Stereoselective radical addition of carbon-centred radicals to the dehydroalanine moiety of the chiral nickel(II) complex of the Schiffprime;s base derived from (S)-2-N-(Nprime;-benzylprolyl)aminobenzophenone and dehydroalanine

摘要

J. CHEM. SOC. PERKIN TRANS. 1 1994 Stereoselective Radical Addition of Carbon-centred Radicals to the Dehydroalanine Moiety of the Chiral Nickel(i1) Complex of the Schiff's Base Derived from (S)-2-[N-(N'-Benzylprolyl)amino] benzophenone and Dehydroalanine Rashid G. Gasanov, Lubov V. Il'inskaya, Michail A. Misharin, Victor 1. Maleev, Nicolai 1. Raevski, Nicolai S. Ikonnikov, Svetlana A. Orlova, Nadezhda A. Kuzmina and Yuri N. Belokon' * Nesme yanov Institute of Organoelement Compounds, Russian Academy of Sciences, 1 17813, Moscow, Vavilov 28, Russia A new approach to the asymmetric synthesis of P-substituted a-aminopropanoic acids by 2.2'- azoisobutyronitrile (A1BN) and Bu,SnH-initiated radical addition of Etl. Pr'Br, Bu'Br, and PhCH,Br to the dehydroalanine moiety of the Ni" complex 1 of a Schiff's base derived from (S)-o-[N-(N'-benzylprolyl)amino] benzophenone (BPB) and dehydroalanine is described. The radical addition produced a mixture of diastereoisomeric complexes 4a-d with a 40-90% excess of (S,S)-diastereoisomers over the (S,R)-ones, giving the reaction products in almost quantitative yields.The diastereoselectivity of the reaction depended on the size of the entering radicals, the most effective asymmetric induction being achieved for the Bur radical addition. Enantiomerically pure '(S)-2-amino-3-(tert-buty1)propanoic acid' [(S)-y-methylleucine] and the chiral auxiliary BPB were recovered from compound 4d after its decomposition with HCI. The reactivities of the carbon-centred radicals towards the carbon-carbon double bond in the amino acid moiety of the complex 1 was quantitatively established by using ESR spectroscopy in the spin-trap technique.It is widely known that organic reactions, proceeding via the key in the solution was finally established (Scheme 1).5 We believed intermediate formation of carbon-centred free radicals, may that, in analogy with other types of chiral dehydroalanine serve as useful methods for C-C bond formation in organic substrates,6 the radical addition of carbon-centred radicals to synthesis.' Recently, it was realized that radical reactions might substrate 1 might also be successfully carried out as outlined in be carried out diastereoselectively 'and there were cases when Scheme 1, and kinetic diastereoselectivities of radical and ionic the stereoselectivities of the radical reaction and the ionic addition to the rigid substrate 1might be compared and a new reaction, leading to the same final compound from similar route to enantiomerically pure non-proteinogenic a-amino starting substrates, either did not differ3 or were giving acids opened.opposite stereochemical result^.^ This was rationalized by postulating stereodetermining attack of the reagents on the diastereotopic faces of the intermediate ionic and radical particles controlled by either similar or different initial Results and Discussion distribution of the rotamers of the intermediate^.^.^ The Alkyl halides EtI 2a, BzlBr or BzlCl2b, Pr'Br 2c and Bu'Br 2d rotamer distribution, in turn, was suggested to be controlled, were chosen as suitable starting materials for the generation of apart from the steric non-bonding interactions, either by the corresponding free radicals, Et', *Pri, 'Bu', and PhCH,' (see chelation (ionic enolate reactions) or by intermolecular dipole- Scheme 1).According to the mechanism of free-radical addition dipole repulsion (radical reaction^).^'^ Stereoelectronic effects reactions,' radicals R', generated in the initial step, add to the (partial sp3 character of the intermediate radicals) were invoked C=C bond of the dehydroalanine moiety of substrate 1 to as a possible source of the unusual 1,2-asymmetric induction in give intermediate a-radicals 3a-d. Then follows the stereo- reactions of nonconjugated acyclic radicals.2b determining attack by HSnBu, on the si- and re-side of We believed that use of rigid chiral substrates in which all the the intermediate adduct-radicals 3a-d (see Scheme 1).The set of groups were fixed in space might allow us to discern subtle the initial carbon-centred radicals generated from alkyl halides differences in the stereochemical behaviour of ionic and radical 2a4 provides a sufficiently diverse series to allow definite particles in the reactions of C-C bond formation. conclusions to be made as to the importance of the newly Earlier we showed that the asymmetric synthesis of P-substi- formed side-chain size and its conformation in adduct-radicals tuted a-amino acids might be carried out by employing 1,4-ad- 3a-d for the final stereodetermining step of a hydrogen-atom dition of nucleophiles to the dehydroalanine moiety of the chiral transfer to the intermediate a-radicals (see Scheme 1) in the Ni" complex 1 of a Schiff's base derived from (S)-0-[N-chain-transfer step, leading to the formation of the complexes 4a-d.(N'-benzylprolyl)amino] benzophenone (BPB) I and dehydro- alanine.Two types of diastereoselectivity were observed in In order to establish the reactivities of the above mentioned the addition. The first had its origin in the relative rates of the free radicals R towards the C=Cbond of the dehydroalanine re- and si-face protonation of the intermediate a-carbanion moiety in the complex 1we have determined the corresponding (kinetic diastereoselectivity) and the other was thermodynamic- rate constants using the spin-trapping method.' ally controlled, as soon as the equilibrium between the resulting diastereoisomers facilitated by the presence of bases Kinetic Data.-For the ESR experiments the radicals Et' and PhCH,' were generated by halogen abstraction from the corresponding alkyl halides by the radical 'Re(CO),, formed t Available from Merck (cat.no. 814473) and Janssen (cat. no. photochemically from Re,(CO),, (Scheme 2).' The Et' and 269 1950). PhCH,' radicals and adduct-radicals 3a and 3b were identified 3344 J. CHEM. SOC. PERKIN TRANS. I 1994 radical or carbanion I 1 3e-d RX 2a EtI 2b BzlBr BzlCl re -face attack ii ii si -face attack 2c Pr'Br 2d Bu'Br I.:,,) 0 A (S,S)-4a-d 64iit 'OH BPB (S,R)-4a-d -Iiii R'k-KO" I NH2 Scheme 1 Reagents: i, RX (2), Bu,SnH, AIBN or BzlMgX; ii, Bu,SnH, or water (D,O); iii, aq.HCI; iv, MeONa in MeOH as their spin-adducts 6a, 6b and 7a, 7b, correspondingly formed as the result of the interaction of the radicals and the adduct- radicals with tri(tert-buty1)nitrosobenzene 5. The experimental ESR spectra of the spin-adducts 6a, 6b, 7a and 7b have been simulated with the hyperfine coupling constants uN 13.42 G, ub-H 18.00 G, um-H 0.8 G for 6a; UN 13.62 G, UN 14.75 G, urn-" 0.83 G for 6b; and uN 10.00 G, urnpH2.06 G for 7a and 7b. The splitting constants are similar to the same constants for the corresponding Et' and PhCH,' radicals in the case of adducts 6a and 6b, and for the tertiary alkyl radicals9 in the case of adducts 7a and 7b.Identification of the signals from adducts 6a and 7a or 6b and 7b simultaneously in the ESR spectra of their admixture allowed us to determine the rate constants k, for the addition of the radicals Et' and PhCH,' to the C=C bond of the complex 1, according to eqn. ( The values of the second-order rate constant k, were calculated on the basis of the simulated ESR spectral data and known' values of k, (2 x lo5 dm3 mol-' s-l for Et', and 1 x lo4 dm3 mo1-' sP1 for PhCH,'). The addition rate constant k, calculated in this manner from eqn. (1) was I .97 x 1O5 dm3 mol-' s-l for Et' and 1 x lo3 dm3 mol-' s-l for PhCH,'. It should be noted that the values of k, have similar magnitudes to the rate constants of the addition of the same radicals to methyl acrylate (2.5 x lo5 dm3 mol-' sP1for Et' and 5.5 x lo3 dm3 mol-' s for PhCH,')and to acryloylpyrrolidine(9.9 x lo5dm3 mol s-' for Et', and 3 x lo3dm3 mol-' s-' for PhCH,').'' In addition, on the basis of the kinetic data" we suggest that the approximate value of k, for the tert-butyl radical addition to substrate 1 may be estimated as lo5 dm3 mol SS1.The salient feature of the kinetic results was that the Ni" ion complexation had the same effect on the carboxyl group as did its esterification, which was revealed by the similarity of the kinetic parameters of the radical addition to the C=C bond of the dehydroalanine moiety of complex 1 and that of methyl acrylate.Summing up the kinetic results obtained for the radicals Et', 'But, and PhCH,', and reactivity data for the radical 'SnBu, towards both the C=C bond in unsaturated compounds and the C=O bond in carbonyl groups, and the rate data for hydrogen transfer from HSnBu, to various free radicals,'2b,' suitable experimental conditions for the synthesis of the complexes 4a4 (see Scheme 1) were chosen. Synthetic Procedure.--The mixture of substrate 1, an alkyl halide, HSnBu,, and AIBN in benzene solution was kept at 80 "C for 4 h in a sealed ampoule under Ar (method A) or in an open vessel (method B). The complexes 4a4 were obtained as an assortment of diastereoisomers of (S,S)-and (S,R)-configuration in good chemical yields (77-9473 after their purification on SiO,.We were unable to separate the diastereoisomers chromatographically. J. CHEM. SOC. PERKIN TRANS. I 1994 3345 Re2(CO)lo -2 'Re(C0)S RX + 'Re(CO)s -R' + XRe(CO)S 0 I 1 + R' k1 3 But R'+ htAN=O 7But -k Bu' 4 N-'Bu' " '0' 5 6 o/ OW,/yg3+5-7 Scheme 2 The ratios of the former isomers were estimated by integrating the areas of the H resonances in the 'H NMR spectra of the mixtures of the corresponding diastereoisomeric (S,R)-and (S,S)-complexes, usually employing the resonances of the diastereotopic (NCH,) protons or some resonances of the protons of the phenyl groups. For complex 4d the signals for Bur (0.65 and 0.73 ppm) might be used. The proportions of diastereoisomers found in the reaction mixture were sometimes corroborated by enantiomeric GLC analysis of the amino acids recovered from the complexes after their decomposition.The use of both methods resulted in similar ratios of the isomers. The absolute conjiguration of the predominant diastereoisomer. This was established by comparing either the 'H NMR spectra of the diastereoisomeric mixtures and that of the pure (S,S)-diastereoisomer (obtained as usual from the corresponding amino acid and BPB under the thermodynamic control conditions, followed by crystallization) ' or by analysing their optical rotatory dispersion (ORD) curves. Chiral GLC analysis data of the amino acids recovered from the complexes, according to Scheme 1, might also be used for the purpose.The synthetic results are summarized in Table 1. The ratio oj the diastereoisomers. This evidently reflects the kinetic preference of the hydrogen-atom transfer from HSnBu, to the intermediate adduct-radicals. Had equilibration of the isomers taken place in the reaction mixture, the proportion of the (S,S)-isomers would have been much greater (at least 90%),5,' as the results of equilibration experiments with some of the complexes indicate (see Table 1, runs 1 and 2). As can be seen, it was always the re-attack of Bu,SnH on the intermediate radicals 3a4, leading to the (S,S)-diastereoisomers, which predominated over the si-face attack, giving the (S,R)-diastereoisomer. The relative rates of the re- and si-attack of the hydrogen-atom transfer, occurring at the second stage of the radical addition (see Scheme l), depended on the size of the Table 1 Radical addition of alkyl halides (RX) 2a-d to complex 1 Yield d.e.(S,S)b Run RX Method" Product (%) (%) 1 EtI PhCH,Cl PhCHiBr A B A 4a 4b 4b 90 94 77 54 (92)' 40(98)' 40 4 Pr'Br A 4c 90 60 5 Bu'Br A 4d 80 90 6 Bu'Br B 4d 90 92 Method A: the reactions were carried out in sealed glass tubes. Method B: the reaction was carried out in an open vessel. bThe diastereoselectivities of the reaction were determined by 'H NMR spectroscopy corroborated via GLC analysis of the amino acids recovered from the complexes. 'Diastereoselectivity after MeONa- catalysed epimerization of the mixture of the diastereoisomers 4.side-chain of the intermediate a-radicals. For example, in the case of 'But radical addition to substrate 1 the best diastereo- isomeric excess (d.e.) (up to 92%) of all the reactions was obtained (Table 1, run 5). At the same time the ethyl and benzyl radical additions to substrate 1 gave only 54 and 40% d.e., respectively, of the product (Table 1, runs 1, 2 and 3). The d.e. ratio may be improved by epimerization of the diastereo- isomeric mixture with MeONa in MeOH at ambient temperature (Table 1, run 1). The stereochemical model of the intermediate radicals. This, as revealed by molecular mechanics calculations, seems to provide some rationale for the observed stereochemistry of the hydrogen-atom transfer.Fig. 1 shows the calculated structure of adduct-radical 3d. The chiral proline moiety causes chiral distortion of the system of the interconnected chelate rings. The metallocycle O(1)C( 1)C(2)N( 1)Ni has a chiral conformation and the bond C(2)-C(22) is tilted downwards. The chiral puckering of the metallocycle formed by C(3)C(4)C(9)-N(2)NiN( l), in turn, causes the phenyl substituent at the C=N bond to turn clockwise. The torsional angle N( l)C(3)C(27)C(32) is 60" and the phenyl substituent shields the re-face of the radical centre at the a-atom of the compound. The same type of shielding in the case of a carbanion centre was responsible for the predominant si-attack of electrophiles on the alanine moiety in the corresponding Schiff's base complexes of Ni".' On the other hand in the case of amino acids with larger side-chains the si-side of the radical (or carbanion) is effectively screened by the bulky side-substituent of the amino acid moiety. The chain may adopt several conformations as a result of the C(22)-C(2) rotation but the conformation depicted in Fig. 1 was found to be at least 3 kcal per mole* more favourable than the nearest one in order of its energy. Evidently, the conformation avoids unfavourable interaction between the side-chain and the phenyl substituent at the C=N bond. On the other hand, it is energetically 'cheap' to move the phenyl substituent in an anticlockwise manner to open the re-side of the radical to the hydrogen-atom attack.Thus, a 10" rotation would require less than 3 kcal energy. As a result the predominant product of the reaction of Bu'Br with substrate 1 was found to be (S,S)-4d, the product of the re-attack of Bu,SnH on the intermediate CC-radical. As the size of the alkyl halide became smaller, the difference in the energies of the transition states of the si-and re- attack on the intermediate radical 3 decreased and the d.e. of the reactions fell in the sequence Pr'Br EtI BzlBr. The diastereoselectivity of deuteriation experiments, con-ducted with the carbanion derived from (S,S)-4a with BuLi in tetrahydrofuran (THF), followed by quenching with DCl in D,O, indicated that the d.e. of the heterolytic deuteriation * 1 cal = 4.184 J. 3346 J.CHEM. SOC. PERKIN TRANS. 1 1994 Fig. 1 The computer-generated (MMX-calculated) structure of the Ni" complex of the carbon a-centred radical of 2-amino-4,4-dimethylpentanoic acid was similar to the d.e. of the homolytic hydrogen-atom transfer, leading to product 4a. Similar diastereoselectivities were found in the BzlMgCl addition to substrate l.5For example, under kinetically controlled conditions the addition gave 50% d.e. in favour of (S,S)-4b whereas the radical addition gave 40% d.e. (see Table 1, runs 2 and 3). Evidently, identical steric factors were responsible for the diastereoselectivities of both homolytic and heterolytic reactions and there are no grounds to invoke any type of specific stereoelectronic effects such as sp3 hybridization of the intermediate radical to account for the observed stereoselectivity of the reactions.Conclusions.-The method elaborated in this work is suitable for the production of new types of a-amino acids with bulky aliphatic side-chains. Although the kinetic stereoselectivity of the addition of smaller alkyl halides to substrate 1 was mostly insignificant, thermodynamic equilibration of the final diastereoisomers with MeONa may bring the quotient of the (S,S)/(S,R)-isomers up to 99/1 and thus a viable asymmetric synthesis of a range of m-amino acids might be achieved. Experimental General.-Reagents and solvents were purified in the usual way. Molecular Mechanics Calculations. -The calculations were performed as previously described.l4 Spectra were recorded with the following instruments: ESR, Radiopan SE/X-2547; 'H NMR, Bruker BM-500 (500 MHz) and Bruker WP-200 (200 MHz); J values are in Hz; ORD, JASCO ORD/UV-5 (specific rotations, in units of lo-' deg cm2 gl,measured with a Perkin-Elmer 241 polarimeter). Enantiomeric analyses of the &-amino acids were carried out by GLC. The amino acids were analysed as the N-trifluoroacetyl derivatives of their isopropyl esters by using a capillary glass column (40 m x 0.23 mm) coated with chiral diamide phase of a Chirasil-Val type (the phase was synthesized and columns produced in our laboratory). Chromatography conditions: column temperature 140 OC, He 1.2 x 1O5 Pa. Rate Measurements.-ESR experiments were carried out in the glass-soldered tubes in the spectometer resonator.First, the reaction mixtures put into the tubes were deoxygenated by the freeze-pumpthaw (FPT) method. In each tube the solution (0.2 cm3), containing Re,(CO),, (4.2 x g, 6.5 x mol), complex 1 (1.1 x g, 2.1 x mol), EtI (0.18 g, 1.18 x mol) or PhCH,Br (0.13 g, 7.5 x mol), and spin trap 5 (2.6 x lop3g, 1.0 x 10 mol) in CH2C12 (0.9 cm3) were mixed. The mixture was irradiated with UV light in the spectrometer resonator (A366 nm). The initial complex 1 was synthesized as described in ref. 5. Epimerization of the Mixture of Diastereois0mers.-A solution of MeONa in MeOH (0.1 mol dm '; 0.1 cm3) was added to a solution of complex 4b (0.1 g, 4.9 x mol) in MeOH (3 x lop3 dm3) under Ar.The mixture has been kept at 20 "C for 24 h, then was neutralized with aq. HCl; the complex was extracted by CHC1, and the organic layers were combined and evaporated. The resulting complex was obtained as almost pure (S,S)-diastereoisomer according to its 'H NMR data. Deuterium Exchange.-To a solution of complex 4a (0.54 g, 1.O x lo-, mol) in THF (10 cm3) in a glass tube, deoxygenated by FPT and cooled to -70 "C, was added a solution of 0.62 mol dm-, BuLi (4.03 cm', 2.5 x lo-, mol) in hexane under Ar. The mixture was kept for 15min at that temperature, allowed to warm up, and then was quenched with DCl (0.35 cm3, J. CHEM. SOC. PERKIN TRANS. 1 1994 3.5 x mol) in D,O (3 cm3). Water (10 cm3) was added to the mixture and then the resulting complex was extracted by CHCl,, and the organic layers were combined, dried with MgSO, and evaporated.The product was purified on a silica gel column [45 x 5 cm; CHC1,-acetone (7: l)]. General Procedure for Syntheses of the Ni" Complexes 4a4 (Method A).-A solution of the complex 1 (0.23 g, 4.5 x lo4 mol), the corresponding alkyl halide (6.0 x lo-, mol), HSnBu, (0.43 g, 1.46 x lop3 mol) and AIBN (0.04 g, 2.6 x lo4 mol) in benzene (5 cm3) was placed in a glass tube. The mixture was deoxygenated by FVD and afterwards the tube was sealed under Ar. The tube was kept in a thermostat at 80 "C for 4 h. According to TLC [SiO,; CHC1,-acetone (7 : l)], in all cases, PhCH,Br being the only exception, the final reaction mixture contained none of the initial complex 1.The reaction mixture was evaporated, and purified on a silica gel column. The first fraction, in the order of its emergence from the chromato- graphic column (as a red band), was the corresponding addition product 4a-d as a mixture of (S,S)-and (S,R)-diastereoisomers in the d.e.s presented in Table 1. The complex 4a had (Found: C, 67.1; H, 5.8; N, 7.5. C,,H,,N,O,Ni requires C, 66.69; H, 5.78; N, 7.78%); G,(CDCl,) for (S,S)-isomer 8.15-6.50 (14 H, m, ArH), 4.38 and 3.52 (2 H, AB, J 12.50, CH,Ph), 3.86 (1 H, m, amino acid a-H), 3.50 (1 H, m, Pro a-H), 3.48 and 3.45 (2 H, m, Pro 6-H,), 2.9-1.0 (4 H, m, Pro p-, y-H,) and 0.8 (3 H, t, Me); and for the (S,R)-isomer 8.60-6.60 (14 H, m, ArH), 4.42 and 3.54(2H,AB,J12.50,CH2Ph),4.18(1H,m,Proa-H),3.78(1 H, m, amino acid wH), 3.62 and 3.48 (2 H, m, Pro 6-H,), 2.9- 1.0 (4 H.m, Pro p-, y-H,) and 0.95 (3 H, t, Me). The c.omplex 4c had (Found: C, 67.3; H, 5.9; N, 7.4. C,,H,,N,O,Ni requires C, 67.17; H, 6.00; N, 7.58%); GH(CDCI,) for (S,S)-isomer 8.20-6.50 (14 H, m, ArH), 4.40 and 3.50 (2 H, AB, J 12.50, CH,Ph), 3.85 (1 H, m, amino acid a-H), 3.48 (1 H, m, Pro a-H), 3.69 and 3.52 (2 H, m, Pro 6-H,), 2.67- 2.08 (4 H, m, Pro p-, y-H,), 1.4 (2 H, m, amino acid p-, y-H) and 1.9 and 0.17 (6 H, d, Me); and for (S,R)-isomer G,(CDCl,) signals were not identified with the exception of 1.95 and 0.13 (6 H, d, Me). Synthesis of the Complex 4d (Method B).-To a solution of complex 1 (3.2 g, 6.3 x lo-, mol) and Bu'Br (1 1.7 g, 8.5 x lo-, mol) in benzene (70 cm3) under Ar was added a solution of AIBN (0.61 g, 4.0 x lo-, mol) and HSnBu, (6.0 g, 2.4 x mol) in benzene (50 cm3) at 80 "C during 0.5 h and then the mixture was boiled for 2 h.The reaction was monitored by TLC (see above). After the disappearance of the initial complex the reaction mixture was evaporated, and purified on a silica gel column [45 x 5 cm; CHC1,-acetone (7 :l)]. Complex 4d was obtained as a mixture of two diastereoisomers in the d.e.s represented in Table 1. The complex 4d had (Found: C, 67.3; H, 6.0; N, 7.2. C,,H,,N,O,Ni requires C, 67.63; H, 6.21; N, 7.38%); dH(CDC13) for (S,S)-isomer 8.10-6.50 (14 H, m, ArH), 4.38 and 3.50 (2 H, AB, J 12.50, CH,Ph), 3.85 (1 H, ABX, JAX 3.68, J,, 1 1.03, amino acid a-H), 3.48 (1 H, m, Pro a-H), 3.69 and 3.43 (2 H, m, Pro 6-H,), 2.75 and 2.20 (2 H, m, Pro y-H,), 2.50 and 2.05 (2 H, m, Pro p-H,), 2.95 and 1.47 (2 H, ABX, JAB 13.27, JAX 3.68, JBX 11.03, amino acid p-H) and 0.72 (9 H, s, Me,); and for (S,R)-isomer G,(CDCl,) 7.97-6.50 (14 H, m, ArH), 4.33 and 3.42 (2 H, AB, J 12.50, CH,Ph), 2.75 and 2.25 (2 H, m, Pro y-H,), 2.60 and 2.03 (2 H, m, Pro p-H2) and 0.64 (9 H, s, Me,).Synthesis of the Complex 4b (Method B).-To a solution of the complex 1 (1.6 g, 3.2 x lo-, mol) and PhCH,Cl(33 g, 0.26 mol) in benzene (20 cm3) under Ar was added a solution of AIBN (0.3 g, 2.0 x lo-, mol) and HSnBu, (3.0 g, 1.2 x lo-, mol) in benzene (20 cm3) at 80°C during 0.5 h and then the 3347 mixture was refluxed for 2 h.The reaction was monitored by TLC(conditions are mentioned above). After the disappearance of the initial complex the reaction mixture was evaporated, and purified on a silica gel column [45 x 5 cm; CHC1,-acetone (7:1)]. The complex 4b was obtained as a mixture of two diastereoisomers in the d.e.s represented in Table 1. The complex4b had (Found: C, 69.4; H, 5.6; N, 6.2. C,,H,,N,O,Ni requires C, 69.79; H, 5.52; N, 6.98%); G,(CDCl,) for (S,S)-isomer 8.05-6.50 (14 H, m, ArH), 4.48 and 3.59 (2 H, AB, J 12.50, CH,Ph), 3.86 (1 H, m, amino acid a-H), 3.56 (1 H, m, Pro a-H), 3.60 and 3.42 (2 H, m, Pro 6-H,), 2.73 and 2.17 (2 H, m, Pro y-H,), 2.38 and 2.05 (2 H, m, Pro P-H,), 3.20 and 2.50 (2 H, m, amino acid p-H,) and 2.27 and 1.88 (2 H, m, amino acid y-H,); the proton chemical shifts for the (S,R)-isomer were not identified with the exception of BH 8.50-6.65 (14 H, m, ArH), 4.49 and 3.59 (2 H, AB, J 12.50, CH,Ph) and 3.77 (1 H, m, amino acid a-H).Procedure for the Recovery of BPB and the x-Amino Acids from the Ni" Complexes.-This was carried out as described in refs. 5, 13 and 14. The recoveries of the amino acids and BPB were in the range 8&98%. The structures of the amino acids recovered from complexes 4a and 4c were established by TLC and GLC, by comparison of the samples with the standard commercially available amino aicds. Recovery of (S)-2-amino-4,4-dimethylpentanoicacid from complex 4d (2.0 g, 3.6 x lo-, mol) was made in the usual way; the amino acid (0.43 g, mol77%) was recrystallized from EtOH- water (7: 3); m.p.235 "C (decomp.) (Found: C. 58.0; H, 10.3; N, 9.6. C7HI5NO2 requires C, 57.95; H, 10.34; N, 9.65%); [ct]If5+ 14.2 (c 1.6 mol dm-, HC1). Enantiomeric purity of the amino acid was higher than 99% according to GLC; G,,(D,O-DCl 1 :1) 4.20 (1 H, ABX, JBX 4.92, JAx 6.89, a-H), 1.93 and 1.64 (2 H, ABX, JAB 14.77, JAX 6.89, JBX 4.92, p-H,) and 0.95 (9 H, s, Me,).Recovery of 2-amino-4-phenylbutanoic acid from complex 4b was made in the usual way. According to GLC the ratio of (S)and (R) isomers was 7 :3; dH(5 mol dmP3 DCl in D,O) 7.22 (5 H,m,Ph),4.11(1 H,AB,J,,6.2,NCH),2.70(2H,m,CH2Ph) and 2.20 (2 H, m, CCH,). Acknowledgements Support has been provided by the Russian Fundamental Research Fund (No.93-03-4497). References 1 B. Giese, Radicals in Organic Synthesis: Formation of Curbon-Curbon Bonds, New York, Pergamon Press, 1986; D. P. Curran, Synthesis, 1988, 489; Selectivity and Synthetic Applications of Radical Reactions; Tetrahedron Symposia in Print, ed. B. Giese, Tetrahedron, 1985,41,3887; P. P. Jasperse, D. P. Curran and T. L. Flevig, Chem. Rev., 1991,91, 1237. 2 (a)N. A. Porter, R. Breyer, E. Swann, J. Nally, J. Pradhan, T. Allen and A. T. McPhail, J. Am. Chem. Soc., 1991, 113, 7002; (6) G. Thoma, D. P. Curran, S. Geib, B. Giese, W. Damm and F. Wetterich, J. Am. Chem. Soc., 1993, 115, 8585 and cited references; (c)D. P. Curran and A. C. Abraham, Tetrahedron, 1993, 49,4821 and cited references. 3 B.H. Kim and D. P. Curran, Tetrahedron, 1993, 49, 293 and cited references. 4 J. G. Stack, D. P. Curran, J. Rebek and P. Ballester, J. Am. Chem. Suc., 1991,113,5918; J. G. Stack, D. P. Curran, S.V. Geib, J. Rebek and P. Ballester, J. Am. Chem. Soc., 1992, 114, 7007. 5 Yu. Belokon', A. Sagyan, S. Djamgaryan, V. Bachmutov and V. Belikov, Tetrahedron, 1988,44, 5507. 6 A. Beckwith and Ch. Chai, J. Chem. SOC.,Chem. Commun., 1990, 1087. 7 R. Gasanov and R. Freidlina, Usp. Khim., 1987,56447 (Russ. Chem. Rev., Engl. Transl., 1987, 56,264). 8 M. Wrighton and D. Ginley, J. Am. Chem. Soc., 1975,97, 2065. 9 S. Terabe and R. Konaka, J. Chem. SOC.,Perkin Trans. 2, 1973, 369. 10 R. Gasanov, A. Tumanskaya and L. Il’inskaya, Zzv. Akad. Nauk SSSR, Ser. Khim., 1992, 136 (Bull. Russ. Acad. Sci., Div. Chem. Sci., 1992,41, 107). 11 R. Gasanov, T. Churkina, L. Il’inskaya, K. Kochetkov and D. Bodrov, Dokl, Akad. Nauk SSSR, 1992,326, 126. 12 (a)K. Ingold, J. Lusztyk and J. Scaiano, J. Am. Chem. SOC.,1984, 106, 343; (b) C. Chatgilialoglu, J. C. Scaiano and K. Ingold, Organometallics, 1982, 1, 466; (c) L. Johnston, J. Lusztyk, D. J. CHEM. soc. PERKIN TRANS. 1 1994 Wayner, A. Abelwickereyma, A. Beckwith, J. Scaiano and K. Ingold, J. Am. Chem. SOC.,1985,107,4594. 13 Yu. N. Belokon’, Janssen Chim. Acta, 1992,10,4. 14 V. Soloshonok, V. Kukhar, S. Galushko, N. Svistunova, D. Avilov, N. Kuzmina, N. Raevski, Yu. Struchkov, A. Pysarevsky and Yu. N. Belokon’, J. Chem. SOC.,Perkin Trans. I, 1993,3143. Paper 4/02422B Received 25th April 1994 Accepted 9th June 1994