Gas Antisolvent Approach for the Precipitation of-Methoxyphenylacetic Acid – ( R )-1-Cyclohexylethylamine Dia stereomeric Salt

A. Zodge,a M. Kőrösi,a M. Tárkányi,a J. Madarász,b I. Miklós Szilágyi,b,c T. Sohajda,d and E. Székelya,* aDepartment of Chemical and Environmental Process Engineering, Budapest University of Technology and Economics, Budapest, Hungary bDepartment of Inorganic and Analytical Chemistry, Budapest University of Technology and Economics, Budapest, Hungary cMTA-BME Technical Analytical Chemistry Research Group, Budapest, Hungary dCYCLOLAB Ltd., Budapest, Hungary


Introduction
The chemical industry market for enantiopure chemicals exhibits a great demand and it is increasing day by day.The preparation technique plays a key role in its economics.2][3][4] These techniques highly depend on reaction conditions such as slight changes in temperature, pressure, etc.Hence, each reaction parameter needs careful optimisation.
Supercritical carbon dioxide (scCO 2 ) might be used as a solvent in chiral resolutions 5,6 as well as an antisolvent in precipitation processes. 7Gas antisolvent (GAS) precipitation is fast and less solvent-intensive than traditional crystallizations.Using carbon dioxide as a solvent or antisolvent has some benefits, especially due to its non-flammable, non-explosive nature while being readily available and non-toxic in trace amounts.][14] α-Methoxyphenylacetic acid (MPAA) is widely used as a resolving agent and building block in various organic syntheses. 15Phenylacetic acids and their derivatives are very versatile platform molecules.Several derivatives show a broad range of biological activity, i.e. antibacterial, herbicide, plant growth regulator, etc.The syntheses of MPAA and its derivatives were briefly summarized by Edward et al. 16 Diastereomeric salt formation of racemic α-methoxyphenylacetic acid (MPAA) using (R)-(−)-1-cyclohexylethylamine (CHEA) as an enantiopure resolving agent is presented (Scheme 1) via GAS method for the first time in this study.

View cell experiments
The variable-volume view cell (New Ways of Analytics, Germany) used in this study is schematically depicted in Figure 1.It is equipped with a transparent zirconium oxide window and a stainless steel piston.The maximum volume of the cell is 70 mL.
Limits for pressure and temperature are 75 MPa and 150 °C, respectively.A known amount of solid sample was placed in the view cell, and in case of using a cosolvent, a known amount of cosolvent was also injected.Then the cell was sealed and pressurized with CO 2 by an ISCO 260D syringe pump (1).The injected mass of CO 2 was calculated from the volume change in the pump and with the CO 2 density obtained from NIST Chemistry Webbook 17 calculated with measured pressure and temperature conditions at (1).Pressure was increased by decreasing the cell volume at constant temperature until a homogeneous solution was observed.After minimum 20 minutes of equilibration, the pressure was slowly decreased by increasing the volume of the cell at constant temperature and molar composition.Cloud points were observed visually.

Gas antisolvent experiments
GAS precipitation experiments were carried out in a high-pressure autoclave shown in Figure 2.
The reactants (100 ± 0.5 mg (0.60 ± 0.003 mmol) of racemic MPAA and a calculated amount of (R)-CHEA (typically 38 ± 0.5 mg equivalent to 0.30 ± 0.03 mmol) were dissolved separately in 1.5 ± 0.006 mL of solvent.After complete dissolution, the solutions were charged into the tempered autoclave.The autoclave was sealed and filled with CO 2 to the

View cell experiments
The variable-volume view cell (New Ways of Analytics Germany) used in this study is schematically depicted in Figure 1.It is equipped with a transparent zirconium oxide window and a stainless steel piston.The maximum volume of the cell is 70 mL.
Limits for pressure and temperature are 75 MPa and 150 °C, respectively.A known amount of solid sample was placed in the view cell, and in case of using a cosolvent, a known amount of cosolvent was also injected.Then the cell was sealed and pressurized with CO 2 by an ISCO 260D syringe pump (1).The injected mass of CO 2 was calculated from the volume change in the pump and with the CO 2 density obtained from NIST Chemistry Webbook 17 calculated with measured pressure and temperature 181 conditions at (1).Pressure was increased by decreasing the cell volume at constant tem homogeneous solution was observed.After minimum 20 minutes of equilibration, t slowly decreased by increasing the volume of the cell at constant temperature and mo Cloud points were observed visually.desired pressure followed by stirring for 1 hour for complete precipitation.Before depressurization, the reactor was washed with three-fold volume (90 mL) of scCO 2 at the pressure and temperature of the reactor in order to extract the CO 2 -soluble components.This extract was trapped in methanol.The solid, crystalline sample recovered from the reactor after depressurization is referred to as the raffinate.The main component of the extract was unreacted MPAA whereas raffinate consists of diastereomeric salts.

Atmospheric reference experiments
The starting reactants were dissolved separately in a minimal amount of solvent.After complete dissolution at 40 °C, both stocks were mixed with excess solvent and kept under magnetic stirring for 1 hour at 50 °C in water bath followed by 1 hour of natural cooling.The solid crystals were filtered using a G-4 Glass filter and air-dried.

Capillary electrophoresis (CE)
The enatiomeric excess of the sample was determined on a Hewlett Packard 3D CE system (Hewlett Packard, Waldbronn, Germany) equipped with a diode array UV detector.Uncoated fused-silica capillaries of 58.5 cm effective and 50.0 cm total length (FSOT, Composite Metal Services Ltd., Worcestershire, UK) were applied throughout the study.Samples and capillaries were tempered at 25 °C; the analytes were detected at 200 nm.The capillary was flushed with 0.1 M NaOH, and purified water for 30 seconds each subsequently, and finally with the running buffer for 60 seconds before every analysis.Samples were injected using 50 mbar pressure for 5 seconds, the applied voltage was +20 kV.Peaks were evaluated using Chemstation (Hewlett Packard) software.
Britton-Robinson buffer (BRB, containing 50 mM boric, acetic and phosphoric acid) was used as a background electrolyte.The desired pH was set to 9.0 by adjusting with 0.2 M NaOH.The final running buffers were filtered through a 0.22 mm Millex-GV syringe filters (Millipore, Bedford, USA).Diastereomeric salt mixture samples were dissolved in methanol to obtain 1 mg mL -1 stock solutions (related to dry material) and diluted further with ethanol:water 50 % v/v 100-fold to achieve optimal peak areas.As a chiral selector, 10 mM of 6-monodeoxy-6-monoamino-β-cyclodextrin was appli ed in all experiments.

Powder X-ray diffraction (XRD)
XRD measurements were done on a PANalytical X'Pert Pro MPD diffractometer (PANalytical, Almelo, The Netherlands), equipped with an X'celerator detector in θ-θ arrangement to the beam source, at the Cu Kα wavelength (1.5408 Å) applying 40 kV tension and 30 mA current.Diffractograms were recorded in the 1°-42° range.

Scanning electron microscopy (SEM)
SEM images were recorded by a JEOL JSM 5500-LV scanning electron microscope using 20 kV voltage and a secondary electron detector.For SEM studies, samples were covered with a 5-10 nm Au layer to make them conductive.

Calculation methods
The appropriate molar ratio (mr) plays a key role in any resolution system.Molar ratio was calculated as mr = n res /n rac , where n denotes the molar quantity and the indices res and rac refer to the resolving agent and racemic compound, respectively.Enantiomeric excess values (Equation 1) were calculated from the peak areas obtained by capillary electrophoresis. (1) In the above equation, A R and A S refer to the areas of respective MPAA enantiomers of capillary electropherograms.
Yields for extracts and raffinates are the ratio of the recovered mass and theoretical mass of the certain fraction estimating full conversion and complete separation.The selectivity (S) of a given fraction was calculated by multiplying the enantiomeric excess and the yield.Indices extr and raff denote extract and raffinate, respectively.

Determination of solubility
For the development of a novel antisolvent based resolution, it is necessary to have preliminary information on the solubility of the components involved.Since there is no literature data available on the solubilities of MPAA, CHEA, or their diastereomeric salts in high pressure CO 2 , the necessary data was measured by cloud point determination.
The solubility of rac-MPAA in CO 2 increases with increasing pressure at a given temperature, e.g. at 37 °C and 8.7 MPa the solubility is 6.3 mg g -1 (mass fraction), while at 11.2 MPa 10 mg g -1 can be dissolved.The cloud point pressure of the same solution increases with temperature, e.g. the 6.3 mg g -1 mass fraction solution is homogeneous above 8.7, 11.2 and 12.7 MPa at 37.1, 44.8 and 51.0 °C, respectively.
While a fair solubility of the rac-MPAA in scCO 2 is required for an efficient extraction of the unreacted enantiomers, an organic solvent is also required to dissolve all components including the diastereomeric salts.Components rac-MPAA and (R)-CHEA are highly soluble in polar solvents and also show reasonable solubility in non-polar solvents (Table 1).The higher solubility in polar solvents is attributed to the presence of carboxylic and amine functional groups in their chemical structure.Those solvents were selected for the further experiments in which both MPAA and CHEA have high (>1 g mL -1 ) solubility.
As (R)-CHEA is a primary amine, it readily reacts with carbon dioxide and forms a carbamate.Its carbamate is white, solid and crystalline.The carbamate formed by the carbon dioxide content of air and that formed in scCO 2 are similar. 17If stirring is not applied, the formed solid saves the remaining liquid (R)-CHEA from being consumed by CO 2 .Neither (R)-CHEA nor its carbamate has any reasonable solubility in pure carbon dioxide.

Selection of solvents for GAS
The solvents, which have good dissolving power for all components involved (Table 1), were also tested for their applicability in gas antisolvent precipitation.These preliminary results are summarized in Table 2.
Although alcohols seemed to be promising in solvent screening (Table 1 and 2), at resolution experiments, ester by-product formation was observed.Experiments with a mixture of ether-acetonitrile (1:1) also showed good results but were excluded due to safety concerns.Mixture of equal amounts of toluene-acetonitrile showed the best results for solubility of rac-MPAA, (R)-CHEA and precipitation of its diastereomeric salt during screening, thus this solvent mixture was selected for optimisation of process conditions.

Powder XRD
The resolving agent reacts with both enantiomers of MPAA and forms the corresponding (R,R)and (S,R)-salts.Since (S)-MPAA is commercially available, while (R)-MPAA is not, (S,S)-diastereomeric salt patterns were used as references of the (R,R)-salt, while (S,R)-salt as the (R,S)-salt's reference, according to the Marckwald principle. 18The XRD patterns of the formed solid product was compared against the reference diastereomeric salts (Figure 3).XRD patterns of the (S,S)-salt and (S,R)salt are different which proves that the two salts are crystallized in different crystalline structures.Comparing the XRD patterns of the same salts prepared with GAS or atmospheric crystallization methods, one may conclude that the peak positions and ratios are the same, suggesting similar crystal structures, but the GAS method results in an increased crystallinity.
Scanning Electron Microscopy (SEM) confirmed the formation of fibrous, elongated needle-like crystals under GAS conditions.Higher crystallinity was observed under scCO 2 conditions than under atmospheric conditions via 4-6 hours crystallization time.Crystallization at atmospheric conditions using Tol-AcN mixed solvent shows different crystal habits (Figure 4); however, the crystal structures were similar for salts prepared with atmospheric and gas antisolvent technique.

Optimization of the gas antisolvent precipitation
After successful screening experiments, detailed optimization of the process was performed regarding the molar ratio of the resolving agent to the racemic acid, pressure and temperature of GAS precipitation.
Regarding the molar ratio of reactants, the best selectivity was achieved using a 0.5 ratio of racemate: resolving agent (Table 3).The half equivalent method or commonly known as the modified Pope-Peachy method is the best suitable for the resolution, as typical in the supercritical resolution techniques. 19The equivalent method, i.e.Pasteur's method, gives a diastereomeric salt with worse optical purity and selectivity.However, it is worthwhile mentioning that the resolution produces solid, crystalline diastereomeric salt with 1:1 molar ratio (mr) as well.Furthermore, since the yield is the mass of recovered crystalline salt versus the theoretical mass of the salt, from the same amount of Scanning Electron Microscopy (SEM) confirmed the formation of fibrous, elongated needle-like crystals under GAS conditions.Higher crystallinity was observed under scCO 2 conditions than under atmospheric conditions via 4-6 hours crystallization time.Crystallization at atmospheric conditions using Tol-ACN mixed solvent shows different crystal habits (Figure 4); however, the crystal structures were similar for salts prepared with atmospheric and gas antisolvent technique.Regarding the molar ratio of reactants, the best selectivity was achi resolving agent (Table 3).The half equivalent method or commonly k  racemic acid, approx.two times as much salt was prepared at mr = 1 than at mr = 0.5.Total recoveries of material at each molar ratio were above 80 %.
Temperature effects at 12 MPa and 0.5 molar ratio are shown in Figure 5.There is a clear optimum of the selectivity at approx.40 °C, which is mainly due to the variation of the ee values.
The effect of pressure on the resolution system was studied at 40 °C, CO 2 pressure ranged between 9 and 21 MPa (Figure 6 and 7).Below 10 MPa pressure, a wet raffinate was recovered with poor enantiomeric excess and slightly lower yield.The raffinates with the highest ee-s (50-55 %) were prepared at 12 MPa.Above 16 MPa, further increase in pressure causes significant decrease in the yield.The dependence of selectivity on pressure is plotted in Figure 7.The increase in CO 2 pressure increases the density of CO 2 .However, as the concentration of the organic solvent was kept constant, the increase in pressure also means an increase in the CO 2 :organic solvent ratio, as at constant temperature by increasing pressure the density of CO 2 increases.Solubilities of the compounds involved increase by increasing pressure but decrease with increasing CO 2 :solvent ratio.The observation that there is an optimum density (dissolving power) is in accordance with our previous results.Please note that in the referred previous work the diastereomeric salts of cis-permetric acid and phenylethyl amine were crystallized from CO 2 only, using carbon dioxide as the reaction medium. 20

Conclusions
Resolution of α-methoxyphenylacetic acid with (R)-cyclohexylethyl amine is possible using gas antisolvent precipitation with carbon dioxide.Half equivalent amount of resolving agent is optimal for maximizing selectivity while keeping the amount of resolving agent as low as possible.Both the temperature and pressure influence the resolution significantly, and the optimal setting was found to be 40 °C and 12 MPa.The obtained diastereomeric salts show similar diastereomeric excess values as optimized atmospheric resolutions, but show higher crystallinity and yields while requiring significantly lower processing time.The results obtained by GAS anti solvent method show a good perspective for development of a semi-continuous process.The effect of pressure on the resolution system was studied at 40 °C, CO 2 pressure range between 9 and 21 MPa (Figure 6 and 7).Below 10 MPa pressure, a wet raffinate was recovered with poor enantiomeric excess and slightly lower yield.The raffinates with the highest ee-s (50-55 %) were prepared at 12 MPa.Above 16 MPa, further increase in pressure causes significant decrease in yield.The dependence of selectivity on pressure is plotted in Figure 7.The increase in CO 2 pressure increases the density of CO 2 .
However, as the concentration of the organic solvent was kept constant, the increase in pressure also means an increase in the CO 2 :organic solvent ratio, as at constant temperature by increasing pressure the density of CO 2 increases.Solubilities of the compounds involved increase by increasing pressure but decrease with increasing CO2:solvent ratio.The observation that there is an optimum density (dissolving power) is in accordance with our previous results.Please note that in the referred previous work the 192 diastereomeric salts of cis-permetric acid and phenylethyl amine were crystallized from CO 2 only, using carbon dioxide as the reaction medium. 20

F i g . 5 -Fig. 5
Fig. 5 Effect of temperature on diastereomeric salt formation.Reaction conditions: 0.5 molar ratio at 12MPa for 1 h.

Fig. 6
Fig. 6 Effect of pressure on diastereomeric salt formation.Reaction conditions: 0.5 molar ratio at 40 °C for 1 h.Empty symbols indicate wet raffinates.The lines are to guide the eye.

Fig. 6
Fig. 6 Effect of pressure on diastereomeric salt formation.Reaction conditions: 0.5 molar ratio at 40 °C for 1 h.Empty symbols indicate wet raffinates.The lines are to guide the eye.

F i g . 6 -
Effect of pressure on diastereomeric salt formation.Reaction conditions: 0.5 molar ratio at 40 °C for 1 h.Empty symbols indicate wet raffinates.The lines are to guide the eye.

F i g . 7 -A
Effect of pressure on selectivity of diastereomeric salt formation.Reaction conditions: 0.5 molar ratio at 40 °C for 1 h.Empty symbols indicate wet raffinates.The line is to guide the eye.192 of cis-permetric acid and phenylethyl amine were crystallized from CO 2 only, using e reaction medium. 20ure on diastereomeric salt formation.Reaction conditions: 0.5 molar ratio at 40 °C ols indicate wet raffinates.The lines are to guide the eye.-area,ee -enantiomeric excess,mr -molar ratio,n -molar quantity, mol S -selectivity, -Y -yield, -S u b s c r i p t s rac -racemic material res -resolving agent ext -extract raff -raffinate A b b r e v i a t i o n s scCO 2 -supercritical carbon dioxide GAS -gas antisolvent SEM -scanning electron microscopy XRD -X-ray powder diffraction CE -capillary electrophoresis