The grating is the core component of the grating interferometer, serving as the reference for displacement measurements. The accuracy of the grating period is a crucial factor influencing the measurement precision of the interferometer. Gratings fabricated using atomic lithography deposition technology can be traced directly to the transition frequency of chromium atoms, providing exceptional accuracy and consistency without the need for additional calibration. However, the limited area of gratings produced by atomic lithography necessitates the use of stitched atomic lithography techniques to expand the grating area for extending the measurement range of the interferometer. Nevertheless, stitched atomic lithography gratings exhibit non-uniform peak-to-valley heights, particularly in the overlapping regions of two depositions, where significant differences in peak-to-valley heights compared to single-deposition areas occur. This non-uniform variation in peak-to-valley height can affect the diffraction performance of the grating. To address this, a simulation model of the stitched atomic lithography grating was established to analyze the impact of peak-to-valley height on the grating’s diffraction performance. Concurrently, a testing setup for assessing the diffraction performance of the grating was constructed for experimental evaluation.A simulation model of atomic lithography gratings with varying peak-to-valley heights was developed, as shown in (Fig.3). Based on Rigorous Coupled Wave Analysis (RCWA), diffraction efficiency variations in stitched gratings with different peak-to-valley heights were simulated under various incident light polarizations. Additionally, diffraction performance testing setups were separately constructed for gratings under TM and TE polarized light. By controlling the translation stage to move the grating, diffraction energy utilization efficiency variations across different grating regions were collected and recorded. Utilizing atomic force microscopy measurements of the peak-to-valley heights in the stitched grating as shown in (Fig.1(b)), a simulation analysis of diffraction efficiency variations was conducted, and a comparative analysis between the simulation and experimental results was performed.Based on rigorous coupled-wave analysis (RCWA), the effects of different peak-to-valley heights on the diffraction efficiency of atomic lithography gratings are shown in Figure 4. The results indicate that as the grating’s peak-to-valley height increases from 30 nm to 85 nm, diffraction efficiency changes accordingly: for TM-polarized incident light, diffraction efficiency rises from 0.3% to 1.9%, while for TE-polarized incident light, diffraction efficiency initially increases from 14.8%, stabilizing at a maximum of 36.8% when the height reaches approximately 80 nm. Further, the diffraction efficiency variation curve of the stitched grating—simulated based on actual peak-to-valley height data—was compared with the diffraction energy utilization efficiency obtained from experimental performance testing (Fig.6). The comparison reveals a high level of consistency in trend between the experimental and simulation curves, though the experimental diffraction energy utilization values are overall lower than the simulated diffraction efficiency. This discrepancy primarily arises from the limited area of the stitched atomic lithography grating, resulting in partial incident light not fully interacting with the effective grating structure. Additionally, the grating’s surface quality is reduced due to particle adsorption in air, deviating from the ideal grating profile used in simulations, which also contributes to the lower experimental results.A stitched atomic lithography grating model was developed based on Rigorous Coupled Wave Analysis (RCWA), enabling simulation analysis of the impact of peak-to-valley height variations in the overlapping regions on diffraction efficiency. A diffraction performance testing setup was also constructed to experimentally measure the diffraction energy utilization efficiency across different regions of the stitched grating. Results indicate a high consistency in trend between the experimental data and simulation results. This study provides critical reference data for the fabrication of large-area atomic lithography gratings and lays a solid foundation for enhancing grating application performance. Future work will further investigate the relationship between diffraction efficiency, peak-to-valley height, and full width at half maximum (FWHM) in grating stitching regions, offering more precise guidance for the optimized fabrication of stitched atomic lithography gratings.