磁控濺鍍氧化物薄膜之製備與其性質之探討

Abstract

氧化物呈現各式各樣之物理特性而廣泛應用於電子和光電元件中,本文中 探討包含釔鋇銅氧,氧化鋯,氧化鈰,和鈦酸鍶鋇等氧化物薄膜在不同基板 上之成長與性質,探討之成長參數包括氧分壓,基板溫度,電源型式,濺鍍槍 內磁鐵組合,和濺鍍槍與基板相對角度等.在成長釔鋇銅氧超導膜方面,如 何去除或避免電漿內的高能粒子撞擊薄膜被發現是最重要的因素.一開始 吾人採用近來廣泛使用的離軸濺鍍法,然該法本質具低成長速率,且實驗顯 示對壓力變化有化學成份比上的偏差,因此吾人變動濺鍍槍內之磁鐵組合, 並使用直流激發電漿,結果顯示此新法可成長更佳之釔鋇銅氧超導膜,壓力 改變下仍具良好當量成份比,且有八公分大之均勻性,同時其薄膜成長速率 亦較快.在成長介電氧化物薄膜方面,吾人發現電漿內的高能粒子撞擊,如 同對釔鋇銅氧超導膜的破壞一樣,會劣化介電氧化物例如氧化鋯與鈦酸鍶 鋇的薄膜,且造成其成份偏差.然而具有簡單與穩定晶體結構之氧化物,在 電漿粒子撞擊下,薄膜仍有結晶相形成,比如在氧化鎂與氧化鋁單晶基板 上,可長出單晶氧化鈰薄膜.避開電漿粒子撞擊而成長的氧化鋯與鈦酸鍶鋇 薄膜,比同樣成長條件而在電漿粒子撞擊下成長的薄膜具更高的介電常數. 電性上吾人成長之介電氧化物其漏電流可控制在每平方公分一微安培,而 在矽晶片上成長之氧化鈰與氧化鋯薄膜,具有空間電荷電流的傳導模式.另 外本文也詳述利用原子力顯微鏡研究A軸釔鋇銅氧超導膜與鈦酸鍶鋇薄膜 的最初成長之成長機制.結果顯示薄膜和基板間的晶格匹配,並非決定其成 長機制之主要因素,相對的,氧化物本質上的異向性,或薄膜與基板間鍵結 的強弱會顯著地影響其成長模式.吾人最後研製超導體熱感測器,探討超導 膜表面形態對感測器的感測能力之影響.結果顯示薄膜表面的氧化銅微粒 對感測是有幫助的.基於上述實驗結果,以磁控濺鍍法成長氧化物薄膜,在 半導體工業與未來超導體的應用與發展上,具有重要的價值. Oxides display a wide variety of physical properties that are well used in electric and optoelectronic devices. In this study, we have explored the growth and properties of YBa2Cu3O7 (YBCO), yttria-stabilized zirconia (YSZ), cerium dioxide, and (Ba,Sr) TiO3 films on various substrates. Deposition variables in sputtering including oxygen partial pressure, total pressure, substrate temperature, power excitation, magnet assembly inside the gun, and relative angle of sputter gun to substrate normal during the growth procedure are changed to explore their effect on the properties of grown films. The elimination of plasma bombardment in the magnetron sputtering is found to be the most important for growing high quality YBCO films. We adopted off- axis sputtering to overcome the plasma bombardment effects at first. However, off-axis sputtering has the inherent disadvantage of a low deposition rate and a compositional nonstoichiometry with pressure. Thus, we modified this sputtering by changing the magnet assembly inside the gun and using dc power excitation. YBCO films grown with this modified sputtering exhibit better superconductivity and better compositional stoichiometry with deposition pressure for substrates over 8cm long. The deposition rate of thin films prepared by this sputtering is higher than that prepared by the conventional off-axis sputtering. In the aspect of the growth of dielectric oxide films, we found that plasma bombardment caused crystalline deterioration and compositional nonstoichiometry of the films like its effect on YBCO growth. However, thin films of dielectric oxides with simple and stable structures may still crystallize even under plasma bombardment and epitaxial cerium dioxide films can be grown on MgO and sapphire substrates. The YSZ and (Ba,Sr)TiO3 films grown without plasma bombardment exhibit higher dielectric constant than those with plasma bombardment. The leakage current of those oxide films can be controlled to around 1uA/cm2. The conduction mechanism in cerium dioxide and YSZ film on Si is dominated by the space-charge- limited current. In addition, we have characterized the initial growth of a-axis YBCO films and (Ba,Sr)TiO3 films on various substrates by using atomic force microscopy. The lattice-match between the film and the substrate is not found to be the dominant factor responsible for the observed growth mechanism. On the other hand, the essential nature of the bulk oxide can be decisive for the initial growth of oxide films. We have also explored the characteristics of a superconductor bolometer. The effect of film's surface morphology on the performance of the bolometer is emphasized. The results show that copper oxide particles on the film's surface are helpful for the absorption and response of YBCO bolometers. The fundamental exploration of oxide thin films leads the author to believe that magnetron sputtering is a feasible method to deposit dielectric oxide films applicable in electronic devices.