I need to measure the optical absorption of a thin film( ~500 nm). Attached are T% and R% spectra. I may use either Beer–Lambert law to calculate A% using R% data or Kubelka-Munk function using both R% and T% data.
However, the calculations give different results. The film is supposed to be TiOx material and the closest spectrum to the phase is show by the results of the calculation using Kubelka-Munk function. However, the film is semi-transparent and it is not clear which theory I should use.
At first glance, your issue is not with the approach you use, but how the two different measurements were done. The %T seems to be cut off
Hi, Dr. Hugh J Byrne,
No, the substrate was quartz and the baseline was measured using bare quartz with the same thickness.
I think you should look into the limitations of both the theories. Your film is not very thin, and if it's described as TiOx, then it is possible you have different phases, some with a very high, metal-like electronic conductivity and some dielectric. Maybe if you you could gather some structural data and check whether these laws can be applied to your materials, and in what range of the spectrum, this would cast some more light on the problem?
Dear Agnieszka Gorzkowska-Sobas, according to XRD and Raman, the film is Ti4O7.
So this is a Magnelli phase, if you had a sample of powder you'd see it's very dark, almost black colour, and you have to see into what are the possible absorption mechanisms in this phase. Did you measure total reflectance? I haven't been working with titanium oxides for quite some time, but I'm pretty sure you could find some reference spectra published. The task won't be trivial though.
Is your film rough?
Are you taking into account the losses due to diffuse scattering also in Beer Lambert's Law experiment?
If your film is rough then there will be diffuse scattering and diffuse scattering is not accounted well in your Beer Lamberts Law experiment. Also the losses due to diffuse scattering are not accounted in your baseline correction too since you are measuring the baseline using a blank substrate. There may not be any diffuse scattering when you have a blank substrate. Hence the values calculated using Beer-Lamberts Law will be different.
Some details are needed.
Are you using the same instrumental setup in both the experiments? If no, then are you using the same lamp source in both the experiments? Are you using the same detectors in both the experiments? If no, then are the sensitivities of both the detectors the same?
I believe you might be using the same sample in both the experiments.
Apologies - I realise I was reading the transmission curve the wrong way - As Agnieszka points oiut, this is in fact a highly coloured phase, and a maximum transmission of 3.7% means that this is not "semi-transparent". The absorbance is saturated in the low wavelength region and looks to be close to saturation from ~700-850nm. The film is too thick for the beer Lambert law to be applied with confidence
@ Mikhail
Have you plotted T or %T? The graph shows it as %T. The limits of both T and %T are different.
For T: 0
Agnieszka Gorzkowska-Sobas, the spectra may be found in doi:10.1016/0022-4596(72)90113-2 or DOI 10.1007/s10854-013-1487-5. My spectra do not look like that, unfortunately.
We use integrating sphere for the measurements.
Hi, V. N. Ravi Kishore V,
The film is ~ 500 nm. The spectra were obtained by diffuse reflectance
measurements. The setups are the same, using the same spectrometer and lamps.
For measuring reflectance, we use the WS-2 standard by Avantes(the technical details are attached). For measuring T% we use bare quartz as a baseline.
The maximum value for both graphs were 100%.
Hugh J Byrne, I should not have called it semi-transparent! Thanks for this note. Though, Abs spectra calculated using LB law look suspicious and remind the spectra of TiO2.
What are you plotting on the y axis in the first graph? The graph looks like an absorption profile. Is it A or % (I0-I)? Both the spectra are normalized to 1.0. Since the absorption maxima of both BL Law and KM law are at different wavelengths hence the differences appear more prominent. Try normalizing the spectra to 1.0 at the peak maxima on the red side.If you do that then there seems to be better agreement between both the theories (better spectral overlap) in the central portion and there are differences on the blue side and the red side.
I'm not sure about your measurement protocol..... for the "blank" (100%) baseline - do you have the WS-2 in both arms of the integrating sphere?
Replacing the sample WS-2 plate with your film should then give you a reflectance measurement.
Putting your sample at the entrance to the integrating sphere will give you a transmission measurement.
Dear Hugh J Byrne,
The optical scheme of our integr. sphere is attached.
What is the principal difference of measuring T% using integrating sphere and just a holder for the measurements?
We use KM function for calculating abs if the T% is negligible. However, what theory we should use if T% and R% are comparable?(attached spectra). We got thin samples (~100 nm) and it may be the reason of high T%. What theory should we use in this case?
- Badges
- Science topic
- Similar topics
- Physical Sciences
- Materials Science