大学化学 >> 2026, Vol. 41 >> Issue (5): 36-49.doi: 10.12461/PKU.DXHX202511137

所属专题： 第5届全国大学生化学实验创新设计大赛

专题 上一篇    

铁配合物合成与性质探究的数智化改进——大学化学实验中“磺基水杨酸合铁配合物的组成与稳定常数的测定”的创新设计

王思寒1, 余晨曦1, 冉术彰1, 陈家伟1, 饶守佟1, 梁昕旖1, 董睿祺2, 曾桂香2, 王国强1, 马晶1   

  1. 1 南京大学化学化工学院, 江苏 南京 210023;
    2 南京大学匡亚明学院, 江苏 南京 210023
  • 收稿日期:2025-11-20 录用日期:2026-02-25 发布日期:2026-05-21
  • 通讯作者: 曾桂香, 王国强, 马晶 E-mail:gxzeng@nju.edu.cn;wangguoqiang710@nju.edu.cn;majing@nju.edu.cn Guixiang Zeng, Guoqiang Wang, Jing Ma
  • 基金资助:
    教育部基础学科和交叉学科突破计划(JYB2025XDXM309)；南京大学本科教育教学改革课题(0205-145011)

Digital and intelligent improvement of the synthesis and property investigation of iron complexes: an innovative approach to determining the composition and stability constant of the sulfosalicylic acid-iron complex in undergraduate chemistry experiment

Sihan Wang1, Chenxi Yu1, Shuzhang Ran1, Jiawei Chen1, Shoutong Rao1, Xinyi Liang1, Ruiqi Dong2, Guixiang Zeng2, Guoqiang Wang1, Jing Ma1   

  1. 1 School of Chemistry and Chemical Engineering, Nanjing University, Nanjing 210023, Jiangsu Province, China;
    2 Kuang Yaming Honors School, Nanjing University, Nanjing 210023, Jiangsu Province, China
  • Received:2025-11-20 Accepted:2026-02-25 Published:2026-05-21
  • Contact: Guixiang Zeng, Guoqiang Wang, Jing Ma E-mail:gxzeng@nju.edu.cn;wangguoqiang710@nju.edu.cn;majing@nju.edu.cn

PDF

322

赞

补充材料

1

摘要： 本项目针对“基础化学实验”中磺基水杨酸合铁配合物测定实验存在的技术瓶颈，如本底吸光干扰、无法确认未知配位体系的配位比n以及实验效率低(单次实验只测1种配体)，开展了系统性改进。在实验方法上，通过建立包含所有组分吸光度的方程组进行求解，有效消除了本底吸光的干扰效应，使单次实验可测配体种类由1种提升至11种。在分析技术方面，引入机器学习算法构建配合物稳定常数预测模型，协助探明实验体系中的配位比n，结合紫外-可见光谱实测数据进行验证分析，通过特征重要性分析揭示配位作用的关键影响因素，为配位场理论的教学提供实证依据。在教学方法方面，构建智能实验教学平台，有效降低师生数智化技术门槛，促进改进实验方法的推广应用。改进后的实验在保持分析精度的同时，显著提高了通量性和普适性，通过探究式教学模式有效培养了学生的数据解析能力和复杂化学问题的解决能力。

关键词: 数智化, 化学实验, 改进创新, 铁配合物

Abstract: This study addresses several technical limitations in the “Basic Chemistry Experiment” involving the determination of iron(III) sulfosalicylate complex, including background absorbance interference, inability to identify coordination ratios (n) in unknown systems, and low experimental throughput (measuring only one ligand per experiment). Methodologically, we developed a system of equations incorporating absorbance data from all components, effectively eliminating background interference and increasing ligand measurement capacity from 1 to 11 per experiment. Analytically, we implemented machine learning algorithms to construct a stability constant prediction model, facilitating determination of coordination ratios (n). The model was validated using UV-Vis spectral data, with feature importance analysis revealing key coordination factors, thereby providing empirical support for coordination field theory instruction. Pedagogically, we established an intelligent teaching platform to reduce technical barriers in digital implementation, promoting wider adoption of these improved methods. The enhanced protocol maintains analytical precision while significantly improving throughput and applicability. This inquiry-based approach effectively develops students’ data analysis skills and problem-solving capabilities for complex chemical challenges.

Key words: Digitalization and intelligence, Chemical experiment, Improvement and innovation, Iron complex