飛秒尺度下的異常聲學反射現象: 親水介面在正己烷覆蓋下引發的異常聲學反射

Abstract

固/液介面的能量傳遞問題一直以來是人們難解的謎團之一，例如著名的Kapitza異常指出:實驗中從固體傳遞到液體的能量總是遠高於理論預測值。特別的是，因為本實驗室的飛秒脈衝音波具有亞兆赫帶寬的特性，正巧與Kapitza問題中1-10K的聲子之頻率重疊。因此，飛秒聲學成了研究此問題的合適技術。 在本論文中，我們將欲探討的固/液介面問題以氮化鎵/正己烷實現。旨在於解決過去固/液介面以水當作液體時所產生的表面水效應(多是因為水分子的強極性與固體彼此作用)。因正己烷擁有在液體中極低的極性，故可以期待其並不會有介面效應的發生，進而實現一個更加理想的固/液介面。而透過觀察飛秒音波在氮化鎵/空氣和氮化鎵/正己烷界面的回波，我們就能夠進一步研究液體對回波的影響。理論上，在觀察飛秒音波脈衝的特性時，如果其是由低損耗、高阻抗(固體)和高損耗、低阻抗(液體)所構成的介面反射，則應該滿足: 一、介面反射係數為負且其絕對值小於1。 二、反射音波的峰值不早於其對於固體/空氣介面的反射。 然而，我們在實驗上的量測得到了違反上述預期的結果(稱之為異常的聲學反射現象):音波回波在將氮化鎵表面加上正己烷之後竟提早回來。在氮化鎵樣品分析上，TEM影像顯示了一0.5奈米的氧化層於樣品表面; XPS分析顯示，除了氧化層，表面上更有額外的氧原子存在; SIMS分析則顯示了表面上的氫原子分布。這些分析共同指出了一個可能:樣品表面上有水分子的存在。此外，在文獻中我們也發現:在各種親水表面上會出現具有類似於冰結構的吸附水層(其機械性質:未知)。 因此，我們藉由模擬音波在表面為氮化鎵/氧化層/水/空氣和氮化鎵/氧化層/水/正己烷的反射來試著分析出其提早回來的成因。在模擬中，為了找尋這層表面吸附水的黏/彈性，我們利用蒙特卡羅法隨機的設定了10,000筆水層的特性，進而從中模擬並找出與實驗軌跡相差最小的結果。透過加入這層水，我們的確開始觀察到音波提早回來的現象。最後的模擬結果顯示:樣品表面上具有約1奈米的吸附水層，其在空氣中的鬆弛時間介在冰和彈性固體(損耗低)之間，而其在加入正己烷後的性質卻有如彈性固體一般。這是首次量測到表面吸附水的機械性質，且此結果成功的解釋了實驗上觀測到的異常現象。我們同時證明了飛秒脈衝音波在探測奈米尺度結構上的能力。最後，我們也深入的探討其他可能出現異常聲學反射的成因，並一一分析後進而回絕，也更加確信了我們所呈現的結果。

Many unresolved mysterious phenomena happened at solid/liquid interfaces, for example, the famous Kapitza anomaly illustrates the abnormal thermal transportation across this interface. The energy transmitted from solid to liquid is always higher in the experiment than theoretical prediction. Here the femtosecond acoustics can be of great help in studying this anomaly since the femtosecond acoustic pulses possess sub-THz bandwidth, which is analogous to the dominant acoustic phonons around 1-10K in the Kaptiza problem. In this thesis, we experimentally model the solid/liquid system as GaN/n-Hexane, since GaN is a well-studied acoustic material and n-hexane (C6H14) has an extremely low polarity to minimize the probable interfacial layering effect, which is usually found in the interfacial region of a solid/water system. By in situ monitoring the acoustic echo from the interface of GaN/Air and GaN/Hexane, we are able to study the influence on the echo that originates from the existence of the liquid. In a general condition, the transient reflection of an acoustic pulse from low-loss high- (solid) to lossy low- (liquid) impedance interface has: I. A negative reflection coefficient with an absolute value less than one. II. The peak of the reflected pulse no earlier than the case of solid/air. However, the acoustic reflection from the studied interface in our experiments show an anomalous behavior: The echo came back earlier after applying hexane onto the GaN surface. Nevertheless, the analysis from XPS shows the extra oxygen existence other than the native oxide layer (Ga2O3), which can be clearly observed in the TEM image. The SIMS analysis reveals the excess hydrogen distribution only at the sample surface. These clues drive us to consider that there were some water molecules adsorbs on the hydrophilic GaN surface. Furthermore, the surface-adsorbed water was found to possess an ice-like structure on various hydrophilic surfaces in the literature. Therefore, we then try to investigate the advanced echo by calculating the acoustic reflection based on the simulation structures of GaN / Ga2O3 / Water / Air and GaN / Ga2O3 / Water / n-Hexane. In order to search for the viscosity and elasticity of the water layers, we adopt the Monte Carlo method to help us find the simulation traces with least error compared to the experiment traces. As a result, the advanced echo can be indeed introduced by appropriate combinations of viscoelasticity in the interfacial water layer either next to the air or hexane. Finally, our result indicates that the adsorbed water layer possesses a relaxation dynamics property between ice and elastic-solid in the air, and it becomes more elastic-solid-like under hexane confinement. This finding successfully explains the acoustic anomaly and provides the new insight into the surface-adsorbed water. In fact, this is the first measurement of adsorbed water layer for its mechanical properties in a noninvasive way under ambient condition. Furthermore, we prove the capability of a femtosecond acoustic pulse to probe nanometer-scale structures. In the end, many other possibilities are also discussed as a reason to explain the acoustic reflection anomaly, while they are all found to be unlikely to resolve this issue.