Graphite Crucibles for E-Beam Sources

Overview

• The crucible forms a thermal barrier between melt and watercooled hearth, so efficient, that deposition rates at the same power can increase by 400%. At the same deposition rate, power may be cut to 25% of its bare hearth value
• Each graphite crucible undergoes high purification before it reaches the customer
• Major contaminants (Fe, Si, Al, and Mg) measure less than 5 ppm, while there are no detectable levels of 16 other common metals (including Na, K, and Pb)
• Withstands as many as 400 evaporation cycles

Materials evaporated from graphite crucibles:

• Al-Ge alloys, Al-Si alloys, B, Be, Bi, CdS, CeO2, Cr, Cu, Ge, In, Mg, Mo, Na3AlF6, Pd, Pt, Sb, SiO, SiO2, Sr, Ta, Ti, Y2O3, Zn
• Choose the correct size for your furnace. Taller models hold more material, but cannot work if put in a multi-hearth furnace not designed for tall crucibles
• Model EVCEB-2 has the same dimensions as EVCEB-4 but includes a top lip

NOTE: Do not confuse this ultra-fine-grain, high-density graphite with "glassy" (or vitreous) carbon which typically cracks after one run.

Common Causes of Crucible Breakage

1. The ramp/soak levels are wrong for that material. A recipe for e-beam consists usually of two ramp/soak power levels. This allows for the gradual heating of the material to a point where deposition can start. When deposition starts there is a transfer of power control from the ramp/soaks to a PID loop control between the e-gun power supply and the quartz crystal controller. If ramp2/soak2 power level is not near the power level to hit the specified deposition rate (say 1 Angstrom/sec) this means the PID loop must try very hard or search very hard to try and get rate. Since the PID loop is in control of the power, power can oscillate dramatically (+90% to 0%) for a while before the PID loop finally gets the power at the right level to control at that 1 Angstrom/sec rate. Wild swings in power rapidly liquefy and then condense the material several times even before the deposition starts. This will crack any crucible. It is best to get the ramp2/soak 2 power level very close to the power needed to hit the desired rate. This way when the PID loop takes over control, it does not need to adjust power that much to hit rate - just a slight tweak. This prevents this rapid liquefy/condense process from occurring. So incorrect Ramp levels can definitely cause crucible cracking.

2. The second reason is much easier - the user shuts the power supply off or has a very short ramp down time for power after the deposition is done. This causes the rapid solidification of the melt, and this stresses out the crucible liner.