Light-induced structural changes

Reversible changes in the optical absorption are accompanied by reversible or irreversible changes in the refractive index [42,43], photoinduced changes in electrical and photoelectrical properties [44,45], volume [32,46], microhardness [47], glass transition temperature [48] and dissolution rate in various solvents [48,49,36], to name a few.

Figure 14: Photoinduced $\Delta n$ versus composition for as-evaporated As-based systems [43].

The totality of these changes has led investigators to the conclusion that the photoinduced changes in the optical absorbtion are caused by the changes in the structure. The first direct proof of structural changes was given by Tanaka (1975), who demonstrated that the "first sharp diffraction peak" in the XRD pattern underwent reversible changes on illumination an annealing. After it was proven by many techniques such as ESR, IR, XPS and Mossebauer spectroscopy. The most widely accepted model for explaining these suggested structural changes is model where is studied the concentration of hetero- and homonuclear bonds [28,29]. According to this model we can have two ideal boundary states.

1. Random Covalent Network - RCN, where the kind and number of surrounding atoms is totaly random and total number of As-S, As-As, S-S bonds depends only on chemical composition of system
2. Chemically Ordered Network - CON, where heterogenous bonds are strongly preferred and homonuclear bond are formed only as result of one element excess

This is two ideal boundary states are depicted on figure 15, where is shown computed number of bonds per atom as function of chemical composition.

Figure 15: Concentration of bonds as function of As-S system chemical composition [29]

The real system is never in one of these two ideal boundary states, it is always the linear combination of both. We just can shift it by illumination or annealing closer to RCN or to CON state. Extended X-ray absorbtion fine-structure (EXAFS) and Raman spectroscopies appear to be the most informative regarding the microscopic change in the structure of amorphous chalcogenides [50]. Both these together with IR spectroscopy results confirms the suggested model.