I have some water-soluble crystalbond that is attached to silicon and, because of some overheating (to about 150 Celcius) during a final processing step, the crystalbond seems to have charred (white/grayish colour, see photo at the bottom) and is no longer dissolvable in water. Since this silicon is important and cannot be replaced, I really need to find a way to chemically remove this overheated crystalbond residue without using abrasives that could scratch of the thin SiO$_2$ coating on silicon.

Looking at the MSDS (see here), crystalbond is composed of the following

  1. 97% Nonylphenol Polyethylene Glycol Ether(127087-87-0) EC-No. 500-315-8
  2. 3% Polyethylene Glycol(25322-68-3)EC-No. 500-038-2

My question is: does anyone have any ideas for what crystalbond decomposes into around 150 C and how one could remove this overheated crystalbond residue?

Things I have tried (unsuccessfully)

  1. Hot DI water (in a cup, and in a ultrasonic bath)
  2. Acetone
  3. 2-propanol/ethanol
  4. "Piranha" etch (asked a lab friend to do this, too scary for me!!)
  5. Photoresist remover (lab friend also tried this)
  6. Hexane
  7. Dunking in LN2
  8. Putting on a hot plate at 225C for about 24 hours
  9. Burning under a 1100C torch for about 30 seconds

Any bright ideas, especially from organic chemists? All I can say is: damn this stuff doesn't want to come off.

Some examples of the residue below.

Lots of crystalbond residue on this Silicon wafer enter image description here

Another silicon wafer with crystalbond residue. Less than the one above, but still significant. enter image description here

  • $\begingroup$ Chloroform - 5% methanol is worth a try. $\endgroup$
    – Waylander
    Feb 25, 2021 at 9:31
  • $\begingroup$ @Waylander interesting suggestion, so is that 5% chloro in 95% methanol or the reverse? Any insight forwhy this suggestion might work? $\endgroup$
    – user157879
    Feb 25, 2021 at 10:07
  • $\begingroup$ 5% MeOH in Chloroform. In my experience this is a pretty good solvent system for a lot of organic matter. The crystalbond is organic and has a lot of lipophilic groups. As for what you made by overheating it I can only speculate. I would guess that you dehydrated some of the alcohol groups perhaps creating higher density polymers $\endgroup$
    – Waylander
    Feb 25, 2021 at 13:41
  • 1
    $\begingroup$ A possible option: paint remover from your local hardware or big box store. These are formulated to remove stuff that was intended to adhere well for a long time. Worth a try. And an ultrasonic bath always helps! $\endgroup$
    – Ed V
    Feb 27, 2021 at 21:00
  • $\begingroup$ Thank you Ed! Do you have any specific paint removers that you know are strong. I.e. what active ingredient to look for $\endgroup$
    – user157879
    Feb 27, 2021 at 22:43

3 Answers 3


Just a quick answer without looking too much into it.

You have been using a lot of polar solvents without any luck. The reason is that the molecule no longer has the groups that make it water-soluble because of the thermal decomposition. Only hydrocarbons chain remain. You need to use an organic solvent (toluene, benzene, hexane). Since you probably don't have access to a lab, I would try with gasoline.

Let me know if it works.


  • $\begingroup$ Thanks! Among the three non-polar solvents you mentioned, which would be best? I do know someone with access to hexane. $\endgroup$
    – user157879
    Feb 27, 2021 at 22:42
  • $\begingroup$ Hey, it seems that you do work in a lab. For the look of your Si wafer (I'm guessing is amorphous Si for the color) you were trying to modify the surface and perform a surface Science Experiment on a heated stage. I might be wrong. Either way, have you tried cleaning on an ultrasonic bath ?. You might not even need the solvent. In any case, I would try to use toluene first. $\endgroup$ Feb 28, 2021 at 2:08
  • $\begingroup$ Somebody recommended a paint remover. That might work as long as it is based on an organic solvent (e.g, xylene). $\endgroup$ Feb 28, 2021 at 2:44
  • $\begingroup$ Thank you! I will try to see which non-polar solvents I can find to try, especially toluene. Yes I did try ultrasonic bath in hot DI water (~75C) for about 20 minutes, but alas not much luck. If the non-polar solvent seems effective, I can try to use ultrasonic cleaning with it. $\endgroup$
    – user157879
    Feb 28, 2021 at 5:11
  • $\begingroup$ I won't have another chance to make these Si wafer devices for about 3 months, so I am desperate to find a solution. Thanks for your help :) $\endgroup$
    – user157879
    Feb 28, 2021 at 5:12

Crystalbond™ 555-HMP is an adhesive. The SDS lists only hazardous ingredients, none of which should decompose at 150$^o$C to char. The char is probably from the non-hazardous ingredients, probably something like a hot-melt adhesive.

The adhesive char now bonded to the SiO$_2$ layer probably involves carboxyl oxygens, and if the char is not removed by the Piranha solution, that solution may have been too weak, or not immersed long enough. Another cleaner for organic residues is potassium dichromate in concentrated sulfuric acid. This may be more potent than Piranha solution.

One easy method to try (probably with a low probability of success) is to drop the wafer into a Dewar of liquid nitrogen: the thermal shock might dislodge a char layer that is distinct from the SiO$_2$ layer, but those final surface atoms may be different from what was there originally. But if any char falls off, the resulting surface may be better cleanable.

Silicon is thermally oxidized at temperatures above 800$^o$C, and any char should dissipate in oxygen (air?) at temperatures significantly below that. Whatever you did to the silicon to produce the oxide, do that at a reduced temperature (so as not to undo what you did originally) for long enough to remove the char without modifying your original treatment.

  • $\begingroup$ Oh, very helpful indeed, thank you James! I will look to see if there are others who can do a longer Piranha or the potassium dichromate methods. But I can certainly try the latter two methods (LN2, heating to high temperature). Would vacuum heating also dissipate the char? Or is the oxygen/air required? $\endgroup$
    – user157879
    Feb 28, 2021 at 15:12
  • $\begingroup$ Probably the char needs something more than vacuum to go away. Oxygen or air could be aggressive enough if you give it some heat. $\endgroup$ Mar 1, 2021 at 16:09
  • $\begingroup$ I tried heating at 225C for about 24 hours, but no changes there. similarly, no changes when held against a torch at ~1100C for about 30 seconds. Liquid nitrogen also showed no changes. This stuff is really tenacious! $\endgroup$
    – user157879
    Mar 2, 2021 at 1:10
  • $\begingroup$ Maybe this is not a lab disaster, maybe you have invented something of great importance, like a permanent Magic Marker. Is silicon carbide black? How about blasting with a NaHCO3 sandblaster until clean? $\endgroup$ Mar 2, 2021 at 14:12

At this point, I would just go and finish burning that char in air. You need air to complete the combustion, don't do it under vacuum. 300-400 C should do the trick. Although, if you are very concerned about the temperature, just do a TGA to the crystal bond under air and determine the temperature where it starts to combust; then apply that temperature to your wafer isothermally for an hour or so as to minimize damage to the surface layer.


Not the answer you're looking for? Browse other questions tagged or ask your own question.