Aluminum Particle Formation in the GaS Phase of a Low Pressure Chemical Vapor Deposition Reactor Using Dimethylethylamin
- PDF / 3,366,935 Bytes
- 6 Pages / 420.48 x 639 pts Page_size
- 49 Downloads / 159 Views
ALUMINUM PARTICLE FORMATION IN THE GAS PHASE OF A LOW PRESSURE CHEMICAL VAPOR DEPOSITION REACTOR USING DIMETHYLETHYLAMINE ALANE (AIH 3 ) AS A PRECURSOR MICHAEL G. SIMMONDS,* WAYNE L. GLADFELTER,* HAOJIANG LI,** AND PETER H. McMURRY** University of Minnesota, Departments of Chemistry* and Mechanical Engineering,** Minneapolis, Minnesota 55455 ABSTRACT Particle formation in the gas phase during the chemical vapor deposition of Al using dimethylethylamine alane was studied. Typical sizes of the crystalline Al particles monitored were in the range 20 nm to 1000 nm. Introducing trace amounts of H2 0, CO and 02 into the reactor during the flow of the precursor caused an increase in the number of particles. Our results suggested that Al particle formation was induced by impurities in the gas phase although competing mechanisms could not be ruled out. INTRODUCTION The incorporation of gas phase particles into a thin film can be deleterious to the properties of the final device. At the very least, particles deposited onto the surface of a wafer or a growing film will contribute to the presence of defects and result in yield losses during device fabrication. Even more troublesome is the premature failure of an already working device. Particles present-pin sufficient number may reduce the adhesion of films to substrates and cause the randomization of grain morphologies in otherwise oriented films. Clearly, an understanding of the events taking place in the gas phase and the factors controlling these events is crucial to the design 6fýa 'clean' system. We have investigated the formation of particles in the gas phase of a hot wall chemical vapor deposition (CVD) system\used to deposit Al films from the new liquid precursor dimethylethylamine alane 2 (DMEAA (Me 2 EtN)AIH 3 ). This report builds on the previous finding 3 that Al particles were indeed generated in our reactor under typical growth conditions and describes our attempt to probe the mechanism responsible for their formation. EXPERIMENTAL Our approach has been to use two independent monitoring devices. The first device was a laser light scattering particle counter (LPC) which was used to count particles > 200 nm in real time. The second device was an inertial impactor used to capture particles on carbon coated transmission electron microscope (TEM) grids. The cut-size of the impactor was estimated at 20 nm. The stainless steel apparatus depicted in Figure 1 was constructed from ultra-high vacuum compatible components. Nylon or copper TEM grids (3 mm diameter) covered with a thin carbon support were attached to a stainless steel moveable collector stage. This stage (which was at room temperature) was mounted to the end of a calibrated linear motion feedthrough attached to a four way cross (using rubber 'o' ring seals) on the downstream side of the reactor. A coaxial 3 mm nozzle was used to direct the major portion of the reactor effluent onto a TEM grid. The spacing between the nozzle and grid was kept at about 2-4 mm. A sampling line was used to withdraw a small sample
Data Loading...
