Characterization of polygrama green photopolymer for Compact Optoelectronic Integrated Neural (COIN) coprocessor applications

Abstract

The research described in this thesis is a portion of a larger project within the Photonic Systems Group at MIT to design Compact Optoelectronic Integrated Neural (COIN) co processor [13]. The choice of photopolymers is critical in determining the performance of COIN processors as we look at ways to increase the diffraction efficiency. The focus of this research was to optically characterize Polygrama Green, a photopolymer that is sensitive to green light (514 nm). We were able to plot diffraction efficiency versus the exposure energy density for a series of gratings. We found the maximum diffraction efficiency to be that of the 678 mJ/cm2 grating with a value of 29.5%. We were able to fit the data to a sin2(x) curve with a X2- value of 20.79. We concluded that this somewhat high X2-value is due to our low number of data points. However, using Kogelnik's equation and the measured diffraction efficiency of each grating, we were also able to calculate the An, of each grating. This analysis shows that Polygrama Green seems to be a promising candidate for the photopolymer used in subsequent optoelectronic neural network applications.