Scholarlink

<p>Nanomaterials are increasingly investigated for dental restorations to improve mechanical reliability. Zinc oxide nanoparticles (ZnO NPs) are attractive for reinforcement and antibacterial potential, yet optimisation is challenging because nanoparticle properties depend on synthesis conditions and experimental datasets are often limited. This study aimed to identify the ZnO loading that yields optimal mechanical performance in a photocured dental resin using an integrated synthesis–testing–modelling workflow. ZnO NPs were synthesised by pulsed laser ablation in liquid (PLAL) using an Nd: YAG laser at three pulse energies (30, 70, and 150&#xa0;mJ) and characterised by SEM, XRD, FTIR, and UV–Vis spectroscopy. A selected ZnO batch was incorporated into a photocured dental resin at 0–0.7569 wt%. Mechanical behaviour was characterized by the maximum fracture load obtained from Universal Testing Machine (UTM) testing, as well as Vickers microhardness (HV) and compressive strength (mean ± SD, <i>N</i> = 3). Five regressors (SVR, random forest, decision tree, KNN, XGBoost) were compared using ZnO loading as the sole input under a hybrid-data screening framework (Latin Hypercube Sampling–based output perturbation), followed by experimental-only LOOCV (<i>n</i> = 7). The optimum loading (0.0381 wt%) increased fracture load, hardness, and compressive strength by ~ 43.8%, ~ 31.3%, and ~ 66.7%, respectively, versus the control; higher loadings reduced performance, consistent with dispersion limitations. Random forest showed the strongest overall screening performance. LOOCV provided a conservative estimate of generalisation on measured datapoints. The integrated PLAL composite workflow combined with data-driven modelling offers a proof-of-concept strategy to optimise ZnO reinforced dental composites and guide follow-up experiments for biomechanics-relevant masticatory load resistance under data scarcity.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

AI-Driven Biomechanics-Relevant Mechanical Characterization of ZnO Nanoparticles Synthesized by Pulsed Nd: YAG Laser for Dental Restoration Applications

  • 期刊论文

  • 发表日期：

    2026年3月21日

查看全文
  • Fatima M. Abdulwahab,
  • Ghaidaa A. Khalid,
  • Ali Al-Naji

摘要

Nanomaterials are increasingly investigated for dental restorations to improve mechanical reliability. Zinc oxide nanoparticles (ZnO NPs) are attractive for reinforcement and antibacterial potential, yet optimisation is challenging because nanoparticle properties depend on synthesis conditions and experimental datasets are often limited. This study aimed to identify the ZnO loading that yields optimal mechanical performance in a photocured dental resin using an integrated synthesis–testing–modelling workflow. ZnO NPs were synthesised by pulsed laser ablation in liquid (PLAL) using an Nd: YAG laser at three pulse energies (30, 70, and 150 mJ) and characterised by SEM, XRD, FTIR, and UV–Vis spectroscopy. A selected ZnO batch was incorporated into a photocured dental resin at 0–0.7569 wt%. Mechanical behaviour was characterized by the maximum fracture load obtained from Universal Testing Machine (UTM) testing, as well as Vickers microhardness (HV) and compressive strength (mean ± SD, N = 3). Five regressors (SVR, random forest, decision tree, KNN, XGBoost) were compared using ZnO loading as the sole input under a hybrid-data screening framework (Latin Hypercube Sampling–based output perturbation), followed by experimental-only LOOCV (n = 7). The optimum loading (0.0381 wt%) increased fracture load, hardness, and compressive strength by ~ 43.8%, ~ 31.3%, and ~ 66.7%, respectively, versus the control; higher loadings reduced performance, consistent with dispersion limitations. Random forest showed the strongest overall screening performance. LOOCV provided a conservative estimate of generalisation on measured datapoints. The integrated PLAL composite workflow combined with data-driven modelling offers a proof-of-concept strategy to optimise ZnO reinforced dental composites and guide follow-up experiments for biomechanics-relevant masticatory load resistance under data scarcity.