Bi1-xSbx on Si(111)

Bi1-xSbx on Si(111)

Day before:

  • HF dip and load sample. I generally run my STOs in the morning because they're short, but you can set up the Bi one to run overnight if you prefer.

  • Determine what the Bi / Sb BEP ratio should be for your targeted composition. The measured composition of Sb is about twice what the Sb composition would appear to be from the BEP ratio. The flux ratio is almost identical to the BEP ratio, and hard to determine from the XRD peak shifting of a poorly-lattice matched substitutional alloy like BiSb, so I prefer to relate the BEP ratio to the measured composition directly.

Growth day:

  • Morning: Set Bi tip to desired temp. Bi base and tip will heat up. Ramp up time ~ 1 hour. Important: also heat up Al (1150 C) or Ga (if the Al shutter is bad). This will be used to getter the Sb.

  • Go to lab

  • Put Sb sublimator in manual, power output to 72

  • Move trolley from LC to BC. Close gate valve. Load first Si wafer onto HTS. Move trolley as far as possible from HTS. Bake for 1 hour at 450 C.

  • Set HVP to 25

  • Record background pressures

  • Make sure CAR is in transfer position

  • Run Sb coarse STO - because the Sb sublimator is optimized for III-V growth with a much higher Sb BEP, the small Sb BEP needed is right around the edge where the valve starts to open. The open position varies quite a bit between individual growth days. The purpose of the coarse STO is to determine where the Sb valve opens.

  • Run Sb fine STO - now that you know where the Sb valve opens, run an STO with a step size of 1 around this range. Because the BiSb band structure is extremely sensitive to the Sb composition, and BiSb is a substitutional alloy, not a stoichiometric compound (i.e. any composition is achievable), it is very important to be precise about the Sb composition.

  • Note - Sb STO recipes open Sb valve to 140, then to desired position. Failing to do this can prevent the Sb valve from opening fully. According to Siff and Rodolfo, the Sb valve experiences hysteresis, so it's important to distinguish between BEPs measured when the valve is closing vs opening. Because we are opening to a higher valve position first, the BEP will always be measured when the valve is closing.

  • Run Bi STO.

  • Double check the BEPs manually. Make sure the BEP ratio is still what you intended using the actual Sb and Bi BEPs.

  • Turn off beam flux IG

  • Idle the Group III cell

  • Ramp up RHEED to 1.5 A and 10 kV

  • After HTS bake, set HTS to 10, wait for temperature to drop below 250 C

  • Move first sample from HTS to GC

  • Move first sample from HTS to trolley

  • Set substrate heater to manual

  • Open GV between GC and BC, then shutter. Make sure to lock shutter in open position

  • Transfer off RHEED block

  • Transfer in first sample

  • Close shutter, then GV

  • Put substrate heater back in auto, set to 450 C for away bake (10 - 15 minutes)

  • Transfer second sample to HTS, start second sample HTS bake

  • After away bake, start "deox" bake (not a true deox, actually desorbing hydrogen passivation + any crap)

    • Rotate CAR to growth position

    • Open RHEED to check wafer has made it

    • Start substrate rotation

    • Set substrate temperature to 750 C (thermocouple)

    • Stop substrate rotation (rotating above 800 C may damage the CAR bearings)

    • Set substrate temperature to 990 C (thermocouple)

    • Deox bake for 15 minutes at 990 C thermocouple. This should be 740-760 C pyro. Since it's the max temperature, and I use it every time, I don't check with the pyro on every growth. The pyro doesn't monitor the low temperatures I need for growth very well, and black body doesn't work well for Si, so I use the thermocouple (TC) for everything.

  • After deox, cool down and face cryoshroud

    • Set substrate temp to 5 degrees below desired growth temp

    • When TC < 750 C, start substrate rotation

    • When TC < 650 C, stop rotation. Manually rotate CAR past transfer position to the hard stop. The CAR is now facing the internal GC cryoshroud. The purpose of this is to prevent crap desorbed from chamber walls from 990 C to land on cooling sample, and to speed up cooling

  • Wait for TC to be about 10 degrees ABOVE desired growth temp. This takes ~1.5 hours.

  • MAKE SURE to come in before TC reaches set temperature - otherwise significant power oscillation will occur and overly increase the temperature

  • Remember to cool down HTS once second sample has finished baking

  • Set substrate heater to manual, power output 0

  • After first sample reaches set growth temperature, prepare for BiSb growth

    • Manually open Sb valve to 140.

    • Manually set Sb valve to valve position calibrated with STO

    • Do not open Sb shutter

    • Rotate CAR in to growth position

    • Start substrate rotation, 5-10 RPM

    • Run BiSb recipe

    • This order is important to avoid blasting the wafer with Sb (at the 140 valve position) in early growth

  • Monitor with RHEED and record the nominal thickness of the double lined to 1x1 RHEED transition. This is the point where the (012) to (001) transition should occur in XRD (corresponding to the A17 to A7 structural transition). Recording this each time can help us determine the transition dependence on substrate temp, growth rate, etc.

  • After growth, promptly transfer first sample to trolley

  • Replace with second sample and repeat as applicable

  • After finishing all growths

    • Transfer RHEED block back in

    • Move trolley with grown wafers to LC

    • Ramp down RHEED

    • Idle Bi cell

    • Set Sb sublimater to auto at same temperature as it is at the end of growth

    • Turn on BF gauge

    • Set HTS to normal temp

    • Any other standard procedures I forgot to list

    • Unload sample and load for next grower