol reaction of chloroacetone with aldehydes has been reported by several groups. [9] We have been interested in an alternate approach, which provides α-chloro-β-hydroxy ketones, for obtaining vic-chlorohydrins. The development of an organocatalyst for the asymmetric aldol reaction of chloroacetone with aldehydes is a formidable task, since the reaction can lead to the formation of regio- and diastereomeric isomers and their enantiomers (Scheme  [¹] ). Nájera and co-workers reported an l-proline-derivative-catalyzed direct asymmetric aldol reaction of chloroacetone, which afforded anti-vic-chlorohydrins in high regio-, ­diastereo-, and enantioselectivity. [9b] Although Gong and co-worker have reported that primary amine catalyzed direct asymmetric aldol reactions of chloroacetone primarily provide syn products, the syn diastereoselectivity was very low. [9e] Despite recent advances in the area of catalytic asymmetric aldol reactions, [¹0] high regioselective, syn-dia­stereoselective, and enantioselective synthesis of α-chloro-β-hydroxy ketones remains challenging. In this paper, we report the highly syn-selective direct asymmetric aldol reaction of chloroacetone catalyzed by l-tert-leucine to afford syn-vic-harohydrins.

Scheme 1 Organocatalyzed asymmetric aldol reaction of chloroacetone (1a)

Initially, a test reaction was carried out using l-proline as a catalyst (20 mol%). The reaction of chloroacetone (1a) with 4-nitrobenzaldehyde (2a) at room temperature gave the anti-selective product 4a in low yield with low regio­selectivity (Scheme  [²] ). This anti selectivity was in agreement with literature data. [9b] Next, we examined the utility of a series of primary amino acids as catalysts for syn-selective aldol reaction of chloroacetone (1a). Natural primary amino acids and commercially available unnatural amino acids were used as catalysts in the reaction of chloroacetone (1a) with 4-nitrobenzaldehyde (2a) at room temperature (Table  [¹] ).

l-tert-Leucine exhibited much higher efficiency than the other primary amino acids, providing syn-aldol product 3a in high yield (89%) with excellent regioselectivity (>99:1), and high diastereo- (syn/anti = 7:1), and enantio­selectivity (83% ee, Table  [¹] , entry 1). Although l-alanine, l-valine, l-leucine, l-threonine, and l-phenylglycine exhibited excellent regioselectivity, only low reaction yields were obtained (Table  [¹] , entries 2-4, 9, and 11). l-Isoleucine, l-phenylalanine, l-tryptophan, and l-methionine gave the syn product 3a in moderate yields and diastereoselectivities (Table  [¹] , entries 5, 7, 8, and 10). l-Tyrosine and l-penicillamine showed no catalytic activity in this reaction (Table  [¹] , entries 6 and 12). All the primary amino acids, except l-tyrosine and l-penicillamine, afforded perfect regioselectivity and the desired aldol product 3a with syn selectivity; of these amino acids, l-tert-leucine was the most effective catalyst in the aldol reaction with chloroacetone (1a). [¹¹] We previously reported that l-tert-leucine exhibits high catalytic activity compared to other amino acids due to the stability of an enamine derived from l-tert-leucine in the presence of aromatic aldehydes. [¹²]

Scheme 2 l-Proline-catalyzed asymmetric aldol reaction of chloroacetone (1a)

Table 1 l-Amino Acid Catalyzed Asymmetric Aldol Reaction of Chloroacetone (1a)

Entrya Catalyst Yield (%)b (3a + 4a)/5a c 3a/4a c ee of 3a (%)d
 1 l-tert-leucine 89 >99:1 7:1 83
 2 l-alanine  7 >99:1 6:1 70
 3 l-valine 14 >99:1 5:1 80
 4 l-leucine 30 >99:1 7:1 73
 5 l-isoleucine 71 >99:1 2:1 38
 6 l-tyrosine trace n.d.e n.d.e n.d.e
 7 l-phenylalanine 66 >99:1 4:1 69
 8 l-tryptophan 64 >99:1 2:1 68
 9 l-threonine  6 >99:1 5:1 n.d.e
10 l-methionine 66 >99:1 4:1 72
11 l-phenylglycine 13 >99:1 3:1 18
12 l-penicillamine trace n.d.e n.d.e n.d.e
a The reaction was performed with 4-nitrobenzaldehyde (2a, 1 equiv), chloroacetonene (1a, 10 equiv), and l-amino acid (0.2 equiv) at r.t. for 7 d.
b The combined isolated yield of the diastereomers.
c Determined by ¹H NMR.
d Determined by HPLC.
e Not determined.

Next, to investigate the generality of this reaction, a variety of aromatic aldehydes 2a-i were used as substrates for the l-tert-leucine-catalyzed aldol reaction (Table  [²] ). [¹¹] In all cases, syn-vic-chlorohydrins 3a-i were obtained with excellent regioselectivity. When aldehydes with an electron-withdrawing substituent were employed, the reaction proceeded smoothly to afford desired syn adducts 3a-g with high stereoselectivities (Table  [²] , entries 1-7). In particular, 2-nitrobenzaldehyde afforded the syn product 3c with excellent yield (99%) and enantioselectivity (95% ee, Table  [²] , entry 3). The reaction of 2-bromobenzaldehyde resulted in the syn product 3f with high diastereo­selectivity (syn/anti = 10:1) (Table  [²] , entry 6). 2-Trifluoromethylbenzaldehyde gave the product 3g with high diastereoselectivity (syn/anti = 7:1) and excellent enantioselectivity (94% ee, Table  [²] , entry 7). On the other hand, benzaldehyde and 1-naphthaldehyde were less reactive and gave 3h and 3i in low yield, but the enantioselectivities remained high (Table  [²] , entries 8 and 9).

Table 2 l-tert-Leucine-Catalyzed Asymmetric Aldol Reaction of Chloroacetone (1a)

Entrya Ar Product (syn) Yield (%)b (3 + 4)/5 c 3/4 c ee of 3 (%)d
1 4-O2NC6H4 3a 89 >99:1  7:1 83
2 3-O2NC6H4 3b 87 >99:1  6:1 87
3 2-O2NC6H4 3c 99 >99:1  8:1 95
4 4-NCC6H4 3d 81 >99:1  5:1 80
5 4-O2CMeC6H4 3e 72 >99:1  7:1 83
6 2-BrC6H4 3f 84 >99:1 10:1 89
7 2-F3CC6H4 3g 80 >99:1  7:1 94
8 Ph 3h 33  94:6  5:1 87
9 1-naphthyl 3i 29 >99:1 16:1 90
a The reaction was performed with arylaldehyde 2 (1 equiv), chloroacetonene (1a, 10 equiv), and l-tert-leucine (0.2 equiv) at r.t. for 7 d.
b The combined isolated yield of the diastereomers.
c Determined by ¹H NMR.
d Determined by HPLC.
Table 3 l-tert-Leucine-Catalyzed Asymmetric Aldol Reaction of α-Substituted Acetone 1a-h

Entrya R Product (syn) Yield (%)b (3+4)/5 c 3/4 c ee of 3 (%)d
1 Cl 3a 89 >99:1 7:1 83
2 OH 3j 85 >99:1 2:1 46
3 OMe 3k 74  19:1 2:1 60
4e F 3l 23   4:1 3:1 85
5 SMe 3m 88  14:1 4:1 64
6 Br 3n trace n.d.f n.d.f n.d.f
7 Me 3o 83   5:1 2:1 68
8 CH2OH 3p 68   2:6 1:1 n.d.f
a Unless otherwise stated, the reaction was performed with 4-nitrobenzaldehyde (2a, 1 equiv), α-substituted acetone 1 (10 equiv), and l-tert-leucine (0.2 equiv) at r.t. for 7 d.
b The combined isolated yield of the diastereomers.
c Determined by ¹H NMR.
d Determined by HPLC.
e The reaction was performed in the presence of 0.35 equiv of l-tert-leucine.
f Not determined.

We then examined l-tert-leucine-catalyzed asymmetric aldol reactions of 4-nitrobenzaldehyde (2a) with α-heteroatom-substituted ketones, 2-butanone, or 4-hydroxy-2-butanone (Table  [³] ). [¹¹] Although hydroxyacetone smoothly underwent the reaction to give syn adduct 3j with excellent regioselectivity, the enantioselectivity was low (Table  [³] , entry 2). Methoxyacetone and methylthioacetone gave syn adducts 3k and 3m, respectively, as a major product with good regioselectivities, but the stereoselectivities were moderate (Table  [³] , entries 3 and 5). The use of fluoroacetone resulted in high enantioselectivity, but the chemical yield was low (Table  [³] , entry 4). In the case of bromoacetone, only trace amounts of the desired product 3n were obtained (Table  [³] , entry 6). Furthermore, the reactions of 2-butanone and 4-hydroxy-2-butanone, which lack a functional group at the α-position, afforded syn adducts 3o and 3p with low and no diastereoselectivity, respectively (Table  [³] , entries 7 and 8).

In previous reports, direct asymmetric aldol reaction of α-hydroxy ketones catalyzed by a secondary amine such as l-proline or its derivatives provided anti-selective aldol adducts, [¹³] whereas the reaction catalyzed by a chiral primary amine provided syn-selective aldol adducts. [9e] [¹4] According to the model proposed by Barbas, [¹4a] in the case of a secondary amine, E-enamine intermediates predominate because of steric repulsion. As a result, secondary amine catalyzed aldol reactions of α-hydroxy ketone 1b afford anti-aldol adducts 4. In contrast, Z-emanine intermediates predominate in the case of a primary amine because of stabilization of the intermediate due to the formation of an intramolecular hydrogen bond. Therefore, primary amine catalyzed reactions afford syn-aldol adducts 3. For the reaction of chloroacetone (1a), we propose a mechanism similar to the case of α-hydroxy ketone (Scheme  [³] ). [¹5] According to the results in Table  [²] , a hydrogen bond can be formed in an enamine generated from a primary amine and chloroacetone (1a). The asymmetric aldol reaction provides syn adducts 3 with high syn diastereoselectivities in the presence of appropriate chiral primary amine catalysts. In addition, the results in Table  [³] suggest that intramolecular hydrogen bonding is necessary for the reaction to proceed in high regio- and stereoselectivity, and only chloro and methylthio groups at the α-position can effectively bond with the neighboring hydrogen. Thus, sufficient electronegativity and an appropriate atomic radius might be prerequisite for the formation of the intramolecular hydrogen bonds necessary to afford syn selectivity.

Scheme 3  A proposed mechanism for the organocatalyzed asymmetric aldol reaction of chloroacetone (1a)

In summary, we have developed a protocol for the highly stereoselective organocatalyzed synthesis of vic-halohydrins. Using l-tert-leucine as the catalyst, the aldol reactions of chloroacetone with aromatic aldehydes were accomplished with excellent regioselectivity (>99%) to generate the corresponding α-chloro-β-hydroxy ketones with high syn selectivity (up to syn/anti = 16:1) and high enantioselectivity (up to 95% ee). Thus, l-tert-leucine-catalyzed reactions hold promise for the synthesis of optically active vic-chlorohydrins. Further applications of this approach to total synthesis are currently under investigation in our laboratory.

Supporting Information for this article is available online at http://www.thieme-connect.com.accesdistant.sorbonne-universite.fr/ejournals/toc/synlett.

Acknowledgment

This work was supported in part by a Grant for the High-Tech ­Research Center Project from the Ministry of Education, Culture, Sports, Science and Technology of Japan.

11

Optimized Procedure for the Synthesis of 3a To a mixture of chloroacetone (1a, 400 µL, 5 mmol) and l-tert-leucine (13 mg, 0.1 mmol), 4-nitrobenzaldehyde (2a, 76 mg, 0.5 mmol) was added, and the mixture was stirred at r.t. The reaction was monitored by TLC analysis. After 7 d, H2O was added and extracted with CH2Cl2 (3×), dried over MgSO4, and concentrated in vacuo. To determine the regioselectivity and the diastereomeric ratio, the remaining residue was analyzed by ¹H NMR. Moreover, the ee value of the product 3a was determined by chiral-phase HPLC analysis of the residue. Then, the residue was purified by column chromatography on silica gel in gradient elution with hexane-EtOAc to give a 7:1 inseparable mixture of the desired products 3a and 4a (108 mg, 89%).
Analytical Data for Compound 3a
¹H NMR (400 MHz, CDCl3): δ = 8.24-8.20 (m, 2 H), 7.61-7.58 (m, 2 H), 5.47 (t, J = 3.6 Hz, 1 H), 4.46 (d, J = 3.2 Hz, 1 H), 3.43 (d, J = 4.0 Hz, 1 H), 2.40 (s, 3 H). ¹³C NMR (100 MHz, CDCl3): δ = 203.4, 147.6, 146.2, 127.3, 123.5, 72.0, 67.3, 28.3. HPLC: 83% ee [Daicel CHRALCEL OJ-H, hexane-iPrOH (9:1), flow rate 1.0 mL/min, λ = 254 nm]: t R(major) = 34.7 min; t R(minor) = 39.1 min.

11

Optimized Procedure for the Synthesis of 3a To a mixture of chloroacetone (1a, 400 µL, 5 mmol) and l-tert-leucine (13 mg, 0.1 mmol), 4-nitrobenzaldehyde (2a, 76 mg, 0.5 mmol) was added, and the mixture was stirred at r.t. The reaction was monitored by TLC analysis. After 7 d, H2O was added and extracted with CH2Cl2 (3×), dried over MgSO4, and concentrated in vacuo. To determine the regioselectivity and the diastereomeric ratio, the remaining residue was analyzed by ¹H NMR. Moreover, the ee value of the product 3a was determined by chiral-phase HPLC analysis of the residue. Then, the residue was purified by column chromatography on silica gel in gradient elution with hexane-EtOAc to give a 7:1 inseparable mixture of the desired products 3a and 4a (108 mg, 89%).
Analytical Data for Compound 3a
¹H NMR (400 MHz, CDCl3): δ = 8.24-8.20 (m, 2 H), 7.61-7.58 (m, 2 H), 5.47 (t, J = 3.6 Hz, 1 H), 4.46 (d, J = 3.2 Hz, 1 H), 3.43 (d, J = 4.0 Hz, 1 H), 2.40 (s, 3 H). ¹³C NMR (100 MHz, CDCl3): δ = 203.4, 147.6, 146.2, 127.3, 123.5, 72.0, 67.3, 28.3. HPLC: 83% ee [Daicel CHRALCEL OJ-H, hexane-iPrOH (9:1), flow rate 1.0 mL/min, λ = 254 nm]: t R(major) = 34.7 min; t R(minor) = 39.1 min.

   Permissions and Reprints All articles of this category

Scheme 1 Organocatalyzed asymmetric aldol reaction of chloroacetone (1a)

Scheme 2 l-Proline-catalyzed asymmetric aldol reaction of chloroacetone (1a)

Scheme 3  A proposed mechanism for the organocatalyzed asymmetric aldol reaction of chloroacetone (1a)