Biocatalysis in Continuous-Flow Mode: a Case- Study in the Enzymatic Kinetic Resolution of Secondary Alcohols Via Acylation

Home , Biocatalysis, Chiral resolution, Kinetic resolution, Racemic mixture

Biocatalysis 2017; 3: 27–36

Communication Open Access

Juliana Christina Thomas, Martha Daniela Burich, Pamela Taisline Bandeira, Alfredo Ricardo Marques de Oliveira, Leandro Piovan* Biocatalysis in continuous-flow mode: A case- study in the enzymatic kinetic resolution of secondary alcohols via acylation and deacylation reactions mediated by Novozym 435®

DOI 10.1515/boca-2017-0003 Received December 4, 2016; accepted February 7, 2017

Abstract: Enzymatic kinetic resolution reactions are a well- established way to achieve optically active compounds. When enzymatic reactions are combined to continuous- flow methodologies, other benefits are added, including reproducibility, optimized energy use, minimized waste generation, among others. In this context, we herein report a case study involving lipase-mediated transesterification by acylation and deacylation reactions of secondary alcohols/esters in batch and continuous-flow modes. 1 Introduction Acylation reactions were performed with high values of Enzymatic kinetic resolution (EKR) is a well-established enantiomeric excess (72 up to >99%) and enantioselectivity approach in the preparation of optically active compounds. (E > 200) for both batch and continuous-flow modes. On Nowadays, enzyme-mediated transformations, such as the other hand, for deacylation reactions using n-butanol EKR by lipases, are present in the toolbox of synthetic as nucleophile, enatiomeric excess ranged between 38 to chemists [1,2]. Enzymes are exceptional catalysts >99% and E from 6 to >200 were observed for batch mode. presenting enormous stereoselectivity for a number of For deacylation reactions in continuous-flow mode, natural and non-natural substrates, which allows the results were disappointing, as in some cases, very low or synthesis of valuable chiral intermediates, building- no conversion was observed. Enantiomeric excess ranged blocks and products with high enantiopurity. Currently, from 16 to >99% and enantioselectivity from 5 to >200 were there is an upward trend in the use and performance observed. In terms of productivity, continuous-flow mode of biocatalytic transformations in continuous-flow reactions were superior in both strategies (acylation: r mode [3]. Application of continuous-flow systems in from 1.1 up to 18.1-fold higher, deacylation: 2.8 up to 7.4- organic reactions offers several advantages, such as fold higher in continuous-flow than in batch mode). reproducibility, efficient control of reaction parameters, fast/homogeneous heating and lower costs in the Keywords: enzymatic kinetic resolution, acylation/ optimization of reaction conditions.[4,5] They are in deacylation, continuous-flow chemistry, lipases, harmony with several principles of green chemistry, such secondary alcohols as optimized energy use, increased processes safety, minimized use of solvents and waste generation.[6] When *Corresponding author: Leandro Piovan, Department of Chemistry, continuous-flow systems are combined with biocatalysis, Universidade Federal do Paraná, Paraná, Brazil, other benefits are added, including no enzyme lixiviation E-mail: [email protected] from support, removing the product almost immediately Juliana Christina Thomas, Martha Daniela Burich, Pamela Taisline from the contact with biocatalyst, especially in cases Bandeira, Alfredo Ricardo Marques de Oliveira, Department of Chemistry, Universidade Federal do Paraná, Paraná, Brazil where they can act as enzyme inhibitors [7], in addition

to the three major types of well-known chemo-, regio- (3 x 10 mL), dried over anhydrous MgSO4, filtered off and and stereoselectivity inherent to enzymes [8,9]. Aldolases dichloromethane was evaporated under reduced pressure [10], amidases [11], transaminases [12], oxidases [13,14], (alcohols 1-3) or distilled (alcohols 4-6). All compounds peroxidases [15] are some enzymes that have been were obtained with high purity and no additional steps of successfully applied to a continuous-flow approach. purification were required. Additionally, lipase-mediated esterification [16-21], Phenylethan-1-ol (1). Yield: 92% (5.620 g). GC-MS (70 interesterification [22,23], transesterification [24-26] and eV), m/z (relative intensity): 122 (M•+, 34%), 107 (90%), 79 enzymatic kinetic resolution (EKR) of racemic compounds (100%), 77 (54%), 51 (21%), 43 (24%). 1H NMR (200 MHz,

[27-32], have demonstrated the potential of combining CDCl3, TMS), δ (ppm): 1.42 (d, J = 6.5 Hz, 3H), 4.79 (q, J = 13 continuous-flow systems and biocatalysis. 6.5 Hz, 1H), 7.20-7.32 (m, 5H). C NMR (50 MHz, CDCl3), δ Recently, we described the EKR of cyanohydrins in (ppm): 25.1, 70.3, 125.4, 127.4, 128.5, 145.9. IR (cm-1): 3352, continuous-flow system [33] and observed a contrasting 2973, 1449, 1074, 698. behaviour between transesterification by acylation (4-Methoxyphenyl)ethan-1-ol (2). Yield: 82% (6.240 g). and deacylation reactions. While a deacylation GC-MS (70 eV), m/z (relative intensity): 152 (M•+, 30%), 137 employing n-butanol led to cyanohydrin derivatives in (100%), 109 (64%), 94 (35%), 77 (32%), 43 (21%). 1H NMR high productivity and enantioselectivity, for acylation (200 MHz, CDCl3, TMS), δ (ppm): 1.45 (d, J = 6.5 Hz, 3H), reaction, in continuous-flow mode, a very low conversion 3.79 (s, 3H), 4.82 (q, J = 6.5 Hz, 1H), 6.86 (d, J = 8.6 Hz, 13 rate was observed, which made this acylation strategy 2H), 7.28 (d, J = 8.7 Hz, 2H). C NMR (50 MHz, CDCl3), δ unfeasible. Due to this observation, we asked ourselves (ppm): 25.0, 55.3, 69.9, 113.8, 126.7, 138.0, 158.9. IR (cm-1): if this behaviour was a particular case to cyanohydrins 3377, 2968, 2056, 1889, 1611, 1458, 1244, 831. or would other organic compounds be susceptible to (4-Nitrophenyl)ethan-1-ol (3). Yield: 88% (7.355 g). the same particular behaviour, according to acylation or GC-MS (70 eV), m/z (relative intensity): 166 (M•+, 1%), deacylation conditions. In this way, secondary alcohols 152 (100%), 122 (25%), 107 (75%), 94 (41%), 77 (89%), 51 1 are an appropriate choice, especially due to high numbers (30%), 43 (58%). H NMR (200 MHz, CDCl3, TMS), δ (ppm): of EKR studies involving acylation and deacylation of this 1.50 (d, J = 6.5 Hz, 3H), 5.00 (q, J = 6.5 Hz, 1H), 7.52 (d, J = 8.6 13 class of organic compounds. Hz, 2H), 8.15 (d, J = 8.6 Hz, 2H). C NMR (50 MHz, CDCl3), In this context, in order to evaluate the differences δ (ppm): 25.3, 22.2, 71.2, 123.8, 126.7, 147.4, 148.9, 170.0. IR between transesterification by acylation and deacylation (cm-1): 3376, 1932, 1803, 1680, 1516, 1343, 1108, 855. reactions under continuous-flow conditions, a series of Octan-2-ol (4). Yield: 89% (5.795 g). 1H NMR (200 MHz, well-known Novozym 435® substrates [34-43], composed CDCl3, TMS), δ (ppm): 0.89 (m, 3H), 1.19 (d, J = 6.2 Hz, 3H), by benzylic and aliphatic alcohols and their respective 1.24-1.50 (m, 10H), 3.78 (sext, J = 6.2 Hz, 1H). 13C NMR (50 esters were evaluated both in batch and continuous-flow MHz, CDCl3), δ (ppm): 14.0, 22.6, 23.4, 25.7, 29.3, 31.8, 39.3, modes. Finally, a multigram scale reaction was performed 68.1. IR (cm-1): 3341, 2973, 2929, 2856, 1459, 1378, 1327. to validate our analytical investigation. 4-Methyl-pentan-2-ol (5). Yield: 94% (4.802 g). 1H NMR

(200 MHz, CDCl3, TMS), δ (ppm): 0.92 (d, J = 6.6 Hz, 6H), 1.19 (d, J = 3.2 Hz, 3H), 1.35-1.48 (m, 1H), 1.63-1.84 (m, 1H), 2 Materials and Methods 13 3.80-3.96 (m, 1H). C NMR (50 MHz, CDCl3), δ (ppm): 22.3, 23.1, 23.9, 24.8, 48.6, 66.1. IR (cm-1): 3348, 2922, 2966, 2863, 2.1 General Experimental Procedure for 1473, 1371, 1327. Syntheses of Racemic Alcohols 1-6 Hexan-3-ol (6). Yield: 92% (4.700 g). 1H NMR (200

MHz, CDCl3, TMS), δ (ppm): 0,93 (t, J = 7.2 Hz, 3H), 0.94 (t, Corresponding ketone (50 mmol) was solubilized in J = 6.9 Hz, 3H), 1.33-1.54 (m, 6H), 3.48-3.59 (m, 1H). 13C NMR methanol (50 mL) in an ice bath and NaBH4 (51 mmol) (50 MHz, CDCl3), δ (ppm): 9.8, 14.1, 18.8, 30.1, 39.1, 73.0. IR was added. After gas evolution stopped, the ice bath was (cm-1): 3348, 2966, 2944, 1459, 1319. removed and the reaction was carried out under magnetic stirring at room temperature. The reaction was monitored 2.2 General Experimental Procedure for by TLC until the consumption of ketone. Then, solvent was Syntheses of Racemic Esters 1a-6a removed under reduced pressure (alcohols 1-3) or distilled under atmospheric pressure (alcohols 4-6). Water and To a solution of corresponding alcohol (25 mmol) in aqueous HCl solution (1 mol L-1) were added until pH 6. dichloromethane (25 mL), acetic anhydride (50 mmol) The reaction media was extracted with dichloromethane and DMAP (1 crystal) were added and the reaction was Biocatalysis in continuous-flow mode 29 maintained under magnetic stirring at room temperature 2.3 General Procedure for Enzymatic Kinetic overnight. Then, the mixture was filtered through silica Resolution in Batch Mode via Acylation and the solvent was evaporated under reduced pressure. 1-Phenylethyl acetate (1a). Yield: 75% (3.079 g). To a solution of substrate (0.1 mmol) in n-hexane (2 mL), GC-MS (70 eV), m/z (relative intensity): 164 (M•+, 23%), vinyl acetate (0.4 mmol) and 20 mg of Novozym 435® 122 (100%), 105 (67%), 104 (89%), 107 (36%), 51 (15%), 43 were added and the reaction was carried out at 50°C 1 (55%). H NMR (200 MHz, CDCl3, TMS), δ (ppm): 1.53 (d, J with magnetic stirring. The progress of the reaction = 6.6 Hz, 3H), 2.06 (s, 3H), 5.88 (q, J = 6.6 Hz, 1H), 7.24-7.31 was monitored by periodic 200 µL aliquots, which were 13 (m, 5H). C NMR (50 MHz, CDCl3), δ (ppm): 21.3, 22.2, 72.3, diluted in 300 µL of n-hexane, and accomplished by 126.1, 127.8, 128.5, 141.7, 170.3. IR (cm-1): 2983, 1743, 1374, chiral GC analysis. Details for GC analyses can be found in 1232, 1069, 760, 696. Supporting Information. 1-(4-Methoxyphenyl)ethyl acetate (2a). Yield: 86% (4.176 g). GC-MS (70 eV), m/z (relative intensity): 194 (M•+, 2.4 General Procedure for Enzymatic Kinetic 27%), 134 (100%), 119 (40%), 105 (22%), 91 (33%), 77 Resolution in Batch Mode via Deacylation 1 (20%), 43 (34%). H NMR (200 MHz, CDCl3, TMS), δ (ppm): 1.51 (d, J = 6.5 Hz, 3H), 2.03 (s, 3H), 3.77 (s, 3H), 5.84 (q, J To a solution of substrate (0.1 mmol) in n-hexane (2 mL), = 6.5 Hz, 1H), 6.87 (d, J = 8.7 Hz, 2H), 7.29 (d, J = 8.7 Hz, n-butanol (0.4 mmol) and 20 mg of Novozym 435® 13 2H). C NMR (50 MHz, CDCl3), δ (ppm): 21.3, 21.9, 55.2, 72.0, were added and the reaction was carried out at 50°C 113.8, 127.6, 133.7, 159.3, 170.3. IR (cm-1): 2981, 2838, 2062, and magnetic stirring. The progress of the reactions 1885, 1733, 1612, 1513, 1242, 829. was monitored by periodic 200 µL aliquots, which were 1-(4-Nitrophenyl)ethyl acetate (3a). Yield: 82% (4.289 diluted in 300 µL of n-hexane, and accomplished by g). GC-MS (70 eV), m/z (relative intensity): 209 (M•+, 1%), chiral GC analysis. Details for GC analyses can be found in 167 (56%), 150 (16%), 119 (23%), 103 (24%), 91 (29%), Supporting Information. 1 77 (30%), 43 (100%). H NMR (200 MHz, CDCl3, TMS), δ (ppm): 1.56 (d, J = 6.7 Hz, 3H), 2.12 (s, 3H), 5.93 (q, J = 6.7 2.5 General Experimental Procedure for Deri- Hz, 1H), 7.51 (d, J = 8.8 Hz, 2H), 8.22 (d, J = 8.8 Hz, 2H). 13C vatization to Propionates 2b and 4b-6b

NMR (50 MHz, CDCl3), δ (ppm): 21.1, 22.2, 71.2, 123.8, 126.7, 147.4, 149.0, 170.0. IR (cm-1): 2966, 1734, 1602, 1517, 1346, Propionic anhydride (5 µL) and DMAP (1 crystal) 1242, 1068, 855. were added directly to the reaction aliquot and it was 2-Octyl acetate (4a). Yield: 84% (3.618 g). 1H NMR (200 maintained under magnetic stirring for 5 min. The aliquot

MHz, CDCl3, TMS), δ (ppm): 0.88 (t, J = 6.3 Hz, 3H), 1.20 was neutralized with aqueous NaHCO3 and dried over

(d, J = 6.3 Hz, 3H), 1.28-1.57 (m, 10H), 2.02 (s, 3H), 4.81-4.97 anhydrous MgSO4 before analysis. Details for GC analyses 13 (m, 1H). C NMR (50 MHz, CDCl3), δ (ppm): 14.0, 20.0, 21.3, can be found in Supporting Information. 22.5, 25.3, 29.1, 31.7, 35.9, 71.0, 170.7. IR (cm-1): 2929, 2863, 1738, 1467, 1371, 1238. 2.6 Continuous-flow system 4-Methyl-2-pentyl acetate (5a). Yield: 81% (2.920 g). 1H

NMR (200 MHz, CDCl3, TMS), δ (ppm): 0.88-0.91 (m, 6H), The continuous-flow system consisted of a syringe pump 1.20 (d, J = 6.2 Hz, 3H), 1.48-1.74 (m, 3H), 2.02 (s, 3H), 4.91- connected to the reactor through a Teflon cannula. The 13 5.07 (m, 1H). C NMR (50 MHz, CDCl3), δ (ppm): 20.4, 21.3, reactor itself was an empty HPLC stainless steel column 22.3, 22.8, 24.7, 45.1, 69.4, 170.6. IR (cm-1): 2958, 1739, 1454, (74.0 x 4.6 mm) that was previously washed to remove the 1370, 1232. stationary phase. It was then filled with the biocatalyst, 3-Hexyl acetate (6a). Yield: 64% (2.307 g). 1H NMR the supported lipase Novozym 435® (200 mg or 100 mg,

(200 MHz, CDCl3, TMS), δ (ppm): 0.84-0.94 (m, 7H), 1.22- internal volume 0.43 mL and 0.40 mL, respectively), 1.63 (m, H), 2.04 (s, 3H), 4.76-4.89 (m, 1H). 13C NMR (50 and deactivated glass wool in both ends, in order to

MHz, CDCl3), δ (ppm): 9.5, 13.9, 18.5, 21.2, 26.9, 35.7, 75.3, prevent enzyme agglomeration on the top of reactor. A 170.9. IR (cm-1): 2966, 1738, 1238, 907, 738. homemade heating block (8.5 x 5.0 x 2.0 cm) controlled by a commercial thermostat was used to set the reaction temperatures. 30 J.Ch. Thomas, et al.

2.7 General Procedure for Enzymatic Kinetic 3.2 Chemical Synthesis Resolution in Continuous-Flow Mode via Acylation Racemic alcohols 1-6 were prepared from the corresponding carbonyl compounds, with yields up to -1 A solution of substrate (0.1 mmol mL ) and vinyl acetate 94%, using conventional reduction with NaBH4. Racemic (4 equivalents) in n-hexane (5 mL) was eluted through the esters 1a-6a were prepared from the corresponding column filled with Novozym 435® (200 mg) maintained alcohols, with yields up to 86%, using acetic anhydride as at 50°C, with a flow rate from 0.1 mL min-1 to 1 mL min-1. acyl donor and 4-(N,N-dimethylamino)pyridine (DMAP) It was collected 0.5 mL of solution in each flow rate to as catalyst (see experimental section and supporting determine the optimal rate for the reactions. information for details).

2.8 General Procedure for Enzymatic Kinetic 3.3 Enzymatic Kinetic Resolution (EKR) Resolution in Continuous-Flow Mode via Reactions Deacylation For reactions in continuous-flow mode, the substrate and A solution of substrate (0.1 mmol mL-1) and n-butanol (4 vinyl acetate, for acylation, or the substrate and n-butanol, equivalents) in n-hexane (5 mL) was eluted through the for deacylation, were dissolved in n-hexane (final volume column filled with Novozym 435® (200 mg) maintained of 5.0 mL), and eluted through the column using an at 50°C, with a flow rate from 0.1 mL min-1 to 1 mL min-1. appropriate flow rate, controlled by the syringe pump. The It was collected 0.5 mL of solution in each flow rate to column was filled with biocatalyst (100-200 mg – reactor determine the optimal rate for the reactions. volume of 0.4 and 0.43 mL, respectively) and maintained under constant temperature (50°C; Table 1 and Table 2). For comparative purpose, batch reactions were carried out 3 Results and Discussion in parallel. Substrate, biocatalyst (20.0 mg), vinyl acetate, for acylation, or n-butanol, for deacylation, and n-hexane 3.1 Selection of Substrates were added to a 4 mL sealed vial (final volume of 2 mL) and the reaction medium was maintained under magnetic A series of benzylic and aliphatic alcohols and their stirring and at 50°C (Table 1 and Table 2). respective esters (Figure 1), was proposed to be evaluated In continuous-flow mode, enzymatic acylation of in Novozym 435®-mediate EKR reaction. For benzylic all benzylic alcohols reached ideal conversion with high alcohols (1-3) and esters (1a-3a) electronic effects of the enantioselectivity (Table 1 – entries 1-3). For alcohol 1, groups at the para position were investigated and for the flow rate of 1 mL min-1 was sufficient to reach ideal aliphatic alcohols (4-6) and esters (4a-6a) the influence conversion (50%) and 99% enantiomeric excess for both of the size of groups attached to the chirality center were (S)-1 and (R)-1a. For EKR of para-substituted alcohols also explored. 2 and 3, decrease in flow rates to 0.5 and 0.7 mL min-1, respectively, was necessary. This modification implied higher residence time of the substrates inside the reactor. Conversion rates reached 50% and (S)-alcohols and (R)-esters were obtained in 99% enantiomeric excess (Table 1 – entries 2-3). For acylation reactions in continuous-flow mode of aliphatic alcohols 4-6, all flow rates (0.1 up to 1.0 mL min-1) were tested and we observed a decrease in enantiomeric excess values (Table 1 – entries 4-6). Aiming to reach higher values of enantiomeric excess in continuous-flow mode, we submitted alcohols 4-6 to EKR using a smaller amount of biocatalyst (100 mg), since alcohols 4-6 are not ideal substrates and there was so much enzyme that Figure 1: Compounds employed as substrates in EKR in batch and continuous-flow modes both enantiomers were transformed (Table 1 – entries 5- 8, values in parentheses). We also observed a decrease in selectivity in EKR reaction in batch mode when a larger Biocatalysis in continuous-flow mode 31

Table 1: EKR of alcohols 1-6 via acylation reaction in continuous-flow and batch mode.

Alcohol / Flow Batch Entry Rate / Time a / c b / % e.e. c / % r d / µmol E f Time / h c b / % e.e. c / % r e / µmol E f mL min-1 min min-1 g-1 min-1 g-1 e.e.s e.e.p e.e.s e.e.p 1 1.0 0.43 50 >99 >99 250.0 >200 1 50 >99 >99 41.7 >200 2 0.5 0.86 50 >99g >99 125.0 >200 6 50 >99g >99 6.9 >200 3 0.7 0.61 50 >99 >99 175.0 >200 3 50 >99 >99 13.9 >200 4 0.3 1.43 51 >99g 93 75.0 >200 1 50 >99g >99 41.7 >200 (0.1)h 4.00 (51) (>99)g (95) (51.0) 5 0.1 4.30 50 96g 96 25.0 194 1 50 >99g >99 41.7 >200 (0.1)h 4.00 (50) (98)g (98) (50.0) >200 6 0.1 4.30 52 88g 82 30.0 29 2 50 >99g >99 20.8 >200 (0.1)h 4.00 (43) (72)g (94) (21.5) 70 (0.1)h,i 8.00 (57) (>99)g (75) (23.5) 35

Reaction conditions: Flow mode: substrate (0.1 mmol mL-1), vinyl acetate (0.4 equivalents) and n-hexane (5 mL) and Novozym 435® (200 mg); Batch mode: substrate (0.1 mmol), vinyl acetate (0.4 mmol), n-hexane (2 mL) and Novozym 435® (20 mg). Temperature for a b c both 50°C; Residence time: reactor volume / flow rate x number of cycles; Conversion: ees / (ees + eep); Enantiomeric excess: (R – S) d e f / (R + S) x 100 (determined by chiral GC analysis); Productivity (flow): [P] f / me; Productivity (batch): nP / t me; Enantiomeric ratio: g h E = ln {[eep (1 - ees)] / (eep + ees)} / ln {[eep (1 + ees)] / (eep + ees)}; Determined by derivatization to corresponding propionate; Reactions carried out with 100 mg of Novozym 435®; i 2 cycles of elution. amount of biocatalyst was used, which suggests an For deacylation reactions in continuous-flow mode, EKR influence of enzyme quantity on enantioselectivity. of benzylic esters 1a and 2a did not reach ideal conversion The slowest flow rate (0.1 mL min-1) gave the best (46 % - Table 2, entries 1a and 2a), which was not observed results for EKR of alcohols 4 and 5 and high enantiomeric in acylation reactions (c = 50% - Table 1, entries 1 and 2). excesses (93 up to > 99%) were obtained. Taking into However, enantioselectivity in continuous-flow mode was account that the same residence time was necessary for as high as was found in batch mode reactions for both alcohols 4 and 5, the influence of chain length attached compounds 1a and 2a. Deacylation of benzylic alcohol 3a to the chiral center was not as pronounced through in continuous-flow mode presented the same behaviour in enzymatic activity and enantioselectivity (Table 1 – entries the acylation reaction - high values of enantiomeric excess 4 and 5). For alcohol 6, which possess ethyl and n-propyl (>99 %) and ideal conversion rate for both acylation and moieties attached to the chirality center, conversion of deacylation reactions were observed (Table 1 – entry 3; 43% was observed (72% e.e. for (S)-6 and 94% for (R)-6a) Table 2 – entry 3a), although, a decrease on flow rate when in the first cycle. To increase the enantiomeric excess, compared to acylation reaction was needed (0.1 mL min-1 the solution was eluted through the column twice at the during 2 cycles for deacylation and 0.7 mL min-1 for acylation). slowest flow rate, resulting in high conversion (57%) and For the deacylation reaction of aliphatic ester 4a, enantiomeric excess for this compound (Table 1 – entry 6). selectivity was as high as in acylation reactions (batch 32 J.Ch. Thomas, et al.

Table 2: EKR of alcohols 1-6 via deacylation reaction in continuous-flow and batch mode.

Ester / Flow Batch Entry Rate / mL Time a / c b / % e.e. c / % r d / µmol E f Time / h c b / % e.e. c / % r e / µmol E f min-1 min min-1 g-1 min-1 g-1 e.e.s e.e.p e.e.s e.e.p 1a 0.1g 8 46 84 >99 11.5 >200 10 49 94 >99 4.1 >200 2a 0.1g 8 46 85 >99h 11.5 >200 24 48 92 >99h 1.7 >200 3a 0.1g 8 50 >99 >99 12.5 >200 24 49 96 >99 1.7 >200 4a 0.1g 8 27 37 >99h 6.8 >200 24 46 85 >99h 1.6 >200 5a 0.1g 8 20 16 64h 5.5 5 24 41 58 82h 1.4 18 6a 0.1g 8 <5 nd nd nd nd 24 38 38 61h 1.3 6

Reaction conditions: Flow mode: substrate (0.1 mmol mL-1), n-butanol (0.4 equivalents) and n-hexane (5 mL) and Novozym 435® (200 mg); Batch mode: substrate (0.1 mmol), n-butanol (0.4 mmol), n-hexane (2 mL) and Novozym 435® (20 mg). Temperature for both 50°C; a Residence b c time: reactor volume / flow rate x number of cycles; Conversion: ees / (ees + eep); Enantiomeric excess: (R – S) / (R + S) x 100 (determined by d e f chiral GC analysis); Productivity (flow): [P] f / me; Productivity (batch): nP / t me; Enantiomeric ratio: E = ln {[eep (1 - ees)] / (eep + ees)} / ln g h {[eep (1 + ees)] / (eep + ees)}; 2 cycles of elution; Determined by derivatization to corresponding propionate; nd: not determined.

and continuous-flow modes), although, conversion rates which makes acylation a more viable strategy to achieve were slower in both modes (46 and 27%, respectively, optically active secondary alcohols and esters. Table 2 – entry 4a). Deacylation of ester 5a presented In summary, the potential of combining lower selectivity and slower conversion rates than for biocatalysis and continuous-flow is indisputable, the acylation reactions in both batch and continuous- however, the particularities of each mode must be flow modes (Table 2 – entry 5a). A loss of selectivity was noted. Transesterification reactions via acylation and observed when the deacylation reaction was carried out deacylation are common in biocatalysis laboratories, in continuous-flow mode (E = 5), compared to batch but changing from batch to continuous-flow mode is not mode (E = 18), which can also be justified due to the straightforward. As cited before, in our previous work larger amount of biocatalyst in continuous-flow than in involving EKR of cyanohydrins [33], a deacylation strategy batch mode. Batch mode deacylation of ester 6a (Table was superior to acylation. In this work, acylation reactions 2 – entry 6a) presented low enantioselectivity (E = 6), were more viable than deacylation reactions, both in which was different from acylation reactions (E > 200). batch and continuous-flow modes, since higher reaction When this reaction was carried out in continuous-flow rates and enantioselectivity were observed. mode, satisfactory conversion was not observed, even in Besides the differences between the two acylation and the slowest flow rate during 2 cycles of elution (Table 2 – deacylation strategies, differences between reactions in entry 6a). Selectivity parameters were not calculated for batch and continuous-flow modes were also observed. The deacylation of ester 6a in continuous-flow mode due to greatest difference between continuous-flow and batch low conversion. modes can be measured by productivity (r). Productivity is It is important to highlight that deacylation reactions a parameter that also considers the quantity of biocatalyst, were slower than acylation reactions for all compounds, measuring how much product can be obtained in 1 min Biocatalysis in continuous-flow mode 33 Page 33 / Line 16 There should be equations below the paragraph that ends in line 16: using 1 g of enzyme and it can be calculated based on two obtained enantiopure compounds, to determine reaction equations below, where [P] is the concentration of the yield and optical rotation [α]D of enantiopure alcohol and product (µmol mL-1), f is the flow rate (mL min-1), n is the ester. Page 33 / Line 16 P amount of product (µmol), t is time (min) and m is the e amountThere of should enzyme be(g) [35].equations below the paragraph that ends in line 16: Page 33 / Scheme 1

Should be: (please, find attached high-resolution version)

For benzylic alcohols and esters, productivity values werePage 2.7 to33 18.1 / Scheme and 2.8 to1 7.4 fold higher in continuous- flow mode than in batch mode, respectively. For aliphaticShould alcohols be: (please, and esters, find in attached the optimized high -reactions,resolution Schemeversion) 1: Preparative scale reaction of alcohol 1. productivity values were 1.1 to 2.3 and 3.9 to 4.3 fold higher than in batch mode, respectively. The productivity value First of all, the flow rate value of 1.0 mL min-1 was for ester 6a in continuous-flow mode was not determined established (according to general procedure in due to low conversion (Figure 2). materials and methods section) as ideal to reach 50% of In continuous-flow mode, even using a larger amount conversion using a concentration of 0.1 mmol mL-1. 50 of enzyme, higher productivity values were observed in mL of this solution was eluted through the column and comparison to batch mode (for ester this parameter was 6a Page 33 / Lines 29the andenantiopure 30 compounds were separate by column not calculated in continuous-flow mode). This observed chromatography (mobile phase n-hexane:ethyl acetate behaviour of productivities in batch andThe continuous-flow symbol is missing.9:1) resulting There in should46% yield, be in 9.7 front (c = 1.0,of 9.7 n-hexane) (line 29) and -6.7 (line 30) mode is in accordance with those already reported in the and e.e. >99% for ester (R)-1a and 43% yield, -6.7 (c = 1.0, literature [34]. n-hexane) and e.e. >99% for alcohol (S)-1. Page 33 / Lines 29 and 30 3.4The Enzymatic symbol is Kineticmissing. Resolution There should (EKR) be in front3.5 ofEnzymatic 9.7 (line 29)Kinetic and -Resolution6.7 (line 30) (EKR) Reaction in Preparative Scale Reaction in a Multigram Scale

As cited before, the coupling of biocatalysis and In order to evaluate the reproducibility of results obtained continuous-flow system is very productive and reactions in the continuous-flow system and also the possible reuse can be performed in an analytical or preparative scale. of the biocatalyst, a multigram reaction was performed. This way, a preparative scale EKR reaction of model For this purpose, a solution with 2.0 g of alcohol 1 in 164 alcohol 1 (Scheme 1) was carried out aiming to isolate the mL of n-hexane (0.1 mol L-1) was prepared and eluted

A B

Figure 2: Comparative productivity values for EKR of alcohols 1-6 (A) and esters 1a-6a (B) in batch and continuous-flow modes.

34 J.Ch. Thomas, et al. through the column at 1 mL min-1 in aliquots of 4.0 mL. At References the end, 41 aliquots were collected and analysed by chiral GC. As can be seen in Figure 3 no decrease of conversion [1] Carvalho, A. C. L. M.; Fonseca, T. S.; Mattos, M. C.; Oliveira, was observed even after the elution of the entire stock M. C. F.; Lemos, T. L. G.; Molinari, F.; Romano, D.; Serra, I. Recent advances in lipase-mediated preparation of solution. pharmaceuticals and their intermediates. Int. J. Mol. Sci., 2015, This fact gives evidence to the high reproducibility of 16, 29682-29716. results, as the same conversion and enantiomeric excess [2] Loughlin, W. A. Biotransformations in organic synthesis. values were observed, even after the elution of 2.0 g of Bioresource Technol., 2000, 74, 49-62. substrate. [3] Junior, I. I.; Miranda, L. S. M.; Souza, R. O. M. A. Towards a continuous flow environment for lipase-catalyzed reactions. J. Mol. Catal. B: Enzym., 2013, 85-86, 1-9. 4 Conclusion [4] Schonstein, L.; Forro, E.; Fulop, F. Continuous-flow enzymatic resolution strategy for the acylation of amino alcohols with a remote stereogenic centre: synthesis of calycotomine In conclusion, it was possible to perform a successful enantiomers. Tetrahedron: Asymmetry, 2013, 24, 202-206. coupling between biocatalysis and flow-chemistry [5] Wiles, C.; Watts, P. Continuous flow reactors: a perspective. employing a simple, cheap and robust homemade Green Chem., 2012, 14, 38-54. continuous-flow system. A series of optically active [6] Machado, A. H. L.; Pandoli, O.; Miranda, L. S. M.; Souza, R. O. secondary alcohols and esters could be obtained M. A. Micro reatores: novas oportunidades em síntese química. Rev. Virtual Quim., 2014, 6, 1076-1085. with high productivity in continuous-flow mode via [7] Souza, R. O. M. A.; Miranda, L. S. M. Reações sob fluxo -1 transesterification by acylation (12.5-250.0 µmol min contínuo: da química verde a um processo verde. Rev. Virtual -1 -1 -1 g ) and deacylation reactions (5.5-12.5 µmol min g ). Quim., 2014, 6, 34-43. Enantioselectivity was higher in acylation than in [8] Kapoor, M.; Gupta, M. N. Lipase promiscuity and its deacylation reactions, which demonstrates that acylation biochemical applications. Process Biochem., 2012, 47, 555-569. reactions are the most viable approach to obtain optically [9] Clouthier, C. M.; Pelletier, J. N. Expanding the organic toolbox: a guide to integrating biocatalysis in synthesis. Chem. Soc. active secondary alcohols employing flow chemistry. Rev., 2012, 41, 1585-1605. [10] Lamble, H. J.; Royer, S. F.; Hough, D. W.; Danson, M. J.; Taylor, Acknowledgements: L. Piovan and A. R. M. Oliveira G. L.; Bull, S. D. A thermostable aldolase for the synthesis thank Brazilian National Council for Scientific and of 3-deoxy-2-ulosonic acids. Adv. Synth. Catal., 2007, 349, Technological Development (CNPq, Conselho Nacional 817-821. [11] Preez, R.; Clarke, K. G.; Callanan, L. H.; Burton, S. G. Modelling de Desenvolvimento Científico e Tecnológico, Brazil) for of immobilised enzyme biocatalytic membrane reactor financial support (Proc. 456834/2014). J. C. Thomas, P. T. performance. J. Mol. Catal. B: Enzym., 2015, 119, 48-53. Bandeira and M. D. Burich thank CAPES and UFPR-TN for [12] Andrade, L. H.; Kroutil, W.; Jamison, T. F. Continuous flow fellowships. synthesis of chiral amines in organic solvents: immobilization

uV (x100,000) 8.0 Chromatogram

7.0

6.0

5.0 rac st

4.0 1 aliquot th 10 aliquot 3.0 th 20 aliquot 2.0 th 30 aliquot 1.0 th 40 aliquot 0.0 2.5 5.0 7.5 10.0 12.5 15.0 min

Figure 3: Chromatograms of racemic mixture of alcohol 1 and ester 1a (red line) and aliquots. Biocatalysis in continuous-flow mode 35

of E. coli cells containing both ω-transaminase and PLP. Org. [27] Manoel, E. A.; Pais, K. C.; Flores, M. C.; Miranda, L. S. M.; Lett., 2014, 16, 6092-6095. Coelho, M. A. Z.; Simas, A. B. C.; Freire, D. M. G.; Souza, R. [13] Zambelli, P.; Tamborini, L.; Cazzamalli, S.; Pinto, A.; Arioli, O. M. A. Kinetic resolution of a precursor for myo-inositol S.; Balzaretti, S.; Plou, F. J.; Arrojo, L. F.; Molinari, F.; Conti, phosphates under continuous flow conditions. J. Mol. Catal. B: P.; Romano, D. An efficient continuous flow process for the Enzym., 2013, 87, 139-143. synthesis of a non-conventional mixture of fructooligosac- [28] Liu, Z.; Burgess, K. Continuous flow biocatalytic resolutions charides. Food Chem., 2016, 190, 607-613. of methyl sulfinylacetates. Tetrahedron: Lett., 2011, 52, [14] Tomaszewski, B.; Schmid, A.; Buehler, K. Biocatalytic 6325-6327. production of catechols using a high pressure tube-in-tube [29] Reetz, M. T.; Wiesenhöfer, W.; Franciò, G.; Leitner, W. segmented flow microreactor. Org. Process Res. Dev., 2014, 18, Biocatalysis in ionic liquids: batchwise and continuous flow 1516-1526. processes using supercritical carbon dioxide as the mobile [15] Ruiz, I. R.; Codina, E. M.; Ackermann, T. N.; Llobera, A. Photonic phase. Chem. Commun., 2002, 992-993. lab-on-chip (PhLOC) for enzyme-catalyzed reactions in [30] Hellner, G.; Boros, Z.; Tomin, A.; Poppe, L. Novel sol-gel continuous flow. Microfluid. Nanofluid., 2015, 18, 1277-1286. lipases by designed bioimprinting for continuous-flow kinetic [16] Madarász, J.; Németh, D.; Bakos, J.; Gubicza, L.; Bakonyi, P. resolutions. Adv. Synth. Catal., 2011, 353, 2481-2491. Solvent-free enzymatic process for biolubricant production [31] Boros, Z.; Falus, P.; Márkus, M.; Weiser, D.; Oláh, M.; in continuous microfluidic reactor. J. Clean. Prod., 2015, 93, Hornyánszky, G.; Nagy, J.; Poppe, L. How the mode of Candida 140-144. antarctica lipase B immobilization affects the continuous-flow [17] Sutili, F. K.; Ruela, H. S.; Nogueira, D. O.; Leal, I. C. R.; Miranda, kinetic resolution of racemic amines at various temperatures. J. L. S. M.; Souza, R. O. M. A. Enhanced production of fructose Mol. Catal. B: Enzym., 2013, 85-86, 119-125. ester by biocatalyzed continuous flow process. Sustain. Chem. [32] Boros, Z.; Weiser, D.; Márkus, M.; Abaháziová, E.; Magyar, Process., 2015, 3, 6. A.; Tomin, A.; Koczka, B.; Kovács, P.; Poppe, L. Hydrophobic [18] Sutili, F.; Nogueira, D. O.; Leite, S. G. F.; Miranda, L. S. M.; adsorption and covalent immobilization of Candida antarctica Souza, R. O. M. A. Lipase immobilized in microemulsion based lipase B on mixed-function-grafted silica gel supports for organogels (MBGs) as an efficient catalyst for continuous- continuous-flow biotransformations. Process Biochem., 2013, flow esterification of protected fructose. RSC Adv., 2015, 5, 48, 1039-1047. 37287-37291. [33] Thomas, J. C.; Aggio, B. B.; Oliveira, A. R. M.; Piovan, L. [19] Junior, I. I.; Flores, M. C.; Sutili, F. K.; Leite, S. G. F.; Miranda, High-throughput preparation of optically active cyanohydrins L. S. M.; Leal, I. C. R.; Souza, R. O. M. A. Lipase-catalyzed mediated by lipases. Eur. J. Org. Chem., 2016, 36, 5964-5970. monostearin synthesis under continuous flow conditions. Org. [34] Csajági, C.; Szatzker, G.; Toke, E. R.; Ürge, L.; Darvas, F.; Poppe, Process Res. Dev., 2012, 16, 1098-1101. L. Enantiomer selective acylation of racemic alcohols by lipases [20] Wang, S. S.; Li, Z. J.; Sheng, S.; Wu, F. A.; Wang, J. Microfluidic in continuous-flow bioreactors. Tetrahedron: Asymmetry, 2008, biocatalysis enhances the esterification of caffeic acid and 19, 237-246. methanol under continuous-flow conditions. J. Chem. Technol. [35] Matsuda, T.; Watanabe, K.; Harada, T.; Nakamura, K.; Arita, Biotechnol., 2015, 91, 555-562. Y.; Misumi, Y.; Ichikawa, S.; Ikariya T. High-efficiency and [21] Woodcock, L. L.; Wiles, C.; Greenway, G. M.; Watts, P.; Wells, A.; minimum-waste continuous kinetic resolution of racemic Eyley, S. Enzymatic synthesis of a series of alkyl esters using alcohols by using lipase in supercritical carbon dioxide. Chem. Novozyme 435 in a packed-bed, miniaturized, continuous flow Commun., 2004, 2286-2287. reactor. Biocatal. Biotransform., 2008, 26, 501-507. [36] Lozano, P.; Verdugo, E. G.; Karbass, N.; Montague, K.; Diego, T.; [22] Varón, E. Y.; Joli, J. E.; Balcells, M.; Torres, M.; Garayoa, R. C. Burguete, M. I.; Luis, S. V. Supported Ionic Liquid-Like Phases Synthesis of poly(ethyl acrylate-co-allyl acrylates) from acrylate (SILLPs) for enzymatic processes: continuous KR and DKR in mixtures prepared by a continuous solvent-free enzymatic SILLP–scCO2 systems. Green Chem., 2010, 12, 1803-1810. process. RSC Adv., 2012, 2, 9230-9236. [37] Reetz, M. T.; Wiesenhöfer, W.; Franciò, G.; Leitner, W. [23] Hernandez, A. L.; Otero, C.; Martín, E. H.; Garcia, H. S.; Hill, C. Continuous flow enzymatic kinetic resolution and enantiomer G. Interesterification of sesame oil and a fully hydrogenated separation using ionic liquid/supercritical carbon dioxide fat using an immobilized lipase catalyst in both batch and media. Adv. Synth. Catal., 2003, 354, 1221-1228. continuous-flow processes. Eur. J. Lipid Sci. Technol., 2007, [38] Wang, B.; Jiang, L.; Wang, J.; Ma, J.; Liu, M.; Yu, H. A tandem 109, 1147-1159. and fully enzymatic procedure for the green resolution of chiral [24] Tran, D. T.; Lin, Y. J.; Chen, C. L.; Chang, J. S. Modeling the alcohols: acylation and deacylation in non-aqueous media. methanolysis of triglyceride catalyzed by immobilized lipase in Tetrahedron: Asymmetry, 2011, 22, 980-985. a continuous-flow packed-bed reactor. Appl. Energ., 2014, 126, [39] Debbeche, H.; Toffano, M.; Fiaud, J. C.; Zouioueche, L. A. 151-160. Multi-substrate screening for lipase-catalyzed resolution of [25] Wang, J.; Gu, S. S.; Cui, H. S.; Wu, X. Y.; Wu, F. A. A novel arylalkylethanols with succinic anhydride as acylating agent. J. continuous flow biosynthesis of caffeic acid phenethyl Mol. Catal. B: Enzym., 2012, 66, 319-324. ester from alkyl caffeate and phenethanol in a packed bed [40] Patel, R. N.; Banerjee, A.; Nanduri,, V.; Goswami, A.; microreactor. Bioresour. Technol., 2014, 158, 39-47. Comezoglu, F. T. Enzymatic resolution of racemic secondary [26] Carnero, A.; Sanghvi, Y. S.; Gotor, V.; Fernández, S.; Ferrero, M. alcohols by lipase B from Candida antarctica. J. Am. Oil Chem. Process development of biocatalytic regioselective 5′-O-levuli- Soc., 2000, 77, 1015-1019. nylation of 2′-deoxynucleosides. Org. Process Res. Dev., 2015, [41] Shang, C. Y.; Li, W. X.; Zhang, R. F. Immobilized Candida 19, 701-709. antarctica lipase B on ZnO nanowires/macroporous silica 36 J.Ch. Thomas, et al.

composites for catalyzing chiral resolution of (R,S)-2-octanol. [43] Ferreira, H.; Rocha, L. C.; Severino, R. P.; Porto, A. L. M. Enzyme Microb. Technol., 2014, 61-62, 28-34. Syntheses of enantiopure aliphatic secondary alcohols [42] Ren, L.; Xu, T.; He, R.; Jiang, Z.; Zhou, H.; Wei, P. A green and acetates by bioresolution with lipase B from Candida resolution–separation process for aliphatic secondary antarctica. Molecules, 2012, 17, 8955-8967. alcohols. Tetrahedron: Asymmetry, 2013, 24, 249-253.