Synthesis, characterization and antibacterial activities of N-tert-butoxycarbonyl-thiazolidine carboxylic acid

Abstract

A possible mechanism of dynamic kinetic resolution by the formation of N-tert-butoxycarbonyl-thiazolidine carboxylic acid was proposed and validated by a quantitative density functional theoretical calculation according to Curtin–Hammett principle. Such a mechanism of action of a nucleophilic substitution reaction through an intramolecular hydrogen bonding could be widely applied in the organic syntheses of particular enantiomer. Antibacterial activities showed that most of the N-tert-butoxycarbonyl-(2R)-arylthiazolidine-(4R)-carboxylic acid derivatives exhibited better antibacterial activities against the four bacterial strains than related (2RS)-arylthiazolidine-(4R)-carboxylic acid derivatives.

---

1 Introduction

Kinetic resolution has long been recognized as an effective tool for the preparation of enantiomerically pure compounds for fine chemicals (i.e. agrochemicals and pharmaceuticals) and material science (i.e. liquid crystals and polymers).^1,2 However, the main drawback of the conventional kinetic resolution was that the obtained maximum theoretical yield of one stereoisomer of the starting material or product could be 50%, which was too low to allow a positive economic and environmental balance for such transformations.³ To overcome this limitation, of all the strategies that allow the transformation of both enantiomers of a racemate into a single enantiomer of the product, dynamic kinetic resolution (DKR) has been the focus of many studies in recent years. With DKR, the reactant isomers must be in rapid equilibrium, making removal of one isomer the rate determining step, and the product could be obtained optically pure in 100% yield (Fig. 1).^4,5

Fig. 1 Schematic representation of a dynamic kinetic resolution (DKR).

The steric course of the reaction between L-cysteine and aldehydes deserved much attention because this reaction had been implicated in several biochemical processes.^6–8 An analogous condensation reaction constituted the first step in the syntheses of important natural products, such as penicillin and biotin.^9,10 Therefore, thiazolidine derivatives have become especially noteworthy in recent years.^11–13

In this paper, because of the wide application of the tert-butoxycarbonyl (Boc) protecting group, under the guidance of the Curtin–Hammett principle, we synthesized a series of N-Boc-(2R)-aryl-thiazolidine-(4R)-carboxylic acids (N-Boc-TCAs) compounds using the control factors of intramolecular hydrogen bond induced-steric-hindrance, combined with dynamic kinetic resolution.^14–18 The corresponding reaction mechanism and the corresponding theoretical explanation were proposed. In order to compare the differences of antibacterial activities of N-Boc-TCAs and TCAs, we tested the antibacterial activities of them against the four bacteria, Gram positive (Bacillus subtilis ATCC 6633 and Staphylococcus aureus ATCC 6538) and Gram-negative (Pseudomonas aeruginosa ATCC 13525 and Escherichia coli ATCC 35218), which often caused infection in burned patients.

2 Results and discussion

2.1 Chemistry

The synthetic route to N-Boc-TCAs (13–24) is shown in Scheme 1. The syntheses of TCAs (1–12) were carried out using our previous literature method,¹⁶ by reaction of commercially available aromatic aldehyde and L-cysteine in the presence of ethanol with excellent yield. Treatment of 1–12 with t-butyloxycarbonyl anhydride ((Boc)₂O) yielded N-Boc-TCAs (13–24).¹⁶ The structures of the synthesized compounds (1–24) were listed in Table 1. All of the synthetic compounds were characterized by spectroscopic data, which were full accordance with their structures.

Scheme 1 Syntheses of 2-aryl-N-Boc-1,3-thiazolidine acid (TCAs) 13–24.

Table 1 The structures of the synthesized compounds (1–24)

Compounds		R1	R2	R3	R4	R5
1	13	H	H	H	H	H
2	14	H	H	OH	H	H
3	15	OMe	H	H	H	H
4	16	H	H	OMe	H	H
5	17	H	F	H	H	H
6	18	H	H	F	H	H
7	19	Cl	H	H	H	H
8	20	H	Cl	H	H	H
9	21	H	H	Cl	H	H
10	22	Br	H	H	H	H
11	23	H	H	Br	H	H
12	24	OH	H	H	F	H

---

From the ¹H NMR results, the unprotected TCA compounds 2-aryl-1,3-thiazolidine acids were obtained two diastereomers. The ratio of the most two diastereomers is nearly to 1 : 1. For example, from the Fig. 2 and 3, the ¹H NMR of the new born 4-position H (CH) of the thiazolidine ring in the compound 10 has two peaks and the chemical shift values are 5.90 and 5.72. However, the ¹H NMR of the 4-position H (CH) of the thiazolidine ring in the compound 22 only has one single peak and the chemical shift value is 6.12. It suggested that the unprotected TCA compounds 2-aryl-1,3-thiazolidine acids have two diastereomers and the protected TCA compound 2-aryl-N-Boc-1,3-thiazolidine acid is single chiral isomer.

Fig. 2 ¹H NMR of the compound 10.

Fig. 3 ¹H NMR of the compound 22.

2.2 Crystal structure analysis

In order to study the absolute structure of the product, we cultured some of the crystal structure of N-Boc-TCAs which are helpful to explain the mechanism of formation of N-Boc-TCAs. Fortunately, X-ray single-crystal diffraction data for compounds 13, 15, 21 and 22 were collected on a Nonius CAD4 diffractometer equipped with graphite-monochromatized MoKα (λ = 0.71073 Å) radiation. The program CAD4 software was used for data collection and cell refinement. Data reduction was solved by XCAD4 program. Structure was solved by direct methods using the SHELXS program of the SHELXTL package and refined by full-matrix least-squares methods with SHELXL.¹⁹

The selected bond distances, bond angles and torsion angles are given in Table 2. Crystal structure data of compounds 13, 15, 21 and 22 are given in Table 3. Fig. 4–7 gave typically perspective view of compounds 13, 15, 21 and 22 with the atomic labeling system, respectively.†

Table 2 Selected bond lengths (Å), bond angles (°) and torsion angles (°) for compounds 13, 15, 21 and 22^a

	13	15	21	22
a #1 represents the calculate value of space distance; #2 represents the mean deviation distances from the surface; #3 represents the dihedral angle of two ring surfaces Φ1 and Φ2.
H(7)⋯H(8)^#1	3.418	3.550	3.456	3.481
H(7)⋯C(10)^#1	4.227	4.447	4.247	4.278
H(8)⋯C(6)^#1	4.444	4.262	4.444	4.397
N(1)–C(7)–C(6)	116.5	116.2	117.2	112.9
N(1)–C(7)–S(1)	103.5	103.3	102.2	107.6
C(6)–C(7)–S(1)	110.8	113.6	109.2	101.9
C(6)–C(7)–H(7)	108.6	107.8	109.3	111.3
N(1)–C(7)–H(7)	108.6	107.8	109.3	111.3
S(1)–C(7)–H(7)	108.6	107.8	109.3	111.3
N(1)–C(8)–C(9)	105.7	105.5	106.6	106.1
N(1)–C(8)–C(10)	114.2	114.0	115.7	113.3
C(9)–C(8)–C(10)	111.1	109.7	109.6	118.3
C(9)–C(8)–H(8)	108.6	109.2	108.2	106.1
N(1)–C(8)–H(8)	108.6	109.2	108.2	106.1
C(10)–C(8)–H(8)	108.6	109.2	108.2	106.1
Φ1[C(1)–C(2)–C(3)–C(4)–C(5)–C(6)]^#2	0.0059	0.0049	0.0052	0.0097
Φ2[C(7)–N(1)–C(8)–C(9)–S(1)]^#2	0.2078	0.2003	0.1972	0.2035
Φ1∩Φ2^#3	98.8	101.2	77.2	69.6

---

Table 3 Crystal structure data of compounds 13, 15, 21 and 22

Compound	13	15	21	22
Formula	C15H19NO4S	C16H21NO5S	C15H18ClNO4S	C15H18BrNO4S
Mr	309.37	339.40	343.81	388.27
Crystal size (mm^3)	0.30 × 0.20 × 0.10	0.30 × 0.20 × 0.10	0.30 × 0.20 × 0.10	0.20 × 0.10 × 0.10
Crystal system	Monoclinic	Orthorhombic	Monoclinic	Monoclinic
Space group	P21	P212121	P21	P21
a (Å)	10.308(2)	6.4240(13)	6.4600(13)	10.485(2)
b (Å)	6.3790(13)	10.255(2)	10.641(2)	6.6020(13)
c (Å)	12.174(2)	27.064(5)	12.411(3)	12.803(3)
α (°C)	90.00	90.00	90.00	90.00
β (°C)	100.71(3)	90.00	94.52(3)	103.81(3)
γ (°C)	90.00	90.00	90.00	90.00
V (Å^3)	786.6(3)	1782.9(6)	850.5(3)	860.6(3)
Z	2	4	2	2
Dc (g cm^−3)	1.306	1.264	1.343	1.498
μ (mm^−1)	0.220	0.204	0.363	2.525
F(000)	328	720	360	396
θ range (°)	1.70/25.26	1.50/25.16	1.65/25.26	1.64/25.16
Index range (h, k, l)	−12/12, 0/7, 0/14	0/7, 0/12, 0/32	−7/7, 0/12, 0/14	−12/12, 0/7, 0/15
Reflections collected/unique	1660/1571	1887/1648	1638/1462	1687/1459
R(int)	0.0406	0.0248	0.0461	0.0362
Observed reflections I > 2σ(I)	1348	1317	1363	1281
Min. and max. transmission	0.9783 and 0.9369	0.9798 and 0.9412	0.9646 and 0.8989	0.7863 and 0.6321
Data/restraints/parameters	1571/49/190	1887/43/207	1638/78/185	1687/61/203
Goodness-of-fit on F^2	1.005	1.092	1.084	1.036
R1, wR2 [I > 2σ(I)]	0.0524/0.1273	0.0726/0.1909	0.0619/0.1453	0.0525/0.1233
R1, wR2	0.0674/0.1558	0.0998/0.2200	0.0761/0.1592	0.0728/0.1363
Absolute structure parameter	0.0(2)	0.1(3)	−0.09(19)	−0.02(2)
Large diff. peak and hole (e Å^−3)	0.354/−0.228	0.428/−0.387	0.435/−0.368	0.383 and −0.480

---

Fig. 4 Molecular structure (thermal ellipsoids at the 35% probability level) of compound 13.

Fig. 5 Molecular structure (thermal ellipsoids at the 35% probability level) of compound 15.

Fig. 6 Molecular structure (thermal ellipsoids at the 35% probability level) of compound 21.

Fig. 7 Molecular structure (thermal ellipsoids at the 35% probability level) of compound 22.

From Table 2 and Fig. 4, for compound 13, we could see that the surface mean deviation angles of benzene ring C1–C2–C3–C4–C5–C6 defined as Φ₁ and ring C7–N1–C8–C9–S1 defined as Φ₂ were 0.0060 Å and 0.2079 Å respectively. In addition, the dihedral angle of surfaces Φ₁ and Φ₂ was 98.8°, demonstrating that the two surfaces was almost in a vertical position, which arranged in the structure of stable conformations, avoiding a large steric effect of the substituents between the benzene ring and the newly formed five-membered ring. The space distance length of H7⋯H8 was 3.418 Å by calculating value, while that of H7⋯C10 was 4.227 Å, which showed that the space distance length of the latter is longer than that of the former, therefore, atom H7 and atom C10 were indicated in the opposite position of ring Φ₂. By comparing the bond angles with C7 and C8 as the centers (see Table 2), every bond angle was almost near to 109.5°, which demonstrated that C7 and C8 were sp³-hybrid. Because of H7 and C10 on both side of ring Φ₂, H7 and H8 were on same side of ring Φ₂. From the starting structure of compound L-cysteine, the configuration of C8 of compound 13 was (R)-configuration. Therefore, the configuration of C7 of compound 13 also was (R)-configuration according to IUPAC nomenclature. Furthermore, from Table 3, we can see that the Flack parameter of crystal 13 is 0.0(2). Consequently to literature, compound 13 can be showed a (2R,4R)-configuration as a single chiral isomer.^20–22 Likewise, from Table 3, Fig. 5–7, the crystal structures of the other three compounds 15, 21 and 22 were analyzed based on the same consideration. The result is that compounds 15, 21 and 22 is (2R,4R)-configuration as a single chiral isomer respectively.

2.3 Mechanism of reaction

To study the cause of the single chiral isomer, we referred corresponding mechanism of reaction of N-Boc-TCAs. As showed in Fig. 8, the reaction of (R)-cysteine and aryl aldehyde produced the Schiff base intermediate I, which had two epimers (Z)-II and (E)-II. (2S,4R)-III and (2R,4R)-III were generated through intramolecular hydrogen bond activation respectively. (2S,4R)-III and (2R,4R)-III were alkalified and then (2S,4R)-IV and (2R,4R)-IV were also formed through intramolecular hydrogen bonding, respectively. Theoretically, (2S,4R)-V and (2R,4R)-V should be rapidly equilibrated into an approximately 1 : 1 mixture like (2S,4R)-III and (2R,4R)-III. However, one pair of electrons of N atom of (2S,4R)-IV or (2R,4R)-IV attacked C atom of the carbonyl of di-tert-butyl carbonate through an S_N2 substitution reaction. This N atom was also a fictitious chiral center. Because of sterical intramolecular hydrogen bond activation, one pair of electrons of N atom of (2S,4R)-IV or (2R,4R)-IV must assault C atom of the carbonyl of di-tert-butyl carbonate from the back of the ring Φ₂. Nevertheless, the sterically crowded effect of suprafacial aryl of (2S,4R)-IV was larger than that of different aryl surface of (2R,4R)-IV. Obviously, the rate of step (2R,4R)-IV → (2R,4R)-V should be much faster than that of step (2S,4R)-IV → (2S,4R)-V and the rate constants named k₂ and k₁ respectively. So, k₂ was much larger than k₁ (k₂ ≫ k₁). Because all the reactions were synthesized at room temperature, the reaction rate was assumed to be controlled by dynamic kinetic resolution effect. Therefore, the reaction rate of (2S,4R)-V → (2S,4R)-IV → (2S,4R)-III → (Z)-II → Schiff base intermediate I → (E)-II → (2R,4R)-III → (2R,4R)-IV → (2R,4R)-V should be much faster than that of inverse sequence. As a result, (2S,4R)-V was almost converted into (2R,4R)-V. As acquired, (2R)-aryl-3-tert-butoxycarbonyl-1,3-thiazolidine-(4R)-carboxylic acid as the single product was obtained because of dynamic kinetic resolution effect.

Fig. 8 The proposed mechanism of formation of compounds 13–24.

A corresponding qualitative explanation from the theory was interpretated according to dynamic kinetic resolution and Curtin–Hammett principle combined with above reaction mechanism, see Fig. 9.^4,5 At first, we found the potential energy was very low in the transformation process of intermediate (2S,4R)-IV ↔ (2S,4R)-III ↔ (Z)-II ↔ Schiff base intermediate I ↔ (E)-II ↔ (2R,4R)-III ↔ (2R,4R)-IV (Fig. 8 and 9). So, the above transformation process could be simplified (2S,4R)-IV ↔ (2R,4R)-IV (Fig. 10).

Fig. 9 Curtin–Hammett mechanism of 2-aryl-N-Boc-1,3-thiazolidine derivatives. ΔG^‡₁ represents the Gibbs free potential energy of (2S,4R)-IV → (2S,4R)-V, ΔG^‡₂ represents the Gibbs free potential energy of (2R,4R)-IV → (2R,4R)-V; ΔG^o represents the Gibbs free potential energy of (2S,4R)-IV → (2R,4R)-IV.

Fig. 10 Simplified reaction mechanism of 2-aryl-N-Boc-1,3-thiazolidine derivatives.

According to the relationship between equilibrium constant and Gibbs free energy, we can get eqn (1). The relationship between equilibrium constant and Gibbs free energy:

In the theoretical model, the Ar group was simplified to Ph, and the (Boc)₂O was simplified to (MeOCO)₂O. All calculations were performed by Gaussian03 (ref. 23) software suite. Geometries of intermediates and transition states were fully optimized without any constraints using B3LYP^24,25 functional with 6-31+G(d) basis set. All geometries were confirmed by frequency analysis in which an intermediate has no negative frequency and a transition state has only one negative frequency corresponding to the reaction coordinate. The thermodynamic corrections for Gibbs free energy at 298.15 K were calculated at the same level of geometrical optimizations. Single point calculations were conducted using B3LYP functional with a larger basis set, 6-311+G(2d,p). The solvent effects were addressed by the polarizable continuum model using the integral equation formalism variant (IEFPCM) with UAHF radii.²⁶ The structures of transition state are depicted in Fig. 11 and 12.

Fig. 11 The structure of transition state leading to (2S,4R)-V. Key bond lengths were shown in the unit of Angstrom.

Fig. 12 The structure of transition state leading to (2R,4R)-V. Key bond lengths were shown in the unit of Angstrom.

According to the calculate results, ΔG^‡₁, ΔG^‡₂ and ΔG^o were 174.7 kJ mol⁻¹, 149.5 kJ mol⁻¹ and 3.3 kJ mol⁻¹, respectively. All data were substituted to eqn (1), and the value of [(2R,4R)-V]/[(2S,4R)-V] was 6705. The percentage of (2R,4R)-V accounted the entire system ((2R,4R)-V and (2S,4R)-V) of 99.99%.

Therefore, the yield of (2R,4R)-V is much more than that of (2S,4R)-V in the case of such high steric hindrance potential energy. As the above analysis, the single chiral isomer 2-aryl-N-Boc-thiazolidine derivatives were gained under the control factors of intramolecular hydrogen bond induced-steric-hindrance.

2.4 In vitro antibacterial assay

The activities of synthesized N-Boc-TCAs were tested against Gram positive (Bacillus subtilis ATCC 6633 (B. subtilis) and Staphylococcus aureus ATCC 6538 (S. aureus)) and Gram-negative (Pseudomonas aeruginosa ATCC 13525 (P. aeruginosa) and Escherichia coli ATCC 35218 (E. coli)), which often caused infection in burned patients by MTT method. The IC₅₀s of the compounds against four bacteria are presented in Table 4. Also included are the activities of reference compounds Kanamycin G and Penicillin B. Antibacterial activities showed that most of the N-Boc-TCAs exhibited better antibacterial activities against the four bacterial strains than relative TCAs.

Table 4 IC₅₀ values of the synthesized TCAs 1–12 and N-Boc-TCAs 13–24 against the four bacteria

Compound	IC50s (μg mL^−1)
	Gram positive		Gram negative
	B. subtilis	S. aureus	P. aeruginosa	E. coli
1	3.125	12.5	>50	25
2	25	>50	25	>50
3	6.25	12.5	12.5	6.25
4	12.5	25	>50	>50
5	3.125	6.25	1.562	6.25
6	6.25	12.5	3.125	25
7	12.5	12.5	12.5	25
8	6.25	12.5	6.25	12.5
9	12.5	12.5	12.5	50
10	25	25	25	50
11	25	25	12.5	25
12	3.125	6.25	1.562	12.5
13	3.125	0.78	0.78	0.39
14	0.39	3.125	0.39	0.78
15	1.562	1.562	0.78	0.39
16	3.125	1.562	0.39	0.78
17	0.39	0.39	0.195	1.562
18	1.562	0.78	0.78	1.562
19	1.562	1.562	0.78	0.78
20	0.78	0.78	0.39	0.39
21	1.562	0.78	0.78	1.562
22	1.562	3.125	0.78	1.562
23	3.125	3.125	0.39	0.39
24	0.39	0.39	0.195	0.78
Kanamycin G	0.39	1.562	3.125	3.125
Penicillin B	1.562	1.562	6.25	6.25

---

3 Conclusions

In summary, the mechanism of dynamic kinetic resolution by the formation of N-Boc-TCAs was proposed and a quantitative theoretical calculation was interpretated by Gaussian03 software suite according to Curtin–Hammett principle. Such a mechanism of a nucleophilic substitution reaction through an intramolecular hydrogen bonding could be widely applied in the organic syntheses of special enantiomers. Antibacterial activities showed that most of the N-Boc-TCAs exhibited better antibacterial activities against Gram positive bacteria (B. subtilis and S. aureus) and Gram-negative bacteria (P. aeruginosa and E. coli) than relative TCAs.

4 Experimental section

4.1 Chemistry

All the reagents and solvents used were analytical reagent grade or were purified by standard methods before use. ESI mass spectra were obtained on a Mariner System 5304 mass spectrometer, and ¹H NMR was recorded at DPX300 in DMSO-d₆ or CDCl₃ on a Bruker spectrometer. Elemental analyses were performed on a CHN-O-Rapid instrument and were within ±0.4% of the theoretical values.

General procedure for the preparation of TCA derivatives 1–12. A mixture of L-cysteine (0.121 g, 1.0 mmol) and appropriate aldehyde (1.0 mmol) in ethanol (25 mL) was stirred at room temperature for 8 h, and the solid separated was collected, washed with diethyl ether and dried to obtain 1–12.
4.1.1 (4R)-2-Phenylthiazolidine-4-carboxylic acid (1). Yield 84%. ¹H NMR (300 MHz, DMSO-d₆) δ 7.53–7.24 (m, 5H), 5.68 (s, 0.5H), 5.50 (s, 0.5H), 4.26–4.22 (m, 0.5H), 3.93–3.88 (m, 0.5H), 3.41–3.27 (m, 1H), 3.17–3.05 (m, 1H). MS (ESI) m/z 210 (M + 1). Anal. calcd for C₁₀H₁₁NO₂S: C, 57.39; H, 5.30; N, 6.69. Found: C, 57.62; H, 5.28; N, 6.71.
4.1.2 (4R)-2-(4-Hydroxyphenyl)thiazolidine-4-carboxylic acid (2). Yield 87%. ¹H NMR (300 MHz, DMSO-d₆) δ 7.32–7.23 (m, 2H), 6.75–6.69 (m, 2H), 5.53 (s, 0.5H), 5.39 (s, 0.5H), 4.27–4.23 (m, 0.5H), 3.86–3.81 (m, 0.5H), 3.37–3.24 (m, 1H), 3.17–3.01 (m, 1H). MS (ESI) m/z 226 (M + 1). Anal. calcd for C₁₀H₁₁NO₃S: C, 53.32; H, 4.92; N, 6.22. Found: C, 53.50; H, 4.94; N, 6.20.
4.1.3 (4R)-2-(2-Methoxyphenyl)thiazolidine-4-carboxylic acid (3). Yield 82%. ¹H NMR (300 MHz, DMSO-d₆) δ 7.51–7.19 (m, 2H), 7.03–6.89 (m, 2H), 5.85 (s, 0.6H), 5.68 (s, 0.4H), 4.18–4.13 (m, 0.6H), 3.87–3.84 (m, 0.4H), 3.79 + 3.78 (s, 3H), 3.36–3.15 (m, 1H), 2.99–2.93 (m, 1H). MS (ESI) m/z 240 (M + 1). Anal. calcd for C₁₁H₁₃NO₃S: C, 55.21; H, 5.48; N, 5.85; found: C, 55.51; H, 5.46; N, 5.87.
4.1.4 (4R)-2-(4-Methoxyphenyl)thiazolidine-4-carboxylic acid (4). Yield 81%. ¹H NMR (300 MHz, DMSO-d₆) δ 7.45–7.35 (m, 2H), 6.93–6.86 (m, 2H), 5.59 (s, 0.5H), 5.44 (s, 0.5H), 4.26–4.22 (m, 0.5H), 3.88–3.83 (m, 0.5H), 3.74 + 3.73 (s, 3H), 3.38–3.25 (m, 1H), 3.17–3.03 (m, 1H). MS (ESI) m/z 240 (M + 1). Anal. calcd for C₁₁H₁₃NO₃S: C, 55.21; H, 5.48; N, 5.85. Found: C, 55.43; H, 5.50; N, 5.83.
4.1.5 (4R)-2-(3-Fluorophenyl)thiazolidine-4-carboxylic acid (5). Yield 87%. ¹H NMR (300 MHz, DMSO-d₆) δ 7.42–7.24 (m, 3H), 7.17–7.05 (m, 1H), 5.69 (s, 0.6H), 5.51 (s, 0.4H), 4.20–4.16 (m, 0.6H), 3.92–3.87 (m, 0.4H), 3.37–3.26 (m, 1H), 3.13–3.05 (m, 1H). MS (ESI) m/z 226 (M − 1). Anal. calcd for C₁₀H₁₀FNO₂S: C, 52.85; H, 4.44; N, 6.16; found: C, 52.66; H, 4.46; N, 6.14.
4.1.6 (4R)-2-(4-Fluorophenyl)thiazolidine-4-carboxylic acid (6). Yield 82%. ¹H NMR (300 MHz, DMSO-d₆) δ 7.62–7.47 (m, 2H), 7.24–7.14 (m, 2H), 5.68 (s, 0.6H), 5.52 (s, 0.4H), 4.24–4.20 (m, 0.6H), 3.93–3.88 (m, 0.4H), 3.41–3.29 (m, 1H), 3.17–3.07 (m, 1H). MS (ESI) m/z 226 (M − 1). Anal. calcd for C₁₀H₁₀FNO₂S: C, 52.85; H, 4.44; N, 6.16; found: C, 52.99; H, 4.42; F, 8.36; N, 6.19.
4.1.7 (4R)-2-(2-Chlorophenyl)thiazolidine-4-carboxylic acid (7). Yield 83%. ¹H NMR (300 MHz, DMSO-d₆) δ 7.84–7.83 (m, 0.3H), 7.58–7.55 (m, 0.8H), 7.45–7.22 (m, 3H), 5.96 (s, 0.7H), 5.77 (s, 0.3H), 4.19–4.15 (m, 0.7H), 4.00–3.95 (m, 0.3H), 3.38–3.22 (m, 1H), 3.16–2.94 (m, 1H). MS (ESI) m/z 242 (M − 1). Anal. calcd for C₁₀H₁₀ClNO₂S: C, 49.28; H, 4.14; N, 5.75; found: C, 49.46; H, 4.17; N, 5.78.
4.1.8 (4R)-2-(3-Chlorophenyl)thiazolidine-4-carboxylic acid (8). Yield 81%. ¹H NMR (300 MHz, DMSO-d₆) δ 7.66–7.34 (m, 4H), 5.71 (s, 0.6H), 5.53 (s, 0.4H), 4.20–4.17 (m, 0.6H), 3.94–3.88 (m, 0.4H), 3.39–3.29 (m, 1H), 3.15–3.07 (m, 1H). MS (ESI) m/z 242 (M − 1). Anal. Calcd for C₁₀H₁₀ClNO₂S: C, 49.28; H, 4.14; N, 5.75; found: C, 49.54; H, 4.16; N, 5.77.
4.1.9 (4R)-2-(4-Chlorophenyl)thiazolidine-4-carboxylic acid (9). Yield 78%. ¹H NMR (300 MHz, DMSO-d₆) δ 7.55–7.53 (m, 1H), 7.46–7.35 (m, 3H), 5.69 (s, 0.6H), 5.50 (s, 0.4H), 4.19–4.15 (m, 0.6H), 3.92–3.86 (m, 0.4H), 3.38–3.26 (m, 1H), 3.13–3.04 (m, 1H). MS (ESI) m/z 242 (M − 1). Anal. calcd for C₁₀H₁₀ClNO₂S: C, 49.28; H, 4.14; N, 5.75; found: C, 49.60; H, 4.12; N, 5.77.
4.1.10 (4R)-2-(2-Bromophenyl)thiazolidine-4-carboxylic acid (10). Yield 73%. ¹H NMR (300 MHz, DMSO-d₆) δ 7.84–7.54 (m, 2H), 7.54–7.34 (m, 1H), 7.27–7.14 (m, 1H), 5.90 (s, 0.7H), 5.72 (s, 0.3H), 4.20–4.16 (m, 0.7H), 4.00–3.95 (m, 0.3H), 3.37–3.21 (m, 1H), 3.08–2.93 (m, 1H). MS (ESI) m/z 286 (M − 1). Anal. calcd for C₁₀H₁₀BrNO₂S: C, 41.68; H, 3.50; N, 4.86; found: C, 41.87; H, 3.48; N, 4.88.
4.1.11 (4R)-2-(4-Bromophenyl)thiazolidine-4-carboxylic acid (11). Yield 87%. ¹H NMR (300 MHz, DMSO-d₆) δ 7.58–7.38 (m, 4H), 5.68 (s, 0.6H), 5.50 (s, 0.4H), 4.19–4.15 (m, 0.6H), 3.93–3.88 (m, 0.4H), 3.39–3.35 (m, 1H), 3.14–3.035 (m, 1H). MS (ESI) m/z 286 (M − 1). Anal. calcd for C₁₀H₁₀BrNO₂S: C, 41.68; H, 3.50; N, 4.86. Found: C, 41.51; H, 3.52; N, 4.88.
4.1.12 (4R)-2-(5-Fluoro-2-hydroxyphenyl)thiazolidine-4-carboxylic acid (12). Yield 84%. ¹H NMR (300 MHz, DMSO-d₆) δ 9.89 (b, 0.6H), 7.32–7.27 (m, 0.4H), 7.16–7.12 (m, 0.6H), 7.01–6.87 (m, 1H), 6.84–6.75 (m, 1H), 5.84 (s, 0.6H), 5.66 (s, 0.4H), 4.22–4.18 (m, 0.6H), 3.90–3.85 (m, 0.4H), 3.26–3.260 (m, 1H), 3.05–2.98 (m, 1H). MS (ESI) m/z 242 (M − 1). Anal. calcd for C₁₀H₁₀FNO₃S: C, 49.37; H, 4.14; N, 5.76. Found: C, 49.56; H, 4.12; N, 5.74.

General procedure for the preparation of TBTCA derivatives 13–24. A mixture of TCA (1.0 mmol) and appropriate NaOH (10%, 1.0 mmol) in dioxane (25 mL) was stirred at ice-water temperature for 2 h. Boc₂O (1.0 mmol) was added and stirred at ice-water temperature for 1 h and then room temperature for 5 h. Most of solvent was extracted and appropriate amount of water was added to adjust to neutral pH values. Ethyl acetate was added and extracted (50 mL × 3), and washed with appropriate saturated aqueous solution of common salt, and dried with anhydrous magnesium sulphate. Solvent was extracted to dry to obtain whiter solids 13–24.
4.1.13 (2R,4R)-3-(tert-Butoxycarbonyl)-2-phenylthiazolidine-4-carboxylic acid (13). Yield 78%. ¹H NMR (300 MHz, DMSO-d₆ + CDCl₃) δ 7.56–7.53 (d, 2H), 7.24–7.14 (m, 3H), 6.04 + 5.85 (2 s, 1H), 4.75–4.57 (d, 1H), 3.23–3.16 (m, 2H), 1.34 + 1.10 (2 s, 9H). MS (ESI) m/z 308 (M − 1). Anal. calcd for C₁₅H₁₉NO₄S: C, 58.23; H, 6.19; N, 4.53. Found: C, 58.05; H, 6.15; N, 4.57.
4.1.14 (2R,4R)-3-(tert-Butoxycarbonyl)-2-(4-hydroxyphenyl)thiazolidine-4-carboxylic acid (14). Yield 76%. ¹H NMR (300 MHz, DMSO-d₆) δ 9.35 (b, 1H), 7.42–7.39 (d, 2H), 6.68–6.65 (d, 2H), 6.01 + 5.84 (2 s, 1H), 4.67–4.51 (m, 1H), 3.44–3.37 (m, 1H), 3.15–3.05 (m, 1H), 1.34 + 1.12 (2 s, 9H). MS (ESI) m/z 324 (M − 1). Anal. calcd for C₁₅H₁₉NO₅S: C, 55.37; H, 5.89; N, 4.30. Found: C, 55.62; H, 6.85; N, 4.32.
4.1.15 (2R,4R)-3-(tert-Butoxycarbonyl)-2-(2-methoxyphenyl)thiazolidine-4-carboxylic acid (15). Yield 84%. ¹H NMR (300 MHz, CDCl₃ + TMS) δ 7.83–7.60 (m, 1H), 7.277 (m, 1H), 7.26–7.23 (m, 1H), 6.98–6.85 (m, 1H), 6.38 + 6.29 (2 s, 1H), 4.84–4.71 (m, 1H), 3.86 (s, 3H), 3.40–3.26 (m, 2H), 1.42 + 1.23 (2 s, 9H). MS (ESI) m/z 338 (M − 1). Anal. calcd for C₁₆H₂₁NO₅S: C, 56.62; H, 6.24; N, 4.13; found: C, 56.85; H, 6.22; N, 4.15.
4.1.16 (2R,4R)-3-(tert-Butoxycarbonyl)-2-(4-methoxyphenyl)thiazolidine-4-carboxylic acid (16). Yield 80%. ¹H NMR (300 MHz, CD₂Cl₂) δ 7.45–7.44 (d, 2H), 6.85–6.84 (d, 2H), 5.95 (s, 1H), 4.83 (s, 1H), 3.77 (s, 3H), 3.38–3.31 (m, 2H), 1.27 (s, 9H). MS (ESI) m/z 338 (M − 1). Anal. calcd for C₁₆H₂₁NO₅S: C, 56.62; H, 6.24; N, 4.13. Found: C, 56.90; H, 6.21; N, 4.17.
4.1.17 (2R,4R)-3-(tert-Butoxycarbonyl)-2-(3-fluorophenyl)thiazolidine-4-carboxylic acid (17). Yield 77%. ¹H NMR (300 MHz, DMSO-d₆) δ 7.64–7.57 (m, 1H), 7.43–7.31 (m, 2H), 7.10–7.05 (m, 1H), 6.13 + 5.97 (2 s, 1H), 4.72–4.52 (m, 1H), 3.49–3.43 (m, 1H), 3.20–3.03 (m, 1H), 1.35 + 1.12 (2 s, 9H). MS (ESI) m/z 326 (M − 1). Anal. Calcd for C₁₅H₁₈FNO₄S: C, 55.03; H, 5.54; N, 4.28; found: C, 55.35; H, 5.56; N, 4.26.
4.1.18 (2R,4R)-3-(tert-Butoxycarbonyl)-2-(4-fluorophenyl)thiazolidine-4-carboxylic acid (18). Yield 75%. ¹H NMR (300 MHz, CD₂Cl₂) δ 8.80 (bs, 1H), 7.59 (s, 2H), 7.04 (t, 2H), 5.96 (s, 1H), 4.94 (s, 1H), 3.37 (s, 1H), 1.25 (s, 9H). MS (ESI) m/z 326 (M − 1). Anal. calcd for C₁₅H₁₈FNO₄S: C, 55.03; H, 5.54; N, 4.28; found: C, 55.33; H, 5.52; N, 4.30.
4.1.19 (2R,4R)-3-(tert-Butoxycarbonyl)-2-(2-chlorophenyl)thiazolidine-4-carboxylic acid (19). Yield 80%. ¹H NMR (300 MHz, CD₂Cl₂) δ 8.29 (b, 1H), 7.32–7.29 (m, 2H), 7.22–7.16 (m, 2H), 6.26 (s, 1H), 4.78–4.77 (m, 1H), 3.34–3.33 (m, 2H), 1.21 (s, 9H). MS (ESI) m/z 342 (M − 1). Anal. calcd for C₁₅H₁₈ClNO₄S: C, 52.40; H, 5.28; N, 4.07; found: C, 52.40; H, 5.28; N, 4.07.
4.1.20 (2R,4R)-3-(tert-Butoxycarbonyl)-2-(3-chlorophenyl)thiazolidine-4-carboxylic acid (20). Yield 81%. ¹H NMR (300 MHz, DMSO-d₆) δ 13.12 (b, 1H), 7.83–7.79 (d, 1H), 7.59–7.57 (d, 1H), 7.39–7.34 (t, 2H), 6.14 + 5.98 (2 s, 1H), 4.75–4.56 (m, 1H), 3.52–3.46 (m, 1H), 3.22–3.05 (m, 1H), 1.38–1.14 (2 s, 9H). MS (ESI) m/z 342 (M − 1). Anal. calcd for C₁₅H₁₈ClNO₄S: C, 52.40; H, 5.28; N, 4.07; found: C, 52.63; H, 5.26; N, 4.09.
4.1.21 (2R,4R)-3-(tert-Butoxycarbonyl)-2-(4-chlorophenyl)thiazolidine-4-carboxylic acid (21). Yield 73%. ¹H NMR (300 MHz, DMSO-d₆) δ 13.08 (b, 1H), 7.67–7.64 (d, 2H), 7.39–7.37 (d, 2H), 6.11 + 5.96 (2 s, 1H), 4.71–4.56 (m, 1H), 3.49–3.43 (m, 1H), 3.36 (m, 1H), 3.17–3.03 (m, 1H),1.34 + 1.12 (2 s, 9H). MS (ESI) m/z 342 (M − 1). Anal. calcd for C₁₅H₁₈ClNO₄S: C, 52.40; H, 5.28; N, 4.07; found: C, 52.16; H, 5.30; N, 4.05.
4.1.22 (2R,4R)-2-(2-Bromophenyl)-3-(tert-butoxycarbonyl)thiazolidine-4-carboxylic acid (22). Yield 84%. ¹H NMR (300 MHz, DMSO-d₆) δ 13.19 (b, 0.5H), 8.24–8.15 (m, 1H), 7.63–7.56 (m, 1H), 7.40–7.39 (m, 1H), 7.24–7.19 (m, 1H), 6.12 (s, 1H), 4.69–4.52 (m, 1H), 3.54–3.44 (m, 1H), 3.18–2.98 (m, 1H), 1.35 + 1.10 (2 s, 9H). MS (ESI) m/z 386 (M − 1). Anal. calcd for C₁₅H₁₈BrNO₄S: C, 46.40; H, 4.67; N, 3.61; found: C, 46.66; H, 4.69; N, 3.59.
4.1.23 (2R,4R)-2-(4-Bromophenyl)-3-(tert-butoxycarbonyl)thiazolidine-4-carboxylic acid (23). Yield 78%. ¹H NMR (300 MHz, DMSO-d₆) δ 13.07 (b, 0.5H), 7.60–7.58 (d, 2H), 7.52–7.50 (d, 2H), 6.09 + 5.95 (2 s, 1H), 4.70–4.54 (m, 1H), 3.48–3.42 (m, 1H), 3.17–3.03 (m, 1H), 1.34 + 1.12 (2 s, 9H). MS (ESI) m/z 386 (M − 1). Anal. calcd for C₁₅H₁₈BrNO₄S: C, 46.40; H, 4.67; N, 3.61. Found: C, 46.72; H, 4.49; N, 3.66.
4.1.24 (2R,4R)-3-(tert-Butoxycarbonyl)-2-(5-fluoro-2-hydroxyphenyl)thiazolidine-4-carboxylic acid (24). Yield 75%. ¹H NMR (300 MHz, DMSO-d₆) δ 7.80–7.76 (m, 1H), 6.91–6.84 (m, 1H), 6.76 (m, 1H), 6.08 (s, 1H), 4.61–4.45 (m, 1H), 3.45–3.39 (m, 1H), 3.11–2.96 (m, 1H), 1.35 + 1.15 (2 s, 9H). MS (ESI) m/z 342 (M − 1). Anal. calcd for C₁₅H₁₈FNO₅S: C, 52.47; H, 5.28; N, 4.08. Found: C, 52.11; H, 5.25; N, 4.12.

4.2 Crystallographic studies

X-ray single-crystal diffraction data for ‘2R, 4R’ isomer of compounds 13, 15, 21 and 22 were collected on a Nonius CAD4 diffractometer equipped with graphite-monochromatized MoKα (λ = 0.71073 Å) radiation. The program CAD4 software was used for data collection and cell refinement. Data reduction was solved by XCAD4 program. Structure was solved by direct methods using the SHELXS program of the SHELXTL package and refined by full-matrix least-squares methods with SHELXL.¹⁹ All non-hydrogen atoms of ‘2R, 4R’ isomer of compounds 13, 15, 21 and 22 were refined with anisotropic thermal parameters. All hydrogen atoms were placed in geometrically idealized positions and constrained to ride on their parent atoms.

4.3 In vitro antibacterial activity

The antibacterial activity of the synthesized compounds was tested against Gram positive bacteria (B. subtilis and S. aureus) and Gram-negative bacteria (P. aeruginosa and E. coli) using MH medium (Mueller-Hinton medium: casein hydrolysate 17.5 g, soluble starch 1.5 g, beef extract 1000 mL). The IC₅₀s of the tested compounds were determined by a colorimetric method using the dye MTT. A stock solution of the synthesized compound (50 μg mL⁻¹) in DMSO was prepared and graded quantities of the tested compounds were incorporated in specified quantity of sterilized liquid medium (MH medium for antibacterial activity). A specified quantity of the medium containing the compound was poured into microtitration plates. Suspension of the microorganism was prepared to contain approximately 10⁵ cfu mL⁻¹ and applied to microtitration plates with serially diluted compounds in DMSO to be tested and incubated at 37 °C for 24 h. After the IC₅₀s were visually determined on each of the microtitration plates, 50 mL of PBS (phosphate buffered saline 0.01 mol L⁻¹, pH 7.4, Na₂HPO₄·12H₂O (2.9 g), KH₂PO₄ (0.2 g), NaCl (8.0 g), KCl (0.2 g), distilled water (1000 mL) containing 2 mg of MTT per mL was added to each well. Incubation was continued at room temperature for 4–5 h. The content of each well was removed, and 100 mL of isopropanol containing 5% 1 mol L⁻¹ HCl was added to extract the dye. After 12 h of incubation at room temperature, the optical density (OD) was measured with a microplate reader at 570 nm. The observed IC₅₀s are presented in Table 4.

Acknowledgements

This work was supported by the National Natural Science Foundation of China (no. 21202153) and Foundation of Zhejiang Educational Committee (no. Y201223263). We also thank Dr Zhe Li (Department of Chemistry, Tsinghua University, Beijing) for the help of the theoretical model calculation.

Notes and references

1. H. Pellissier, Recent developments in dynamic kinetic resolution, Tetrahedron, 2008, 64, 1563–1601.
2. O. Pamies and J. E. Backvall, Chemoenzymatic dynamic kinetic resolution, Trends Biotechnol., 2004, 22, 130–135.
3. F. F. Huerta, Y. R. S. Laxmi and J. E. Backvall, Dynamic Kinetic Resolution of α-Hydroxy Acid Esters, Org. Lett., 2000, 2, 1037–1040.
4. A. Berkessel, F. Cleemann, S. Mukherjee, T. N. Muller and J. Lex, Bifunktionale Organokatalysatoren auf Harnstoff-Basis für die effiziente dynamische kinetische Racemattrennung von Azlactonen, Angew. Chem., 2005, 117, 659; A. Berkessel, F. Cleemann, S. Mukherjee, T. N. Muller and J. Lex, Highly Efficient Dynamic Kinetic Resolution of Azlactones by Urea-Based Bifunctional Organocatalysts, Angew. Chem., Int. Ed., 2005, 44, 807–811.
5. V. Chan, J. G. Kim, C. Jimeno, P. J. Carroll and P. J. Walsh, Dynamic Kinetic Resolution of Atropisomeric Amides, Org. Lett., 2004, 6, 2051–2053.
6. S. L. You and J. W. Kelly, Highly Efficient Biomimetic Total Synthesis and Structural Verification of Bistratamides E and J from Lissoclinum bistratum, Chem.–Eur. J., 2004, 10, 71–75.
7. L. Szilagyi and Z. Gyorgydeak, Comments on the putative stereoselectivity in cysteine-aldehyde reactions. Selective C(2) inversion and C(4) epimerization in thiazolidine-4-carboxylic acids, J. Am. Chem. Soc., 1979, 101, 427–432.
8. R. G. Kallen, Mechanism of reactions involving Schiff base intermediates. Thiazolidine formation from L-cysteine and formaldehyde, J. Am. Chem. Soc., 1971, 93, 6236–6248.
9. M. Seki, M. Hatsuda, Y. Mori, S. I. Yoshida, S. I. Yamada and T. Shimizu, A Practical Synthesis of (+)-Biotin from L-Cysteine, Chem.–Eur. J., 2004, 10, 6102–6110.
10. A. Wagner, R. Flaig, B. Dittrich, H. Schmidt, T. Koritsánszky and P. Luger, Charge Density and Experimental Electrostatic Potentials of Two Penicillin Derivatives, Chem.–Eur. J., 2004, 10, 2977–2982.
11. A. Serganov, A. Polonskaia, A. T. Phan, R. R. Breaker and D. J. Patel, Structural basis for gene regulation by a thiamine pyrophosphate-sensing riboswitch, Nature, 2006, 441, 1167–1171.
12. V. Sriramurthy, G. A. Barcan and O. Kwon, Bisphosphine-Catalyzed Mixed Double-Michael Reactions: Asymmetric Synthesis of Oxazolidines, Thiazolidines, and Pyrrolidines, J. Am. Chem. Soc., 2007, 129, 12928–12929.
13. V. Gududuru, E. Hurh, J. T. Dalton and D. D. Miller, Discovery of 2-Arylthiazolidine-4-carboxylic Acid Amides as a New Class of Cytotoxic Agents for Prostate Cancer, J. Med. Chem., 2005, 48, 2584–2588.
14. A. Traff, K. Bogar, M. Warner and J. E. Backvall, Highly Efficient Route for Enantioselective Preparation of Chlorohydrins via Dynamic Kinetic Resolution, Org. Lett., 2008, 10, 4807–4810.
15. A. Klapars, K. R. Campos, J. H. Waldman, D. Zewge, P. G. Dormer and C. Y. Chen, Enantioselective Pd-Catalyzed α-Arylation of N-Boc-Pyrrolidine: The Key to an Efficient and Practical Synthesis of a Glucokinase Activator, J. Org. Chem., 2008, 73, 4986–4993.
16. Z. C. Song, G. Y. Ma, P. C. Lv, H. Q. Li, Z. P. Xiao and H. L. Zhu, Synthesis, structure and structure–activity relationship analysis of 3-tert-butoxycarbonyl-2-arylthiazolidine-4-carboxylic acid derivatives as potential antibacterial agents, Eur. J. Med. Chem., 2009, 44, 3903–3908.
17. M. Seki, Y. Mori, M. Hatsuda and S. Yamada, A Novel Synthesis of (+)-Biotin from L-Cysteine, J. Org. Chem., 2002, 67, 5527–5536.
18. M. E. F. Braibante, H. S. Braibante and E. R. Costenaro, The Use of Curtius Rearrangement in the Synthesis of 4-Aminothiazolidines, Synthesis, 1999, 6, 943–946.
19. G. M. Sheldrick, SHELXTL-97, Program for Crystal Structure Solution and Refinement, University of Göttingen, Göttingen, Germany, 1997.
20. H. D. Flack, On enantiomorph-polarity estimation, Acta Crystallogr., Sect. A: Found. Crystallogr., 1983, 39, 876–881.
21. J. Zhang, S. Chen, T. Wu, P. Feng and X. Bu, Homochiral Crystallization of Microporous Framework Materials from Achiral Precursors by Chiral Catalysis, J. Am. Chem. Soc., 2008, 130, 12882–12883.
22. D. E. Williams, J. Davies, B. O. Patrick, H. Bottriell, T. Tarling, M. Roberge, R. J. Andersen and A. G. Cladoniamides, Tryptophan-Derived Alkaloids Produced in Culture by Streptomyces uncialis, Org. Lett., 2008, 10, 3501–3504.
23. M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, J. A. Montgomery Jr, T. Vreven, K. N. Kudin, J. C. Burant, J. M. Millam, S. S. Iyengar, J. Tomasi, V. Barone, B. Mennucci, M. Cossi, G. Scalmani, N. Rega, G. A. Petersson, H. Nakatsuji, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, M. Klene, X. Li, J. E. Knox, H. P. Hratchian, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, P. Y. Ayala, K. Morokuma, G. A. Voth, P. Salvador, J. J. Dannenberg, V. G. Zakrzewski, S. Dapprich, A. D. Daniels, M. C. Strain, O. Farkas, D. K. Malick, A. D. Rabuck, K. Raghavachari, J. B. Foresman, J. V. Ortiz, Q. Cui, A. G. Baboul, S. Clifford, J. Cioslowski, B. B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I. Komaromi, R. L. Martin, D. J. Fox, T. Keith, M. A. Al-Laham, C. Y. Peng, A. Nanayakkara, M. Challacombe, P. M. W. Gill, B. Johnson, W. Chen, M. W. Wong, C. Gonzalez, and J. A. Pople, Gaussian 03, revision C02, Gaussian, Inc, Wallingford, CT, 2004.
24. A. D. Becke, Density-functional thermochemistry. III. The role of exact exchange, J. Chem. Phys., 1993, 98, 5648–5652.
25. C. Lee, W. Yang and R. G. Parr, Development of the Colle–Salvetti correlation-energy formula into a functional of the electron density, Phys. Rev. B: Condens. Matter Mater. Phys., 1988, 37, 785–789.
26. S. Miertuš, E. Scrocco and J. Tomasi, Electrostatic interaction of a solute with a continuum – a direct utilization of ab initio molecular potentials for the prevision of solvent effects, Chem. Phys., 1981, 55, 117–129.

---

Footnote

† Electronic supplementary information (ESI) available. CCDC 760948, 703635, 703637 and 703638. For ESI and crystallographic data in CIF or other electronic format see DOI: 10.1039/c4ra15284k

---