Catalytic asymmetric addition of diorganozinc reagents to N ... - PNAS

Download PDF
2 downloads 0 Views 329KB Size Report
Comment
Apr 13, 2004 - enantioselectivities in the presence of a catalytic amount of the novel Me-DuPHOS monoxide [(R,R)-BozPHOS] chiral ligand. (Scheme 1) (18 ...
SPECIAL FEATURE

Catalytic asymmetric addition of diorganozinc reagents to N-phosphinoylalkylimines Alexandre Coˆte´, Alessandro A. Boezio, and Andre´ B. Charette* Department of Chemistry, University of Montreal, P.O. Box 6128, Station Downtown, Montreal, QC, Canada H3C 3J7 Edited by Jack Halpern, University of Chicago, Chicago, IL, and approved February 4, 2004 (received for review October 31, 2003)

he synthesis of ␣-chiral amines using the catalytic asymmetric addition of diorganozinc reagents has produced very exciting results in recent years (1–3). This very important subunit is commonly found in many pharmaceuticals and other biologically important compounds. Some specific examples include Flomax (antihypertensive) and Mexitil (antiarrhythmics) (Fig. 1) (4). The most important methodologies developed to prepare ␣-chiral amines in enantiomerically pure form rely on either chemical resolution or on the use of readily available chiral synthons as building blocks. More recently, the development of efficient chiral auxiliaries has led to the extensive use of chiral imines as precursors to ␣-chiral amines by alkylation chemistry (5–9) or hydrogenation (10). Catalytic asymmetric nucleophilic addition reactions of organometallic reagents to imines have been reported with Ntosylimines (11, 12), N-arylimines (13–15), and N-acylimines (16, 17) as amine precursors. The cleavage of the protecting group leads to the ␣-chiral amine; however, in the former two cases, harsh conditions are usually necessary [N-tosyl, SmI2; N-aryl, ceric ammonium nitrate or PhI(OAc)2; N-formyl, H3O⫹, heat]. Furthermore, the reactions usually require excess of the diorganozinc reagent, and in a number of cases, the addition of dimethylzinc is unsuccessful or affords to the amine in much lower yields and enantiomeric excesses (ee). Recently, we reported that the copper-catalyzed addition of diorganozinc reagents to N-phosphinoylarylimines proceeds in high yields and enantioselectivities in the presence of a catalytic amount of the novel Me-DuPHOS monoxide [(R,R)-BozPHOS] chiral ligand (Scheme 1) (18, 19). In addition to being readily available, this hemilabile bidentate ligand offers superb catalytic activity, broad substrate generality, and mild reaction conditions. Furthermore, the increase in imine electrophilicity imparted by the N-phosphinoyl protecting group, combined with its ease of cleavage under mildly acidic conditions, makes this method very attractive to prepare ␣-chiral amines. The one major limitation of this methodology is our inability to prepare alkylaldehyde-derived N-phosphinoylimines bearing ␣-enolizable protons in reasonable yields. This has been a limitation not only in diorganozinc addition chemistry (20–25), but also in catalytic asymmetric nitro-Mannich (26), Mannich (27), and Strecker (28) processes that use similar electrophilic precursors. This inherent limitation is also present in the preparation of N-acylimines, and several strategies have been developed to circumvent this problem. The in situ generation of

T

www.pnas.org兾cgi兾doi兾10.1073兾pnas.0307096101

Fig. 1.

Bioactive ␣-chiral amines.

sensitive imines from stable precursors has been a strategy that has been quite successful in a number of cases. Typically, a stable imine adduct is used as a precursor and is converted to the imine in situ (Scheme 2). The method involves the use of a leaving group (LG) on the ␣-carbon of the N-protected amine. Several leaving groups including benzotriazolates (29), sulfinates (30), and succinimidates (31) have been used for the in situ preparation of N-acylimines. In this article, we describe the feasibility of the in situ synthesis of N-phosphinoylimines and we show that these precursors are also suitable reaction partners for the preparation of ␣-chiral amines in high yields and ee in the (R,R)-BozPHOS䡠Cu catalyzed addition of diorganozinc reagents (Scheme 3). Experimental Methods General. All nonaqueous reactions were run under an inert atmosphere (nitrogen or argon) with rigid exclusion of moisture from reagents and glassware by using standard techniques for manipulating air-sensitive compounds. All glassware was stored in the oven and兾or was flame-dried before use under an inert atmosphere of gas. Anhydrous solvents were obtained either by filtration through drying columns (ether, toluene) on a GlassContour system (Irvine, CA) or by distillation over sodium兾 benzophenone (toluene). Analytical TLC was performed on precoated, glass-backed silica gel (Merck 60 F254). Visualization of the developed chromatogram was performed by UV absorbance, aqueous cerium molybdate, ethanolic phosphomolybdic acid, iodine, or aqueous potassium permanganate. Flash column chromatography was performed by using 230–400 mesh silica [EM Science or Silicycle (Quebec City, QC, Canada)] of the indicated solvent system according to standard technique. Melting points were obtained on a Buchi (Flawil, Switzerland) melting-point apparatus and are uncorrected. NMR spectra (1H, 13C, 31P) were recorded on Bruker AV 300, AMX 300, AV 400, or ARX 400 spectrometers. Chemical shifts for 1H NMR spectra are recorded in ppm with the solvent resonance as the internal standard (chloroform, ␦ 7.27 ppm, or DMSO, ␦ 2.50 ppm). Data are reported as follows: chemical shift, multiplicity (s, singlet; d, doublet; t, triplet; q, quartet; qn, quintet; sept, septuplet; This paper was submitted directly (Track II) to the PNAS office. Abbreviation: ee, enantiomeric excess. *To whom correspondence should be addressed. E-mail: [email protected]. © 2004 by The National Academy of Sciences of the USA

PNAS 兩 April 13, 2004 兩 vol. 101 兩 no. 15 兩 5405–5410

CHEMISTRY

The synthesis of ␣-chiral amines bearing two alkyl groups has been hampered by the accessibility and stability of the alkylimine precursor. Herein, we report an efficient strategy to generate the alkyl-substituted imine in situ that is compatible with the MeDuPHOS monoxide䡠Cu(I) catalyzed addition of diorganozinc reagents. The sulfinic acid adduct of the imine is readily prepared by mixing diphenylphosphinic amide, the aldehyde, and sulfinic acid. The sulfinic acid adduct is generally isolated by filtration. The addition of diorganozinc reagents in the presence of Me-DuPHOS monoxide䡠Cu(I) and the in situ-generated imines affords the corresponding ␣-chiral amines in high yields and enantiomeric excesses.

Scheme 3.

1 h at room temperature, diethylzinc (205 ␮l, 2.0 mmol, 2.5 eq) was added, and the resulting dark-brown suspension was stirred for 20 min and cooled to the desired temperature. A cold suspension of sulfinyl adduct (0.80 mmol, 1.0 eq) in toluene (3 ml ⫹ 2 ml to rinse ⫹ 1 ml to rinse) was transferred (Teflon cannula; o.d., 1兾8th of an inch; i.d., 1兾16th of an inch) to the catalyst solution. After stirring, the reaction was quenched with saturated aqueous NH4Cl and the aqueous layer was extracted with CH2Cl2 (3 ⫻ 40 ml). The combined organic layers were dried over Na2SO4. Concentration and purification by silica gel column chromatography (100% EtOAc) gave the pure protected amine. Scheme 1.

m, multiplet; br, broad), coupling constant in Hz, and integration. Chemical shifts for 13C NMR spectra are recorded in ppm from tetramethylsilane by using the central peak of deuterochloroform (77.00 ppm) or deuterated DMSO (39.52 ppm) as the internal standard. All spectra were obtained with complete proton decoupling. Optical rotations were determined with a Perkin–Elmer 341 polarimeter at 589 or 546 nm. Data are reported as follows: [␣]␭temp, concentration (c in g兾100 ml), and solvent. High-resolution mass spectra were performed by the Centre re´gional de spectroscopie de masse of the University of Montreal. Combustion analyses were performed by the Elemental Analysis laboratory of the University of Montreal. Analytical HPLC was performed on a Hewlett–Packard analytical instrument (model 1100) equipped with a diode array UV detector. Data for determination of the enantiomeric excess is reported as follows: column type, eluent, flow rate, and retention time (tr). Reagents. Cu(OTf)2 was purchased from Strem Chemicals (New-

buryport, MA). All starting materials were purchased from Aldrich or Alfa Aesar (Ward Hill, MA). Unless otherwise stated, commercial reagents were used without purification. Diethylzinc was purchased neat from Akzo Nobel and used without purification. Racemic samples for HPLC analysis were prepared by addition of Et2Zn兾CuCN in toluene to the sulfinyl adducts. Ligand (R,R)-BozPHOS was prepared according to literature procedure (18). General Procedure for Diethylzinc Addition on Sulfinyl Adducts.

(R,R)-BozPHOS (13 mg, 0.040 mmol, 0.050 eq) and Cu(OTf)2 (13 mg, 0.036 mmol, 0.045 eq) were dissolved in toluene (3 ml). The resulting dark-green heterogeneous solution was stirred for

Scheme 2. 5406 兩 www.pnas.org兾cgi兾doi兾10.1073兾pnas.0307096101

p-Toluenesulfinic Acid. p-Toluenesulfinic acid was formed by dissolving its hydrated sodium salt in a minimum of hot 10% vol兾vol HCl (the resulting pH must be lower than 3) and crystallization at 4°C. Further filtration and drying under vacuum led to white crystals. General Procedure for the Synthesis of Sulfinyl Adducts. To a sus-

pension of P,P-diphenylphosphinic amide (1.00 g, 4.6 mmol, 1 eq) and sulfinic acid (1.08 g, 6.9 mmol, 1.5 eq) in anhydrous diethylether (40 ml, ACS grade) was added the freshly distilled aldehyde (6.9 mmol, 1.5 eq) at room temperature. The mixture was stirred for 15 h, during which a white precipitate was slowly formed. On completion of the reaction, the solution was filtered, and the white solid was washed with anhydrous diethylether (15 ml, ACS grade) and dried under vacuum. N-[(4-methylphenyl)sulfonyl](phenyl)methyl]-P,P -diphenylphosphinic amide (3). The general procedure was followed. The crude

compound (71% yield) was used without purification for the next step: (white powder) mp 122–124°C (dec.) HRMS (FAB) m兾z calc. for C26H25NO3PS [M ⫹H]⫹: 462.1293 found: 462.1305. P,P-diphenyl-N-[(1S)-1-phenylpropyl]phosphinic amide. The product 2 was obtained following the general procedure (specific conditions: 0°C for 24 h). Yield: 87%. The ee (97%) was determined by HPLC analysis [Chiralpak AD, 80:20 hexanes:i-PrOH, 1.0 ml兾min: (R)-2 tr ⫽ 9.6 min, (S)-2 tr ⫽ 13.0 min]. The S configuration was assigned by comparison of the optical rotation of the deprotected amine hydrochloride with the literature (32). N-[1-methyl-(toluene-4-sulfonyl)-methyl]-P,P-diphenylphosphinic amide.

The general procedure was followed. The crude compound 9 (91% yield) was used without purification for the next step: (white powder) mp 120–121°C (dec.); 1H NMR (DMSO-d6, 400 MHz) ␦ 7.68–7.63 (m, 4H), 7.54–7.47 (m, 4H), 7.37–7.29 (m, 6H), 6.37 (t, J ⫽ 10.6 Hz, 1H), 4.46 (m, 1H), 2.41 (s, 3H), 1.48 (d, J ⫽ 6.8 Hz, 3H); 13C NMR (DMSO-d6, 75 MHz) ␦ 145.2, 135.0 (d, JC-P ⫽ 68.4 Hz), 134.1, 133.3 (d, JC-P ⫽ 72.7 Hz), 132.6 (d, JC-P ⫽ 2.7 Hz), 132.3 (d, JC-P ⫽ 3.0 Hz), 132.2 (d, JC-P ⫽ 10.2 Hz), 132.0 (d, JC-P ⫽ 10.2 Hz), 130.4, 130.2, 129.2 (d, JC-P ⫽ 12.5 Hz), 129.0 (d, JC-P ⫽ 12.6 Hz), 69.0, 21.9, 16.6 (d, JC-P ⫽ 1.9 Hz); 31P NMR (DMSO-d6, 162 MHz) ␦ 26.0; HRMS (APCI) m兾z [MSO2Tol]⫹ calc.: 244.1 found: 244.2; Anal. calc. for C21H22NO3PS: C 63.14, H 5.55, N 3.51 found: C 63.10, H 5.70, N 3.60. Coˆ te´ et al.

N-{2-methyl-1-[(4-methylphenyl)sulfonyl]propyl}-P,P-diphenylphosphinic amide. The general procedure was followed. The crude com-

pound 8 (88% yield) was used without purification for the next step: (white powder) mp 116–117°C (dec.); 1H NMR (DMSO-d6, 400 MHz) 7.81–7.69 (m, 2H), 7.58 (d, J ⫽ 8.3 Hz, 2H), 7.54–7.43 (m, 4H), 7.38–7.25 (m, 4H), 7.22 (d, J ⫽ 8.3 Hz, 2H), 6.25 (t, J ⫽ 13.0 Hz, 1H), 4.56 (dt, J ⫽ 11.8, 2.9 Hz, 1H), 2.49 (m, 1H), 2.35 (s, 3H), 1.03 (d, J ⫽ 6.8 Hz, 3H), 0.90 (d, J ⫽ 7.0 Hz, 3H); LRMS (APCI) m兾z [M - SO2Tol]⫹ calc.: 272.1 found: 272.1; Anal. calc. for C23H26NO3PS: C 64.62, H 6.13, N 3.28 found: C 64.61, H 6.32, N 3.39.

N-[(1S)-1-ethyl-2-methylpropyl]-P,P-diphenylphosphinic amide. The product 15 was obtained by following the general procedure (specific conditions: room temperature for 24 h). Yield 86%, ee (96%) was determined by HPLC analysis [Chiralpak AD-H, 90:10 hexanes:i-PrOH, 1.0 ml兾min: (S)-15 tr ⫽ 13.1 min, (R)-15 20 tr ⫽ 14.7 min]. mp 135–136°C; Rf 0.43 (100% AcOEt); [␣]D ⫹7.7° 1 (c 1.04, CHCl3); H NMR (300 MHz, CDCl3) ␦ 7.92 (m, 4H), 7.45 (m, 6H), 2.82 (br s, 1H), 2.70 (br s, 1H), 1.88 (m, 1H), 1.52 (m, 2H), 0.89 (m, 9H); 13C NMR (75 MHz, CHCl3) ␦ 134.3 (d, JC-P ⫽ 10.3 Hz), 132.7 (d, JC-P ⫽ 3.4 Hz), 132.6 (d, JC-P ⫽ 3.3 Hz), 132.2 (d, JC-P ⫽ 9.4 Hz), 132.1 (d, JC-P ⫽ 1.0 Hz), 132.1 (d, JC-P ⫽ 1.0 Hz), 128.9 (d, JC-P ⫽ 2.9 Hz), 128.7 (d, JC-P ⫽ 2.9 Hz), 58.6 (d, JC-P ⫽ 2.2 Hz), 32.1 (d, JC-P ⫽ 8.7 Hz), 26.7 (d, JC-P ⫽ 14.0 Hz), 19.0, 18.2, 10.9; 31P NMR (121 MHz, CDCl3) ␦ 23.4; HRMS (MAB) m兾z calc. for C18H24NOP [M]: 301.1587 found 301.1596. N-{3-methyl-1-[(4-methylphenyl)sulfonyl]butyl} -P,P -diphenylphosphinic amide. The general procedure was followed. The crude

compound 10 (84% yield) was used without purification for the next step: (white powder) mp 112–114°C (dec.); 1H NMR (400 MHz, DMSO-d6) ␦ 7.72–7.67 (m, 2H), 7.68 (d, J ⫽ 8.2 Hz, 2H), 7.53–7.40 (m, 8H), 7.28 (d, J ⫽ 8.3 Hz, 2H), 6.32 (dd, J ⫽ 13.7, 11.0 Hz, 1H), 4.49 (ddd, J ⫽ 23.0, 11.1, 3.8 Hz, 1H), 2.36 (s, 3H), 1.79–1.57 (m, 3H), 0.79 (d, J ⫽ 6.2 Hz, 3H), 0.64 (d, J ⫽ 6.0 Hz, 3H); 13C NMR (100 MHz, CHCl3) 135.6 (d, JC-P ⫽ 90.9 Hz), 134.5, 134.4 (d, JC-P ⫽ 94.3 Hz), 132.8 (d, JC-P ⫽ 14.6 Hz), 132.0 (d, JC-P ⫽ 10.4 Hz), 130.5 (d, JC-P ⫽ 2.8 Hz), 130.1, 129.2 (d, JC-P ⫽ 11.8 Hz), 129.0 (d, JC-P ⫽ 12.5 Hz), 72.2 (d, JC-P ⫽ 10.4 Hz), 38.6, 24.6, 24.3, 22.1, 21.4; 31P NMR (162 MHz, CDCl3) ␦ 26.0; LRMS (APCI) m兾z [M - SO2Tol]⫹ calc.: 286.1 found: 286.1; Anal. calc. for C24H28NO3PS: C 65.29, H 6.39, N 3.17 found: C 65.18, H 6.36, N 3.20. N-[(1S)-1-ethyl-3-methylbutyl]-P,P-diphenylphosphinic amide. The

product 17 was obtained by following the general procedure (specific conditions: ⫺20°C for 16 h). Yield 97%, ee (96%) was determined by HPLC analysis [Chiralpak AD-H, 80:20 hexanes:i-PrOH, 1.0 ml兾min: (S)-17 tr ⫽ 7.2 min, (R)-17 tr ⫽ 11.0 20 min]. mp 103–104°C; Rf 0.49 (100% AcOEt); [␣]D ⫹16.0° (c 1.05,

Coˆ te´ et al.

SPECIAL FEATURE

16 was obtained by following the general procedure (specific conditions: ⫺60°C for 48 h). Yield 97%, ee (90%) was determined by HPLC analysis [Chiralpak AD-H, 80:20 hexanes:iPrOH, 1.0 ml兾min: (S)-16 tr ⫽ 7.8 min, (R)-16 tr ⫽ 8.7 min]. mp 20 128–130°C; Rf 0.25 (100% AcOEt); [␣]D ⫺18.7° (c 1.27, CHCl3); 1H NMR (300 MHz, CDCl ) ␦ 7.94–7.83 (m, 4H), 7.48–7.36 (m, 3 6H), 3.11 (m, 1H), 2.73 (br s, 1H), 1.59 (m, 1H), 1.46 (m, 1H), 1.19 (d, J ⫽ 6.4 Hz, 3H), 0.88 (t, J ⫽ 7.4 Hz, 3H); 13C NMR (75 MHz, CHCl3) ␦ 133.9 (d, JC-P ⫽ 8.6 Hz), 132.1 (d, JC-P ⫽ 8.6 Hz), 132.0 (d, JC-P ⫽ 4.6 Hz), 131.9 (d, JC-P ⫽ 4.6 Hz), 131.5, 128.4, 128.2, 48.8 (d, JC-P ⫽ 1.7 Hz), 32.2 (d, JC-P ⫽ 6.1 Hz), 23.1 (d, JC-P ⫽ 4.4 Hz), 10.2; 31P NMR (162 MHz, CDCl3) ␦ 23.1; LRMS (APCI) m兾z [M ⫹ H]⫹ calc.: 274.2 found: 274.1; Anal. calc. for C16H20NOP: C 70.31, H 7.38, N 5.12 found: C 70.32, H 7.44, N 5.15.

CHCl3); 1H NMR (300 MHz, CDCl3) ␦ 7.93 (dd, J ⫽ 8.0, 1.6 Hz, 2H), 7.89 (dd, J ⫽ 8.1, 1.6 Hz, 2H), 7.53–7.38 (m, 6H), 3.04 (m, 1H), 2.64 (br s, 1H), 1.75 (sept, J ⫽ 6.6 Hz, 1H), 1.66–147 (m, 2H), 1.47–1.23 (m, 2H), 0.90 (t, J ⫽ 7.4 Hz, 3H), 0.78 (t, J ⫽ 6.9 Hz, 6H); 13C NMR (75 MHz, CHCl3) ␦ 134.3 (d, JC-P ⫽ 20.0 Hz), 133.7 (d, JC-P ⫽ 9.3 Hz), 132.6 (d, JC-P ⫽ 9.3 Hz), 132.1 (d, JC-P ⫽ 1.4 Hz), 128.9 (d, JC-P ⫽ 1.1 Hz), 128.7 (d, JC-P ⫽ 1.2 Hz), 51.2 (d, JC-P ⫽ 2.0 Hz), 46.5 (d, JC-P ⫽ 5.9 Hz), 30.2 (d, JC-P ⫽ 4.0 Hz), 25.1, 23.2, 23.0, 9.7; 31P NMR (121 MHz, CDCl3) ␦ 22.5; LRMS (APCI) m兾z [M ⫹ H]⫹ calc.: 316.4 found: 316.1; Anal. calc. for C19H26NOP: C 72.36, H 8.31, N 4.44 found: C 72.10, H 8.54, N 4.46. N -{cyclopentyl[(4-methylphenyl)sulfonyl]methyl}-P,P -diphenylphosphinic amide. The general procedure was followed. The crude

compound 4 (95% yield) was used without purification for the next step: (white powder) mp 111–113°C (dec.); 1H NMR (DMSO-d6, 300 MHz) 7.77–7.65 (dd, J ⫽ 11.3, 7.7 Hz, 2H), 7.56 (d, J ⫽ 7.7 Hz, 2H), 7.50–7.40 (m, 4H), 7.38–7.29 (m, 4H), 7.18 (d, J ⫽ 7.7 Hz, 2H), 6.31 (t, J ⫽ 13.1 Hz, 1H), 4.66 (dt, J ⫽ 11.5, 5.4 Hz, 1H), 2.31 (s, 3H), 1.72–1.55 (m, 2H), 1.50–1.15 (m, 7H); 13C NMR (DMSO-d , 75 MHz) ␦ 144.8, 136.1 (d, J 6 C-P ⫽ 59.9 Hz), 135.4, 134.4 (d, JC-P ⫽ 62.0 Hz), 132.7 (d, JC-P ⫽ 2.5 Hz), 132.3 (d, JC-P ⫽ 2.4 Hz), 131.8 (d, JC-P ⫽ 11.3 Hz), 131.6 (d, JC-P ⫽ 10.4 Hz), 130.2, 129.6, 129.0 (d, JC-P ⫽ 12.5 Hz), 128.7 (d, JC-P ⫽ 12.8 Hz), 75.8, 30.4, 28.7, 25.7, 25.5, 22.0; 31P NMR (DMSO-d6, 162 MHz) ␦ 25.0; HRMS (FAB) m兾z calc. for C25H29NO3PS [M ⫹ H]⫹: 454.1606 found: 454.1601. N-[(1S)-1-cyclopentylpropyl]-P,P-diphenylphosphinic amide. The

product 12 was obtained by following the general procedure (specific conditions: 0°C for 24 h). Yield 92%, ee (95%) was determined by HPLC analysis [Chiralpak AD-H, 80:20 hexanes:i-PrOH, 1.0 ml兾min: (S)-12 tr ⫽ 7.3 min, (R)-12 tr ⫽ 9.2 min]. The S configuration is tentatively assigned based on compounds 3, 5, 8, 9 and 10. mp 145–146°C; Rf 0.45 (100% 20 AcOEt); [␣]D ⫹8.8° (c 1.03, CHCl3); 1H NMR (300 MHz, CDCl3) ␦ 7.90 (dt, J ⫽ 11.7, 5.9 Hz, 4H), 7.46 (m, 6H), 2.92 (br s, 1H), 2.72 (br s, 1H), 1.96 (m, 1H), 1.82 (m, 1H), 1.58 (m, 7H), 1.27 (m, 2H), 0.91 (t, J ⫽ 7.4 Hz, 3H); 13C NMR (75 MHz, CHCl3) ␦ 134.4 (d, JC-P ⫽ 33.8 Hz), 132.7 (d, JC-P ⫽ 6.2 Hz), 132.7 (d, JC-P ⫽ 32.5 Hz), 132.6 (d, JC-P ⫽ 6.2 Hz), 132.1 (d, JC-P ⫽ 2.3 Hz), 132.0 (d, JC-P ⫽ 2.3 Hz), 128.9, 128.7, 56.9 (d, JC-P ⫽ 2.3 Hz), 45.0 (d, JC-P ⫽ 6.3 Hz), 29.8 (d, JC-P ⫽ 8.8 Hz), 28.6 (d, JC-P ⫽ 3.2 Hz), 25.8 (d, JC-P ⫽ 14.0 Hz), 9.5; 31P NMR (121 MHz, CDCl3) ␦ 22.5; LRMS (APCI) m兾z [M ⫹ H]⫹ calc.: 328.2 found: 328.2; Anal. calc. for C20H26NOP: C 73.37, H 8.00, N 4.28 found: C 73.37, H 8.14, N 3.93. N -{1-[ (4-methylphenyl)sulfonyl]-3-phenylpropyl}-P,P -diphenylphosphinic amide. The general procedure was followed. The crude

compound 5 (97% yield) was used without purification for the next step: (white powder) mp 118–120°C (dec.); 1H NMR (DMSO-d6, 300 MHz) ␦ 7.78 (m, 2H), 7.49 (m, 10H), 7.28 (d, J ⫽ 8.2 Hz, 2H), 7.18 (m, 3H), 7.0 (d, J ⫽ 6.9 Hz, 2H), 6.45 (t, J ⫽ 11.6 Hz, 1H), 4.39 (tdd, J ⫽ 21.2, 21.2, 2.1 Hz, 1H), 2.65 (m, 1H), 2.48 (m, 1H), 2.35 (s, 3H), 2.27 (m, 1H), 1.93 (␮, 1H); 13C NMR (DMSO-d6, 75 MHz) 145.3, 141.4, 136.0, 135.1, 134.3, 133.4, 132.6 (J ⫽ 2.4 Hz), 132.4 (d, J ⫽ 2.4 Hz), 132.3, 132.2, 132.1, 132.0, 130.5, 130.1, 129.4, 129.3, 129.2, 129.2, 129.1, 129.0, 128.9, 127.0, 72.9, 31.9, 31.7, 22.0; 31P NMR (DMSO-d6, 121 MHz) ␦ 24.7. LRMS (APCI) m兾z [M - SO2Tol]⫹ calc.: 334.1 found: 334.1; Anal. calc. for C28H28NO3PS: C 68.69, H 5.76, N 2.86 found: C 68.56, H 5.94, N 2.93. N-[(1S)-1-ethyl-3-phenylpropyl]-P,P-diphenylphosphinic amide. The product 11 was obtained by following the general procedure (specific conditions: ⫺20°C for 16 h). Yield 98%, ee (96%) was PNAS 兩 April 13, 2004 兩 vol. 101 兩 no. 15 兩 5407

CHEMISTRY

N-[(1R)-1-methylpropyl]-P,P-diphenylphosphinic amide. The product

determined by HPLC analysis [Chiralpak AD-H, 80:20 hexanes:i-PrOH, 1.0 ml兾min: (R)-11 tr ⫽ 9.2 min, (S)-11 tr ⫽ 12.8 20 min]. mp 138–139°C; Rf 0.54 (100% AcOEt); [␣]D ⫺20.3° (c 1 1.32, CHCl3); H NMR (300 MHz, CDCl3) ␦ 7.91 (dt, J ⫽ 11.6, 5.9 Hz, 4H), 7.52–7.39 (m, 6H), 7.28–7.21 (m, 2H), 7.20–7.13 (m, 3H), 3.09 (br s, 1H), 2.85–2.56 (m, 3H), 1.85 (q, J ⫽ 7.4 Hz, 2H), 1.64 (qn, J ⫽ 6.9 Hz, 2H), 0.93 (t, J ⫽ 7.4 Hz, 3H); 13C NMR (75 MHz, CHCl3) ␦ 142.4, 134.3 (d, JC-P ⫽ 2.1 Hz), 133.6 (d, JC-P ⫽ 6.0 Hz), 132.6 (d, JC-P ⫽ 5.9 Hz), 132.1 (d, JC-P ⫽ 2.2 Hz), 128.9 (d, JC-P ⫽ 1.1 Hz), 128.8 (d, JC-P ⫽ 0.7 Hz), 126.2, 53.0 (d, JC-P ⫽ 1.5 Hz), 38.3 (d, JC-P ⫽ 5.2 Hz), 32.4, 29.9 (d, JC-P ⫽ 4.5 Hz), 10.1; 31P NMR (121 MHz, CDCl3) ␦ 22.9; HRMS (MAB) m兾z calc. for C23H26NOP [M] calc.: 363.1752 found: 363.1748.

Table 1. Effect of additives in the catalytic asymmetric synthesis of amines

Entry

Additive (LG-ZnEt)

Yield*, %

ee, %

⬎95

98

2

44–95

78–94

3

⬎95

68

4

⬎95

97

5

71

95

6

⬎95

96

⬍5

N兾A

1

None

N-{1-[(4-methylphenyl)sulfonyl]heptyl}-P,P-diphenylphosphinic amide.

The general procedure was followed. The crude compound 7 (87% yield) was used without purification for the next step: (white powder) mp 117–118°C (dec.); 1H NMR (DMSO-d6, 400 MHz) 7.71 (dd, J ⫽ 11.9, 7.37 Hz, 2H), 7.59 (d, J ⫽ 8.0 Hz, 2H), 7.54–7.35 (m, 8H), 7.27 (d, J ⫽ 7.9 Hz, 2H), 6.32 (t, J ⫽ 12.3 Hz, 1H), 4.47 (dd, J ⫽ 11.0, 2.2 Hz, 1H), 2.34 (s, 3H), 1.95–1.86 (m, 1H), 1.73–1.62 (m, 1H), 1.32–1.20 (m, 1H), 1.15–0.85 (m, 7H), 0.75 (t, J ⫽ 7.2 Hz, 3H); 13C NMR (DMSO-d6, 100 MHz) ␦ 145.0, 135.6 (d, JC-P ⫽ 72.2 Hz), 134.4, 134.4 (d, JC-P ⫽ 75.9 Hz), 132.3 (d, JC-P ⫽ 3.3 Hz), 131.9 (d, JC-P ⫽ 10.0 Hz), 131.8 (d, JC-P ⫽ 10.3 Hz), 130.3, 130.0, 129.1 (d, JC-P ⫽ 12.5 Hz), 128.9 (d, JC-P ⫽ 12.7 Hz), 72.9, 31.7, 28.5, 25.5, 22.7, 22.0, 14.7; 31P NMR (DMSO-d6, 162 MHz) ␦ 26.0; LRMS (APCI) m兾z [M - SO2Tol]⫹ calc.: 314.2 found: 314.1; Anal. calc. for C26H32NO3PS: C 66.50, H 6.87, N 2.98 found: C 66.43, H 6.87, N 2.92.

7†

None 31P

N-[(1S)-1-ethylheptyl]-P,P-diphenylphosphinic amide. The product

14 was obtained by following the general procedure (specific conditions: ⫺20°C for 16 h). Yield 98%, ee (95%) was determined by HPLC analysis [Chiralpak AD-H, 90:10 hexanes:iPrOH, 1.0 ml兾min: (S)-14 tr ⫽ 10.3 min, (R)-14 tr ⫽ 14.7 min]. The S configuration is tentatively assigned based on the compounds 3, 5, 8, 9 and 10. mp 110–111°C; Rf 0.51 (100% AcOEt); 20 [␣]D ⫺5.4° (c 1.06, CHCl3); 1H NMR (300 MHz, CDCl3) ␦ 7.93 (d, J ⫽ 7.6 Hz, 2H), 7.90 (d, J ⫽ 7.1 Hz, 2H), 7.52–7.35 (m, 6H), 3.01 (br s, 1H), 2.68 (br s, 1H), 1.62–1.41 (m, 4H), 1.41–1.11 (m, 8H), 0.93–0.77 (m, 6H); 13C NMR (75 MHz, CHCl3) ␦ 134.4 (d, JC-P ⫽ 9.2 Hz), 133.6 (d, JC-P ⫽ 16.5 Hz), 132.6 (d, JC-P ⫽ 1.9 Hz), 132.1 (d, JC-P ⫽ 1.9 Hz), 128.9, 128.7, 60.8, 53.1 (d, JC-P ⫽ 2.0 Hz), 36.6 (d, JC-P ⫽ 5.3 Hz), 32.2, 29.7 (d, JC-P ⫽ 5.5 Hz), 25.9, 23.0, 14.5, 10.0; 31P NMR (121 MHz, CDCl3) ␦ 22.8; LRMS (APCI) m兾z [M ⫹ H]⫹ calc.: 344.4 found: 344.2; Anal. calc. for C21H30NOP: C 73.44, H 8.80, N 4.08 found: C 73.11, H 9.08, N 3.84. N-{cyclohexyl[(4-methylphenyl)sulfonyl]methyl}-P,P-diphenylphosphinic amide. The general procedure was followed. The crude com-

pound 6 (72% yield) was used without purification for the next step: (white powder) mp 113–115°C (dec.); 1H NMR (DMSO-d6, 300 MHz) 7.71 (dd, J ⫽ 11.4, 7.8 Hz, 2H), 7.56 (d, J ⫽ 8.2 Hz, 2H), 7.52–7.41 (m, 4H), 7.35–7.24 (m, 4H), 7.20 (d, J ⫽ 7.8 Hz, 2H), 6.28 (t, J ⫽ 12.9 Hz, 1H), 4.49 (dt, J ⫽ 11.9, 2.7, 1H), 2.34 (s, 3H), 2.18 (dt, J ⫽ 10.5, 2.1 Hz, 1H), 1.90 (d, J ⫽ 11.7 Hz, 1H), 1.52 (dt, J ⫽ 35.9, 10.8, 4H), 1.35–0.95 (m, 4H), 0.95–0.77 (m, 1H); 31P NMR (DMSO-d6, 121 MHz) ␦ 25.0; HRMS m兾z (FAB) calc. for C26H31NO3PS [M ⫹ H]⫹: 468.1762 found: 468.1767.

N -[(1 S )-1-cyclohexylpropyl]- P , P -diphenylphosphinic amide. The product 13 was obtained by following the general procedure (specific conditions: 0°C for 24 h). Yield 89%, ee (96%) was determined by HPLC analysis [Chiralpak AD-H, 90:10 hexanes:i-PrOH, 1.0 ml兾min: (S)-13 tr ⫽ 10.4 min, (R)-13 tr ⫽ 14.7 min]. The S configuration is tentatively assigned based on the 5408 兩 www.pnas.org兾cgi兾doi兾10.1073兾pnas.0307096101

*Yields determined by NMR with an internal standard. †Et Zn replaced by EtZnOTf (4 eq). 2

compounds 3, 5, 8, 9 and 10. mp 151–152°C; Rf 0.43 (100% 20 ⫺5.7° (c 0.99, CHCl3); 1H NMR (400 MHz, AcOEt); [␣]D CDCl3) ␦ 7.92 (m, 4H), 7.47 (m, 6H), 2.80 (br s, 1H), 2.72 (br s, 1H), 1,80–140 (m, 8H), 1.29–1.07 (m, 4H), 1.03–0.92 (m, 1H), 0.90 (t, J ⫽ 7.4 Hz, 3H); 13C NMR (75 MHz, CHCl3) ␦ 134.3 (d, JC-P ⫽ 23.7 Hz), 132.7 (d, JC-P ⫽ 20.5 Hz), 132.7 (d, JC-P ⫽ 1.8 Hz), 132.2 (d, JC-P ⫽ 20.5 Hz), 132.0 (d, JC-P ⫽ 2.5 Hz), 128.9 (d, JC-P ⫽ 2.1 Hz), 128.7 (d, JC-P ⫽ 2.1 Hz), 58.1 (d, JC-P ⫽ 2.3 Hz), 42.2 (d, JC-P ⫽ 5.1 Hz), 29.7, 28.8, 27.0 (d, JC-P ⫽ 3.0 Hz), 26.8 (d, JC-P ⫽ 3.7 Hz), 10.6; 31P NMR (121 MHz, CDCl3) ␦ 22.3; HRMS (MAB) m兾z calc. for C21H28NOP [M]: 341.1909 found: 341.1905. Results One important aspect of the in situ generation of Nphosphinoylimines from stable precursors is the concomitant formation of a stoichiometric amount of an EtZnLG species (Scheme 3). Because the basic nature of this by-product may seriously hamper the subsequent catalytic asymmetric nucleophilic addition step, we initially elected to test a standard catalytic asymmetric addition on a preformed and stable N-phosphinoylimine in the presence of several additives (Table 1). These reactions were then compared to the standard one in the absence of additives to identify which additives had the minimal detrimental effect on the efficiency and level of enantioselection for the reaction. The data collected in Table 1 illustrate that the benzotriazoyl group (entry 2) resulted in irreproducible yields depending on the reaction time (longer times led to higher yields and ee of the desired product: 78% ee after 6 h and 94% ee after 20 h). Conversely, the succinimidate group (entries 3) allowed for an excellent yield but eroded ee when compared to the standard reaction run in the absence of the additive (entr y 1). The ethylzinc ptoluenesulfinyl was fully compatible with the reagents and Coˆ te´ et al.

Scheme 4.

Entry

catalytic species, and high yield and enantiocontrol was observed in the presence of a stoichiometric amount of this additive (entry 4). It is noteworthy that high enantiocontrol was also achieved in the presence of ethylzinc phenoxide (entry 5) or ethylzinc methoxide (entry 6). To further corroborate these findings, the benzotriazoyl adduct 1 of the imine derived from benzaldehyde was prepared and submitted to the reaction conditions (Scheme 4). Although a low conversion to the desired amine 2 was observed after 6 h (43% conversion and 38% isolated yield), the ee of the product was excellent, perhaps indicating that the benzotriazoyl group may be a very good choice for the in situ preparation of imines. However, the data presented in Table 1 led us to focus initially on the p-toluenesulfinyl adduct of the imine because it appeared to be the most promising precursor for the in situ preparation of dialkyl-substituted ␣-chiral amines. The synthesis of the sulfinic acid adduct of N-phosphinoylimines could not be accomplished by using standard procedures that involve acidic conditions and heat. However, we found that they could be prepared simply by stirring an aldehyde, diphenylphosphinic amide, and p-toluenesulfinic acid in ether (Table 2). The mixture is stirred for several hours (16–36 h) during which the resulting sulfinic acid adduct precipitates and is isolated by filtration in high yield. The reaction proceeds well with several aryl- and alkyl-substituted aldehydes. More importantly, the reaction involving aldehydes containing enolizable protons proceeded smoothly and without any significant self-condensation of the aldehyde. Typically, the solid that is recovered on filtration is sufficiently pure to be used directly in the catalytic asymmetric alkylation reaction with diorganozinc reagents. One limitation of this approach is that ␣,␤-unsaturated aldehydes are not compatible with this procedure as the sulfinic moiety is prone to undergo conjugate addition reaction with this substrate. The next step was to test several reaction conditions and other parameters to identify any peculiar aspect of copper catalyzed addition of diorganozinc reagents to the sulfinic acid adduct 5 (Table 3). Based on our previous work on this reaction, the first important variable to be tested was the optimal stoichiometry

1 2 3 4 5 6

SPECIAL FEATURE

Cu salt, mol%

mol% (R,R)-BozPHOS

Yield, %

ee, %

Cu(OTf)2 (10) Cu(OTf)2 (6) Cu(OTf)2 (4.5) Cu(SO2Tol)2 (6) Cu(OTf)2 (6) CuOTf (10)

5 5 5 5 3 5

97 93 98 90 90 83

94 97 96 82 98 93

between the copper salt and (R,R)-BozPHOS, as well as the optimal copper salt. The Cu兾ligand stoichiometry is important because it defines the amount of ethylzinc triflate present in the mixture. Although ethylzinc triflate is not sufficiently nucleophilic to add to the imine (Table 1, entry 7), we believe its role as a Lewis acid to activate the imine is crucial in this reaction. We also elected to use 2.5 eq of diethylzinc because 1 eq is consumed to form the imine, leaving 1.5 eq for the nucleophilic addition process. The first three entries illustrate the effect of the stoichiometry between (R,R)-BozPHOS and copper(II) triflate on the efficiency of the reaction. It appears that a 1:1 ratio of Cu and the chiral ligand is optimal for this reaction. Excellent ee of the product was observed when the amount of the chiral ligand was lowered to 3 mol%; however, the yield was slightly lower (Table 1, entry 5). A similar reduction in yield was observed if copper(I) triflate was used as the catalyst precursor (Table 1, entry 6). Because a stoichiometric amount of the sulfinyl counterion is formed in the reaction, we were intrigued by the possibility of using copper p-toluenesulfinyl as the catalyst precursor. Unfortunately, much lower ee resulted with this copper salt (Table 1, entry 4). With the optimal conditions in hand, the copper catalyzed addition reactions of diethylzinc to sulfinic acid adducts 3-10 were tested (Table 4). We were quite pleased to find that the additions proceeded smoothly in all of the cases and that, more importantly, excellent enantiocontrol was observed. Furthermore, the reaction was shown to be highly effective with ␣-branched imines (Table 4, entries 2, 4, and 6). In all cases, the Table 4. Catalytic asymmetric synthesis of amines

Table 2. Synthesis of the sulfinic adduct of N-phosphinoylimines

Entry Entry 1 2 3 4 5 6 7 8

Coˆ te´ et al.

R1

Yield, %

Ph c-C5H11 PhCH2CH2 c-C6H13 C6H13 i-Pr Me i-Bu

71 (3) 95 (4) 97 (5) 72 (6) 87 (7) 88 (8) 91 (9) 84 (10)

1 2 3 4 5 6 7 8

R1

Yield, %

ee, %, (er)

Ph c-C5H11 PhCH2CH2 c-C6H13 C6H13 i-Pr Me i-Bu

87 (2) 92 (12) 98 (11) 89 (13) 98 (14) 86 (15) 97 (16) 97 (17)

97 (98.5:1.5) 95 (97.5:2.5) 96 (98:2) 96 (98:2) 95 (97.5:2.5) 96 (98:2) 90 (95:5) 96 (98:2)

er, Enantiomeric ratio. PNAS 兩 April 13, 2004 兩 vol. 101 兩 no. 15 兩 5409

CHEMISTRY

Table 3. Optimization of the catalytic asymmetric addition

N-phosphinoyl protecting group could be cleaved under mild conditions to liberate the amine (HCl, MeOH).

advantages of this process are high yields and enantioselectivities, as well as the mild conditions for the deprotection of the N-protecting group.

Conclusion In conclusion, we have shown that Me-DuPHOS monoxide [(R,R)-BozPHOS] is a very effective ligand in the coppercatalyzed addition of diethylzinc to N-phosphinoylalkylimines prepared in situ from their p-toluenesulfinyl adducts. The major

This work was supported by the Natural Sciences and Engineering Research Council (NSERC), Merck Frosst Canada, Boehringer Ingelheim (Canada), and the University of Montreal. A.A.B. is grateful to NSERC (PGF B) and Fonds pour la Formation de Chercheurs et l’Aide `a la Recherche (B2) for postgraduate fellowships.

Enders, D. & Reinhold, U. (1997) Tetrahedron Asymmetry 8, 1895–1946. Bloch, R. (1998) Chem. Rev. 98, 1407–1438. Kobayashi, S. & Ishitani, H. (1999) Chem. Rev. 99, 1069–1094. Kleemann, A., Engel, J., Kutscher, B. & Reichert, D. (2001) Pharmaceutical Substances: Syntheses, Patents, Applications (Thieme, Stuttgart). Ellman, J. A., Owens, T. D. & Tang, T. P. (2002) Acc. Chem. Res. 35, 984–995. Liu, G. C., Cogan, D. A. & Ellman, J. A. (1997) J. Am. Chem. Soc. 119, 9913–9914. Enders, D., Schubert, H. & Nu ¨bling, C. (1986) Angew. Chem. Int. Ed. Engl. 25, 1109–1110. Denmark, S. E., Weber, T. & Piotroswski, D. W. (1987) J. Am. Chem. Soc. 109, 2224–2225. Sibi, M. P. & Asano, Y. (2001) J. Am. Chem. Soc. 123, 9708–9709. Overberger, C. G., Marullo, N. P. & Hiskey, R. C. (1961) J. Am. Chem. Soc. 83, 1374–1378. Fujihara, H., Nagai, K. & Tomioka, K. (2000) J. Am. Chem. Soc. 122, 12055–12056. Hayashi, T. & Ishigedani, M. (2000) J. Am. Chem. Soc. 122, 976–977. Porter, J. R., Traverse, J. F., Hoveyda, A. H. & Snapper, M. L. (2001) J. Am. Chem. Soc. 123, 984–985. Porter, J. R., Traverse, J. F., Hoveyda, A. H. & Snapper, M. L. (2001) J. Am. Chem. Soc. 123, 10409–10410. Wei, C. & Li, C.-J. (2002) J. Am. Chem. Soc. 124, 5638–5639. Hermanns, N., Dahmen, S., Bolm, C. & Bra¨se, S. (2002) Angew. Chem. Int. Ed. Engl. 41, 3692–3694.

17. Dahmen, S. & Bra¨se, S. (2002) J. Am. Chem. Soc. 124, 5940–5941. 18. Boezio, A. A. & Charette, A. B. (2003) J. Am. Chem. Soc. 125, 1692–1693. 19. Boezio, A. A., Pytkowicz, J., Co ˆte´, A. & Charette, A. B. (2003) J. Am. Chem. Soc. 125, 14260–14261. 20. Wipf, P., Kendall, C. & Stephenson, C. R. J. (2003) J. Am. Chem. Soc. 125, 761–768. 21. Shi, M. & Wang, C.-J. (2003) Adv. Synth. Cat. 345, 971–973. 22. Jimeno, C., Reddy, K. S., Sola, L., Moyano, A., Pericas, M. A. & Riera, A. (2000) Org. Lett. 2, 3157–3159. 23. Sato, I., Kodaka, R. & Soai, K. (2001) J. Chem. Soc. Perkin Trans. 1, 2912–2914. 24. Pinho, P. & Andersson, P. G. (2001) Tetrahedron 57, 1615–1618. 25. Zhang, H. L., Zhang, X. M., Gong, L. Z., Mi, A. Q., Cui, X., Jiang, Y. Z., Choi, M. C. K. & Chan, A. S. C. (2002) Org. Lett. 4, 1399–1402. 26. Yamada, K., Harwood, S. J., Groger, H. & Shibasaki, M. (1999) Angew. Chem. Int. Ed. Engl. 38, 3504–3506. 27. Matsunaga, S., Kumagai, N., Harada, S. & Shibasaki, M. (2003) J. Am. Chem. Soc. 125, 4712–4713. 28. Masumoto, S., Usuda, H., Suzuki, M., Kanai, M. & Shibasaki, M. (2003) J. Am. Chem. Soc. 125, 5634–5635. 29. Katritzky, A. R. & Harris, P. A. (1992) Tetrahedron Asymmetry 3, 437–442. 30. Mecozzi, T. & Petrini, M. (1999) J. Org. Chem. 64, 8970–8972. 31. Kohn, H., Sawhney, K. A., Robertson, D. W. & Leander, J. D. (1994) J. Pharm. Sci. 83, 689–691. 32. Wu, M.-J. & Pridgen, L. N. (1991) J. Org. Chem. 56, 1340–1344.

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16.

5410 兩 www.pnas.org兾cgi兾doi兾10.1073兾pnas.0307096101

Coˆ te´ et al.