A Procedure for Computing Hydrocarbon Strain Energies Using Computational Group Equivalents, with Application to 66 Molecules †
Round 1
Reviewer 1 Report
I think this article is interesting from my point of view.
It is obvious that NMR and X-Ray studies are the best methods to determine the structure of compounds, but the calculations presented in this article can significantly save time and costs associated with the study of structures and intra and intermolecular interactions in organic compounds.
The author, using detailed group equivalents, provides a modernized version of the computational group equivalent approach for hydrocarbon strain energies.
The author uses computational methods for computing very accurate energies (W1, G4, CBS and M062X) and compares his results with 66 model hydrocarbons It is noteworthy, computational group equivalent approach proposed by the author may be used in other studies using other quantum-chemical methods, if necessary.
The results of these studies can be used to predict the structures of organic compounds and, consequently, to predict their chemical and biochemical applications.
I believe that the article may be appreciated by the readers. It is written clearly and briefly. Most of the information is arranged in tables.
I suggest publishing the article in its current form.
Author Response
Thank you for the positive review.
Reviewer 2 Report
The author has presented an extended set of 'group equivalents' for 5 high-level methods focused in the theoretical calculation of strain energies. The work is clear, wide, and deals acceptably well with other approaches. The work is interesting and deserves to be published in Chemistry after some minor details be corrected:
1) Remove from the Title 'with Numerous Examples'.
2) Begin, please, the introduction clearly defining 'Strain energy'.
3) In Table 2, something in methylstyrene is wrong. At least in the pdf version.
4) Despite the Table S1, and the text in lines 108-131, explain graphically in the main text with an additional scheme with colors the 'group equivalents' present in the three cases exposed : bicyclobutane, [2.1.1]-propellane and [4.4.4.4]-fenestrane.
5) Please, include in the Discussion some other interesting cases, such as spiro-compounds and also comment how this method deals with stereochemistry and aromaticity.
6) Extend a little bit the Conclusions with a better explanation of the goals and findings of this work.
Author Response
Thank you for the thoughtful review.
1) “With Numerous Examples” has been replaced with the more informative “with Application to 66 Molecules” (Page 1, line 4). It is important for the title to indicate that the paper provides not only a method, but also a modest compendium of calculated strain energies.
2) A few sentences have been added to the first paragraph of the introduction defining strain. (Page 1, lines 24-33). However, it is worth pointing out that strain is ultimately a somewhat fuzzy concept that defies precise definition, except through definition of a procedure for computing it (as is, for instance, done in this manuscript).
3) The misprinted “alpha” character in Table 2 has been fixed.
4) A graphic representation (Figure 2) has been added.
5) The primary purpose of the paper is to present a method. While there is doubtless much that is interesting to discuss, detailed analysis of all the cases is beyond the intended scope. However, a brief discussion of the spiropentanes has been added, as suggested, as well as brief discussions of the propellanes and fenestranes, all in the context of the ring strain additivity principle (page 11, lines 239-271).
It is not clear to me what is meant by how the method deals with stereochemistry or aromaticity. Whatever stereochemistry a compound has, the method works the same way. As for aromaticity, there is a stabilizing interaction, of course. That is inherent in the “CB” atom types for a benzene ring. For other sorts of aromatic rings, new parameters would need to be developed.
6) The goals and findings are really quite limited, and I believe are adequately described in the Conclusions section in the present form. To wit, a computational version of the Benson Group Equivalents method for computing a reference energy, against which to compute strain, has been developed, using modern electronic structure methods of the highest quality, and the approach is illustrated through calculation of strain energies for 66 compounds of general interest to physical organic chemists. However, I have also added one more sentence to the Conclusions, mentioning both the general agreement of calculated strain energies with those obtained using earlier methods, and also the more sensible results obtained than with previous methods for the propellanes (page 12, lines 330-332).
Reviewer 3 Report
The manuscript submitted by Rablen reports a modified computational strategy for the computation of hydrocarbon strain energies starting from Wiberg’s and Schleyer’s approaches based on group equivalents. The slight novelties of this work are the following: i) the author adopted the more detailed group equivalents defined by Benson and ii) he used highly accurate quantum mechanical methods to perform the calculations.
Other than considering the strategy presented in this work only a slight improvement compared to similar existing techniques, in my opinion, in its present form, the paper is not acceptable for publication. Anyway, the following points should be addressed before the paper can be reconsidered.
1) A part of the Introduction should be moved to a dedicated Theory section, particularly when the author discusses the approach they followed in their work. It could be useful to add some figures that will definitely help the reader to better understand how the method works.
2) The three examples given at the end of page 3 are not very clear at a first sight. They should be integrated in the Theory section (see point 1 above) and also illustrated with some figures, showing the global molecules under exam and the corresponding group equivalents taken into account. This will greatly improve the quality and the readability of the manuscript.
3) The current version for the “Materials and Methods” section is too short for a scientific paper reporting computational results. The author should give more details here. For example, he should specify here (and not in other points of the text) i) on which molecules he tested the method (and why he chose them) and ii) which levels of theory he decided to use (and why he used them). I am aware that this information is also given in other points of the text, but it should be specified also here and with more details.
4) At page 10 (lines 180-183), the author states that the estimated strain energy for norbornadiene is substantially lower than a previous consensus estimate. The author justifies this result by stating that this is probably due to the fact that “a proper estimate requires a reference value for carbon of the type C-(Cd)2(C)(H), but since such a value was not computed as part of the present study, the value for C-(Cd)(C)2(H) was used instead.” Since the author suspects what is the origin of the problem, he should also try that possibility to fully confirm his statement.
5) Page 1 - line 27. I do not understand the following sentence that should be probably rephrased: “Most ultimately rely on comparison to some notion of a “strain-free” reference system.”
6) Page 4 – Line 137. “We have purposely included” instead of “We have purposely include”.
7) Please delete all the exclamation marks in the manuscript (even in the dedication of the paper).
Author Response
Thank you for the thoughtful review, especially the suggestion in point #4.
1) I do not agree that a dedicated theory section makes sense; the approach is quite straightforward, and does not require detailed explanation, especially since such explanations can be found elsewhere, such as in the advanced organic chemistry textbook that is referenced.
2) A new figure (Figure 2) has been added (as reviewer #2 also suggested) to aid the reader in understanding the examples that show how the strain energy calculations are carried out.
3) The Materials and Methods section has been expanded to list what methods were used, why they were chosen, and what molecules were computed, as well as a few other details (page 3, lines 108-124). Also, a new Figure 1 has been added, showing graphically the structures of the molecules used to define the group equivalents.
4) The reviewer raises a good point, that the hypothesis is quite easily tested. A value for the carbon type C-(Cd)2(C)(H) has now been included, and the strain energy for norbornadiene accordingly recomputed (Additional line in Tables 2-4; changed values for norbornadiene in Table 5 and Figure 3). In fact, this modification only changes the computed strain energy by about 0.5 kcal/mol, and in the “wrong” direction. The discussion of this point has been completely changed. In fact, the value calculated here agrees very closely with the original strain energy estimate of Doering based on experimental data, but is lower than some of the more recent, mostly computational estimates (page 11, lines 266-271).
5) The statement has been expanded and, I hope, made clearer (page 1, lines 36-39).
6) The typographical error has been corrected (Table 2).
7) The exclamation points have been deleted (page 1, line 20; page 12, line 316).
Reviewer 4 Report
The author has shown how strain energies for hydrocarbons can be obtained through group equivalents derived from computational data on strain free hydrocarbons using 5 different methodologies. The strain energies for the different levels of theory are very similar, demonstrating the robustness of the procedure.
I would like the author to comment on the following:
What about non-next-nearest neighbor interactions, interactions between atoms separated by at least 2 atoms? They have been addressed by Benson and Cohen, and also Sabbe et al. have estimated the NNIs, besides Group Additive Values for the gas phase standard enthalpy of formation of hydrocarbons and hydrocarbon radicals using CBS-QB3 level of theory (J. Phys. Chem. A 2005, 109, 33, 7466-7480) and later also for the standard entropy and heat capacity of hydrocarbons and hydrocarbon radicals (J. Phys. Chem. A 2008, 112, 47, 12235-12251). Are all of your strain free reference molecules lacking these kind of interactions. Are all those interactions part of the strain energy? (Examples being 1,4-gauche interaction; single cis interaction; double cis interaction; ene-yne cis interaction; ortho interaction; tert- butyl-cis interaction; alkyl 1,5 interaction)
Maybe a comment on this matter can be added to the manuscript.
Finally, only one typo needs to be corrected:
p4, line 137: include => included
Author Response
Thank you for the thoughtful and positive review.
1) Non-next-nearest neighbor interactions have not been explicitly included. As the reviewer suggests, however, the reference cases for developing the group equivalents have carefully been selected to avoid such interactions wherever possible. The fact that the equivalents are based on the lowest energy single conformation makes such an approach practical. There are no alkane gauche interactions in the reference molecules assumed to be strain free (Figure 2), and no cis alkenes. There is likely some 1,3-allylic strain in 3,3-dimethyl-1-butene and in t-butylbenzene, as well as perhaps in 2-methy-1,3-butadiene and alpha-methylstyrene. That is not accounted for, and would presumably lead to a slight underestimate of strain energies when using the corresponding parameters. A paragraph has been added to the Discussion explaining these considerations (page 12, lines 299-313).
2) The typo has been corrected (Table 2).
Round 2
Reviewer 3 Report
The author has satisfactorily addressed all the points that I have previously raised about his manuscript. For this reason, now I can finally recommend the publication of the paper.
