Negative charge state of theDXcenter inAlxGa1−

Abstract

We present experimental data for the thermally activated capture of electrons on Si-induced impurity states in Al x Gaix As in the strong-lattice-relaxation regime {DX centers). Experiments have been performed after photoionization in the region of transition from the normal to the metastable state of the defect, using hydrostatic pressure up to 8 kbar. An analysis of the isothermal as well as the thermostimulated capture kinetics strongly supports the hypothesis of the negative charge state of the DX center.PACS numbers: 61.70. Tm, 71.55.Eq, 72.20.Jv The DX centers associated with tf-type substitutional impurities in many III-V semiconductors constitute a class of defects characterized by the dominant role of the lattice relaxation in capture and emission transitions. * It has been recognized that the substitutional impurity is, in itself, responsible for the existence of the defect. In spite of this apparent structural simplicity, and of the general nature of the defect, no definitive model for its description has been proposed to date. A point of particular controversy is the charge of the defect in the ground state: Is it neutral (DX°) or negatively charged {DX')! 2 -1Chadi and Chang's 2 pseudopotential calculations for substitutional donors in GaAs described a metastable resonant state associated with a large lattice relaxation. In the same work, they predicted that DX is a negatively charged center with a negative Hubbard correlation energy; i.e., in its ground state it binds two electrons. The lack of an EPR signal has been advanced as a confirmation of the negative-*/ character of substitutional donors in Al-Ga-As and Zn-Cd-Te by Khachaturyan, Weber, and Kaminska. 3 These authors also used the arguments of Toyozawa's self-trapping theory and Anderson's theory of the electron pairing due to strong lattice relaxation (see Ref.3) to show that the centers showing self-trapping and persistent photoconductivity are expected to possess a negative-^/ character. Various attempts were made to discriminate between the DX° and the DX~ model by interpreting mobility, 4 Hall density, 5 or capture 6 results, without conclusive success until now. In a recent paper, Khachaturyan et al. 1 showed that the DX centers are paramagnetic and concluded from their measurements that the ground state of the defect is neutral {DX°). Recently, Katsumoto et al} carried out similar experiments in Te-doped AUGaix As and found that the amount of paramagnetism was significantly less than that claimed by Khachaturyan et al. 7 In addition, Fockele, Spaeth, and Gibart 9 presented magneto-optical and optically detected magnetic-resonance data in Sn-doped Al x Gaix As, giving evidence for the existence of a paramagnetic singly ionized state Sn° which implies a negative ground state DX~.To discriminate between the DX° and DX ~ hypotheses, we study the thermally activated capture of electrons on Siimpurity states.The samples investigated were Si-doped Al x Gaix As layers, grown in a dedicated molecular-beam-epitaxy (MBE) setup used for indus...

DX centers in GaAs/Si-δ/AlAs heterostructure

“…Silicon impurity in GaAs as well as in AlGaAs seems to be the negative charge state responsible for the deep donor, which leads to a negative-U behavior. [9][10][11] However, low dimensional systems like Si-␦-doped GaAs have been shown that both neutral (DX 0 ), and charged (DX Ϫ ) DX-center models, contribute to the measured mobility changes. 12 In this letter we determined the microscopic mechanism of the formation of DX centers in GaAs/Si-␦/AlAs heterostructure.…”

DX Puzzle: Where Are We Now?

“…Jantsch et al [23,32] showed that the barrier height per one electron is two times smaller than the barrier per capture or emission event. Mosser et al [33] measured directly the capture efficiency showing that the capture energy varies with two Fermi energies during the capture transients. Both experiments were done for Si doped AlGaAs.…”