The definition of absorbance as -lgT means that at absorbance 4 only 0.01 % of the light reaches the detector - very demanding...-> dilute your sample!

Thank you very much for your answer. I diluted the sample. Signals after 250 nm look very good. Below 250 nm, signals are still broken, even in places where absorbance was below 1. What could be the reason? Is this wavelength simply not suitable for analysis?

Is something wrong with the sample measurement procedure? I am preparing a solution of compound A in DMF. I have a double beam spectrophotometer. In stage 1, in the chamber for the standard and the sample I put cuvette in DMF and perform baseline correction. In the 2nd stage in the sample chamber I put a solution of compound And in DMF and I do spectrum measurement. Is this procedure correct? The sample is very poorly soluble, dissolves only in DMF.

What kind of cuvette are you using?

You need to measure the single-beam intensity (energy) spectra to begin with.

1) Has everything been calibrated? Light sources can become unreliable over time. 2) Is the light source hooked up to fiber? Sometimes, that can be faulty near the tip closest to the sample.

I am using a standard optical glass cuvette.

Which kind of glass? The kind determines the lower wavelength limit...

Jadwiga Skóra - you should use a quartz cuvette below ~300nm.

Thanks for the answers.

In fact, until now I used the wrong cuvette made of optical glass (symbol G) instead quartz cuvette.

I made the spectrum again using the cuvette with the symbol Q (quartz cuvette for measurements at 190-2500 nm) and I diluted the sample. However, the problem still exists. It's probably not the fault of the wrong cuvette. Any ideas what else can I check?

Looks much better. Your cut off in now probably your solvent.