Generally, if your material can be dispersed well in deionized water, ethanol, chloroform etc, you can use any of these solvents.
For UV-Vis-NIR absorption spectroscopy, whatever solvent you are using, a baseline is first made with its reference. So, choice of solvent is in general not crucial, provided the material under observation does not encounter prior physical or chemical reaction with it; and quality of dispersion or solution is good.
Depends on the method of your synthesis mostly in the cases of green Synthesis, the blank solution is prepared by taking same amount (with deionized H2O)of extract used for synthesis of nanoparticles.