聚(3,4-乙烯二氧噻吩)固態接觸式離子選擇電極在水耕物聯網監測與電流記憶效應之研究

Abstract

本論文聚焦於以固態接觸式離子選擇電極（solid-contact ion-selective electrodes, SCISEs）開發的原位 (in situ) 即時多離子監測系統應用於水耕養液。同時，在系統開發過程中，我們發現若能達到超能斯特斜率的靈敏度（super-Nernstian sensitivity），將有助於提升感測解析度，進而偵測養分濃度的微小變化。考量到 SCISE 在連續監測應用中的需求日益增加，若能直接透過感測裝置本身進行原位性能控制，將是具有潛力的策略。因此，我們開發出一種無記憶效應（memory-free）之極化量測法 (polarization measurement)，結合定電流極化與定電位去極化，不僅可提升電位訊號的重現性與靈敏度，也改善其長期穩定性。 水耕栽種因植物養分吸收的動態變化及環境波動，常面臨養分管理上的挑戰。傳統的監測工具（如導電度與 pH 測定）無法量化單一離子濃度。為克服此限制，我們開發了物聯網養分監測系統（IoT Nutrient Sensor System, IoNSS），可即時偵測關鍵巨量離子（鉀、硝酸根、銨根和鈣離子）並將數據上傳至雲端平台。該系統已應用於溫室與植物工廠，並能偵測特定離子吸收情況，同時觀察其與環境因子（如蒸氣壓差 (vapor pressure deficit, VPD）之關聯性。值得注意的是，IoNSS 使我們得以觀察特定離子失衡與生理障礙（如高蒸氣壓差下的鹽分累積或頂燒）之潛在關聯。此外，結果也顯示鉀離子濃度變化可能作為作物蒸散效率的間接指標。此發現突顯了將特定離子感測與微氣候數據整合的重要性，有助於提升診斷準確度並支持精準農業中的決策制定。 雖然物聯網養分監測系統 (IoNSS) 已應用於實際農場中，其解析度與長期穩定性仍受到環境雜訊與 SCISE 本身特性的限制。為解決此問題，我們進一步探索透過定電流極化進行原位電調控，藉由促進離子通量以達成超能斯特靈敏度。然而，我們發現文獻未提及過的現象–電流極化誘導的PEDOT 氧化還原記憶效應，並造成感測訊號重複性下降。藉由原位光電化學 (in situ optoelectrochemical) 分析，我們開發出結合定電流極化與定電位去極化的無記憶效應極化量測方法。定電位去極化能有效恢復 PEDOT 的氧化還原狀態，且不顯著干擾離子選擇膜。最後，無記憶效應極化量測方法提升了SCISE的靈敏度與長期偵測極限（經 31 天連續浸泡後可達 2.8 μM）。此外，我們還探討了定電流與定電位控制對 PEDOT 與離子選擇膜行為的影響。結果也顯示 PEDOT 的可逆氧化還原特性與 ISM 的再平衡能力，對整體感測動態具有關鍵作用。此研究驗證了原位電調控策略的可行性與彈性，未來可望作為提升 SCISE 感測表現之有效方法，並擴展應用至連續、即時監測系統。 總體而言，本論文整合了多離子感測、物聯網自動化與電調控技術，以強化 SCISE 感測平台之發展。我們證明了即時多離子資料可與特定應用領域的參數整合，實現多維度的感測架構。此整合資訊能展現過去技術難以觀察的系統動態，並具備在生醫診斷、環境監測與智慧製程控制等領域的應用潛力。我們認為，基於 SCISE 的物聯網監測系統在解析離子組成與其他關鍵變因的複雜關係上具有極大潛力，有利應用於多維參數整合的前瞻技術。

This dissertation focuses on the development of solid-contact ion-selective electrodes (SCISEs) for real-time, in situ, multi-ion monitoring system in hydroponic nutrient solutions. During the development of the system, we found that achieving a super-Nernstian sensitivity could enhance the sensing resolution, which is essential for detecting subtle variations in nutrient concentrations. Given the increasing demand for SCISEs in continuous monitoring applications, direct in situ control of SCISE performance using the device itself emerged as a promising strategy. To realize this, we developed a memory-free polarization measurement that combines galvanostatic polarization with potentiostatic depolarization, allowing us to improve sensor reproducibility, sensitivity, and long-term stability. Hydroponics face challenges in nutrient management due to dynamic nutrient absorption of plants and environmental fluctuations. Traditional monitoring tools (EC and pH meters) fall short in quantifying individual ion concentrations. To overcome this limitation, we developed an IoT Nutrient Sensor System (IoNSS), capable of detecting key macronutrients (K+, NO3-, NH4+, Ca2+) and transmitting real-time data to a cloud-based platform. The system was deployed in greenhouses and a plant factory. It enabled the detection of ion-specific uptake patterns and their correlation with environmental parameters, such as vapor pressure deficit (VPD). Notably, the IoNSS enabled observation of possible correlations between specific ion imbalances and physiological disorders such as excess salt accumulation and tip burn under high VPD conditions. Additionally, the result demonstrated that K+ concentration may serve as an indirect indicator of transpiration efficiency. These findings highlight the value of integrating ion-specific sensing with microclimate data to improve diagnostic accuracy and support informed decision-making in precision agriculture. Despite the practical deployment of the IoNSS, sensor resolution and long-term stability remained limited by environmental noise and the intrinsic properties of SCISEs. To address these limitations, we explored in situ electrical conditioning via galvanostatic polarization, aiming to achieve super-Nernstian sensitivity by enhancing ion flux. However, this approach revealed a previously unreported memory effect stemming from the polarization-induced redox behavior of PEDOT, which compromised signal reproducibility. Through in situ optoelectrochemical analysis, we developed a memory-free polarization measurement by combining galvanostatic polarization with potentiostatic depolarization. The potentiostatic depolarization enables effective restoration of the PEDOT redox state without significantly perturbing the ion-selective membrane (ISM). The method improved SCISE sensitivity, and long-term detection limits—reaching 2.8 μM after 31 days of continuous immersion. Furthermore, we investigated the respective roles of galvanostatic and potentiostatic control over PEDOT and ISM behavior. The results revealed how the reversible redox properties of PEDOT and the re-equilibration capacity of the ISM contribute to the overall sensing dynamics. These findings highlight the feasibility and flexibility of in situ electric conditioning as a useful strategy for enhancing SCISE performance, with a potential for future applications in continuous, real-time monitoring systems. Overall, this dissertation integrates multi-ion sensing, IoT-based automation, and electrical modulation to advance SCISE technology. We show that real-time multi-ion data can be coupled with domain-specific parameters to enable multidimensional sensing. Such integration reveals previously inaccessible system dynamics and offers strong potential for applications in biomedical diagnostics, environmental monitoring, and smart process control. We believe SCISE-based cyber-monitoring systems hold great promise for uncovering complex interrelationships between ionic composition and critical variables in real-world scenarios.