Answers are grounded in this paper and cite their sources.
The development of efficient low-loss electro-optic and nonlinear components based on silicon or its related compounds, such as nitrides and oxides, is expected to dramatically enhance silicon photonics by eliminating the need for non-CMOS-compatible materials. While bulk Si is centrosymmetric and thus displays no second-order (χ(2)) effects, a body of experimental evidence accumulated in the last decade demonstrates that when a strain gradient is present, a significant χ(2) and Pockels coefficient can be observed. In this work we connect a strain-gradient-induced χ(2) with another strain-gradient-induced phenomenon, the flexoelectric effect. We show that even in the presence of an extremely strong strain gradient, the degree by which a nonpolar material like Si can be altered cannot possibly explain the order of magnitude of observed χ(2) phenomena. At the same time, in a polar material like SiN, each bond has a large nonlinear polarizability, so when the inversion symmetry is broken by a strain gradient, a small (few degrees) reorientation of bonds can engender χ(2) of the magnitude observed experimentally. It is our view therefore that the origin of the nonlinear and electro-optic effects in strained Si structures lies not in the Si itself but in the material providing the strain: the silicon nitride cladding.
No takes yet. Share an insight, caveat, or question.
Khurgin et al. (2015) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: