The stability of covalent dative bond significantly increases with increasing solvent polarity

It is generally expected that a solvent has only marginal effect on the stability of a covalent bond. In this work, we present a combined computational and experimental study showing a surprising stabilization of the covalent/dative bond in Me3NBH3 complex with increasing solvent polarity. The results show that for a given complex, its stability correlates with the strength of the bond. Notably, the trends in calculated changes of binding (free) energies, observed with increasing solvent polarity, match the differences in the solvation energies (ΔEsolv) of the complex and isolated fragments. Furthermore, the studies performed on the set of the dative complexes, with different atoms involved in the bond, show a linear correlation between the changes of binding free energies and ΔEsolv. The observed data indicate that the ionic part of the combined ionic-covalent character of the bond is responsible for the stabilizing effects of solvents.


REVIEWERS' COMMENTS
Reviewer #1 (Remarks to the Author): This work reports results of fundamental research on the influence of solvents on the stabilities of compounds featuring dative bonds. High level theoretical calculations in combination with Raman and NMR spectroscopy experiments have been used to study the Me3NBH3 complex, which shows a surprising increase in stability as solvent polarity is raised. The authors argue, soundly, that the wavefunction for the covalent component of the dative bond does not change appreciably with increasing solvent polarity but that of the ionic portion does. Several other examples featuring different types of dative bonds are also reported in the study. This dependence is unusual and, given the widespread occurrence of dative bonds, can be consequential in determining functions of such bonds in solvent media. Publication is recommended after minor revisions noted below.
(a) Provide T1 diagnostic values for CCSD(T) calculations to check for multireference character in the wavefunction.
(b) The isotopically labeled 15N sample contains CD3 groups instead of CH3 in Scheme S1. But the caption in Figure S1 says 15N-labeled Me3NBH3. Please clarify the use of deuterium labeling to avoid confusion.
(c) Include methods and basis sets in Table/Figure captions. (d) In several places the language was a bit hard to follow so please reword for clarity where necessary.
(e) Minor point: The top panel in Figure S7 needs to have the formula format for Me3NBH3.
Reviewer #2 (Remarks to the Author): The manuscript of Lo et al. discusses the stability of dative bonds of molecules in solution with respect to the polarity of the solvent. The project nicely combines experiment with theory in order to prove the initial hypothesis of the authors, i.e. a more polar solvent results to a stronger dative bond. However, I am not so sure if this discovery justifies the publication of this work as a "communication". The authors discuss extensively a simple molecular system (Me3NBH3) at both experimental and computational level, and they extrapolate their finding with additional computations on other molecular species with variable complexity. Here are my comments that the authors should address before this article is considered for publication: -Since the authors have considered both ammonia borane and Me3NBH3 in this study, it would be interesting to examine what is the effect of the substituted species in the bond strength under different solvents. The discussion can be extended to the BH3 and BF3 species.
-In Table 4, is there any correlation between delta G of solvation and the delta delta G (difference between gas phase and in solution)?
-Have the authors used the Becke-Johnson dumping to the D3 dispersion correction?
-What is the reason of such a large deviation between the CCSD(T)/cc-pVTZ and the CCSD(T)/CBS that is mentioned in the Methods section?
-Please fix typo CCSDT(T) on page 13 Reviewer #3 (Remarks to the Author): The manuscript presents experimental and computational data elucidating the effect of solvation on the stability and the bond strength of dative bonds. The authors find that both stability and bond strength for dative bond increase with the dielectric constant of the solvent. This is in important finding that does not seem to known. It differs strongly from the behavior of normal covalent bonds. It is laudable that the authors make a clear distinction between the bond energy or stability and the bond strength which is independent of the fragments.
The paper should be published after a revision addressing these issues: (1) It is rightfully pointed out that "bond strength" cannot be determined directly, and is also not very well defined. They also point out that stability and bond strength do not necessary correlate. Unfortunately, the manuscript is not really clear on which measures are used to assess bond strength. One good measure is the force constant of the bond, or the frequency, which is found in Table 2. The NMR J-coupling is the first measure given in Tab. 1. The authors write (p.4) "It thus reflects chemical bonding". Do they mean: "reflects bond strength"? Is there further evidence for this relation? If not what conclusion can be drawn from the NMR data?
In the conclusion, the "intrinsic energy" is stated as measure for "bond strength". This is not elaborated on in the Introduction or Results section.
(2) The calculated data from Table 3 are very important for this work. Unfortunately, one can only  guess from the Methods section what calculations are behind Table 3.
(3) The English could be improved

Authors' replies to the Reviewers' remarks and questions
Referee: 1 This work reports results of fundamental research on the influence of solvents on the stabilities of compounds featuring dative bonds. High level theoretical calculations in combination with Raman and NMR spectroscopy experiments have been used to study the Me3NBH3 complex, which shows a surprising increase in stability as solvent polarity is raised. The authors argue, soundly, that the wavefunction for the covalent component of the dative bond does not change appreciably with increasing solvent polarity but that of the ionic portion does. Several other examples featuring different types of dative bonds are also reported in the study. This dependence is unusual and, given the widespread occurrence of dative bonds, can be consequential in determining functions of such bonds in solvent media. Publication is recommended after minor revisions noted below.

Reviewer's Point #1:
Provide T1 diagnostic values for CCSD(T) calculations to check for multireference character in the wavefunction.

Response:
T1 diagnostic values for CCSD(T) calculation are now reported in the revised manuscript. The T1 diagnostic values suggest the systems are predominantly single reference in nature. We have included the following sentence in the revised manuscript (section Methods, pg 7) "The T1 diagnostic value of 0.008, indicates that the complex is predominantly single reference in nature."

Reviewer's Point #2:
The isotopically labelled 15N sample contains CD3 groups instead of CH3 in Scheme S1. But the caption in Figure S1 says 15N-labeled Me3NBH3. Please clarify the use of deuterium labelling to avoid confusion.

Response:
We have corrected the caption in Figure S1.

Reviewer's Point #3:
Include methods and basis sets in Table/Figure

Response:
In the revised version we have incorporated methods and basis sets in Table/

Reviewer's Point #4:
3 In several places the language was a bit hard to follow so please reword for clarity where necessary.

Response:
The language was corrected by the native English-speaking chemist.

Reviewer's Point #5:
Minor point: The top panel in Figure S7 needs to have the formula format for Me3NBH3.

Response:
The formula format for Me3NBH3 is now incorporated in the top panel in Figure S7.

Referee: 2
The manuscript of Lo et al. discusses the stability of dative bonds of molecules in solution with respect to the polarity of the solvent. The project nicely combines experiment with theory in order to prove the initial hypothesis of the authors, i.e. a more polar solvent results to a stronger dative bond. However, I am not so sure if this discovery justifies the publication of this work as a "communication". The authors discuss extensively a simple molecular system (Me3NBH3) at both experimental and computational level, and they extrapolate their finding with additional computations on other molecular species with variable complexity.
Here are my comments that the authors should address before this article is considered for publication.

Reviewer's Point #1:
Since the authors have considered both ammonia borane and Me3NBH3 in this study, it would be interesting to examine what is the effect of the substituted species in the bond strength under different solvents. The discussion can be extended to the BH3 and BF3 species.

Response:
We have now included the discussion regarding the effect of the substituted species in the bond strength under different solvents in the current version of the manuscript. We have included the following sentence in the revised manuscript, pg 6. "The functionalization of the electron donor (NH 3 ) with electron-donating CH 3 group leads to a larger complex stabilization in both gas phase and o-DCB. This effect is more pronounced in the gas phase. The functionalization of the electron acceptor (BH 3 ) with electron-withdrawing F atoms destabilizes the complex in both environments." Reviewer's Point #2: