以固態接觸式離子選擇電極陣列試片應用於葉菜汁液離子組成分析

智慧農業應用物聯網的概念及技術，在農場中導入感測技術收集農場的數據，並透過與大數據分析及人工智慧技術的整合，提供農民更有效率的農場經營管理模式。為了收集蔬菜的營養素數據並將其轉化為對農民作物生長管理上有用的資訊的，本研究開發以固態式離子選擇電極陣列試片 (solid-contact ion-selective electrode array chip, SCISE array chip)為基礎，搭配可攜式的多通道電位感測模組感知蔬菜葉片汁液中離子的訊號，並將其發展成電子舌 (electronic tongue)透過對收集到的離子訊號資料進行圖譜辨識，識別不同的蔬菜樣本。SCISE為以導電高分子作為固態接觸層輔助離子電子轉換的全固態式離子選擇電極。藉由此一導電高分子固態接觸層，SCISE擁有更快的響應時間及更高的電位穩定性。本研究開發的感測陣列試片包含了鉀 (靈敏度為51.9 mV/decade)、鈉 (64.2 mV/decade)、銨 (59.3 mV/decade)、鈣 (45.6 mV/decade)以及鎂 (33.6 mV/decade)離子的選擇電極以及一固態參考電極。為了測試其應用，本研究以開發好的電子舌感知七種不同種類的蔬菜葉片汁液中的離子訊號 (濃度範圍：10-3〜10-4 M)，並將結果與離子色譜法進行比較。結果顯示電子舌能夠獲取每種蔬菜汁獨特並且可被識別的離子組成圖譜，暗示其圖譜辨識應用的可行性。而進一步的研究成果顯示，透過主成份分析 (PCA)和K-均值集群分析離子濃度的感測資料，可以成功識別不同耕種條件下 (正常條件、缺乏氮以及鉀元素)生長的蔬菜。綜合以上，此研究展示一基於固態式離子選擇電極陣列的電子舌，具有模組化便於攜帶、快速且直接的量測以及量測範圍廣等優點，除了識別不同的蔬菜樣本亦可同時提供蔬菜樣本營養素含量的診斷。

關鍵字

電子舌；固態式離子選擇電極；主成份分析；圖譜辨識；蔬菜營養素分析

並列摘要

Smart agriculture applies the concept and technology of internet of things (IoT) by introducing sensing technology to collect data from the farm; moreover, through combination with big data analysis and artificial intelligence, it provides farmers a more efficiency way to manage the farm. To collect the nutrient data of vegetables and transform it into meaningful information for farmers to manage the growth of crops, this study developed an electronic tongue based on a solid-contact ion-selective electrode (SCISE) array chip combined with a portable multi-channel potentiometric model for on-site analysis of vegetable samples. SCISEs are all-solid-state ion-selective electrodes with a conducting polymer for ion-to-electron transduction. With the conducting polymer solid-contact, the electrodes have faster response time and higher stability during potentiometric sensing. The developed SCISE array chip consists of potassium (SEN=51.9 mV/decade), sodium (64.2 mV/decade), ammonium (59.3 mV/decade), calcium (45.6 mV/decade), and magnesium (33.6 mV/decade) selective electrodes. To prove the application, seven types of fresh crude vegetable leaf juices (concentration range: 10-3~10-4M) were measured with the electronic tongue, and the result was compared with ion chromatography. It was found that the electronic tongue was able to obtain the unique, differentiable ion composition profile (a radar plot) of each vegetable juice, implying the fingerprint application of the present technology. Furthermore, using the electronic tongue and analyzing the multiplex ion concentration data by principle component analysis (PCA) method and K-means clustering, lettuces grown in different environments (deficiency in potassium and nitrate) are able to be discriminated. In summary, we demonstrate an electronic tongue based on SCISE array chip having the advantages of modulation, portable, fast and direct measurement, and have a board sensing range and emerging as a powerful tool for rapid ion composition profiling.

並列關鍵字

solid-contact ion-selective electrode (SCISE) ； electronic tongue ； pattern recognition ； principle component analysis (PCA) ； vegetable nutrient analysis