Integrating epitaxial barium titanate on various substrates for silicon nitride photonics

The study of thin film Barium Titanate (BaTiO3) is critical for advancing the current state of photonic devices for communication technologies. Furthermore, realization of these devices can help provide efficient solutions in the fields of quantum and high-performance computing. BaTiO3 is a leading candidate in achieving high modulation speeds at low powers due to its exceedingly high linear electro-optic (EO) coefficient, also known as the Pockels coefficient. Due to the anisotropic nature of the EO effect, developing highly crystalline barium titanate with its c-axis in plane (and hence, its polarization vector) is crucial to utilize the highest value of the Pockels coefficient. The initial challenge is to promote growth of in-plane oriented BaTiO3 while minimizing the formation of out-of-plane oriented BaTiO3. Further challenges involve addressing temperature-based integration incompatibilities of BaTiO3 with current state of the art photonics utilizing silicon nitride waveguides.

Our research explores the epitaxial growth of BaTiO3 using pulsed laser deposition on three different substrates in parallel as alternative strategies for integration on Si3N4. One strategy is to integrate BaTiO3 (Tetragonal, a = b = 0.399 nm, c = 0.4035) [1] on MgO (Cubic, a = 0.421 nm) in order to apply in-plane tensile stress and promote in-plane oriented growth, with the additional dual purpose of MgO acting as a lower refractive index material for greater light confinement in the EO BaTiO3 layer. Another option is to integrate BaTiO3 on silicon using a 10 nm buffer layer of Strontium Titanate (SrTiO3). The SrTiO3 buffer layer induces an in-plane compressive stress which initially promotes out-of-plane oriented BaTiO3. However, by tuning oxygen pressure during growth, we have successfully promoted the relaxation and re-orientation of BaTiO3 from out-of-plane to in-plane. Finally, we attempt deposition of BaTiO3 on amorphous substrates using Ca2Nb3O10 nanosheets deposited via the Langmuir-Blodgett method as a template layer to promote epitaxial growth [2]. This is crucial for sidestepping temperature-based monolithic integration incompatibilities, as the BaTiO3 layer is grown directly on the waveguide or cladding material. Moreover, mode confinement optimization simulations have been performed as a function of various cladding and BaTiO3 film thicknesses, along with simulations aimed at extracting optimal embedded waveguide geometry.

Ongoing research involves drawing links between tetragonality of the BaTiO3 thin films and its ferroelectric properties. Subsequently, measurements regarding the electro-optic behaviour and related parameters such as the half wave voltage and propagation loss exhibited by BaTiO3 thin films on a silicon nitride platform will be performed, enabling steps towards high speed, low power EO modulators.