I have done deconvolution of the UV Vis spectra of SnS nanoparticles dispersed in water. The UV Vis spectra contains a continuous decreased graph from lower wavelength to higher wavelength. So does deconvolution help to interpret this graph or not?
How did you deconvolute (van Cittert mthod, other method ?), what Kernel did you use for deconvolution and wahat was the idea or aim behind doing this? I doubt that without a demonstration (graph) of what you did anyone shall be able to voice an opinion.
You might call this a decomposition (into a bunch of peaks of whatever shape), but technically/mathematically this is certainly not a deconvolution. Is there a reason to pick this particular set of resonance-like features?
IMHO the spectrum is not sufficiently structured to maintain it must be decomposed this way (anyway, i can't see what quantity is being measured. Dou you understand the processes by which the signal is generated?). I think you first need an idea / a rough model which tells you what kind of behavior you might expect. if it leads to a decomposition into a bunch of peaks as you did it, then the parameters of the peaks (position, width, area) specify the results within this model. You may then (try to) cross check whether these parameters make sense.
If you don't know what to expect, then i'd say the data you show give no clear evidence about how to attack a decomposition (and whether it is justified at all).