PeakFit v.12 is an excellent software for baseline correction. You can download a 30 day trial. In the simplest case, you can do the same in Excel, by subtracting baseline column from the chromatographic signal column.

Note that the term deconvolution has a very different mathematical meaning: https://terpconnect.umd.edu/~toh/spectrum/Deconvolution.html

Thank you very much for the input. I will have a look :)

If you want to subtract the baseline, you need to make sure that the system have reached equilibrium already. Minor retention shifting or baseline shifting can occur from run to run. Also, its peak area might not be the same. I used Excel to subtract baseline before, which works well. Besides, if you need to calculate baseline noise manually, then you should use the original data. The subtracted one will increase its noise at 1.4 times. The best way is to find the reason and to get rid of this blank peak. 'inject nothing' means inject 'mobile phase' or ' a blank sample' or 'no action of your injector'?      

I mean blank sample just methanol for example

If you blank sample is not the mobile phase, it can happen. I saw the same thing before. You can check the specification of your methanol. They are some impurities in your reagent. If you really want to confirm it, you can compare the UV-vis spectrum of that peak with your blank sample with a PDA detector. Since it is in the end of the chromatogram,  you should not be bothered.

You can subtract one chromatogram from another in Xcalibur. Provided you run the sample and the blank under identical conditions, the system is equilibrated properly, and you are getting stable retention times, then you get a reasonable subtraction (at least for LC-MS chromatograms this works). There are some things in the subtraction that don't seem to work as they are supposed to, but generally I get OK results (I have been subtracting chromatograms before and after derivatization with thiols to identify Michael-acceptors in my extracts). But as others have commented, why bother artificially removing this peak if it is at the end of the chromatogram?

@Christopher: Can you please tell me where to find the option for subtracting 2 LC/UV chromatograms as you mentioned above?

I would like to remove the peak to estimate purity

In Xcalibur, choose tools then background subtract. A dialogue bxx opens where you can choose which file to subtract from which, and also you can scale the amount of subtraction, choose which detector files are subtracted (e.g. UV, MS, or both), and you are supposed to be able to move the subtracted file forward or back in retention time during the subtraction. The latter is to help in case you have retention time drift, I guess, but I recall finding that it only actually worked in one direction and not the other when I tried using this feature.

Thank you very much all. I will check it out and let you know the results