for me this seems anything but the right way. I wonder, how you obtained what you refer as absorption? For me it seems that you show Absorbance.
I suspect the negative values you obtained are due to the reflectance values higher than 1 (or 100 %). However, in contrast with my statement above, what you show is not the Absorbance, but Optical density. Your spectrometer software just made a negative logarithm of values it measured.
Thanks for your answer.@ Tomáš Halenkovič sorry for my ambiguous explanation in the question.
As you said that the reflectance value is higher than 100%, so when I converted it to Abs, the value became negative. however, it is this phenomena that confuses me, how could the absorbance be negative? it looks abnormal to me.
And I wonder if it is because of my wrong handling of sample or machine problem?
values of reflectance higher than 100 % means, that your sample reflects more than your reference. There can be other reasons, but I don't know your spectrometer and your measurement conditions.
Thanks for all the answers, I have solved the problems. Here I want to share my experience, it may be helpful to someone who faces similar issue. I borrowed the reference sample from my friend at first time, but the surface was not very clean and flat. Then I made a new reference sample again, just simply pressed the BaSO4 into a holder(I used the holder we check for XRD...), everything works fine after that. I also compared the data with different sampling method (dropping onto Si wafer vs scattering onto BaSO4 reference), seems doesn't really matter in my case. So make sure the reference surface CLEAN AND FLAT.