Poor optical property extraction when using 1mm glass slides

[anishabahl]

Dear Sir/Madam/Mx,
I have found the IAD software to be invaluable to my PhD (in King's College London supervised by Prof Tom Vercauteren) and am very grateful for its availability and documentation! We have found that using the IAD version 3.11.4 on a phantom with known absorbance and reduced scattering coefficients works well if the phantom is measured directly, however it returns a largely overestimated absorbance coefficient if 1mm glass slides are used experimentally and specified in the IAD routine. Example data and a notebook demonstrating this and detailing the parameters used (MWE_phantom_IAD_glass_effect.ipynb) can be found here, and examples of the optical parameter outputs are attached. We would be very grateful for any advice for how to improve these results!
Many thanks,
Anisha Bahl

[scottprahl]

@anishabahl Not sure what is going on here. I went all the way back to v3.1.25 and still seem the same results.

Did you add some sort of index matching fluid between the slides and the phantom that you used?

What is very strange is that the scattering comes out perfectly.

Anyhow, I am still investigating, but nothing obvious has shown up yet.

[anishabahl]

@scottprahl Thank you for your response. We used a few drops of water to create uniform connection with the glass slides aiming to avoid air bubbles, however we did not use any specific index matching fluid. Is it likely that this is the problem? It is interesting that the scattering is recovered well regardless! Thank you very much for your help with this!

[ccampb19]

Hi all,
Just glancing at this, the discrepancy resembles the absorption one might expect from 2mm of typical microscope slide (soda lime glass), which bottoms out around 500-550 nm [1]. Could that be the issue? Do the slides have a greenish tint?

[1]https://rocoes.com.tw/2glass/sodalime2.html#notes

[scottprahl]

@anishabahl

@ccampb19 you were right! The discrepancy can be directly attributed to the absorption by soda lime glass. If I take the data for clear soda lime glass from Rubin, "Optical Properties of Soda Lime Silica Glasses" 1985 then I get the following graphs.

This is all done with v3.16.0 using iad -i 8 -X file.rxt

If we assume BK7 for the microscope slides then the absorption is still not consistent

The absorption of soda lime glass looks like

When the analysis is done with absorbing slides then we consistent results with a soda lime glass that absorbs 75% of the clear soda lime glass that Rubin measured. (clear = 100% and is too much absorption, 50% is too little)

Finally, we can compare the absorption with that from PDMS measurements by Cai 2008. We see that there is small constant offset of 0.017mm⁻¹ needed to make the measured spectrum match.

The .rxt files are here (remove the .txt)

phantom-with-75%-clear-soda-lime.rxt.txt

phantom-with-no-slides.rxt.txt

[scottprahl]

@tvercaut you also might be interested in the resolution of this issue.

[tvercaut]

Brilliant, thanks! We will come back to this with @anishabahl and @ssegaud