Talking about a Double Beam UV-Vis spectrophotometer:
I have always placed water in a cuvette to zero the instrument. (Question 1 Please tell if there are cases in wich it has to be done different)
I have always setted up the baseline correction, and to run the baseline, I place a cuvette with all except the analyte, for example if my analyte is benzene and my sample is orange juice with benzene, my baseline would be orange juice without benzene.
(Imagine I will add different concentrations of benzene to the orange juice and analyze them in the UV-Vis) I would run just orange juice without nitrobenzene as baseline, and i would put the same in the reference holder (the rear cell holder) and then analyze the different concentrations of benzene. Question 2. What is the purpose of placing a reference when I have already ran a baseline? are baseline and the reference always the same? On wich cases are not they?
From my personal experience, I have ran differente scan experiments just to see how the baseline correction works, and I basically noticed that the reference is what really matters. I ran a baseline with water and 5 ppm of an analyte, then I placed the same in the reference cell and in the sample cell, I just got a straight horizontal line at 0 abs, just as expected. Then a ran a scan of the same sample with the same baseline but in the reference cell I put water, and I got a characteristic peak even though the baseline and the sample was the same. Then I ran a scan with the same sample and reference (water with 5 ppm of analyte), and water as a baseline and a got a straight line at 0 abs . Then a ran a scan, with water as a baseline and as a reference, and a sample of water with ppm of analyte and I got a characteristic peak as expected. After these experiments I would say that the reference sample is what really matters.
I really have been struggling with this, I hope I made myself clear