Steps toward Rationalization of the Enantiomeric Excess of the Sakurai–Hosomi–Denmark Allylation Catalyzed by Biisoquinoline N,N’-Dioxides Using Computations

Round 1

Reviewer 1 Report

This work by Roberto Peverati et al. described “Steps Towards Rationalization of the Enantiomeric Excess of the Sakurai-Hosomi-Denmark Allylation Catalyzed by Biiso-quinoline N,N’-Dioxides Using Computations”. The paper is complete and well structured. The Computational Methods are detailed and results critically analyzed and commented. I think the manuscript can be accepted in this form and it needs no revision.

Author Response

We thank the referee for taking the time to review our manuscript, no further action required.

Reviewer 2 Report

In this manuscript, Roberto Peverati and co-workers describe computational studies of the Sakurai-Hosomi-Denmark allylation reaction using Biiso-quinoline N,N’-Dioxides. The authors clear explained the reasons for the enantioselective excess using computational studies. The authors properly characterized all transition states and the reason for the relative energies. This manuscript can be published in Catalysts journals.

Additional questions.

1. It could be easy to understand the readers if authors could provide the reaction conditions and yield and ee of the reported reaction instead of giving the general reaction.
2. Authors need to provide compound numbers for Scheme-1.
3. In figure-1 when generalizing the catalyst structure authors could provide the reported C2 symmetric catalyst.
4. Authors could provide an expert conclusion and future development of this work.
5. Authors could refine and rewrite the conclusion.

Author Response

In this manuscript, Roberto Peverati and co-workers describe computational studies of the Sakurai-Hosomi-Denmark allylation reaction using Biiso-quinoline N,N’-Dioxides. The authors clear explained the reasons for the enantioselective excess using computational studies. The authors properly characterized all transition states and the reason for the relative energies. This manuscript can be published in Catalysts journals.

We thank the reviewer for the comments. We will address them below.

Additional questions.

• It could be easy to understand the readers if authors could provide the reaction conditions and yield and ee of the reported reaction instead of giving the general reaction.

In the new version of the manuscript, we included an additional scheme (Scheme 2 on page 5) in which we explicitly reported the molecules used in the computational studies, as well as the reaction conditions and yield from the synthesis previously reported by our groups.

• Authors need to provide compound numbers for Scheme-1.

We added the numbers as requested. The aldehyde is now structure 1, allyltrichlorosilane is structure 2, and the allylalcohol is structure 3. All the other molecules have been re-numbered accordingly.

• In figure-1 when generalizing the catalyst structure authors could provide the reported C2 symmetric catalyst.

Pictures of all the transition structures involving catalyst 8 are reported in the Supporting Information. We now stress this detail in the captions of Figure 1 and Figure 2.

The new caption to Figure 1 reads:

“The five possible arrangements of the ligands and chlorine atoms around a hexavalent silicon atom when a C2-symmetric Lewis base (like structures 2–6 in Figure 2) is used. See also Section S2 in the Supporting Information for a visual representation.”

The new caption to Figure 2 is:

“Structures of the catalysts that have been computationally characterized in the literature (2–7, 9), as well as Takenaka’s catalyst studied in this work (8). Structures 2–6 are C2-symmetric, see also the label of Figure 1 and Section S2 of the Supporting Information for additional visual representations.”

•  Authors could provide an expert conclusion and future development of this work.

We added a paragraph before the conclusions as advice for practitioners. The new paragraph reads:

“Modern functionals should always be preferred, as they have been designed to make up for the deficiencies of older approximations and to have a wider range of applicability.[38] We also advise caution when interpreting the computational results, especially if they disagree with the experimental findings. The reasons behind the failure of a certain approximation are not always easy to understand, and comparison with different approximations can guide towards the choice of a better one.[22,27,32,33,38] For more complicated cases—as in this study—a comprehensive analysis including multiple functionals provides a way to validate the results, especially when the agreement with the experimental findings is questionable.”

We also added an additional sentence to address the future development of this work. It reads:

“We plan on expanding our study in the future to include a larger number of aldehydes and different catalysts, and to eventually consider alternative reaction mechanisms.”

• Authors could refine and rewrite the conclusion.

We rephrased the concluding paragraph, which now reads:

Motivated by the disagreement between the computational and experimental results, as well as similar ones reported in the literature,[17] we also performed an assessment of 34 different density functional methods, with the goal of understanding the applicability of DFT as a general tool for studying this chemistry. We found that the DFT results are—in general—consistent, as long as functionals that correctly account for dispersion interactions are used. Agreement with the experimental results, however, is not always found, for reasons that are likely not attributable to a deficiency of the DFT methods. As such, we advise caution in the interpretation of computational results. Our results question to some degree the ability to obtain computational results that are generalizable to several substitution pattern and different reactions. As such, modelling of the relevant species is always recommended in conjunction with experimental effort for a thorough rationalization of new catalysts.