Characterization of energy filtering slit widths for MicroED data collection

A favorable signal-to-noise ratio is essential for obtaining high-quality diffraction data in macromolecular electron crystallography. Inelastic scattering contributes significantly to the noise, reducing contrast between diffraction peaks and background, which complicates peak detection and compromises the accuracy of intensity integration. Energy filtering mitigates these challenges and enhances diffraction data quality by removing inelastically scattered electrons, leading to reduced background noise and sharper Bragg peaks. Previously, we reported a substantial improvement in MicroED data quality and resolution with energy filtering. Here, we systematically evaluate the impact of different energy filter slit widths for optimal MicroED data collection. Data from proteinase K lamellae were collected using 5, 10, and 20 eV energy filter slits. Our results show that the optimized energy filtering reduces background counts and improves the signal-to-noise ratio, enhancing the precision of intensity measurements and resulting in improved structural models. These findings provide insight into the optimization of energy filter slit settings that, when paired with direct electron detection, enhance data collection strategies by improving the signal-to-noise ratio, supporting higher quality data, and ultimately enabling more precise structure determination.