Semiconductors adsorption bonds

Structural parameters for atomic adsorption on the (110) and (1010) surfaces of zincblende and wurtzile structure compound semiconductors. The bond length is that of the anion-cation dimer in the first layer. D is the tilt angle in the top layer. The other parameters arc defined by fig. 16. 

Whereas the type of bonding in normal compounds is determined by the nature of the reactants, the types and relative amounts of adsorption bonds on semiconductors are governed by the presence of free electrons and holes, by the concentration of adsorbate on the surface, and by the nature and concentration of impurities in the bulk of the crystal. 

For rather ionic semiconductors, it appears that oxygen uptake is predominantly by electron transfer adsorption in which the electronic charge participating in the adsorption bond comprises electrons from the semiconductor conduction band. The principal evidence for this is the frequent observation that the kinetics for this tyipe of gas-solid interaction are governed by an Elovich rate law typical of a reaction whose rate is limited by the diffusion of electronic majority carriers through a Schottky-type barrierSemiconductivity modulation has also been observed to accompany this type of gas-solid interaction. 

If the above comparison of the properties of metal atoms with those of hydrogen deposited on the surface of a solid body (semiconductor) is correct, the effect of their adsorption on electric properties of semiconductor oxide films will be similar to features accompanying adsorption of hydrogen atoms. The atoms of hydrogen are very mobile and, in contrast to metal atoms, are capable of surface recombination resulting in formation of saturated molecules with strong covalent bond. 

The dangling and the surface ion-induced states are intrinsic surface states that are characteristic of individual semiconductors. In addition, there are extrinsic surface states produced by adsorbed particles and siuface films that depend on the enviromnent in which the siuface is exposed. In general, adsorbed particles in the covalently bonded state on the semiconductor surface introduce the danglinglike surface states and those in the ionically bonded state introduce the adsorption ion-induced surface states. In electrochemistiy, the adsorption-induced surface states are important. 

Fig. 6-68. Surface states created by oovsdently adsorbed particles on semiconductor electrodes BL = bonding level in adsorption = electron donor level D ABL = antibonding level in adsorption = electron acceptor level A W. = probability density of adsorption-induced surface state.

The second important difference is that the interface potential is present at the (outer) Helmholtz layer of the semiconductor/soiution interface. The interface potential is produced by surface dipoles of surface bonds as well as surface charges due to ionic adsorption equilibria between the semiconductor surface and the solution. If the interface potential can be regulated by a change in the chemical structure of the semiconductor surface, then the semiconductor band energies can be shifted to match the energy levels of the solution species (oxidant or reductant). This is another advantage of the semiconductor system because this enables improvement of the electron transfer rate at the semiconductor/soiution interface and the energy conversion efficiency. 

We assumed in Fig. 4.2 that no surface charge or surface dipole is present in the semiconductor. In general, however, both surface charges and surface dipoles are present in the semiconductor owing to adsorption equilibria for various ions between the electrolyte and the semiconductor surface as well as formation of polar bonds at the semiconductor surface. Such surface charges and surface dipoles change the potential difference in the (outer) Helmholtz layer and thus cause shifts in the surface band positions, as shown schematically in Fig. 4.3. The shifts can be expressed as changes in 0(0) or in the above equations, with the... 

All the inelastic scattering data discussed in this part of the review demonstrate the differences between H2 adsorption at a metallic and a semiconductor surface. They show the former preferentially occurs at multiply bonded sites located between the surface atoms whereas, in constrast, on MoS2 H2 adsorption is on top of a single atom. 

This is different at semiconductor surfaces where the covalent bonds between the substrate atoms are often strongly perturbed by the presence of adsorbates. This can result in a significant surface restructuring. Hence the dynamics of the substrate atoms has to be explicitly taken into account which of course increases the complexity of the modelling of the adsorption/desorption dynamics, as will be shown below for the H2/Si system. 

As already mentioned, in the case of semiconductor surfaces there is often a strong surface rearrangement upon adsorption due to the covalent bonding of the semiconductor substrate. The benchmark system for the study of the adsorption and desorption dynamics at semiconductor surfaces is the interaction of hydrogen with silicon surfaces [2, 61]. Apart from the fundamental interest, this system is also of strong technological relevance for the growth and passivation of semiconductor devices. 

For the elementary group IV semiconductor (001) surfaces there are in principle 3 plausible adsorption sites for ad-dimers on-top of the substrate rows with the dimer bond of the ad-dimer either parallel (A configuration) or perpendicular (B configuration) to the substrate dimer bonds and a third adsorption site in the trough with the dimer bond of the ad-dimer aligned parallel to the substrate dimer bonds (C configuration). For Ge/Ge(0 01) and Ge/Si(0 01) ad-dimers is also some intermediate configuration on-top of the substrate dimer rows labeled A/B. Whether this is really a stable adsorption site or just due to the fact that the ad-dimer rapidly rotates back and forth from A to B and vice versa is not yet settled. 

Mui et al.36 report a comparative experimental - theoretical study of amines on both the Si(001)-(2x 1) and the Ge(001)-(2x 1) surface. Both substrates were modeled by X9H12 (X = Si, Ge) clusters, utilizing DFT at the BLYP/6-31G(d) level of theory. For both, the Si and the Ge substrate, formation of a X-N dative bond (X = Si, Ge) is the initial step of the reaction between the considered amine species and the semiconductor surface. Flowever, while primary and secondary amines display N-H dissociation when attached to Si(001)-(2 x 1), no such trend is observed for the Ge counterpart of this system. This deviating behavior may be understood in terms of the energy barrier that separates the physisorption from the chemisorption minimum, involving the cleavage of an H atom. For dimethylamine adsorption, this quantity turned out to be about 50% higher for the Ge than for the Si surface. The authors relate this characteristic difference between the two substrates to the different proton affinities of Si and Ge. 

---