CO 2 Laser Cleaning of Hydrophilic Oxidized Silicon Surfaces
- PDF / 486,998 Bytes
- 6 Pages / 414.72 x 648 pts Page_size
- 50 Downloads / 217 Views
0.95 J and duration of 0.2 Vs, was steered using a mirror and a focusing lens to irradiate the substrate surface. The substrate was mounted face down on an X YZ stage whose displacements were computer controlled. The X Y axes permitted the scanning of the surface to perform the cleaning of large areas, while the Z axis was used to vary the laser beam energy flux. To deposit the energy transfer water film at the irradiated spot before the laser pulse triggering, a stainless steel chamber was half-filled with deionized (DI) water. This water was heated to 37' C using a stainless steel isolated heater. The temperature was measured using a stainless steel isolated thermometer. In this way, the water was in equilibrium with its saturated vapor. A nitrogen gas input of 4700 ml/min, connected to a flowmeter and valves, was used to carry a controlled volume of water vapor toward the gas output. The gas output was connected to a stainless steel nozzle, heated to a temperature slightly higher than that of the water, whose end was held near the surface to be cleaned. When reaching the colder target surface, the vapor condensed to a water film. A pulse-timing unit permitted the water vapor to be deposited for periods ranging between 0.5 and 3.5 s; after 0.1 s, 4 pulses were triggered at intervals of 0.1 s to assure that all the water had been evaporated. To perform a cleaning over a sufficiently large area, the wafer was linearly moved by 3 mm after each vapor burst/4-pulses sequence. The overlap of the cleaned spots varied between 2 and 9 mm when varying the laser energy density from 0.5 to 3 J/cm2, respectively. The area cleaned was simply a square 24x24 mm2 , whose center corresponded to that of the wafer. Improvements made to the experimental set-up since our previous reports [20-21] concerned the replacement of components by stainless steel and particle-free parts (water chamber, heater and thermometer isolation, nozzle and its silicone heating tape). Our purpose was to ensure the harmlessness of these components with regard to the cleaning process. The substrates used were 100 mm silicon wafers whose surfaces were first cleaned and made hydrophilic using a modified RCA recipe [2]. This consisted of (i) 0.05:1:5 SC1, 80' C, 10 min; (ii) 1:1:6 SC2, 80' C, 10 min; (iii) 0.5 % HF etch; (iv) boiling isopropyl alcohol, 2 min; (v) 0.05:1:5 SCI, 80' C, 10 min. An X-ray photoelectron spectroscopy (XPS) investigation of the resulting surfaces revealed silicon dioxide with negligible carbon contamination. To artificially contaminate the surfaces, we used a particle generator (Particle Measuring Systems Inc.) similar to those used to generate monodisperse PSL particles for calibrating particle counters. A filtered air is driven through a nebulizer that has particles suspended in DI water. The contaminated droplets so generated are carried through a long tube and a drying chamber, resulting in particle-laden dry air. This air exits through a nozzle which could be manually moved over the wafer surface. The particles used to contaminat
Data Loading...
