Review Reports - Ring-Opening of Epoxides with Amines for Synthesis of β-Amino Alcohols in a Continuous-Flow Biocatalysis System

Round 1

Reviewer 1 Report

“But so far synthesis of β-amino alcohols from the ring-opening of epoxides with aromatic amines catalyzed by enzyme is rarely studied during recent years.” Phrasing is a bit awkward, maybe “But so far the synthesis of β-amino alcohols from the ring-opening of epoxides with aromatic amines catalyzed by enzymes has been rarely studied”.

“Lipase from Aspergillus Oryzae” should be “Lipase from Aspergillus oryzae”

Paragraph between lines 44 to 47 is too long, please split in two

Lines 49-50, “The enzymatic reactions are relatively mild…[…]… or require longer reaction time usually 12 hours”, longer than what? Please provide term for comparison

Lines 60-61 “which established foundation for further drug screening.” Please rephrase, awkward sentencing, and please check verb tense

Experimental setup

Please provide more details on the material and configuration of the microreactor, e.g. metal, polymeric (and specifically which)? Was it coiled? Was the Y-shaped junction commercially available (supplier? Material?)

Lipozyme TL IM is provided as a silica granulate, how was is kept inside the tubing? Was a frit used? Other? Please clarify

The authors state that “resulting stream (31.4 μL min-1) was passed through microchannel reactor, reaction at 35 ºC for 20 minutes, the solution was collected in a glass vessel”. According to the text, “Microchannel reactor (inner diameter ID= 2.0 mm, length = 100 cm)” hence the free volume of the reactor is 3.14 cm³, hence 3140 µL. If there was no packing 3140 µL /31.4 μL min^-1 would result in a residence time of 100 min. So, were the 20 minutes of residence time determined based on a void volume and how was this determined?

Effect of reaction media and catalyst. Did the authors perform duplicates of these experiments and (thus provide error intervals), so that one can establish that the yield differences highlighted in Table 1 have statistical significance. Same comment applies to yield data presented in other Figures Could the authors add a column in Table 1 with a parameter such as log P that could easily support the claim that the key parameter highlighted by the processed data is solvent polarity?

Please rephrase and clarify “When the molar ratio reached 1:1, the highest yield reached 91%, further change in molar ratio decreased the yield of product. Due to the amount of enzyme and the number of catalytically active sites on the enzyme was limited in the fixed-length microreactor”

Please rephrase and clarify “Besides, we found that the bind of the substrates by the enzyme forming complexs, which more inclined to nucleophilic attack at the less hindered carbon atom of the epoxide ring form the major products 3 (Table 2), results were 126 confirmed by NMR”. Awkward sentencing

“Thermomyces Lanuginosus” should be “Thermomyces lanuginosus”

Author Response

Dear reviewer:

Thank you very much for your valuable suggestions on our work “Ring-Opening of Epoxides with Amines for Synthesis of β-Amino Alcohols in a Microfluidic Biocatalysis System” (Manuscript ID: catalysts-991641). We have made major revisions according to your opinion. We hope that with your help and advice, our work can be greatly improved. Our responses to your comments are as follows (all the changes made in the paper I have marked with red font). Thanks again.

Q1.“But so far synthesis of β-amino alcohols from the ring-opening of epoxides with aromatic amines catalyzed by enzyme is rarely studied during recent years.” Phrasing is a bit awkward, maybe “But so far the synthesis of β-amino alcohols from the ring-opening of epoxides with aromatic amines catalyzed by enzymes has been rarely studied”.

A1: Thank you for your suggestions, we have revised this sentence. (Line 43-44)

Q2.“Lipase from Aspergillus Oryzae” should be “Lipase from Aspergillus oryzae”.

A2: Thank you for your suggestions, we have revised this word. (Line 44)

Q3. Paragraph between lines 44 to 47 is too long, please split in two.

A3: Thank you for your suggestions, we split the paragraph into two sentences. (Line 44-47)

Q4. Lines 49-50, “The enzymatic reactions are relatively mild…[…]… or require longer reaction time usually 12 hours”, longer than what? Please provide term for comparison.

A4: Thank you for your question. We compare the enzymatic synthesis of β-amino alcohols with chemical catalytic methods, we have provide term for comparison in our manuscript. Thank you. (Line 49-52)

Q5. Lines 60-61 “which established foundation for further drug screening.” Please rephrase, awkward sentencing, and please check verb tense.

A5: Thank you for your question, we have rechecked grammar of the sentence and given a more appropriate description (Line 62-64). Thank you.

Experimental setup

Q6. Please provide more details on the material and configuration of the microreactor, e.g. metal, polymeric (and specifically which)? Was it coiled? Was the Y-shaped junction commercially available (supplier? Material?)

A6: Thank you for your question. Flow reactor and Y-mixer were made of perfluoroalkoxy (PFA) and purchased from Beijing Haigri Medical Engineering Design Co., Ltd. The flow reactor was made of coiled PFA tube. We have explained the materials, composition and commercial sources of the reactor and Y-mixer in the manuscript (Line 157-160). Thank you.

Q7. Lipozyme TL IM is provided as a silica granulate, how was is kept inside the tubing? Was a frit used? Other? Please clarify.

A7: Thank you for your question. First, we used absorbent cotton at both ends of the micro-pipeline to keep the immobilized enzyme particles inside the tubing, secondly, because our reaction flow rate is relatively slow, the enzyme particles are evenly filled in the micropipeline with a diameter of 2 mm, the enzyme particles will not flow out of the micropipeline with the flow of the reaction solution. We have made additional explanations in the manuscript (Line 165-167). Thank you.

Q8. The authors state that “resulting stream (31.4 μL min^-1) was passed through microchannel reactor, reaction at 35 ºC for 20 minutes, the solution was collected in a glass vessel”. According to the text, “Microchannel reactor (inner diameter ID= 2.0 mm, length = 100 cm)” hence the free volume of the reactor is 3.14 cm³, hence 3140 µL. If there was no packing 3140 µL /31.4 μL min^-1 would result in a residence time of 100 min. So, were the 20 minutes of residence time determined based on a void volume and how was this determined?

A8: Thank you for your question. We control the residence time by setting different flow rates. The volume of a 1m PFA tube is 3.14 mL, and the void volume filled with enzyme is 628 µL. The void volume of the enzyme tube was determined by the following experiments. We set a flow rate by the microfluidic pump and use the timer calculated the residence time: Start with the first drop of the reaction solution entering the enzyme tube to the first droplet of product flow out of the enzyme tube of microreactor, the void volume is obtained by multiplying the flow rate by the residence time, repeat the experiment three times and take the average value to obtain the void volume of enzyme tube was 628 µL.

Residence time=628 µL /31.4 µL min^-1 =20 min

Entry	Residence time (min)	Flow rate (µL min^-1)	V (µL )
1	15.74	40.0	629.6
2	15.72	40.0	628.8
3	15.66	40.0	626.4
V (average )			628.2

Q9. Effect of reaction media and catalyst. Did the authors perform duplicates of these experiments and (thus provide error intervals), so that one can establish that the yield differences highlighted in Table 1 have statistical significance. Same comment applies to yield data presented in other Figures Could the authors add a column in Table 1 with a parameter such as log P that could easily support the claim that the key parameter highlighted by the processed data is solvent polarity?

A9: Thank you for your question. In order to examine the accuracy the product yields, we repeated each experiment for three times and calculate the average yield. We calculated the standard deviation of three yields and added the standard deviation to each table (Table 1). The data are presented as Average ± SD. Thank you.

Table 1. Effect of reaction solvent and catalysts on the synthesis of β-amino alcohols under continuous-flow reactors^a


Entry	Solvent	Log p	Catalysts	Yield^b (%)
Entry	Solvent	Log p	Catalysts	3	4
1	Methanol	-0.76	None	n.d.	n.d.
2	Methanol	-0.76	Lipozyme TL IM	85.2±1.2	3.1±1.1
3	Ethanol	-0.24	Lipozyme TL IM	80.1±1.6	2.9±1.2
4	Acetonitrile	-0.33	Lipozyme TL IM	78.8±2.1	1.8±1.5
5	Toluene	2.5	Lipozyme TL IM	76.2±0.8	2.3±1.6
6	n-Hexane	3.94	Lipozyme TL IM	71.5±1.5	1.7±1.2
7	Methanol	-0.76	Subtilisin	n.d.	n.d.
8	Ethanol	-0.24	Subtilisin	n.d.	n.d.

^aReaction conditions: continuous-flow reactor, feed 1, 5.0 mmol aniline (1a) was dissolved in 10 mL solvent, feed 2, 5.0 mmol epichlorohydrin (2a) was dissolved in 10 mL solvent, Lipozyme TL IM 870 mg (174 mg/mmol), flow rate 20.9 µL min^-1, residence time 30 min, total reaction time 15.9 h, 40 °C. ^b Isolated yield. Yield: 100 × (actual received quality/ideal calculated quality). The data are presented as average ± SD of triplicate experiments.

Q10. Please rephrase and clarify “When the molar ratio reached 1:1, the highest yield reached 91%, further change in molar ratio decreased the yield of product. Due to the amount of enzyme and the number of catalytically active sites on the enzyme was limited in the fixed-length microreactor”.

A10: Thank you for your suggestions, we have redescribed this sentence (Line 110-114). Since the determined volume of the microchannel reactor and the quality of the enzyme, the active sites of the enzyme involved in the catalysis are certain. As the epoxide increases, the contact between the amine and the active site of the enzyme is reduced, resulting in a decrease in yield.

Q11. Please rephrase and clarify “Besides, we found that the bind of the substrates by the enzyme forming complexs, which more inclined to nucleophilic attack at the less hindered carbon atom of the epoxide ring form the major products 3 (Table 2), results were 126 confirmed by NMR”. Awkward sentencing.

A11: Thank you for your suggestions. In order to clarify the process of the reaction, we analysis the mechanism of lipase catalyzed ring-opening reaction (Figure 1). We have revised the previous description and added it to Supporting Information. Due to the steric effect, amines are inclined to nucleophilic attack at the less hindered carbon atom of the epoxide ring form the major products. We have rephrased the sentence (Line 137-139). Thank you very much.

Figure 1. Mechanism of lipase catalyzed ring-opening reaction of epoxide.

Q12. “Thermomyces Lanuginosus” should be “Thermomyces lanuginosus”

A12: Thank you for your suggestions, we have revised this word (Line 16 and 272).

Finally, thank you very much for your careful review and your valuable comments and suggestions. It is of great help to our work and our future work. Your serious and rigorous attitude towards scientific research is an example for us to learn from and we hope that our work can be accepted by high-level experts like you and high-level professional journals like Catalysts.

Best wishes,

Sincerely yours,

Lihua Du

Author Response File: Author Response.docx

Reviewer 2 Report

In this manuscript Li-Hua and authors have described an efficient method for synthesis of β-amino alcohols in a continuous flow microreactor. Lipozyme TL IM from Thermomyces Lanuginosus was used for the first time in the β-amino alcohol synthesis. The authors have studied and characterized this enzyme by understanding the effects of reaction medium, reaction temperature, substrate ratio, residence time and the structure of substrate. Comment: The authors claim the enzyme reported to be efficient. It will be greatly interesting to the readers, if authors can make a table with different enzymes (natural or synthetic) currently used to produce aminoalchols and compare it with their studied enzyme on the efficiency of amino alcohol production.

Author Response

Dear reviewer:

Q: In this manuscript Li-Hua and authors have described an efficient method for synthesis of β-amino alcohols in a continuous flow microreactor. Lipozyme TL IM from Thermomyces Lanuginosus was used for the first time in the β-amino alcohol synthesis. The authors have studied and characterized this enzyme by understanding the effects of reaction medium, reaction temperature, substrate ratio, residence time and the structure of substrate. Comment: The authors claim the enzyme reported to be efficient. It will be greatly interesting to the readers, if authors can make a table with different enzymes (natural or synthetic) currently used to produce aminoalchols and compare it with their studied enzyme on the efficiency of amino alcohol production.

A: Thank you for your suggestions. According to your suggestion, we have added different enzymes for the production of amino alcohols, and compared them with those of the enzymes studied by us in the production efficiency, temperature, time and solvent (Scheme 1). Thanks again.

Scheme 1. Strategies for the synthesis of β-amino alcohols by ring-opening of epoxides by lipase.

Best wishes,

Sincerely yours,

Lihua Du

Author Response File: Author Response.docx

Reviewer 3 Report

In this communication, Du et al. report the development of an enzymatic continuous-flow approach for the synthesis of β-amino alcohols.

The epoxide taken in consideration on this study (epichlorohydrin) exists in two different chiral forms, or as the corresponding racemic mixture. Consequently, if the authors are starting from the (±)-epichlorohydrin (I am assuming this since the authors point that their starting materials are "low cost"), the reported reaction is also enantioselective and not just regioselective. The authors should state which substrate they are using and if applicable, the enantiomeric excess (ee) of the obtained products.

This is extremely important not only for the accurate description of the work that was done, but also to access the impact and relation to the state of the art. For example, the comparison with other methods it is not fair when product selectivity was not assessed. Other aspect is that many applications (like drug development) are reliable pure chiral compounds. For this reason, I recommend major revisions.

Other comments:

Figure 1, 1^st structure: Cardiovascular misspelled.

Figure 1, 3^rd structure: The represented structure is not chloroquine (also is it misspelled) but hydroxychloroquine.

Line 32: ring opening (ring-opening) is not consistent throughout the text.

Line 42: condiments? Food?

Line 44: consider substitute “enzyme is” for “enzymes are”

Line 71: “We performed a group of blank control trial and found that this reaction did not occur 72 without the enzyme involved” are the reaction conditions and results presented in SI? What was the reactants recovery rate?

Line 74: “It was observed that the reaction could be promoted by using strong polar solvent” This is not entirely true, since n-hexane and toluene yielded 76 and 71 % (table1) and are considered non-polar solvents.

Table 1, caption: suggestion to add the amount of enzyme used in mmol. Easier to the reader see the equivalents of enzyme used in relation to the other reactants.

Line 88: “it is obvious that as the temperature increases, the yield of the product increases steadily” complicated and confusing phrase to say that yield increase from 30 to 35 ºC.

Figure 2, title: the product 3a name “1-chloro-3-(phenylamino)propan-2-ol” is presented for the first time here. Would be easier for the reader to see just “product 3a” or at least “1-chloro-3-(phenylamino)propan-2-ol 3a” as figure title. The technical sheet from Steam Chemical report an optimum temperature between 50-75 ºC (https://www.strem.com/uploads/technical_notes/06-3120tech.pdf green table entry2). Did the authors explore this range of temperature? Also, how long were the reactants left to react and why? This is important to correlate with Figure 3.

Figure 3, title: the same of Figure 2 title.

Figure 4: indicate reaction time.

Line 121: “that the introduction of electron donor can effectively improve the reaction activity of the enzyme” what would be the impact of amines with electron-withdrawing group(s)? Here, only are reported the synthesis with electron-donor amines, which does not represent a “general method”.

Lines 173 and 177: The names of the two first compounds are wrong

Line 262: substitute “the opening reaction” to “the ring-opening reaction”.

Author Response

Dear reviewer:

Q1. In this communication, Du et al. report the development of an enzymatic continuous-flow approach for the synthesis of β-amino alcohols. The epoxide taken in consideration on this study (epichlorohydrin) exists in two different chiral forms, or as the corresponding racemic mixture. Consequently, if the authors are starting from the (±)-epichlorohydrin (I am assuming this since the authors point that their starting materials are "low cost"), the reported reaction is also enantioselective and not just regioselective. The authors should state which substrate they are using and if applicable, the enantiomeric excess (ee) of the obtained products. This is extremely important not only for the accurate description of the work that was done, but also to access the impact and relation to the state of the art. For example, the comparison with other methods it is not fair when product selectivity was not assessed. Other aspect is that many applications (like drug development) are reliable pure chiral compounds. For this reason, I recommend major revisions.

A1: Thank you for your suggestions. We agree that pure chiral compounds are widely used in drug synthesis, and many chemical studies are devoted to the synthesis of enantiomers with high selectivity. In our study, we used the racemic compound (±)-epichlorohydrin (purchased from Aladdin company, product number: E108182). All the β-amino alcohols are racemic mixtures except that the products obtained by ring opening of cyclohexene oxide are pure trans-diastereomers. We have redescribed the stereo configuration of epichlorohydrin, styrene oxide and reaction products in the manuscript (Line 157-160, Scheme 1 in manuscipt). In the future, we will make further efforts to explore the enzymatic synthesis of highly stereoselective compounds and their application in medicine. Thank you very much for your professional advice.

Other comments:

Q2. (1) Figure 1, 1^st structure: Cardiovascular misspelled. (2) Figure 1, 3rd structure: The represented structure is not chloroquine (also is it misspelled) but hydroxychloroquine. (3) Line 32: ring opening (ring-opening) is not consistent throughout the text. (4) Line 44: consider substitute “enzyme is” for “enzymes are”

A2: Thank you for your suggestions, we have revised in the manuscipt. (1) line 37; (2) line 37; (3) line 32; (4) line 44.

Q3. Line 42: condiments? Food?

A3: Thank you for your question. The application of enzymes in food processing has been widely reported. For example, Supeeraya Arsa used Alcalase enzyme preparated hydrolyzed rice bran protein concentrate (HRPC) as a flavoring agent in 2018. (J. Sci. Food Agric., 2018, 98(12), 4479–4487. doi:10.1002/jsfa.8972). We have added relevant reference to the manuscript (Line 42). Thank you.

Q4. Line 71: “We performed a group of blank control trial and found that this reaction did not occur 72 without the enzyme involved” are the reaction conditions and results presented in SI? What was the reactants recovery rate?

A4: Thank you for your suggestions, we have supplemented the reaction conditions and results of the blank control experiment in SI. Through silica gel column chromatography of the collected blank test reaction solution, the recovery rates of reactants are as follows: amine: 94.2%; epichlorohydrin: 96.4%. Thank you.

Q5. Line 74: “It was observed that the reaction could be promoted by using strong polar solvent” This is not entirely true, since n-hexane and toluene yielded 76 and 71 % (table1) and are considered non-polar solvents.

A5: Thank you for your suggestions. Considering that this reaction can also obtain good yields in non-polar solvents such as toluene and n-hexane, so we have revised this sentence in the manuscript. (Line 73-74)

Q6. Table 1, caption: suggestion to add the amount of enzyme used in mmol. Easier to the reader see the equivalents of enzyme used in relation to the other reactants.

A6: We use a commercial immobilized enzyme, which is a mixture of silica gel and lipase, which is usually measured by mass. In order to clarify the quantitative relationship between catalyst and reactant, we added the amount of enzyme used in mg/mmol in table footnotes. (enzyme amount=870mg/5.0 mmol=174 mg/mmol)

Q7. Line 88: “it is obvious that as the temperature increases, the yield of the product increases steadily” complicated and confusing phrase to say that yield increase from 30 to 35 ºC.

A7: Thank you for your suggestions, we have revised this sentence. (Line 88-89)

Q8. Figure 2, title: the product 3a name “1-chloro-3-(phenylamino)propan-2-ol” is presented for the first time here. Would be easier for the reader to see just “product 3a” or at least “1-chloro-3-(phenylamino)propan-2-ol 3a” as figure title. The technical sheet from Steam Chemical report an optimum temperature between 50-75 ºC (https://www.strem.com/uploads/technical_notes/06-3120tech.pdf green table entry2). Did the authors explore this range of temperature? Also, how long were the reactants left to react and why? This is important to correlate with Figure 3. Figure 3, title: the same of Figure 2 title.

A8: Thank you for your question. In order to make the chart more clear, so as not to mislead the reader, we deleted "1-chloro-3-(phenylamino) propan-2-ol" on the top of figure and added the "3a" to the title of the figure. Lipase is commonly used to catalyze reactions such as lipolysis, transesterification, and ester synthesis. Based on the catalytic promiscuity of enzyme, we try to use lipase to catalyze other unconventional reactions, such as ring-opening reaction of epoxides. The optimum conditions of the same enzyme in different chemical reactions and reaction devices are different. According to your suggestion, we further explored the temperature range of 50-75 °C and found that the yield did not increase, the results are shown in Table 1 and Figure 1. A remarkable feature of continuous-flow microreactor is that it can obtain better yield in a shorter residence time. In the past, we have tried Michael addition, regioselective acylation of sugar and other reactions with continuous-flow microreactor, most of them can obtain good yields within a residence time of tens of minutes, so we chose 30 min residence time to initially explore the temperature. The residence time was added to our reaction parameter exploration experiment (Line 89). Thank you for your professional advice.

Table 1 Reaction yield at different temperatures

T (°C )	50	55	60	65	70	75
Yield (%)	68.3±2.3	65±1.7	60.7±3.2	53±1.9	44±2.1	32±1.6

Figure 1 Reaction yield at different temperatures

Q9.; Figure 4:indicate reaction time.

A9 : Thank you for your suggestions, we have added time information in the text (Line 108).

Q10. Line 121: “that the introduction of electron donor can effectively improve the reaction activity of the enzyme” what would be the impact of amines with electron-withdrawing group(s)? Here, only are reported the synthesis with electron-donor amines, which does not represent a “general method”.

A10: Thank you for your question. We have studied the reactivity of different aniline and found that 4-chloroaniline as an electron-withdrawing amine has lower reactive than aniline. The impact of amines with electron-withdrawing group has been added in the manuscript (Line 132-135). Since this reaction is applicable to different amines (electron-withdrawing amines, electron-donor amines, primary amines, secondary amine) and epoxides, the yield reached to 65.0%-93.8%, so we present it as a “general method”. Thank you again.

Q11. Lines 173 and 177: The names of the two first compounds are wrong

A11: Thank you for your suggestions, we have checked the manuscript and Supporting Information and revised these names. (Line 185 and 189)

Q12. Line 262: substitute “the opening reaction” to “the ring-opening reaction”.

A12: Thank you for your suggestions, we have revised this phrase in this manuscript. (Line 276)

Best wishes,

Sincerely yours,

Lihua Du

Author Response File: Author Response.docx

Reviewer 4 Report

The manuscript submitted by Du, Luo, and co-workers describes the use of an immobilised lipase for the synthesis of beta-amino alcohols from aromatic amines and epoxides in a flow system. Many beta-amino alcohols are of interest in medicinal chemistry or as chiral auxiliaries in asymmetric synthesis, and the authors demonstrate a relatively broad scope for their method, which makes it potentially interesting not only for biocatalysis researchers but also for synthetic chemists in general. Nevertheless, I cannot recommend to accept this submission for publication, since the work presented is too shallow (entirely neglecting mechanistic and stereochemical aspects), the experimental setup is not described in sufficient detail, and the manuscript contains several errors and poorly supported statements. Detailed suggestions for improving the manuscript are provided below:

(1) Microfluidics/microreactor: Microreactors are usually defined as flow reactors that have structural features smaller than 1 mm (see, for example: https://doi.org/10.1039/B609428G, https://doi.org/10.1021/op034193x, https://doi.org/10.1002/anie.200300577, https://doi.org/10.1016/j.ces.2018.03.026). With an inner diameter of 2 mm, the setup described here does not qualify as a microreactor.

(2) Stereochemistry: Epichlorohydrin and styrene oxide as well as all reaction products (beta-amino alcohols) are chiral molecules, and the lipase is a chiral catalyst. Therefore, the authors should investigate whether the reaction shows any degree of enantioselectivity, in particular since Gupta et al. (ref. 26) report a high enantioselectivity for similar reactions catalysed by Candida rugosa lipase. Besides, the products obtained by ring opening of cyclohexene oxide are expected to be pure trans-diastereomers, which is supported by the NMR spectra. The relative stereoconfiguration should be indicated in the structural drawings.

(3) Regioselectivity: In Scheme 1 and in the reaction schemes in the table headers the authors show the regioisomeric product 4 along with the major product 3. However, there is no information in the manuscript on whether the formation of product 4 was indeed observed, and if yes, in what relative amount. This is of particular relevance because Borude et al. (ref. 25) and Gupta et al. (ref. 26) report exclusive regioselectivity in related reactions catalysed by other lipases, while ref. 26 also states that the non-catalysed reaction affords a mixture of regioisomers. Therefore, the regioselectivity of the reactions reported in the present manuscript should be quantified.

(4) Lack of mechanistic understanding: No effort is made to understand the mechanism of the reported lipase-catalysed ring-opening reaction, for instance by kinetic analysis, yet the authors invoke mechanistic and/or kinetic reasoning in attempts to explain some of their experimental findings, e.g.: "When the concentration of epichlorohydrin was high, the substrate saturated the active sites of the enzyme, resulting in the decrease of the yield.", "Besides, we found that the bind of the substrates by the enzyme forming complexs, which more inclined to nucleophilic attack at the less hindered carbon atom of the epoxide ring form the major products 3 ..." Without a solid understanding of the enzymatic reaction mechanism, such statements are pure speculation.

(5) Operational stability/reusability: One of the major advantages of flow chemistry is the possibility to "scale up" transformations by extending the time of continuous operation. However, a prerequisite for this is a good operational stability of the flow setup. The authors should investigate for how long their reaction setup can be operated continuously without loss of yield, or - if continuous operation is not possible due to technical restrictions - whether the enzyme-packed reactor can be reused several times in non-continuous flow reactions.

(6) Reaction time, residence time and calculation of yield: Throughout the manuscript, no clear distinction is made between residence time and total time of the reaction. For instance, in section 2.3 the authors speak exclusively of residence time (determining a residence time of 20 min as optimal), but in the experimental section they state that "the resulting stream ... was passed through microchannel reactor, reaction at 35 ºC for 20 minutes...", suggesting that the total reaction time was 20 min. At the indicated flow rate, this would mean that only 628 µL of product solution were collected. If the reaction is stopped before all the feed solution has been consumed, calculation of the yield is obviously dependent on an exact determination of the collected volume of product solution. The authors should clearly distinguish between residence time and total reaction time and report both values for all investigated reactions (e.g., in Table footnotes). Moreover, they should describe precisely how the residence time was determined, how they have quantified the volume of collected product solution, and how they have calculated the yields. The isolated yields should be reported not only as percentages but also as absolute amounts (weight of isolated material after chromatography).

(7) Data presentation: The 3D bar charts make it unnecessarily difficult to read exact values from the graphs. Use standard 2D bar charts instead. Also, remove the legend (1-Chloro-3-(phenylamino)propan-2-ol), since it is not necessary for understanding the graphs.

(8) Choice of references: The choice of references in the introduction is poor. For instance, for the first sentence I would expect references to review articles that summarise the applications of amino alcohols, rather than selected examples of methods for their synthesis (refs. 1-6). For the second sentence, I would either expect references to reviews on the general medicinal use of amino alcohols or references to papers that describe the medicinal use of the four concrete examples shown in Figure 1. Of the references 14-19 for the use of bismuth(III) salts in ring opening of epoxides with amines, four (refs. 14, 15, 18, 19) do not deal with bismuth compounds at all, and one (ref. 17) describes the use of an organobismuth complex, not a bismuth(III) salt. Refs. 20-24 should be supplemented or replaced with review articles covering the respective topics. Ref. 37 is cited as an example of the use of a microreactor, but no microreactor is involved in this study (although a flow setup is being used).

(9) line 49: "The enzymatic reactions are relatively mild, however they always need other auxiliary reagents, or require longer reaction time usually 12 hours or more to achieve the desired yield [27-29]." What enzymatic reactions are the authors referring to? The three examples cited here have no relevance to the work presented in the manuscript.

(10) line 85: "Temperature has an important effect on the stability and catalytic activity of the enzyme [40]." This is a very general finding and I don't see why this particular reference (whose only connection to the present manuscript is that it also uses a lipase) was chosen to support it.

(11) line 106: "In enzymatic reactions, the relative content of substrates is the key to determine the composition of the final product and the efficiency of enzyme catalysis [41]." The cited reference is irrelevant, because the influence of relative concentrations of substrates on enzyme performance is not discussed therein.

(12) Experimental details: What is a "reactant injector", what is a "product collector". The authors should provide more details and/or provide a photograph of the reaction setup in the Supporting Information. What material is the "microchannel reactor" made of? Is it purpose-made or commercially abvailable?

(13) Compound characterisation: The names of products 3a, 3b, 3e, 3f, 3g, 3l, 3m, and 3r are incorrect. Indeed, the reported names seem to have no connection to the actual structures at all. This also applies to the Supporting Information.

(14) SI: Some of the NMR spectra (e.g., the ¹³C spectra of products 3b, 3f, 3g) show additional signals that can't be ascribed to the product. This suggests that the isolated products still contain impurities or residual solvent, leading to inaccuracies in the isolated yields.

Author Response

Dear reviewer:

Q(1) Microfluidics/microreactor: Microreactors are usually defined as flow reactors that have structural features smaller than 1 mm (see, for example: https://doi.org/10.1039/B609428G, https://doi.org/10.1021/op034193x, https://doi.org/10.1002/anie.200300577, https://doi.org/10.1016/j.ces.2018.03.026). With an inner diameter of 2 mm, the setup described here does not qualify as a microreactor.

A1: Thank you for your question. The literature you provided gives us a more comprehensive understanding of the definition of microreactor. We read some literature on flow chemistry and found that some reactors with an inner diameter greater than 1 mm are more accurately described as "flow reactor" (https://pubs.acs.org/doi/full/10.1021/ol502712v; https://pubs.rsc.org/en/content/articlepdf/2018/cs/c7cs00906b; https://pubs.rsc.org/en/content/articlehtml/2019/cy/c8cy02192a?page=search; https://chemistry-europe.onlinelibrary.wiley.com/doi/epdf/10.1002/cctc.201900085). We have revised “microreactors” to “flow reactor” in the manuscript. Thank you again for your advice. (Line 16, 21, 58, 61, 68, 79, 80, 93, 102, 104, 113, 117, 128, 141, 143, 147, 149, 159, 164, 165, 168, 171, 176, 272, 281)

Q(2) Stereochemistry: Epichlorohydrin and styrene oxide as well as all reaction products (beta-amino alcohols) are chiral molecules, and the lipase is a chiral catalyst. Therefore, the authors should investigate whether the reaction shows any degree of enantioselectivity, in particular since Gupta et al. (ref. 26) report a high enantioselectivity for similar reactions catalysed by Candida rugosa lipase. Besides, the products obtained by ring opening of cyclohexene oxide are expected to be pure trans-diastereomers, which is supported by the NMR spectra. The relative stereoconfiguration should be indicated in the structural drawings.

A2: Thank you for your suggestions. We quite agree with you that enzymes are widely used as chiral catalysts for the selective synthesis of chiral molecules. We used the racemic compound (±)-epichlorohydrin (purchased from Aladdin company, product number: E108182), the racemic compound (±)-styrene oxide (purchased from Aladdin company, product number: S165282). (Line 157-160) In our study, all the β-amino alcohols are racemic mixtures except that the products obtained by ring opening of cyclohexene oxide are pure trans-diastereomers. We have supplemented the stereo configuration of epichlorohydrin, styrene oxide and reaction products in the manuscript (Table 2 in the manuscript). In the future, we will make further efforts to explore the enzymatic synthesis of highly stereoselective compounds and their application in medicine. Thank you very much for your professional advice.

Q(3) Regioselectivity: In Scheme 1 and in the reaction schemes in the table headers the authors show the regioisomeric product 4 along with the major product 3. However, there is no information in the manuscript on whether the formation of product 4 was indeed observed, and if yes, in what relative amount. This is of particular relevance because Borude et al. (ref. 25) and Gupta et al. (ref. 26) report exclusive regioselectivity in related reactions catalysed by other lipases, while ref. 26 also states that the non-catalysed reaction affords a mixture of regioisomers. Therefore, the regioselectivity of the reactions reported in the present manuscript should be quantified.

A3: Thank you for your suggestions. Indeed, we agree with you very much. For a chemical reaction, in addition to observing the main product, it is also necessary to quantify the secondary product. In our study, due to the two reaction selection, in addition to the main product 3, we also observed a small amount of secondary product 4. The specific quantitative data of 4 have been supplemented in the manuscript (Table 1 and Table 2 in manuscript). Thank you very much.

Q(4) Lack of mechanistic understanding: No effort is made to understand the mechanism of the reported lipase-catalysed ring-opening reaction, for instance by kinetic analysis, yet the authors invoke mechanistic and/or kinetic reasoning in attempts to explain some of their experimental findings, e.g.: "When the concentration of epichlorohydrin was high, the substrate saturated the active sites of the enzyme, resulting in the decrease of the yield.", "Besides, we found that the bind of the substrates by the enzyme forming complexs, which more inclined to nucleophilic attack at the less hindered carbon atom of the epoxide ring form the major products 3 ..." Without a solid understanding of the enzymatic reaction mechanism, such statements are pure speculation.

A4: Thank you for your suggestions. In order to clarify the process of lipase catalyzed ring opening reaction, the mechanism of lipase catalyzed ring opening reaction was supplemented and explained (Figure 1). Through the analysis of the reaction mechanism, we have a clearer understanding of the previous description. Since the determined volume of the flow reactor and the quality of the enzyme, the active sites of the enzyme involved in the catalysis are certain. As the epoxide increases, the contact between the amine and the active site of the enzyme is reduced, resulting in a decrease in yield. Due to the steric effect, amines are inclined to nucleophilic attack at the less hindered carbon atom of the epoxide ring form the major products. We have rephrased these sentences in manuscript (Line 137-139). Thank you very much.

Figure 1. Mechanism of lipase catalyzed ring-opening reaction of epoxide.

A5: Thank you for your advice. We quite agree that the remarkable advantage of flow chemistry is that it can amplify chemical reactions. In the flow reactor, the product can be continuously synthesized by extending the time of continuous operation. In addition, the process can be enlarged by paralleling several reaction pipelines. So we tried to synthesize β-amino alcohols by using continuous flow reactor. Operational stability and reusability are essential for flow chemistry. In order to explore the operational reusability of the flow reactor, we select the ring-opening reaction of aniline and epichlorohydrin as the template reaction, and repeated the experiment 10 times on the same enzyme-packed reactor, and the results were shown in Figure 2 (Figure 5 in manuscript). Increased the number of cycles is accompanied by a progressive decrease in yields, 43% yield was obtained during the tenth cycle. The result indicated that the lipozyme TL IM has sufficient reusability. Thank you very much.

Figure 2. The influence of enzymes reusability on the enzymatic synthesis of β-amino alcohol 3a in reactors.

A6: Thank you for your question. We control the residence time by setting different flow rates. Catalysts was filled into a flow reactor made of a PFA reactor coil (inner diameter ID = 2.0 mm, length = 100 cm). When the flow reactor is full with catalysts, the void volume of the enzyme-packed reactor is about 628 µL.

Residence time=628 µL/31.4 μL min⁻¹=20 min

All the feed solution has been consumed and finally collected (≈20 mL), the product was purified by silica gel flash chromatography. Yield=100 × (actual received quality/ ideal calculated quality). The data are presented as average ± SD of triplicate experiments. The residence time was expressed uniformly and the total reaction time was added. The quality of the isolated material after chromatography was reported in the manuscript and Supporting Information. Thank you again. (Table 1 and Table 2 in the manuscript)

Q(7) Data presentation: The 3D bar charts make it unnecessarily difficult to read exact values from the graphs. Use standard 2D bar charts instead. Also, remove the legend (1-Chloro-3-(phenylamino)propan-2-ol), since it is not necessary for understanding the graphs.

A7: Thank you for your suggestions. We have replaced the 3D bar chart in our manuscript with 2D bar chart and removed the legend (Figure 1-6 in the manuscript). Thank you again.

Q(8) Choice of references: The choice of references in the introduction is poor. For instance, for the first sentence I would expect references to review articles that summarise the applications of amino alcohols, rather than selected examples of methods for their synthesis (refs. 1-6). For the second sentence, I would either expect references to reviews on the general medicinal use of amino alcohols or references to papers that describe the medicinal use of the four concrete examples shown in Figure 1. Of the references 14-19 for the use of bismuth(III) salts in ring opening of epoxides with amines, four (refs. 14, 15, 18, 19) do not deal with bismuth compounds at all, and one (ref. 17) describes the use of an organobismuth complex, not a bismuth(III) salt. Refs. 20-24 should be supplemented or replaced with review articles covering the respective topics. Ref. 37 is cited as an example of the use of a microreactor, but no microreactor is involved in this study (although a flow setup is being used).

A8: Thank you for your suggestions. We have supplemented and replaced the references. Thank you.

Q(9) line 49: "The enzymatic reactions are relatively mild, however they always need other auxiliary reagents, or require longer reaction time usually 12 hours or more to achieve the desired yield [27-29]." What enzymatic reactions are the authors referring to? The three examples cited here have no relevance to the work presented in the manuscript.

A9: Thank you for your suggestions. The “enzymatic reaction” refers to the “enzymatic synthesis of β-amino alcohol”, this information was added in the manuscript. We have deleted the irrelevant literature [27-29] and the different strategies of enzymatic synthesis of β-amino alcohol were showed in Scheme 1. Thank you very much. (Line 50-53)

Q(10) line 85: "Temperature has an important effect on the stability and catalytic activity of the enzyme [40]." This is a very general finding and I don't see why this particular reference (whose only connection to the present manuscript is that it also uses a lipase) was chosen to support it.

A10: Thank you for your suggestions. Temperature is an important factor affecting enzymatic reaction. We have replaced reference [40] by consulting the literature (ref. 38 in the manuscript, line 86). Thank you very much.

A11: Thank you for your suggestions. We have replaced reference [41] by consulting the literature (ref. 39 in the manuscript, line 107). Thank you very much.

Q(12) Experimental details: What is a "reactant injector", what is a "product collector". The authors should provide more details and/or provide a photograph of the reaction setup in the Supporting Information. What material is the "microchannel reactor" made of? Is it purpose-made or commercially abvailable?

A12: Thank you for your suggestions. Flow reactor and Y-mixer were made of perfluoroalkoxy (PFA) and purchased from Beijing Haigri Medical Engineering Design Co., Ltd. The flow reactor was made of coiled PFA tube. We have explained the materials, composition and commercial sources of the reactor and Y-mixer in the manuscript (Line 157-160). The details of the reaction setup are shown in Figure 3 and 4, we also add them to Supporting Information. Thank you.

Figure 3 Syringe pump and reactant injector.

Figure 4 Photograph of the continuous flow system.

Q(13) Compound characterisation: The names of products 3a, 3b, 3e, 3f, 3g, 3l, 3m, and 3r are incorrect. Indeed, the reported names seem to have no connection to the actual structures at all. This also applies to the Supporting Information.

A13: Thank you for reminding. We have revised the wrong name in the manuscript (Line 185, 189, 202, 207, 211, 236, 241, 265). Thank you very much.

Q(14) SI: Some of the NMR spectra (e.g., the 13C spectra of products 3b, 3f, 3g) show additional signals that can't be ascribed to the product. This suggests that the isolated products still contain impurities or residual solvent, leading to inaccuracies in the isolated yields.

A14: Thank you for your suggestions. Since β-amino alcohol are easy to metamorphose under high temperature conditions, so we separate and purify the products under low temperature conditions (≤ 45 °C)，which result in a small amount of solvent remained in the product. In the future, we will improve the treatment process of the reaction liquid to ensure the purity of the products as far as possible. If necessary, these compounds will be synthesized again and characterized by NMR. Thank you very much for your advice.

Best wishes,

Sincerely yours,

Lihua Du

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

Supplementary material

Has the reaction mechanism presented in Figure 1 been established previously? Was this carried out in the present work? The authors should provide references in the literature for the former, or detailed experimental evidence for the later case

There are still several English glitches, thus the authors are advised to have the manuscript double checked by a proficient English speaker. Some examples are nevertheless given

“The enzymatic synthesis of β-amino alcohols reactions are relatively mild, however they need other auxiliary reagents or longer reaction time to achieve the desired yield compared to chemical catalysis (Scheme1)” please rephrase

“influencing factors which affect” should be “influencing factors that affect”

Title of Table 1 continuous-flow conditions would be more adequate than continuous-flow reactors. A similar comment applies to the captions of some figures, e.g. 2 and 3 where “reactors” could be replaced by “continuous-flow conditions”

“Yield: 100 × (actual received quality/ideal calculated quality).”, “quantity” should be used rather than “quality”

Lines 112 to 114, phrasing is a bit awkward. Moreover, the authors should further doublecheck that saturation of the active sites of the enzyme was the reason underlying the decrease in yield, or if this could result from other factors

(lines 124 to 126, new text added) the authors justify the observed decrease in activity of the enzyme formulation with the distortion of the active site due to frequent encounters with the substrate. One of the papers referenced to support this claim (ref 40) assigns changes in the active site to interactions with the enzyme carrier, which is common, rather than to interaction with the substrate. However, reference 41 actually claims that “The decreases in enzymatic activity during repeated use can be related to repetitive encountering of substrate to the active site of immobilized enzyme” and later that “Furthermore, the recurrent encountering of substrate with the active site of immobilized enzyme causes its distortion and leads to loss of activity.” To this reviewer knowledge, this paper quoted as reference 41 is the first that claims that loss of enzyme activity can be due to regular catalytic activity, and the authors of said paper provided no evidence to support such claim. Exception for suicide substrates but this is not the case. Regarding distortion of the active site, one must bear in mind that the induced fit model commonly used evidences the need for some distortion for catalysis actually happen. One may notice extreme distortion leading to inactivation due to heat, or organic solvents for instance. It is more likely that enzyme leakage may have occurred, or other phenomena, e.g fouling , to justify the decrease while reusing the catalyst formulation

Author Response

Dear reviewer:

Supplementary material

Q1. Has the reaction mechanism presented in Figure 1 been established previously? Was this carried out in the present work? The authors should provide references in the literature for the former, or detailed experimental evidence for the later case

A1: Thank you for your question. Actually, the reaction mechanism shown in Figure 1 was not established previously. Based on the mechanism of enzymatic reaction in the reference (https://doi.org/10.5267/j.ccl.2012.10.002), we speculated on the mechanism of enzymatic ring-opening reaction. In order to make readers more clear, we have added the reference to Figure 1 and changed the title to “proposed mechanism”. Thank you for your professional advice.

There are still several English glitches, thus the authors are advised to have the manuscript double checked by a proficient English speaker. Some examples are nevertheless given.

Q2.“The enzymatic synthesis of β-amino alcohols reactions are relatively mild, however they need other auxiliary reagents or longer reaction time to achieve the desired yield compared to chemical catalysis (Scheme1)” please rephrase.

A2: Thank you for your suggestion. We have rephrased this sentence in the manuscript. (Line 50-51), thank you again.

Q3.(1) “influencing factors which affect” should be “influencing factors that affect”; (2) Title of Table 1 continuous-flow conditions would be more adequate than continuous-flow reactors. A similar comment applies to the captions of some figures, e.g. 2 and 3 where “reactors” could be replaced by “continuous-flow conditions”; (3) “Yield: 100 × (actual received quality/ideal calculated quality).”, “quantity” should be used rather than “quality”

A3: Thank you for your suggestions. We have rechecked the manuscript and corrected the above errors. (1) line 69; (2) line 78, 93, 105, 117, 126, 140, 178;(3) line 82, 151. Thank you.

Q4. Lines 112 to 114, phrasing is a bit awkward. Moreover, the authors should further doublecheck that saturation of the active sites of the enzyme was the reason underlying the decrease in yield, or if this could result from other factors

A4: Thank you for your suggestions. Our statements are probably lacking some of suppport from detailed kinetic studies evidence. In order to avoid speculative statements, we have removed this part from the manuscript. Thank you again for your professional advice.

Q5. (lines 124 to 126, new text added) the authors justify the observed decrease in activity of the enzyme formulation with the distortion of the active site due to frequent encounters with the substrate. One of the papers referenced to support this claim (ref 40) assigns changes in the active site to interactions with the enzyme carrier, which is common, rather than to interaction with the substrate. However, reference 41 actually claims that “The decreases in enzymatic activity during repeated use can be related to repetitive encountering of substrate to the active site of immobilized enzyme” and later that “Furthermore, the recurrent encountering of substrate with the active site of immobilized enzyme causes its distortion and leads to loss of activity.” To this reviewer knowledge, this paper quoted as reference 41 is the first that claims that loss of enzyme activity can be due to regular catalytic activity, and the authors of said paper provided no evidence to support such claim. Exception for suicide substrates but this is not the case. Regarding distortion of the active site, one must bear in mind that the induced fit model commonly used evidences the need for some distortion for catalysis actually happen. One may notice extreme distortion leading to inactivation due to heat, or organic solvents for instance. It is more likely that enzyme leakage may have occurred, or other phenomena, e.g fouling , to justify the decrease while reusing the catalyst formulation

A5: Thank you for your question. In the process of reuse, the reduction of yields are affected by many factors. Heat and solvent can distort and inactivate the enzyme, in addition, enzyme leakage may have occurred, or other phenomena, e.g fouling, while reusing the catalyst formulation. Our explanation based on reference 41 that “The decreases in enzymatic activity during repeated use can be related to repetitive encountering of substrate to the active site of immobilized enzyme” lacks experimental confirmation, it may be a combination of many factors. We have removed this part from the manuscript. Thank you again for your professional advice.

Best wishes,

Sincerely yours,

Lihua Du

Author Response File: Author Response.docx

Reviewer 3 Report

My major concern was not addressed. In the new scheme 1, the authors now state that compound 3 exists in two different enantiomeric forms, but nothing is said or analysed in the text with regards to the enantiomeric excess. If they do get both enantiomers (R and S), then they should report the ee for their synthetic process.
In the response, they state that they obtain compounds 3 and 4 and that those are diastereomers, but they are structural isomers. Each of these compounds has two stereoisomers, but this fact is basically neglected through the text. This is a very fundamental aspect that should be considered for the article to be suitable for publication.

Author Response

Dear reviewer:

My major concern was not addressed. In the new scheme 1, the authors now state that compound 3 exists in two different enantiomeric forms, but nothing is said or analysed in the text with regards to the enantiomeric excess. If they do get both enantiomers (R and S), then they should report the ee for their synthetic process.
In the response, they state that they obtain compounds 3 and 4 and that those are diastereomers, but they are structural isomers. Each of these compounds has two stereoisomers, but this fact is basically neglected through the text. This is a very fundamental aspect that should be considered for the article to be suitable for publication.

A: Thank you for your suggestion. Indeed, the ee is essential for the compounds containing two enantiomers forms. In order to determine the relative amount of the two enantiomers, we measured % ee with chiral HPLC, and measurements showed that the products are racemic mixtures. We have reported % ee in our manuscript (Table 2), some HPLC spectra were added to Supporting Information. Thanks again for the professional advice.

Best wishes,

Sincerely yours,

Lihua Du

Author Response File: Author Response.docx

Reviewer 4 Report

In the revised version of their manuscript, the authors have made a good effort to improve their article by incorporating the suggestions of the reviewers. In particular, the quality of data presentation and the accuracy of the experimental details have been significantly improved. There are, however, still some critical shortcomings that have not been resolved, and further revision is therefore required. Detailed suggestions for improving the manuscript further are provided below.

(1) Figure 1: I forgot to mention this in the original reviewer report: The (R)-enantiomers of Naftopidil and Salbutamol are shown in the figure, but both compounds are marketed as racemates. I think it would be more appropriate to show the racemic compounds in the figure.

(2) line 40: Of the new references 19-21, two (20, 21) are of limited relevance. I suggest to cite more general reviews, particularly those that focus on the sustainability aspect of biocatalysis. Personally, I can recommend: https://doi.org/10.1021/acs.chemrev.7b00203, https://doi.org/10.1002/cssc.201900351, https://doi.org/10.1016/j.cogsc.2019.08.006

(3) line 78: "under continuous-flow reactors" should be changed to "under continuous-flow conditions".

(4) line 83: In this and all following table footnotes, "quality" should be replaced with "quantity".

(5) line 93: Here and in the following Figure captions, "in reactors" should be changed to "under continuous-flow conditions".

(6) lines 112-114: "Due to the amount of enzyme and the number of catalytically active sites on the enzyme was limited in the flow reactor, when the concentration of epichlorohydrin was high, the active sites of the enzyme were saturated resulted in lower yields." As pointed out in my original report, this statement is pure speculation. The authors need to understand that the statement "the active sites of the enzyme were saturated" has a very specific meaning in the context of Michaelis-Menten kinetics, the kinetic model most commonly used to describe enzymatic reactions, and that this statement cannot be made without support from experimental evidence in the form of detailed kinetic studies. It is a common phenomenon that enzymatic activity decreases at high substrate concentrations, and there are several possible reasons for this, including substrate inhibition, enzyme inactivation at a liquid-liquid interface (if the substrate has low aqueous solubility), etc. In many cases, such a decrease in activity occurs at substrate concentrations that are well below the level needed for the enzyme to be saturated with substrate.
To summarise, I urge the authors to avoid speculative statements such as this one. On the basis of the presented data, it is simply impossible to identify the reason for the observed optimal substrate ratio of 1:1.

(7) lines 124-126: "Moreover, the frequent interaction of the substrate with the active site of the enzyme causes distortion of the active site, leading to reduction of activity [40-41]." This sentence, which is taken in almost identical form from ref. 41, is naive and speculative. Enzymes "distort" all the time, as they are flexible molecules and minor structural fluctuations are part of their catalytic cycle. This leads to inactivation only when the structural changes are so severe that the active conformation of the enzyme is permanently lost. Such unfolding can be triggered by heat, mechanical stress, high concentrations of cosolvents, ionic compounds, etc., but to the best of my knowledge there are no data showing that regular catalytic turnover leads to deactivation.

(8) Table 2: The 0 values in the yield (4) column of entries 13–18 should be replaced with "n.a." (not applicable), since no regioisomeric product is possible in these reactions.

(9) SI: The reaction mechanism that the authors have added is purely speculative. Identifying the reaction mechanism of an enzymatic transformation is a difficult task that often requires a combination of protein crystal structure analysis, site-directed mutagenesis, kinetic investigations, and computational chemistry. On the basis of the data presented in this manuscript, no mechanistic conclusions can be drawn. Indeed, it is not even certain whether the reaction takes place in the enzyme's active site or perhaps on the enzyme surface. If the obtained products are indeed racemic, as the authors claim, this would actually hint at a reaction outside of the active site.
I suggest to remove the mechanism if it cannot be supported with evidence, or, at the very least, to label it as a "proposed mechanism" to make its speculative nature clear to the readers.

(10) General comment: In response to the original reviewer reports, the authors state that "all the β-amino alcohols are racemic mixtures", but they do not present any evidence for this. Of course, the high yields suggest that both enantiomers of the starting material are consumed at comparable rates, but still the enantiomeric excess of the products needs to be analysed. The simplest option for this would be to determine the optical rotation of the products, but a more reliable approach would be the synthesis of racemic reference material by an independent method, followed by analysis of the reference material and the obtained product using chiral stationary phase GC or HPLC. Such an analysis should be performed for at least one product, but preferably for all of them.

(11) General comment: The tables have been much improved by giving the yields as ranges (mean and standard deviation) of three independent experiments. If the data in the figures are also based on triplicate experiments, error ranges should be shown in the bar charts to indicate the range of obtained results.

Author Response

Dear reviewer:

Q(1) Figure 1: I forgot to mention this in the original reviewer report: The (R)-enantiomers of Naftopidil and Salbutamol are shown in the figure, but both compounds are marketed as racemates. I think it would be more appropriate to show the racemic compounds in the figure.

A1: Thank you for your suggestion. We have revised in the manuscript. (Figure 1 in the manuscript). Thanks again.

Q(2) line 40: Of the new references 19-21, two (20, 21) are of limited relevance. I suggest to cite more general reviews, particularly those that focus on the sustainability aspect of biocatalysis. Personally, I can recommend: https://doi.org/10.1021/acs.chemrev.7b00203, https://doi.org/10.1002/cssc.201900351, https://doi.org/10.1016/j.cogsc.2019.08.006.

A2: Thank you very much for your suggestion. We carefully read the literature you provided, which is tremendously helpful to us. We have supplemented and replaced the references in the manuscript. Thank you again. (Line 40)

Q(3) line 78: "under continuous-flow reactors" should be changed to "under continuous-flow conditions"; line 83: In this and all following table footnotes, "quality" should be replaced with "quantity"; line 93: Here and in the following Figure captions, "in reactors" should be changed to "under continuous-flow conditions".

A3: Thank you for your suggestions. We have rechecked the manuscript and corrected the above errors. Thank you. (Line 78; line 82 and 151; line 93, 105, 117, 126, 140, 178)

Q(4) lines 112-114: "Due to the amount of enzyme and the number of catalytically active sites on the enzyme was limited in the flow reactor, when the concentration of epichlorohydrin was high, the active sites of the enzyme were saturated resulted in lower yields." As pointed out in my original report, this statement is pure speculation. The authors need to understand that the statement "the active sites of the enzyme were saturated" has a very specific meaning in the context of Michaelis-Menten kinetics, the kinetic model most commonly used to describe enzymatic reactions, and that this statement cannot be made without support from experimental evidence in the form of detailed kinetic studies. It is a common phenomenon that enzymatic activity decreases at high substrate concentrations, and there are several possible reasons for this, including substrate inhibition, enzyme inactivation at a liquid-liquid interface (if the substrate has low aqueous solubility), etc. In many cases, such a decrease in activity occurs at substrate concentrations that are well below the level needed for the enzyme to be saturated with substrate.
To summarise, I urge the authors to avoid speculative statements such as this one. On the basis of the presented data, it is simply impossible to identify the reason for the observed optimal substrate ratio of 1:1.

Q(5) lines 124-126: "Moreover, the frequent interaction of the substrate with the active site of the enzyme causes distortion of the active site, leading to reduction of activity [40-41]." This sentence, which is taken in almost identical form from ref. 41, is naive and speculative. Enzymes "distort" all the time, as they are flexible molecules and minor structural fluctuations are part of their catalytic cycle. This leads to inactivation only when the structural changes are so severe that the active conformation of the enzyme is permanently lost. Such unfolding can be triggered by heat, mechanical stress, high concentrations of cosolvents, ionic compounds, etc., but to the best of my knowledge there are no data showing that regular catalytic turnover leads to deactivation.

A5: Thank you for your suggestions. In the process of reuse, the reduction of yields are affected by many factors. The change of enzyme structure was affected by heat, mechanical stress, high concentrations of cosolvents, ionic compounds, etc. Our explanation based on reference 41 that “The decreases in enzymatic activity during repeated use can be related to repetitive encountering of substrate to the active site of immobilized enzyme” lacks experimental confirmation, it may be a combination of many factors. We have removed this part from the manuscript. Thank you again for your professional advice.

Q(6) Table 2: The 0 values in the yield (4) column of entries 13–18 should be replaced with "n.a." (not applicable), since no regioisomeric product is possible in these reactions.

A6: Thank you for your suggestions. We have replaced in the manuscript (Table 2). Thank you again.

Q(7) SI: The reaction mechanism that the authors have added is purely speculative. Identifying the reaction mechanism of an enzymatic transformation is a difficult task that often requires a combination of protein crystal structure analysis, site-directed mutagenesis, kinetic investigations, and computational chemistry. On the basis of the data presented in this manuscript, no mechanistic conclusions can be drawn. Indeed, it is not even certain whether the reaction takes place in the enzyme's active site or perhaps on the enzyme surface. If the obtained products are indeed racemic, as the authors claim, this would actually hint at a reaction outside of the active site.
I suggest to remove the mechanism if it cannot be supported with evidence, or, at the very least, to label it as a "proposed mechanism" to make its speculative nature clear to the readers.

A7: Thank you for your suggestions. We have labeled the reaction mechanism as "proposed mechanism". Thank you again.

Q(8) General comment: In response to the original reviewer reports, the authors state that "all the β-amino alcohols are racemic mixtures", but they do not present any evidence for this. Of course, the high yields suggest that both enantiomers of the starting material are consumed at comparable rates, but still the enantiomeric excess of the products needs to be analysed. The simplest option for this would be to determine the optical rotation of the products, but a more reliable approach would be the synthesis of racemic reference material by an independent method, followed by analysis of the reference material and the obtained product using chiral stationary phase GC or HPLC. Such an analysis should be performed for at least one product, but preferably for all of them.

A8: Thank you for your suggestions. In order to determine the relative amount of the two enantiomers, we used chiral HPLC column to analyzed the enantiomeric excess of the products, and found they are approximately equal (ee≤1%, Table 2 in our manuscript). Some HPLC spectra were added to Supporting Information. Thanks again for the professional advice.

Q(9) General comment: The tables have been much improved by giving the yields as ranges (mean and standard deviation) of three independent experiments. If the data in the figures are also based on triplicate experiments, error ranges should be shown in the bar charts to indicate the range of obtained results.

A9: Thank you for your suggestions. In order to shown the error range more accurately, we have added error bars to all figures (Figure 2-6). Thank you again.

Best wishes,

Sincerely yours,

Lihua Du

Author Response File: Author Response.docx