Approach: We want to pass (non-parallel) broad-range (400-1100nm) light through milk and collect ALL of it and concentrate it onto a slit of dimension 50umx1mm.
Question: Is it possible? Can we do it through a self designed optical guide? Any other way?
Why not an integration sphere (https://en.wikipedia.org/wiki/Integrating_sphere)
Dear Shubham,
I agree with Holger and I also see the obstacles that Sergei suggests. I will only add the following:
You could use a commercial integrating sphere (IS) , or depending on your span of wavelengths of light of interest, you even might construct a cheap one with modest efficiency. Web searches should reveal a few "build-it-yourself" plans for IS's.
If you still need to take a fraction of the output from the IS and focus it down to a narrow slit, then you can use standard optics to accomplish that from an IS exit port. You sacrifice light intensity by using the IS, but you gain a wide angle of acceptance for the forward scattered in your milky sample. Much will also depend on the path length of sample you intend to study - if you haven't the option of making that path length small then you start running out of photons quickly.
Best regards,
Mike Bender
Thanks Holger, Sergie and MIke. I am definitely considering integrating sphere and have searched about do-it-yourself works too. My next plan is :
1) Take a light bulb, cut its metal part so that we get transparent ~sphere.
2) Silver (mirror) coat the sphere from outside and paint it to protect silver.
3) Fill the bulb with milk. scratch the bulb to make light inlet and outlet.
4) Shine a LED through inlet and use the other scratch itself as slit for spectrometer.
Please give your input about this idea.
Please see: I'll be using 600-1100nm range in the final version.
Thanks.
Dear Shubham,
Good for you! Always nice to see somebody try an experiment.
If barium sulphate is available to you then you might also try:
http://www.triticeaecap.org/wp-content/uploads/2011/12/Barium_Sulfate.pdf
Even painting the inside of a simple metal paint can with such a modified paint would make a nice IS, I think. There will be some wavelengths, potentially, where you may have slightly stronger absorption than you would like, but the metal paint can approach makes a durable device.
If your sample has low scattering you should perhaps consider not having a direct path from the sample entrance port straight across to your outlet port. IS's often use "baffles" to avoid direct transmission of light. Shouldn't be a big problem with high scattering samples, I believe.
Be aware that your white light LED is not a broad, smooth distribution of wavelengths and intensities. Colors are mixed to produce the appearance of white to the human eye based on its sensitivity to wavelength.
https://www.maximintegrated.com/en/app-notes/index.mvp/id/3070
Good hunting!
Mike Bender
Wow that's something new. Thanks. I have a few questions now.
1) Given high turbidity of milk, would there be any difference between mirror coating and white painting?
2) I read a few literature and there seems to be some significance of spherical shape that it gives uniform radiance across the surface area. Wouldn't cylindrical shape of a can create some problem with MAXimum light collection at a point? Because I can easily take a transparent circular or rectangular pipe and go on with it.
Yes I'm aware of the emission spectrum of LED. In the final version I'll use many infrared LEDs different spectrum, to mock a continuous source like tungsten lamp. I'm not using tungsten lamp as it takes over 10min for its spectra to stabilize. LED too change spectrum with different temperatures, but that's a problem I choose to tackle later.
Dear Subham,
You are right --the sphere is highly preferred. But finding a right circular cylinder is sometimes easier and not an extremely terrible approximation. If you can find a good sphere, then certainly go with that.
Mike
Thank you Sir. Also, could you please comment on the metal coating vs white paint given highly turbid sample.
Thanks Dr. Sergie for the data. I tried googling the terms and guess light scattering length is the distance light travels before getting reflected/refracted off a particle?
What I personally feel is that shape of the reflective container doesn't matter in the case of turbid liquids, be it sphere, or cylindrical or any other shape. In order to get high output intensity we need to keep the volume of the container as low as possible. Is that what you mean by "light integration is compromised due to turbidity" ?
That's actually a good idea. Do you mean like in the attached photo? Though I'm not much confident if the lens would be able to focus a non-parallel light coming from turbid milk.
Thanks for the clarification.
Objective: To perform the spectroscopy of the unabsorbed light through milk sample, to determine the concentration of fat and protein in it.
Challenge: After passing light through milk, very low intensity is left to enter the slit of spectrometer, due to scattering. We want to increase the amount of this light.
Approach: We want to pass (non-parallel) broad-range (400-1100nm) light through milk and collect ALL of it and concentrate it onto a slit of dimension 50umx1mm.
Question: Is it possible? Can we do it through a self designed optical guide? Any other way?
Okay i'll do it.
I am working in the range 600-1100nm.
Recently, there have been a lot of papers describing the usability of the range 600-1100nm for fat detection. Bogomolov et al has in fact shown a very high accurate level of prediction using this range.
Article Building global models for fat and total protein content in ...
Finally achieved it! Made a cylindrical container of inner dia 5mm out of aluminum and made a tiny flowcell in it with 1mm pathlength. Used 1W LED bulb as source and got an awesome intensity of light after passing through milk at the slit (made by two razor slits).
To Learn:
Thanks everyone for their support. :)
- Badges
- Science topic
- Similar topics
- Physical Sciences
- Optics