/
Atomic Hydrogen Station

Atomic Hydrogen Station

Atomic Hydrogen Station Procedure

A combination of the Demo Procedure and Procedure on the Wiki from Rodolfo on 10/12/11 (http://lase.mer.utexas.edu/wiki/index.php/Atomic_Hydrogen_Manifold, past edits), to make a Master Procedure Compiled NTS, 7/16/2015; Updated NTS/DJI 20171116;

Note: Must have sample/s wafer loaded on trolley and in buffer chamber. AHS is done for quaternary InP and regrowth samples.

Note: This is assuming the AHS filaments don’t need to be de-gassed, which is needed after system opening, BC pressure loss, the filaments get turned off (0 amps), etc.

Before operating H2 station

  1. The Hydrogen purifier should have the [N2 Feed] button pressed and lit

  2. All other buttons should be off

  3. Valve 5 (labeled E in Fig. 2) should be closed

  4. Check H2 station config in Amber, see Degas below

  5. Determine whether you will pump out through the Bravo LC or Echo LC. The steps are slightly different for each. The procedure is written for pumping out through Bravo’s LC. Pumping out through Echo’s LC takes about 15 minutes longer for each AHS bake.

Setup

  1. Close the GV between Bravo BC and Bravo GC to full close

  2. Press [N2 Feed] button (turns LED off) to stop the flow of N2

    1. Wait for the pressure gauge (D in Fig. 2) on the manifold to go to ~0 psi, or just dip a bit

  3. Press the [H2 Feed] button to start the flow of H2 into the purifier.

    1. Press the [H2 Pure] button to start the flow of clean H2 up to Valve 5 (but do NOT open Valve 5 yet), which goes into the H2 cell

      1. Always turn on the H2 feed first to pressurize line into purifier

      2. They used to be finicky

      3. If buttons are finicky, once the line is pressurized, you can turn feed off temporarily to mess with buttons to get both Feed and Pure on

      4. Watch filament temperature (should be 400 C) to make sure it comes back down after spiking (thermal load changing)

  4. Make sure the flow meter on the Hydrogen Station is reading ~25-30 (this is the Bleed Flow) [I am not sure what this is, psi?]

    1. Adjust the knob until ~25 if not set already

    2. Check water line flow for AHS, flowmeter is on Bravo manifold

  5. Setup the N2 purge/dilution flow to the scroll on the LC to dilute pumped out H2

  6. In operation (later on), pumping out through the Bravo LC means the Bravo LC and BC are connected, and closed off to the Echo BC. Pumping out through the Echo LC means the Bravo BC is connected to the Echo BC and LC, and closed off to the Bravo LC.

    1. Turn on Bravo (or Echo) Chamber UHP N2 to barely on, about 10-12 psi. There is a sign you can put on the Chamber N2 knob to prevent someone from accidentally turning it off.

    2. Double check both RM valves are closed (green and black)

    3. Bottom of N2 line stack goes to flowmeter for AHS purge/dilution flow. Flowmeter/valve max is set to scroll pump max.

    4. Check out Bravo (or Echo) backing pressure (BF PG1)

    5. Turn up flowmeter til it starts chirping, then turn it down til it stops

    6. Flowmeter ball should float around 1, wiggles a bit

    7. Backing pressure should be ~9 mT now

  7. Echo N2 purge setup is similar but a bit tricky and it has a wonky flowmeter (poor control)

    1. On Echo, remember to close both ion pump GVs. Both Echo buffer line GVs are open. Both GC GVs are closed.

    2. Can’t hide the Echo trolley, so keep it by the Echo HTS

    3. Follow above procedure halfway, but for Echo components: Steps 6b, 6c, 6e.

    4. Barely open up Echo N2 purge flowmeter

    5. Slowly open the N2 purge line quarter-turn valve (used for control) under the Echo table til flowmeter chirps, then back off a bit

    6. Flowmeter should be showing flow

    7. Backing pressure should be increased 50-100%

  8. To turn off (later on): close flowmeter, quarter-turn valve (Echo only), and N2 line.

  9. Prep trolley and blocks w/ samples as needed

    1. Trolleys can’t both fit in one BC where sample moving has to happen. Have to take blocks off, store them on places in BC, then move a trolley to a LC and close GV.

    2. When we get started, we will start off with a dummy block bake every time (empty block, just faceplates and backing wafer)

    3. 30 minutes for daily use; 50 minutes for used a month ago; adjust as needed

    4. Time all times on a timer, the actual bake time at 7 A is 6-10 minutes less due to current ramp up and down time (timer gets started at 5 A)

    5. DJI said that a dummy block bake is only needed for finicky samples (quaternary on InP) or if the AHS hasn’t been used in a few weeks. It will still clean samples for regrowth without doing an initial dummy block bake (as long as the AHS was used recently).

Move sample into Position

  1. Open Bravo/Echo Buffer GV

  2. Move the block onto Bravo HTS

  3. Move Bravo and Echo trolleys into Echo BC, unless Echo trolley is already isolated in the Echo LC. Be careful not to run the trolleys into each other.

  4. Close Bravo/Echo Buffer GV

  5. Note we don’t want chambers with active independent pumps open to each other for more than a few seconds due to cross-pumping.

  6. Open Bravo LC/BC GV to pump out the H2 through the turbo pump in the Bravo LC. Or setup the systems to pump out through the Echo LC.

  7. Close Bravo Ion Pump GV, quickly. And Echo’s if needed.

    1. The idea is to always have something pumping on each chamber without having pumps pumping on each other

    2. We want the ion pump pumping the BC as much as possible, but not when there’s H2 in the BC

  8. Ion pump WILL be damaged if exposed to H2

  9. System check:

    1. The sample should be on the Bravo HTS at this point, dummy block if first bake of day

    2. The trolleys are in the Echo BC and the GV is closed

    3. The ion pump(s) is isolated—GV(s) is closed!

    4. The Bravo LC turbo pump is pumping on Bravo LC and BC. Or Echo turbo Echo LC, BC, and Bravo BC.

    5. There is N2 flow through Bravo’s LC scroll

    6. Make sure the H cell filament current is at or below 2 A.

Ramp up and Clean

  1. Begin AHS filament current ramp up to 5 A, slowly

    1. Starts at 1 A, ramp to 5 A at 0.5 A/min

      1. This assumes the filament has already been out-gassed so the pressure in the BC does not increase above 1E-7 torr. Using the system without a proper filament degas is bad and can at minimum ruin your growth.

    2. Monitor the Bravo BC pressure and AHS voltage, current, and power output during this Step. We want to avoid spikes and jumps. Make sure the Eurotherm is tracking the actual set pt.

  2. Note: Ion pump GV(s) must be closed, this is critical

  3. At 5 Amps, slowly open (CCW) the flow of H2 into the H2 cell (i.e. open Valve 5) and observe the BC ion gauge.

    1. Be very careful with the valve, it is connected to delicate fittings and no longer produced. Don’t put weight on it. You may want to use two hands to open it, like an AMV.

    2. Valve starts to open at ~86.5, then more with each bit, notably more at 87 and so on. Needs to go to ~88.

  4. Check the H2 flowmeter is at the correct value, ~25-30.

  5. Increase and adjust the flow via Valve 5 until the BC ion gauge reads about ~1E-6 torr

    1. Or 5 to 8e-7 torr to prevent going too high when sample is heated. Or 7e-7 to 1.2e-6 torr.

    2. Be patient, and account for lag in the pressure changing. It’s helpful to look at the valve gauge see it move a bit then check the pressure.

    3. Roughly 86: 2e-7; 87: 4e-7; 88: 7e-7 torr;

  6. Start timer and monitor pressure, as well as current, voltage, and output power.

    1. The pressure may creep up (or down) a bit over time, but it’s generally a small amount.

    2. The current may spike when opening the valve. The output power may go to a higher or lower plateau during the bake.

  7. You can do a few small adjustments as needed during use, but don’t try to be too accurate or picky. The valve has hysteresis, slip and delay.

  8. Always watch for runaway pressure, can be very bad.

  9. Increase the H cell filament current up to 7 Amps at 0.5 A/min

  10. Keep at 7A for nearly rest of bake (til only 4 minutes left)

  11. Eurotherm example at 7 A: 14.5 V (to 20 V), 7A, output 18.2

  12. Target pressure should be ~1E-6 torr for the BC, keep an eye on the pressure gauge in case the pressure begins to creep up too high (or low). Adjust the H2 flow using Valve 5 so that the final pressure is no more than ~2E-6 torr.

  13. Check the H2 flowmeter is at the correct value, ~25-30, and monitor during use

  14. Seriously, don’t exceed ~2e-6 torr at any point

  15. Heat the Buffer HTS to required cleaning temperature (substrate dependent)

    1. Immediately turn up buffer heater station (HTS) temp to 450 C (or the necessary T)

    2. Prevents stuff sticking to it

    3. The oxygen-hydrogen reaction makes water and it’s critical it leaves and doesn’t stick

  16. Clean for required time at 7A

  17. Adjust H2 to target ~1e-6 torr if needed (<2e-6 torr!)

  18. Usually 45-60 minutes, with 50-60 min being common (30 minute minimum)

  19. Monitor all systems during bake (pressures, power, flow), and watch for runaway

  20. When bake gets to 4 minutes remaining, set current to 5 A (from 7 A)

  21.  

  22. Note that the timer starts when the H2 valve is opened and the current is at 5 A, so the time at 7 A is ~8 minutes less. The time at BC Bake T is ~18 minutes less.

End a Clean

  1. Close the flow of H2 into the cell (close Valve 5), which occurs at 80 (reads 99980). Don’t go beyond 80.

    1. The valve can go beyond 80, which can damage it and alter the new fully closed position.

  2. The BC pressure should drop to low -8s, or high -8s at worst

  3. Set the temperature of the HTS to 10 C

  4. Lower the current through the H2 filament to 2 A slowly (from 5 A) so it’s faster to go up to 5 A.

    1. If after last sample, set to 1 A and go to Step 23b

  5. Wait until the pressure is at least in the 7-8E-9 torr range / <5e-9 torr is better. Signifies most hydrogen is pumped out.

    1. This takes ~30+ / 45+ min from the low -8s. Add 25 minutes from the high -8s.

    2. Do not open Echo BC GV before 7-8e-9 torr because the Echo BC ion pump is on and lingering H2 can damage it. Recommended to wait til <5e-9 torr.

Changing samples

  1. Open Bravo BC ion pump GV and close Bravo LC GV

    1. Open Bravo/Echo BC GV and bring in Trolley

  2. Remove block from HTS and replace with next block

    1. First real sample if you just did dummy sample

  3. Move Bravo and Echo trolleys into Echo BC. Be careful not to run the trolleys into each other

    1. Close Bravo/Echo BC GV

  4. Note: For Steps 12-15, first bake (dummy) was 50 minutes on timer, which is 42 minutes actual at 7 A, 4 min ramp up/down on each side. For example, the next bake for this demo is 33 min timer, which is 25 min at 7 A.

  5. Go back to Step 9 and repeat as needed (for each sample)

    1. Pause at Step 18 for last sample, return here

  6. Note: You can run the Hydrogen station during growth, but you cannot open the GC to the BC until the BC pumps out spare H2 (Step 19). Also can’t open the GC to BC when BC is open to the LC (Step 19).

AHS Cleans Done

  1. After last sample and getting to Step 18, set current to 1 A

  2. Press the [H2 Pure] button (light should turn off) to isolate Valve 5 to the main manifold

    1. Press the [H2 Feed] button (light should turn off) to close the hydrogen feed into the main manifold

  3. Press the [N2 Feed] button (light should turn on) to allow the flow of Nitrogen to the main manifold

    1. Adjust H2 bleed flowmeter to 40ish (+/- 10), as it will have jumped up

      1. Follow Steps 19-21 (without putting anything to bake on the HTS)

  4. Return the trolleys to their respective systems and close the Bravo/Echo Buffer GV

    1. Check all GVs, and what’s pumping on what

    2. Turn off N2 dilution flow, see Step 11

AHS Degassing

  1. AHS filaments need to be degassed after: system opening, BC pressure loss, the filaments get turned off (0 amps), etc.

  2. Hydrogen is not turned on during this filament degassing!

  3. Check Software and Config file: this just means Amber is talking to Eurotherm and current source reacts

  4. Address: 20, Comport: 5, Modbus 2400 series

  5. Turn on AHS power supply

  6. Make sure Amber AHS setpt is 0.1 or below, so that Eurotherm can be switched to auto

  7. Switch AHS Eurotherm to auto

  8. Ramp up from 0 to 5 A at 0.2 A/min, with pauses each 1A for BC pressure rollover

  9. At low power, the power supply bounces between ~0-2 A

    1. On July 19th, 2015, took about 4 hours:

      1. At 2 A, got two peaks of pressure

      2. At 3 A, got two peaks

      3. At 4 A, got two little peaks and a big peak (1 hr)

      4. At 5 A, got a short long peak and a large peak (1 hr)

  10. Ramp up from 5 to 7 A at 0.5 A/min

    1. Wait 1 min

    2. Ramp down to 1 A at 0.5 A/min

Notes

  • H2 filament idles at 1 A, for however long as needed (months+)

  • N2 Feed stays on for however long (months+ as needed), at 20-40 psi, whatever it is set at

  • Block can go on whenever convenient as long as the block is on the heater station before you start opening the cracked H.

  • Check N2 scroll dilution flow before opening H to the block

  • Minimize Bravo load seeing Echo buffer, but it's OK for short times.

  • To shut off AHS filament:

    • Set current to 0.1 A in Amber, let it reach 0.1 A

    • Turn off power supply

    • Check Eurotherm is stable, can keep Eurotherm in Auto as long as it’s stable, otherwise turn it to manual.

AHS Manifold and Figures

Figure 1: AHS manifold schematic
Valve 1 = [N2 Feed] button Valve 2 = [H2 Feed] button Valve 3 = [H2 Pure] button Valve 5 = Needle valve between hydrogen purifier and Buffer chamber
Figure 2: Pictures of AHS manifold and current controller
Figure 3: Nitrogen ballast flow and scroll pump
[N2 Feed] open/close Valve 1 [H2 Feed] open/close Valve 2 [H2 Pure] open/close Valve 3 Pressure gauge in the manifold Valve 5 controls the H2 flow into the H2 cell Nitrogen ballast flow control

AHS Bake-out

EMK 2/2011

  • Valve positions:

    • OPEN: Manual valves: 6, 7, 8, 9, 10; Pneum valves: 3

    • CLOSED: Manual valves: 0, 4 (open during purge w/bleed flowmeter), 5; Pneum valves: 1,2

  • Pump/purge with N2 on both sides of the purifier

  • Pull vacuum on both sides of purifier through LC

  • Heat up Pd, ramp rate: ~8C/min , ramp to 400C in 45 min

  • Heat up heater-wire by variac, 90-100C

  • Monitor pressure via ion gauge on LC

  • Valve 6 should always remain closed when not in bake-out!!!

make sure 0 is open

  • Pull vacuum:

    • Valve positions:

      • CLOSED: bleed flowmeter, valve 4 (miniAMV), 1 & 2, 5

      • OPEN: 3, 6, 7, 10 then 8,9,

    • pump/purge with N2 on both sides of purifier

    • pull vacuum on both sides of purifier

    • close 8,9, 6

    • heat up Pd, ramp rate: ~8C/min, ramp to 400C in 45 min

  • Start flow:

    • close 10

    • open bleed flowmeter, 4 (miniAMV)

    • open 2

    • slowly open 10

    • adjust flowmeter

    • slowly leak 5, while checking pressure of BC

  • Stop flow:

    • close 2

      • allow H2 flow to cell decrease

    • close 5

      • pause.

    • open 1

    • adjust flowmeter to allow higher % of bleed

  • Idle

    • keep Pd hot

  • Shut down

    • open 1

    • increase bleed rate

    • evacuate H2 from pure side into BC (if not already done)

      • open 5, 3, 10

      • close 10, 3, 5

    • cool down Pd, ramp rate: ~8C/min, ramp to 20C in 45 min

Confluence Documentation | Web Privacy Policy | Web Accessibility