Dopant Quantification by X-Ray Absorption Spectroscopy: Zn in InP
- PDF / 319,729 Bytes
- 5 Pages / 414.72 x 648 pts Page_size
- 110 Downloads / 216 Views
capacitance-voltage profiling) or secondary ion mass spectrometry (SIMS). The latter technique is most widely used, but it has the drawback of requiring ion implant standards for each material studied (the sputtered ion yield depends on the matrix, i.e., material composition, being evaluated). Furthermore, while both SIMS and Polaron profiling techniques provide information on Zn concentration versus depth in the grown layers, neither can be easily be applied to wafers actually used for device fabrication because of their destructive nature. The goal of this study is to investigate whether another technique, x-ray absorption spectroscopy, could be a practical alternative to SIMS for determining Zn concentrations (a) at least as reliably, (b) non-destructively, and (c) on wafers actually used for device fabrication. EXPERIMENTAL X-ray absorption measurements from the Zn K edge (- 9.6 keV) were obtained using fluorescence detection on the AT&T Bell Laboratories X 15B beamnline at the National Synchrotron Light Source. Details of the experimental setup appear elsewhere [3,4]. The epitaxial structures were prepared by low-pressure MOCVD under typical growth conditions for materials used in fabricating laser diodes. The structures consisted of n/p/n InP layers grown on 50-mm-diameter InP:S (n-doped) substrates. Silicon and zinc were the n and p dopants grown in layers of 0.4 and 0.5 pm thickness, respectively. The Zn levels ranged between -7x10 16 - 3x10 1 8 by varying the flow of dimethylzinc into the reactor during growth. The parameters of InP growth were 4 im/hr, 60 Torr, and 650' C. 509 Mat. Res. Soc. Symp. Proc. Vol. 406 01996 Materials Research Society
Samples were cleaved into pieces amenable for SIMS measurements, which were performed on a CAMECA IN4S-4f instrument using a Cs÷ primary ion beam. The x-ray measurements were obtained from identical samples. RESULTS Figure 1 shows Zn K-edge absorption spectra from an InP sample uniformly doped with Zn and from two
I Zn-doped InP (bulk)
p-InP(Zn)7
n/p/n layered InP
structures grown C1 with (CH 3)2Zn flow rates of 40 and 100 -& standard cubic cen-0 timeters per minute (sccm). Although the average Zn concentrations in the bulk-doped sample and 100-sccmgrown structure are
Zn flow rate:
02 Zn-doped InP (layer) ," 40 sccm ,'n-InP(S),;
comparable, about
2-3x10 atoms per cm3, the statistics in 9700 9800 9900 the former are superior because it Photon energy (eV) is not average atom concentration per cm 3 that is import- Fig. 1.Zn K-edge absorption data from three different Zn-doped InP samples. ant but rather the The bottom two scans, obtained from 0.5-pm-thick p-type layers buried betotal number of neath 0.4-pm-thick n-type layers, have poorer statistics than that from the absorbing upper scan of the bulk-doped sample but all exhibit similar structure in the 2.• Zn atoms per cm , i.e., aver- near-edge region. age dose. Thus, the effective sampling depth probed in the x-ray absorption measurement is key in determining effective dose. Taking -10 pm as a simple and
Data Loading...
