Search for content and authors

Time-resolved differential transmission and photoluminescence studies of recombination mechanisms in Mg-doped InN 

Tim Veal 1Chito Kendrick 2Maurice Cheung 3Young-Wook Song 4Phil D. King 1Chris F. McConville 1Alex Cartwright 3Roger J. Reeves 4Steven M. Durbin 2

1. University of Warwick, Department of Physics, Gibbet Hill Road, Coventry CV4 7AL, United Kingdom
2. Department of Electrical and Computer Engineering, The MacDiarmid Institute for Advanced Materials and Nanotechnology, University of Canterbury, Christchurch, New Zealand, Christchurch, New Zealand
3. State University of New York, Department of electrical Engineering, Buffalo 14260-1900, United States
4. Department of Physics and Astronomy, The MacDiarmid Institute for Advanced Materials and Nanotechnology, University of Canterbury, Christchurch, New Zealand, Christchurch, New Zealand

  InN is a narrow-gap (~0.65 eV) semiconductor which has generated interest for its potential in applications ranging from photovoltaics to terahertz detectors to high-speed transistors. The realization of p-type InN through in-situ Mg doping has recently been reported by a number of different groups [1-3]. Two distinct and unrelated phenomena initially interfered with attempts to determine whether doping experiments yielded p-type conductivity: a high density surface electron accumulation layer [4], and unexpected quenching of the photoluminescence [1, 2]. Verification of p-type material has since been performed using a variety of techniques including electrochemical capacitance-voltage, valence band x-ray photoelectron spectroscopy, and variable magnetic field Hall effect. The origin of the apparent photoluminescence quenching, however, has remained an open question.  We have investigated recombination dynamics of Mg-doped InN epilayers grown by plasma-assisted molecular beam epitaxy using a combination of photoluminescence and time-resolved differential transmission (TRDT) measurements. The Mg concentration was determined by secondary ion mass spectrometry using an ion-implanted standard, and found to be in the range of 3×1017 to 1×1020 cm-3. Photoluminescence was performed using an argon ion laser and either an InSb or InGaAs detector, resulting in detectable signals from all samples. TRDT was performed at room temperature using a pump-probe technique on three separate samples, with doping densities of 6.2×1017 cm-3 (SL), 3.6×1019 cm-3 (SM) and 1.0×1020 cm-3 (SH), respectively.As has been reported by others, the photoluminescence of the films generally consisted of a single peak, with lightly doped films typically exhibiting a single feature near the bandedge. Quantum efficiency decreased markedly with increased Mg content, and many of the moderately doped films were characterized by a lower energy peak near 0.6 eV. For TRDT sample SL, the decay was single-exponential, as would be expected from a combination of radiative and Shockley-Read-Hall recombination. Samples SM and SH, however, were characterized by a different type of decay, indicating that a different recombination mechanism dominates. Sample SL, which in many respects has the same characteristics of undoped (n-type) InN, does not exhibit this different decay behaviour until higher excitation. Thus, Mg doping leads to a dramatic shift in the dominant recombination pathway, which has profound implications for devices.

[1] R. E. Jones, et al., Phys. Rev. Lett. 96, 125505 (2006).

[2] P.A. Anderson, et al., Appl. Phys. Lett. 89, 184104 (2006).

[3] X. Wang, et al., Appl. Phys. Lett. 91, 242111 (2007).

[4] I. Mahboob, et al., Phys. Rev. Lett. 92, 036804 (2004).


Legal notice
  • Legal notice:

Related papers

Presentation: Poster at E-MRS Fall Meeting 2009, Symposium A, by Tim Veal
See On-line Journal of E-MRS Fall Meeting 2009

Submitted: 2009-04-18 12:15
Revised:   2009-06-07 00:48