Carbon dioxide capture and functionalization by bis(N-heterocyclic carbene)-borylene complexes

Derivatives of free monocoordinated borylenes have attracted considerable interest due to their ability to exhibit transition-metal-like reactivity, in particular small molecules capture. However, such complexes are rare as the formation is either endergonic, or the resulting adduct is a transient intermediate that is prone to reaction. Here, we present the synthesis of two bis(N-heterocyclic carbene)-borylene complexes capable of capturing and functionalizing carbon dioxide. The capture and subsequent functionalization of CO2 by the bis(NHC)-disilylamidoborylene 1 is demonstrated by the formation of the bis(NHC)-isocyanatoborylene-carbon dioxide complex 3. Reversible capture of CO2 is observed using the bis(NHC)-mesitylborylene 2, and the persistent bis(NHC)-mesitylborylene-carbon dioxide adduct 4 can be stabilized by hydrogen bonding with boric acid. The reactions of 4 with ammonia-borane and aniline demonstrate that the captured CO2 can be further functionalized.


Reviewers' Comments:
Reviewer #1: Remarks to the Author: the manuscript submitted by Su, So and co-workers reported the synthesis of Bis(N-Heterocyclic Carbene)-Borylene Complexes and their application in CO2 capture and functionalization.Almost all borylene structures were isolated and well-defined by X-ray chromatograph.Especially, the key intermediate 3-bis(NHC)-isocyanatoborylene-carbon dioxide adduct, which is challenging, and has been isolated and characterized by X-ray chromatograph.Additionally, the electronic structure of 2 and 3 has been elucidated based on a DFT calculation, such as frontier orbital analysis, and bond order analysis.Meanwhile, the process of CO2 activation was also studied by computational study.Overall, this is a solid work, and this reviewer recommends its publication in this journal after the following minor revisions: 1. Regarding the choice of calculation level, especially for the study of the reaction mechanism of carbon dioxide activation, the selected B3LYP-D3(BJ)/def2SVP theroy of level is too low.It is not very suitable to use for the functional.M062X with a triple-zeta basis set, such as M06-2x/def2-TZVP, should be selected as the optimal choice for calculating the electronic energy.2. The isovalue of the orbital surface should be noted in the footnote or SI. 3. the DFT calculations should consider the solvent model or the author should note the gas phase or the solvent model.4 .The layout of the pictures in the article is a bit messy, and there are some small problems that should be avoided to increase the beauty of the article.For example, there are no spaces on either side of the equal sign in Figure 3 and so on.
Reviewer #2: Remarks to the Author: Professor So and coworkers have presented the synthesis of two new bis(carbene)borylenes and their reactivity with CO2.The first examples of boron bound CO2 featuring an η1 κC coordination mode are crystallographically characterized, and are of significant fundamental interest.The stoichiometric functionalization of borylene-bound CO2 with NH3BH3 and aniline is also presented.The work described herein is well supported by spectroscopic and crystallographic data.The crystallographic data is of good quality and is well-treated.The mechanism for the formation of the isocyanatoborylene CO2 adduct has been studied computationally and supports its isolation.
However, the mode of stoichiometric CO2 hydrogenation presented in this manuscript is well known with ammonia-borane in the presence of a strong base.Given that low valent boron, in the form of a diazadiborine, has recently been reported to catalyze the N-formylation of amines with CO2, I don't believe that the impact of this work rises to the bar of Nature Communications.However, I would enthusiastically support its publication another journal such as Angewandte Chemie or Chemical Science.
A few additional points to consider: In the mechanism, have the authors also considered elimination of hexamethyldisiloxane from Int2 to form and isocyanato-borylene intermediate, which then could in turn activate additional CO2?
The mimicry of CO dehydrogenase by compound 4 is somewhat overstated.Hydrogen bonding in carbonates is well known.
Minor corrections: Line 12-13 "...enabling the latter to be functionalized by disilylamido to form..." is somewhat confusingly worded.perhaps something like "..functionalized by he boron-bound disilylamido moiety to form.." would be clearer Line 80. Include esd on the dihedral angle Line 223.The word 'activates' is used to describe the scission of an O-H bond in the deprotonation of a carboxylic acid by a borylene.I would argue that this reaction likely deactivates these atoms toward further reactivity.
Reviewer #3: Remarks to the Author: The manuscript from Su, So and coworkers presents two highly nucleophilic borylenes with two NHCs.They then treat them with CO2, whereby the borylene attacks at the carbon atom of the CO2, an unprecedented reactivity pattern.In further reactivity, they hydrogenate and functionalize the CO2, both unknown reactions with borylenes.The synthesis and characterization appears to have been done competently.
The work thus presents the first examples of a long-awaited class of compounds (bis(NHC) borylenes), expected to be very electron-rich species.They also demonstrate likely the first true example of a monovalent boron atom acting as a strong nucleophile.As such, the work is a significant achievement, and I believe worthy of publication in Nature Communications.I have a few small points to raise before acceptance, however, as listed below.
-Introduction: some mention of conventional FLPs might be in order here, since the nucleophilic part of a conventional FLP attacks the C atom of CO2.That the boron is usually the electrophilic part of an FLP makes this work a nice example of "umpolung".
-The abstract doesn't seem to provide much insight and doesn't seem to fit with the journal's normal style.It is more full of compound formulae and compound numbers, rather than an insightful description of what is happening.
-The conclusions are similar to the abstract -lots of formulae and no insight.A summary of the advances (first bis(NHC) borylene, first single-site CO2 activation with boron, etc.) would be much more useful here than merely a summary of the reactions performed.
1 Dear Reviewers, Thank you for reviewing the manuscript.It has been revised according to the Author Checklist as well as the following Reviewers' comments.Please kindly refer to the manuscript file with tracking for your reference.
DFT calculations were performed to demonstrate that regeneration of compound 2 in the functionalization of the captured CO2 is feasible.The DFT calculations are shown in Figure 5b and Supplementary Fig. 39.The molecular structures of 4•B(OH)3, 5 and 6 are given in Figure 6.The Figure captions and manuscript are revised accordingly.

Reviewer 2
1. Comments: At several points in the text (line 20, 72, and 228), regeneration of 2 by further reaction of coordinated CO2 in 4 is implied, which in turn is used to claim the viability of 2 in catalytic transformations with CO2.It appears that this reasoning is based on speculation, and the authors have not presented evidence for the reformation of 2 in these reactions.In the absence of data or computational support, the authors should probably avoid speculating on this point, or use much softer language in doing so.The regeneration of 2 in these reactions is even mentioned in the abstract, despite the lack of evidence for it.Certainly, the strongest proof would be in the form of a report of compound 2's catalytic activity.

Reply:
We have removed the mention of regeneration of 2 in lines 20 and 70 but kept the one in line 227 in the revised manuscript.The mention in line 227 refers to a proposed mechanism, which is now supported by DFT calculations (Figure 5b, Supplementary Fig. 39).We have edited the sentences (lines 227 -232) relating to the mechanism to reflect this.Thank you for the suggestions.
3. Comments: 115 it seems a bit strange to say that a pi-acidic substituent is 'essential' to the stability of bis(NHC)borylenes and then present two examples to the contrary

Reply:
The two examples presented both feature π -acidic substituents, which are the Cym ligand and isocyanide ligand.The examples serve to demonstrate that in previously reported bis(NHC)borylenes with two weak π-accepting NHC ligands, the third ligand has to be π-electronic withdrawing.We have edited the phrasing in that paragraph (line 113 -115) to better explain this."essential" is removed in the revised manuscript.