Advisor:	Paul R. Berger
Students:	Wei Gao (graduated with Ph.D. in September 1995)
Collaborators:	Matthew H. Ervin (Fort Monmouth Army Research Lab)
	Jagadeesh Pamulapati (Fort Monmouth Army Research Lab)
	Richard T. Lareau (Fort Monmouth Army Research Lab)
	Stephen Schauer (Motorola)

The In_0.53Ga_0.47As material system is very important for modulation-doped field-effect transistors (MODFET) used in monolithic microwave integrated circuits (MMIC) and avalanche photodiodes (APD), p-i-n photodiodes and metal-semiconductor-metal (MSM) photodiodes used in long wavelength (1.3 $\mu$ m $\leq$ $\lambda$ $\leq$ 1.55 $\mu$ m) lightwave communications. For these applications, the In_0.53Ga_0.47As material needs to be of high quality and low carrier concentration. Liquid phase epitaxy (LPE) is a well developed and proven technique to grow compound semiconductor materials. High quality layers are possible with LPE with a minimum of capital investment when compared to alternative techniques such as molecular beam epitaxy (MBE) and metalorganic chemical vapor deposition (MOCVD). LPE inherently provides high quality material since growth is at reduced temperatures and closer to chemical equilibrium, thus achieving near perfect crystallinity and purity. LPE remains a viable option for the production of optoelectronic materials and devices.

However, one problem with LPE growth is the unavoidable impurities which arise from the melt and growth ambient. It has now been recognized that Si and S are the two major background donor impurity species present in the In and InP, and that C is the principal acceptor, originating from the graphite boat. Various methods have been used to reduce the concentration of the residual impurities. Extended hydrogen baking of the In melt and the source solution (formed by adding InAs and GaAs) is widely known to lower the levels of both donor (Si and S) and acceptor (Al, Mg, Zn) impurities in the grown epitaxial layer. P. K. Bhattacharya et al. showed electron mobilities significantly improved upon hydrogen baking the entire melt after the In slug was individually prebaked with hydrogen. Also, Bagraev et al. used a sapphire boat to reduce the background doping and further enhance the electron mobility. But Bagraev et al. also discovered that the addition of trace amounts of rare earth elements to LPE growth melts could significantly reduce the background doping level and increase the electron mobility.

Recent studies have shown that impurities (for example, S, Se, Si, C, Te, O etc.) in III-V semiconductors form stable compounds with rare earth elements and that these rare earth compounds are insoluble in indium (In), the common solvent used in LPE growth of InGaAs. Thus, rare earth elements can be used to remove these impurities from the InGaAs LPE growth melts and reduce the background doping of grown epilayers. Also, it was found that the rare earth elements do not incorporate into the epitaxial layer at reduced growth temperatures, similar to those used in this study. Thus, the reader is cautioned that the LPE melt is doped with rare earth elements, but the epitaxial layer is not. The epitaxial layer will henceforth be referred to as undergoing a rare earth treatment. One dilemma of this study is the limited purity of commercially available rare earth elements (see Table IV). This leads to new impurities being introduced into the grown layers. The first rare earth elements studied were La and Ba by Haigh, and examples of more recent investigations of rare earth doped semiconductors concerning defects, doping, growth, theory and optoelectronics can be found in the Materials Research Society Proceedings.

High quality In_0.53Ga_0.47As epilayers have been grown on semi-insulating (100) Fe-doped InP substrates. The growths were performed by liquid phase epitaxy (LPE) using rare earth doped melts in a graphite boat. The rare earth elements studied were Yb, Gd and Er which act as gettering agents of impurities. Hall measurements show an elevated electron mobility for rare earth treated samples over undoped samples, $\mu_e$ =11,470 cm²/V $\cdot$ s at 300K and reduced carrier concentration (n-type), 9.33 $\times10^{13}$ cm^-3. The Hall results indicate an improvement in layer quality, but suggests that the treated layers are compensated. Photoluminescence (PL) studies show that the layers grown from rare earth doped melts have higher integrated PL efficiency with narrower PL linewidths than the undoped melt growths. The grown materials were fully characterized by Fourier transform infrared spectroscopy, double crystal x-ray diffraction, energy dispersive spectroscopy, secondary ion mass spectroscopy and deep level transient spectroscopy (DLTS). Compositional measurements reveal no measurable incorporation of rare earth elements into the grown epilayers. DLTS measurements indicate the creation of two deep levels with rare earth treatment, which is attributed to either the rare earth elements or impurities from within the rare earth elements. Subsequent, glow discharge mass spectrometry measurements revealed many impurities within the rare earth elements which preferentially might lead to p-type doping centers and/or deep levels. Thus, rare earth doping of LPE melts clearly improves epitaxial layer quality, however, the purity of commercially available rare earth elements hinders optimal results.

Paul R. Berger

Professor
Electrical and Computer Engineering
Physics

Director
Nanofabrication and Materials Processing Center (NanoMPC)
Nanoscale Patterning Laboratory
Nanoelectronics and Optoelectronics Laboratory (NOEL)
Polymer Device Laboratory (PDL)

Campus Address:
201 Caldwell Laboratory

Mailing Address:
Department of Electrical and Computer Engineering
The Ohio State University
205 Dreese Laboratory
2015 Neil Avenue
Columbus, OH 43210 USA

Direct phone: (614) 247-6235
EE Dept. FAX: (614) 292-7596
Email: pberger@ieee.org