How the Lewis Base F- Catalyzes the 1,3-Dipolar Cycloaddition between Carbon Dioxide and Nitrilimines.

The mechanism of the Lewis base F- catalyzed 1,3-dipolar cycloaddition between CO2 and nitrilimines is interrogated using DFT calculations. F- activates the nitrilimine, not CO2 as proposed in the literature, and imparts a significant rate enhancement for the cycloaddition. The origin of this catalysis is in the strength of the primary orbital interactions between the reactants. The Lewis base activated nitrilimine-F- has high-lying filled FMOs. The smaller FMO-LUMO gap promotes a rapid nucleophilic attack and overall cycloaddition with CO2.

The general use of small, highly abundant organic molecules, such as carbon dioxide (CO2), as building blocks in organic synthesis by activation and selective transformation to useful chemicals is highly attractive from both an economic and societal point of view.[1] Carbon dioxide is a desired feedstock for more complex hydrocarbon derivatives, since it is an abundant green-house gas that is cheap.[2] Chemists have engaged in significant efforts toward crafting methods to activate these otherwise unreactive molecules; however, it is in many cases a trial-and-error process, as the factors that play a role in the bond activation are not entirely understood.

In 2017, Lu and co-workers developed a convenient methodology to efficiently activate CO2 by F– (used as CsF/18-crown-6) and subsequently trap it with nitrilimines via a 1,3-dipolar cycloaddition to access 1,3,4-oxadiazole-2(3H)-ones (Scheme ).[3] In this reaction, F– first acts as a base to produce the nitrilimine intermediate (i.e., a 1,3-dipole) from hydrazonyl chloride. The nitrilimine then undergoes a cycloaddition reaction with CO2 (i.e., dipolarophile) to form the oxadiazolone. It was found that the presence of a base (e.g., amines or carbonates) alone does not facilitate efficient conversion to the product, but instead generates a large amount of the (undesired) dimerized dipole product.

Scheme 1

Proposed Mechanism of the Fluoride-Catalyzed 1,3-Dipolar Cycloaddition between CO2 and Nitrilimines by Lu and Co-workers, in Which the Lewis Base F Together with CO2 Forms the Activated Dipolarophile (i.e., CO2F);[3] R1 and R2 = Aryl or Alkyl

The authors proposed the formation of the known and stable fluorocarbonate (CO2F–),[4] formed from CO2 and F–. Fluorocarbonate has been suggested to enhance the reactivity of CO2 in some reactions (Scheme ).[5] The use of simple Lewis bases (e.g., F–) as catalysts for the activation of small molecules, which allows for efficient transformations under mild conditions, is of high interest.[6] Merino and co-workers previously investigated this reaction computationally and provided evidence for an operative mechanism based on the analysis of a number of possibly competing potential energy surfaces; however, they did not comment on the origin of the catalytic effect of the F– Lewis base.[7] Lu and co-workers experimentally showed that a base alone does not efficiently facilitate the cycloaddition reaction, so that F– is playing another role, presumably as a catalyst.[3] In contrast to the work of Merino et al., this indicates that F– does not solely act as a base. It is widely accepted that these 1,3-dipoles are readily formed from the hydrazonyl chloride, even in the presence of weak bases.[8] We have revisited the mechanism of this formal 1,3-dipolar cycloaddition between CO2 and nitrilimine to identify the catalytic role of the F– in this transformation.

Three possible reaction pathways (Scheme ) have been investigated, using state-of-the-art DFT calculations, to unravel the physical mechanism behind the Lewis base F– activation in 1,3-dipolar cycloaddition reactions: (i) the uncatalyzed cycloaddition, (ii) the activation of the dipolarophile, by the generation of CO2F– by F–, followed by the cycloaddition, and (iii) the activation of the dipole by the addition of F– to nitrilimine 1 forming activated nitrilimine–F–2, followed by the cycloaddition. In order to pinpoint the actual role of F– in lowering the reaction barrier, we selected the model system depicted in Scheme . We found that the inclusion of Cs+ and 18-crown-6, as Merino and co-workers did, slightly raises the reaction barrier due to the fact that F– is less Lewis basic because of the interaction with the Cs+ (Supporting Information Figure S1), while the overall mechanism remained unchanged.

Scheme 2

Possible Reaction Pathways for the 1,3-Dipolar Cycloaddition between Carbon Dioxide and Nitrilimine 1 in the Presence of the Lewis Base F

To identify the origin of the catalytic effect of the Lewis base, we employed the distortion/interaction–activation strain model[9] in combination with Kohn–Sham molecular orbital (KS-MO)[10] theory and energy decomposition analysis (EDA).[11] This methodological approach facilitates the analysis of the potential energy surface and, more importantly, the activation barrier, by decomposing the total energy of the system into chemically meaningful and easily interpretable terms and has been used by us to study other related 1,3-dipolar cycloadditions.[12]

The reaction profiles of the three reaction pathways of the studied 1,3-dipolar cycloaddition between carbon dioxide and nitrilimine 1 in the presence of the Lewis base F–, as well as their key transition state structures, are shown in Figure . The uncatalyzed pathway (i.e., pathway I; black) follows a concerted cycloaddition reaction with a barrier of ΔG‡ = 22.7 kcal mol–1 leading to product 3 (Figure a). The transition state, I-TS, is highly asynchronous, but is still concerted, with a C···N distance of 1.57 Å and a C···O distance of 2.33 Å (Figure b). This relatively high computed reaction barrier is consistent with the experimentally observed dimerization of nitrilimine 1 in the absence of F–, a process that goes with a more favorable barrier of ΔG⧧ = 20.1 kcal mol–1 (see SI Figure S2).[3]

Figure 1

(a) Reaction profiles (ΔGtoluene in kcal mol) for the three reaction pathways of the studied 1,3-dipolar cycloaddition between carbon dioxide and nitrilimine 1, including (i) uncatalyzed cycloaddition (black), (ii) formation of CO2F followed by stepwise cycloaddition (blue), and (iii) formation of 2 followed by stepwise cycloaddition (red), computed at SMD(toluene)-M06-2X-D3/def2-TZVP; TS = transition state and INT = intermediate. (b) Key transition state structures with key bond lengths (in Å) for the three reaction pathways.

Pathway II (blue) begins with the exergonic (ΔGrxn = −18.9 kcal mol–1) formation of CO2F– from CO2 and F–. Coodination of F– to CO2 is driven by stabilizing covalent and electrostatic interactions and induces a buildup of electron density on the oxygens of CO2F– compared to CO2 (see SI Table S1 and Figure S3). This leads, in contrast to pathway I, to a stepwise mechanism, whereby addition of the oxygen of the CO2F– to the electrophilic imine carbon center in II-TS occurs first and leads to II-INT (Figure b). This is the rate-limiting step of pathway II and goes with a reaction barrier of ΔG⧧ = 16.2 kcal mol–1, which is 6.5 kcal mol–1 lower than the uncatalyzed cycloaddition. Next, II-INT undergoes ring closure via a near barrierless pathway through II-TS2 (ΔG‡ = 1.2 kcal mol–1). Formation of product 3 and F– from II-INT2 is endergonic; however, the F– transfer, either to a second molecule of CO2 leading to 3 and CO2F– or to a second nitrilimine leading to 3 and 2, is exergonic. Lastly, pathway III (red), which like pathway II proceeds via a stepwise process, first begins with the barrierless formation of a Lewis base (or nucleophilic catalyst) activated nitrilimine–F– (i.e., 2) by the coordination of F– on the electrophilic nitrile carbon of nitrilimine 1 (see SI Figure S6). Formation of 2 is highly exergonic (ΔGrxn = −34.2 kcal mol–1) and is more than twice as favorable compared to the formation of CO2F– due to even more stabilizing covalent and electrostatic interactions (pathway II; blue, ΔΔGrxn = −15.3 kcal mol–1; see Table S1). The imine nitrogen of nitrilimine–F–2 exhibits an increased electron density, compared to dipole 1 (see SI Figure S3), and can engage in an efficient addition to CO2. Activated nitrilimine–F–2 can then proceed through transition state III-TS with a very low barrier of only 6.6 kcal mol–1, resulting in III-INT. This intermediate then undergoes intramolecular ring formation, through III-TS2 with a barrier of 14.3 kcal mol–1, to then form product 3 and regain the Lewis base F–. We have also computed pathway III with the corresponding Cl– adduct of 1, which is a similar intermediate to that proposed by Merino and co-workers,[7] and we found that F– adduct 2 follows a lower energy pathway in the reaction with CO2 (see Supporting Information, Schemes S1–S3 and Figure S1).

In order to gain quantitative insight into the physical factors why Lewis base catalyzed (i.e., nucleophilic-catalyzed) reaction pathway III is highly favored over the uncatalyzed pathway I, we turned to the distortion/interaction–activation strain model (D/I-ASM).[9] The D/I-ASM decomposes the electronic energy (ΔE) into two distinct energy terms, namely, the strain energy (ΔEstrain) and the interaction energy (ΔEint). The strain energy results from the deformation of the individual reactants and the interaction energy accounts for all chemical interactions between the deformed reactants along the reaction coordinate, defined, in this case, as the forming N···C bond.[12] As previously discussed, pathway I goes with a reaction barrier of 22.7 kcal mol–1, while the first step of pathway III proceeds with a barrier of only 6.6 kcal mol–1. Figure a shows the D-I/ASM analysis of pathway I (black) and pathway III (red). The origin of the lower barrier, in terms of electronic energy (trends are consistent for ΔE‡ and ΔG‡) for Lewis base catalyzed pathway III, can be traced exclusively to a more stabilizing interaction energy, while having a comparable strain energy.

Figure 2

(a) Distortion/interaction–activation strain model analysis; and (b) energy decomposition analysis of the cycloaddition reaction of 1 (black) and 2 (red) with CO2 (transition states indicated with a dot).[13] (c) Frontier molecular orbital diagram of the most important FMO-LUMOCO2 orbital interaction with the calculated energy gaps, orbital overlaps, and the S2/Δε terms, at consistent geometries with a N···C bond length of 2.14 Å. Computed at SMD(toluene)-M06-2X-D3/def2-TZVP using autoDIAS[14] for (a) and M06-2X-D3/TZ2P//SMD(toluene)-M06-2X-D3/def2-TZVP using PyFrag[15] for (b,c).

The interaction energy between the deformed reactants can be further analyzed in terms of quantitative Kohn–Sham molecular orbital theory (KS-MO)[10] together with a canonical energy decomposition analysis (EDA).[11] The EDA decomposes the ΔEint into the following four physically meaningful energy terms: electrostatic interactions (ΔVelstat), (steric) Pauli repulsion (ΔEPauli), orbital interactions (ΔEoi), and disperision interactions (ΔEdisp). The EDA (Figure b) shows that the more stabilizing ΔEint for the Lewis base catalyzed pathway III originates from a significantly more stabilizing ΔEoi. The ΔVelstat is slightly more stabilizing and the ΔEPauli remains almost unchanged. The origin of the more stabilizing of the Lewis base catalyzed pathway can be analyzed and explained by means of a Kohn–Sham molecular orbital analysis (KS-MO). We have quantified the key occupied–unoccupied orbital interaction between the FMO of 1 and 2 and the antibonding unoccupied orbital of CO2 at consistent geometries with a N···C bond length of 2.14 Å (Figure c). The stronger orbital interaction of the Lewis base catalyzed pathway could be traced back to the smaller FMO energy gap for the normal electron demand (NED) orbital interactions with the LUMOCO2. This originates from the much higher energy of the filled FMOs of 2 (i.e., nitrilimine–F–), compared to 1 (i.e., uncatalyzed pathway). The HOMO of 2 also shows a similar trend, but the overlap is much lower since the orientation of this orbital is nearly orthogonal to the LUMOCO2 (see SI Figure S4). In all, the presence of the negatively charged Lewis base F– in 2 causes a significant negative external potential that destabilizes the FMOs (see SI Figure S5). Additionally, the FMOs, especially the HOMO–1, are further destabilized as a result of (steric) Pauli repulsion with the filled FMOs of F– (see SI Figure S5).

In conclusion, we have investigated the Lewis base, nucleophile, F– catalyzed 1,3-dipolar cycloaddition between CO2 and nitrilimine. In contrast to the previous proposed mechanism,[3] we find that the reaction actually proceeds via the addition of the Lewis base F– to the dipole (i.e., nitrilimine), thereby activating the dipole, which rapidly engages in nucleophilic attack and overall cycloaddition with CO2 (Scheme ). Our distortion/interaction–activation strain analysis revealed that the mechanism behind the Lewis base catalysis was driven by the more stabilizing interaction energy between the reactants. This could be traced back to the stronger normal electron demand (NED) orbital interactions, as a result of the higher-lying donor orbitals of 2 (i.e., nitrilimine–F– species). This leads to smaller NED energy gaps and, thus, more stabilizing orbital interactions with the LUMO of CO2. F– destabilizes all FMOs of 2 by (i) the presence of a negative potential of the anion and (ii) the Pauli repulsion between the filled FMOs of the nitrilimine and F–. In all, this showcases the potential of Lewis base catalyzed small molecule activation, in which one can tune the reactivity of the reactants by the Lewis base.

Scheme 3

Novel Mechanism Emerging from Our Study for the Lewis Base F Catalyzed 1,3-Dipolar Cycloaddition of CO2 to Nitrilimines, Where F Activates the Dipole, Instead of the Dipolarophile

Methods

Computational Details

Conformer searches were performed using Grimme’s CREST 2.7.1[16] using default settings and toluene as solvent. DFT calculations were performed using Gaussian 09 Rev. D.01[17] employing the M06-2X density functional[18] in combination with the def2-TZVP[19] basis set. Solvent effects were included by using the SMD model[20] as implemented in Gaussian with toluene as a solvent. Empirical dispersion was included using Grimme’s D3 model[21] without additional dampening as proposed by Grimme and co-workers. Quasi-harmonic correction[22] was applied to all frequencies by raising all vibrations below 100 cm–1 to 100 cm–1. All computed stationary points have been verified by performing a vibrational analysis calculation, to be energy minima (no imaginary frequencies) or transition states (only one imaginary frequency). The character of the normal mode associated with the imaginary frequency of the transition state has been inspected to ensure that it is associated with the reaction of interest. The potential energy surfaces of the studied cycloaddition reactions were obtained by performing intrinsic reaction coordinate (IRC) calculations. The distortion/interaction–activation strain model (D/I-ASM)[9] was performed by the use of autoDIAS,[14] followed by an energy decomposition analysis (EDA)[11] in the gas phase, based on the solution PES, using PyFrag[15] and the Amsterdam Density Functional (ADF2018.106) software package at M06-2X-D3/TZ2P.[23] The optimized structures were illustrated using CYLview.[24]

Distortion/Interaction–Activation Strain and Energy Decomposition Analysis

The distortion/interaction–activation strain model[9] is a fragment-based approach in which the potential energy surface (PES) can be described with respect to, and understood in terms of the characteristics of, the reactants. It considers the rigidity of the reactants and to which extent they need to deform during the reaction plus their capability to interact with each other as the reaction proceeds. With the help of this model, we decompose the total energy, ΔE, into the strain and interaction energy, ΔEstrain and ΔEint, respectively [eq ].In this equation, the strain energy, ΔEstrain, is the energy required in order to deform the reactants from their equilibrium to the geometry they adopt over the course of the reaction. On the other hand, the interaction energy, ΔEint, accounts for all the chemical interactions that occur between these two deformed reactants along the reaction coordinate.

The interaction energy between the deformed reactants can be further analyzed in terms of quantitative Kohn–Sham molecular orbital (KS-MO)[10] theory together with a canonical energy decomposition analysis (EDA).[11] The EDA decomposes the ΔEint into the following three energy terms [eq ]:Herein, ΔVelstat is the classical electrostatic interaction between the unperturbed charge distributions of the (deformed) reactants and is usually attractive. The Pauli repulsion, ΔEPauli, includes the destabilizing interaction between the fully occupied orbitals of both fragments due to the Pauli principle. The orbital interaction energy, ΔEoi, accounts for, among others, charge transfer between the fragments, such as HOMO–LUMO interactions. Finally, the ΔEdisp term accounts for the interactions coming from disperion forces. In the herein presented distortion/interaction–activation strain and accompanied energy decomposition diagrams, the energy terms are projected onto the forming bond (N···C) distance. This critical reaction coordinate undergoes a well-defined change during the reaction from the reactant complex via the transition state to the product.[12]

Voronoi Deformation Density

The atomic charge distribution was analyzed by using the Voronoi Deformation Density (VDD) method.[25] The VDD method partitions the space into so-called Voronoi cells, which are nonoverlapping regions of space that are closer to nucleus A than to any other nucleus. The charge distribution is determined by taking a fictitious promolecule as reference point, in which the electron density is simply the superposition of the atomic densities. The change in density in the Voronoi cell when going from this promolecule to the final molecular density of the interacting system is associated with the VDD atomic charge Q. The VDD atomic charge QA of atom A is calculated according to eq .So, instead of computing the amount of charge contained in an atomic volume, we compute the flow of charge from one atom to the other upon formation of the molecule. The physical interpretation is therefore straightforward. A positive atomic charge QA corresponds to the loss of electrons, whereas a negative atomic charge QA is associated with the gain of electrons in the Voronoi cell of atom A.