Progression of the Surface Roughness of N + Silicon Epitaxial Films as Analyzed by AFM
- PDF / 1,552,598 Bytes
- 6 Pages / 414.72 x 648 pts Page_size
- 94 Downloads / 169 Views
Mat. Res. Soc. Symp. Proc. Vol. 399 01996 Materials Research Society
value of 2 to 3 mTorr for disilane and 12 to 13 mTorr for silane. The morphology of Si films grown using disilane was studied for doping levels of 2 x 1019 to 1 x 102- phosphorous atoms/cm3 . It was difficult to obtain a constant P doping level for silicon grown using silane because of surface segregation of the dopant, but comparable partial pressures of phosphine were used. The crystallinity of the samples was determined using reflection high energy electron diffraction (RHEED) immediately after deposition. The typical energy of the incident electrons was approximately 15 keV, and a spot size of approximately 1 mm was used. A Digital Instruments Nanoscope H atomic force microscope (AFM) under tapping mode was used to study the surface morphology of the deposited wafers. A polysilicon-on-oxide stack covering a portion of the wafer was used to determine the deposition rate. An initial polysilicon layer of -1000 A was deposited using low pressure CVD. After epitaxial growth, the change in the polysilicon layer was measured using infrared ellipsometry. This was used, along with the change in the stack height, measured using profilometry, to determine the epitaxial growth. The error in determining the total deposition is approximately 100 A, which for the deposition times used in this investigation, yielded deposition rate errors of 3 A/min or less. RESULTS We have studied the variation of the surface morphology of highly doped epitaxial silicon. For silane, a constant partial pressure of 1.2 mTorr was used while the phosphine partial pressure was varied from 1.7 x 10-4 to 8 X 10-4 mTorr. For disilane experiments, a partial pressure of 1.2 mTorr was used with the phosphine partial pressure being adjusted from 2.66 x 10'4 mTorr to 1.33 x 10-3 mTorr. We first discuss the Si deposition rate as a function of the phosphine partial pressure and deposition time and then the morphology of growth using silane and disilane. Figure 1 shows the deposition rate reduction with phosphine partial pressure using silane as the silicon source gas. A dramatic decrease in deposition rate was observed even for PH 3 partial pressures of 1.7 x 10'4 mTorr. SIMS analysis indicates that this corresponds to an initial P doping level of approximately 8 x 1018 cm-3, higher than levels where Si deposition rate reduction is commonly observed in the literature. As the duration of deposition increases, the amount of phosphorus atoms incorporated into the film increases although partial pressures of the gases remained constant. Furthermore, once the phosphine flow into the deposition chamber was terminated, only after 500 A of further Si growth did the phosphorus content show any reduction. Also included in Figure I is the Si deposition rate using disilane as the phosphine partial pressure was increased. A decrease in deposition rate was again observed though it was less dramatic than for silane. Phosphorus SIMS data for disilane shows a constant dopant concentration in the film, and
Data Loading...
