Abstract: Biodegradable Zn-based alloys have gained increasing attention as bone implant materials due to their moderate degradation rates, bone-like mechanical properties, and excellent biocompatibility. Selective laser melting (SLM) has emerged as a promising technique for producing customized metallic bone im-plants, offering high-quality prints and precise geometric control. However, process optimization for SLM Zn alloys, which have only recently been developed, typically relies on trial and error. In this study, we applied machine learning to optimize the SLM parameters for a Zn-2Cu alloy for the first time. A su-pervised Gaussian Process Regression (GPR) method was used to predict the optimal high-density pro-cess window. Notably, a rarely utilized combination of high-power and low-speed (HPLS) parameters was identified and experimentally verified. The microstructures, mechanical properties, degradation perfor-mance, biological properties, and antibacterial properties of Zn-2Cu specimens fabricated using three representative SLM parameter sets were systematically compared. The SLM Zn-2Cu alloy exhibited re-fined Zn grains and randomly distributed ε-CuZn5 phases. Among the parameter sets, the HPLS group demonstrated the best mechanical properties, with an ultimate tensile strength of 119.00 ±1.73 MPa, a tensile elongation of 3 %, and an ultimate compressive strength of 681.39 ±7.41 MPa. The degrada-tion rate of the SLM Zn-2Cu alloy remained moderate at approximately 0.16 mm/year, with no significant differences between parameter sets. Additionally, 10 % and 20 % diluted extracts of SLM Zn-2Cu speci-mens exhibited favorable biocompatibility and alkaline phosphatase (ALP) activity in vitro using MC-3T3 cells. Furthermore, the SLM Zn-2Cu demonstrated superior antibacterial properties compared to SLM Zn. This study highlights the potential of additively manufactured Zn-2Cu alloys as promising bone implant materials and illustrates how machine learning can enhance the process optimization of SLM Zn-based alloys.

Key words: Selective laser melting, Machine learning, Zn-2Cu alloy, Mechanical properties, Degradation, Biocompatibility