藉由石墨烯與膜蛋白發展生物電子相關應用

Abstract

隨著人工智能與智慧醫療的發展，找尋合適的材料去溝通生物系統與電子產品已經成為重要的目標。電子產品一般在水溶液的環境容易受到潮濕的環境或是電解質水溶液存在的影響，導致漏電或腐蝕進而影響其功能的運作。在本論文中，我們利用具有生物相容性的材料來發展多元的生物電子的相關應用，期望未來能應用在穿戴式裝置或是人機介面的溝通工具。 首先，我們利用石墨烯電晶體來發展水相記憶體。我們發現在石墨烯與支撐的玻璃之間的奈米厚度水層會影響石墨烯的導電特性。我們藉由控制正負不同的閘極電壓來完成寫入和刪除的過程，並成功達成穩定的高導電度態(開)與低導電度態(關)。此石墨烯水相記憶體也展現能至少進行9次的重複寫入-讀取-刪除-讀取，且其高低導電狀態都至少能維持104秒。我們更進一步使用原子力顯微鏡與拉曼光譜來檢驗此兩狀態下水層的厚度的差異，也使用了延伸式DLVO理論去解釋水層厚度變化的形成原因。此利用奈米水層來調控石墨烯與玻璃間相互作用的技術，未來有機會應用於發展水中操作之極薄二維型態的非揮發性記憶體。 接著，我們結合先前在本實驗室研究的蛋白質明膠記憶體與視紫蛋白來增加其導電度和提升其記憶體表現。視紫蛋白為光驅動氫離子幫浦，在吸收光能後可以製造細胞膜兩側的氫離子濃度差異。我們的實驗結果顯示在明膠記憶體中間入一層具有視紫蛋白的脂質膜可以顯著的增加蛋白-結和水網絡的導電能力。我們認為視紫蛋白在光驅動後，會讓脂質膜兩側的電性因為氫離子的移動，而發生兩側非電中性的狀態，我們認為這樣的狀態會更進一步造成明膠整體結構的變化，進而產生更多的蛋白-結和水來增加電子傳遞的能力。加入視紫蛋白的明膠記憶體具有穩定且可重複操作的能力，且其開關電流比純明膠記憶體高了2倍。 最後，我們利用視紫蛋白的脂質膜改變膜電位的能力，來研究並控制目標神經細胞的活性。我們將具有視紫蛋白的脂質膜覆蓋住神經細胞群並使用雷射做為光源來刺激視紫蛋白改變神經細胞外的電位，並使用鈣離子影像來觀測螢光亮度變化並推測神經細胞是否因電位改變而發生變化。為了驗證視紫蛋白改變電位的能力，我們使用COMSOL軟體去建構模型，其結果說明紫蛋白能在至少10微米的位置內製造出0.06V的電位變化。這些初步的研究結果都支持我們建構之具有視紫蛋白的脂質膜能影響神經細胞活性。我們相信這個嶄新、非侵入性的方法具有可以控制神經細胞的潛力，以幫助建構電子元件與人體系統之間溝通的橋樑。

With the development of artificial intelligence and health sensing, developing suitable materials to bridge between biological systems and electronics has become an important goal. Electronic devices are generally susceptible to the presence of a moist environment or an aqueous electrolyte solution, resulting in leakage current or corrosion that affects the electronic function. In this thesis, we use flexible, biocompatible materials to develop a variety of bioelectronic applications and show them potentially to be applied to wearable devices and human-machine interfaces in the future. First, we demonstrated a solution-gated graphene memory device. We found that the nano-thickness water layer between graphene and the silica support can greatly affect the conductivity of graphene. We underwent the process of writing and erasing by applying the positive and negative gate voltages to the device and successfully achieved a stable high conductivity state (on) and low conductivity state (off). The water-based graphene memory also showed that it can undergo at least 9 WRER cycles, and its high and low conductivity states were maintained at least to 104 s. We further used AFM and Raman spectroscopy to confirm the difference of the water layer thickness between two states and also applied the extended DLVO theory to explain the formation and variation of the water layer thickness. The manipulation of the interaction between a graphene sheet and silica support through a thin water layer can also facilitate the realization of a 2D water-based non-volatile memory. Second, we combined a previously studied gelatin memory with bacteriorhodopsin (BR) to increase its conductivity and enhance its memory performance. After the light illumination, BRs can create additional protons on one side of the membrane and proton vacancies on the other side. We hypothesize the proton accumulation or proton deficiency could affect the structure of the gelatin-bound water metastructure by creating electrical non-neutrality. The non-neutrality could cause gelatin polypeptides to unwind to a greater extent and therefore cause more bound water to be trapped in the polypeptide-bound water network, enhancing the electron transport in the device. The BR-incorporated gelatin memory device also features long-term stability and reprogrammability. The ON/OFF current ratio is increased by 2 folds compared to the gelatin memory device with no BR-lipid-membrane. Using BR-incorporated gelatin thin film as a memory provides a new pathway for writing, reading, and erasing by controlling the structure and the water content in the gelatin, which could be further used for various bioelectronic applications. In addition, we prepared a BR membrane sheet to change the electric potential around target neuron cells to control their activity. We covered the neuronal cell with a BR membrane sheet and applied a laser beam as a light source to stimulate BRs to change the electrical potential around target neurons. Our calcium imaging results show that the neuron state did change after the stimulus. To verify the capability of the electric potential change by BRs, we used COMSOL simulation to construct a model, and the results suggest that the BR membrane sheet could produce a potential change of 0.06 V when it is placed at a position of 10 μm from the membrane. We believe that this new, non-invasive method could have potential to control neuron activities and bridge the gap between electronics and biological systems.