Membrane Transfer
Membrane Transfer Process
NTS, 07/01/2013, special thanks to Jonathan Kurvits at Brown for most of the procedure. Updated NTS 20141030; 20160224, 20170608.
[Note: This process requires an AlAs etch release layer, so there is no easy way to make dummy samples— they must be grown.]
[Note: Be mindful of the gear you use for this process, and when and where you do it. HF is involved, as is grown III-As, and the etch can take a long time.]
Note: There are two Wax Application methods.
Apply Wax: Method 1 Solvent
There are two solvents you can use: Xylene and TCE.
The below procedure was written for xylene. The basic idea is the same for TCE, except you use the TCE labware and only apply one thin layer of dissolved black wax.
To use TCE, sub in TCE wherever you see xylene below.
Prep
Takes four hours with xylene, and one hour with TCE
Dissolve Apiezon W black wax in xylene
Put small piece of wax into beaker
Pour from the xylene bottle to a xylene-only beaker first (to prevent the wax solution from being too thin)
Pour xylene (~1-5 mL) into wax beaker and wait for wax to mostly dissolve (proceed on boat)
Make foil boat for samples (of proper size)
Be sure to have level surface and boat for applying wax
Cleave samples if not already done
Place samples into foil boat, membrane side up (so the wax is dropped onto the membrane), ~3 cm apart. They will remain in the boat while applying wax and heating.
Wax App
Pipette the solution onto a freshly cleaved sample, carefully. One drop will spread, but 2+ drops are probably needed (for a 1 cm x 1 cm sample).
Pay attention to see if top-side edge-adjacent (TSEA) material or corners are having trouble getting coverage.
Let the sample dry in the fume hood for ~20 minutes
Place it on a hotplate at 110C for ~30 minutes to bake out the rest of the xylene.
[Note: Generally after the first application and baking the wax is pitted in the middle.]
[Note: The heating process lets the wax flow a little and cover missed corners.]
Repeat 4-6 times to get uniform coverage, and the desired thickness and a slightly dome shaped wax layer. Do not repeat for TCE, as we only want one thin layer of wax, and it will not be dome shaped.
For a 1cm^2 sample this usually takes 4-6 applications
If not flat or dome shaped by layer 3, you may want to slightly tilt the sample boat after dropping the wax on while it is air drying.
Putting a few flat wipes or a small but sturdy piece of foil may work.
Angles in the range of 5-15 degrees are permissible. Any more may risk excessive side wax.
We don’t want the wax too thick or it might crack the membrane, so I lowered the application rec from 6-8 to 4-6.
To get OK coverage at all corners you might have used a lot of solution, which may lead to overflow. Side swabbing can fix this, so care more about getting good top side coverage.
You don’t need perfect coverage at all edges and corners.
[Note: Overflow in the boat can carefully be soaked up using paper kimwipes folded very thin and thick.]
Inspect the sample, paying attention to the wax on the top and sides.
Clean sides and bottom
Make sure no wax got on the sides of the sample (it will). If it did (it does), clean it off with xylene and a swab after all layers are applied.
You must wait for the wax to harden considerably before cleaning it, otherwise the wax will be soft and sticky and become hard to work with, and potentially be damaged.
This might mean waiting days for the wax to harden for the Solvent method. Xylene can take a while, TCE may be ready in under 12 hours (or not). The Heat method may be read to swab just one hour after cooling.
Cleaning the sample sides will degrade the TSEA wax, so you must be very careful to degrade this wax as little as possible.
You have already inspected the TSEA wax, so make sure that when you’re done cleaning the sides, the TSEA hasn’t degraded much.
It is most important to completely clean the side so the etch layer can be etched. A good way to ensure this is to swab the sides until clean and accept the degraded TSEA wax.
If you can see a tiny TSEA halo on your sample and the sides are clean, you should be good.
Generally, I start on the back side with a larger swab, and then some smaller swabs. I then move around to each side. I then go back to the back, and then to each side again, and repeat as necessary, usually just a third time. Each cycle should be a lower degree of cleaning needed. I look at the sample under the microscope when almost done with each side on cycle two and three.
Apply Wax: Method 2 Heat
Prep
Takes ~one hour
Prepare little pieces of Apiezon W black wax
Size depends on the size of the sample you want to cover. Generally you will want somewhere from rice to pea sized, and sort of in the shape of your sample.
The wax is sort of brittle, and slightly malleable, so use a careful combination of cutting and breaking off
The wax can go flying, so containing the piece you're working with is important. A kimwipe or UHV foil can help with this. We don't want the wax getting into other items or processes.
Make foil boat for samples (of proper size) with a level bottom
Cleave samples if not already done
Place samples into foil boat, membrane side up (so the wax is placed onto the membrane), ~2 cm apart. They will remain in the boat while applying wax and heating.
Wax App
Using tweezers, carefully place each little piece of wax onto the intended sample
Place the boat on a hotplate set to wax melting temperature, which will allow the wax to slowly melt
Al top hotplates should do this at 150 C, and the old ceramic top may have to go to 250 C
The wax shouldn't melt right away, it should take a bit of time. We want to have the wax flow but keep it viscous.
If the wax isn't melting and flowing slightly within 5 minutes, bump up the temperature-- don't add more wax
A good amount of wax and proper temperature should cover most of the sample in 30 minutes
Once wax flow and some coverage are achieved, you can assess whether more wax must be added to get proper thickness of wax or to get decent edge coverage. Remove the sample from the hotplate and let it cool before adding more tiny pieces of black wax.
We don't need perfect to the edge/corner coverage, but do want to maximize the transferred membrane area
I have seen that the wax in a flat boat tends not to go running over the sample edge, due to cohesion
A small piece of UHV foil under the boat may help steer the wax flow, but you have to be more careful about the heat and where samples are in the boat
The wax should be somewhat dome shaped, and tend toward having more round topography than abrupt topography
We don’t want the wax too thick or it might crack the membrane. The proper range is ~1-2 mm.
To get OK coverage at all edges (and corners) you might have used some tilting or extra wax, which may lead to overflow. Side swabbing can fix this, so care more about getting decent top side coverage.
You don’t need perfect coverage at all edges and corners.
[Note: Overflow in the boat can carefully be soaked up using paper kimwipes folded very thin and thick.]
Inspect the sample, paying attention to the wax on the top and sides.
Clean sides and bottom
See the Clean sides and bottom section for the above section, Method 1 Solvent
Release epitaxial membrane from substrate
Takes Overnight (18+ hours), leave a proper note x2: HF, Name, PI, phone number, date, time
Setup acid bench, put dipper in bench, and place samples wax side UP into dipper.
Pour 49% HF into HF Etch beaker, carefully
MRC provides HF at 49% (~50%)
Put dipper into Etch beaker with 49% HF solution
The lateral etch rate of AlAs in 10% HF is ~1 mm per hour supposedly, so a 1 cm^2 sample should take 5-6 hours to etch in 10% HF, but it takes ~18 hours in 49%
Ideally, after the etch is complete, the wax layer and membrane will float to the surface and you can handle it carefully with tweezers.
Usually, you will have to jiggle the dipper in the acid after leaving it overnight, and then the membrane will release (if the etch is done) and float
Once the membrane is released, prepare the Rinse beaker with water about 6-10 mm from the brim and carefully move the dipper from the Etch to the Rinse beaker, capturing the membrane in the dipper and letting it float on the Rinse water.
After a minute use tweezers to take the membrane out, dunk it 3x, and put it on wipes.
Flip it over. This lets the wipes absorb-dry the sample. Repeat flipping it over if needed.
Repeat above two steps if needed & membrane isn’t drying (dunk 3x and flipping over to dry).
Put it into your sample holder. Repeat as needed for liftoff samples.
Use a very light N2 spray on each side of it if desired.
Rinse the substrates and gear to clean up.
If it doesn't release and float to the surface, give it more time. If it has been a long time (~72 hours), the wax layer can be removed from the substrate using a pair of tweezers after removing it from the HF.
Work a pair of Delrin tweezers under a receptive corner and move the tweezer tip around the edges. Work on getting the tweezer further under the membrane, and so on. Be careful.
Carefully rinse sample and substrate to remove HF (as well as tweezers and gear)
Your best option is probably keeping the sample and substrate in the dipper and doing a dunk rinse, or light cascade rinse or spray. A normal sprayer is probably too forceful.
Transfer membrane
Prepare target substrate
PDMS or Kapton
Right now we have used PDMS and Kapton tape, both of which are sticky(ish).
PDMS needs to be (made and) cleaned of dust by kapton tape ‘exfoliation’. Then put it on a glass slide and activate it by performing a 5 minute UV exposure on the MJB4 (365 nm) or shorter wavelengths.
Kapton tape needs to be cut to size and rectangular, and have no bunching anywhere or defects in the center. Put it on a glass slide, adhesive side up. Then put some solvent on a wipe corner and wipe it on the tape ends to make the adhesive less effective (but not sludgy or totally gone). This is for easier handling.
We have copper tape, but have not attempted BETA transfer yet.
Put target substrate onto clear glass slide for handling
Put membrane (with wax) onto target substrate, with membrane facing down.
Apply gentle but firm pressure so the membrane attaches to the target substrate
Lift up the glass slide and checkout your progress from the top, side, and bottom (don’t flip the glass slide over, lift it up).
Repeat as needed, make sure the center, sides, and corners are down. All of it. Especially the corners and sides.
Bake in 90 C oven when done: PDMS 2 hours; Kapton Tape ~45 minutes.
Let cool for 2+ hours
van der Waal’s method
Put a small drop(s) of DI water on the target substrate and set the film + wax on top
One drop per ~3.5 mm x 3.5 mm
Press down gently but firmly—some water might leave, some will resist
Use a corner of lens tissue paper to slowly remove the excess water from under the sample
Bake at 105 C for 1-2 hours
Let dry in fume hood for a few hours so that the water evaporates and the film bonds to the glass by Van der Waals forces.
[Note: Typically let it dry for a day to be sure the bond is complete.]
Rasha has found that putting the sample under a 2 kg weight for 24 hours helps with the bonding.
Remove Wax
This procedure was written for the traditional ~1 mm domed wax on a sample. For thin wax skip, down to the TCE + wipe step below, or possibly even the solvent dunk.
Prepare a beaker with xylene, a pipette, and wipes for soaking up wax
Get a large paper kim wipe and keep cutting it in half until you get 32 ~4.5” x 2.25” pieces
Drop two (or one as appropriate) drops of xylene onto the wax, and wait for ~10 seconds. *While waiting fold up (thrice) the wipe pieces along the length direction so they’re now thick and long.
Hold the wipe piece at the ends and slowly lower it to the now dissolved top layer of wax so that it can suck up the wax solution. Don’t press down on the wax.
The glass slide should be turned to minimize how much wipe is over the substrate while still being over the whole membrane.
Repeat a few times to get a lot of the wax off, perhaps half or most. Once you can start to see membrane (or 15 times, whichever comes first), stop.
If you continue to suck up the wax solution you will also suck up the membrane, damaging it.
Do the same solvent dropping and wipe folding, but now use TCE, and instead of sucking up the wax solution, tilt the glass slide to one side (a short side) so the solution flows that way and off the membrane. Then suck up the solution with a wipe. Repeat til done.
Clean the target substrate where there is no membrane carefully if needed, e.g. a TCE or xylene swab or two, or soaking up wax solution underneath it or at the ends.
There have been some variants of this cleaning technique, with submersion of the sample in solvent at various times. If trying this, I’d recommend doing so only after the initial xylene + wipe steps, so your solvent doesn’t turn dark brown right away. Preferably you can wait til most of the wax is off and use the solvent dunk as a nearly final clean.
Appendix
Notes from John Kurvits
“Let me start by saying that although I've had enough success with this process for my purposes, the films still end up cracked, which could be a problem depending on your collaborator's application. If they have issues with cracking I'd suggest they try using thicker wax layers or even sticking a Teflon sheet on top of the wax to use as a handle after etching so the sample doesn't have to be directly handled with tweezers. A summary of my lift off and transfer process is written below. Let me know if you have any questions.”
“Dissolve Apiezon W black wax in xylene and pipette the solution onto a freshly cleaved sample. Let the sample dry in the fume hood for ~20 minutes and then place it on a hotplate for ~20 minutes to bake out the rest of the xylene. You’ll have to repeat this process a few times to get the desired thickness and a slightly dome shaped wax layer. For a 1cm^2 sample this usually takes 3 applications. Finally, make sure no wax got on the sides of the sample. If it did, you can clean it off with xylene and a swab.
Once the sample is prepared, place it wax side up on the bottom of a beaker with 10% HF. Ideally, after the etch is complete, the wax layer, along with the thin film, will float to the surface and you can handle it with tweezers. If it doesn't release and float to the surface, the wax layer can still be removed from the substrate using a pair of sharp tweezers after removing it from the HF. The lateral etch rate of AlAs in 10% HF is ~1 mm per hour, so a 1 cm^2 sample takes 5-6 hours to etch.
To bond the film to glass put a small drop of DI water on the glass, set the film on top and use lens tissue to remove the excess water. After a few hours the water evaporates and the film bonds to the glass via van der Waals forces. I typically let it dry for a day to be sure the bond is complete, and then use xylene to remove the wax layer.”
“Typically I make my samples are 1cm^2. Samples rarely release from the host substrate on their own, so I slide a pair of sharp tweezers or a razor between the film and the substrate to remove them. I'm fairly certain that this is what causes cracking in the sample. Most cracks are fairly parallel to each other and span the distance of the sample. Therefore, I think they form due to bending the sample slightly while peeling it off the substrate. If you're having significant issues with cracking I'd suggest that you suspend the sample in HF with a Teflon string bonded to the wax. This way, the weight of the substrate will help with separation.”
“I haven't had any issues with bond strength. For all of my samples, the bond has been strong enough that I can do drop and drag cleaning with a solvent without damaging the film.”
“I'm not sure what the concentration of the black wax xylene solution is. I made it by adding xylene to a beaker with a small piece of wax until the wax was fully dissolved. I bake out the xylene at ~110C for 30 minutes. Also, note that I usually apply the wax several times. Generally after the first application and baking the wax is pitted in the middle. Typically I bake out the xylene and add more wax four or five times to get uniform coverage. Finally, no I don't use a hot plate or sonication during the HF etch, in fact some sources suggest chilling the solution to 0C during the HF etch to prevent bubble formation that can prevent undercut.”
Notes from NTS
Membrane Liftoff
Trying to transfer membrane
Trying alternate etch handles
Wax App
Cleave or grind edges - not needed
Liftoff
Strong acid - YES!
Heat acid - untried
stir bar - untried
Etch Handles
Copper Tape - not really
Kapton - no
Nail polish - curls up
Optical glue (w/ Feng?)
Bonding to Substrates
Pressure + bake – yes for PDMS (+UV) and Kapton
water droplet – yes, onto silicon
Nail polish - untried, will retard stress transfer
PR - untried, will retard stress transfer
Removing Wax
Drop xylene and suck up with wipe
Can damage membrane
Wax release layer? – untried (what do I mean by this?)
Melt or melt w/ solvent?
Ablation (w/ Feng?)