My lab is using an in-vacuum CCD camera to detect XUV light. The sensor is bare, it has no glass cover due to needing to detect XUV. Somehow this sensor got an oily film deposited on it in the past, probably from oil diffusing from a vacuum pump.

I believe I can just carefully clean off the oil with lens tissue and isopropanol or a similar solvent. It is silicon so it should be relatively robust. The biggest danger will be damaging electrical leads. Does this sound right?

I was wondering if anyone had any experience with this and words of wisdom. Unfortunately the camera is old and the company we bought it from wants nothing to do with it.

  • 1
    \$\begingroup\$ The surface is probably at least passivated with a thin layer of amorphous SiO2, so like a thin film of quartz. So I would guess reagent grade isopropanol would be okay, but would wait for a more authoritative answer. \$\endgroup\$ Jun 3, 2021 at 14:59
  • \$\begingroup\$ You may find this interesting, they used vapour-phase cleaning for CCDs contaminated with vacuum grease: eso.org/sci/libraries/SPIE2014/9151-202.pdf I suspect the laser-optics technique of laying a low-lint wiper on the offending surface, adding solvent on top, and then gently dragging off across the surface would work acceptably too. \$\endgroup\$
    – Cursorkeys
    Jun 3, 2021 at 16:05
  • \$\begingroup\$ Vendors like Thorlabs and Newport have instructions for cleaning precision optics with subwavelength thick coatings on them. This involves using forceps, pure alcohol and very clean lens paper folded WITHOUT touching it. The same principle can be applied to silicon sensors. Look up their instructions. \$\endgroup\$ Jun 3, 2021 at 16:19
  • \$\begingroup\$ thanks for the feedback so far. user1850479, you are referring to the drop and drag method, right? I agree that would probably be the way to go for using lens tissue \$\endgroup\$ Jun 3, 2021 at 17:04
  • \$\begingroup\$ You probably cannot use the drag method on a lot of sensor circuit boards since it requires an exposed, flat surface, so I was suggesting using the forepes method, which can work on curved or recessed surfaces. If you have a sensor that is exposed enough to drag and still get the corners, then I would try that first. \$\endgroup\$ Jun 3, 2021 at 17:11

1 Answer 1


I would use a method similar to the best cleaning methods DSLR cameras, there are many kits available. Many have no residue solvents that are for sensors (and many of those are isoproponol based).

There are two problems with cleaning sensors:

  1. Removing all the residue
  2. Not adding any additional material from swabs or tissues

The residue can't be radially removed without wiping it off or soaking it up, solvent will move the oil or residue around but won't remove it. Whatever tissue or swab will need to be the high grade cleaning variety with fibers are built to stay intact. Even then some fibers may be left on the sensor and can be removed with clean compressed air (many DSLR kits use a puffer to gently blow any remaining material off the sensor). Also watch the corners of the sensor or anything with edges as it will have a tendency to catch fibers from swabs (and like you say, don't catch any leads or wires). If there is any dust on the sensor make sure that you don't push it into the sensor as it could 'scratch' the sensor.

A few fears would be that there is some material that is sensitive to solvents, I'd check with the manufacturer for solvent compatibility if possible. If not then test a corner and see if there is any degradation, if not then clean the whole sensor. I'd imagine that the oil is probably just as bad as any solvent will be anyway.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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