Evidence for substitutional-interstitial defect motion leading toDXbehavior by donors inAlxGa

Metastability of Mn3+ in ZnO driven by strong d (Mn) intrashell Coulomb repulsion: Experiment and theory

“…In the case of the As antisite in GaAs studied in the past (the EL2 center), optical excitation is followed by a large displacement, exceeding 1Å, of the defect towards the metastable interstitial site. 8,9 A similar mechanism is operative also in the case of donors, which can acquire the DX configuration when a shallow donor captures an electron and becomes a deep one with a strongly localized electronic state in the band gap, [10][11][12][13][14] and in the case of native defects, 15 where the (meta)stability is responsible for quenching of doping efficiency. A metastable configuration can also consist in a breathing-like displacement of the surrounding host atoms.…”

DX Puzzle: Where Are We Now?

“…It might explain the inconsistency obtained by us when we tried to constuct "universal" Arrhenius plot (similar to that in Fig. 2) using the available experimental data for tellurium doped AlGaAs [26,29].…”

“…Originally, the authors interpreted the results as a "proof' of the negative U character of the DX-center claiming that the two stages correspond to photoemission of the first and of the second electron. New data which show strong alloy splitting of the emission energy, done also by Dobaczewski et al [26], indicate now a new possibility of interpretation of the old photoionization results [35]. A preliminary analysis shows that the characteristic family of transients can originate from alloy splitting of the top and of the ground state of the DX-center.…”

“…measured by Piotrzkowski et al [25] during a warming process, after pressure induced filling of the DX-states, can be described by common parameters c0e and Ee. The latest data of Dobaczewski et al [26], where a Laplace analysis of the DLTS transients [27] has been done, are not presented in this figure. All these results show a well-resolved alloy splitting effect, however, the latter measurements were performed with constant voltage technique which could lead to some ambiguity.…”

“…Firstly, its occurrence supplied the cucial arguments proving the correctness of the model of Chadi and Chang [14]. Recent papers report on four and seven components of the emission processes in AlGaAs samples doped with Si [24][25][26][27][28] and Te [26,29] respectively. Secondly, the variety of DX levels and activation energies resulting from alloy splitting increases the difficulties in treating of the experimental data.…”