Speaker
Description
In this study, we introduce cutting-edge neural network algorithms that precisely predict crystal unit cell parameters and contact planes from Grazing Incidence X-Ray Diffraction (GIXD) data. Our method processes a list of q-positions and delivers predictions of the unit cell with exceptional accuracy—better than 0.1 Angstrom in dimensional precision and sub-degree in angular measurements. It eliminates the necessity to measure the 'missing wedge' in a specular scan, simplifying the experimental setup. Our AI method facilitates the rapid, autonomous processing of complex GIXD patterns without user intervention. It enables a detailed evaluation of the analysis's sensitivity to missing or spurious peaks due to its ability to predict structures for a large number of GIXD patterns quickly. Indeed the analysis copes well with patterns that miss peaks or contain only few peaks in total. These advancements present a substantial improvement in efficiency and reliability for researchers utilizing intricate GIXD patterns in crystallographic surface science studies.
| Please select the related topic from the list below | Instrumentation and methods |
|---|
