We will present the proof of concept for a next- generation Shashlyk-type calorimeter based on the use of the novel opaque scintillation technique, implemented in this scenario using scintillating powder. At present, electromagnetic calorimeters based on the sampling technique generally provide good granularity for a limited cost but feature a rather poor photon energy resolution, compared to uniform crystals detectors, at energies below hundreds of GeV. For this reason, we investigated the possibility of using sub- millimetric-sized crystals for extremely fine sampling to improve the energy resolution. This work is inspired by the new LiquidO detection technology which, breaking the conventional paradigm of transparency, makes use of opaque scintillation. With a short scattering length, scintillation light is stochastically confined in the region of its creation point and can be detected with wavelength shifting (WLS) fibers. In our case, we foresee the use of high-Z scintillator powder, for good stopping power, immersed in a transparent liquid in order to increase the density and better match the refraction index that controls the optical path, thus playing, in this case, the role of the diffusion material in the LiquidO systems. In this communication, we will discuss the results of a Geant4 Monte-Carlo simulation performed to compare the energy resolution of a conventional lead-scintillator Shashlyk-type calorimeter and a powder one, with same geometry, composed of ZnWO4 grains immersed in a CH2I2 bath. Furthermore, we will present the results of the tests aimed at demonstrating the possibility to collect blue light in a small volume of white powder by means of WLS fibers, readout by SiPMs. The amplitude and the time properties of the signals collected with the WLS fibers have been measured as a function of the distance from the light injection, for different filling mixture in the reference volume.