AJA systems
07-Apr-2014:This is the wikipedia page for the 2 AJA systems located on the 2nd floor room D218. The system is maintained by the technical staff: Shiv, Claus and Nader. Willy and Morten K are the Czars/super users. Contact Shiv for technical questions and basic training. Contact Willy or Morten if you can't get hold of Shiv.
Systems 1 and 2 are identical in terms of operating procedures. Password for logging in is 'apex' (no quotation marks). They differ slightly in their outfitting:
- System 1: Two DC sputtering targets, heater in sample mount and Kaufmann ion source for cleaning. Mechanically clamped to the loading arm as well as the rotation holder inside the chamber.
- System 2: One DC sputtering target, one RF sputtering target, an RF supply to substrate as well as a Kauffman ion source for cleaning. No heater. Mechanically clamped to the loading arm and magnetically clamped to the rotation holder inside.
Czar's log (use this to catch up on updates/changes to the AJA systems):
19-May-2014: AJA2 was opened up and a new shutter for the Kaufman ion source has been installed. To use the shutter, toggle the 'Shutter' button next to the button that turns on the plasma for Kaufman ion source.
This is the ion source shutter in action.
The shutter can be used to protect your sample from the ion bombardment during the filament warmup or during the sample cool down period between bursts, if you intend to mill for a long period (>5 mins). It is analogous to the substrate shutter and metal evaporation.
The Kaufman source still needs to be shutdown after you're done using the system.
To toggle the shutter open/close, you need to hit the 'shutter' button right next to the button marked 'Output' that turns on the Kaufman ion source itself. The button is highlighted below.
01-Mar-2014: AJA1 was opened up and the loading arm alignment was adjusted.
- To load/unload:
- Rotate the stage till the permanent mark around 270
- Tighten the brass knob until the permanent markers meet
- Load/Unload! The loading should be smooth enough to push the pins in with two fingers. If you force it, you are directly responsible for pushing the stage out of alignment.
System 1
This system is expected to at least reach a vacuum of about 5.0x10-9 Torr after pumping for 24 hours on main chamber.
Current materials loaded:
| 1 | 2 | 3 | 4 | 5 | 6 | DC 1 | DC 2 |
|---|---|---|---|---|---|---|---|
| Ti | Cr | Pt | Au | Ag | Al | Ti | Nb |
Various procedures on the system are described in a step by step guide below:
Loading your sample
- Check the cryo-pump monitor. It should be between 12-17 K. If it's higher than 20 K, grab an ansvarlig or a technical staff member.
- Check that the turbo frequency is 1500 Hz. The turbo pumps on the load lock.
- Check that the pressure in the main chamber (ion gauge sensor) is <1x10-7 Torr
- Check that the load lock gate valve (connects the load lock to the main chamber) is closed
- Push down the 'Load Lock' switch in order to vent the load lock
- Once the load lock pressure reaches ~740 Torr, the load lock lid pops out a bit and can be rotated freely.
- Do not apply force and pull the load lock lid out. The load lock may not be vented yet.
- Rotate the lid until the permanent markers meet and twist the lid out, pivoting about the two permanent marks on the left.
- The permanent marks indicate the position of spring loaded ball bearing that hold the lid in place, preventing it from falling out.
- Ideally, you want rotate the lid so as to pivot against two bearings.
- Place the load lock lid, handle up, on the three rubber bumps.
- Remove the sample cassette. It's held in place by three pins that lock into a groove
- Examine the sample cassette. There are three small pins. These pins, via a similar mechanism lock onto the holder inside the main chamber.
- Grab a fresh cleanroom wipe, place the sample cassette on the wipe.
- Load your chip either using the mechanical clamps or the double sided Kapton tape. Carbon tape is being phased out as it may be a source of unwanted residues inside the chamber and on the metals.
- Load the sample cassette inside the load lock, rotate the holder to confirm all three pins are locked in place, replace the lid and flick the 'load lock' switch up to pump out.
- Wait till the chamber pressure goes down to 3x10-6 Torr. The turbo will have revved up to 1500 Hz by now. Confirm this.
- Open the gate valve between the main chamber and the load lock.
- Load your sample. The mechanism which locks the sample cassette on to the holder in the main chamber, also simultaneously unlocks it from the loading arm.
- Take note of the sample cassette orientation on the holder as well as the rotation/orientation of the loading arm. In principle, you should unload using the same orientation as this will be the easiest.
- Close the load lock valve.
- Check main chamber vacuum
- Check cryo pump temp
Evaporating metal
- Choose the material on the linear crucible drive as well as on the deposition controller.
- Everything is manual. We no longer run automated processes
- Open the ebeam shutter by flicking the physical ebeam shutter switch to open.
- This exposes the metal to be evaporated.
- If you don't open this shutter, the accelerated focussed electron beam will hit the shutter instead of the metal and drill a hole through it!
- Switch on the Carrera Ferro Tec high voltage power supply. The switch is green and is labelled 'Main'.
- Turn on the high voltage. It sets the beam acceleration voltage to 10 kV. This is fixed and not changed by the users.
- Be very sure that you have opened the ebeam shutter.
- Two clicks of the knob and the current set point is set to 4 mA.
- Wait till the current increases to this value.
- Can you see the bright spot where the beam hits the metal in the crucible.
- Center the beam and make sure the beam is neither too focused nor too defocused.
- Now, consult the excel log sheet to determine the typical current needed to get a finite evaporation rate.
- Ramp the current up at about 20 mA/min to half the value. Let it sit at that value for 2 mins while the metal soaks and thermally equilibrates.
- Too fast and you'll crack the crucible liner
- or your evaporated metal film will be rough
- 1 A/s is a good rate for metal film evaporation. Try and stay around this value.
- When you are ready to evaporate, zero the counter on the deposition controller and open the sample shutter.
- The shutter takes about 1-2 secs to open, so you don't have to be paranoid about synchronizing the zero with the shutter opening.
- Wait until the right thickness is evaporated.
- Close the substrate shutter.
- Ramp the beam down to 0 in a period of a min. Don't be too quick about it. We want the metal and the liner to cool down slowly to stop the liner from cracking due to stress.
- Turn off high voltage.
- Wait 2 mins for the metal to cool down before moving over to the next metal
- If you're done, turn off the Carrera voltage supply.
- Close the ebeam shutter if the metal is no longer red hot.
Unloading your sample
- Rotate the sample to the correct position (same orientation as during the loading procedure)
- Open the load lock gate valve, and unload your sample
- Close the load lock valve.
- Vent the load lock using the load lock switch on the main rack.
- As before, wait till the load lock reaches ~740 Torr and pops out a bit.
- Twist and pull the lid out if the permanent marks are lined up.
- Rest the lid on rubber knobs, handle up
- Fresh cleanroom wipe!
- Get the sample cassette out, unload your sample.
- If you used a double sided tape, wipe off the residue with IPA or Ethanol
- Put the sample cassette back and pump out the load lock.
- Fill out the excel log file
- Clean up after yourself. If the work station is found untidy, the last user will be held accountable.
Using the Kaufman ion source (ion milling)
To utilize the ion milling, rotate your sample to point towards the Kaufman source and power it on. In some cases it is beneficial to have your sample at an angle and/or rotate the sample while milling. The figure below shows how to use the software on the control computer to turn on the ion flow.
When operating normally, the chamber should light a clear whiteish hue, and the kaufman power source should read numbers similar to these:
You have to time the milling 'by hand', i.e. use a stopwatch. To turn of the ion source simply reverse the steps outlined in the above figure.
Miscellaneous notes / values for milling
Please update this list with good tips / mill rates for materials:
- The approximate mill rate for InSb heterostructure is 15 nm/min. It is advised to tilt the sample to 30 degrees and use 30 speed on the rotating engine. This gives a cleaner and more smooth surface.
- Photolith AZ1505 millrate is approximately 15nm/min (at angle 30 degrees and rot speed 30).
- The optimum milling time for making superconducting contact to InSb nanowires at Harvard was 30 seconds (at 300 V, 13 (?) mA). We tried various things, like milling straight on for 10 seconds, and milling at 45 degrees on both sides of the wire for 10 seconds each, but the total time is the most important thing. Too much and you destroy the wire, too little and contact is bad. There was some variability in the optimum time over a period of 6 months, so don't assume the recipe will be constant.
- The beam current density depends on the beam voltage, beam current and the accelerator voltage. Profiles of the dependence can be found in the ion source manual supplement, which can be found here: File:KDC40 4cm DCS Sup Micro Moly V2.pdf. For example, for a beam current density of 0.14mA/cm^2 with 150V beam voltage, Kaufman recommends a beam current of 17 mA with a 300mV accelerator voltage.
- As of 6-11-2013 the ion source needs to be warmed up before use. If you experience variable result when contacting (!!) or a power law like law for your experimentally determined milling rate it means that you were using the ion source when it was still ramping up the emission ( 46 mAmps is the default values and it takes 3 minutes to reach it).
Milling rates
Here are some approximate milling rates measured for AJA 1 at 300V:
| SiO2 | PMMA | Al/AlOx | Al2O3 (ALD) | Gold | User | |
|---|---|---|---|---|---|---|
| 1 min | 3 nm | 23 nm | 3 nm | Nino | ||
| 1:30 min | 3.5 nm | Merlin | ||||
| 1:30 min - 30° + rot. | 2.0 nm | Merlin | ||||
| 2:30 min | 5 nm | Merlin | ||||
| 3 min | 8 nm | 55 nm | 8 nm | Nino | ||
| 6 min | 15 nm | 9 nm | 42 nm | Anders J | ||
| 8 min - 30° + rot. | 31.5nm | Merlin | ||||
| 2 min (Fil. warmed up before) | 9 nm | Rawa/Shiv | ||||
| 5 min (Fil. warmed up before) | 25 nm | Rawa/Shiv |
Using the sputtering system
Because of the software upgrade on system 1, when the system was moved from harvard to cph, the old recipes from harvard no longer work.
Anders has made a .pdf of many of the old recipes. If you want to use any, you need to type them in by hand. The .pdf can be found here: AJA sputtering processes.pdf
Using the heater
As of today (2/11/2013) the heater is broken, at max it goes up to 50 and then sops heating. Willy is working on this.
Igniting plasma using DC
System 2
This system is expected to at least reach a vacuum of about 3.5x10-8 Torr after pumping for 24 hours on main chamber.
For safe operation of the rf ALWAYS enter a ramp rate such that the rf circuitry never ramps faster than 1W/second. Enter ramp rate BEFORE changing wattage!
Current materials loaded:
| 1 | 2 | 3 | 4 | 5 | 6 | RF 2 | DC 1 |
|---|---|---|---|---|---|---|---|
| Ti | Cr | Au | SiO2 | Al | Pd | Re | Mo |
The mirror is a piece of Si/SiO2 wafer (150nm).
Igniting plasma using RF
The RF plasma can be ignited using the following guidelines:
- Argon: 80sccm, set cryopump valve controller to pressure mode and set pressure to 30mTorr. The RF1 (stage) should be set to 25 and RF2 (sputtering target) set to 50. Once the plasma ignites, the pressure can be lowered to 3mTorr. Check that the plasma is visible (a faint purple blueish hue inside the main chamber). There is a little pink indicator in the software that should light up in the control software congruent with when the plasma is visible in the chamber. Remember to check that the power is on for the RF circuitry!
Purchasing
Materials for evaporation can be bought from many sources, e.g.
The boats we use approximately take 50g of material. Material should be high purity (about 3-4N ~ 9.999% purity). Preferably buy nuggets or pellets, sizes 1/8" ~ 3-6mm in size.
For ordering of copper gaskets:
Type: G-1650 16.5" OD CF Copper Gasket, Package of 1
Maintenance
The standard configurations on the systems are:
If you are unable to ignite the plasma (either dc, rf or ion plasma) start by checking for shorts between pins on the powersupply input on the sputtering arm / ion source.