Title: 過濾式電弧沈積非晶質類鑽碳膜的拉曼與電性分析
Raman and electrical analyze of amorphous diamond-like carbon films deposited by filtered cathodic arc system
Authors: 曾裕仁
Yu-Jen Tseng
Dr. Chia-Fu Chen
Keywords: 過濾式電弧沈積系統;粗粒子;非晶質類鑽碳膜;磁碟片;拉曼光譜;電阻;Filtered arc deposition system;Macro-particle;Amorphous diamond-like carbon film;Magnetic disk;Raman Spectrum;Resistance
Issue Date: 1998
Abstract: 硬碟片之磁區密度每年都會以60﹪的速度提高,為了達到更高的容量密度,磁頭與碟片間的飛行距離也必須越來越接近,如此可增加訊號雜訊S/N比值,降低失真。以10G/in2的容量密度來說,此距離(磁頭與碟片間的飛行距離)必須低於25nm,如此一來碟片和磁頭間的保護膜厚度也必需薄於5nm的範圍。以目前生產硬碟的技術而言,大部份都選擇利用濺鍍法沈積含氫或含氮非晶質類鑽碳膜(a-C:H or a-C:N)以保護磁性層,磁頭和碟片上的含氫非晶質類鑽碳膜約在10-12nm厚度的範圍,當薄膜厚度低於10nm時其抗磨耗的特性會降低,幾乎已到達極限程度,這種技術對未來的發展受到了限制。最近陰極電弧類鑽碳膜技術因此成為受注目的焦點。換言之,利用陰極電弧法沈積超薄(2-10nm)非晶質類鑽碳膜有許多優點例如鍍膜中含有高量的sp3組成、高的硬度、抗磨耗性及很低的摩擦係數,這些性質恰好很適合運用在磁頭/碟片的介面被覆。陰極電弧沈積技術,在近年來的研究結果發現此方法可以沈積非晶質類鑽碳膜,甚至可以達到鑽石膜的標準,惟陰極電弧沈積之缺點是會產生粗粒子(數μm)而影響薄膜品質及平滑度。因此為了解決上述缺點,近年來的研究方向大多投入在陰極電弧電漿粒子的過濾技術發展,其中以磁控弧管粗粒子過濾法最具效果,可徹底改善薄膜品質,並且可得到很高的粗粒子過濾效果。這篇論文簡要的描述了45°弧管過濾式陰極電弧沈積製程並討論偏壓效應、氫氣和氮氣摻雜對陰極電弧沈積非晶質類鑽碳膜的影響和特性分析及概述。 經實驗的結果顯示,沈積後的非晶質類鑽碳膜經由拉曼光譜儀、n & k Analyzer、電阻和AFM的量測發現:當無偏壓、無任何反應氣體摻雜時薄膜的拉曼Raman I(D)/I(G) 比值為最低,表示此膜有最高的sp3組成。薄膜品質會隨著偏壓或氮氣含量的增加而變差擁有較高的拉曼Raman I(D)/I(G)比值與較低的電阻,換言之,有較高的sp2組成使薄膜較趨向於"類石墨"。在基材偏壓較高時,氫氣含量的增加有助於薄膜形成sp3組成,隨著氫氣含量的增加而有較低的拉曼Raman I(D)/I(G)比值與較高的電阻使薄膜較趨向於"類鑽石"。偏壓越高可得到越平整的薄膜表面形態,但卻會有較高的sp2組成。在不改變引弧電流的條件下增加基材偏壓或提高H2氣流量皆會造成薄膜表面上粗粒子含量的增加。
The areal density of magnetic storage is increasing at a blistering pace of 60% annually. Reaching higher areal density targets dictates that magnetic spacing between heads and disks are reduced. For example, to achieve an areal density of 10 Gb/In2 for the next-generation magnetic recording media, the head-to-magnetic layer spacing must be scaled down to about 25nm. Consequently, in budgeting this magnetic spacing, it is required that disk and slider air bearing surface overcoats thickness be decrease to ~5nm from the current 10nm range. Thinner overcoat helps reduce the head-to-magnetic layer spacing, resulting in narrower read-back pulse width as well as higher signal-to-noise ratio and, ultimately, enabling higher areal density, i.e., more data bits to be written on the same disk surface area. Present choice of carbon overcoat in the magnetic storage hard disk drive industry is sputter deposited, hydrogenated amorphous diamond-like carbon (a-C:H). The typical thickness of sputtered carbon overcoats decreases gradually over the years and now stands at ~10nm. At such an ultra-thin regime (~5nm), the performance and reliability of the sputtered carbon films cause grave concerns discontinuity. The cathodic arc evaporated carbon films are developed for future application, cathodic arc can deposite ultra-thin amorphous hard carbon films of high sp3 content, high hardness, and low coefficient of friction. These properties make it of great interest for head/disk interface application, in particular for contact recording. The drawback of cathodic arc deposition is the production of macro-particle of the cathode material which have a typical size of microns and contaminate the films. Much effort has been put into the improvement of the deposition process by filtering the macro-particles from the cathodic arc plasma. A high quality filter system has been developed which satisfy the requirements of magnetic storage devices. In the present study, we briefly describe the 45-degree angle magnetic filter cathodic arc deposition process and investigate the influence of substrate bias, hydrogen-doping and nitrogen-doping on properties of amorphous diamond-like carbon films. After deposition, the film properties were analyzed by Raman spectroscopy, n & k Analyzer, resistance meter (Keithley 617). Film surface was examined by OM and AFM. It was found that without substrate bias and any reactive gas produces the best quality film with lowest Raman I(D)/I(G) ratio which indicate highest sp3 content in the film. On the other hand, increase the substrate bias and/or incorporates nitrogen doping into the film during the process will degrade the film properties. This type of film posses higher Raman I(D)/I(G) ratio with the G-peak shifts to higher wavenumber, plus low optical band gap energy and low resistance. All these evident indicate the film is more "graphite-like". The effect of hydrogen flow in the film is less significant, it is believed that the hydrogen can prevent the nucleation of graphite phase, and stabilize of sp3 bonding. With careful examination by AFM, it was fount that higher substrate bias can produce smoother film. Moreover, high substrate bias and/or high hydrogen gas flow rate will generate significantly more macro-particles on the surface.
Appears in Collections:Thesis