Collimate Light from Mercury Arc Bulb for 5cm Circular Beam

[sbcbmx112]

I need a help collimating light from a mercury arc bulb

i need the light to be collimated into a circular beam with a diameter of 5cm
i know that i need to use a fused silica lens because the wavelength of the light i need is in the 320nm range
buying a collimated beam arc lamp housing is out of the question because they are too expensive
i haven't taken physics in a few years


first i need a basic walkthrough of how to collimate light
any tips, resources, places to buy fused silica lens, people to talk to would be greatly appreciated.

thanks

 

[mgb_phys]

How collimated do you need ?
You can't perfectly collimate an extended incoherent source.

If you just need roughly colimated any lens will do, it needs to be 5cm diameter any focal length will do although something like 200-300mm will be easiest (shorter than this will be sensitive to position, longer is expensive)

If you need better then you might have to reimage it to a pinhole first.

 

[sbcbmx112]

we want to use it for UV-initiated polymerization so as collimated as possible to ensure that the intensity is uniform over the whole area we are polymerizing.

what about a diffuser? could i diffuse the light and then collimate it?

 

[AJ Bentley]

we want to use it for UV-initiated polymerization so as collimated as possible to ensure that the intensity is uniform over the whole area we are polymerizing.

what about a diffuser? could i diffuse the light and then collimate it?

I would have thought that was the last thing you wanted to do!

The idea is to get the 'rays' all parallel or the wavefront perfectly flat - same thing.

Randomising the wavefront wouldn't be a good start. That's why starting with a pinhole was suggested - that at least gives you a spherical wavefront.

 

[sbcbmx112]

okay, so no diffuser. wouldn't a pinhole greatly reduce the intensity of the light? right now we use mercury arc bulbs that are around 100W-300W and if we could stick with the same size bulb that would be ideal

 

[AJ Bentley]

Yeah, I don't think a pinhole would be a good idea.

To be honest, I leapt in with a flip response because you mentioned collimation and diffusion in the same context.

I know nothing about the techniques you are looking at. It was Just a high-school physics comment.

I'm not sure that diffusion wouldn't be the way to go - but not as a prelude to collimation.

 

[Andy Resnick]

I would not try and put the Hg light into a spatial filter (pinhole)- you will likely destroy the pinhole in a matter of seconds.

You say you want a uniform spot of light, 5 cm in diameter, from a (short?) arc source.

You don't say how uniform (10%? 1%? 0.1%?), and you should realize that 10% uniformity is easy, while 1% is very hard. You also don't specify the bulb housing- is the bulb open to the environment? Is it contained within some enclosure (which often has a place for a condenser lens)?

What I would do is get a (fused silica) plano-convex lens with an f-number between 2 and 3 (the f-number is simply the focal length divided by the diameter). For example, if you get a 5 cm diameter lens, it must have a focal length of about 10 cm. Note, the amount of light you are capturing will depend on the diameter of the lens and how far the lens is from your source, so think about how much light you need. And the larger the lens, the more uniform (over the central 5cm area) your light will be. So maybe get a 10 cm diameter lens, although the larger the lens the more expensive it will be.

In any case, orient the lens with the flat side toward the bulb and separated by the focal length, turn it on, do some alignment (centering, tip/tilt, etc), and you're done.

I'm not sure how much a diffuser will help, but if you want to try, place the diffusing screen on the curved side of the lens.

You may need to filter out some of the visible or short-wave UV for your application, depending on the photoresist you are using.

 

[Marcot9]

This applies exactly to me as well :D I've been been having constant problems overexposing in the center... and I've been told again and again it's probably because my photoresist isn't level. Brainstorming I'm sure it's the light source...

You say you want a uniform spot of light, 5 cm in diameter, from a (short?) arc source.

You don't say how uniform (10%? 1%? 0.1%?), and you should realize that 10% uniformity is easy, while 1% is very hard. You also don't specify the bulb housing- is the bulb open to the environment? Is it contained within some enclosure (which often has a place for a condenser lens)?

I imagine for me 10% is alright... I was investigating the encasing. It has already some type of filter, but I don't think it's doing any more than cutting out short wavelengths. (If it helps UVP B-100YP lamp)..

What I would do is get a (fused silica) plano-convex lens with an f-number between 2 and 3 (the f-number is simply the focal length divided by the diameter). For example, if you get a 5 cm diameter lens, it must have a focal length of about 10 cm. Note, the amount of light you are capturing will depend on the diameter of the lens and how far the lens is from your source, so think about how much light you need. And the larger the lens, the more uniform (over the central 5cm area) your light will be. So maybe get a 10 cm diameter lens, although the larger the lens the more expensive it will be.

In any case, orient the lens with the flat side toward the bulb and separated by the focal length, turn it on, do some alignment (centering, tip/tilt, etc), and you're done.

I'm not sure how much a diffuser will help, but if you want to try, place the diffusing screen on the curved side of the lens.

You may need to filter out some of the visible or short-wave UV for your application, depending on the photoresist you are using.

So: I would place the lens (if I were using f/2.. 10cm focal length) 10cm from the plane of the bulb... with the filter in between?

The lens would then be held with some sort of mount?

Finally: fused silica is necessary for it's UV transparency?

I apologize if that's a lot to clarify.. but the biggest feat is convincing my PI to purchase the lens! Appears to run in the $70 area...

Thanks so so much!

 

[Andy Resnick]

I imagine for me 10% is alright... I was investigating the encasing. It has already some type of filter, but I don't think it's doing any more than cutting out short wavelengths. (If it helps UVP B-100YP lamp)..


So: I would place the lens (if I were using f/2.. 10cm focal length) 10cm from the plane of the bulb... with the filter in between?

The lens would then be held with some sort of mount?

Finally: fused silica is necessary for it's UV transparency?

I apologize if that's a lot to clarify.. but the biggest feat is convincing my PI to purchase the lens! Appears to run in the $70 area...

Thanks so so much!

It's tough to tell from the website [http://www.uvp.com/highintensitylamps.html] , but based on the illumination at the base of the illumination box (pic #4), I'd say you are correct- the irradiance appears to be highly non-uniform.

I think your first order of business is to figure out a way to (safely) see the UV light, so you can see what is going on- can it make a white cotton shirt glow? The bluish light may be good enough to use on its own.

You have a different problem than the OP- your bulb already has a lens attached- at least, it seems to be a lens and not just a protective cover. Are there any alignment controls? Maybe play with the bulb-lens distance to see if you can get a more uniform output. Otherwise, you have a medium-difficulty-level project on your hands, starting with figuring out what lens is already there. I would spend some time on the phone with UVP- get a tech person who is familiar with the product and the application.

But yes, fused silica is needed for UV applications- at least, that's the least expensive UV transmissive material commonly available.

Good luck- remember, you are going for 'good enough', not 'perfect'.

 

[cragar]

Fresnel lens

 

[palane]

As noted below, you need a uniform light source rather than a collimated one. You're hitting a fairly standard problem that comes up for anyone testing solar cells. A diffuser is exactly what you need. Ground glass diffusers are relatively cheap, though there is the problem that glass absorbs UV. Engineered diffusers are better and provide a more uniform light source. The very best option is an integrating sphere, but throughput is low and they're expensive.

ThorLabs is my usual source, but I didn't see any silica diffusers except as a custom option. Edmund Optics has both ground silica diffusers ($20-$100) or engineered silica diffusers ($350+). Knight Optical looked to be a good source--they're UK based.

This might be a bit late, but I happened by while searching for other information.

I need a help collimating light from a mercury arc bulb

i need the light to be collimated into a circular beam with a diameter of 5cm
i know that i need to use a fused silica lens because the wavelength of the light i need is in the 320nm range
buying a collimated beam arc lamp housing is out of the question because they are too expensive
i haven't taken physics in a few years


first i need a basic walkthrough of how to collimate light
any tips, resources, places to buy fused silica lens, people to talk to would be greatly appreciated.

thanks