How do you clean glass optical windows?

Cleaning the Optical Window and Reflectors

WARNING! 

For more information, please visit our website.

Use ONLY Industrial Strength Windex® with Ammonia D (General Monitors PN -1) to clean the optical window. The optical window material is sapphire, not glass. Using any other commercial glass cleaner will cause damage to the optical window. The warranty does not supply coverage for optical windows that are damaged in this way.

Failure to follow this warning can result in serious personal injury or death.

Remove debris and film buildup on the optical window and reflectors regularly to make sure that the device has the correct sensitivity and to prevent COPM Faults. General Monitors recommends that you clean the optical window and reflectors every 30 days at a minimum. Clean the optical window and reflectors more frequently for devices that are installed in dirty areas.

1.

Additional resources:
How to Choose Optical Mirrors?
Transmission efficiency - acrylic vs 50/50 BK7 glass prism

For more information, please visit ruiqi.

Use a clean, soft, lint-free cloth, tissue, or cotton swab soaked with Industrial Strength Windex with Ammonia D to clean the optical window and UV/IR reflectors.

2.

Use a clean, dry, cloth to rub the optical window until it is clean. Do NOT touch the lens with your fingers.

3.

Let the optical window and UV/IR reflectors air-dry fully.

Post by Doug Coulter » Fri May 14, 12:04 pm

I do use a sacrificial layer in front of my viewports. Sometimes it's just a sheet of mica stuck on with a tiny bit of apeizon wax, sometimes another full-thickness piece of pyrex. I'd rather have to toss a ten buck piece of that than replace the window in my viewport/door at vacuum supplier prices -- they use odd sizes that are hard to find elsewhere.

More often than not, plain old HCL will do the trick on many things. For ceramic, I follow with a rinse in something basic (ammonia or NaOH solution) then a final set of rinses with Di water/alcohol in an ultrasonic cleaner, but for the ceramics, the acid doesn't always do. I use both tungsten and titanium grids here with sometimes some exposed Cu or Al for the HV stalk. The shadow method is good -- prevention is worth more than cure to the extent it works.

Particularly if the stuff doing the shadowing is disposable or arranged in such a way that it doesn't hurt things if it becomes conductive. I have had mica fail horribly in this case because once it got somewhat conductive, it attracted enough ions to heat up and thermally decompose. Gives a pretty light show while it dies and takes the fusor down -- you'll be getting inside again to fix that before you can run again. Mica just doesn't work well in that place, but seems fine in front of a viewport.

Sandblasting works well on ceramics, especially with the little gun like Steven pictured. It's really cheap, the whole kit is on the order of 10 bucks at Harbor freight and some very cheap compressors meant for airbrush painting will do in this case -- you don't need 120 psi shop air or even a portable tank you fill at the gas station, and the media that comes with the kit works, as does other sandblasting media. You may want to avoid silicon carbide though -- conductive. Since I do have a shop compressor, I never considered that as a problem though, and they are getting cheap for the crummy and noisy ones anyway. If you do sandblasting -- do it outdoors, as it can really make a mess in the shop...you will have sand down your back, and regular safety glasses may not keep it out of your eyes, either, as the stuff bounces around everywhere. I use a cheap hood when doing that.

I too have had interesting problems with browned glass (but interestingly not with sapphire, which is getting hit enough to glow a pretty blue during runs). Luckily it was the 3/8" thick sacrificial piece of pyrex from McMaster -- cheap, and they sell it in round pieces the right sizes. When nothing seemed to work to remove it, I broke it to see how deep the brown went and it was virtually all the way through the glass (which had no doubt been placed in there either way facing in at one time or another), so I am not sure which theory is correct there -- X rays could be part of it too, as well as things driven deeply into the glass, but I've not run energies here that wold drive things that deep. I bet someone here who does ion implant could chime in with a number for expected penetration depth vs the few tens of Kv we tend to run -- making the assumption that any ions actually reach that going *away* from the grid (reaction products might, and they can be plenty hot). To me it would seem more reasonable that it is electrons because of other things we see -- as in the glass lighting up white under their bombardment and a faraday probe saying that it's electrons coming that way (or negative ions?).

Warm ammonium bifluoride solution works pretty well on things other than glass....and if you are desperate, even glass, which you follow with cerium oxide and elbow grease to un-frost it again. This is pretty tame stuff compared to real HF. I happened to have a quantity of it provided by my electroplating supplier (Caswell Plating) as a pickle/cleaning solution mix.

One thing that may lend some credence to the carbon theory is that I had a lot more trouble when the fusor was pretty new, and presumably had more contamination in it. However, for the last fairly long time, I've been using high grade graphite (McMaster) for endcaps on my cylindrical grids to hold the tungsten rods and not noticed any big increase in deposits, even though I can see some degradation on the graphite. Basically it started off with a mirror polish, and got more like flat black over time, a somewhat roughened surface. Nothing else seemed specially affected at least so far. The McMaster graphite rod stock is really good stuff -- not carbon, but really graphite, no binders to outgas (or mess up lathe tools), and of course it easily takes incandescent heat in fusor conditions when you put in too much power.

My policy, which seems to be working out, is to prevent deposition on sensitive stuff as far as possible,
following the old saw, an ounce of prevention....is worth a lot of effort in a cure. I don't worry about the tank walls at all. If I sputter some Ti or W onto them, it only seems to help, though I'd avoid getting too much W on there as it will then make X rays more efficiently under electron bombardment than Ti or stainless steel elements. Not an issue here, we bit the bullet and totally shielded our fusor with thick lead sheet. A lot of work, but no worries now.

A lot can be solved by getting the geometry right so the things that get coated don't matter, but that takes significant skull sweat and a few tries, usually. My new HV feedthrough design accomplishes this pretty well. This:

If that fails, harbor freight sells some nifty small diamond hole saws that easily make clean holes in glass or quartz (not so good on alumina sheet) using the normal procedures -- slow drill press, water lube in a dam made of modeling clay, and patience - slow rotation and light pressure gets it done. You can put them over the HV stalk and they catch the stuff before it gets on the harder to clean ceramic, and no compunction about either putting them into nasty chemicals to clean or simply replace them.



I find it interesting that you can pretty much tell who really runs their gear by the problems they run into, and if they don't see this (or some other issues) -- they must not be running very much!

I am wondering where the carbon might be coming from, assuming Chris is correct, as I only generally see the usual small amounts of CO and CO2 on the mass spec. Could it be impurities in the normal CP grades of D that people are using? There is always a tiny bit in the glass already, but I'd think that it may not be the first thing to be reduced by electron bombardment. I do know that I am often making enough X rays to do the "color center" thing in glass and brown it that way. I am finding that really pure quartz (as obtained at quartz.com, not vycor but the good stuff) gets reduced *or* browned a lot less easily than most ceramics or glass, for whatever reason - I am finding a lot of good uses for that material around here. It is one heck of a lot trickier to do glass work on than pyrex, though. The score and snap technique does NOT work with quartz, and the melting point and boiling points are fairly close together - it takes good tools and some practice to work with even once you get good at doing the same things with pyrex, but drilled holes are easy in either case. To cut tubing I pretty much have to use a thin diamond wheel in a toolpost grinder on the lathe to get good results every time. As usual, use water lube and go slow and it works fine.

I have used, and I believe posted prior to this, all 3 main methods. I think anyone who actually runs a fusor for many hours runs into these things.I do use a sacrificial layer in front of my viewports. Sometimes it's just a sheet of mica stuck on with a tiny bit of apeizon wax, sometimes another full-thickness piece of pyrex. I'd rather have to toss a ten buck piece of that than replace the window in my viewport/door at vacuum supplier prices -- they use odd sizes that are hard to find elsewhere.More often than not, plain old HCL will do the trick on many things. For ceramic, I follow with a rinse in something basic (ammonia or NaOH solution) then a final set of rinses with Di water/alcohol in an ultrasonic cleaner, but for the ceramics, the acid doesn't always do. I use both tungsten and titanium grids here with sometimes some exposed Cu or Al for the HV stalk. The shadow method is good -- prevention is worth more than cure to the extent it works.Particularly if the stuff doing the shadowing is disposable or arranged in such a way that it doesn't hurt things if it becomes conductive. I have had mica fail horribly in this case because once it got somewhat conductive, it attracted enough ions to heat up and thermally decompose. Gives a pretty light show while it dies and takes the fusor down -- you'll be getting inside again to fix that before you can run again. Mica just doesn't work well in that place, but seems fine in front of a viewport.Sandblasting works well on ceramics, especially with the little gun like Steven pictured. It's really cheap, the whole kit is on the order of 10 bucks at Harbor freight and some very cheap compressors meant for airbrush painting will do in this case -- you don't need 120 psi shop air or even a portable tank you fill at the gas station, and the media that comes with the kit works, as does other sandblasting media. You may want to avoid silicon carbide though -- conductive. Since I do have a shop compressor, I never considered that as a problem though, and they are getting cheap for the crummy and noisy ones anyway. If you do sandblasting -- do it outdoors, as it can really make a mess in the shop...you will have sand down your back, and regular safety glasses may not keep it out of your eyes, either, as the stuff bounces around everywhere. I use a cheap hood when doing that.I too have had interesting problems with browned glass (but interestingly not with sapphire, which is getting hit enough to glow a pretty blue during runs). Luckily it was the 3/8" thick sacrificial piece of pyrex from McMaster -- cheap, and they sell it in round pieces the right sizes. When nothing seemed to work to remove it, I broke it to see how deep the brown went and it was virtually all the way through the glass (which had no doubt been placed in there either way facing in at one time or another), so I am not sure which theory is correct there -- X rays could be part of it too, as well as things driven deeply into the glass, but I've not run energies here that wold drive things that deep. I bet someone here who does ion implant could chime in with a number for expected penetration depth vs the few tens of Kv we tend to run -- making the assumption that any ions actually reach that going *away* from the grid (reaction products might, and they can be plenty hot). To me it would seem more reasonable that it is electrons because of other things we see -- as in the glass lighting up white under their bombardment and a faraday probe saying that it's electrons coming that way (or negative ions?).Warm ammonium bifluoride solution works pretty well on things other than glass....and if you are desperate, even glass, which you follow with cerium oxide and elbow grease to un-frost it again. This is pretty tame stuff compared to real HF. I happened to have a quantity of it provided by my electroplating supplier (Caswell Plating) as a pickle/cleaning solution mix.One thing that may lend some credence to the carbon theory is that I had a lot more trouble when the fusor was pretty new, and presumably had more contamination in it. However, for the last fairly long time, I've been using high grade graphite (McMaster) for endcaps on my cylindrical grids to hold the tungsten rods and not noticed any big increase in deposits, even though I can see some degradation on the graphite. Basically it started off with a mirror polish, and got more like flat black over time, a somewhat roughened surface. Nothing else seemed specially affected at least so far. The McMaster graphite rod stock is really good stuff -- not carbon, but really graphite, no binders to outgas (or mess up lathe tools), and of course it easily takes incandescent heat in fusor conditions when you put in too much power.My policy, which seems to be working out, is to prevent deposition on sensitive stuff as far as possible,following the old saw, an ounce of prevention....is worth a lot of effort in a cure. I don't worry about the tank walls at all. If I sputter some Ti or W onto them, it only seems to help, though I'd avoid getting too much W on there as it will then make X rays more efficiently under electron bombardment than Ti or stainless steel elements. Not an issue here, we bit the bullet and totally shielded our fusor with thick lead sheet. A lot of work, but no worries now.A lot can be solved by getting the geometry right so the things that get coated don't matter, but that takes significant skull sweat and a few tries, usually. My new HV feedthrough design accomplishes this pretty well. This: http://www.coultersmithing.com/AuxCP/FT.html describes that, and a couple samples of this have run hundreds of hours without any problems or noticeable degradation. The very tip of the quartz tube gets a little deposition on it, that's all, and it doesn't affect how well it works at all, it seems. Key are the dimensions near the end that sticks into the tank -- and the aluminum design pictured there is the best so far -- note the stalk is turned down thinner near the end where the action is -- that turns out to solve a lot of things. Small changes there make large differences.If that fails, harbor freight sells some nifty small diamond hole saws that easily make clean holes in glass or quartz (not so good on alumina sheet) using the normal procedures -- slow drill press, water lube in a dam made of modeling clay, and patience - slow rotation and light pressure gets it done. You can put them over the HV stalk and they catch the stuff before it gets on the harder to clean ceramic, and no compunction about either putting them into nasty chemicals to clean or simply replace them. www.quartz.com is where I get my quartz, good people to deal with, but quartz isn't cheap no matter where you get it -- I generally buy in bulk to get the discount. I sometimes resort to McMaster for small round quartz windows even though they tend to be a little pricey there, still not too bad when you only need one or two small ones.I find it interesting that you can pretty much tell who really runs their gear by the problems they run into, and if they don't see this (or some other issues) -- they must not be running very much!I am wondering where the carbon might be coming from, assuming Chris is correct, as I only generally see the usual small amounts of CO and CO2 on the mass spec. Could it be impurities in the normal CP grades of D that people are using? There is always a tiny bit in the glass already, but I'd think that it may not be the first thing to be reduced by electron bombardment. I do know that I am often making enough X rays to do the "color center" thing in glass and brown it that way. I am finding that really pure quartz (as obtained at quartz.com, not vycor but the good stuff) gets reduced *or* browned a lot less easily than most ceramics or glass, for whatever reason - I am finding a lot of good uses for that material around here. It is one heck of a lot trickier to do glass work on than pyrex, though. The score and snap technique does NOT work with quartz, and the melting point and boiling points are fairly close together - it takes good tools and some practice to work with even once you get good at doing the same things with pyrex, but drilled holes are easy in either case. To cut tubing I pretty much have to use a thin diamond wheel in a toolpost grinder on the lathe to get good results every time. As usual, use water lube and go slow and it works fine.

For more information, please visit Optical Glass Window.