Synthesis and Characterization of Some New C2 Symmetric Chiral Bisamide Ligands Derived from Chiral Feist’s Acid

The hemilabile chiral C2 symmetrical bidentate substituted amide ligands (1R,2R)-5a-d and (1S,2S)-6a-d were synthesized in quantitative yield from (1R,2R)-(+)-3-methylenecyclo-propane-1,2-dicarboxylic acid (1R,2R)-3 and (1S,2S)-(-)-3-methylene-cyclopropane-1,2-dicarboxylic acid (1S,2S)-3, in two steps, respectively. The chiral Feist’s acids (1R,2R)-3 and (1S,2S)-3 were obtained in good isomeric purity by resolution of trans-(±)-3-methylene-cyclopropane-1,2-dicarboxylic acid from an 8:2 mixture of tert-butanol and water, using (R)-(+)-α-methylbenzyl amine as a chiral reagent. This process is reproducible on a large scale. All these new synthesized chiral ligands were characterized by 1H-NMR, 13C-NMR, IR, and mass spectrometry, as well as elemental analysis and their specific rotations were measured. These new classes of C2 symmetric chiral bisamide ligands could be of special interest in asymmetric transformations.


Introduction
Over the past few decades, C 2 symmetric chiral amides and sulfonamides have proven to be efficient ligands for several asymmetric transformations [1][2][3][4], due to their great potential of binding OPEN ACCESS with the metal alkoxides, especially with Ti(IV) alkoxides, through the nitrogen atom. Therefore considerable efforts have been devoted to the synthesis of a variety of substituted C 2 symmetric chiral amide ligands [5][6][7][8]. Recently a series of new chiral sulfonamides with a rigid cyclohexyl backbone were introduced by Wals and co-workers for the asymmetric addition of diethyl zinc to aldehydes [8]. In addition, all these C 2 symmetric chiral amides and sulfonamides are capable of forming five-and six-membered rings with metal chelates, in which the transition states only allow the approach of incoming groups from the less hindered side by blocking the highly hindered face. Moreover, asymmetric addition of organozinc to aldehydes is probably the most successful and still vigorously pursued area in asymmetric C-C bond formation [9][10][11][12][13]. Despite of the enormous success of chiral ligands in asymmetric reactions, a limited number of amides with 1,1'-biaryl backbones are reported for the organozinc addition [14][15][16][17][18][19][20][21][22]. In addition, the rational design of new chiral ligands for enantioselective conjugate alkylation has achieved limited success, presumably due to several factors that have to be taken into consideration. In case of C 2 symmetric chiral amide ligands, these factors could be explained as follows: firstly, due to the presence of C 2 symmetric axis in the chiral ligand, a number of possible transition states, in particular chiral transformations could be minimized [23]. Secondly, in the amide functional group, the carbonyl group has a potential donor site -NH group and therefore elucidation of their protonation behavior and related phenomena such as hydrogen bonding and Lewis acid complexation has drawn a good deal of attention. An amide molecule may have a dual role of both proton acceptor and donor, conferring a dual nature to the amide functionality [24]. Nevertheless, the size of the chelate ring has also proven to be important, since it controls the orientation of the substituents around the metal center. Hence, the bulkiness of the substituent in the amide ligands could be adjusted by changing the amide chain in order to achieve a better ligand structure for a particular reaction simply by selecting the appropriate stereochemistry and bulkiness [23].
In addition, Ikeda reported that the presence of stereogenic centers on the backbone of the ligands, introduces an extra element of complexity in the ligand structure, and special effect will arise when such ligands are employed in asymmetric catalysis [25]. Therefore C 2 symmetric chiral ligands have been used in the past few decades for catalytic asymmetric processes with a high degree of enantioselectivity [26][27][28][29], although chiral amides still remain an attractive choice for highly selective catalytic reactions due to their ready availability and simple reaction conditions. Hence, the development and application of C 2 symmetric chiral amides are still interesting and limited.
In this article, we report the preparation of (±)-Feist's acid as a chiral precursor in order to introduce a cyclopropane framework into the C 2 symmetric chiral ligands. For this purpose we need an effective resolution of this (±)-Feist's acid. There are many protocols in the literature for the resolution of (±)-Feist's acid. In the earliest, Doering and Roth described the resolution of Feist's acid, using L-(−)-quinine as the resolving reagent [30], whereas Al- Majid et al., used L-(−)-menthol as a resolving reagent [31]. Recently, Godfrey et al. have used (R)-(+)-α-methylbenzyl amine as a chiral reagent [32]. Although a generally applicable method is still lacking, our attention, in this context, has focused on the effective modification of (±)-Feist's acid resolution, and the general method for the synthesis of novel C 2 symmetric chiral bisamide ligands with a rigid cyclopropane framework.
The syntheses of the acid chlorides (1R,2R)-4 and (1S,2S)-4 were carried out using two different procedures A and B, although the better yields of the ligands (up to 85%) were achieved by using the acid chloride obtained from procedure B, as described in the Experimental section. The formation of the acid chloride intermediates (1R,2R)-4 and (1S,2S)-4 was confirmed by IR spectroscopy, where the absence of a broad band around 3,452 cm -1 (due to OH stretching frequency) was observed, along with the shifting of the C=O stretching frequencies from 1,710 to 1,800 cm -1 , indicating the formation of the acid chlorides. Furthermore, appearance of a sharp peak at 764 cm -1 was attributed to C-Cl stretching frequency.
The ligands (1R,2R)-5b and (1S,2S)-6b with a sec-butyl amide side chain, were obtained in good yield, in a similar fashion as described in the above paragraph (Schemes 4 and 5). Formation of these compounds were confirmed by IR and 1 H-NMR spectroscopy. The IR spectra of these two ligands, showed sharp peaks at 3,273 and 3,269, 1,632 and 1,630 and 1,557 and 1,560 cm -1 , which were assigned to the N-H (sec) stretching, C=O stretching and C-N stretching vibrations, respectively. In case of their 1 H-NMR spectra, the presence of a triplet at δ 0.81, a doublet at δ 1.02 and another doublet at δ 8.05, which were assigned to the protons of two sets of primary methyls, secondary methyls and two sets of amide groups, respectively. The splitting of the NH proton is mainly due to the presence of an adjacent tertiary proton in the sec-butyl chain. The calculated specific rotation of the ligands (1R,2R)-5b and (1S,2S)-6b were found to be [α] 20 D + 129 (c, 0.31%, 10% MeOH/CHCl 3 ) and [α] 20 D − 112 (c, 0.31%, 10% MeOH/CHCl 3 ), respectively, which matching well with their configuration. The ligands with the isobutyl chain (1R,2R)-5c and (1S,2S)-6c were obtained in very good yield, as described in the Experimental section (Schemes 4 and 5). The structures of these amide ligands were elucidated from their IR and 1 H-NMR spectra. In the IR spectra, three sharp peaks were observed for both the ligands at 3,296 and 3,288, 1,632 and 1,634, and 1,558 and 1,559 cm -1 , due to the stretching frequencies of sec-N-H, C=O and C-N groups, respectively. In the 1 H-NMR, observation of a doublet at δ 0.84, and a triplet at δ 8.21, are indicative of the presence of four sets of primary methyls and two sets of amide groups, accordingly. The presence of two methylene protons in the iso-butyl chain adjacent to the NHCO, are responsible for the splitting of the NH proton signal into a triplet. The configuration of these ligands were determined from their specific rotation values, which were found to be [α] 20  Finally, the ligands (1R,2R)-5d and (1S,2S)-6d with benzyl groups in amide side chain were prepared in very satisfactory yield by reaction of benzylamine (2 eq.) with the intermediates (1R,2R)-4 and (1S,2S)-4, respectively. These ligands were also characterized from the IR and 1 H-NMR spectra. In the IR spectra of these two ligands, three characteristic sharp peaks at 3,289 and 3,297, 1,631 and 1,630 and 1,540 and 1,541 cm -1 were observed for the stretching frequencies of sec-N-H, C=O and C-N, respectively. The 1 H-NMR of these ligands revealed characteristic peaks at δ 7. 25-7.32 and 8.78 as a multiplet and triplet, which were assigned to the corresponding aromatic and amide protons, respectively. Again the splitting of the NH proton into a triplet was due to the presence of two benzylic protons in the amide chain. The configurations of these ligands were determined from their specific optical rotations. Thus the specific rotations for these two ligands were found to be [α] 20 D + 100 (c, 0.4%, 10% MeOH/CHCl 3 ) and [α] 20 D − 166 (c, 0.31%, 10% MeOH/CHCl 3 ), respectively.
In 1 H-NMR spectra, for both types of ligands (1R,2R)-5 a-d and (1S,2S)-6 a-d , we found a shift of the NH proton's δ values downfield to δ 7.87, 8.05, 8.21, and 8.78, and this downfield shift is due to the decreasing +I effect for the substituent groups in the amide chain in going from tert-butyl to a benzyl group.

General
All the moisture and air sensitive reactions were carried out under an inert atmosphere using an argon filled glove box and standard Schlenk-line techniques. All the chemicals were purchased from Aldrich, Sigma-Aldrich and Fluka and were used as received without purification, unless otherwise stated. TLC plates were used for monitoring the reactions. Flash chromatography was carried out with silica gel (100-200 mesh). Pyridine, triethylamine and diisopropylamine were dried over sodium hydroxide. Diethyl ether and tetrahydrofuran were distilled from sodium benzophenone ketyl. Hexane, heptane and pentane were distilled by using sodium/triglyme benzophenone ketyl. Chloroform, dichloromethane, benzene, toluene and dimethyformamide were dried using calcium hydride. Deuterated solvents were dried over calcium hydride and deoxygenated prior to use. 1 H and 13 C-NMR spectra were recorded on Jeol-400 spectrometer ( 1 H 400 MHz, 13 C 100 MHz): using deuterated CHCl 3 or DMSO as solvent. The chemical shifts (δ in ppm) for 1 H-and 13 C-NMR were referenced internally using residual non deuterated solvent resonance shift and reported to trimethylsilane (TMS). Coupling constants (J) are taken in Hertz (Hz). Specific rotations were measured by using ATAGO-POLAX-2L. Elemental analyses were performed on a Perkin Elmer 2400 Elemental Analyzer. IR spectra were recorded on a Model FTIR-800 Infrared FT-IR Spectrometer using KBr pellets for solids or neat for liquids. Mass spectrometric analysis was conducted by using ESI mode on AGILENT Technologies 6410-triple quad LC/MS instrument. (2), 20 g, 0.14 mol) was added to an 8:2 mixture of tert-butanol-water (160 mL) and the suspension was heated to 90 °C on a steam bath until it had completely dissolved. The resulting solution was then removed from the steam bath and (R)-(+)-α-methylbenzamine (17 g, 0.14 mol) was added slowly over a period of 20 min. The reaction mixture was stirred for 10 min and left to stand for 24 h to give a mixture of (1R,2R,1'R)-(+)-3a and (1S,2S,1'R)-(+)-3b in 1:1 ratio.

Conclusions
In conclusion, the synthesis of the new class of C 2 symmetric chiral bisamide ligands with high specific rotation values were carried out. All these ligands were derived from trans-(±)-3-methylenecyclopropane-1,2-dicarboxylic acid (Feist's acid), which served as the key precursor. In order to obtain the chiral backbone of the ligands, we have resolved trans-(±)-3-methylenecyclopropane-1,2dicarboxylic acid in a new modified method, hence (1R,2R)-(+)-3-methylenecyclopropane-1,2dicarboxylic acid and (1S,2S)-(−)-3-methylenecyclopropane-1,2-dicarboxylic acid were obtained with a high degree of enantioselectivity. The applications of these chiral ligands are under investigation and the results will be reported in the nearest future. On the basis of the chiral Feist's acid, we are working to explore a large variety of novel classes of chiral C 2 symmetrical bidentate, or tetradentate ligands, with bulky environments.