Optical Approaches to Real-Time Analysis and Control of Crystal Growth
- PDF / 752,537 Bytes
- 11 Pages / 420.48 x 639 pts Page_size
- 77 Downloads / 141 Views
OPTICAL APPROACHES TO REAL-TIME ANALYSIS AND CONTROL OF CRYSTAL GROWTH
D. E. ASPNES Bellcore, Red Bank, NJ 07701-7040 ABSTRACT
A variety of optical methods are now available for studying surface processes and for monitoring layer thicknesses and compositions during semiconductor crystal growth by molecular beam epitaxy (MBE), organometallic chemical vapor deposition (OMCVD), and related techniques. Spectroellipsometry (SE) and spectroreflectometry (SR), the older, primarily bulk-sensitive probes, are now augmented by new, primarily surface-sensitive probes such as reflectance-difference spectroscopy (RDS), second-harmonic generation (SHG), and laser light scattering (LLS). Examples of real-time growth studies now include SE determinations of thicknesses and compositions of Al.Gal-xAs layers on GaAs by organometallic molecular beam epitaxy (OMMBE) to 10 A thickness scales, RDS determinations of surface dielectric anisotropy spectra of various (001) GaAs surfaces relevant to crystal growth by MBE, and LLS determinations of the evolution of surface roughness during chemical vapor deposition (CVD) growth on Si. Proven capabilities suggest new applications, particularly to growth-interrupted and metastable systems. INTRODUCTION The need to process information faster, cheaper, and more efficiently is placing new demands on crystal growth in the form of more stringent constraints on layer thicknesses, interface smoothness and abruptness, and dopant location, and of an increased need for heteroepitaxy. At the same time yields must be maintained at high levels to prevent costs from escalating. As a result, a substantial effort is now underway to develop a better understanding of growth mechanisms and improved methods of monitoring and controlling crystal growth in real time. As conventional surface-science probes are of limited use in real-time growth environments, interest is centering on optical approaches, which are intrinsically noninvasive and can be used in any transparent ambient. However, optical probes have had limited success in surface studies for two good reasons: accessible spectral ranges are narrow, and surface sensitivities are low. Common visible-near uv (v-nuv) quartz-optics systems only cover 2 octaves, from 1.5 to 6.0 eV. Absorption coefficients rarely exceed 106 cm- 1 (104 cm- 1 in the infrared). Consequently, optical penetration depths are rarely less than 100 A, so surfaces rarely contribute more than 1% to the total optical signal. Yet the emphasis on surface analysis has caused many to lose sight of the importance of obtaining information integrated over depth as opposed to information integrated over time. In the three years that we have been studying crystal growth with reflectance-difference spectroscopy (RDS), a surface-sensitive optical probe, the primary question that I have been asked is, how thick is my film? In fact, the technology that answers that question, ellipsometry, has been available for years and was even briefly applied a decade ago by Theeten, Hottier, and co-workers to monitor
Data Loading...
