Theoretical Research On The Reactivity Of Solid Surface

Research on the subject of "interaction" between two bodies is of fundamental importance either in theoretical physics or chemistry. Within the framework of the latter field, the most general representation of "interaction" is the formation (or breaking) of chemical bond. Right after the concept of chemical bond was proposed, people started to think hardly how atoms were glued together to form molecules. The knowledge on the "Holy Grail" is still limited even after hundreds of years have passed. Many fundamental questions, such as what factors on earth determine the strength of chemical bond, are still largely unresolved and far from perfectly answered. However, things are not getting better since the arising of surface science in the latest decades, in which case, the chemical bonding are considerably more complex compared to the classic chemical bond in molecules. Exploring the accompanying new concepts/rules governing the behavior of the extended system would benefit both the in-depth understanding of chemical bond and the development of new conceptual models concerning interaction strength (the so-called "chemical reactivity theory") from the theoretical aspect, and the understanding of the behavior of certain solid catalyst and further rational design from the application aspect.Although the methods/tools regarding analysis of chemisorption bond has been well established a long time ago, systematic studies on different solid surfaces and different adsorbate/surface systems are severely scarce. Here we applied various analytical methods/tools based on the LCAO implementation of DFT, such as the population analysis, COOP/COHP and Wannier function analysis, etc. to unravel the nature of chemisorption bond and have obtained several intriguing new insights so far. Among these discoveries, three ones were listed below:Firstly, electrons in simple metals of "s" symmetry only could be well described by the nearly free electron model and play the role of electron donor when interacting with an adsorbate (ionic bond); electrons in the same system of "p" symmetry, however, are not that simple due to the orientation-dependent orbital overlapping, and are likely to form covalent bonds. Furthermore, the electronic structures of different simple metals (e.g., Na, Mg, Al) are hardly the same and this would make their reactivity differ from one another;Secondly, there are vital differences between the sp electrons of simple metal surfaces (denoted as "spM") and that of transition metal surfaces (abbr. "TM"). Specifically, the bonding between sp electrons in TM resembles typical covalent character, rather than metallic character in spM. The difference of electronegativity between spM and TM is solely determined by the different nature of s electrons. Besides, the electronegativity of s electron in TM is generally more positive than that in spM;Thirdly, the chemisorption bond on solid surface and in molecule involving the same species (such as H, F, O, etc.) are very similar from the perspective of extremely localized orbital, as revealed by the bond order analysis and maximally localized Wannier function analysis.The aforementioned new insights and comprehension of the existing chemical reactivity theories in molecules/solid surfaces are the MUST towards developing new conceptual models fit for the description of chemical reactivity of solid surfaces. Through an enlightening bonding-antibonding analysis based on the perturbation theory, we found that the relevant most important single electron states are those that are very close to Fermi level in energy, and the lesser contribution the state will make if its corresponding eigen-energy drops/rises further relative to the Fermi level. This finding leads us to filter the energy window by a weighting function centered at Fermi level. When the weighing function is the derivative of Fermi-Dirac distribution, the finally obtained picture is exactly the softness of solid surface at finite electronic temperature and deteriorates to the weighted (by DOS at Fermi level) local density of states at zero Kevin.In order to fully exposing the key idea of the as-mentioned refinement, we devised a new concept termed the "Fermi softness". The Fermi softness is capable of describing both the relative global activity of different solid surfaces and the relative local activity over the same species. More significantly, the local picture could even predict the adsorption geometry of small (or large) atom/molecule on complex solid surfaces.The Fermi softness, as we found, could partially describe the activity of both molecules and solid surfaces in a unified manner to a certain degree. But it’s far from the very unification of description of chemical interaction we demand. If we are "tempted" into a unified description of chemical bond from the conceptual perspective, we have to decouple it. But unlike conventional physical interactions, which can be quantified in strength by certain intrinsic property, such as the electrostatic force under given relative distance being determined by the electric charge and the gravity force by the mass, the chemical bond cannot be explicitly described by any known property of the participating species. However, we found there is likely to exist a hidden property related directly to the bond energy (EAB), a quantitative measure of the bond strength. Such a property is represented by a characteristic complex quantity R (=(?)εe1θ), termed chemical amplitude (abbreviated as "chemplitude") in the present work, which connects to EAB through a simple relationship EAB=|RA-RB*|2. This extraordinary principle has been verified by more than 800 typical bonds, including covalent bonds in molecules (discrete case), metallic bonds in alloys and adsorption bonds on metal surfaces (extended case). The principle can be rationalized as the quantum interference of two contemporary energy-releasing events, each with an energy amplitude RA and RB, whereby the energy factor ε and the phase factor θ of R can be understood as the energy "stored" in the valence electron (VE) and the space condensation (due to bonding) ratio y (defined as Δr/rα, with Ar being ra-rc, in which ra is the size of free state and rc is the corresponding contracted size in chemical bond) of VE of the given species, respectively.