Scholarlink

<p>Mg<sub>x</sub>Zn<sub>1-x</sub>O (x = 20% and 1-x = 80%) TFT-based UV photodetector was simulated using TCAD tool. The performance of simulated UV photodetector was calculated using spectral response measurements. Responsivity of photodetectors was 0.28 A/W, 0.30 A/W, 0.32 A/W, and .32 A/W with corresponding External quantum efficiency of 351%, 140%,109%, and 109%. respectively. The changes in detectivity and noise equivalent power are also calculated.&#xa0;These TFT-based UV photodetector results show excellent responsivity, photosensitivity making them compatible for optoelectronic devices. Also, this UV photodetector parameter prediction is done using Linear Regression and Neural Network machine leaning model. The parameters R<sup>2</sup> and RMSE for linear regression model are .85 and .144 and for Neural Network model .96 and .030 calculated. Observed results show that Neural Network model gives better results as compared to linear regression model.</p> Graphical abstract <p></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

Machine learning on simulation study of UV photodetector parameter based on MgZnO thin-film transistor

  • 期刊论文

  • 发表日期：

    2025年5月13日

查看全文
  • Manju Choudhary,
  • Lava Bhargava,
  • C. Periasamy

摘要

MgxZn1-xO (x = 20% and 1-x = 80%) TFT-based UV photodetector was simulated using TCAD tool. The performance of simulated UV photodetector was calculated using spectral response measurements. Responsivity of photodetectors was 0.28 A/W, 0.30 A/W, 0.32 A/W, and .32 A/W with corresponding External quantum efficiency of 351%, 140%,109%, and 109%. respectively. The changes in detectivity and noise equivalent power are also calculated. These TFT-based UV photodetector results show excellent responsivity, photosensitivity making them compatible for optoelectronic devices. Also, this UV photodetector parameter prediction is done using Linear Regression and Neural Network machine leaning model. The parameters R2 and RMSE for linear regression model are .85 and .144 and for Neural Network model .96 and .030 calculated. Observed results show that Neural Network model gives better results as compared to linear regression model.

Graphical abstract