Resist Handling - Best Practices Guide
Storing and Handling Resist
Resist should NEVER, under any circumstances, be pipetted directly from the stock (factory) bottle! Doing this risks contamination of hundreds (sometimes thousands) of dollarsí worth of resist, not to mention potentially ruining many peoplesí processes by unknowingly altering or contaminating the stock resist. Always use a small bottle labeled with your name, the date, and the type of resist to store your own supply. Any contamination from particles on the pipette or other sources is contained this way.
A supply of small glass bottles is available on the wire rack near the Vistec; feel free to use these as needed for personal resist storage. Labels are kept either near the personal resist bottles in the hood or on the wire rack near the sample bottles. The bottles come straight from a factory and may be contaminated with any number of things, so itís a good idea to thoroughly rinse them out with the solvent for the resist youíll be putting in them (e.g. anisole/A Thinner for A-series PMMA) to remove any residual particles and chemicals.
Once you have a clean personal bottle, carefully decant (pour) a small amount of resist from the stock bottle into your bottle (assuming you want to use the stock concentration; if you don't, see the section on thinning below). Label your personal bottle with your name, the date, and the type and concentration of the resist (e.g. PMMA C4).
All e-beam resists except HSQ can be safely stored at room temperature in glass bottles; the primary storage area for e-beam resists is in the drawers underneath the spinner (stainless steel) fume hood in bay 5. For HSQ, there are pre-poured small plastic bottles inside the refrigerator in bay 5 (on the floor near the SEM sputter coater). Take one and put a label with your name, the resist type, etc on it; that's your personal bottle. Don't even touch the stock HSQ bottle, as it contaminates very easily.
Diluting resist from the stock concentration without contaminating it can be deceptively tricky. Dedicated glassware is your friend; try to keep a supply of beakers, graduated cylinders, funnels, etc that are only used for handling and diluting a specific resist/solvent combination. If forced to use non-dedicated glassware, thoroughly rinse and solvent-clean it before starting.
There are dedicated glassware kits for some common e-beam resist types available to Vistec users. These are kept in locked toolboxes on the wire rack by the system. To get a key, see Bryan. Anyone caught cross-contaminating any glassware in the kits will have their kit key revoked immediately.
Using the compatible solvent/thinner for your resist, thoroughly rinse all the glassware you plan to use before starting. This will ensure that any residual liquid in the glassware is resist-thinner and not some other chemical.
When mixing small amounts (<100 mL) of diluted resist, a graduated cylinder is the easiest way to measure quantities. A funnel can be used if the cylinderís opening is too small to accurately pour into.
Before starting, calculate exactly how much resist and thinner youíll need to get to your desired quantity and dilution. For example, to make 100 mL of A2 PMMA youíd need 20 mL of A10 PMMA and 80 mL of anisole/A Thinner, assuming equal densities.
Measure out the resist first, carefully pouring it into your graduated cylinder or other measurement glassware of choice. Keep in mind that thicker resists will be viscous and slow to pour. Go slowly and allow any resist on the glassware sidewalls to drip down to avoid overshooting your desired quantity.
Pour the measured quantity of resist into your storage bottle. Itís not necessary to get every drop of resist out of the cylinder, as weíll see in the next step.
Using the same graduated cylinder, measure out the quantity of thinner you calculated out in the previous step and pour it into the bottle with the resist. The solvent should dissolve any resist remaining in the cylinder and transfer it into the bottle, which is why itís important to pour the resist before the solvent.
If you plan to use your diluted resist in the next 24 hours, itís a good idea to use a magnetic stirrer bead to thoroughly mix it up. Make sure the stirrer bead is clean, rinse it with thinner (again, itís a good idea to use resist-dedicated stirrer beads), and stir the resist for 5-10 minutes. The bead can be left in the resist bottle if itís difficult to get it out. If you let the resist sit for more than a day, diffusion will generally take care of the mixing process for you and eliminate the need for this step.
Triple-rinse all the glassware you used in the dilution process with thinner in order to remove all traces of resist from the glass. Following this, perform a full solvent rinse of your glassware and allow it to dry thoroughly.
If the resist is not fully rinsed off the glassware prior to the solvent rinse, certain resists (such as PMMA and PMGI) can form gummy precipitates when exposed to acetone, methanol, and/or isopropanol. If this happens the precipitate can usually be dissolved away by soaking it in resist-thinner for 5-30 minutes.
Make sure you label your bottle of resist with your name, the resist type and concentration, and todayís date. Incorrectly-labeled bottles are subject to being thrown away without warning.
Since youíll never get the concentrations exactly right when diluting resist, itís a good idea to spin up a sample or two and measure the thickness with the ellipsometer, profilometer, or film-thickness analyzer to see how it compares to the manufacturersí spin curve for the concentration you were aiming for. Once you know the offset, it should be relatively simple to get the thickness you need by varying the spin speed.
Spinning resist is accomplished by transferring a small amount of resist onto your sample via plastic pipette (located under the bay 5 spinner hood) and running the spin recipe of your choice (see the bay 5 spinner SOP? for operation details). Blowing off the tip of the pipette with a nitrogen gun and avoiding letting the tip come into contact with anything but your resist will minimize particle contamination. If you need to set the pipette down, do it so the tip is hanging over the sink, propped up, or otherwise suspended in midair.
For small pieces, a few drops of resist will probably be enough to cover the whole sample with liquid prior to spinning. For 4Ē wafers and other large samples, a puddle of resist in the center about the size of a half-dollar is generally enough to uniformly coat the sample. Note that for thicker resists and slower spin speeds youíll need to put more resist on the sample to get an even coat. Streaks of uncoated area at the edge of a wafer are a good indication that you didnít use enough resist.
The pipettes are disposable; throw them away in the resist-waste bin when youíre finished with them. If you accidentally touch something with the pipette tip, throw it out and get a new one; a wasted pipette is much less of a problem than potential particles in your resist film.
E-beam resist, with a few exceptions, is fairly robust when it comes to post-application baking. Hot plate baking generally works well, except in cases of a combination of a large sample (e.g. a 4Ē wafer) and a very bake-sensitive resist (e.g. NEB-31). In those situations oven-baking will give better uniformity but takes much longer (30-60 min., vs. 5-15 min. for a typical hot plate bake).
Post-application-baking resist has two purposes: to drive out the solvent in the film via evaporation or boiling, and to allow the film to relax into a stable configuration. For best results, youíll want to use bake temperatures that are well above the solvent boiling point and glass transition temperature for your resist but still below the thermal degradation temperature. All of these temperatures should be in the datasheet for the resist.
For resists with particularly volatile solvents (HSQ, for example), a post-application bake isnít generally necessary, since the solvent will boil/evaporate off and the film will relax on its own at room temperature.
Resist development is another process step that seems simple, but where quite a few things can go wrong and affect the repeatability of your process.
One of the most important things to keep in mind when developing resist is that almost all development reactions have temperature-dependent rates, sometimes to the point where a difference of a few degrees can measurably alter the development rate. In general, developer stored in the same place in a climate-controlled clean room will always be at the same temperature so in many cases you can get away with ignoring this, but it never hurts to measure and record the developer temperature, particularly if youíre doing very high-resolution work with a small process window.
An important thing to keep in mind is that mixed-solution developers (e.g. IPA:MIBK solutions for PMMA development) will usually react exothermically when first mixed, which can raise their temperature a few degrees. This has confused many, many process engineers over the years, as their freshly-mixed developer does not behave the way they expect it to. Make sure any newly-mixed developer has had at least a few hours (preferably a full day) to reach thermal equilibrium before using it (you can shortcut this by pouring out a small amount, which will usually take less than an hour to thermalize), and be sure to check the temperature to be safe.
As with everything else resist-related, cleanliness goes a long way toward process repeatability. Use dedicated glassware (glass petri dishes are ideal for immersion development) for each type of developer/rinse, and solvent-clean all glassware with acetone, methanol, and isopropanol before and after each use. Finally, rinse the glassware out with a small amount of developer after cleaning it and prior to using it to ensure that any residual liquid stuck to the glass is developer and not a possibly process-contaminating solvent.