Very strong chalcogen bonding: Is oxygen in molecules capable of forming it? A First-Principles Perspective
Pradeep R. Varadwaj 1* Arpita Varadwaj1,2, Helder M. Marques3
1 Department of Chemical System Engineering, School of Engineering, The University of Tokyo 7-3-1, Tokyo 113-8656, Japan
2 The National Institute of Advanced Industrial Science and Technology (AIST), Tsukuba 305-8560, Japan
3 Molecular Sciences Institute, School of Chemistry, University of the Witwatersrand, 1 Jan Smuts Avenue, Braamfontein, Johannesburg 2050, South Africa.
* Corresponding authors address: pradeep@t.okayama-u.ac.jp
Abstract
There are views prevalent in the noncovalent chemistry literature that i) the O atom in molecules cannot form a chalcogen bond, and ii) if formed, this bond is very weak. We have shown here that these views are not necessarily true since the attractive energy between the oxygen atom of some molecules and several electron-rich anionic bases examined in a series of 34 ion-molecule complexes varied from the weak (ca –2.30 kcal mol-1) to the ultra-strong (–90.10 kcal mol-1). The [MP2 /aug-cc-pVTZ] binding energies for several of these complexes were found to be comparable to or significantly larger than that of the well-known hydrogen bond complex [FH···F] (~ 40 kcal mol-1). The nature of the intermolecular interactions was examined using the quantum theory of atoms in molecules, second-order natural bond orbital and symmetric adaptive perturbation theory energy decomposition analyses. It was found that many of these interactions comprise mixed bonding character (ionic and covalent), especially manifest in the moderate to strongly bound complexes. All these can be explained by an n (lone-pair bonding orbital) → σ* (anti-bonding orbital) donor-acceptor charge transfer delocalization. This study, therefore, demonstrates that the covalently bound oxygen atom in molecules can have a significant ability to act as an unusually strong chalcogen bond donor.
  1. IntroductionChalconide (Ch) oxygen is generally regarded as the least polarizable element in Group 16 of the periodic table.1,2Because of this, several studies have demonstrated that it does not form a chalcogen bond.3-7 (A chalcogen bond is formed when there is evidence of a positive site on the Ch atom in a molecule that interacts attractively with a negative site on a Lewis base in another molecule.8-15) The argument is that the oxygen atom in molecules is often negative and therefore does not feature a positive σ-hole on its electrostatic surface that can attract the negative site on a base.3-15 (A σ-hole16,17 is an electron density deficient region on the surface of the atom Ch along the outer portion of the R–Ch bond axis, where R is the remainder part of the molecule.18) While this may indeed be so in many cases (such as H2O and H2CO),8,9 it should not always be taken for granted. We have recently shown that when the oxygen atom is covalently bonded to an electron-withdrawing group X (X = F, Cl, Br) and –CN, the group draws the electron density to the bonding region and generates a weak to a moderately strong electron density deficient-region (σ-hole) on the surface of the atom lying opposite to the bond.8,9 The interaction energy (also called the binding energy) of putative 1:1 complexes formed by the O atom and the Lewis bases was found to be small (< 3 kcal mol-1).8,9 This has led to the interpretation that the aforementioned complexes may be regarded as being formed by van der Waals or weak interaction. These studies on O-centered chalcogen bonding have been recognized by others, both experimentally and theoretically.19,20 However, some have claimed that these complexes may not involve true “chalcogen bonding” since the intermolecular interactions in them are weakly bound and “chalcogen bonding” should be regarded as an “electrostatically driven” interaction – reminiscent of a debate that is very common in the area of halogen bonding.21,22 We counter this view by noting that there is no hard and fast rule in which a “weakly bound” or “van der Waals” complex cannot be regarded as truly “chalcogen bonded”. The minimal criterion for the recognition of a “Type-II chalcogen bond”11,23 is that there must be an attraction between the electrophilic region on the Ch atom and a negative site, and that the angle of approach of the electrophile should be such that ∠X⋯Ch–R = 140–180o.8-15We add further that the importance of weak interactions should not be underestimated.24,25 They appear in many different flavors,25-32 yet an understanding of their physical and chemical behavior in chemical systems to date is not complete. They are important factors, for example, not only in the theoretical and experimental design of drug33,34 and polymer network structures,35-37 but also for developments in the fields of crystal engineering38 and molecular recognition.24,39 The hydrogen bond between two H2O molecules is also weak, and is never stronger than about a twentieth the strength of the O–H covalent bond. Such bonds are structure determining, and have profound significance in fields as diverse as biology and materials science.40 The energy of a van der Waals interaction is very weak, only about 1 kcal mol−1, comparable with the average kinetic energy of a molecule in solution (approximately 0.4 kcal mol−1). It is significant only when many of them are combined so they contribute to the overall structure of a chemical system (as in interactions of complementary surfaces).41,42 Accordingly, and based on their energy of stability preferences, intermolecular interactions have been classified as van der Waals (energy < 1 kcal mol−1),43 weak (1–4 kcal mol-1), 43,44 moderate (4 –15 kcal mol-1),45 strong (15–40 kcal mol−1),44-46 very strong (40–60 kcal mol-1)47 and ultra-strong (>> 60 kcal mol-1).47-51 The first four have been recognized in many systems, while the last two have been identified in singly- and doubly charge-assisted composite systems, respectively.48,51-53Whereas thousands of studies have been reported centering discussion on the chemical physics and physical chemistry of halogen bonding and chalcogen bonding interactions between molecules containing heavy atom donors, the exploration of the chemistry of O-center chalcogen bonding from a theoretical modeling perspective is very limited. To this end, we report here an investigation of the structure, energy, electronic, orbital, and topological properties of 34 ion-molecule complexes formed by the attractive engagement between the positive sites (positive σ-holes) on the O atoms of some O-containing molecules and a series of anions. Analogous molecule-anion complexes, formed by hydrogen- and halogen-bonds are well known and hundreds of structures featuring these have been deposited in the Cambridge Structure Database.54 Kumar and coworkers, for instance, have examined the robustness of stable benzylic selenocynates for halide ion recognition in the solid-state and in solution.55 Their XRD analysis of various cocrystals reveals that the NCSe···X systems are driven by structurally important Se···X⁻ (X = Cl, Br, I) chalcogen bonds. Similar studies have been reported by ohers.56,57Similarly, Galmés and coworkers have recently performed a combined Cambridge Structural Database and theoretical DFT study of charge assisted chalcogen bonds involving sulfonium, selenonium, and telluronium cations in which divalent chalcogen atoms typically have up to two σ‐holes and form up to two chalcogen bonds; the same holds for tetravalent chalcogens which adopt a seesaw arrangement.58 Analogous studies have been reported elsewhere.59 However, the complexes examined in this study are uncommon; they are promoted by O-centered chalcogen bonding. The results of the ab initio first-principles MP2 method60 show that the binding energy of many of these complexes can be unusually high, comparable to or greater than that of various halogen- and hydrogen-bonded systems already reported in the noncovalent chemistry literature. In addition, we used the results of symmetry adapted perturbation theory (SAPT),61,62 the second-order perturbative estimates of ’donor-acceptor’ (bond-antibond) interaction energies in the natural bond orbital (NBO) basis,63 and the quantum theory atoms in molecules (QTAIM)64 to show that the intermolecular interactions responsible for the formation of the ion-molecule complexes contain appreciable covalent character. Of course, this adds to their inherent ionic character, which can well be rationalized by the Coulomb’s law.
  2. Chemical model systems and computational details
The binary complexes of OX2 (X = F, Cl, Br, CN) with the anions A (A = F, Cl, Br, CN, Br3, SCN, NCO, NO3) were fully energy minimized using MP2 (fc), in conjunction with the aug-cc-pVTZ basis set; the reliability of this and other theoretical approaches to study noncovalent interactions has been discussed elsewhere.8,9,65-67 The calculation of the Hessian second derivative of the energy with respect to the fixed nuclear coordinates of the atom was performed for all cases to ensure that a true minimum was found; all eigenvalues were found to be positive, and the structures reported here are not transition states. All calculations were formed using Gaussian 09.68
The nature of electrostatic surface of each OX2 molecule was examined using the popular molecular electrostatic surface potential (MESP) approach.69 As has been done elsewhere,8,69 the 0.001 a.u. isodensity envelopes of these molecules were used on which to compute the potential. The local maxima and minima of potential, often referred to asVS,max and VS,min , respectively, were used to identify the positive and negative regions, respectively. It should be kept in mind that the sign ofVS,max and VS,min is not always positive or always negative. The positive/negative sign associated with these two properties depends on the nature of the nucleophilicity/electrophilicity of a specific region on an atom or fragment in a molecule. Nevertheless, when VS,max> 0 (or VS,max < 0) on atom Ch along the outer extension of the R–Ch bond, it identifies a positive (or a negative) σ-hole.17 Similarly, whenVS,min > 0 (orVS,min < 0) on a specific region, it signifies an electrophilic (or nucleophilic) site. BothVS,max and VS,min were calculated using Multiwfn,70 and the MESP plots were generated using AIMAll71 with the wavefunctions generated using the [MP2/aug-cc-pVTZ] geometries. The counterpoise method of Boys and Bernardi was invoked to account for the effect of Basis Set Superposition Error (BSSE) on energy.72
QTAIM calculations were performed with [MP2/aug-cc-pVTZ] – a theory that relies on the zero-flux boundary condition to partition atomic domains in real space.64,73,74 The typical topological properties such as the gradient paths, bond paths, bond critical points (bcps) of the charge density (ρ b), the Laplacian of the charge density (∇2ρ b) and the total energy density (H b) were evaluated. In addition, the delocalization indices, δ , between various atom-atom pairs were evaluated for each system to gain insight into the covalent nature of the various bonding interactions involved.75,76
We examined the nature of the charge transfer delocalization energiesE2 between ”filled” (donor) Lewis-type NBOs and ”empty” (acceptor) non-Lewis type NBOs in several complexes – all within the second-order framework of NBO analysis (Eqt. 1).63 In Eqn. 1, qi is the donor orbital occupancy, εi andεj are diagonal elements (orbital energies) associated with each donor NBO (i ) and acceptor NBO (j ), respectively, and F(i,j) is the off-diagonal NBO Fock matrix element. These calculations were carried out within the Hartree–Fock (HF) level theory using Gaussian 09’s NBO Version 3.1.63