Optical Resolution of Rimantadine

This work discloses a new procedure for the resolution of commercially available racemic rimantadine hydrochloride to enantiomerically pure (S)-rimantadine using (R)-phenoxypropionic acid as a recyclable resolving reagent. Good chemical yields, operational ease, and low-cost structure underscore the preparative value of this method for the production of enantiomerically pure rimantadine for medicinal or synthetic studies.

In the preliminary research [47,48], ligand (S)-3 was found to perform potentially better, as compared with previous results [49][50][51], in the chemical dynamic kinetic resolution and (S)/(R)-interconversion of unprotected AAs. Specifically, due to the extreme bulkiness and lipophilicity of the rimantadine residue, intermediate complexes 5 of particular (S C ,R N ,R C ) absolute configuration were easy to precipitate, allowing the realization of a second-order asymmetric transformation protocol for the preparation of tailor-made AAs 8. It is, of course, understood that application of (R)-configured rimantadine will give rise to Ni(II) complexes 5 of (R C ,S N ,S C ) absolute stereochemistry. Therefore, to continue our exploration of the synthetic potential of ligands of type 3, we needed reliable

Results and Discussion
Since its approval by the Food and Drug Administration (FDA) in 1994, rimantadine (α-methyl-1-adamantane-methalamine hydrochloride) (Scheme 2) is a marketed drug prescribed for the treatment of influenza virus A infection [52]. Despite some subtle differences in the binding properties of the (S)-and (R)-enantiomers 1 [53,54], in all its medical applications, rimantadine is used as a hydrochloric salt of the racemic form. Consequently, the research data for enantiomerically pure rimantadine and its commercial availability are quite limited. For example, the recent asymmetric synthesis of enantiomerically pure 1 [55] involves the reductive amination of the corresponding ketone using rather expensive Ellman's chiral sulfinimide [56][57][58] as a chiral auxiliary. Accordingly, with the aim to develop an economically sound process, and drawing inspiration from the recent publications of optical resolutions of various chiral amines [59][60][61][62][63] and, in particular, patent data [64], we focused our attention on the optical resolution approach, the science of which is well understood and can be performed on an industrial scale [65][66][67][68].
At the outset, we needed to find a high-yield procedure for the preparation of racemic free-base rimantadine from the commercially available hydrochloric salt 9 (Scheme 2). as a hydrochloric salt of the racemic form. Consequently, the research data for enantiomerically pure rimantadine and its commercial availability are quite limited. For example, the recent asymmetric synthesis of enantiomerically pure 1 [55] involves the reductive amination of the corresponding ketone using rather expensive Ellman's chiral sulfinimide [56][57][58] as a chiral auxiliary. Accordingly, with the aim to develop an economically sound process, and drawing inspiration from the recent publications of optical resolutions of various chiral amines [59][60][61][62][63] and, in particular, patent data [64], we focused our attention on the optical resolution approach, the science of which is well understood and can be performed on an industrial scale [65][66][67][68].
At the outset, we needed to find a high-yield procedure for the preparation of racemic free-base rimantadine from the commercially available hydrochloric salt 9 (Scheme 2). Scheme 2. Transformation of rimantadine HCl salt 9 to free amine 10.
This goal was achieved using aqueous 1 N NaOH as a base in a biphasic system with dichloromethane, allowing for simultaneous extraction of free amine 10 into the organic layer. Target free rimantadine 10 was obtained in >99% yield and isolated simply by evaporating the organic solvent, before it was used for the consecutive experiments without additional purification.
Our next task was to find a suitable standard compound for the convenient determination of an enantiomeric composition of rimantadine-derived diastereomeric salts. Following the literature [64], we prepared racemic amide 12 to have a reference for retention times of the corresponding (S)-and (R)-enantiomers. The target transformation was achieved via the treatment of free amine 10 with acyl chloride 11, as presented in Scheme 3. The corresponding HPLC tests confirmed that p-t-Bubenzoylated rimantadine 12 can perfectly serve the purpose of a reliable standard compound for routine HPLC analyses. With these results in hand, we proceeded with the main objective of this work, looking for a resolving reagent to achieve the optical resolution of racemic rimantadine 10. For this task, we selected chiral acids 13-26 presented in Figure 1. All of these compounds are commercially available in both enantiomeric forms and were successfully used for the separation of various chiral amines [65][66][67][68].
Scheme 2. Transformation of rimantadine HCl salt 9 to free amine 10.
This goal was achieved using aqueous 1 N NaOH as a base in a biphasic system with dichloromethane, allowing for simultaneous extraction of free amine 10 into the organic layer. Target free rimantadine 10 was obtained in >99% yield and isolated simply by evaporating the organic solvent, before it was used for the consecutive experiments without additional purification.
Our next task was to find a suitable standard compound for the convenient determination of an enantiomeric composition of rimantadine-derived diastereomeric salts. Following the literature [64], we prepared racemic amide 12 to have a reference for retention times of the corresponding (S)and (R)-enantiomers. The target transformation was achieved via the treatment of free amine 10 with acyl chloride 11, as presented in Scheme 3. The corresponding HPLC tests confirmed that p-t-Bu-benzoylated rimantadine 12 can perfectly serve the purpose of a reliable standard compound for routine HPLC analyses. as a hydrochloric salt of the racemic form. Consequently, the research data for enantiomerically pure rimantadine and its commercial availability are quite limited. For example, the recent asymmetric synthesis of enantiomerically pure 1 [55] involves the reductive amination of the corresponding ketone using rather expensive Ellman's chiral sulfinimide [56][57][58] as a chiral auxiliary. Accordingly, with the aim to develop an economically sound process, and drawing inspiration from the recent publications of optical resolutions of various chiral amines [59][60][61][62][63] and, in particular, patent data [64], we focused our attention on the optical resolution approach, the science of which is well understood and can be performed on an industrial scale [65][66][67][68].
At the outset, we needed to find a high-yield procedure for the preparation of racemic free-base rimantadine from the commercially available hydrochloric salt 9 (Scheme 2). Scheme 2. Transformation of rimantadine HCl salt 9 to free amine 10.
This goal was achieved using aqueous 1 N NaOH as a base in a biphasic system with dichloromethane, allowing for simultaneous extraction of free amine 10 into the organic layer. Target free rimantadine 10 was obtained in >99% yield and isolated simply by evaporating the organic solvent, before it was used for the consecutive experiments without additional purification.
Our next task was to find a suitable standard compound for the convenient determination of an enantiomeric composition of rimantadine-derived diastereomeric salts. Following the literature [64], we prepared racemic amide 12 to have a reference for retention times of the corresponding (S)-and (R)-enantiomers. The target transformation was achieved via the treatment of free amine 10 with acyl chloride 11, as presented in Scheme 3. The corresponding HPLC tests confirmed that p-t-Bubenzoylated rimantadine 12 can perfectly serve the purpose of a reliable standard compound for routine HPLC analyses. With these results in hand, we proceeded with the main objective of this work, looking for a resolving reagent to achieve the optical resolution of racemic rimantadine 10. For this task, we selected chiral acids 13-26 presented in Figure 1. All of these compounds are commercially available in both enantiomeric forms and were successfully used for the separation of various chiral amines [65][66][67][68]. With these results in hand, we proceeded with the main objective of this work, looking for a resolving reagent to achieve the optical resolution of racemic rimantadine 10. For this task, we selected chiral acids 13-26 presented in Figure 1. All of these compounds are commercially available in both enantiomeric forms and were successfully used for the separation of various chiral amines [65][66][67][68].
The screening of resolving reagents 13-26 was conducted under standard conditions to compare the results and, thus, to find some promising candidates for an in-depth study. To this aim, we conducted a series of experiments using acetone as a solvent and using chiral acids 13-26 in 1.0, 0.5, and 0.25 molar stoichiometric ratios [65][66][67][68] relative to racemic rimantadine 10. At the end of each experiment, the diastereomeric salt was treated with base, similar to the process presented in Scheme 2, to generate free rimantadine, following its conversion to the corresponding derivative 12 for the determination of the stereochemical outcome. The results are presented in Table 1. It should be noted that the absolute configuration of less soluble salts derived from the resolving reagent and rimantadine is always unpredictable [65][66][67][68]. Thus, in many cases (S)-configured acid gave preference for the (S)-enantiomer of rimantadine; however, in other cases, the relationships of the absolute configurations were opposite.
As one can see from Table 1, the results were generally very poor with the exception of (R)-2-phenoxy propionic acid 13, (S)-naproxen 14, (2S,3S)-dibenzoyltartaric acid 24, and (2R,3R)-dibenzoyltartaric The screening of resolving reagents 13-26 was conducted under standard conditions to compare the results and, thus, to find some promising candidates for an in-depth study. To this aim, we conducted a series of experiments using acetone as a solvent and using chiral acids 13-26 in 1.0, 0.5, and 0.25 molar stoichiometric ratios [65][66][67][68] relative to racemic rimantadine 10. At the end of each experiment, the diastereomeric salt was treated with base, similar to the process presented in Scheme 2, to generate free rimantadine, following its conversion to the corresponding derivative 12 for the determination of the stereochemical outcome. The results are presented in Table 1. It should be noted that the absolute configuration of less soluble salts derived from the resolving reagent and rimantadine is always unpredictable [65][66][67][68]. Thus, in many cases (S)-configured acid gave preference for the (S)-enantiomer of rimantadine; however, in other cases, the relationships of the absolute configurations were opposite.    The screening of resolving reagents 13-26 was conducted under standard conditions to compare the results and, thus, to find some promising candidates for an in-depth study. To this aim, we conducted a series of experiments using acetone as a solvent and using chiral acids 13-26 in 1.0, 0.5, and 0.25 molar stoichiometric ratios [65][66][67][68] relative to racemic rimantadine 10. At the end of each experiment, the diastereomeric salt was treated with base, similar to the process presented in Scheme 2, to generate free rimantadine, following its conversion to the corresponding derivative 12 for the determination of the stereochemical outcome. The results are presented in Table 1. It should be noted that the absolute configuration of less soluble salts derived from the resolving reagent and rimantadine is always unpredictable [65][66][67][68]. Thus, in many cases (S)-configured acid gave preference for the (S)-enantiomer of rimantadine; however, in other cases, the relationships of the absolute configurations were opposite. Firstly, we decided to explore (2R,3R)-dibenzoyl tartaric acid 25 as a resolving reagent due to its ready evaluability and reasonably low cost. Table 2 summarizes a series of experiments conducted using 0.25 equivalents of resolving compound 25 in different organic solvents.
Firstly, we decided to explore (2R,3R)-dibenzoyl tartaric acid 25 as a resolving reagent due to its ready evaluability and reasonably low cost. Table 2 summarizes a series of experiments conducted using 0.25 equivalents of resolving compound 25 in different organic solvents. The data obtained were evaluated from the standpoint of both isolated yield of the crystalline precipitate of salt 31 and enantiomeric purity of the amine component in it. In our opinion, the best result was obtained in the case of the use of aqueous acetone as a solvent (entry 1) allowing isolation of salt 31 with excellent yield (46.8%) and reasonable enantiomeric purity (60.4% ee). Further attempts to improve this stereochemical outcome, using various amounts of water, concentrations, and temperature, were unfortunately without success. Therefore, we considered an option of additional purification of salt 31 obtained from the base experiment presented in entry 1.
As shown in Table 3, some success was achieved by crystallizing salt 31 from the same solvent (entry 1) or aqueous ethyl acetate (entry 2) and THF (entry 3).
The data obtained were evaluated from the standpoint of both isolated yield of the crystalline precipitate of salt 31 and enantiomeric purity of the amine component in it. In our opinion, the best result was obtained in the case of the use of aqueous acetone as a solvent (entry 1) allowing isolation of salt 31 with excellent yield (46.8%) and reasonable enantiomeric purity (60.4% ee). Further attempts to improve this stereochemical outcome, using various amounts of water, concentrations, and temperature, were unfortunately without success. Therefore, we considered an option of additional purification of salt 31 obtained from the base experiment presented in entry 1.
As shown in Table 3, some success was achieved by crystallizing salt 31 from the same solvent (entry 1) or aqueous ethyl acetate (entry 2) and THF (entry 3). However, in all cases, the results neither offered the enantiomeric composition nor the yields that can be considered as satisfactory for a sound practical procedure. For example, the additional, second crystallization from either solvent system did not bring the products much closer to enantiomerically pure (>99% ee) form. The best value achieved after the second crystallization was 94.8% ee, which was obtained using THF as a solvent.
Therefore, we decided to focus our attention on phenoxypropionic acid 13 as a resolution reagent. The search for a better-performing solvent brought about the same outcome as in the case of dibenzoyl tartaric acid 25, pointing to aqueous (5%) acetone as the best solvent system. However, in sharp contrast to the latter, the option of additional purification via crystallization from a different solvent showed rather encouraging results (Table 4). However, in all cases, the results neither offered the enantiomeric composition nor the yields that can be considered as satisfactory for a sound practical procedure. For example, the additional, second crystallization from either solvent system did not bring the products much closer to enantiomerically pure (>99% ee) form. The best value achieved after the second crystallization was 94.8% ee, which was obtained using THF as a solvent.
Therefore, we decided to focus our attention on phenoxypropionic acid 13 as a resolution reagent. The search for a better-performing solvent brought about the same outcome as in the case of dibenzoyl tartaric acid 25, pointing to aqueous (5%) acetone as the best solvent system. However, in sharp contrast to the latter, the option of additional purification via crystallization from a different solvent showed rather encouraging results ( Table 4).
As shown in Table 4, we started with the mixing of racemic rimantadine 10 and (R)-phenoxypropionic acid 13 in aqueous (5%) acetone, resulting in the precipitation of diastereomeric salt 32, which was isolated with 34.7% yield and showed 88.4% (S)-enantiomeric excess of the amine residue in it. Thus, obtained product 32 was next crystallized from aqueous ethyl acetate allowing precipitation of 87.3% of the original material 32 with noticeably increased enantiomeric purity (98.7% ee) of the constituent residue of rimantadine. One more crystallization procedure was performed using the same solvent system, affording salt 32 with 92% yield and enantiomeric purity of the target amine exceeding 99% ee. The overall yield of this three-step procedure was 28.1%, which we considered as adequate to provide reliable access to enantiomerically pure rimantadine.
Finally, we needed to develop a protocol for the isolation of free rimantadine (S)-1 from diastereomerically pure salt 32. As one can see from Scheme 4, salt 32 was treated with NaOH in a biphasic system with dichloromethane, allowing the extraction of the released rimantadine (S)-1 in the organic layer. Expectedly, this simple procedure afforded target free rimantadine (S)-1 with nearly quantitative chemical yield and of uncompromised (99.7% ee) enantiomeric purity. Table 4. Application of (R)-phenoxypropionic acid 13 as a resolution reagent.  As shown in Table 4, we started with the mixing of racemic rimantadine 10 and (R)phenoxypropionic acid 13 in aqueous (5%) acetone, resulting in the precipitation of diastereomeric salt 32, which was isolated with 34.7% yield and showed 88.4% (S)-enantiomeric excess of the amine residue in it. Thus, obtained product 32 was next crystallized from aqueous ethyl acetate allowing precipitation of 87.3% of the original material 32 with noticeably increased enantiomeric purity (98.7% ee) of the constituent residue of rimantadine. One more crystallization procedure was performed using the same solvent system, affording salt 32 with 92% yield and enantiomeric purity of the target amine exceeding 99% ee. The overall yield of this three-step procedure was 28.1%, which we considered as adequate to provide reliable access to enantiomerically pure rimantadine.
Finally, we needed to develop a protocol for the isolation of free rimantadine (S)-1 from diastereomerically pure salt 32. As one can see from Scheme 4, salt 32 was treated with NaOH in a biphasic system with dichloromethane, allowing the extraction of the released rimantadine (S)-1 in the organic layer. Expectedly, this simple procedure afforded target free rimantadine (S)-1 with nearly quantitative chemical yield and of uncompromised (99.7% ee) enantiomeric purity.  As shown in Table 4, we started with the mixing of racemic rimantadine 10 and (R)phenoxypropionic acid 13 in aqueous (5%) acetone, resulting in the precipitation of diastereomeric salt 32, which was isolated with 34.7% yield and showed 88.4% (S)-enantiomeric excess of the amine residue in it. Thus, obtained product 32 was next crystallized from aqueous ethyl acetate allowing precipitation of 87.3% of the original material 32 with noticeably increased enantiomeric purity (98.7% ee) of the constituent residue of rimantadine. One more crystallization procedure was performed using the same solvent system, affording salt 32 with 92% yield and enantiomeric purity of the target amine exceeding 99% ee. The overall yield of this three-step procedure was 28.1%, which we considered as adequate to provide reliable access to enantiomerically pure rimantadine.
Finally, we needed to develop a protocol for the isolation of free rimantadine (S)-1 from diastereomerically pure salt 32. As one can see from Scheme 4, salt 32 was treated with NaOH in a biphasic system with dichloromethane, allowing the extraction of the released rimantadine (S)-1 in the organic layer. Expectedly, this simple procedure afforded target free rimantadine (S)-1 with nearly quantitative chemical yield and of uncompromised (99.7% ee) enantiomeric purity. Importantly, the resolving reagent (R)-13 was also recovered with excellent chemical yield. It should be noted that recycling of chiral acid (R)-13 is very important for the overall low-cost preparation of enantiomerically pure rimantadine (S)-1.

General Methods
All reagents and solvents were used as received. Reactions were monitored by thin-layer chromatography on Merck silica gel 60-F 254 coated 0.25-mm plates, detected by ultraviolet (UV). Flash chromatography was performed with the indicated solvents on silica gel (particle size 0.064-0.210 mm). Yields reported are for isolated, spectroscopically pure compounds. HPLC was performed on a SHIMADZU LC-2010CHT chromatography system and a CLASS-VPTM analysis data system. 1 HNMR spectra were recorded on a Brüker AVANCE III-400 spectrometer. Chemical shifts are given in ppm (d), referenced to tetramethylsilane (TMS). The letters s, d, t, q, m, and br stand for singlet, doublet, triplet, quartet, multiplet, and broad, respectively. Melting points were recorded on a Mettler Toledo MP70 Melting Point System and are not corrected.

Transformation of Rimantadine HCl Salt 9 to Free Amine 10
To a mixture of racemic rimantadine HCl 9 (100 g, 463.5 mmol) in CH 2 Cl 2 (1000 mL), 1 N NaOH (1000 mL) was added. The reaction mixture was stirred at room temperature for 30 min. The resultant mixture was separated. The organic layer was washed with 1 N NaOH (500 mL) and water (3 × 500 mL) then dried over Na 2 SO 4 . The organic layer was evaporated to afford racemic rimantadine 10 as a white residue (84.1 g, yield: >99%). 1

General Procedure for Resolution
To the solution of racemic rimantadine 10 in acetone with 5% H 2 O, resolving reagent 13−26 was added in acetone with 5% H 2 O at 50 • C. The amount of solvent was adjusted to 12 volumes. The mixture was stirred for 30 min at 50 • C, then cooled to room temperature and stirred for 24 h to form the corresponding salt. The salt was filtered, washed with acetone with 5% H 2 O, and dried under vacuum (<0.5 mmHg) at room temperature (rt). The solvent for crystallization is indicated in Tables 1-4  To a mixture of salt 32 (40 g, 116 mmol) in CH 2 Cl 2 (200 mL), 1 N NaOH (200 mL) was added. The reaction mixture was stirred at room temperature for 30 min. The resultant mixture was separated. Organic layer was washed with 1 N NaOH (100 mL) and water (2 × 80 mL), then dried over Na 2 SO 4 . The organic layer was evaporated and dried under vacuum to afford enantiomerically pure rimantadine (S)-1 as a white solid (21.5 g, yield: >99%).

Conclusions
In summary, we developed a new procedure for the resolution of commercially available racemic rimantadine hydrochloride to enantiomerically pure (S)-rimantadine free base. The resolving reagent (R)-phenoxypropionic acid, used in this method, can be conveniently recycled and reused for continuous preparation of the target chiral amine. Reasonably good chemical yields and operational ease of all transformations, coupled with low overall cost, bode well for its synthetic value for the preparation of enantiomerically pure rimantadine.