The new generation of non-linear optical devices, based on periodically-poled lithium niobate thin film platforms, aims at achieving record-high conversion efficiencies and a formerly unknown capability for integration into modern quantum optical systems. It thus not only suits the demands of high-end telecommunication applications, but also provides striking potential for further performance enhancement. However, most of these record-setting new developments lack in in-depth analysis of their domain grid structures to determine whether the key requirements for efficient non-linear optical conversion are ideally fulfilled. Established analysis techniques for the exploration of bulk lithium niobate, like Raman spectroscopy, have not been adapted for these thin film structures, because the low interaction volumes require long acquisition times. In this work, the importance of Raman spectroscopy analysis for thin film lithium niobate device optimization is demonstrated in cross-correlated co-imaging with second-harmonic imaging and piezoresponse force microscopy. Key performance indicators of quasi-phase matching are identified and specifically investigated. The experiments show that Raman spectroscopy is capable of detecting all these indicators and is unique in its ability to identify performance-inhibiting mechanical and electrical stress fields. In an attempt to establish high-speed Raman imaging on lithium niobate structures, the coherent four-wave mixing method of broadband coherent anti-Stokes Raman scattering (B-CARS) is, for the first time, systematically introduced for fundamental solid state analysis by theoretically and experimentally addressing all special implications of crystalline material systems via the model material lithium niobate, e.g. phase matching conditions, phase retrieval, and complex selection rules. It is shown that the resonant B-CARS signal can be retrieved in post-processing and allows a direct comparison with spontaneous Raman spectroscopy. The predicted CARS selection rules are experimentally confirmed, and phonons are assigned to their respective B-CARS peaks. Furthermore, B-CARS signals are shown to be predominantly generated by scattering in forward direction. The insights of these fundamental investigations are applied for lithium niobate domain wall imaging via B-CARS. Hyperspectral spontaneous Raman spectroscopy and B-CARS images are compared with regard to imaging speed, signal-to-noise ratio, domain wall contrast mechanism, and signature width. The experiments prove that B-CARS allows at least a 20-fold speed increase with an improved signal-to-noise ratio as compared to spontaneous Raman spectroscopy. The domain wall signature is of similar nature for both techniques, and is not changed via phase retrieval, thus allowing high-speed B-CARS domain wall imaging in lithium niobate without post-processing. The massive B-CARS domain wall signal is attributed to a Čerenkov-like effect analogous to second-harmonic imaging. This is the first time that such an effect has been shown by a Raman scattering technique. These findings show the importance of Raman scattering investigations for the optimization of modern non-linear optical devices, and outline a way to massively increase the speed of Raman crystal analysis via B-CARS. Thus, with further studies that quantify the effects detailed qualitatively in this work and take B-CARS analyses to a broader range of crystalline samples, this work can form the basis towards establishing the high-speed and in-depth analysis of modern non-linear optical platforms via coherent Raman imaging.:1. Motivation .......... 1 2. Fundamentals .......... 4 2.1 Lithium Niobate .......... 4 2.1.1 Ferroelectric Structures .......... 4 2.1.2 Domain Engineering .......... 6 2.1.3 Thin Film Lithium Niobate (TFLN) .......... 8 2.2 Spontaneous Raman Spectroscopy .......... 10 2.2.1 Raman Spectroscopy of Lithium Niobate Crystals .......... 14 2.2.2 Domain Imaging with Spontaneous Raman Spectroscopy .......... 18 2.3 Coherent anti-Stokes Raman Scattering (CARS) .......... 20 2.3.1 Signal Generation .......... 21 2.3.2 Broadband Coherent anti-Stokes Raman Scattering (B-CARS) .......... 24 2.3.3 The Non-Resonant Background .......... 25 2.3.4 Selection Rules .......... 27 2.3.5 Phase Matching .......... 30 3. Experimental Methods .......... 32 3.1 Spontaneous Raman Spectroscopy (SR) Setup .......... 32 3.2 B-CARS Setup .......... 34 3.3 Raman Data Analysis .......... 35 3.4 B-CARS Phase Retrieval .......... 36 3.5 Second-Harmonic Generation Microscopy .......... 38 3.6 Piezoresponse Force Microscopy .......... 40 4. Thin Film Lithium Niobate Co-Imaging .......... 42 4.1 TFLN Sample and Co-Imaging Approach .......... 45 4.2 Piezoresponse Force Microscopy Imaging .......... 47 4.3 Second-Harmonic Generation Imaging .......... 49 4.4 Hyperspectral Raman Imaging .......... 53 4.5 Conclusions on Cross-Correlated Co-Imaging .......... 58 5. Fundamental Aspects of B-CARS on Lithium Niobate .......... 61 5.1 Phase Matching .......... 62 5.2 Phase Retrieval .......... 65 5.3 Selection Rules and Phonon Assignment .......... 67 5.4 Conclusions .......... 71 6. B-CARS Domain Wall Analysis in Lithium Niobate .......... 73 6.1 Preliminary Considerations .......... 73 6.2 Sample Preparation .......... 74 6.3 Spectral Analysis .......... 75 6.4 Determination of Acquisition Times .......... 78 6.5 Domain Wall Signatures .......... 79 6.6 Domain Imaging .......... 82 6.7 Conclusions .......... 83 7. Summary and Outlook .......... 85 Appendix .......... I References .......... XIII Own Publications .......... XXIV Acknowledgements .......... XXV