A Spectroscopic Investigation of the Amorphous to Microcrystalline Transition in Silicon Prepared by Reactive Magnetron

  • PDF / 1,193,200 Bytes
  • 6 Pages / 414.72 x 648 pts Page_size
  • 52 Downloads / 163 Views

DOWNLOAD

REPORT


A SPECTROSCOPIC INVESTIGATION OF THE AMORPHOUS TO MICROCRYSTALLINE TRANSITION IN SILICON PREPARED BY REACTIVE MAGNETRON SPUTTERING

W.A. Turner and G. Lucovsky, N.C. State University, Raleigh, NC 27695. ABSTRACT

Reactive magnetron sputtering of a c-Si target in an atmosphere containing argon and a varying partial pressure of hydrogen has been used to prepared a series of samples which span the transition from amorphous to microcrystalline silicon. Samples prepared at a substrate temperature of 200 'C, power of 100 W, and Ar partial pressure of 2.10 mtorr can be divided according to the partial pressure of H2, PH2, used in the deposition. Samples deposited using PH2 < 1.00 mtorr are amorphous, those deposited using PH2 > 1.25 mtorr are microcrystalline. Raman scattering is used to confirm amorphousity/microcrystallinity. Two types of changes are observed using infrared, IR, spectroscopy. First, the IR absorption spectrum changes at the onset of microcrystallinity. The absorption centered at 630 cm-1 changes shape while the feature at 2000 cm- 1 appears to shift upward to 2090 cm-1. Second, microcrystalline samples deposited using PH2 Ž1.50 mtorr show significant post-deposition contamination upon atmospheric exposure. This is most clearly noted by the strengthening, over time, of Si0 2 related absorption features which appear at 1050 and 1150 cm- 1, and a weakening of a narrow component of the 630 cm- 1 absorption. This narrow component is not observed in amorphous material. The strength of the 0-related absorption suggests that the oxide is a bulk property. The porous, low density network which is necessary for this contamination to occur is confirmed using HRTEM. INTRODUCTION Amorphous hydrogenated silicon, a-Si:H, has wide ranging technological applications.

It has long been the standard in low cost, high efficiency solar cell

technologies. However, a-Si:H has several characteristics making it less than ideally suited to this purpose. Its band-gap is too wide to optimally match the solar spectrum in terms of the competition between the absorption of phonons and the subsequent energy available in a device. In addition, a-Si:H suffers from the Staebler-Wronski effect, its photoconductivity degrades when it is illuminated. Considerable investigation has focused on finding materials having more suitable properties and/or to engineering around these difficulties. Microcrystalline hydrogenated silicon, Ac-Si:H, and its alloys have been proposed as alternates for a number of the components of solar cells. In particular, when lightly B-doped it forms a near-intrinsic, highly photoconductive material which does not suffer from light degradation and would hence be a superior candidate for use as the active layer in photovoltaic devices'. Both amorphous and microcrystalline materials may be prepared by any number of deposition methods including reactive magnetron sputtering, RMS. RMS was chosen because it is possible to introduce dopants from the solid phase and thus avoid residual doping effects that are inherent to pro