Broadband coherent anti-Stokes Raman scattering for crystalline materials

Broadband coherent anti-Stokes Raman scattering (B-CARS) has emerged in recent years as a promising chemosensitive high-speed imaging technique. B-CARS allows for the detection of vibrational sample properties in analogy to spontaneous Raman spectroscopy, but also makes electronic sample environments accessible due to its resonant excitation mechanism. Nevertheless, this technique has only gained interest in the biomedical field so far, whereas CARS investigations on solid-state materials are rare and concentrate on layered, two-dimensional materials such as graphene and hexagonal boron nitride . In this work, we discuss the specific properties of this technique when applied to single-crystalline samples, with respect to signal generation, phase matching, and selection rules in the model systems lithium niobate and lithium tantalate. Via polarized B-CARS measurements and subsequent phase retrieval, we validate the predicted selection rules, unequivocally assign the phonons of the ${\mathrm{A}}_{1}$(TO), E(TO) and ${\mathrm{A}}_{1}$(LO) branches to the detected CARS peaks, and address differences in spontaneous Raman spectroscopy concerning peak frequencies and scattering efficiencies. We thus establish this technique for future investigations of solid-state materials, specifically in the field of ferroelectric single crystals.