Microstructure and Hydrogen Dynamics in a-Si 1-x C x :H

  • PDF / 435,700 Bytes
  • 6 Pages / 382.5 x 615.6 pts Page_size
  • 78 Downloads / 178 Views

DOWNLOAD

REPORT


EXPERIMENT Bilayer RFS a-Sij.,C,:(H,D)/a-Si,.,C.:H films were fabricated by rf sputtering a Si target [6,8] in an Ar, H 2, and CH 4 plasma. Undoped and boron doped bilayer ECR films were fabricated using a system described elsewhere [9]. SAXS measurements were conducted on the RFS films codeposited on Al foil; the samples were annealed in a tube furnace under flowing high purity He. For IR and DSIMS measurements, films deposited on Si wafers were annealed in evacuated pyrex tubes. Csi.H (with an estimated error of ±15%) was determined from the 640 cm-1 IR wagging mode [10,11]. Initial values are given in Table I. Since the 3000 cm' C-H stretch band was unobservable, the C-bonded H content Cc.H could not be determined directly. The Si-CH. and C-H wagging modes at -770 and -1000 cm', respectively, provide only a qualitative measure of Cc.H, due to overlap with other modes. The 2000-2150 cm 1 Si-H stretch bands provide insight into the nature of the Si-H bonding configuration. The 2000 cm'1 band 1is the vibration frequency of a Si-H embedded in the a-Si network [10,11]. The 2080- 2150 cm- band is associated with the vibration of Si-H bonds on internal surfaces of voids, Si-H bonds in SiHt configurations, or H bonded to Si-O or Si-C. The C content x was determined by Auger electron spectroscopy (AES) or electron probe microanalysis (EPMA). In the ECR samples x14 at.%, consistent with the measured energy gap of-2 eV. The boron levels were determined using SIMS. Table I. The carbon content x, boron level CB, and the initial Si-bonded H-content CQsjof the RFS- and ECR-deposited films. Csl:n CB Cs1:H Sample x (at%) Sample (a)From EPMA (at.%) (at.%) (at.%) 0")From AES 12 0 ECR1 2.2 (b) RFS1 0.01 12 ECR2 2.0(a), 3.0O0) RFS2 12 0.2 12.3 ECR3 1.6(a) RFS3 0.6 12 ECR4 14.0 3.1 (a) RFS4 0 6 ECR5 12.5 5.7(a) RFS5 2.9(a) RFS6 16.8 6.0(a), 6.20') RFS7 9.0 7.1a), 7 .4(b) RFS8 19.0(a) RFS9 RESULTS AND DISCUSSION a. SAXS and the Microstructural Dynamics of the RFS Films. The SAXS intensity I(q), where the scattering wavevector q = (47d-)sinO (A = 1.54 A is the x-ray wavelength and 20 is the scattering angle), of the RFS films increased steeply with decreasing q at low q (see Figure 1). In this region, the scattering curve I.(q) results generally from structural features which are large relative to the instrumental resolution of -20 nm. The existence or absence of preferred orientation of the structural features may be obtained from tilting SAXS measurements, in which 1(q) is measured at different angles of the film with respect to the incident x-ray beam [1]. Figure 2 shows the effects of the tilt angle on the SAXS of RFS3 and RFS4, which were also observed in other samples. The large tilt-angle dependence at q < 0.6 nm', which corresponds to the large features, indicates that these features are highly

330

elongated and have a preferred orientation. We suggest that they are due to residual columnarlike domains [1,2] and are defined by lower density regions at their boundaries. In addition to IL(q), all RFS samples showed a clear q-i