Chapter
6.15
ULTEK E-Beam Evaporator
(ultek)
1. Introduction
The ultek is an e-beam evaporation system. This
system is cryo pumped and can attain a base pressure of 2e-7 Torr. The ultek is
set up to hold one 4" wafer. Users may attach smaller samples to the
existing 4" sample holder, which has screws and washers, or they can
design their own sample holder. An Inficon, thin-film crystal monitor measures
deposition rate and accumulated thickness of the deposited films. Evaporation
at hi vac is a line of sight process. The material travels from the source in a
straight line and stops on the first surface it hits. Sidewalls of structures
or topography do not get coated.
E-beam evaporation uses an electron beam to heat
a source of material in a crucible (hearth liner). Electrons are emitted
from a heated filament and are accelerated to a high velocity by several kV.
Electron current is typically small, around 100 mA. A permanent, horseshoe
magnet bends and guides the e-beam in a circular path from the filament to the
source.
Material selection of the crucible should be
appropriate for the material being evaporated. Crucibles made of Cu or
graphite are stocked in the Microlab Office, Room 406. Users should
purchase their own crucibles, and their own source material.
The most common error is to have too much or too
little material in the crucible. The crucible should be at least 1/3 full, but
not more than ¾ full. If there is not enough material, it is possible to burn a
hole through the bottom of the crucible and supporting copper hearth. If there
is too much material it can escape from the crucible and contaminate the
system.
Material can be added to the crucible as needed.
Typically the material is in the form of 1/8 inch or ¼ inch pellets (not
powder, which is bad for contaminating the system).
For the cleanest deposition, the beam current
should be adjusted to just melt a spot in the middle of the material. If the
power is too high, all the material will melt, and then the crucible material
can diffuse into it (may be ok, depending on what you are doing). As power is
increased and more material is melted, the deposition rate increases, but also
the amount of spitting of droplets of material increases. You need to look
through the glass to see if this is happening, and then find a compromise
setting that minimizes spitting, but still gives a reasonable dep rate. Typical
dep rates are 1 to 5 Angstroms / second, but can be higher for some materials.
Some materials have a high enough vapor pressure that a spot does not need to
be melted, and it can sublimate directly from the solid. In order to look at
the hot material, you need to used the welding glasses. The darker ones
(rated”12”) are needed for the high melting point materials (e.g., Ta), while
the one rated “5” can be used for lower melting materials (e.g., Au, Si).
After buying high purity pellets of material for
evaporation, you do not want it to get contaminated. Therefore you need to
check to see if the chamber is clean. If you need to clean it, then first cover
the pump port with a clean piece of aluminum foil so particles don’t fall down
in it since they can prevent the hi vac valve or roughing valve from sealing.
Check the bottom of the shutter for built up deposits that may be flaking off
and may fall into your crucible. Check aluminum foil that is wrapped around the
components that need protection. If the deposits look like they may flake off,
replace with new foil. If there is material on the copper hearth, you can
carefully remove it with a clean room cloth, or if it is stuck on, very
carefully use a clean razor blade. Finally, vacuum out all loose debris with
the vacuum cleaner. Put in new clean microscope slides to protect the viewing
areas of the glass bell jar from being covered over by deposits. There are two
2 in x 3 in slides mounted vertically on the side, and one 1x3 inch slide
mounted at 45 degrees at the top of the chamber. Remove the Al foil that you
put over the pump port.
You need to have clean gloves. Gloves quickly
get dirty since people unconsciously touch their noses, cell phones, etc, and
then transfer the oily, sodium rich goop to all the door handles, deck hoses,
keyboards, and equipment knobs in the lab.
Therefore, put on new gloves, and then do not touch anything with your
finger tips. Reserve the finger tips for handling things in the chamber (and
use tools for handling things when possible).
Another common source of contamination is kapton
tape. This will outgas under the radiant heating from the evaporating material
source. You should only use metal clips, springs, and screws to mount your
samples (no tape, or rubber bands).
Correct heat flow is needed for
consistent results. The most consistent setup is with the thermal contact only
between the copper hearth and the bottom of the crucible (not the sides). This
assumes that you have cleaned out the bottom of the hearth with a cotton swab
and vacuum cleaner, so no particles are there to hold the crucible out of
contact with the hearth. If your carbon
crucible cracks, you probably had nonsymmetric heating and contact stress on
the sidewall.
2. Stand-by Condition
When the ultek is not
being used, it should be left under hi vac, in the following standby mode:
AUTO button on
VENT button off
Ion gauge off
Cooling water on
e-beam power supply off
Inficon off
3 Operating Procedure
(1)
Verify that the cooling water is on: the
water wheel is turning.
(2) Check
to see that the crystal monitor is functioning.
Turn on the crystal monitor. Press the XTAL button. The first readout
will tell you the percentage of the crystal life that has been used up.
Readings are unreliable when crystal use is above 90%. Replace the
crystal if needed. Spare crystals are kept between the nrc and v401 or may be
procured in the Microlab office. They are not recharged. Lab members are
responsible for crystal changes and should make sure a spare is also available
for the next user in case the office is closed when they need it.
(3) Check System base pressure (Granville Phillips Gauge
controller)
The DEGAS function is intended to improve the
accuracy of readings by desorbing absorbed gases from the ion gauge anode. Do
not leave the ultek unattended while the ion gauge controller is in DEGAS mode.
Excessive DEGAS will eventually sag the ion gauge, requiring expensive replacement. First turn
off the filament. Next, flip up the
switch labeled DEGAS. This will
begin heating the ion gauge tube's anode. Turn degas off after 5-10
minutes. When the filament is turned back on the pressure should be 2 - 4
× 10E-7 Torr. Ion gauges should not be used to measure pressures
greater than 10E-4 Torr. Ion gauge controllers are equipped with
circuits that automatically turn off above this pressure. High system base
pressure may mean the cryopump needs regeneration.
(4)
Turn off the ion gauge filament current.
(5)
Close the gate valve to the cryo pump.
1.
Push the lighted AUTO button. The AUTO button
will go dark and the hi vac valve will close. Listen for the hi vac valve to
close.
2.
Push the VENT button. The vent button lights up.
N2 enters the chamber. Venting is finished when the bell jar is
loose and can be moved.
(6) Raise
the bell jar manually.
Move slowly, be careful not to bang the bell jar
against the other components. Once the bottom of the bell jar is higher than
the structures in the chamber you can rotate the column (pull the red handle on
the column) so the bell jar goes to the back, out of the way. (Safety tip:
before opening the bell jar, look at the high voltage power supply to make sure
the main power switch is off. Although there is an interlock to prevent it, you
just want it to be impossible for the high voltage to be on when your hands are
in the chamber).
(7) Put
on new gloves.
Keep the finger tips clean as described above
(8) Lift
up the stainless steel shroud.
Again,
move slowly and guide it gently over the structures so you don’t apply high
forces to anything. Set the shroud on clean aluminum foil.
(9) Clean
the chamber.
As described above.
(10) Load
your material source(s).
As described above. Use clean tweezers. Turn the red handle of the hearth
positioning screw to align the center of the crucible to be adjacent to the
cathode of the e-gun. If you will be evaporating from more than one crucible,
make sure you know where the marks on the leadscrew (positioning screw) are
when the different crucibles are lined up with the e-gun. Make sure the shutter
is closed.
(11) Protecting
the bell jar.
If your target substrate (e.g., wafer or die)
has any apertures that would let evaporated material travel to the bell jar,
then you need to cover it with aluminum foil. Otherwise deposits will build up
on the bell jar. Make sure two new 2" x 3" glass slides are in the
window, and a new 1” x 3” slide is in the 45 degree holder on top. These slides
allow viewing the sources and must be
replaced after each evaporation.
Failure to protect the bell jar will make it
increasingly difficult to see into the system.
(12) Replace
the stainless steel shroud.
Slowly, carefully,
guide the shroud over the chamber structures so you don’t damage anything.
(13) Make
sure the gasket on the bottom of the bell jar and the mating surface on the
baseplate are clean.
(Wipe with clean glove finger) and then lower
the bell jar. The N2 vent pressure holds the bell jar up a little
bit above the base plate, so push down on it with your left hand, so you can
operate the valve control buttons with your right hand.
(14) Pump
down.
Push the vent button (lighted). The vent button
will go dark.
Push the AUTO button. The AUTO button will light
up.
The system will now pump down automatically, but
you have to stay there to make sure it works: the roughing valve will open.
When the convectron gauge reaches 200 mT, the roughing valve will close. The
system will wait about 1 minute, then open the HI VAC valve to the cryo. After
that happens you can turn on the ion gauge, and wait for base pressure to be
reached (or, depending on what you are doing, it may be ok for you to do the
evaporation when it reaches 3x10-6 T).
(15) Program
the crystal monitor.
(See the Inficon manual which is kept on top of
the power supply cabinet.) Push the PROGRAM button, then step through the
variables using the E (enter)
button. The Z-ratio and densities of different metals are listed in the Inficon
manual and are posted on the right side of the NRC. The tooling ratio for the
ultek has been estimated at 145%. All other parameters can be left at the
default value. Press the PROGRAM button again to exit program mode.
(16) Turn
on the e-beam power supply.
Flip the
circuit breaker switch labeled “main power” to the on position. Set
the beam current control variac to 0. Turn the variac on. Turn on the
high voltage (push the high voltage “ON” button). The red light will turn on
(may take a minute). Slowly raise the e-beam current by turning the knob on the
variac clockwise. Watch all of the following: the milliammeter on the high
voltage front panel, the brightness of the e-gun cathode, and the location of
the brightest spot on the evaporation source (where the e-beam is hitting, will
be hottest and looks brighter than the surrounding area). Put the beam on the
center of the material by adjusting the leadscrew, and by flipping the toggle
switch labeled “toward filament”/ “away” (whichever you want) and then turning
the knob on the adjacent front panel variac to move the beam in the direction
perpendicular to the leadscrew (If the voltage adjustment for the this
direction does not work, press the white buttons labeled 3 next to the
voltage knob. This resets the circuit breakers. If it still doesn’t work,
report it on FAULTS).The main thing is that the beam should not be hitting the
crucible, but only the source material. Increase the beam current slowly enough
so the material has time to heat up and not lag too far below equilibrium
temperature for a given beam current setting. Continue increasing the beam current
until you get the desired amount of material melted, and a dep rate reading on
the crystal monitor. Wait 5-10 minutes, allowing the source to outgas. Open the
shutter and adjust the beam current so that deposition rate is about 1 to 5
angstroms per second (can be higher for some materials).
(17) Shutdown.
When the desired thickness has been deposited,
rotate the shutter over the crucible. Turn beam current control variac to 0.
Turn the variac power off. Push the high voltage “off” button. Switch the high
voltage power supply “main power” circuit breakers off. Let cool 20 minutes. Vent the chamber. Raise
the bell jar. Remove your wafer. Lift out the stainless steel shroud. Remove
your crucible(s). Clean the chamber (as described earlier). Replace the 3
microscope slides. Put the stainless steel shroud back in. Pump down. Turn off
the Inficon. Turn off the ion gauge. Leave the water on (it is on the lab
recirc, so it isn’t just going down the drain). Make sure you leave the area
clean. Some users think that it is ok for them to leave their used glass slides
lying around, or even leave kapton tape with evaporated films on them stuck on
the bench. These are not civilized people, and have no right to be in the lab.
5.
Regeneration
Cryopumps are captive pumps and require periodic
regeneration. In addition to normal regeneration cycles, mistakes and
mechanical failures may necessitate regeneration of the cryopump. An
indication of a need for regeneration is a higher than normal base pressure or
excessive cropump temperature. A cryopump temperature monitor is located in the
instrument. The cryopump should be less than 20 Kelvin for proper operation. If
the cryopump needs regeneration file a FAULT report via the WAND.
Refer to the comments section of the wand
for additional suggestions and lab member entries.
6. Checklist
1.
Verify water flow (waterwheel turning)
2.
Verify crystal function
3.
Verify system base pressure
4.
To vent:
1.
Turn off ion gauge.
2.
Push AUTO.
3.
Wait for valve to close.
4.
Push VENT.
5.
Manually raise the bell jar.
6.
Lift out stainless steel shroud.
7.
Clean the chamber (should already be
clean if last user is a good citizen).
8.
Verify clean glass microscope slides are
in place.
9.
Put crucible(s) of material in hearth.
10.
Align crucible with cathode.
11.
Position shutter over crucible.
12.
Put stainless steel shroud back in.
13.
To pump down:
1.
Hold bell jar down.
2.
Push VENT to turn it off.
3.
Wait for valve to close.
4.
Push AUTO.
14.
Program Inficon
1. Enter
density.
2. Enter
z-factor.
15.
Evaporate
1. Verify
variac set to 0, turn it on.
2. Turn
on power supply.
3. Push
high voltage ON button, then wait for the red light to go on.
4. Slowly
turn up variac to melt some of the material.
5. Look
through glass and adjust position of hot spot as needed.
6. Adjust
variac to get desired dep rate (look to make sure material is not spitting
excessively, use welding glasses).
7. Open
shutter.
16.
Shutdown
1. Turn
variac to 0, turn off.
2. Push
high voltage “off” button.
3. Turn
off main power switch.
4. Let
cool 20 minutes.
5. Turn
off ion gauge.
6. Vent.
7. Remove
wafer.
8. Remove
stainless steel shroud.
9. Remove
crucible(s).
10. Clean
chamber.
11. Replace
the 2 x 3-inch glass slides (2) and the 1 x 3-inch slide.
12. Replace
the stainless steel shroud.
13. Pump
down.
14. Leave
water running.
15. Leave
the whole work area clean.
7. Crucibles
The following recommendations (from Poco Graphite, a
specialty material supplier) address several common crucible problems.
7.1
Melt Levels
The most common cause
of crucible failures is overfilling. Overfilling can cause the melt to spill
over the edge of the crucible. When a spillover occurs, a thermal hort is
created between the crucible and the hearth. The resultant thermal stress
causes the crucible to crack. For this reason a maximum melt level of 80% of
the crucible capacity and a minimum melt level of 30% of the crucible capacity
are recommended.
7.2
Crucible Contact
Another significant cause of crucible failures is cracking due to the improper seating of the crucible in the hearth. Out of round or chiseled hearths often create nonuniform mechanical stresses on the crucible walls. For the longest crucible life and for the most reproducible evaporation results, contact between the graphite crucible and the copper hearth should be restricted to the bottom of the crucible and the bottom on the hearth cavity. A circular graphite or copper shim is frequently used to achieve proper contact.
7.3
Handling
Improper crucible handling and storage also can be the source of crucible life problems. Crucibles should be handled with tongs, gloves or finger cots; never with bare hands or fingers. Used crucibles available for reuse should be stored in a dry, oxygen-free environment.
7.4
Aluminum Melts
Aluminum
carbide formation affects the life of crucibles used for aluminum evaporation.
The aluminum carbide forms a transparent, yellowish film on the surface of the
aluminum. When the film covers the entire surface of the aluminum, the
evaporation rate is reduced to near zero. The presence of this phenomenon is
indicative of excessive crucible temperature. The beam power should be reduced
to minimize the formation of aluminum carbide.
Rev. 00 – 1/07, C. Keller – Rewrite of the chapter.
Rev. 01 – 1/07, W. Flounders – Added
four informational paragraphs from Poco Graphite as Section 7 – Crucibles.