In this project, funded by the Chemical Structure, Dynamics and Mechanisms Program of the Chemistry Division, Professor Alexander Boldyrev of Utah State University will develop new theoretical tools to rationalize bonding, structure, stability, and reactivity of novel and unusual chemical species. The existing Adaptive Natural Density Partitioning (AdNDP) method applicable to closed-shell finite molecules will be advanced to include a wider variety of species. The new software will be written and used for deciphering chemical bonding in: 1) new open-shell pure and mixed main group and transition metal clusters; 2) chemical species upon transformation in chemical reactions; 3) biomolecules; 4) condensed phase materials. The broader impacts will involve incorporation of the AdNDP method into teaching computational and quantum chemistry to undergraduate and graduate students, as well as to high school students during the annual Utah State University High School Summer Internship. Teaching the AdNDP method is beneficial to students as it is the only tool capable of rendering a complete chemical bonding picture for molecules featuring both localized and delocalized bonding.
The newly developed AdNDP method for open-shell species, as well as solids, biomolecules and reaction intermediates will advance the ability to decipher chemical bonding in clusters, nanoparticles, solids, biomolecules, catalysts, as well as to study mechanisms of chemical reactions. The results obtained in these studies can have significant potential for future advancement of nanotechnology, catalysis, material science and biotechnology.
In general chemistry classes we were taught that atoms are bound to each other either through one pair of electrons (single two-center two-electron bond: 2c-2e bond), or through two pairs of electrons (double 2c-2e bond), or through three pair of electrons (triple 2c-2e bond). If an atom has extra electrons, which do not participate in chemical bonding, we say they form lone pairs. These are essential elements of the Lewis model of chemical bonding, which is a backbone of our description of chemical bonding in chemistry. For example, a water molecule has two 2c-2e O-H bonds and oxygen has two lone pairs. However, there are many cases, where Lewis model fails to describe chemical bonding, for example in aromatic molecules. In that case we invoke either to the resonance of a few Lewis structures or we use the delocalized bonds, which spread one or more electron pairs over more than two atoms (therefore we call them multicenter bonds). During our previous NSF grant (CHE-1057746, 2/1/10-1/31/13), we developed a theoretical method and the corresponding software, which for the first time allows to recover all classical bonding elements (lone pairs and 2c-2e bonds) as well as multicenter bonds This method is called Adaptive Natural Density Partitioning (AdNDP) and it is an extension of Natural Bond Orbital (NBO) analysis, popular in chemistry. Our AdNDP method is now widely used in chemical research and we aware that it is already used in classes at a few Universities. In the current proposal in collaboration with the professor J.R. Schmidt group at University of Wisconsin, Madison, we developed a Solid State Adaptive Natural Density Partitioning (SSAdNDP) method and the corresponding software. This method for the first time allows us to rationalize chemical bonding in solids and novel nano-materials such as one- (1D), two- (2D), and three-dimensional (3D) materials and nanoparticles. As an example of the efficiency of our SSAdNDP method we will briefly discuss chemical bonding in one of the 2D-materials, the so-called boron α-sheet (the most stable forms of the 2D-boron) and in solid crystal (3D-material) - Zintle phase Na8BaSn6. The SSAdNDP method allows us to explain a peculiar spotted structure of the α-sheet with empty and filed hexagons (Figure 1a). The overall SSAdNDP results are shown in Figure 1b. The SSAdNDP analysis revealed that α-sheet does not have any σ- or p- 2c-2e classical bonds. Instead, all bonding is muticenter. There are 8 boron atoms and 24 valence electrons per unit cell, thus we anticipate 12 two-electron bonds. Six 3c-2e s-type bonds were found on every boron triangle bordering a vacant hexagon. Three 4c-2e s bonds were revealed in the rhombi connecting two centered hexagons. Thus, nine electron pairs were found over three and four centers, leaving three more to be accounted for. Next we found a 6c-2e p-bond over the hexagonal hole and a 7c-2e p-bond over each centered hexagon in the unit cell. The bonding motifs found in SSAdNDP explains why a-sheet has this peculiar structure. With this chemical bonding for each B7 fragment, we have six valence electrons coming from three 3c-2e s-bonds, three electrons coming from three 4c-2e s-bonds, and two electrons coming from the 7c-2e p-bond, for a total of eleven electrons, while the filled hexagon has 12 electrons on this fragment. The "extra" electron, which is not needed for bonding in the field hexagon goes to form 6c-2e p-bond over empty hexagons. Those empty hexagons serve as electron scavengers of extra electrons. Thus, this spotted structure of α-sheet is no longer a mystery. The second example of the application of our SSAdNDP method is the Na8BaSn6 crystal. With two formula units per unit cell, this results in 42 valence electron pairs per unit cell (Figure 2a). The SSAdNDP analysis revealed 4 core electron pairs on each Ba atom, ON=2.0 |e| (not shown), one lone pair on each of the Sn atoms, and a 2c-2e s bond between every pair of Sn atoms around the pentagonal rings (Figure 2b). Three 5c-2e p bonds on each of the Sn5 moieties were also found, highlighting each ring’s aromaticity. The remaining six electron pairs could not be localized on the two non-ring Sn atoms, which would satisfy the Sn4- octet configuration of a purely ionic system. However, only the single s-type lone pair was found on each of the two atoms; no p-type lone pairs could be localized, even with low ON thresholds. We believe this is due to the metallic nature of these remaining electrons, making them intrinsically non-localizable. We think that our SSAdNDP method has a great potential in chemical research, including rational design of new materials, and in chemical education. This method and the software can be used in classes such as computational chemistry, quantum chemistry, advanced inorganic chemistry, nanotechnology, and material chemistry.
