A New Complex Borohydride LiAl(BH₄)₂Cl₂

Round 1

Reviewer 1 Report

The authors report the synthesis and structural characterization of the new compound LiAlCl₂(BH₄)₂. The hydrogen storage properties are investigated. In general the manuscript is well structured and well written.

The author state that the main problem with complex metal hydrides is the high release temperatures and the emission of other gases than H₂. Another big issue is also the reversibility of the hydrogen absorption/desorption. Even in the cases where pure H₂ is released in endothermic reactions, reversibility is an issue and requires extreme conditions. This should be addressed in the introduction as the main focus is towards functional (and reversible) hydrogen storage.

In line 25, the absorption properties of mixed cat/anion borohydrides are mentioned. But has there been any successful rehydrogenation that resulted in the reformation of the bimetallic metal borohydride compound?

Line 30-34: In the case of most mixed-cation borohydrides, the mixed-cation compound decomposes into the individual components upon heating to the stability of the lesser stable compound, and the two metal borohydrides decompose more or less individually. This section suggests that this is different for the LiBH₄ compounds. Is the decomposition temperature of LiBH₄ really lowered? In line 160-161 the authors also mention that LiBH₄ is formed from the decomposition of Li₄Al₃(BH₄)₁₃, suggesting that there are no destabilization of LiBH₄.  

Table 1 does not really provide much information. If all phases are known in the diffraction pattern, it should be possible to add information on the amounts of the individual components. Also, the refined values for the lattice parameters of the compounds that undergo substitution reactions, i.e. Li₄Al₃(BH₄)₁₃ and LiAlCl₂(BH₄)₂, could be interesting. Furthermore, occupancy refinements of the BH₄ deficient structures could provide information on preferred sites for the Cl (or BH4?) in the structures and possibly determining x in Li₄Al₃(BH₄)_13-xCl_x and LiAlCl_2-x(BH₄)_2+x.

Figure 1: Please assign the remaining peaks as well. Do they correspond to LiCl and LiAlCl₄?

The spacegroup of LiAl(BH4)2Cl2 (C222₁) is a non-centrosymmetric spacegroup, so the authors should check if it could be better described in a centrosymmetric spacegroup (e.g. Cmcm?), which can often be realized by reorientation of the BH4 or by assuming disordered H positions. There are also several zero-intensity reflections in the Rietveld plot, which could indicate a higher symmetry.

Table 2 and line 101: The author state that the best fit is obtained for a random occupancy of the cationic sites, but according to the table there appear to be a slight preference for Li on site M1 and Al on site M2 (occupancy of 0.55 vs 0.45).

Table 3: I doubt that you can trust the H positions based on PXD data. Please comment on this. If so, I would suggest to remove the table and just mention the relevant bond distances (M-B and M-Cl) in the text. I would expect the coordination of the BH4 to be bidentate since it is bridging, as you also see for other metal borohydrides and the ammonia derivatives.

I feel that the crystal structure description on line 114-121 can be improved. It is e.g. not mentioned what the local coordination of M is (tetrahedral). I would rather describe the structure as built from connected tetrahedral [M(BH4)2Cl2] complexes via alternating bridges units of M-BH4-M and M-Cl-M along the a-axis.

It would be interesting with a comparison with the structure of LiAl(BH₄)₄. Are the structures related? See https://doi.org/10.1002/cssc.201701629

 Line 123: Is it BH4 or Cl substituting into the structure in LiAl(BH₄)_2+δCl_2+δ. I assume one of the “+” should be a “-“

In the dehydrogenation section, the authors state that LiBH4 is stable at 385 °C, and use this as an explanation for why the weight loss is less than expected for Li₄Al₃(BH₄)₁₃ (4.33:1 mixture). However, LiBH4 is not stable at 385 °C, and in case it was present, it should be visible from PXD. Except for the 4:33:1 ratio, there appear to be a trend that decreasing LiBH4 content results in lower H2 release. So is it possible that the 4.33:1 sample has partly decomposed prior to the measurement?. Or can the lower loss of H2 be explained by the higher amount of B2H6 being released?

The slow gradual release of H2 is interesting. Usually the decomposition of metal borohydrides happens abruptly. Any suggestions on a potential decomposition pathway and possible intermediate products, that can explain this behavior?

Author Response

1. The author state that the main problem with complex metal hydrides is the high release temperatures and the emission of other gases than H₂. Another big issue is also the reversibility of the hydrogen absorption/desorption. Even in the cases where pure H₂ is released in endothermic reactions, reversibility is an issue and requires extreme conditions. This should be addressed in the introduction as the main focus is towards functional (and reversible) hydrogen storage.

 

Author’s response: A sentence addressing reversibility has been added at the end of the first paragraph of the introduction.

 

2. In line 25, the absorption properties of mixed cat/anion borohydrides are mentioned. But has there been any successful rehydrogenation that resulted in the reformation of the bimetallic metal borohydride compound?

 

Author’s response: To the best of our knowledge - no; “and absorption” has been deleted from line 25.

 

3. Line 30-34: In the case of most mixed-cation borohydrides, the mixed-cation compound decomposes into the individual components upon heating to the stability of the lesser stable compound, and the two metal borohydrides decompose more or less individually. This section suggests that this is different for the LiBH₄ Is the decomposition temperature of LiBH₄ really lowered? In line 160-161 the authors also mention that LiBH₄ is formed from the decomposition of Li₄Al₃(BH₄)₁₃, suggesting that there are no destabilization of LiBH₄.  

Author’s response: Since formation of mixed cation borohydrides lowers onset temperatures of hydrogen release compared to pristine LiBH₄, we believe that removing “of LiBH₄” and adding “onset” resolves the issue.

4. Table 1 does not really provide much information. If all phases are known in the diffraction pattern, it should be possible to add information on the amounts of the individual components. Also, the refined values for the lattice parameters of the compounds that undergo substitution reactions, i.e. Li₄Al₃(BH₄)₁₃ and LiAlCl₂(BH₄)₂, could be interesting. Furthermore, occupancy refinements of the BH₄ deficient structures could provide information on preferred sites for the Cl (or BH4?) in the structures and possibly determining x in Li₄Al₃(BH₄)_13-xCl_x and LiAlCl_2-x(BH₄)_2+x.

Author’s response: We updated Table 1 with lattice parameters of the observed phases. Quantitative phase analysis is complicated because in addition to crystalline products, the mixtures may contain variable amounts of X-ray amorphous phases.

5. Figure 1: Please assign the remaining peaks as well. Do they correspond to LiCl and LiAlCl₄?

Author’s response: Figure 1 has been updated with the requested information.

6. The space group of LiAl(BH₄)₂Cl₂ (C222₁) is a non-centrosymmetric space group, so the authors should check if it could be better described in a centrosymmetric space group (e.g. Cmcm?), which can often be realized by reorientation of the BH₄ or by assuming disordered H positions. There are also several zero-intensity reflections in the Rietveld plot, which could indicate a higher symmetry.

Author’s response: The referee is correct in that, for example (201) Bragg reflection that is possible in C222₁ at 2q = ~19.3° but forbidden in Cmcm has zero intensity.  However, weak, but measurable (401), (403), and (601) are observed at 2q = ~32.9°, ~47.83, and ~48.3°, respectively. Thus, C222₁ is the only space group symmetry possible.  Further, all refinements assuming higher symmetry lead to inferior fits.

7. Table 2 and line 101: The author state that the best fit is obtained for a random occupancy of the cationic sites, but according to the table there appear to be a slight preference for Li on site M1 and Al on site M2 (occupancy of 0.55 vs 0.45).

Author’s response: Deviation from random occupancy is within two standard deviations, hence it may be considered random, now clarified: “The best fit (Fig. 2) was obtained for random (within two standard deviations) occupancy of the two available 4a sites with Al and Li atoms.”

8. Table 3: I doubt that you can trust the H positions based on PXD data. Please comment on this. If so, I would suggest to remove the table and just mention the relevant bond distances (M-B and M-Cl) in the text. I would expect the coordination of the BH4 to be bidentate since it is bridging, as you also see for other metal borohydrides and the ammonia derivatives.

Author’s response: We agree, and as clearly stated in the Structural characterization section, only centers of gravity of [BH₄]^- groups were refined. The denticity of the bridging [BH₄]^- groups is now correctly shown in the revised Fig. 3. The caption of Table 3 has been updated with the following note: “Since only the centers of gravity of two symmetrically independent [BH₄]^- groups were refined, the M-H bond lengths are tentative.”

9. I feel that the crystal structure description on line 114-121 can be improved. It is e.g. not mentioned what the local coordination of M is (tetrahedral). I would rather describe the structure as built from connected tetrahedral [M(BH4)2Cl2] complexes connected along the a-axis via alternating bridging M-BH4-M and M-Cl-M unites.

Author’s response: The suggested description has been added: “The structure can also be described as a network of tetrahedral [M(BH₄)₂Cl₂] complexes connected along the a-axis via alternating bridging M-BH₄-M and M-Cl-M units.“

10. It would be interesting with a comparison with the structure of LiAl(BH₄)₄. Are the structures related? See https://doi.org/10.1002/cssc.201701629

Author’s response: The relation of the new structure to the earlier known Li[Al(BH₄)₄] complex, which crystallizes in the monoclinic space group P2₁/c (#14), is in the coordination environment of the cations. All of the Li⁺ and Al³⁺ cations in the Li[Al(BH₄)₄] complex are surrounded by four BH₄^- anions, coordinated via edges forming a 3-D framework constructed from distorted [Al(BH₄)₄]^- “supertetrahedra.” In the new Cl-containing complex, there is also coordination of metals M1 and M2 to four anions, two of which are BH₄^- and two Cl^-. The difference is that the Li[Al(BH₄)₄] structure model includes four independent cations and complex anions, while unit cell of the LiAl(BH₄)₂Cl₂ compound consists of two cations, one Cl^- and two BH₄^- anions. As a result, the Cl-containing compound crystallizes in a higher symmetry (space group C222₁ (#20)), where M-BH₄-M and M-Cl-M layers alternate.

11. Line 123: Is it BH4 or Cl substituting into the structure in LiAl(BH₄)_2+δCl_2+δ. I assume one of the “+” should be a “-“.
    
    Author’s response: Thank you for pointing this out. The formula should be LiAl(BH₄)_2±δCl_2±δ, now corrected.
    
    
12. In the dehydrogenation section, the authors state that LiBH4 is stable at 385 °C, and use this as an explanation for why the weight loss is less than expected for Li₄Al₃(BH₄)₁₃ (4.33:1 mixture). However, LiBH4 is not stable at 385 °C, and in case it was present, it should be visible from PXD. Except for the 4.33:1 ratio, there appear to be a trend that decreasing LiBH4 content results in lower H2 release. So is it possible that the 4.33:1 sample has partly decomposed prior to the measurement? Or can the lower loss of H₂ be explained by the higher amount of B2H6 being released?

Author’s response: We did not detect LiBH₄ after the ball milling. A small quantity may still be present as X-ray amorphous LiBH₄ phase, yet this is unlikely because the richest in LiBH₄ reaction mixture was stoichiometric 4.33:1 with respect to Li₄Al₃(BH₄)₁₃. There was no decomposition of the 4.33:1 sample prior to heating. The lower loss of H₂ for this sample with the following increase for the next (3.67:1) composition can be explained by the presence of the Cl^- anion in the latter, which partially destabilizes the structure and leads to release of higher H₂ amount. Hence, formation of LiBH₄ upon the decomposition of the 4.33:1 sample along with release of a significant amount of B₂H₆ leads to decrease of the amount of hydrogen when compared to other samples with a low LiBH₄ content.

 

Below ~400°C LiBH₄ releases hydrogen in three steps:

1 - When LiBH₄ transforms from orthorhombic to hexagonal structure at 105∼112°C (∼0.1 wt.%).

2 - At its melting point of 275∼278 °C (~0.5-1.0 wt. % H₂₎;

3 - Between 400°C and 680°C (~9.0 wt. %).

As our experiments were limited to 385°C, it is likely that the third decomposition step may proceed with slow kinetics in high LiBH₄-content systems, where the majority of the newly formed LiBH₄ is still present in the mixture (except the 2:1 mixture, where we have a rapid gas desorption). The ball-milled sample has poor crystallinity, which makes it difficult to distinguish the LiBH₄, which is not fully decomposed.

 

13. The slow gradual release of H2 is interesting. Usually the decomposition of metal borohydrides happens abruptly. Any suggestions on a potential decomposition pathway and possible intermediate products, that can explain this behavior?

 

Author’s response: Major hydrogen release by LiBH₄ occurs above 400°C. As our experiments were limited to 385°C, it is likely that the newly formed LiBH₄ decomposes very slowly just below 400°C, except for the 2:1 mixture where rapid gas desorption occurs.

Reviewer 2 Report

The submitted study is of a very high quality. The Authors have successfully synthesized and studied the structure and properties of a new complex borohydride. The newly synthesized material is very promising and it may found practical application in the nearest future.The manuscript surely deserves to be published, however I have three questions that I would like the Authors to answer.

Does either AlCl₃ or LiBH₄ (substrates) exhibit polymorphism? If yes, it would be interesting to check how the reaction goes when another form of a substrate is being used. If no, please state it somewhere in the introduction. This is important as it is purely solid-state synthesis.

Table 2 and crystal structure. Is it common for other similar compounds to crystallize in that way? I.e., for NaY(BH₄)₂Cl₂?  mean with the mixture of atom (Li and Al in your case) at 4a site? I also wonder what was the Rw and Rwp in a models where there was no such structural disorder. There should be two of them with either Al or Li present at (0.633(2); 0; 0). I believe that the solved stricture is the correct one but it would be worth to show the results of the refinement without the disorder, maybe in the supplementary materials?

Figure 4, I agree that in the 11B spectrum there is a single peak. But in the 27Al spectra it seems like there is another peak at c.a. 60 ppm. Can you please comment on this?

Author Response

The submitted study is of a very high quality. The Authors have successfully synthesized and studied the structure and properties of a new complex borohydride. The newly synthesized material is very promising and it may found practical application in the nearest future. The manuscript surely deserves to be published, however I have three questions that I would like the Authors to answer.

1. Does either AlCl₃ or LiBH₄ (substrates) exhibit polymorphism? If yes, it would be interesting to check how the reaction goes when another form of a substrate is being used. If no, please state it somewhere in the introduction. This is important as it is purely solid-state synthesis.

Author’s response: LiBH₄ has four structural modifications: two at ambient and two at high pressure. Solid AlCl₃ exhibits no polymorphism. In our experiments, we used orthorhombic LiBH₄ stable at room temperature and ambient pressure. When ball-milled individually at the ambient temperature, no structural polymorphism was observed in LiBH₄. When ball milled together with AlCl₃, the components react. While we agree that it would be interesting to examine reactivity of, for example, hexagonal LiBH₄ with AlCl₃, this requires milling above ~100 °C, but we lack this capability.

2. Table 2 and crystal structure. Is it common for other similar compounds to crystallize in that way? I.e., for NaY(BH₄)₂Cl₂?  mean with the mixture of atom (Li and Al in your case) at 4a site? I also wonder what was the Rw and Rwp in a models where there was no such structural disorder. There should be two of them with either Al or Li present at (0.633(2); 0; 0). I believe that the solved stricture is the correct one but it would be worth to show the results of the refinement without the disorder, maybe in the supplementary materials?

Author’s response: It is common for cations in mixed borohydrides to occupy same sites. However, the structure of the LiAl(BH₄)₂Cl₂ is unique because of practically random occupancy of both available sites by Li and Al. When refined without disorder, the R_p and R_wp increase from 4.78 and 6.49 % to 5.84 and 8.47 % for the model with Al in M1 and Li in M2 and to 5.81 and 8.51 % for Li in M1 and Al in M2. The sentence ”The refinements of the models where Al ad Li are ordered in M1 and M2 positions leads to increase of R_p and R_wp, respectively, from 4.8 and 6.5 % to ~5.8 and ~8.5 % in both of the ordered models.” was added to manuscript.

 

3. Figure 4, I agree that in the 11B spectrum there is a single peak. But in the 27Al spectra it seems like there is another peak at c.a. 60 ppm. Can you please comment on this?

 

Author’s response: We added the assignment of ²⁷Al signal at 60 ppm as follows (page 6): "The minor ²⁷Al signal at 60 ppm was also identified to Al(BH₄)₂Cl₂ whose coordination geometry is slightly different from that yielding the main ²⁷Al signal at 74 ppm, suggesting the presence of a disordered phase as an impurity [20]."

Round 2

Reviewer 2 Report

The Authors have made the necessary changes and now I can strongly suggest publication of this manuscript "as it is now". Good job! Well done!