Enhanced Biohydrogen Production Through Continuous Fermentation of Thermotoga neapolitana: Addressing By-Product Inhibition and Cell Viability in Different Bioreactor Modes

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The paper entitled “Enhanced Biohydrogen Production through Continuous Fer- mentation of Thermotoga neapolitana: Addressing By-Product Inhibition and Cell Viability in different Bioreactor Modes” evaluated the efficient biogenic production of hydrogen via the thermophilic bacterium Thermotoga neapolitana, focusing on optimizing process configurations to maximize yield and productivity. There are several suggestions as follows:

1. Condense the introduction section, too many words in this section making it tedious. In addition, please rearrange the formation based on the Journal rather than a thesis.
2. The title mentioned that this paper would forces on the addressing by-product inhibition and cell viable in different bioreactor mode, while no information was given in the last paragraph of the introduction, only the aim of differences between batch, fed-batch and continuous operation under otherwise identical conditions was mentioned.
3. Improve the quality of the figures.
4. How did you decide which mode was best? What’s your basis? Did you systematically compare these three modes?
5. Why did the carbon balance carry out only for continuous mode?
6. Please improve the English.
7. The conclusion section was too long, please give the most important conclusions.
8. Update the referents with the newest papers.

Author Response

Dear reviewer, thank you very much for taking time to review this manuscript and for providing us with this very helpful feedback on our work. Below are our responses to each of your comments. We have done our best to address the comments to your convenience.

Comment 1: Condense the introduction section, too many words in this section making it tedious. In addition, please rearrange the formation based on the Journal rather than a thesis.

Answer 1: Thank you for your feedback. We have tried to shorten and omit unnecessary descriptions so that the relevant information is more prominent. Overall, the introduction section has also been combined into a single subchapter, resulting in a coherent text that provides the reader with all the necessary information for the following sections.

Comment 2: The title mentioned that this paper would forces on the addressing by-product inhibition and cell viable in different bioreactor mode, while no information was given in the last paragraph of the introduction, only the aim of differences between batch, fed-batch and continuous operation under otherwise identical conditions was mentioned.

Answer 2: Thanks, this topic deserves more attention. When adjusting the introduction, we referred to the viability of the cells, taking care not to make the introduction unnecessarily long. There are indications in the literature that bacteria inevitably enter a stationary phase over time and the number of CFUs decreases, even though sufficient nutrients are available and by-products are not responsible for this. There are also papers that describe in detail the inhibitory effects of increased by-product formation.

Comment 3: Improve the quality of the figures.

Answer 3: Thank you for your comment. We always strive to use well-designed, high-resolution graphics to present our research results. To meet this requirement, the images were re-uploaded in higher DPI after being adjusted.

Comment 4: How did you decide which mode was best? What’s your basis? Did you systematically compare these three modes?

Answer 4: We took various factors into account when assessing which mode was considered particularly suitable. As mentioned in the conclusion, the H₂ production rate is higher in continuous reactor operation compared to batch and fed-batch operation. In addition, the H₂ production rate can be kept constant over the steady state period, whereas the batch and fed-batch data only refer to a short period of maximum production rate, while before and after this period the rate is sometimes significantly lower. Therefore, compared to continuous fermentation, the substrate can be converted to H₂ more efficiently in terms of time. Particularly with regard to possible pilot plants, which are intended to demonstrate that Thermotoga neapolitana can reliably produce hydrogen over a longer period of time, continuous reactor operation offers the highest productivity and robustness. We have also described this in the text.

 

Comment 5: Why did the carbon balance carry out only for continuous mode?

Answer 5: We also discussed this in advance and made a conscious decision to only prepare the carbon balance for continuous mode. This is due to several advantages. There was a constant supply and removal of medium, resulting in a constant supply of substrate. In steady state in the continuous reactor, the glucose consumption rate was constant, as were the biomass formation rate and the production rate of all analysed products. This meant that a sample from the feed and the effluent allowed clear conclusions to be drawn about what was currently happening in the reactor. In batch and fed-batch processes, on the other hand, conditions inside the reactor are constantly changing, which is why growth and production rates vary continuously. This means that comparability with the continuous process is limited. Another important factor is CO₂ production. Since fresh medium was continuously added and the used medium was removed in the continuous process, CO₂ formation had little effect on the pH value of the medium. In batch and fed-batch processes, however, the production of organic acids and CO₂ causes the pH value to drop sharply despite the buffered system, and the described regulation with NaOH was necessary to maintain a constant pH value. However, since the addition of NaOH is also known as a CO₂ capturing method because sodium carbonate is produced, which cannot be recorded in the balance, distortions occur.

 

Comment 6: Please improve the English.

Answer 6: We have revised the manuscript in this regard and have attempted to elevate the wording in several places to a higher level in order to meet scientific standards.

Comment 7: The conclusion section was too long, please give the most important conclusions.

Answer 7: We strive to present readers with all the important findings that can be derived from this work, and we have apparently been a little too detailed in the conclusion. As a result, we have now revised the conclusion so that only the most important conclusions are listed there, so that it does not appear too long.

Comment 8: Update the referents with the newest papers.

Answer 8: Thank you for pointing this out. We have rewritten the references section, added further sources and checked the existing sources.

 

 

Reviewer 2 Report

Comments and Suggestions for Authors

This manuscript compares batch, fed-batch, and continuous operation for biohydrogen production by Thermotoga neapolitana and shows stable, high volumetric productivities in chemostat mode. The study is technically solid and relevant for Fermentation.

Comments:

1. Sharpen the Abstract with quantitative scope. Please state the explicit objective in one sentence and add key numbers: steady-state duration (≥56 h per D), best D (=0.07 h⁻¹), H₂ rate (96.1 ± 1.7 Nml·L⁻¹·h⁻¹), and yield range (2.7–3.0 mol H₂·mol⁻¹ glucose). This is already in the main text; mirror it in the abstract.
2. Gas analytics & “standard volume” definition. Specify STP conditions used for “Nml”, calibration gases/protocol for BlueVary/BlueVCount. Report N₂ sparge rate and headspace pressure to support gas–liquid mass transfer claims.
3. Viability assay robustness. DiBAC₄ primarily marks depolarised membranes; please report counting method (fields, cells counted) and replicates. Include error bars on Figure 3.
4. Fed-batch inhibition analysis. You discuss acetate/ionic-strength effects; quantify base addition (NaOH consumption) and, if possible, conductivity to support the ionic-strength argument. Consider a simple electron/redox balance (including ethanol) to explain the H₂ decline, please.
5. Carbon balance uncertainty. The balance closes at ~103%. Provide an uncertainty estimate and clarify contributions from yeast extract, please.
6. Language polish. Fix minor wording/typos (“gas stream”, not “steam”), unify decimal separators (comma vs dot) and define units consistently, please.
7. The work is strong on reactor-level comparison and viability; adding small pieces of quantitative mass-transfer/statistics would make it exemplary.

Author Response

Dear reviewer, thank you very much for taking time to review this manuscript and for providing us with this very helpful feedback on our work. Below are our responses to each of your comments. We have done our best to address the comments to your convenience.

Comment 1: Sharpen the Abstract with quantitative scope. Please state the explicit objective in one sentence and add key numbers: steady-state duration (≥56 h per D), best D (=0.07 h⁻¹), H₂ rate (96.1 ± 1.7 Nml·L⁻¹·h⁻¹), and yield range (2.7–3.0 mol H₂·mol⁻¹ glucose). This is already in the main text; mirror it in the abstract.

Answer 1: Thank you very much for your feedback on the abstract. I agree with you that, as the highlight of the publication, it should provide a concise overview of the most important results of the work and be substantiated with corresponding data. We have adapted the abstract and tried to limit and focus on the essentials, thereby providing the reader with a clear insight into the most important results of this work.

Comment 2: Gas analytics & “standard volume” definition. Specify STP conditions used for “Nml”, calibration gases/protocol for BlueVary/BlueVCount. Report N₂ sparge rate and headspace pressure to support gas–liquid mass transfer claims.

Answer 2: Thank you for your comment. It is, of course, important to record the exact details of how the gas analysis and calibration were carried out. In this work, we deliberately chose to use the standard volume unit (Nml) to enable comparability. This specifies the volume of the gas at 273.15 K and 1 atm (equal to 101325 Pa). Since the volume of gases depends heavily on pressure and temperature, this is particularly important for volumetric data. We have added this information to the Materials and Methods section, along with the N₂ sparging rate during fermentation. BlueSens sensors are delivered calibrated, and according to the manufacturer, the BlueVCount does not require further calibration. The BlueVary for gas concentration determination was checked with a one-point calibration in intervals of 30 days of active measurement time, as specified in the protocol. The manufacturer's protocol specifies that the CO₂ sensor must be gassed with N₂ for 30 minutes to verify the zero point. The H₂ sensor is gassed with CO₂ for 30 minutes and then set to 0%.

Comment 3: Viability assay robustness. DiBAC₄ primarily marks depolarised membranes; please report counting method (fields, cells counted) and replicates. Include error bars on Figure 3.

Answer 3: We have added a note in the methods section stating that four images per time point were evaluated for the count and added the corresponding error bars in Fig. 3. Since the aim was not to determine the cell count but only the ratio of dead (stained) cells to living (unstained) cells, a standard microscope slide with a cover glass was used. However, the total number of cells counted varies due to the biomass density at the respective time point in the bioreactor. If the cell density appeared too high, it was diluted accordingly before a sample was examined under the microscope. The total number of cells evaluated per data point ranged between 91 and 646 cells.

Comment 4: Fed-batch inhibition analysis. You discuss acetate/ionic-strength effects; quantify base addition (NaOH consumption) and, if possible, conductivity to support the ionic-strength argument. Consider a simple electron/redox balance (including ethanol) to explain the H₂ decline, please.

Answer 4: Thank you very much for this interesting comment. The influence of ionic strength on bacterial performance is a very exciting aspect. We have given some thought to how we can better quantify ionic strength. Unfortunately, we cannot specify an exact conductivity, as complex reaction sequences with multiple by-products make it almost impossible to calibrate the conductivity, but by adding NaOH, the acids produced and ethanol, we can specify a total substance concentration in mmol/l, which was higher in the fed batch compared to the batch.  In addition, due to the metabolism of glucose as C6 compound into smaller carbon compounds at the end of each discontinuous fermentation, the molar concentration is higher than at the beginning. We also attempted to calculate the change in osmolarity of the medium from the start of fermentation to the end of fermentation in order to provide information about the osmotic forces acting on the cells. However, since it is unclear to what extent components and nutrients other than glucose are absorbed by the cells, the uncertainty was too great. Furthermore, it is not possible to clearly determine which substances in the medium the added NaOH could still react with.

Commet 5: Carbon balance uncertainty. The balance closes at ~103%. Provide an uncertainty estimate and clarify contributions from yeast extract, please.

Answer 5: We added the carbon balance uncertainty. Since there is no defined composition for yeast extract, it is difficult to estimate the extent to which carbon from this source has been utilised. However, since the modified version of the TBGY medium contains 0.5 g/L yeast extract and 5 g/L glucose, a maximum deviation of up to 10% cannot be ruled out in theory.

Comment 6: Language polish. Fix minor wording/typos (“gas stream”, not “steam”), unify decimal separators (comma vs dot) and define units consistently, please.

Answer 6: Thank you for pointing out this error. We have reviewed the manuscript again and made the necessary adjustments. The decimal separation is now uniformly marked with a dot. The comma in the remaining numbers is only used to improve readability for numbers that have at least four figures.

Comment 7: The work is strong on reactor-level comparison and viability; adding small pieces of quantitative mass-transfer/statistics would make it exemplary

Answer 7: Thank you very much for this feedback. We have tried to focus more on mass transfer in some places, and high glucose conversion rates were achieved in the continuous bioreactor compared to the batch fermentation approaches, particularly with regard to performance. Similarly, biomass formation in continuous operation at a dilution rate of D=0.07 is only slightly higher compared to batch operation in the first 24 hours of fermentation. The first 24 hours of fermentation are particularly relevant here, as glucose was present in the medium until this point and from then on only less than 1 mM.

 

Reviewer 3 Report

Comments and Suggestions for Authors

This study aimed to investigate the efficient bio-hydrogen production by the anaerobic cultivation of thermophilic bacteria, Thermotoga neapolitana focusing on the optimization of process configurations such as batch, fed-batch, and continuous fermentations.

 

1. English and MS writing must be significantly improved.
2. More detailed table and figure captions be given for self-understood.
3. Statistical analysis method must be explained, and any comparison must be made based on statistical analysis. Otherwise, no statistically meaningful comparison be made.
4. The result of Fig.2 is time-dependent and culture condition-dependent, and thus more careful analysis be made.
5. The complete sets of time-profile data should be compared in Fig.3 in relation to the culture condition and nutrient availability.
6. More careful and rational experimental design must be considered for Figs. 4, 5, and 6 in view of culture condition and nutrient availability.
7. Many similar papers have already been published on the biohydrogen production by thermophilic bacteria, and therefore, the present approach/method and the result must be compared with others, and show the essential novelty/academic originality as compared to the others.
8. In particular, the mechanistic (physiological/metabolic) details on how the culture condition and nutrient availability in batch, fed-batch, and continuous culture affected the metabolism and bio-hydrogen production must be examined with additional supporting data of metabolic flux analysis etc.
9. The performance of biohydrogen production must be compared with others.

 

Overall, it is not clear about novelty/academic originality to meet the standard of the journal.

 

 

Author Response

Dear reviewer, thank you very much for taking the time to review this manuscript and for providing us with this very helpful feedback on our work. Below are our responses to each of your comments. We have done our best to address the comments to your convenience.

Comment 1: English and MS writing must be significantly improved.

Answer 1: Thank you very much for your feedback. I agree with you. We have revised the manuscript in this regard and have attempted to elevate the wording in several places to a higher level in order to meet scientific standards.

Comment 2: More detailed table and figure captions be given for self-understood.

Answer 2: Thank you very much for pointing that out. We have gone through the tables and illustration captions and made the necessary adjustments. The captions for the tables and illustrations have been expanded and described in more detail without becoming too long and confusing.

Comment 3: Statistical analysis method must be explained, and any comparison must be made based on statistical analysis. Otherwise, no statistically meaningful comparison be made.

Answer 3: Thank you very much for your comment. Where a statistical comparison was mentioned, the method used for validation was also included. For this purpose, a modified Welch's t-test was applied, in which two mean values with different unknown variances were compared with each other in order to prove a significant difference. The remaining statements in which statistical significance was mentioned refer to the literature, where this was described accordingly.

Comment 4: The result of Fig.2 is time-dependent and culture condition-dependent, and thus more careful analysis be made.

Answer 4: Thank you for your comment. You are right, the results of vital staining depend on the condition of the culture. Figure 2 is primarily intended to show how the cell staining and its evaluation took place. Figure 2a shows a dead control. As mentioned in the figure caption, this is a freshly autoclaved sample of Thermotoga neapolitana. The cultivation was carried out in the same way as for the production of a fresh preculture to ensure that viable cells grow until the time of autoclaving and that as few cell fragments as possible in the solution can distort the result. After autoclaving, 1 ml aliquots with 25% glycerol were prepared and frozen so that a fresh dead control could be used for each vital staining. The dead control served as a reference to verify that the staining was working. Figure 2c shows the live control, which was also incubated overnight like a fresh preculture and was then either used directly as a live control for the staining or stored in 1 ml aliquots with 25% glycerol until use. This sample was also stained with DiBAC₄ (3) dye, with the expectation that no/hardly any fluorescent cells would be visible under the microscope. Since fresh controls were used at each point in time when the viability of a fermentation sample was determined, the viability determination procedure was verified repeatedly. Fig. 2b shows a real fermentation sample from the batch after 24 hours of fermentation, which shows that most of the cells are viable, as only a few of them were stained by the dye, while most remained unstained. We have included the explanation that the culture depicted in figure 2b is at the beginning of the stationary phase due to high viability, but with only a slight increase in biomass thereafter.

Comment 5: The complete sets of time-profile data should be compared in Fig.3 in relation to the culture condition and nutrient availability.

Answer 5: I fully agree with you on this point. Cell viability must be considered in relation to the availability of nutrients. Glucose is particularly important here, as it is essential as an energy source for high cell viability. To better illustrate this relationship, we have made a reference to Figures 4 and 5. Since these figures show the available glucose, a difference in cell viability can be seen between the batch and fed-batch processes. While low cell viability is associated with low nutrient availability in batch, viability in fed-batch decreased over time despite the availability of nutrients, suggesting that nutrient availability alone is not sufficient for high cell viability over a longer period of time.

Comment 6: More careful and rational experimental design must be considered for Figs. 4, 5, and 6 in view of culture condition and nutrient availability.

Answer 6: Thank you for pointing this out. We have tried to refer more to nutrient availability and general culture conditions.

Comment 7: Many similar papers have already been published on the biohydrogen production by thermophilic bacteria, and therefore, the present approach/method and the result must be compared with others, and show the essential novelty/academic originality as compared to the others.

Answer 7: Thank you very much for this feedback.  We have tried to emphasise the scientific novelty value of this work more strongly. To our knowledge, this is the first study to compare hydrogen production by T. neapolitana in batch, fed-batch and continuous modes within the same reactor setup, while also presenting data on the cell viability of Thermotoga neapolitana at different stages of fermentation and evidence of glutamate formation as a novel fermentation by-product.

Comment 8: In particular, the mechanistic (physiological/metabolic) details on how the culture condition and nutrient availability in batch, fed-batch, and continuous culture affected the metabolism and bio-hydrogen production must be examined with additional supporting data of metabolic flux analysis etc.

Answer 8: Thank you for addressing this exciting topic. Establishing the connection between nutrient availability and hydrogen production in the different reactor modes is very important, especially since nutrient availability varies over time in all three systems. A detailed metabolic flux analysis was beyond the scope of this study, but our C-balance and by-product profiles allow qualitative conclusions on electron distribution and pathway activity. We have attempted to present this topic more clearly. In addition, we have drawn on literature in which some metabolic analyses have already been carried out using suitable methods, such as 13C labelling.

Comment 9: The performance of biohydrogen production must be compared with others.

Answer 9: Thank you for pointing this out. Comparing our own results with those of other research groups is absolutely essential. When writing the manuscript, we made effort to compare comparable approaches in the literature with our fermentation approaches. As you mentioned in a comment above, we came across many papers dealing with biohydrogen production. However, we did not initially include all of these in our paper, as the methodology naturally always involves certain differences, which meant that we only considered a direct comparison to be meaningful in some cases, as different media and reactor designs could lead to different results. Therefore, the comparisons should be viewed with caution. Nevertheless, we are constantly striving to improve the quality of our scientific work and have compared our data with the results of other research groups where we felt that a comparison was meaningful due to similar methodologies.

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

The qusetions have been solved, I recommend to accept the paper 

Author Response

Dear reviewer,

we thank you for your constructive feedback and are very pleased that we were able to implement your comments in a reasonable manner. We are convinced that this has improved the quality of the manuscript.

Reviewer 3 Report

Comments and Suggestions for Authors

This study demonstrates that the continuous culture is better than batch and fed batch culture for the cell growth and H2 production. However, conclusion must be careful.

 

1. The medium composition with carbon and nitrogen sources should be specified.
2. It is not clear how the pH values were controlled constant from Fig.1. The pH values are important in the present cases, and therefore, the pH values should be shown for all the three cases.
3. The result of Fig.3 depends on the culture condition, substrate concentration and metabolite concentrations, and therefore more careful analysis be made.
4. In the fed-batch culture how the substrate concentration was controlled constant ? The effect of this set point on the fermentation result must be examined. If the substrate concentration was kept low, the acetate formation may be lower, and less inhibitory to the cell growth and H2 production. Then it becomes comparable to the continuous culture.
5. The effect of acetate concentration on the (maximum) specific growth rate and the specific metabolite production rates must be examined, and shown.
6. In Figure 6, the concentrations of substrate and acetate should also be compared.
7. More detailed metabolic (flux) analysis must be made for the cell growth and metabolite production.
8. Many statements are speculative without showing the confirming data, and thus this must be significantly improved.

 

Overall, it is not clear about novelty/academic originality to meet the standard of the journal.

 

 

Author Response

Thank you very much for taking the time to review the manuscript. You will find our detailed responses to your comments below.

Comment 1: The medium composition with carbon and nitrogen sources should be specified.

Answer 1: The exact details of the media composition are already provided in section 2.1 Bacterial strain and precultivation, lines 134 to 137. No other carbon or nitrogen sources were used. The only test component for which this composition did not apply is the feed in the fed-batch. As mentioned in the manuscript, the composition is the same except for the glucose concentration. The glucose concentration was 100 g·l^-1 (equals 555 mM), i.e. 20 times as much as in the modified TBGY medium according to Childers described above.

 

Comment 2: It is not clear how the pH values were controlled constant from Fig.1. The pH values are important in the present cases, and therefore, the pH values should be shown for all the three cases.

Answer 2: We strive to present the methodology as clearly as possible, but it appears that the details regarding pH regulation have not been described sufficiently for the sake of shortness (as we determine pH regulation as a basic bioprocess standard procedure). We have made additions to the caption of Figure 1 to make it clear that the pH value was measured continuously during fermentation using a pH electrode. As soon as the pH value of the bioreactor suspension fell below 7.35, counter-regulation of a standard PID controller with 2 M NaOH solution started automatically. The pump rate was 0.5 mL/min and stopped automatically as soon as the pH value was greater than or equal to 7.35. This value was chosen to prevent possible over-regulation. It was found that the pH value was kept stable, fluctuating by <0.02 during the entire fermentation process. We therefore believe that adding explicit graphs for a pH value curve would not add any scientific value to this work. However, if a pH value curve is desired, we can provide this data as an supplementary on request.

 

Comment 3: The result of Fig.3 depends on the culture condition, substrate concentration and metabolite concentrations, and therefore more careful analysis be made.

Answer 3: We thank the reviewer for this comment. We fully agree that the results shown in Fig. 3 are dependent on culture conditions, substrate concentration and the accumulation of metabolites. To address this, we have clarified in the revised manuscript that Fig. 3 must be interpreted in the context of these parameters. Specifically, we highlight that in batch culture the decline in viability correlates with decreasing substrate availability, whereas in fed-batch cultures the decline occurs despite sufficient substrate, indicating that metabolite accumulation and changing culture conditions may also play a role. The observation that viability decreases over time in the fed-batch despite a constant feed supply is consistent with the observation by Ughy et al. that nutrient availability alone does not ensure long-term viability. As we do not have complete metabolite profiles beyond those already presented, we have deliberately phrased our interpretation with appropriate caution and have pointed to the relevant figures (Figs. 4 and 5) where glucose availability is shown. In sections 3.2 Bacterial cell growth, 3.3 Fed-batch, and 3.4 Continuous fermentation, viability is revisited and considered in the context of substrate availability and byproduct accumulation. For better illustration, we have gone into more detail on the individual values in these sections. We believe this provides a careful and balanced interpretation of the data within the scope of the study.

Comment 4: In the fed-batch culture how the substrate concentration was controlled constant ? The effect of this set point on the fermentation result must be examined. If the substrate concentration was kept low, the acetate formation may be lower, and less inhibitory to the cell growth and H2 production. Then it becomes comparable to the continuous culture.

Answer 4: I agree in principle that a permanently low substrate concentration increases the comparability of fed-batch and continuous fermentation. This was also our intention. As described in section 3.3 Fed-batch, the feed rate was reduced due to the observation of increasing substrate concentrations in the reactor. The substrate concentration was not actively controlled, it resulted from the feed rate and the consumption of the substrate by the bacteria. The feed rate for the fed-batch was correspondingly lower than the feed rate for the continuous fermentation approaches and was adjusted via a controllable peristaltic pump on the BioFlo 120 control unit. This is primarily due to the limited head space in the reactor and the fact that, in contrast to continuous operation, no medium was removed from the reactor during the experiment. The initial feed rate used for the fed-batch was based on the glucose consumption rate determined in previous experiments. The exact feed rate was set using the pumps on the BioFlo 120 Control Unit from Eppendorf, as described in the text. Since no medium was removed from the fermenter during fed-batch, once glucose was added, it remained in the fermenter until it was consumed. In order to better classify the substrate increase in the fermenter, the substrate concentration was lower than the starting concentration in the fermenter up to the measurement point after 214 h, and also in comparison to the start of batch fermentation, which showed high biomass growth and H₂ production. After this point, the concentration continued to rise, indicating that substrate consumption was lower than supply, but at the end of 360 h, the glucose concentration was 37.5 mM, which was not many multiples higher than in the batch. We only partially agree with the assumption that a lower substrate concentration leads to less acetate and, accordingly, less H₂ inhibition. Of course, it can be assumed that with an even more limited substrate supply and the substrate concentration in the reactor kept at a constant level of almost 0, less acetate would be produced, or at a lower rate, as this is the main by-product. However, hydrogen production only occurs if there is high metabolic activity. This is the case, for example, when glucose is converted to pyruvate, which then degrades into acetate and CO₂. The reduction equivalents NADH produced during glycolysis can be oxidised back to NAD⁺ during hydrogen production. Thus, high acetate production is generally associated with high hydrogen production, even if excessive acetate concentrations have inhibitory effects. A lower feed supply would therefore have led to a slowdown in acetate and hydrogen production, but would not have prevented inhibitory effects, as the acetate produced is not removed from the reactor anyway. Nor is it necessarily to be expected that the acetate will be consumed by capnophilic lactic acid fermentation, as the reactor was gassed with nitrogen.

 

Comment 5: The effect of acetate concentration on the (maximum) specific growth rate and the specific metabolite production rates must be examined, and shown.

Answer 5: The effect of acetate on the maximum growth rate is, of course, important. However, the aim of the present study was not to present the inhibition kinetics of various metabolites, but to compare different reactor operating modes. Nevertheless, we have referred to existing results on the inhibition of acetate on biomass formation. The work of Dreschke et al 2019 showed that acetate reduced both biomass formation and hydrogen production in batch experiments. Concentrations of 0–240 mM acetate were tested as additives, with the result that biomass growth decreased by up to 43%. The concentrations tested there cover the entire concentration range observed in this study. The highest acetate concentration was 130 ± 4 mM in the fed-batch within this range, which supports the assumption that the increasing acetate concentration in the fed-batch negatively affects the further performance of the cells. In addition, this publication shows that acetate in a continuous fermentation process at up to 240 mM in a feed, i.e. also in the reactor, has no negative effects on biomass concentration and hydrogen production. Since no acetate was explicitly added to the feed in this study, the reactor only contained the acetate produced by the bacteria themselves. As the medium was continuously replaced during continuous operation, the acetate concentration in steady state was always below 45 mM in the reactor at all dilution rates tested. This, together with the lower acetate concentration compared to fed-batch, supports the assumption that inhibition by by-products such as acetate does not have a limiting effect in continuous operation.

Comment 6: In Figure 6, the concentrations of substrate and acetate should also be compared.

Answer 6: For continuous fermentation, graphics showing the substrate concentration together with the acetate concentration were deliberately not produced separately. The respective glucose feed rates are described in the text. Since the glucose that was added was almost completely consumed in steady state at the three dilution rates D=0.03 h^-1, D=0.05 h^-1 and D=0.07 h^-1, most glucose concentration measurements in the reactor are close to 0 g/L and thus at the lower limit of measurability with the HPLC method described. Furthermore, Fig. 6 does not show a concentration curve, but rather the production rate of the various products, including acetate. Steady state was defined, among other things, by the fact that the substrate concentration in the reactor was observed to be constantly low with fluctuations of less than 0.5 mM. In addition, the production rates also showed constant production. This could be observed particularly in hydrogen production, as this was measured continuously online and thus indicated a clear range for constant production rates.

Comment 7: More detailed metabolic (flux) analysis must be made for the cell growth and metabolite production.

Answer 7: As already noted in the first review round, the topic of metabolic flux analysis is a very exciting one, while not within the scope of our current research with an emphasize on bioprocess development. We have included data from the literature and have now added the outlook that analyses such as these at different points in time and in different reactor operating modes and dilution rates can provide further insights into metabolic behaviour. Unfortunately, however, we are currently unable to perform our own metabolic flux analyses as we have a focus on bioprocess enineering. I sincerely hope you understand this point.

Comment 8: Many statements are speculative without showing the confirming data, and thus this must be significantly improved.

Answer 8: From the outset, we have taken great care to support the statements made in this study with both our own data and values from literature. Wherever we expressed a point as a hypothesis, we explicitly referred to the relevant scientific literature. For analyses or observations not performed within the scope of this study, we consistently cited published work. In addition, during the first round of review we carefully revised our wording so that significant changes are only claimed where supported by statistical evidence. For other statements, we deliberately avoided claiming absolute certainty, as batch and fed-batch systems are subject to continuously changing conditions, and it cannot be unambiguously attributed that a single factor alone is responsible for a given observation. In light of this, we respectfully maintain that our statements are already formulated with due caution and with appropriate references, and therefore we do not consider a substantial revision of the manuscript necessary on this point.