Silicon-based implantable neural devices have great translational potential as a means to deliver various treatments for neurological disorders. However, they are currently held back by uncertain longevity following chronic exposure to body fluids. Conventional deposition techniques cover only the horizontal surfaces which contain active electronics, electrode sites, and conducting traces. As a result, a vast majority of today’s silicon devices leave their vertical sidewalls exposed without protection. In this work, we investigated two batch-process silicon dioxide deposition methods separately and in combination: atomic layer deposition and inductively-coupled plasma chemical vapor deposition. We then utilized a rapid soak test involving potassium hydroxide to evaluate the coverage quality of each protection strategy. Focused ion beam cross sectioning, scanning electron microscopy, and 3D extrapolation enabled us to characterize and quantify the effectiveness of the deposition methods. Results showed that bare silicon sidewalls suffered the most dissolution whereas ALD silicon dioxide provided the best protection, demonstrating its effectiveness as a promising batch process technique to mitigate silicon sidewall corrosion in chronic applications.

基于硅的可植入神经装置具有巨大的翻译潜力，可作为对神经系统疾病进行各种治疗的一种手段。但是，由于长期暴露于体液后，它们的寿命目前仍不确定，因此无法忍受。常规沉积技术仅覆盖包含有源电子器件，电极位置和导电迹线的水平表面。结果，当今绝大多数的硅器件都暴露出其垂直侧壁而没有保护。在这项工作中，我们分别研究了两种批处理二氧化硅沉积方法，并结合在一起进行了研究：原子层沉积和电感耦合等离子体化学气相沉积。然后，我们利用涉及氢氧化钾的快速浸泡测试来评估每种保护策略的覆盖质量。聚焦离子束横截面，扫描电子显微镜和3D外推法使我们能够表征和量化沉积方法的有效性。结果表明，裸露的硅侧壁受到的溶解最多，而ALD二氧化硅提供了最好的保护，证明了其作为缓解长期应用中硅侧壁腐蚀的有前途的批量处理技术的有效性。