• PDF / 147,857 Bytes
• 7 Pages / 612 x 792 pts (letter) Page_size
• 17 Downloads / 296 Views

DOWNLOAD

REPORT

---

B8.11.1

Phosphorus and Boron Implantation into (100) Germanium Y. S. Suh, M. S. Carroll1, R. A. Levy, A. Sahiner2 and C. A. King3 Materials Science and Engineering, New Jersey Institute of Technology, University Heights, Newark, NJ 07102 1

Agere Systems, Allentown PA 18109 (current address: Sandia National Laboratories, P.O. Box

5800, M.S. 1077, Albuquerque, NM 87185) 2

Evans East, East Windsor NJ 08520 (current address: Department of Physics, Seton Hall

University, South Orange, NJ 07079) 3

Noble Device Technologies, Newark, NJ 07103

ABSTRACT Boron and phosphorus were implanted into (100) Ge with energies ranging from 20-320 keV and doses of 5×1013 to 5×1016 cm-2. The as-implanted and annealed dopant profiles were examined using secondary ion mass spectrometry (SIMS) and spreading resistance profiling (SRP). The first four moments were extracted from the as-implanted profile for modeling with Pearson distributions over the entire energy range. The samples were annealed at 400, 600, or 800°C in nitrogen ambient. The dopant activation and diffusion were also examined and it was found that p-type sheet resistances immediately after boron implantation as low as 18 ohms/sq could be obtained without subsequent annealing. INTRODUCTION Recent advances in process technologies, such as high-κ gate dielectrics used for high mobility Ge MOSFETs [1] and relaxed graded buffer layers for fabrication of high quality Ge photodetectors on Si [2], have created renewed interest in device fabrication with Ge. Implantation and dopant activation are crucial steps for these device structures, for which critical modeling parameters that are typically available for Si fabrication (e.g. Pearson’s implant parameters, diffusivities, and solid solubilities) are either not reliable [3,4] or are not readily available for Ge. In this study, boron and phosphorus implants in Ge before and after dopant activation anneals were characterized by secondary ion mass spectrometry (SIMS) and spreading resistance profiling (SRP). EXPERIMENTAL DETAILS Ge (100) n- and p-type substrates were implanted at 7° tilt, no rotation and using 31P+ or 11B+ with energies ranging from 20-320 keV and doses of 5×1013 to 5×1016 cm-2. Si substrates were also implanted with the same conditions. Implanted samples were annealed for 3 hours at either 400,

B8.11.2

600 or 800°C in ultra high purity N2. Annealing samples at 800°C for 3 hours roughened the uncapped Ge surfaces and is perhaps due to nitridation, which is reported to commence as low as 700°C [5,6]. All samples were subsequently analyzed using SRP and SIMS, figure 1 & 2. Unless otherwise noted, uncertainties in SIMS depths should be within ±3% and uncertainties in SIMS concentrations were estimated to be within 20-30%. Bevel angles between 0.00676–0.706 degrees were used for SRP leading to ±3% uncertainty in depths of SRP, and ±15% uncertainty has been estimated for the accuracy of the measured resistivity. Except in the highest boron dose case, the implanted doses in Ge were within 22% of that measured in the