Hydrogen is known to strongly affect the physical properties of amorphous semiconductors. Indeed hydrogen is introduced during the growth of amorphous silicon films, used in active matrix displays and solar cells, to passivate silicon dangling bonds and to relax the lattice thereby reducing the density of states in the energy gap by several orders of magnitude and giving rise to device grade material. Ideally, hydrogenated amorphous silicon (a-Si:H) is a continuous covalently bonded random network of silicon-silicon and silicon-hydrogen atoms, with the predominant nearest neighbour environment similar to that of crystalline silicon. a-Si:H typically contains about 10 atomic percent hydrogen.

Tritium can readily substitute for hydrogen in a-Si:H without altering the physicochemical properties of the material. Tritium decay leads to a change in the local bond structure of the material as helium detaches from bonds leaving behind dangling bonds. The decay rate of tritium and therefore the rate of dangling bond formation is determined by the half-life of tritium. Hence, tritium provides a unique avenue to dynamically study the effect of dangling bonds on the density of states in the energy gap and therefore on the optoelectronic properties of a-Si:H. Tritiated hydrogenated amorphous silicon (a-Si:H:T) was deposited using mixtures of tritium and silane gases in a dc saddle-field glow-discharge deposition system. The amount of tritium in the films was controlled by adjusting the relative flow of tritium and silane gases into the deposition chamber.

Photoluminescence, isothermal capacitive transient spectroscopy and constant photocurrent spectroscopy were used to measure defect concentration as a function of time in the films. The defect concentration was found to increase between 1 and 2 orders of magnitude, in about 300 hours. Thermal annealing decreased the defect concentration. It was found that tritium permits a study of the change in the density of defect states due to dangling bond formation in a-Si:H without the uncertainties introduced by the use of multiple samples.