Composition changes in bombarded oxides and carbides - the distinction between ballistic, chemically guided, and chemically random behavior

When oxides or carbides are bombarded it is commonly observed that new phases or compositions appear. Like ion beam mixing, the changes fall into categories which we term ballistic, chemically guided motion of defects, and chemically random motion of defects. An example of a ballistic change is a near-threshold process, as when 0 is lost from Ta2O5 or C is lost from TaC due to 1 keV He+ impact. Chemically guided defects are inferred when the changes correlate with thermodynamics (i.e. with the enthalpy increase), volatility, Gibbsian segregation, or defect-induced transport. For example, 0 loss from transition metal oxides is conventionally taken as chemically guided by virtue of correlating with thermodynamics. Likewise, C loss from SiC in some cases correlates with the C being transported to the surface and then being subject to sputtering. Chemically random processes, in which thermodynamics, etc. are not relevant, are believed to occur when oxides are bombarded with N+ or N2+. A typical example is found with TiO2, where O2+ impact leads to TiO2.00 and Ar+ impact to TiO1.66, the latter being equivalent to the Ti2O3 seen by electron diffraction. Finally, N2+ impact leads to TiO1.12N0.60, and since the sum O + N is similar to O alone after Ar+ impact we infer a random replacement of 0 by N. A similar pattern is shown by B2O3, Al2O3, SiO2, GeO2, ZrO2, V2O5, and Nb2O5. A somewhat different response is shown by SiC. Here N+ impact leads to a replacement of C by N but the sum C + N significantly exceeds the initial C content. Evidently the replacement is not stoichiometric but it still constitutes a clear example of chemically random behavior.