Characterization of a conglomerate-forming derivative of (±)-milnacipran and its enantiomeric resolution by preferential crystallization

Abstract

Among some derivatives of milnacipran, a last-generation antidepressive drug with useful application in the treatment of fibromyalgia, N-isobutyramide (±)-4 was identified as a conglomerate-forming compound and characterized using a binary melting diagram and X-ray analysis. Exploiting the conglomerate nature of (±)-4, its enantiomeric resolution by preferential crystallization was shown to be feasible and allowed us to obtain both enantiomers in satisfactory yield and excellent optical purity over a few cycles of entrainment.

---

1. Introduction

The pharmacological activity of a drug mainly depends on its ability to interact with biological systems that are able to selectively recognize and bind molecules with a suitable three dimensional arrangement. In the case of chiral drugs, different interactions of each enantiomer with the target could occur, resulting in therapeutic or antagonist and/or adverse effects, depending on the stereochemistry.¹ When the activity resides in a single enantiomer, the use of the racemate is only allowed when (a) in vivo racemization occurs or (b) the enantiomers display similar toxicological or (c) positive synergic effects.^2,3 Most new chiral synthetic drugs are currently marketed as single enantiomers and the production of enantiopure active pharmaceutical ingredients (API) is therefore a primary focus of industry.^4–6 Among the applied chiral technologies, stoichiometric or catalytic asymmetric synthesis and diastereoselective methods using chiral auxiliaries have been intensively developed from academic and industrial laboratories as effective tools for obtaining enantiomerically pure compounds on a large scale, but the use of expensive chiral auxiliaries and catalysts, together with the need for additional steps for their removal, often affects the global costs.^7–10 Over the past decade, biocatalytic methods based on purified enzymes or whole cells have been significantly exploited for their high efficiency in terms of regio-, stereo-, and chemo-selectivity and contribution to reduced time, costs and synthetic steps.^11–13

Besides all these methods, the resolution of a racemate by crystallization is still the most used approach in the pharmaceutical industry for large scale preparation of enantiopure compounds.¹⁴ Crystallization can be classified into two main types, the fractional crystallization of diastereoisomeric salts, which implies the use of a chiral enantiopure resolving agent in equimolecular ratio, and the direct crystallization of one enantiomer from a racemic mixture by preferential crystallization of a conglomerate, which is actually the most economical route to achieve pure enantiomers but is limited in its applicability due to the rather rare occurrence of conglomerates.^15–17 Indeed, about 5–10% of racemic mixtures crystallize as homochiral crystals and most of them are salts.

Milnacipran, Z-(±)-2-(aminomethyl)-N,N-diethyl-1-phenylcyclopropane, (±)-1 is an active drug for the treatment of major depressive disorders and has recently attracted interest due to its painkiller effects in the treatment of fibromyalgia.^18,19 Milnacipran is marketed in racemic form as IXEL® or SAVELLA®, but recent pharmacokinetic studies on single enantiomers showed significantly higher activity for (1S,2R)-levomilnacipran (F2695), (−)-1 as a serotonin–norepinephrine reuptake inhibitor than the racemic mixture, with less risk of cardiovascular disorders and toxicity.^20,21 For this reason, in 2014 the FDA agency approved the use of the enantiopure drug and it was launched on the market under the registered mark FETZIMA® by Forest in a partnership with Pierre Fabre Laboratories.

The preparation of enantiomerically pure levomilnacipran (−)-1 by different procedures, including synthesis starting from chiral substrates, asymmetric catalysis and resolution of the racemic precursors, has been reported.^22–26 The enantiomers of (±)-cis-milnacipran (±)-1 were separated by high performance capillary electrophoresis or HPLC on cellulose-based stationary phases and, on a preparative scale, by diastereoisomeric multiple crystallization using commercially available L-(+)-mandelic or L-(−)-tartaric acid derivatives as resolving agents.^27–30 Biocatalytic procedures include the enantioselective cyclopropanation of N,N-diethyl-2-phenylacrylamide promoted by a cytochrome P450 mutant and the kinetically controlled resolution of the (±)-1 through lipase-catalyzed enantioselective N-acylation in an organic solvent.^31,32

Here, we describe a simple alternative procedure for the preparation of enantiopure levomilnacipran based on the resolution of its N-iso-butyramide derivative (±)-4 by preferential crystallization. This amide was characterized as a conglomerate-forming compound by its solubility and melting point properties as well as by X-ray crystallographic analysis, and an effective protocol for its chiral separation by entrainment was proposed.

2. Results and discussion

In our recent work, we developed a biocatalytic process for the kinetic resolution of (±)-1 based on the enantioselective formation of the corresponding amide catalyzed by lipase B from Candida antarctica (Novozyme 435) in methyl tert-butyl ether (MTBE) in the presence of a suitable acyl donor (Scheme 1).³² Due to the high reactivity of the aminomethyl group of (±)-1 toward carboxylic esters as well as its distance from the stereogenic center(s), the reaction proceeded with moderate enantioselectivity and the enantiomeric purity of the amide products was increased by crystallization.

Scheme 1 Lipase-catalyzed kinetic resolution of racemic milnacipran.

All the scalemic amides 2–4 obtained from the enzymatic resolution gave enantiopure crystals from their saturated solutions in MTBE, ethyl acetate or diisopropyl ether at room temperature. However, during the crystallization of 4, we detected an inversion in the sign of the enantiomeric excess of the mother liquor with respect to the starting solution. This anomalous behavior suggested a possible conglomerate nature for crystals of (±)-4 and led us to better investigate its physical properties in view of a possible resolution of this amide by preferential crystallization.

In the crystallization of a chiral compound, three types of crystals can originate: (1) racemic crystals, when the heterochiral (R–S) interactions are stronger than the homochiral (R–R/S–S) ones. This is the most common case (90–95% occurrence), characterized by an ordered and equimolecular mixture of both enantiomers in the crystal lattice; (2) conglomerate crystals (5–10% of cases), in which a single enantiomer is present in the unit cell of the lattice, in consequence of the preferred homochiral interactions, and the whole solid can be viewed as a mechanical mixture of R–R and S–S crystals; (3) a racemic solid solution (or pseudo-racemate), in which both enantiomers are present in the lattice, but are randomly arranged and are not in an equimolecular ratio (very low occurrence), (Fig. 1).³³

Fig. 1 Representation of possible crystalline solids formed by chiral compounds.

The crystalline nature of a chiral compound has important implications on the possibility to perform a resolution by crystallization. Indeed, with compounds that form racemic crystals, an enantiopure product can be obtained only in the presence of a suitable enantiomeric excess in the starting material or by resorting to the formation of diastereomeric derivatives, while racemic mixtures giving solid solutions require several crystallization steps to be resolved. On the contrary, racemic mixtures that crystallize as conglomerates offer a valuable opportunity for complete and inexpensive enantiomeric separation on both laboratory and industrial scales, through the preferential crystallization of the desired enantiomer without the need of any resolving agent. In this context, the search for suitable counterions or chemical derivatization, resulting in the change from a racemic compound-forming system into a conglomerate-type, is of practical significance.^34,35

2.1. Investigations on the conglomerate nature of milnacipran isobutyramide

As a preliminary screening, (±)-4 was crystallized from MTBE at 20 °C and different samples of prismatic crystals with variable dimensions were collected and analyzed by chiral HPLC. Most of the chromatograms showed different ratios of (1S,2R)-(+)-4 to (1R,2S)-(−)-4 peaks, indicating that (±)-4 did not crystallize as a racemic compound. Further support for this hypothesis came from a collection of some enantiopure crystals grown from a dilute solution of (±)-4 at 27 °C for three days.

For the sake of comparison, when the same experiments were performed on (±)-2 and (±)-3, we always observed a racemic composition of the crystals (Fig. 2).

Fig. 2 Amide derivatives of milnacipran.

Rod-shaped crystals from separate crystallizations of (+)-4 and (−)-4 were compared with respect to their morphology by optical microscopy, but no appreciable differences were detected between the two enantiomers. X-ray analysis was then performed on single crystals taken from both the enantiopure samples and the conglomerate precipitate. In the latter case, a monocrystal of large dimensions (2.0 × 1.2 × 0.9 mm) was chosen as a sample suitable for data collection. Both diffractometric investigations provided the same homochiral composition of the single crystals, confirming that (±)-4 crystallizes as a conglomerate rather than a racemic compound. The monocrystal chosen for X-ray data collection of the conglomerate was then submitted to HPLC analysis on a chiral column to assign its absolute configuration, which was (1S,2R). Its X-ray structure is reported in Fig. 3.

Fig. 3 ORTEP diagram of the molecular structure of (1S,2R)-(+)-4 (thermal ellipsoids set at the 30% probability level).

Binary melting point diagrams are a useful tool to discriminate racemic compounds from conglomerates on the basis of their typical and different shapes. When the enantiomers form a racemic compound, the diagram displays the maximum value of the melting point for the racemic composition and two symmetrical minima for a eutectic composition, whereas in a conglomerate, a single minimum, coinciding with the 1 : 1 eutectic composition, is observed.³⁶

A binary diagram of 4 was built by determining the melting points of pure enantiomers, a racemic conglomerate and some scalemic mixtures. A plot of the data gave a V-shaped graph with a eutectic composition at (+)-4/(−)-4 = 1 : 1 and T_eut = 112 °C, while pure enantiomers melted at T = 137 °C in agreement with the typical values of ΔT_enant-rac observed for conglomerate-forming crystals (Fig. 4(a)). Melting enthalpies ΔH_enant = 56.2 kJ mol⁻¹ and ΔH_rac = 53.1 kJ mol⁻¹ for pure (+)-4 and (±)-4, respectively, were then determined by differential scanning calorimetry (DSC) and the observed difference is in line with values reported for a conglomerate-forming system (Fig. 4(b)).

Fig. 4 (a) Binary melting diagram of mixtures of 4 with different enantiomeric compositions. (b) DSC thermograms of (±)-4 and (+)-4.

In order to determine the solubility of (±)-4 and (+)-4 at different temperatures, an amount of solid sufficient to achieve saturation was suspended in MTBE, chosen as the most suitable solvent for crystallization, and the suspension was stirred for 24 h. Quantitative measurements were then carried out by chiral HPLC on the filtered solutions and the data are shown in Table 1.

Table 1 Temperature-dependent solubilities of 4^a

Temp. (°C)	(±)-4^b (mg mL^−1)	(±)-4^b (mole fraction × 10^3)	(+)-4^b (mg mL^−1)	(+)-4^b (mole fraction × 10^3)	α^c
a Experimental conditions: 30 mg mL^−1 of (+)-4 or 60 mg mL^−1 of (±)-4, MTBE, 600 rpm stirring, 24 h.b Determined by chiral HPLC.c Meyerhoffer's coefficient.
10	5.2	1.95	2.2	0.828	2.35
20	9.0	3.3	4.7	1.77	1.86
30	12.7	4.76	6.7	2.52	1.88

---

The solubility of (±)-4 was about twice that of the pure enantiomer at all temperatures, in good agreement with the so-called “double solubility rule”, S_rac = 2S_en (where S_rac and S_en are the solubilities of the racemic mixture and the pure enantiomer, respectively), valid for a conglomerate due to the lack of interactions between the enantiomeric species. Meyerhoffer's coefficient, α, given by the solubility ratio of the racemic mixture to the pure enantiomer (expressed in mole fractions), is an useful parameter for setting a successful enantiomeric separation by crystallization and a value of α ≤ 2 is usually considered optimal.³⁷

2.2. Preferential crystallization of milnacipran isobutyramide

Different procedures, in batch or continuous mode, have been developed and applied to optimize the performance in the resolution of a racemic compound by entrainment and a survey of literature revealed that, in most cases, the crystallization is triggered by an enantiopure crystal added to a racemic or scalemic mixture.^38–40 Provided that the solid is filtered prior to the nucleation of the undesired enantiomer, crystals of the enantiomer that is in excess can be obtained in high optical purity, while the solution displays an enantiomeric excess opposite in sign to that of the starting material. After the conditions of supersaturation required to activate a new crystallization are restored, the entire initial amount of racemic mixture can be, in theory, quantitatively resolved through successive steps of recycling of the mother liquors, whose enantiomeric excess switches the sign at each cycle. The process can be carried out under isothermal conditions when α ≤ 2, or by applying temperature gradients if α > 2, and in some instances, seeding with enantiopure crystals has proved unnecessary.

In the search for the most suitable conditions for the enantiomeric resolution of (±)-4 by entrainment we screened some highly supersaturated solutions with different enantiomeric excesses and each one of them was seeded with enantiopure crystals (1% w/w) of (+)-4 (Table 2). The experiments were carried out in MTBE by dissolving the solid under reflux until a clear solution was obtained, then leaving to cool spontaneously to 25 °C and seeding.

Table 2 Determination of initial enantiomeric excess optimal for preferential crystallization of 4^a

Entry	Starting material^b (mg, % ee)	Seed (mg)	Time (h)	Collected crystals^b (mg, % ee)	Mother liquor^b (% ee)	% EY^c	% RY^d
a Experimental conditions: (±)-4 or (+)-4 with the indicated positive enantiomeric excess (45 mg mL^−1), MTBE, 25 °C, no stirring, seeding with (+)-4 enantiopure crystal (1% w/w).b Enantiomeric excess determined by chiral HPLC.c Enantiomeric yield, % EY = 100 × (Wcrystals × eecrystals)/[Winit × (1 + eeinit)/2 + Wseed].d Resolution yield, % RY = 100 × [(Wcrystals × eecrystals) − (Winit × eeinit) − Wseed]/[Winit × (1 − eeinit)/2].e Without seeding.
1	500, 0	5	1	38, +80	–6	11.9	10.2
2	500, +2	5	1	35, +82	–4	11.0	2.3
3	460, +6	4.6	2	62, +84	–6	21.0	4.6
4	470, +12	4.7	1.5	80, +99	−12	29.6	8.7
5^e	450, +12	—	2.5	72, +99	−6	28.2	8.7

---

Crystallizations were performed over a period of 1–3 h and the efficiency of the process was compared in terms of enantiomeric yield, EY, expressed by the ratio of the pure enantiomer recovered in crystals to the available mass of this enantiomer in the starting solution. The amount of pure enantiomer extracted from the racemic mixture by entrainment was given by the resolution yield, RY, in which the mass of the enantiomer responsible for the initial enantiomeric excess and for the seed are taken into the account. When RY = 0, it means that only the excess of the enantiomer is recovered in the crystals and no resolution takes place.

The enantiomeric yield as well as the optical purity of the obtained crystals progressively increased with the enantiomeric excess in the starting mixture and, in all the cases, the enantiomeric imbalance in the mother liquor of crystallization was comparable or higher, but opposite in sign, with respect to that in the initial solution. Starting from mixtures with sufficiently high % ee values (8–12%), seeding was shown to be unnecessary and preferential crystallization proceeded without loss of optical purity in the recovered crystals (Table 2, entries 4 and 5).

In order to check the feasibility of the resolution of (±)-4 by means of successive cycles of entrainment, a highly supersaturated solution of (±)-4 was seeded with 1% of enantiopure (+)-4, and after 2 h, crystals of (+)-4 with 80% ee settled, leaving a solution enriched in the opposite enantiomer, (−)-4. To this solution, an amount of (±)-4 equal to the mass of the collected crystals was added in order to restore the initial supersaturation, and the procedure was repeated again to obtain crystals of (−)-4. After four cycles of crystallization (two for each enantiomer), 96 mg of (+)-4 with 84% ee and 117 mg of (−)-4 with 85% ee were recovered (Table 3). The collected crystals were then recrystallized from MTBE to give enantiopure (+)-4 and (−)-4 in 22% and 26% EY, respectively.

Table 3 Resolution of (±)-4 by entrainment^a

Cycle	Added mass of (±)-4 (mg)	Crystals (+)-4		Crystals (−)-4
		(mg)	(% ee)^b	(mg)	(% ee)^b
a Conditions: (±)-4 (50 mg mL^−1), seed crystals 5 mg of (+)-4 or (−)-4; MTBE; 25 °C, no stirring; 2–4 h.b Determined by chiral HPLC.
1	500	38	80
2	38			57	84
3	62	58	86
4	58			60	86
Total	658	96		117

---

Although the described procedure can be advantageously carried out over several cycles, the data in Table 2, entry 5, prompted us to develop an alternative protocol based on the preferential crystallization of a suitable optically enriched mixture without any addition of (±)-4 or enantiopure seed in the successive cycles, by only exploiting the enantiomeric inversion produced in the mother liquor in each step and restoring the initial supersaturation of the solution through the appropriate reduction of the solvent.

The procedure could be conveniently applied for cases in which the resolution process is mainly targeted to one enantiomer for its added value, for example as a pharmaceutical active principle, since the only enantiomeric imbalance required can be supplied by the unwanted enantiomer.

In the case in hand, we were interested in (+)-4 as direct precursor of levomilnacipran and the starting racemic mixture was then enriched with (−)-4. In a fine optimization, working parameters T = 27 °C, initial concentration 40.0 mg mL⁻¹ and an initial 12% ee allowed us to obtain the best compromise between the crystallization rate, crystal size and enantiomeric yield.

Our protocol proved to be very effective, leading to the isolation of (+)-4 at 49% of the theoretical yield in just four sequential crystallizations (two for each enantiomer). On a laboratory scale, starting from 440 mg of (±)-4 enriched with 60 mg of (−)-4 (ee₀ = 12%), we were able to isolate 160 mg of (−)-4, which corresponds to 100 mg of enantiomer extracted from the initial racemic mixture, and 115 mg of (+)-4, both with 94% optical purity (Table 4).

Table 4 Sequential preferential crystallizations of scalemic 4^a

Cycle	Initial mixture			Recovered crystals			EY
	mg	(+)-4/(−)-4	% ee	mg	(+)-4/(−)-4	% ee
a Experimental conditions: (+)-4 or (−)-4 (40 mg mL^−1), MTBE, 27 °C, no stirring, 18 h.
1	500	44/56	−12	107	4/96	−92	35.1
2	388	55/45	+10	77	97.5/2.5	+95	34.3
3	306	44/56	−12	53	1/99	–98	30.6
4	245	54/46	+8	38	96/4	+92	26.4

---

It is noteworthy that the whole amount of (+)-4 comes merely from the racemic mixture and further sequential crystallization steps can be carried out with high efficiency up to complete resolution of the initial racemic mixture, provided that the enantiomeric excess in the mother liquor is maintained in the range 8–12% ee.

3. Conclusions

The conglomerate nature of the crystals of an N-isobutyramide derivative of milnacipran, (±)-4, was assessed by the study of its physical properties and crystallographic X-ray analysis. On this basis, the resolution of 4 by preferential crystallization of the racemic or a suitable scalemic mixture was shown to be feasible and both enantiomers were obtained in good chemical and enantiomeric yield in a few cycles of entrainment. Through management of the crystallization conditions, the process could be scaled-up, allowing convenient access to the active enantiomer of milnacipran, (−)-1, or its pharmaceutically acceptable salts after chemical hydrolysis of (+)-4.

In comparison with the reported methods for the preparation of (−)-1, which employ chiral reagents or a chiral acid for the formation of diastereoisomeric salts, only a small amount of enantiopure amide is required as a source of chirality and the process can be made more convenient by using the unwanted enantiomer (−)-4 to create the initial enantiomeric excess necessary to trigger the entrainment.

4. Experimental

4.1. Materials

Chiral HPLC analyses were carried out with a thermostatted (23 °C) Phenomenex Lux® Cellulose-4 column using an n-hexane/2-propanol 70 : 30 (v/v) mixture as the mobile phase at a flow rate of 0.5 mL min⁻¹ and UV-detection at λ 225 nm: t_R 12.1 min for (1S,2R)-(+)-4 and 14.2 min for the (1R,2S)-(−)-4. Samples of enantiopure (+)-4 and (−)-4 were obtained from the biocatalyzed resolution of (±)-4.³²

Melting points were measured using a Mel Temp II Laboratory Device, repeating the measures in triplicate.

4.2. Preparation of (±)-4

Commercial tablets (1 g) of the drug containing milnacipran hydrochloride were ground in a mortar and suspended in trimethylamine (4 mL). Iso-butyric anhydride in a twofold molar excess was added and the suspension was stirred at 40 °C overnight. After dilution with water, the mixture was extracted with AcOEt (3 × 20 mL) and the organic layers were collected, washed twice with brine, dried over Na₂SO₄ and taken to dryness to give (±)-4 as a white solid in 92% theoretical chemical yield. Crystallization from MTBE gave pure (>99%) amide (±)-4, whose ¹H-NMR spectrum was in agreement with literature data.³²

4.3. DSC analyses

Differential scanning calorimetry (DSC) profiles were obtained with a TA Instruments Q100 calibrated with melt purity indium standard (156.6 °C and ΔH = 28.45 J g⁻¹). For DSC analyses, the samples (4–6 mg) were heated at 10 °C min⁻¹ from 40 °C to 160 °C, under a nitrogen atmosphere (flow rate 50 mL min⁻¹). Analyses were performed in triplicate and the melting temperature corresponds to the extrapolated onset temperature.

4.4. Solubility measurements

A suspension of 30 mg of (+)-4 (or 60 mg of (±)-4) in 1 mL of MTBE was maintained at 10, 20 or 30 °C using an electronic Peltier thermostat and magnetically stirred for 24 h to reach equilibrium. After centrifugation, an aliquot of the clear supernatant was diluted with a known amount of 2-PrOH and analyzed by chiral HPLC for quantitative determination of the dissolved amide, by reference to a calibration curve previously built with pure standards. All the measurements were repeated three times and averaged values were considered.

4.5. Entrainment of 4 from solution with different enantiomeric compositions

In a screw-capped vial, (±)-4 and the appropriate amounts of (+)-4 required to obtain different initial enantiomeric excesses (in the 0–12% ee range) were mixed and MTBE was added to a concentration of 45 mg mL⁻¹.‡ After complete dissolution of the whole solid by heating to reflux, the solution was left to cool spontaneously to 25 °C and maintained at this temperature in a thermostat. The solution was seeded with enantiopure crystals of (+)-4 (1% w/w) and, after the appearance of a precipitate, the enantiomeric composition of the mother liquors was monitored by HPLC. At a suitable time, the crystals were filtered, washed with cold MTBE, then dissolved and the solution was analyzed by chiral HPLC.

4.6. Resolution of (±)-4 by preferential crystallization

Racemic (±)-4 (500 mg, 1.58 mmol) was dissolved in 10.0 mL of MTBE by heating to reflux in a screw-capped vial until a clear solution was obtained. The solution was left to cool to 25 °C and then seeded with (+)-4 (5 mg, 1% w/w). Crystallization was carried out isothermally in a thermostat without stirring. Crystals (38 mg) were then separated by filtration and 38 mg of (±)-4 were added to the mother liquor and the mixture was heated to reflux until the complete dissolution of any solid. This solution was then seeded with (−)-4 (5 mg, 1% w/w) and left to crystallize again. The procedure was repeated by separating the crystals at each cycle and restoring the initial supersaturation by addition of an amount of (±)-4 equal to the mass of the collected crystals.

After four crystallizations (two for each enantiomer) the collected crystals were dissolved in MTBE and recrystallized to give pure (+)-4 (75 mg, ee > 99%, 22% EY) and (−)-4 (90 mg, ee > 99%, 26% EY).

4.7. Resolution of (±)-4 by sequential preferential crystallizations

A mixture of (±)-4 (440 mg, 1.39 mmol) and (−)-4 (60 mg, 0.19 mmol) was dissolved in 12.5 mL of MTBE by heating to reflux in a screw-capped vial until a clear solution (12% ee) was obtained. The solution was maintained at 27 °C in a thermostat for 18–24 h without stirring and the crystalline solid (−)-4 (107 mg, 92% ee) was collected by filtration. For the next run, the mother liquor was concentrated to restore the initial supersaturation conditions (40 mg mL⁻¹) and then heated to reflux until the complete dissolution of any solid. Once the crystallization temperature was reached, the solution was left without stirring for 18–24 h and crystals of (+)-4 (77 mg, 95% ee) were separated by filtration. This cycle was repeated again and after four alternate crystallizations (two for each enantiomer) 160 mg of (−)-4 (94% ee, 32% yield, 54% EY) and 115 mg of (+)-4 (94% ee, 23% yield, 49% EY) were obtained.

4.8. Crystallographic data for (+)-4

The intensity data were collected on a Bruker Smart Apex CCD area detector using graphite-monochromated Mo Kα radiation (λ = 0.71073 Å). Data reduction was made using SAINT programs; absorption corrections based on multiscan were obtained by SADABS.⁴¹ The structures were solved by SHELXS-97 and refined on F² by full-matrix least-squares using SHELXL-97.⁴² All the non-hydrogen atoms were refined anisotropically, and hydrogen atoms were included as ‘riding’ and not refined. The isotropic thermal parameters of H atoms were fixed at 1.2 (1.5 for methyl groups) times the equivalent thermal parameter of the atoms to which they are bonded. Crystal data and results of the refinement: colorless prism 0.45 × 0.35 × 0.30 mm, M_r = 316.43, orthorhombic, space group P2₁2₁2₁, a = 10.3704(6) Å, b = 11.6346(6) Å, c = 15.6401(9) Å, V = 1887.06(18) ų, Z = 4, T = 296(2) K, μ = 0.072 mm⁻¹, 31 195 measured reflections, 2687 independent reflections, 2469 reflections with I > 2σ(I), 4.36 < 2θ < 56.86°, R_int = 0.0230. Refinement on 2687 reflections, 215 parameters. Final R = 0.0345, wR = 0.0966 for data with F² > 2σ(F²), S = 1.040, (Δ/σ)_max = 0.001, Δρ_max = 0.157, Δρ_min = −0.137 e Å⁻³. CCDC 1469407 contains the supplementary crystallographic data for this paper.†

Acknowledgements

Thanks are due to Dr Salvatore Battiato, Institute for Polymers, Composites and Biomaterials of CNR, for differential scanning calorimetry analyses.

Notes and references

1. J. Gal, Helv. Chim. Acta, 2013, 96, 1617–1657.
2. K. Singh, P. Shakya, A. Kumar, S. Alok, M. Kamal and S. P. Singh, IJPSR, 2014, 11, 4644–4659.
3. L. Nguyen, H. He and C. Pham-Huy, Int. J. Biomed. Sci., 2006, 2, 85–100.
4. M. C. Nunez, M. E. Garcia-Rubino, A. Conejo-Garcia, O. Cruz-Lopez, M. Kimatrai, M. A. Gallo, A. Espinosa and J. M. Campos, Curr. Med. Chem., 2009, 16, 2064–2074.
5. M. Breuer, K. Ditrich, T. Habicher, B. Hauer, M. Keßeler, R. Stürmer and T. Zelinski, Angew. Chem., Int. Ed. Engl., 2004, 43, 788–824.
6. F. Faigl, E. Fogassy, M. Nógrádi, E. Pálovics and J. Schindler, Tetrahedron: Asymmetry, 2008, 19, 519–536.
7. H. Lorenz and A. Seidel-Morgenstern, Angew. Chem., Int. Ed. Engl., 2014, 53, 1218–1250.
8. J. Alemán and S. Cabrera, Chem. Soc. Rev., 2013, 42, 774–793.
9. G.-Q. Lin and X.-W. Sun, in Chiral Drugs: Chemistry and Biological Action, ed. G.-Q. Lin, Q.-D. You and J.-F. Cheng, John Wiley & Sons, 2011, pp. 29–76.
10. V. Farina, J. T. Reeves, C. H. Senanayake and J. J. Song, Chem. Rev., 2006, 106, 2734–2793.
11. A. C. L. M. Carvalho, T. S. Fonseca, M. C. Mattos, M. C. F. Oliveira, T. L. G. Lemos, F. Molinari, D. Romano and I. Serra, Int. J. Mol. Sci., 2015, 16, 29682–29716.
12. D. Muñoz Solano, P. Hoyos, M. J. Hernáiz, A. R. Alcántara and J. M. Sánchez-Montero, Bioresour. Technol., 2012, 115, 196–207.
13. A. Ghanem and H. Y. Aboul-Enein, Chirality, 2005, 17, 1–15.
14. B. Y. Shekunova and P. Yorka, J. Cryst. Growth, 2000, 211, 122–136.
15. T. Vetter, C. L. Burcham and M. F. Doherty, AIChE J., 2015, 61, 2810–2823.
16. Y. Wang and A. Chen, in Stereoselective Synthesis of Drugs and Natural Products, ed. V. Andrushko and N. Andrushko, John Wiley & Sons, 2013, 56, pp. 1663–1682.
17. G. Coquerel, in Novel Optical Resolution Technologies, K. Sakai, N. Hirayama and R. Tamura, Springer, 2007, 269, pp. 1–51.
18. C. M. Spencer and M. I. Wilde, Drugs, 1998, 56, 405–427.
19. J. A. Kyle, B. D. A. Dugan and K. K. Testerman, Ann. Pharmacother., 2010, 44, 1422–1429.
20. M. M. Saraceni, J. V. Venci and M. A. Gandhi, J. Pharm. Pract., 2014, 27, 389–395.
21. G. M. Asnis and M. A. Henderson, Neuropsychiatr. Dis. Treat., 2015, 11, 125–135.
22. J. Alliot, E. Gravel, F. Pillon, D.-A. Buisson, M. Nicolas and E. Doris, Chem. Commun., 2012, 48, 8111–8113.
23. Y. Ishizuka, H. Fujimori, T. Noguchi, M. Kawasaki, M. Kishida, T. Nagai, N. Imai and M. Kirihara, Chem. Lett., 2013, 42, 1311–1313.
24. S. Shuto, S. Ono, Y. Hase, N. Kamiyama and A. Matsuda, Tetrahedron Lett., 1996, 37, 641–644.
25. M. P. Doyle and W. Hu, Adv. Synth. Catal., 2002, 343, 299–302.
26. P. Viazzo, V. Alphand and R. Furstoss, Tetrahedron Lett., 1996, 37, 4519–4522.
27. S. Grard, P. Morin, M. Dreux and J.-P. Ribet, Electrophoresis, 2000, 21, 3028–3034.
28. A. Patti, S. Pedotti and C. Sanfilippo, Chirality, 2008, 20, 63–68.
29. L. Ai and X. Liu, US Pat., 8, 222, 454 B2, 2012.
30. J. Deregneaucourt and R. Grosse, US Pat., RE43879, 2012.
31. Z. J. Wang, H. Renata, N. E. Peck, C. C. Farwell, P. S. Coelho and F. H. Arnold, Angew. Chem., Int. Ed. Engl., 2004, 53, 6810–6813.
32. C. Sanfilippo, G. Nicolosi and A. Patti, J. Mol. Catal. B: Enzym., 2014, 104, 82–86.
33. G. Coquerel, Enantiomer, 2000, 5, 481–498.
34. J. Vávra, L. Severa, I. Císařová, B. Klepetářová, D. Šaman, D. Koval, V. Kašička and F. Teplý, J. Org. Chem., 2013, 78, 1329–1342.
35. K. Nemák, M. Ács, D. Kozma and E. Fogassy, J. Therm. Anal., 1997, 48, 691–696.
36. H. G. Brittain, Pharm. Res., 1990, 7, 683–690.
37. A. Collet, M.-J. Bienne and J. Jacques, Chem. Rev., 1980, 80, 215–230.
38. M. P. Elsner, G. Ziomek and A. Seidel-Morgenstern, Chem. Eng. Sci., 2011, 66, 1269–1284.
39. M. P. Elsner, G. Ziomek and A. Seidel-Morgenstern, AIChE J., 2009, 55, 640–649.
40. C. Rougeot and J. E. Hein, Org. Process Res. Dev., 2015, 19, 1809–1819.
41. Bruker, SMART, SAINT and SADABS, Bruker AXS Inc., Madison, Wisconsin, USA, 1997.
42. G. M. Sheldrick, Acta Crystallogr., 2008, A64, 112–122.

---

Footnotes

† CCDC 1469407. For crystallographic data in CIF or other electronic format see DOI: 10.1039/c6ra07745e

‡ The amount W_{pure enant} (g) of enantiopure enantiomer required to obtain a desired enantiomeric excess ee_f starting from a mass W_rac (g) of racemic mixture is given by the formula: W_{pure enant} = (W_rac × ee_f)/(1 − ee_f).

---