Scholarlink

<p>The influence of control parameters was studied on silicon (Si) nanoparticle synthesis using tandem modulated induction thermal plasmas (Tandem-MITP) with time-controlled feeding of feedstock (TCFF) method to optimize the modulation conditions on the basis of machine learning technique. This novel method, developed by our group, creates a time-varying high-temperature thermofluid field that facilitates efficient nanoparticle synthesis, with numerous control parameters influencing the process. To optimize the synthesis conditions, a comprehensive numerical thermofluid model was developed to simulate thermal plasma fields, feedstock dynamics, and nanoparticle formation and transport. Using this model, we applied a machine learning-based sequential approximate optimization (SAO) method with a radial basis function (RBF) network to identify optimal modulation conditions for maximizing nanoparticle production with smaller particle sizes. The results demonstrate that higher modulation amplitudes induce greater fluctuations in the plasma temperature and gas flow fields, leading to an increased quantity of smaller Si nanoparticles. Results showed that larger modulation condition provides larger variation in temperature-gas flow field, which results in larger quantities of smaller nanoparticles.</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

Optimization for modulation conditions in nanoparticle synthesis using tandem modulated induction thermal plasmas with intermittent synchronized feeding by machine learning

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年3月6日

查看全文
  • Yasunori Tanaka,
  • Yurina Nagase,
  • Rio Okano,
  • Yusuke Nakano,
  • Tatsuo Ishijima,
  • Satoshi Kitayama,
  • Siori Sueyasu,
  • Shu Watanabe,
  • Keitaro Nakamura

摘要

The influence of control parameters was studied on silicon (Si) nanoparticle synthesis using tandem modulated induction thermal plasmas (Tandem-MITP) with time-controlled feeding of feedstock (TCFF) method to optimize the modulation conditions on the basis of machine learning technique. This novel method, developed by our group, creates a time-varying high-temperature thermofluid field that facilitates efficient nanoparticle synthesis, with numerous control parameters influencing the process. To optimize the synthesis conditions, a comprehensive numerical thermofluid model was developed to simulate thermal plasma fields, feedstock dynamics, and nanoparticle formation and transport. Using this model, we applied a machine learning-based sequential approximate optimization (SAO) method with a radial basis function (RBF) network to identify optimal modulation conditions for maximizing nanoparticle production with smaller particle sizes. The results demonstrate that higher modulation amplitudes induce greater fluctuations in the plasma temperature and gas flow fields, leading to an increased quantity of smaller Si nanoparticles. Results showed that larger modulation condition provides larger variation in temperature-gas flow field, which results in larger quantities of smaller nanoparticles.