Preparation of PIT-Free Hydrogen-Terminated Si(111) in Deoxygenated Ammonium Fluoride
- PDF / 2,477,749 Bytes
- 6 Pages / 414.72 x 648 pts Page_size
- 104 Downloads / 182 Views
diffusion of oxygen to the silicon surface over several minutes, the consequence is that the total etch time is very important. EXPERIMENTAL Silicon (111) wafers (Siltec, n-type, 0.1-0.9 Q*cm, < 0.50 miscut, with an undefined misorientation angle) were cleaned in "piranha" solution, consisting of 4 volumes of sulfuric acid (J.T. Baker, CMOS electronic grade) to 1 volume of 30% by weight aqueous hydrogen peroxide (J.T. Baker, CMOS electronic grade), for 10 minutes at 100°C, followed by an "SC2" clean, consisting of 1 volume 30% by weight aqueous hydrogen peroxide to 1 volume of concentrated hydrochloric acid to 4 volumes of water, for 10 minutes at 75°C, and finally a second 10 minutes in piranha solution at 100°C. The wafers were rinsed with ultra-pure water (Millipore filtration system, 18 MQ*cm resistivity) between and after each cleaning step. The clean silicon wafers were then immersed and held vertically in ammonium fluoride solutions contained in Teflon vials. A 40% by weight aqueous ammonium fluoride solution (Ashland Chemical Co., cleanroom grade) was used as purchased. Deoxygenation of the fluoride solution was achieved by bubbling with argon for 30-60 min. through a Teflon TM tube inserted into the solution. The argon was purified through a titanium-gettering furnace (Centorr Model 2A) to remove trace amounts of oxygen. The STM studies reported here were performed immediately following emersion from the fluoride solutions. We have previously found by XPS that the least amount of oxygen, and no detectable fluoride, are obtained when the hydrophobic H-Si(1 11) samples are emersed into air from the NH 4F etch and not rinsed with water [2]. The ex-situ STM measurements were performed under purified argon with a custom-designed instrument [2]. STM tips were electrochemically etched from 0.25 mm Pt/Ir wire in calcium chloride solution. The H-Si(1 11) surfaces were typically imaged at Esample- -1.5 V with a tunneling current of 50 pA. These tunneling parameters were empirically determined and appeared to give the best images. RESULTS Figures 1 and 2 compare STM images of H-Si(1 11) surfaces prepared in deoxygenated 40% NH 4 F (Figure 1) and in air-saturated 40% NH 4F (Figure 2) for 15 min. Notice the large terraces and uniform step distribution in both sets of images. The observed step height is the expected 3.1 A. The small etch pits (1-2 nm dia.) with a ca. 1% surface coverage seen on H-Si(1 11) surfaces prepared for 15 minutes in air-saturated 40% NH4 F (Figure 2b) are no longer present when the solution is deoxygenated (Figure lb). In addition, saturation of the etch solution with oxygen gas greatly increases the etch pit density (data not shown). These results indicate that dissolved oxygen is involved in the initiation of etch pits. We undertook a similar experiment in more dilute fluoride in order to amplify the effect of pit formation. H-Si( 111) surfaces were prepared in deoxygenated 20% NH4F and in air-saturated 20% NH 4F for 15 min. STM images of the H-Si(111) surface prepared in air-saturated 20% NH
Data Loading...
