基於經修飾的阻抗分析和太赫茲時域頻譜的高敏感度與解析度的介電質顆粒感測器

Abstract

隨著環境保護逐漸受到重視，空氣汙染由於與公共衛生與健康成為了熱門討論的議題，介電質懸浮微粒排放隨著工業的快速發展成為了一個重點，因應該問題，用於監測懸浮微粒排放的感測器成為了一個需求，大部分狀況下，這些微粒會由過濾器防止排放到空氣中，因此此研究使用有介電質顆粒散佈在表面的濾紙來模擬該裝置，為了減少不必要的過濾裝置拋棄和有效監測排放，需要非接觸、非破壞性與即時監測，介電質品譜因此成為一個解決方案，除此之外，因應嚴格的工業廢棄物質排放標準，監測系統需要高敏感度，相較於訊號的大小或是能量密度，此研究提出兩種基於相位偵測的感測方法，第一種是經修飾的四線式阻抗量測，透過添加一些離散的電路元件到介面夾具上，即使電容大小變化極小仍然能帶來較大的相位轉變，使其能夠感測微小的介電質物質量，第二種是基於太赫茲時域頻譜，此方法使我們可以偵測太赫茲波的相位，使用相位而非接收到的能量作為感測工具帶來一些顯著的優勢，此文章將會詳盡討論此兩種感測方法的理論基礎與驗證實驗，最後將會有兩種方法的比較以及總結。

As environmental issues has been increasingly recognized, air pollution has induced much active discussion due to its strong relationship to public health. Dielectric yield in floating particle form has become serious issue since the rapid development of industry. To improve such issues, sensor for such particle discharge is in demand. In most scenario, particles are filtered by filters and are prevented from discharged into the air, so this research models them with a filter with salt particles scattered on it. In order to reduce unnecessary disposal of filtering devices and to monitor mission system effectively, non-attaching, non-destroying and real-time sensor is needed. Dielectric spectroscopy is a potential solution for these demands. Besides properties mentioned previously, high sensitivity should be met due to strict regulatory standard when it comes to industrial yields. This research proposes two methods based on phase detection instead of magnitude or power. The first one is modified 4-terminal impedance analysis. With addition of some discrete circuit components to interface of Kelvin clip, phase change becomes large despite small increment of the capacitor under test, making it possible to detect slight difference in the amount of dielectric substances. The second method is based on terahertz time-domain spectroscopy(THz-TDS). This technique makes detecting phase of terahertz waves achievable. There are some advantages employing phase detection instead of power absorption to develop particle sensors. Working principles and experiments are discussed in detail. There is also a comparison between these two methods and a summary of them in the end of this article.