<p>Single-crystal aluminum nitride (AlN), which is distinguished by its wide bandgap and outstanding thermal conductivity, is considered a promising option for optoelectronic uses among third-generation semiconductor materials. However, the material's inherent high hardness and notable chemical inertness bring about substantial hurdles in respect of machining efficiency and surface finish quality. In this article, single-crystal AlN was processed by chemical mechanical polishing utilizing a UV photocatalytic reaction. The effect of process parameters, comprising H₂O₂ concentration, TiO₂ concentration, polishing pressure, and polishing pad speed, as well as their interactive effects on polishing performance, was systematically investigated using Box-Behnken response surface methodology. Regression prediction models for material removal rate (MRR) and surface roughness (Ra) are constructed respectively, and the process parameters are optimised on this basis. Analysis of the results shows that H₂O₂ concentration, TiO₂ concentration, polishing pressure, polishing pad speed, and the interplay between H₂O₂ and TiO₂ concentrations exerted a highly significant impact on the MRR (p &lt; 0.01). For the surface roughness index, in addition to the four process parameters, the interaction between polishing pad speed and polishing pressure also had a highly significant influence, while the interaction between H₂O₂ and TiO₂ concentrations had a significant influence (p &lt; 0.05). The models of the regression equations constructed all have a prediction error of less than 5%. Under optimal process conditions: H₂O₂ concentration 8.74&#xa0;wt.%, TiO₂ concentration 4.67&#xa0;g/L, polishing pressure 0.0446&#xa0;Mpa and polishing pad speed 40&#xa0;rpm, material removal rate 199.7&#xa0;nm/h, surface roughness reduced to Ra 0.249&#xa0;nm. The study establishes a theoretical foundation and provides process guidance for applying UV photocatalytic reactions to wide bandgap semiconductor polishing.</p>