• PDF / 439,616 Bytes
• 6 Pages / 414.72 x 648 pts Page_size
• 100 Downloads / 227 Views

DOWNLOAD

REPORT

---

HYDROGEN AND LITHIUM DIFFUSION IN AMORPHOUS SILICON

WOLFHARD BEYER AND UWE ZASTROW ISI-PV, Forschungszentrum Juelich, D-5170 Juelich, Germany

ABSTRACT Diffusion processes of hydrogen and lithium in amorphous silicon are compared. While similarities in the diffusion energies and in a doping dependence exist, the charge state of the diffusing particles and the type of bonding of diffusing particles in traps differ. In case of Li diffusion, oxygen contamination plays an important role. INTRODUCTION Hydrogen stability is of fundamental importance for the stability of hydrogenated amorphous silicon (a-Si:H) based devices. Lithium stability and diffusion, on the other hand, is of interest for potential application as an effective interstitial donor [1]. Since H and Li are elements close to each other in the periodic system and since they are both monovalent, a comparison of diffusion effects may serve for a better understanding of diffusion processes in a-Si:H in general. In this article we report on results of a systematic investigation of hydrogen and lithium diffusion in a-Si:H films prepared under various deposition conditions with various doping levels. We confine ourselves to films prepared at substrate temperatures above about 200 0 C. The diffusion coefficients were obtained by SIMS profiling of lithium and deuterium implanted material as well as of layered structures of hydrogenated and deuterated material prior and after heat treatment. EXPERIMENTAL Most films investigated .tere prepared in a conventional plasma deposition system with a base pressure of about 10- mbar. Silane (SiH 4 ) (or silane-diborane (B2 H6)_ and silanephosphine (PH 3 ) mixtures) were decomposed in a capacitive reactor at antrffrequency of 13.5 MHz. Typical deposition conditions involved a substrate temperature Ts of 250 - 280 °C, a pressure of about 0.5 mbar, a silane gas flow of 2-5 sccm and an rf power of 5W resulting in deposition rates of 1-5 A/s. Film thickness was 0.5-3 pm. Some films were also deposited under similar conditions in a plasma deposition system built from UHV components. The base pressure here was < 10- mbar. Films deposited by vacuum evaporation at a substrate temperature of 300°C were also investigated. Deuterium and lithium ion implantation was performed at room temperature using a mass separator. Implantation energy was 30-60 keV. The samples were cut into pieces which were annealed in vacuum at various temperatures and for different time intervals. For SIMS profiling we used an oxygen (0, +) sputtering beam at almost normal incidence at a beam energy of 3 keV for lithium defection and 6-9 keV for deuterium implanted samples. The low beam energy for Li detection was necessary to minimize the decay length in the Li depth profiles [2]. A comparison between implantation dose and SIMS depth profile allowed an absolute calibration of the SIMS signals for hydrogen (deuterium) and lithium concentrations. RESULTS AND DISCUSSION Typical SIMS profiles of implanted deuterium and lithium in undoped a-Si:H prior and after anne