Due to their slight structural differences, the rapid identification of benzimidazole fungicides is challenging. In this study, three benzimidazole fungicides (FBZ, CBZ and TBZ) with similar structures were added to the 2C30@3Q[8] host–guest complex. The solutions changed color, suggesting that a reaction occurred. Based on this host–guest interaction, a supramolecular colorimetric sensor array was constructed by extracting the red, green and blue (RGB) color change of the host–guest complex in response to fungicides. Linear discriminant analysis (LDA) identified and distinguished the benzimidazole fungicides with similar structures. The supramolecular colorimetric sensor array was in the 1–20 μM range. The Euclidean distance of the array showed excellent linear correlation with the fungicide concentration, with a 10 mol/L detection limit and a signal-to-noise ratio of 3. The colorimetric sensor array clearly distinguished between the binary and ternary fungicide mixes. The sensor array was integrated with machine learning to predict TBZ concentrations in river water. The applicability of the supramolecular colorimetry sensor array was verified by distinguishing and identifying three benzimidazole fungicides in river water and fruit skin.

中文翻译：

---

超分子比色传感器阵列结合机器学习用于苯并咪唑类杀菌剂判别与预测


由于苯并咪唑类杀菌剂的结构略有不同，快速鉴定苯并咪唑类杀菌剂具有挑战性。本研究将三种结构相似的苯并咪唑类杀菌剂（FBZ、CBZ和TBZ）添加到2C30@3Q[8]主客体复合物中。溶液变色，表明发生了反应。基于这种主客体相互作用，通过提取主客体复合物响应杀菌剂的红、绿、蓝（RGB）颜色变化，构建了超分子比色传感器阵列。线性判别分析（LDA）识别并区分了具有相似结构的苯并咪唑类杀菌剂。超分子比色传感器阵列的范围为 1–20 μM。该阵列的欧氏距离与杀菌剂浓度呈现良好的线性相关性，检测限为10 mol/L，信噪比为3。比色传感器阵列可以清楚地区分二元和三元杀菌剂混合物。传感器阵列与机器学习相结合，可预测河水中的 TBZ 浓度。通过区分识别河水和果皮中的3种苯并咪唑类杀菌剂，验证了超分子比色传感器阵列的适用性。