possibly Raman scattering if it's a sharp peak unlike broad emission bands. Change the excitation wavelength to check if the so-called Raman peak position changes nearly proportionately. Usually the shift between excitation wavelength and Raman peak position lies within a fixed range for every solvent. It's usually weak and does not matter if the fluorophore you are going to measure in the solvent is highly fluorescing. You can try making the instrument slits smaller or you can also do an easy background subtraction.
possibly Raman scattering if it's a sharp peak unlike broad emission bands. Change the excitation wavelength to check if the so-called Raman peak position changes nearly proportionately. Usually the shift between excitation wavelength and Raman peak position lies within a fixed range for every solvent. It's usually weak and does not matter if the fluorophore you are going to measure in the solvent is highly fluorescing. You can try making the instrument slits smaller or you can also do an easy background subtraction.
Genrally when a molecule is excited at lower wavelength the molecule would have gone to higher exicted which wolud decayed to first singlet excited state in non radiative mode. In view of losing of some energy in the non radiative mode, the fluorescence emmision wave length shift to higher wavelenght. This shift is known as stokes shift .
Raman scattering is just an inelastic scattering of incident photons so it can appear in any solvent. It's not absorbing the photon. The best way to determine is changing the excitation wavelength. Whereas fluorescence appears for only certain excitation wavelength, Raman scattering will occur for any excitation wavelength always maintaining a definite Stokes shift (or atleast in a definite range).
Stokes and Anti-Stokes shift both can occur in a Raman interaction. Any fundamental spectroscopy book can give you a basic understanding of the different parameters.
It appears a Raman scattered peak as you told it is pure solvent(not impurity) and this scattering can occur in any solvent as Rupashree explained nicely. Unlike fluorescence spectroscopy in which peaks are generally Stoke shifted, both type of shifts are possible in a Raman effect, but at room temperature usually Anti-Stoke peaks are very feeble and these are the stoke shifted peaks, which are generally observable.
The wavenumber difference between photons at 283 nm and 314 nm is 3488 cm-1.
This does not appear to be Raman scattering on ethanol. Are you sure that your ethanol is pure? In fact there is no such thing as 100% ethanol! "Pure" ethanol always contains about 5% of water and/or benzene. I would suspect that your ethanol contains some water which exhibits Raman scattering approximately at this wavenumber. To verify whether this is indeed Raman scattering follow the advice of Rupashree Singh - change the excitation wavelength and see if the wavenumber difference of excitation and the Stokes-shifted peak remains constant. It should be like that for Raman scattering.