可攜式高壓模組之開發和微電漿光譜在有機氣體檢測之應用

Abstract

電漿作為物質的第四態，在分析化學的檢測應用上利用其發射光譜而具有廣泛的應用，而微電漿因體積小、設備成本低、功率消耗低等優勢，所以在可攜式的檢測應用上也佔有一席之地。本研究設計兩種可攜式高壓模組，分別建構出兩套氣體檢測平台用以檢測揮發性有機物蒸氣，為了測試此兩套檢測平台作為有機氣體檢測器的表現，本研究使用乙醇蒸氣作為待測氣體，並且針對此兩種高壓模組的電性特徵、光譜表現等實驗結果進行評估。 本研究的第一套檢測平台使用藍芽遙控高壓模組產生DBD電漿，其中的設計概念為透過手機將正弦波聲音訊號遠端輸出，利用藍芽耳機模組將該訊號輸入至LM386功率放大器，透過將正弦波過度放大飽和而修飾成方波並用來控制BJT的開關，BJT在瞬間開關時將使變壓器獲得電壓而能使MGD產生電漿。在該模組的設計過程中，因選用的藍芽耳機模組為PWM擬合波形，所以須使用低通濾波器修飾波形，此外LM386具有內建直流偏壓且較小的電壓增益，因此須使用分壓器消除直流偏壓，並在增益腳位間連接電容以提升電壓增益，最後使用加速電容提升BJT開關速度後即可實現以手機聲頻訊號進行控制的DBD電漿系統。由於藍芽耳機模組內部為全橋式換流電路且BJT開啟時僅具有內電阻而降低電源電壓輸出，於是採用三個獨立電源運作以免損壞元件並維持電漿的穩定性。 本研究提出三種操作模式控制藍芽遙控高壓模組，掃頻模式使電漿開關頻率呈線性變化，可找出使乙醇特徵峰具較強響應的條件；調節模式使電漿在特定開關頻率下運作，能夠對乙醇蒸氣定量檢測，於本研究中發現使用開關頻率7 kHz時監測CH特徵峰後可得R2為0.9969、LOD達791.7 ppb的檢量結果；突衝模式使電漿在設定的脈衝間隔下以特定的操作點條件運作，能夠節省能源並減緩電極老化、損壞，因此可對乙醇蒸氣進行長期監控。在掃頻模式繪製的甲醇、乙醇、異丙醇二維光譜中，由於特徵峰強度、相對高度不同，因此若能建立有機氣體二維光譜資料庫，則此裝置具有發展為電子鼻的潛力。然而此裝置僅在氣氛環境為氬氣時才有對乙醇蒸氣的檢測能力，若在空氣環境中改變電漿開關頻率也僅能延長MGD的壽命，光譜則無明顯的CN特徵峰，因此現階段此模組並不適用於空氣環境的有機氣體檢測。 本研究的第二套檢測平台使用自激振盪高壓模組產生直流微電漿，希望透過串聯不同鎮流電阻值改變電漿性質以找出在空氣環境中有利於有機氣體檢測的條件。實驗中測試的鎮流電阻值範圍落在1到100 kΩ之間，並且在電性量測中發現串聯不同鎮流電阻時均產生自脈衝放電，且具有鎮流電阻值越小、峰值電流越大、放電持續時間越短的特性。在電漿光譜測試中也發現鎮流電阻越小，氮氣放光被抑制且具有較長的特徵峰拖尾，足以說明電漿性質在串聯不同鎮流電阻時會產生變化，且鎮流電阻越小時乙醇造成的CN特徵峰越明顯，因此在統整電性與光譜性質後可推測出高峰值電流、短放電持續時間造成高峰值功率時有利於空氣環境下偵測乙醇蒸氣。最後在串聯1 kΩ鎮流電阻後收取光譜並計算CN淨強度進行定量則可得R2為0.9963、LOD達37 ppm的檢量結果。

Plasma, as the fourth state of the matter, is widely used in the detection in analytical chemistry by optical emission spectroscopy. Besides, microplasma has the advantages of being small in size, low equipment cost, and low power consumption, which makes it has a place in portable detection applications. In this study, two portable high-voltage modules were designed to construct two sets of gas detection platforms to detect vapor of volatile organic compounds. In order to test the performance of these two gas detection platforms, ethanol vapor was used as the testing gas, and the experimental results, such as the electrical characteristics and spectral performance of the two high-voltage modules, were evaluated. The first gas detection platform developed by this research uses a Bluetooth remote control high-voltage module to generate DBD microplasma. By remotely transmitting the sine wave signals from a mobile phone through a Bluetooth headphone module and inputting the signals to the LM386 power amplifier, the sine wave could be saturated into a square wave that controls the switch of the BJT. Next, with the instant switch of the BJT, the transformer will obtain voltage and generate microplasma on the MGD. However, because the output waveform of the Bluetooth headphone module is PWM, a low-pass filter must be used to block high-frequency signals. Besides, due to the built-in DC bias and the small voltage gain of the LM386 chip, a voltage divider must be used to eliminate the DC bias, and a capacitor must be connected between the gain pins to increase the voltage gain. A speed-up capacitor is also added to increase the switching speed of the BJT to realize control of the DBD plasma system with audio signals from a mobile phone. Nevertheless, 3 independent power supplies are needed to avoid damage to the components and maintain the stability of the DBD due to the full-bridge converter inside the Bluetooth headphone module and the low impedence of the circuit that causes the output voltage decrease when BJT switches on. In this study, three operation modes are proposed to control the Bluetooth remote control high voltage module. The first operation mode is sweep mode which makes the switching frequency of the plasma change linearly and can be used to find the optimal switching frequency that makes the characteristic peaks of ethanol vapor have the strongest response. The second operation mode is regulate mode which operates plasma at a specific operation point and can be used for ethanol vapor calibration. In this study, by operating the switching frequency of 7 kHz and monitoring the CH characteristic peak, the calibration result showed that LOD was 791.7 ppb and R2 value was 0.9969. The third operation mode is burst mode which operates plasma at a specific operation point between the set up interburst intervals. This mode enables the possibility of long-term monitoring of ethanol vapor as it saves energy and prolongs the lifespan of electrodes. In addition, the two-dimensional spectra of methanol, ethanol, and isopropanol drawn in sweep mode have different intensities of characteristic peaks and relative heights. That means this device could be developed into an electronic nose by establishing a two-dimensional spectral database of different organic vapors. However, this device only has the ability to detect ethanol vapor when the atmosphere is argon. It could only change the lifespan of the MGD and have little CN characteristic peak responses in the spectra caused by ethanol vapor in air. Therefore, the vapor detection of volatile organic compounds with this module is still not applicable in air environment. The second gas detection platform developed by this research uses a self-oscillating high voltage module to generate DC microplasma. The purpose of constructing this gas testing platform is to change the plasma properties by connecting different values of ballast resistors in series and finding out the conditions that are beneficial to the detection of organic vapors in air. The selected values of ballast resistors in this work range from 1 to 100 kΩ. In the electrical measurements, it is found that self-pulsation discharge occurs at any selected value of ballast resistors. While connecting smaller values of ballast resistor, the characteristics of higher peak current and shorter discharge duration would be found. In addition, the performance of plasma spetra reveals that the characteristic peaks of nitrogen are suppressed and drag longer emission tails while connecting smaller values of ballast resistor, which indicates that the plasma properties change with different values of ballast resistor connecting. It is also found that the CN characteristic peak caused by ethanol vapor is more obvious when applying smaller values of ballast resistor. Therefore, by integrating electrical and spectral properties, it can be inferred that higher peak current and shorter discharge duration, which causes higher peak power, would be the conditions that are suitable for ethanol vapor detection in air. Finally, by connecting 1 kΩ ballast resistor in series, the net intensity of CN could be acquired for quantification, and the calibration result showed that LOD was 37 ppm and R2 value was 0.9963.