Convenient Asymmetric Synthesis of Fmoc-(S)-6,6,6-Trifluoro-Norleucine

In this work we report a convenient asymmetric synthesis of Fmoc-(S)-6,6,6-trifluoronorleucine via alkylation reaction of chiral glycine equivalent. The target amino acid of 99% enantiomeric purity was prepared with 82.4% total yield (three steps).

In this paper we describe a convenient asymmetric synthesis of Fmoc derivative 6,6,6-trifluoronorleucine 1 via CF 3 (CH 2 ) 3 I alkylation of the recently rationally designed chiral equivalent of nucleophilic glycine 4. The method is operationally convenient, robust, scalable, and can be recommended for practical preparation of enantiomerically pure (~99% ee) derivative of this important tailor-made AA.

Materials and Methods
General Methods. All solvents and reagents were used as purchased without further purification. All reactions were conducted by magnetically stirring and detected by routine chromatography on TLC plates. Flash chromatography was carried out using the corresponding solvents on silica gel (0.064-0.210 mm). The reported yields are for isolated and chemically pure compounds. HPLC experiments were performed on a standard equipment using the Inertsil TM ODS-3 column (3 µm, 150 × 4.6 mm) ran at 1.0 mL/min, 30 • C; monitoring was set at 254 nm with a gradient of 10 mM aqueous HCOOH/NH 3 containing 0.1% HCOOH (eluent A) and MeCN (eluent B) from A: B = 95:5 to 20:80 and 20:80. 1 H-, 19 F-, and 13 C-NMR data were recorded on Bruker AVANCE III-400 instrument. Chemical shifts are presented in ppm (d), referenced to SiMe 4 (TMS). Optical rotations data were conducted on a DIP-370 instrument. Melting points were taken as usual.
To the above (S)-6,6,6-trifluoro-norleucine HCl green solution were added ethylenediaminetetraacetic acid disodium salt hydrate (10.5 g, 1.0 equiv) and acetonitrile (60 mL) and the mixture was stirred for 0.5 h at room temperature; 48% NaOH (9.5 g, 4.1 equiv) was added. Then, sodium carbonate (3.87 g, 1.3 equiv) and Fmoc-OSu (9.48 g, 1.0 equiv.) were added to the resulting mixture. The mixture was stirred for 3 h at room temperature, and then was concentrated. To the residue was added ethyl acetate (100 mL) and HCl (6N, 20.0 mL), and the phases were separated. The aqueous layer was extracted with ethyl acetate (40 mL) and the combined organic layer was washed with water (40 mL) and 10 % brine (40 mL). The organic solution was dried with Na 2 SO 4 , and then the filtrate was concentrated to dryness and dried in vacuo at 50 • C to afford (S)-9 (11.45 g, a white powder) (see Supplementary Materials).

Results and Discussion
In line with our longstanding curiosity in synthesis of several types of tailor-made AAs, in particular trifluoromethyl-and [49,50] phosphorus-containing [51,52], sterically constrained [53,54], and nonlinear optical properties of AA and their derivatives, such as self-disproportionation of enantiomers [55][56][57], we were contributing to the chemistry of Ni(II) complexes of Schiff bases of AA as a general methodology for synthesis of tailor-made AAs [58][59][60]. Over the last several years, we were focusing on the modular design [61,62] of chiral tridentate ligands used for preparation of the corresponding Ni(II) complexes of AA Schiff bases. Among other advances [63][64][65], recently we developed a strategically trichloro-substituted ligand 5 (Scheme 2) [66,67], which showed excellent stereocontrolling properties in the dynamic kinetic resolution of unprotected α- [68,69] and β-AAs [70]. It was shown that the presence of strategically positioned chlorine atoms favorably influence the parallel displaced type of aromatic stacking interactions between the proline N-benzyl and o-amino-benzophenone ring [66]. The quality of these aromatic stacking has important synthetic consequences [67] enhancing the stereochemical preferences at the a-position of the amino acid residue. It should be mentioned that the strategically chlorinated ligand 5 was developed by and commercially available from Hamari Chemicals. However, other methodological avenues of ligand 5 applications for the asymmetric preparation of tailor-made AAs still remain unexplored [71]. Ligands (S)-or (R)-5 are commercially available and can be conveniently prepared [72] starting form (S)-or (R)-proline and transformed to the Ni(II) complexes of glycine Schiff base 4 [71,73]. As presented in Scheme 2, ligand 5 is reacting with glycine and Ni(II) ions in basic methanol solution to afford complex 4. for the asymmetric preparation of tailor-made AAs still remain unexplored [71]. Ligands (S)-or (R)-5 are commercially available and can be conveniently prepared [72] starting form (S)-or (R)-proline and transformed to the Ni(II) complexes of glycine Schiff base 4 [71,73]. As presented in Scheme 2, ligand 5 is reacting with glycine and Ni(II) ions in basic methanol solution to afford complex 4.

Scheme 2. Synthesis of chiral Ni(II) complex of glycine Schiff base 4.
Alkylation of the glycine moiety in complexes of type 4 with alkyl halide as alkyl precursor can be conducted under homogeneous [74] as well as phase-transfer catalysis (PTC) conditions [75]. The latter are usually preferred, due to the low byproducts formation, but can be realized only for activated alkyl halides. Thus, under standard PTC conditions [75], CF3(CH2)3I was found to be totally inefficient for alkylation of complex (S)-4, resulting in noticeable decomposition of the alkylating reagent. In sharp contrast, under the homogeneous conditions (Scheme 3), by use of DMSO as a solvent and NaOH as a base (Table 1), the expected alkylation products were isolated and fully characterized. Alkylation of the glycine moiety in complexes of type 4 with alkyl halide as alkyl precursor can be conducted under homogeneous [74] as well as phase-transfer catalysis (PTC) conditions [75]. The latter are usually preferred, due to the low byproducts formation, but can be realized only for activated alkyl halides. Thus, under standard PTC conditions [75], CF 3 (CH 2 ) 3 I was found to be totally inefficient for alkylation of complex (S)-4, resulting in noticeable decomposition of the alkylating reagent. In sharp contrast, under the homogeneous conditions (Scheme 3), by use of DMSO as a solvent and NaOH as a base (Table 1), the expected alkylation products were isolated and fully characterized. for the asymmetric preparation of tailor-made AAs still remain unexplored [71]. Ligands (S)-or (R)-5 are commercially available and can be conveniently prepared [72] starting form (S)-or (R)-proline and transformed to the Ni(II) complexes of glycine Schiff base 4 [71,73]. As presented in Scheme 2, ligand 5 is reacting with glycine and Ni(II) ions in basic methanol solution to afford complex 4.

Scheme 2. Synthesis of chiral Ni(II) complex of glycine Schiff base 4.
Alkylation of the glycine moiety in complexes of type 4 with alkyl halide as alkyl precursor can be conducted under homogeneous [74] as well as phase-transfer catalysis (PTC) conditions [75]. The latter are usually preferred, due to the low byproducts formation, but can be realized only for activated alkyl halides. Thus, under standard PTC conditions [75], CF3(CH2)3I was found to be totally inefficient for alkylation of complex (S)-4, resulting in noticeable decomposition of the alkylating reagent. In sharp contrast, under the homogeneous conditions (Scheme 3), by use of DMSO as a solvent and NaOH as a base (Table 1), the expected alkylation products were isolated and fully characterized. Scheme 3. Alkylation of (S)-3 with CF3(CH2)3I under homogeneous conditions. Scheme 3. Alkylation of (S)-3 with CF 3 (CH 2 ) 3 I under homogeneous conditions. HPLC analysis of the reaction mixture revealed rather low rate of the alkylation and disappointing diastereoselectivity. As shown in Table 1, in entries 1-3, after 4.5 h of the reaction time, over 43% of starting complex 4 was still intact. Near-complete consumption of glycine complex 4 was observed after 24 h (entry 4) along with a large amount of various byproducts. One of the major byproducts was identified as previously described [73] compound 8 resulting from oxidative decomposition of stating glycine complex 4 [76,77]. Table 1. Reaction of complex (S)-4 with CF 3 (CH 2 ) 3 I in DMSO using solid NaOH as a base [a] . One of the critical notes made in this series of experiment was the observation that solid NaOH is not the best choice of introducing the base into the reaction mixture. After a series of experiments focused on solvent/base issue, we found that combination of DMF as a reaction solvent and solution of NaOMe in MeOH as a base allows for a dramatically improved outcome. Thus, as presented in Table 2, the alkylation of (S)-4 with CF 3 (CH 2 ) 3 I performed in DMF and using NaOMe/MeOH (28% solution) proceeded with high rate providing for virtually complete (>99%) consumption of the starting materials within about 30 min (entry 1). Importantly, the amount of byproducts was also dramatically reduced, albeit the stereochemical outcome was rather marginal (90:10 dr). Interestingly, extension of the reaction time from 0.5 to 2.0 h did not result in any visible changes of the chemical or stereochemical outcome (entry 2). Additional experiments with combination of DMF/NaOMe indicated that the application of less concentrated solution of NaOMe, has some advantageous effect on the reaction outcome. As shown in Table 2 (entry 3), the use of 10% NaOMe solution in MeOH as a base resulted in almost complete alkylation of glycine complex (S)-4 with CF 3 (CH 2 ) 3 I in less than 30 min of the reaction time. Similar to the previous experiments (entries 1 and 2) the alkylation proceeded rather cleanly, but most notably with rather improved diastereoselectivity (97:3 dr). Also in this case, the extended reaction time has no detrimental effect on the overall outcome. Using these conditions we were able to isolate diastereomerically pure major product 6 with reasonably good chemical yield of 87.5%. Diastereomers The obtained data are in agreement of general trends in optical rotation observed for diastereomeric Ni(II)-complexes of this type [1f,5i,20]. As presented in Scheme 4, the disassembly of purified diastereomerically pure (>98% de) major product (S,2S)-6 was performed under the action of 3N aqueous HCl at 60 • C using dimethoxyethane (DME) as organic solvent. Virtually complete disappearance of complex (S,2S)-6 was observed within about 2 h of the reaction time. Upon cooling of the reaction mixture, the precipitate of salt of ligand (S)-5 was conveniently removed by filtration. The aqueous solution of the Ni(II) ions and free (S)-1, was concentrated, and treated with Fmoc-OSu in MeCN/H 2 O to provide the N-Fmoc protected amino acid (S)-9.

Compounds
Symmetry 2019, 11, x FOR PEER REVIEW 6 of 10 As presented in Scheme 4, the disassembly of purified diastereomerically pure (>98% de) major product (S,2S)-6 was performed under the action of 3N aqueous HCl at 60 °C using dimethoxyethane (DME) as organic solvent. Virtually complete disappearance of complex (S,2S)-6 was observed within about 2 h of the reaction time. Upon cooling of the reaction mixture, the precipitate of salt of ligand (S)-5 was conveniently removed by filtration. The aqueous solution of the Ni(II) ions and free (S)-1, was concentrated, and treated with Fmoc-OSu in MeCN/H2O to provide the N-Fmoc protected amino acid (S)-9.

Conclusions
In summary, it was found that the application of new generation of chiral glycine equivalent (S)-4, prepared from commercially available ligand (S)-5, allows for convenient preparation of Fmoc-(S)-6,6,6-trifluoro-norleucine via alkylation reaction with CF3(CH2)3I. This protocol was consistently reproduced for synthesis of the target AA on ~10 g scale.
Supplementary Materials: The following are available online at www.mdpi.com/xxx/s1, Experimental procedures, full spectroscopic data for compounds 6, 7, and 9, and copies of 1 H NMR, 13 C, 19 F NMR, and HPLC spectra (PDF).

Conclusions
In summary, it was found that the application of new generation of chiral glycine equivalent (S)-4, prepared from commercially available ligand (S)-5, allows for convenient preparation of Fmoc-(S)-6,6,6-trifluoro-norleucine via alkylation reaction with CF 3 (CH 2 ) 3 I. This protocol was consistently reproduced for synthesis of the target AA on~10 g scale.