Ion Implantation In III–V Compounds

Section: Methodsmentioning

confidence: 99%

An Abrupt Dopant Profile in GaAs Produced by Te Implantation

Shin

Park

1976

“…The experiment described here was developed considering discrepancies in past research (6) and the availability of ion implantation equipment (used in the experiment) to industry (7). Research conducted in the mid-1960s (8,9) has indicated that certain dopants could be ionimplanted into diamond, but the relative amounts of radiation damage and the apparently small amount of boron that was electrically active redirected most diamond research efforts toward other, more promising techniques (10,11). However, the results of recent experimental research indicate the implantation and electrical activation of typical III-IV dopants in diamond is possible (12,13).…”

mentioning

confidence: 99%

Characterization of the Ion Implantation and Thermal Annealing of Boron in <100> Diamond

Baginski

Davidson

1990

“…Implantation is considered to be an attractive alternative to diffusion for the creation of thin heavily doped n-type regions in GaAs (1)(2)(3)(4)(5)(6)(7)(8)(9)(10)(11)(12)(13)(14)(15). We have utilized ion implantation to fabricate high-low READ GaAs IMPATT diodes (6) and state-of-the-art Ka-band GaAs Gunn effect diodes (7).…”

mentioning

confidence: 99%

Annealing of Ion‐Implanted GaAs in a Controlled Atmosphere

Malbon

Lee

Whelan

1976

Implantation is considered to be an attractive alternative to diffusion for the creation of thin heavily doped n-type regions in GaAs (1-15). We have utilized ion implantation to fabricate high-low READ GaAs IMPATT diodes (6) and state-of-the-art Ka-band GaAs Gunn effect diodes (7). However, the results reported to date indicate that, in general, the doping efficiencies of the n-type implanted species are low unless long annealing periods are used. Lengthy annealing times are unattractive because diffusion tends to excessively broaden the impurity doping gradients and thus impair device performance. Even for short anneal times within the range of commonly used annealing temperatures, 700~176it is necessary to take special precautions to minimize erosion of the GaAs surfaces. A variety of approaches have been used. These include sealing the GaAs both with and without a dielectric encapsulant in evacuated quartz ampuls prior to annealing (4, 8). The encapsulants have included SiO2 (3, 4, 6, 9), Si3N4 (8, 9), A12Oa (11, 12), and 14). However various problems associated with uncontrolled contamination and interdiffusion at the encapsulant-GaAs interface have been noted (15).This paper describes the results of substituting a controlled atmosphere for the dielectric encapsulant to protect the GaAs surface during high temperature anneals. Epitaxial materials technology has demonstrated that polished substrates can be maintained at elevated temperature with minimal surface degradation in a flowing high purity hydrogen atmosphere. Taking advantage of this experience, an anneal system was designed which utilizes a controlled atmosphere of high purity hydrogen (He) in conjunction with an arsenic (As) source to protect the GaAs surface during annealing. The system was used to anneal epitaxial GaAs films implanted with either silicon ions or sulfur ions to create thin, highly doped n-type layers. The epitaxial surfaces exhibited no degradation after anneals at 800~ for 20 min. The electrical results indicate that the apparent electrical conversion efficiencies achieved for the implanted layers were as high as 85% for the 800~ 20 min anneal. ExperimentalThe GaAs epitaxial material used in this investigation was grown by the He-Ga-AsCts vapor phase technique on heavily Te-doped n+-substrates oriented 3 ~ Off the <100> toward the <111>. The epitaxial layers were doped with sulfur and were electrically uniform in depth with net donor concentrations within the range 2-8 X 1015 cm -~. The epitaxial layers were implanted at room temperature with either 120 keV silicon ions or 140 keV sulfur ions. In the case of the silicon implants it was concluded after careful analysis of the separated ion spectrum that the implanted species was 28Si+ with less than 1% of other possible molecular components (e.g., N2+). To minimize channeling effects, the GaAs lattice was oriented to appear random to the incident ion beam. Under this condition, a first order approximation to the implanted ion distribution is Gaussian with a mean projected range Rp...