We want to measure CIELab values in liquid samples using the reflectance spectrophotometer, and we are not sure about how we could do it having reliable results. Have any of you tried using the Konika-Minolta CM-600d with a CM-A514 Sample Holder and a specific cuvette? Any other options?
Can you say more about the liquid you are measuring? Is it opaque, translucent, or transparent?
Kind of transparent/translucent. We could play with the concentration and make it more transparent
Transparency makes it very difficult for a reflectance measurement device either for an integrating sphere or 0/45 geometry. In a reflectance device, light has to reflect from something in order to get measured. If the sample is transparent, then the only thing that comes back will be light that is reflected from behind the sample. So, the reflectance of the backing material is crucial. The best results will be achieved with something white and matte.
Integrating sphere devices (such as your KM 600d) work best when the reflective surface is right at the plane of the sample port. So, you want the cuvette to be thin with respect to the size of the sample port. Of course, that means that the absorbance of the sample will be small, so you need to have a high concentration.
I think a better option would be to use a transmissive measurement device. In this device, the illumination is on one side of the cuvette, and the detector is on the other, measuring the amount of light that passes through the sample.
I am not as familiar with transmissive spectros, but I see that Konika-Minolta has two such instruments, the CM-5 and the CR-5.
Thanks John,
One of my colleagues has also suggested the white tile in the bottom and in a dark room as a potential way to do it. We will give it a try.
We will also measure transmittance in the very same samples, but we are not familiar with converting the spectra measurement into CIELab values. Any suggestion for this?
I had a thought last night... Instead of using a white tile, you could paint the outside of the cuvette with white paint. This will improve several things:
1. Any optical transition gets in the way of measuring your sample. In particular, the transition from the glass or plastic of the cuvette to an air gap to the white tile and back is a problem. There is a reflection back into the cuvette as light goes from the glass to the air gap. This could be appreciable - maybe 40% to 60%. There is also a reflection as light tries to re-enter the glass after reflecting from the tile - maybe 4%. Since the paint would be optically connected to the cuvette, these optical transitions (and the reflections caused by them) are eliminated..
2. The cuvette can be painted on all surfaces (except the aperture) so that ambient light is much less of an issue. This will likely require multiple coats of paint.to achieve opacity. I recommend shining a flashlight into the cuvette to test for opacity.
3. Painting the cuvette all the way around makes the chamber similar to an integrating sphere, although the shape is not quite spherical. :) This has the effect of capturing a greater portion of the light. On the downside, the light will travel through your liquid multiple times, so that it makes the results more difficult to model mathematically. You go from possibly being able to use Beer-Lambert-Bouger as a model to needing the Saunderson correction, and perhaps the Kubelka-Munk equation.
(If you would like more detail on the modelling, please contact me at [email protected].)
4. Painting the outside of the cuvette is far more practical than painting the inside, since you avoid contamination of the paint into the liquid and the liquid into the paint. From an optical standpoint, the two are very similar.
I am assume your question about computing CIELAB values in transmission mode is about the step of normalizing the XYZ values (dividing by Xn, Yn, Zn). I would recommend capturing a measurement with clear liquid in the cuvette, and using this as the white point.(This approach also applies for measuring with a reflectance device.)
Normally for reflectance measurement, an illumination spectrum is applied. For transmission, the reasonable illumination is 100% at all wavelengths.
I hope this answers the question you were asking? Or did I misunderstand?
When you ask about converting to CIELAB, I am left trying to understand why you would want to. CIELAB is useful because it predicts (to some extent) what a sample looks like to a human observer. Why is that useful? Does anyone really care what the liquid looks like in a cuvette? It seems to me that a spectrum could be more appropriate, but of course, I don't understand your application.
Thanks a lot for you advice John, we are now trying to set up a system in which we minimize the air gap and the influence of the ambient light.
On top of the transmittance spectra, it is important to us to know how the samples would be perceived by human sight, since it is a product that will eventually be transformed into an ingredient.
We are still trying to find ways to transform the transmittance full visible spectra into CIELab values, but apparently this is not that common in the literature. However, some transmittance spectrophotometers do offer the possibility of such values (RGB or Lab), so that I guess such a conversion is possible. Have you heard about that?
Thanks
The conversion from spectra to CIELAB can easily be done. I can provide a spreadsheet, if you need it.
Doing the conversion in a way that makes sense for your application is a bit harder, since the measurement that a transmittance device includes only a single pass through the liquid, and viewing would generally involve light passing through twice. Also, the viewing color will include the color of whatever is behind the liquid.
Am I understanding the question?
Sure you have,
I understand the drawbacks, but it will probably be enough for our purposes. At least, we would very much like to give it a try and compare to the reflectance values with the setup we are trying to build. So, if it is not too much bothering, it would be nice to have the spreadsheet.
Thank you so much John
- Badges
- Science topic
- Similar topics
- Geoscience
- Atmospheric Sciences
- Atmosphere
- Air