用於測定食物中黃麴毒素的樣品檢液製備之自動化預處理機性能優化

Abstract

台灣溫暖濕熱的環境，容易使得黃麴毒素於花生、穀類、豆類等農產品中孳生。其中，黃麴毒素B1有強烈的肝毒性與致癌性，且其耐高熱性也導致完全消除毒素有一定的困難。因此，事先預防受汙染的農產品進入市場，為保障消費者健康的唯一方法。本研究旨在完成開發一項針對固體樣品(以花生為例)的自動化前處理機，並優化其性能，以符合法規對於製備黃麴毒素B1檢液的步驟要求，從而確保後續檢驗結果之有效性。該前處理機主要由重量感測、研磨萃取以及流道系統組成，以實現樣品的磨碎、均質、過濾步驟。透過修改研磨機構設計以及參考過往文獻中提及的濕研磨方法，將磨碎效能由過篩率(孔徑：1 mm) 30%提升至99%。隔離流道系統中感測器之訊號干擾，將350 ml下的流量誤差百分比控制於1%，以確保萃取效率的一致性。重量感測系統則透過更換合適感測器以及實施自動校正程序，將重量感測誤差百分比由0.2%降低至0.02%。進樣機構與進樣管道的設計確保了量測完畢之樣品可掉落至研磨機構內。注水口位置的設計充分清除前處理過程中，噴濺於進樣管道內的樣品。探討清洗程序內的各項參數，包括清洗時間、清潔溶液體積、清洗次數，並驗證了系統在重複使用的情況下，經過清洗程序後仍能將殘留樣品的含量清洗至0.1%以下。本研究使用甲烯藍代替黃麴毒素B1塗抹在樣品上，並進行樣品前處理。對前處理後的檢液進行量測，結果顯示回收率為78%，表明了本系統能夠處理固體樣品，並有效萃取出目標物。本研究以原系統設計之基礎進行性能上的優化，且結果均滿足法規對於樣品前處理之要求。此外，自動化可行性的驗證確保了優化後的系統能夠重複處理固體樣品且不影響後續檢液之有效性。

The warm and humid climate in Taiwan makes it susceptible to the growth of aflatoxin in agricultural products such as peanuts, grains, and beans. Among them, aflatoxin B1 is highly hepatotoxic and carcinogenic, and its heat resistance makes complete toxin elimination challenging. Therefore, the safeguarding of consumer health relies on the prevention of contaminated agricultural products from entering the market in advance. The objective of this study was to develop an automated pre-treatment apparatus for solid samples (using peanuts as an example) and optimize its performance to meet the official requirements for aflatoxin B1 detection, ensuring the effectiveness of subsequent analysis results. The pretreatment apparatus was mainly comprised of weight sensing, grinding & extraction, and flow control systems to achieve sample crushing, homogenization, and filtration steps. By modifying the grinding mechanism design and adopting wet grinding methods, the grinding efficiency was improved from 30% (sieve size: 1 mm) to 99%. The flow control system ensured that the flow error percentage under 350 ml was controlled within 1% by isolating the signal interference of the flow sensor. The weight sensing error percentage of the weight sensing system was reduced from 0.2% to 0.02% by an appropriate sensor module and by implementing an automatic calibration procedure. The design of the sample injection mechanism and pipeline ensured that the measured samples could be completely dropped into the grinding cup after the weighting procedure. The design of the water injection port effectively removed the sample splashed inside the sample injection pipeline during pretreatment. Parameters within the cleaning procedure, including cleaning time, cleaning solution volume, and the number of cleaning cycles, were discussed and validated. The residual sample content was cleaned to below 0.1% even after repeated use. Finally, methylene blue was used as a substitute indicator for aflatoxin B1 applied to the validation of pretreatments. The result showed a recovery rate of 78%, indicating that the system could handle solid samples and effectively extract the target substance. The optimization of performance was conducted based on the pre-version prototype design, and the results met the requirements for sample pretreatment according to official regulations. Furthermore, the feasibility verification of automation ensured that the optimized system could handle solid samples repeatedly without affecting the effectiveness of subsequent analysis.