.
WHERE TO BUY HSQ
.
DisChem Inc. (USA)
EM Resist LTD (UK)
Applied Quantum Materials (Canada)
See also links at AQM for distributors in China and Japan.
This presentation shows a comparison between Dow and Dischem HSQ. They are very similar. Dow no longer produces HSQ.
.
HOW TO STORE HSQ
.
You can purchase dry powder HSQ, mixing it in MIBK as needed. However, it is much more convenient to buy it in solution and then to store it in liquid nitrogen.
Dispense the liquid HSQ into little 4 ml bottles (VWR 414004-151) which are made of HDPE. Rinsing the HDPE bottles is not helpful, since MIBK itself is often contaminated with metal ions. At VWR, search for part 414004-151
HSQ can be stored indefinitely in liquid nitrogen. We use the Worthington HC35 dewar, which holds around 400 4ml bottles in ten bins immersed in liquid nitrogen. If you prefer to use small 2ml vials, then you could use a storage dewar such as the Worthington LS4800 or the MVE Cryosystem 4000, which hold boxes of vials in cold nitrogen vapor.
When thawing the little HDPE bottles, the caps (low density poly) expand at a different rate, and so the bottles can leak a little as they warm up. You can place the plastic bottle inside a small glass “scintillation” vial while it warms up to room temperature, or simply let it sit in a fume hood.
We use smaller 4ml HDPE bottles, each of which can be used for ~2 wafers. Each person takes their own little bottle, and warms it up just once. We do not stack the bottles neatly in the dewar- they tumble about randomly. We keep different types in different tubes, but inside the tube it’s just a jumble.
An old data sheet from Dow tells you to store it above 0C, but that’s wrong. A freezer is better, and liquid nitrogen is the best, as you can see in this presentation.
Syringe filters (0.22 um, with 5ml syringe ref. 309647) help a lot to reduce streaking, since some of the little bottles are not as clean as they should be.
You might also want a stock of transfer pipettes, such as 14670-147-PK.
.
WARMING IT UP
.
You can warm up the bottle in your hands (with gloves on) or you can put the little 4ml bottle inside a glass vial to carry it around while it thaws. It takes just a few minutes to reach room temperature.
We have never seen the caps pop off the small 4ml bottles, but that might happen with the larger bottles. It’s best to make sure the cap is loose as it warms. The poly caps expand faster than the HDPE bottles, and a little MIBK will escape while the bottle warms up, but this is a good feature. If the vial forms a little crack as it cools, then air and water could enter as the vial freezes. When the vial warms up, you could get an “ice dam” over the crack and then the gas inside the vial might cause it to burst. So, it’s a good thing that the bottles leak a little.
.
SPINNING
.
We often use syringe filters to dispense the HSQ onto the wafer. Sometimes the little HDPE bottles have a bit of junk inside, so filtering can reduce streaking quite a lot. Using filters to dispense the HSQ also eliminates the need to wash the little bottles before use. Washing the bottles with cheap MIBK is probably worse than having a bit of dust. Also this is less work.
We do not bake the Dow HSQ after spinning, but it will do no harm. We found that the Dischem HSQ should be baked for 2 minutes at 120ºC to improve adhesion.
You should not measure the thickness until after exposure and development, since the film condenses considerably. Print some large rectangles that you can later scan with a profilometer. HSQ will shrink about 6% when it is cured by exposure.
.
EXPOSURE & DEVELOPMENT
.
At 100 kV the dose will be around 1 mC/cm2 (that is, 1000 uC/cm2) – very roughly, when you are using MF312 (TMAH) as the developer.
Development for 4 minutes in MF312 is our standard procedure. Rinse in water.
For low-stress HSQ films, develop in a weak developer such as MF319. This will require a lower dose.
For high-stress and high resolution, develop for 10 minutes in the MIT salty solution:
1 liter water
40 g NaCl
10 g NaOH
This will require a higher dose. Here is another paper about salty developer. This salty developer will attack bare silicon and lithium niobate.
25% TMAH can be used as a high-resolution developer (room temperature or heated to 90°C) but this will attack silicon aggressively. Normally, the MIT salty developer is better. The only exception is when using lithium niobate, which etches in NaOH.
If these strong developers etch your substrate, then keep in mind that HSQ can be annealed at up to 900ºC after development. You can develop with a weak TMAH solution, then later bake the wafer to increase the resist’s density (and stress).
.
INSULATING SUBSTRATES
.
You can use HSQ on quartz or sapphire wafers, if you cover it with a conducting layer. There are several ways to do this- the expensive way and the cheap way. See this presentation, which is a bit out of date, since we should also mention DisChem’s conducting polymer.