Application Note: Evaporating Aluminum in an e-Gun™ source
Aluminum (Al) is highly reactive when molten. Al will react with all liner materials. Al also getters Oxygen to form oxide layers if any oxygen is present during deposition. Aluminum Oxide (Al2O3) has a much higher melting point and vapor pressure when compared to Al (2,045C M.P., 1550C V.P. for Al2O3 compared to 660C M.P., 1,010 V.P. for Al). Further Al2O3 is an excellent insulator compared to Al which is a fairly good conductor. These properties make electron beam evaporation of Al more complicated than most metals.
When evaporating Aluminum DO NOT SWEEP the beam! Sweeping the beam produces more heat closer to the liner / material interface. This will greatly speed up the reaction process. You can sweep to initially melt down the charge if needed (larger crucibles will require this- a 2.2cc typically will not). Place the beam as close to the center of the crucible as possible during evaporation. Slowly ramp power up and down. It should take 1-3 minutes to reach evaporation and after finishing the run, you should come back to a low power level (20-30 mA) and soak for 30-60 seconds to “cool down” the melt. This will reduce liner breakage from expansion.
Do not overfill liner. Fill the liner 50%-70% (when melted). Aluminum will “climb” the walls of a liner and if it spills over and touches the water-cooled crucible the liner will break. In larger sources you can use a liner inside a liner to eliminate this possibility. Overfilling a liner is the #1 cause of failure with AL.
Liner choices:
Graphite: Pure electrical grade graphite will work for AL deposition. Straight Graphite is messy (generates particulates) and will react very quickly with Al to form Aluminum carbide. Typically, carbide will form after just one or two short runs. You will see a brownish / gold formation at the Al/Graphite interface. We normally do not recommend straight graphite for Al deposition due to this failure mode. Please note that our standard Graphite liners are not electrical grade graphite.
Coated graphite: Coated graphite (CG) or Glassy Carbon is a baked graphite finish with a smooth glass like surface. It is similar to a baked enamel finish on pottery. This thin smooth coating resists chemical reaction and also reduces surface area. The smooth surface also helps with a material that expands and contracts giving less mechanical resistance to movement. Although the coating reduces chemical reaction Al will still react over time. It is typical to get anywhere from 4 to 12 runs with this type liner before failure.
FabMate®: (FM) is a patented process similar to coated graphite. It is a graphite-based coating thicker than the glassy coating. It has excellent chemical reaction resistance. The finish is more porous then glassy carbon adding additional reaction surface area. It is also more prone to breakage from expanding materials.
**NOTE** Users have mixed opinions when comparing CG to FM. We have had favorable results with CG and prefer it to FM. Our experience with FM is slightly less favorable. However, customers are split about 50%/50% in overall responses.
Boron Nitride: Boron Nitride (BN) offers the best chemical resistance to AL. However, its insulating nature makes evaporation more complex. You must soak the material at low power ~30mA for a period of 1 to 5 minutes to generate heat into the liner to force electrical conduction. This process is tricky for a novice user and often results in a broken liner. The cost of BN is double the cost of CG or FM and with the ease of breakage not typically recommended.
Intermetallic: Intermetallic liners are Ti impregnated BN. They are available in the following sizes: 1.5cc, 2.24cc, 4cc, 7cc, 10cc, 15cc, 25cc and 40cc. They work well with Al and offer the improved chemical resistivity of BN while being electrically conductive.
Bare hearth / No Liner: Al evaporation is difficult without a liner. It requires high power and oxides can form very quickly on the top surface of the melt. If the oxide forms and power is not immediately reduced the entire charge of aluminum will often explode completely out of the source. If the crucible is to be used for only AL deposition the copper crucible can be treated with Hydrogen Peroxide to form a copper oxide which reduces thermal transfer from the AL melt. This technique works well for high volume production systems with large capacity crucibles. Simply swab the copper with peroxide (use a cotton ball) multiple times until the copper forms a dull dark brown finish
