RESUME / CONTACT

RESEARCH
. . NiAl by CVD
. . . . System
. . . . Publications
. . . . Ph.D.
. . Al2O3 by MOCVD
. . . . System
. . . . Publications
. . . . Masters
. . Laser Interferometry
. . . . System
. . . . Publications
. . . . Senior Report

. . Lab Management
. . SEM Micrographs

TEACHING

RESIDENCE LIFE

STUDENT ORGS.

HOBBIES & INTERESTS

PROJECTS

PROF. SOCIETIES,
AFFILIATIONS
& LINKS

 

DOWNLOADS

DOCUMENT INDEX

WHAT'S NEW?

 

 


In this system, we endeavor to create thin Al2O3 (aluminum oxide or alumina) films. Most of our substrates were Rene N5 (a nickel-based superalloy created by GE Aircraft Engines) or SN282 (Si3N4, silicon nitride, a very hard ceramic). We also put a 1cm^2 piece of polished silicon wafer in each run to check for batch-to-batch consistency and quality.


This process begins with the controlled flow of the Argon carrier gas from the left. The gas is passed through two bubblers where it picks up precursor vapor and carries it out and to the reactor. In the reactor, the gasses mix and react with any material heated to a sufficient temperature. The used gasses exit the system towards the pump, which is holding the system at about 1/8 of atmoshperic pressure, and are handed off to the air handling system.

Click on the system diagram for a larger version.


From the left, we begin with the Argon (Ar) carrier gas supply. This gas is metered by mass flow controllers through two parallel paths going to the reactor chamber. During system warm-up, the gas flows through the bypass valves (the path not through the bubbler) and straight into the reactor. Once the reactor is heated up, the bypass is closed and the bubbler (the gray and blue rectangles) entrance and exit valves are opened. Due to pressure, the gas is forced down a tube to the bottom of bubbler, where it exits and bubbles up to the surface and out the exit valve.

During its time in a bubbler, the gas picks up a small amount of water (at room temperature) or aluminum acetylacetonate (at 135 degrees C) vapor. This is how a liquid precursor (starting ingredient) is conveyed into a reaction chamber in a CVD system. Since the aluminum precursor must be held at 135 degrees C to liquify it, the tubing from the bubbler to the chamber and part of the chamber must also be heated (indicated by the red background) to at least the same temperature to prevent the precursor from condensing and solidifying in the piping.

Once the vaporized precursors reach the chamber with the carrier gas, they spread out to fill the reactor, mixing as they do so. In some cases, it is desireable to install a "guide" tube to confine the gasses as long as possible before they reach the substrates one wishes to coat. This can reduce the mixing effect, though.

Pretty much all CVD reactions are endothermic, which means they require some heat/energy to start or sustain the reaction(s). This means that any item heated to the minimum temperature will be coated. In this case, the sample is heated to 400-600degC on a small heater inside the reactor. This is called a "cold-wall" reactor, as the walls are not heated (for coating, only to prevent condensation). In a "hot-wall" reactor, the walls are heated to above the reaction temperature, and heat is radiated from the walls to the samples inside, heating them.

Once the gas passes the substrate (reacting or not), it is pulled out of the system by the mechanical pump. Since our reaction takes place fastest at a pressure below that of the normal atmospheric pressure, we regulate the system pressure (from the bubbler exit points to after the reactor) at about 1/8 of an atmosphere, or a little under 2psi. Since the pump runs continuously, a throttle valve is used to regulate the strenght of the pump's pull. This is accomplished through the use of a feedback loop. The pressure transducer puts out a signal based on the pressure measured, and this signal is read by the controlling unit (not in the diagram) which either opens the throttle valve a little if the pressure is too high or closes the throttle valve a little if the pressure is too low.

The pump exhaust is then convveyed to a specialized air handling system which treats the exhaust before it is released to the atmosphere.

As you may have noticed, there are a few extra valves (circles with an 'x' through them) in the diagram and system. These are generally for safety, maintenance, and other control-oriented purposes. Also, as may be observed in the photo below, the gas is evacuated from the top of the sytem, inverted relative to the diagram. Samples were attached to the heater with silver paste to conduct heat.

Click on the system photo for a larger version.


This system was actually built in the same system cabinet as my earlier Laser Interferometry (Senior Design) project. In fact, it uses the same chamber, heater, and downstream assembly (the pressure transducer, throttle valve, and pump). For this project we changed the precursors and the delivery system. We also removed the laser interferometry and other in-situ characterization equipment. No photo of the Laser Interferometry system was available.

A picture of the heater removed from the system, with a few samples atop it.

     
Best viewed at 1280 x 1024.   This page was last updated on 20Mar04.