Halogen bonding, RAXÁÁÁB, and hydrogen bonding, RAHÁÁÁB, are electrostatically-driven noncovalent interactions of covalently-bonded halogens RAX and hydrogens RAH with negative sites B. A significant difference between halogen and hydrogen bonding is that the former is typically near-linear (the R-X-B angles are close to 180 ), whereas the latter is more likely to be nonlinear (R-H-B angles sometimes considerably less than 180). In this work, we have looked at the properties of several RABrÁÁÁB and RAHÁÁÁB complexes as functions of these angles. The differences in the directional tendencies in these interactions can be attributed to the presence of nonbonding valence electrons on the bromines, and their absence on hydrogen. We also found that for a given negative site, the halogen and hydrogen bonding interaction energies correlate very well with the positive electrostatic potentials created at it by RAX and RAH. This attests to the electrostatically-driven nature of these interactions. Overall, this study provides support for regarding both halogen and hydrogen bonding as subsets of r-hole interactions.

No abstract

Historically, two important approaches to the concept of electronegativity have been in terms of: (a) an atom in a molecule (e.g., Pauling) and (b) the chemical potential. An approximate form of the latter is now widely used for this purpose, although it includes a number of deviations from chemical experience. More recently, Allen introduced an atomic electronegativity scale based upon the spectroscopic average ionization energies of the valence electrons. This has gained considerable acceptance. However it does not take into account the interpenetration of valence and low-lying subshells, and it also involves some ambiguity in enumerating d valence electrons. In this paper, we analyze and characterize a formulation of relative atomic electronegativities that is conceptually the same as Allen's but avoids the aforementioned problems. It involves the property known as the average local ionization energy, I̅(r), defined as [Formula: see text], where ρi(r) is the electronic density of the i(th) orbital, having energy εi, and ρ(r) is the total electronic density. I̅(r) is interpreted as the average energy required to remove an electron at the point r. When I̅(r) is averaged over the outer surfaces of atoms, taken to be the 0.001 au contours of their electronic densities, a chemically meaningful scale of relative atomic electronegativities is obtained. Since the summation giving I̅(r) is over all occupied orbitals, the issues of subshell interpenetration and enumeration of valence electrons do not arise. The procedure is purely computational, and all of the atoms are treated in the same straightforward manner. The results of several different Hartree-Fock and density functional methods are compared and evaluated; those produced by the Perdew-Burke-Ernzerhof functional are chemically the most realistic.

In the present work, aged cotton linters have been analyzed for their chromophore content according to the CRI (''chromophore release & identification'') method. Despite the very low contents in the ppb range, nine chromophores have been unambiguously identified, which makes this account the first one on defined chromophoric structures isolated from cotton. A common feature of the chromophores are 2-hydroxy-[1,4]benzoquinone, 2-hydroxyacetophenone and 5,8-dihydroxynaphthoquinone moieties, which resemble chromophoric structures found in other cellulosic substrates, such as bleached pulps or fibers. The finding of these compounds in lignin-free cotton linters confirms the previous hypothesis that those chromophores are formed from (oxidized) carbohydrate structures rather than from lignin fragments.

In a continuing effort to further explore the use of the average local ionization energy [Formula: see text] as a computational tool, we have investigated how well [Formula: see text] computed on molecular surfaces serves as a predictive tool for identifying the sites of the more reactive electrons in several nonplanar defect-containing model graphene systems, each containing one or more pentagons. They include corannulene (C20H10), two inverse Stone-Thrower-Wales defect-containing structures C26H12 and C42H16, and a nanotube cap model C22H6, whose end is formed by three fused pentagons. Coronene (C24H12) has been included as a reference planar defect-free graphene model. We have optimized the structures of these systems as well as several monohydrogenated derivatives at the B3PW91/6-31G* level, and have computed their I(r) on molecular surfaces corresponding to the 0.001 au, 0.003 au and 0.005 au contours of the electronic density. We find that (1) the convex sides of the interior carbons of the nonplanar models are more reactive than the concave sides, and (2) the magnitudes of the lowest I(r) surface minima (the I S, min) correlate well with the interaction energies for hydrogenation at these sites. These I S, min values decrease in magnitude as the nonplanarity of the site increases, consistent with earlier studies. A practical benefit of the use of I(r) is that a single calculation suffices to characterize the numerous sites on a large molecular system, such as graphene and defect-containing graphene models.