Heterogeneous Copper Catalysts in the Aqueous Phase Hydrogenation of Maltose to Sorbitol

Round 1

Reviewer 1 Report

The interesting results were obtained in the manuscript, specific items required to be handled are listed below:

1. The selectivities of products were listed in Table 1, the total was not 100%, what is the rest?

2. As authors describing, the acid is important for the catalysts, and the Py-IR was performed, Lewis acid sites was detected. How about the Bronsted acid sites in silica support and catalysts? Authors stated that lots of –OH groups were found in silica support, which acted as Bronsted acid sites.

3. It is interesting to discover the large difference in catalytic performance for the alumina and silica catalysts. Reduction degree of Cu is important for the hydrogenation reaction. TPR is suggested to be performed to the alumina and silica catalysts to verify the CuO is reduced completely under the pre-reduction conditions.

4. TGA was described in 3.3. Catalyst characterization section, while no TGA results was listed in the manuscript.

5. Minor grammar errors and sentence error were found in manuscript, such as in line 172 “in figure 2…” in line 257 “ then….”

English language and style are fine/minor spell check required

Author Response

We warmly thank the reviewer for his/her comments and suggestions. We amended the manuscript as described below:

1. The selectivities of products were listed in Table 1, the total was not 100%, what is the rest?  In Table 1 yields obtained by HPLC in the raw reaction mixture are reported. The main by-product quantified is HMF, the others being obtained in small amounts. Among these we identified maltitol, mannose and fructose, while levulinc acid, glycerol and lactic acid have been excluded among the standard injected. In the reaction giving poor yield in sorbitol also the unsoluble humins were obtained. A comment about this has been added in the manuscript.
2. As authors describing, the acid is important for the catalysts, and the Py-IR was performed, Lewis acid sites was detected. How about the Bronsted acid sites in silica support and catalysts? Authors stated that lots of –OH groups were found in silica support, which acted as Bronsted acid sites. Actually with Py as the probe molecule no Bronsted acidity was observed on silica materials, as already stated in the previous literature [Dalton Trans., 2013, 42, 1319–1328]
3. It is interesting to discover the large difference in catalytic performance for the alumina and silica catalysts. Reduction degree of Cu is important for the hydrogenation reaction. TPR is suggested to be performed to the alumina and silica catalysts to verify the CuO is reduced completely under the pre-reduction conditions.  The reducibility of the different catalysts is certainly an important aspect as already pointed out in the manuscript. The TPR profiles have been already reported in previous papers. In particular the TPR of CuO/SiO₂ catalyst presents a sharp and low temperature peak that can be ascribed to the reduction of a well dispersed and highly uniform CuO phase into Cu(0) [Dalton Trans., 2013, 42, 1319–1328 and  Chinese Journal of Catalysis 40 (2019) 1788–1794]. On the other hand, both CuO/Al₂O₃ and Cu/SiO₂-Al₂O₃ show broader peaks not ascribable to a single copper phase. [Catal. Sci. Technol., 2017,7, 1386; Inorganica Chimica Acta 380 (2012) 194–200 and  J. Phys. Chem. B 2006, 110, 7851-7861]. A comment in this respect has been added in the main manuscript.
4. TGA was described in 3.3. Catalyst characterization section, while no TGA results was listed in the manuscript. TGA analysis was used to evaluate the -OH density. This aspect has been clarified in the main text.
5. Minor grammar errors and sentence error were found in manuscript, such as in line 172 “in figure 2…” in line 257 “ then….”Errors have been corrected.

Reviewer 2 Report

The objective of the work is to obtain sorbitol from maltose. To promote the reaction, copper catalysts supported on oxides (silica, alumina, silica-alumina, etc.) were used. The results are interesting, but I consider that, before its publication,  this work should be substantially improved. The work lacks a good correlation between the catalytic results and the results of the characterization of the catalysts (whose physicochemical properties have been widely reported in the bibliography). Some catalyst characterization techniques results could be added to the work. Probably, a complete reaction scheme could be added, which could help to understand the formation of by-products. And on the other hand, the selectivity of the reaction and the identification of by-products should be added in the revised version.

-Could the authors report the selectivity of the reaction?

-Line 137: With the latter oxides almost complete conversion of glucose was observed differently to the 35-45% remaining at the end of the reaction carried out with silica or without any catalyst. This means that silica-alumina and alumina favor as themselves the decomposition or isomerization of glucose.

What is expressed in this text is not well understood. Why talk about glucose conversion? Again, selectivity results should be added to the work (in the liquid phase, ruling out the formation of humic). The authors observed hydrolysis products of maltose, and also 5HMF, which comes from the dehydration of fructose. Did the authors observe fructose as a reaction by-product? This sugar comes from the isomerization of glucose.

-Table 2. Please change commas to points in the last column

-Line 149. Do the authors rule out the presence of Bronsted acid sites in the catalysts? To support this result, the FTIR diagrams should be added to the work as supplementary material.

-Line 154: Please review this text:

, thus suggesting that the acidity of the system promotes side reactions but the hydrogenation one.

 

- The catalysts were pre-reduced before the catalytic tests at 270 °C for 20 min under 250 air and for 20 min under vacuum and then reduced through three hydrogen (1 bar) and 251 vacuum cycles (5 mins each) at the same temperature. What is the oxidation state of the copper species when used in the reaction? Could the authors add XPS analysis to this work? Why is CuO mentioned in the caption of Figure 2? Do the authors consider the coexistence of metallic phases and oxidic phases in the catalyst? For each catalyst, the authors must establish the nature of the copper species.

-Line 206: Were the authors able to observe the presence of products of hydrogenolysis?

-What was the TGA technique used for?

Author Response

We warmly thank the reviewer for his/her comments and suggestions. The manuscript has been revised as described below:

The objective of the work is to obtain sorbitol from maltose. To promote the reaction, copper catalysts supported on oxides (silica, alumina, silica-alumina, etc.) were used. The results are interesting, but I consider that, before its publication, this work should be substantially improved. The work lacks a good correlation between the catalytic results and the results of the characterization of the catalysts (whose physicochemical properties have been widely reported in the bibliography). Some catalyst characterization techniques results could be added to the work. Probably, a complete reaction scheme could be added, which could help to understand the formation of by-products. And on the other hand, the selectivity of the reaction and the identification of by-products should be added in the revised version.

-Could the authors report the selectivity of the reaction? Selectivity to sorbitol has been added in Table 1

-Line 137: With the latter oxides almost complete conversion of glucose was observed differently to the 35-45% remaining at the end of the reaction carried out with silica or without any catalyst. This means that silica-alumina and alumina favor as themselves the decomposition or isomerization of glucose. What is expressed in this text is not well understood. Why talk about glucose conversion? Again, selectivity results should be added to the work (in the liquid phase, ruling out the formation of humic). The authors observed hydrolysis products of maltose, and also 5HMF, which comes from the dehydration of fructose. Did the authors observe fructose as a reaction by-product? This sugar comes from the isomerization of glucose.

A complete scheme with the main by-products observed has been added in the manuscript.

Fructose was detected in very few amounts in the reactions carried out with Al₂O₃, Cu/SiO₂-Al₂O₃ and SiO₂-Al₂O_3, being also the systems promoting in higher extent the formation of unsoluble humins.This comment has been added in the main text. Moreover the sentence relying on “glucose conversion” has been rewritten.

-Table 2. Please change commas to points in the last column The correction has been done.

-Line 149. Do the authors rule out the presence of Bronsted acid sites in the catalysts? To support this result, the FTIR diagrams should be added to the work as supplementary material

Figure S1 reporting the FT-IR spectra of pyridine desorbed at 150°C has been added and also a comment in the main text describing the bands of the spectra. In particular the bands around 1610 cm^-1 and 1450 cm^-1 unambiguously show the presence of Lewis acid sites. Only the Cu/SiO₂-Al₂O₃ sample shows the presence of a band at 1620 cm^-1 that can be diagnostic of both Lewis and Brønsted sites. On the other hand, the absence of the 1550 cm^-1 band, the one traditionally used to quantify Brønsted sites supports the hypothesis of the presence of the sole Lewis acidity also for the silica alumina based catalyst.  

-Line 154: Please review this text:, thus suggesting that the acidity of the system promotes side reactions but the hydrogenation one. The sentence has been rewritten.

 - The catalysts were pre-reduced before the catalytic tests at 270 °C for 20 min under 250 air and for 20 min under vacuum and then reduced through three hydrogen (1 bar) and 251 vacuum cycles (5 mins each) at the same temperature. What is the oxidation state of the copper species when used in the reaction? Could the authors add XPS analysis to this work? Why is CuO mentioned in the caption of Figure 2? Do the authors consider the coexistence of metallic phases and oxidic phases in the catalyst? For each catalyst, the authors must establish the nature of the copper species.

The reducibility of the different catalysts is certainly an important aspect as already pointed out in the manuscript. The TPR profiles have been already reported in previous papers. In particular the TPR of CuO/SiO₂ catalyst presents a sharp and low temperature peak that can be ascribed to the reduction of a well dispersed and highly uniform CuO phase into Cu(0) [Dalton Trans., 2013, 42, 1319–1328 and  Chinese Journal of Catalysis 40 (2019) 1788–1794]. On the other hand, both CuO/Al₂O₃ and Cu/SiO₂-Al₂O₃ show broader peaks not ascribable to a single copper phase. [Catal. Sci. Technol., 2017,7, 1386; Inorganica Chimica Acta 380 (2012) 194–200 and J. Phys. Chem. B 2006, 110, 7851-7861]. The reduction of the copper phase, confirmed by the disappearance of the shake-up peak typical of the CuO species and by the changing of the contribution peak areas [Molecular Catalysis 528 (2022) 112462]

 A comment in this respect has been added in the main manuscript

-Line 206: Were the authors able to observe the presence of products of hydrogenolysis?

The main products observed and identified have been put in light in the main text.

-What was the TGA technique used for?

TGA analysis was used to evaluate the -OH density. This aspect has been clarified in the main text.

Reviewer 3 Report

The topic of the manuscript is of interest, however, several aspects should be improved.

1.      Materials and methods section should be more detailed, since several procedures are missing or uncomplete.

a.      Quantification of Cu on the catalysts

b.      Surface morphology of the catalysts

c.      Determination of crystallite size

d.      Determination of surface hydrophobicity

 

2.      The percentage of Cu on each catalysts should be measured

3.      Add to the surface area of the supports, also of the catalysts should be included.

4.      Crystallite size of all catalysts, not only one of them is required. At this point, the increased dispersion on SiO2A could be logical, from the higher surface area of the silica in comparison to SiO2B.

5.      Quantification of weak, medium and strong acid sites is convenient, since not all of them participate in the reactions.

It could be improved.

Author Response

We warmly thank the reviewer for his/her comments. The manuscript has been corrected as descrobed below:

The topic of the manuscript is of interest, however, several aspects should be improved.

1. Materials and methods section should be more detailed, since several procedures are missing or uncomplete.
2. Quantification of Cu on the catalysts: The Cu loading of the catalysts is 16%. It was determined by ICP analysis and this datum has been added in the experimental section.
3. Surface morphology of the catalysts. Acidity was determined from pyridine analysis, mean particles size from TEM and -OH loss from TGA.
4. Determination of crystallite size. A comment was added into the experimental section.
5. Determination of surface hydrophobicity. As measure of hydrophobicity we calculated the -OH surface density from TGA analysis: the water loss was evaluated in the range of 250–900 °C and mmol_OH/g_cat was calculated by the following formula: [(2 × Δwt %) × 10]/18 (g/mol).
6. The percentage of Cu on each catalysts should be measured. The % of Cu by weight was determined by ICP and the value, approximated to the unit, is 16% for all the catalysts.
7. Add to the surface area of the supports, also of the catalysts should be included. The surface areas of the supports have been added in table 2.
8. Crystallite size of all catalysts, not only one of them is required. At this point, the increased dispersion on SiO2A could be logical, from the higher surface area of the silica in comparison to SiO2B. Mean particles size is reported in Table 2.
9. Quantification of weak, medium and strong acid sites is convenient, since not all of them participate in the reactions. FT-IR spectra of the adsorbed pyridine cannot give information about the strength of the sites, but only about their nature, in particular in the discrimination between Lewis and Bronsted ones. On the other hand some indications can be drawn by the shift of the bands to lower wavenumbers. In this respect in the case of the Cu/Al₂O₃ sample we observe (Figure S1, added in the revised manuscript) a small shift suggesting the presence of weaker Lewis sites with respect to the other materials.

 

Round 2

Reviewer 1 Report

Minor grammar errors and sentence error were found in manuscript, such as in line 152 and Table 2.

-

Author Response

We thank the reviewer and we corrected the suggested errors.

Reviewer 2 Report

The work was improved and can be published in the current form.

Author Response

We warmly thank the reviewer for his/her reply.

Reviewer 3 Report

Effectively, FT-IR does not give information of the amount of acid sites, but other methods such as NH3-TPD allows this determination.

Table 2 entry is not well described.

What is the role of the different surfaces area in the catalytic behaviour?

....

Author Response

Effectively, FT-IR does not give information of the amount of acid sites, but other methods such as NH3-TPD allows this determination.

W e thank the reviewr for his/her comment. Actually NH₃-TPD would allow one to gain different informations. We will take this suggestion into account for a future deeper study about these systems.

Table 2 entry is not well described.

We detailed the Table 2 entries.

What is the role of the different surfaces area in the catalytic behaviour?

The role of catalyst surface area in promoting the formation of sorbitol is certainly important. Thus, the two Cu on silica materials show both the best selectivity and the highest SA among the  catalysts tested. On the other hand the presence of the other features mentioned above has also a strong influence. Cu/Al₂O₃ with acidity comparable to Cu/SiO₂ A but a much lower SA show poor selectivity, whereas the significant presence of  acidic sites on Cu/SiO₂-Al₂O₃ results in no selectivity nothwistanding a good SA.

We added this comment in the main text and we thank the reviewr for his/her suggestion