Journal of Crystal Growth 300 (2007) 277–283 The kinetics of parasitic growth in GaAs MOVPE A.J. Clayton, S.J.C. Irvine à Chemistry Department, UWB, Deiniol Road Bangor, Gwynedd LL57 2UW, UK Received 19 August 2005; received in revised form 16 November 2006; accepted 17 November 2006 Communicated by G. Muller Available online 21 December 2006 Abstract Gallium arsenide (GaAs) deposition was carried out in a horizontal quartz reactor tube with trimethylgallium (TMGa) and arsine (AsH 3 ) as precursors, using a hydrogen (H 2 ) carrier gas. Temperatures were in the range 400–500 1C, where surface reactions limit deposition rate. Nucleation time and deposition rate were monitored using laser interferometry, optimum reflectance was gained by aligning a quartz wafer to back reflect the incident beam. The 980 nm infrared laser beam was sufficiently long in wavelength to be able to penetrate the wall deposit. Results showing the effect of temperature and V/III ratio on the nucleation time and deposition rate are presented, where with temperature the nucleation delay was observed to reduce and the growth rate to increase. The nucleation delay is consistent with a thermally activated surface nucleation for the parasitic GaAs. A theoretical growth rate model, based on a restricted set of reaction steps was used to compare with the experimental growth rates. Without any free parameters, the growth rates from theoretical calculation and experiment agreed within a factor of two and showed the same trends with V/III ratio and temperature. The non-linearity of the theoretical growth rates on an Arrhenius plot indicates that there is more than one dominant reaction step over the temperature range investigated. The range of experimental activation energies, calculated from Arrhenius plots, was 17.56–23.59 kJ mol 1 . A comparison of these activation energies and minimum deposition temperature with the literature indicates that the wall temperature measurement on an Aixtron reactor is over 100 1C higher than previously reported. r 2007 Elsevier B.V. All rights reserved. PACS: 81.05.Ea; 82.30.b; 81.15.Gh; 81.15.Kk Keywords: A1. Growth kinetics; A1. In situ monitoring; A3. MOVPE; B2. GaAs 1. Introduction Metal organic vapour-phase epitaxy (MOVPE) is a well- established process for production of high-quality semi- conductor devices. In particular, light-emitting diode (LED) production using MOVPE has increased in recent years. The use of multi-wafer reactors has increased production and reduced cost in the MOVPE industry, making it an attractive technique for large-scale manufac- turing. However, loss of precursors to deposition at reactor walls has increased production costs and reduced yield. Bergunde et al. [1] investigated gallium arsenide (GaAs) MOVPE from trimethylgallium (TMGa) and arsine (AsH 3 ) in a planetary Aix-200/4 multi-wafer reactor, where the quartz top plate was cooled using a mixture of hydrogen (H 2 ) and Argon (Ar). The effect of cooling gas mixture on the growth rate, and type of parasitic deposit was investigated. A minimal parasitic deposit was achieved by optimising the H 2 /Ar mixture to 63% H 2 and 37% Ar. When Bergunde et al. [2] used a variable H 2 /Ar cooling gas mixture, the measured maximum growth rate of the parasitic growth was 0.19 nm s 1 . The minimum was 0.01 nm s 1 , where the top plate temperature was calculated to be 300 1C. A thermocouple was fixed at the outer ceiling wall 15mm from the inlet, in a different work by Bergunde et al. [3]. The temperature was measured at 240 1C, but no measurement of the inside ceiling wall was made. Weeks et al. [4] demonstrated that laser interferometry could be used to monitor top plate deposition on an ARTICLE IN PRESS www.elsevier.com/locate/jcrysgro 0022-0248/$ - see front matter r 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.jcrysgro.2006.11.311 à Corresponding author. Tel.: 44 1248 382383. E-mail address: sjc.irvine@bangor.ac.uk (S.J.C. Irvine).