雷射剝削法製作碳奈米管輔助LiCoOx電極之製程及性質

Abstract

本實驗企圖利用脈衝式雷射共沉積碳奈米管和LiCoOx電極，以改善電極的電化學性質。採用均是波長10.6μm的脈衝式CO2雷射系統，其靶材是由LiCoO2 90wt%；graphite powder 9.3wt%；Indium powder 0.7wt%，粉末經由混合、球磨、壓製和燒結而成。以氬氣(Ar)與氧氣(O2)為濺鍍氣體﹔利用不同的沉積順序、壓力、氣體流量比、基板溫度和沉積時間，將電極沉積在已鍍上Pt的矽晶圓上。所沉積電極的結構與性質，將藉由掃描式電子顯微技術(SEM)、穿透式電子顯微技術(TEM)、能量散佈光譜技術(EDS)、拉曼光譜技術(Raman spectroscopy)、.感應耦合電漿質譜儀(Inductively coupled plasma-mass spectrometer)、X光繞射儀(X-ray diffractometer)、原子力顯微技術(AFM)和三極系統的循環伏安法(Cyclic voltammetry)測量分析。從實驗結果中可獲得下列結論。 就壓力對脈衝式雷射沉積(PLD)系統沉積電極的影響，結果顯示當壓力大於100 Torr時，有明顯的MWNTs被發現在電極之中，其管徑介於10 ~ 15 nm，且當壓力從0.1 Torr增加至100 Torr ，Li/Co的莫耳比例從1.4改變至241。而CNTs沉積對壓力的相依性已在文獻中被確立，亦即SWNTs只有當壓力大於50 Torr時才可以藉由CO2雷射剝削法產生。換言之，沉積具有CNTs輔助的LiCoOx電極，且Li/Co比例接近 1.4，需要藉由多重沉積過程，在不同的條件下分別沉積CNTs和LiCoOx。在靶材中不同元素被選擇性剝削的效果是由於具有較高熔點和較大原子尺寸的元素較難被雷射光束所激發。 就基板溫度對電極沉積的影響，在基板溫度大於300℃時，由Raman光譜分析看不到純氧化鈷的訊號。換言之使用較高的基板溫度可促使沉積的電極，其成分和結晶度更接近在600℃之LiCoO2化合物。其原因為較高的基板溫度具較高原子移動率，有利於形成熱力學上較穩定的相。另一方面，高的基板溫度會使得電極的表面粗糙度由Ra = 2.7 nm (室溫) 變成20.6 nm (600oC)，這是由於高溫下晶粒粗化所造成。從應用層面來看，對於製造小尺寸電池是一項不利的因素。 藉由改變壓力和溫度的雙重作用下而言，在相同的沉積溫度利用交互改變壓力以製造富CNTs和富LiCoOx的沉積層。XRD結果顯示LiCoO2相的(003)平面峰值(2θ=18.7o) 隨著基板溫度上升而增加；且代表碳管管狀石墨層的(100)平面峰值(2θ=42.7o)除了基板600oC以外都檢測不出來。這可能是較高基板溫度使CNTs具較佳的結晶性，且去除電極中非晶質的部分，使X光更易穿透電極內部所造成。另外，基板溫度對LiCoOx電極沉積的影響在沒有或有CNTs之複合電極時有著相同的趨勢。 對於沒有和有CNTs輔助的LiCoOx電極，其循環伏安法分析顯示，代表每一迴路電極能量密度變化之CV迴路曲□面積隨基板溫度增加而增加(0.028 ~ 0.078；0.01 ~ 0.130 mA*V*cm-2，分別代表沒有和有CNTs輔助的LiCoOx電極在基板溫度由300℃到600℃)。在較低基板溫度時，鋰相關氧化還原反應的峰值較弱，一來由於其LiCoOx電極結晶性較差，二來作為鋰離子傳導通道的CNTs的量較少所造成。隨著基板溫度增加到600℃，LiCoOx電極和CNTs具有較好的結晶性時，在沒有CNTs輔助的LiCoOx電極中，其氧化還原峰主要發生在3.3 ~ 3.4 V之間，相對於具有CNTs輔助的複合電極，其氧化還原峰發生在2.2 V、3.2 ~ 3.3 V、3.7 ~ 3.9 V附近。換言之，在基板溫度為600℃時，具有CNTs輔助複合電極的能量密度可改善為無CNTs輔助時的1.7倍。因此，調整在電極中CNTs的比例，或許可以得到更佳的改善。 考慮雷射剝削成長多壁碳奈米管的機制，可發現有兩種CNTs形貌，亦即包覆或無包覆觸媒之MWNTs兩種，其對應之直徑分別為40 ~ 50 nm和10 ~ 15 nm。圓頂小帽(Yarmulke)機制和唇對唇反應橋接模式(Lip-lip interaction bridging model)或許可以個別被用來解釋上述MWNTs的成長機制。有觸媒輔助的CNTs尺寸受頂端觸媒尺寸所決定，本文中鈷為觸媒。

In this work, the feasibility to use the pulsed laser deposition (PLD) method to co-deposit carbon nanotubess (CNTs) and the LiCoOx electrodes to improve their electrochemical properties was investigated. The CO2 pulsed laser ablation system (λ= 10.6 µm) was adopted, where the sputtering target is made of 90wt%LiCoO2, 9.3wt%graphite, and 0.7wt%In through mixing, milling, compressing and sintering processes. The electrodes were deposited on Pt-coated Si substrates with Ar and O2 as sputtering gases under different deposition sequences with various pressures, gas ratios, substrate temperature and time. The structures and properties of the deposited electrodes were characterized by AFM, SEM, TEM, HRTEM, EDX, Raman spectroscopy, ICP-MS, XPS and 3-electrodes CV (cyclic voltammetry) measurements. From the experimental results, the following conclusions can be drawn. Regarding effect of pressure in PLD system on electrode deposition, the results indicate that the significant MWNTs (multi-walled carbon nanotubes, 10 ~ 15 nm in diameter) in the electrodes can be observed for pressure > 100 Torr, and the Li/Co ratios of the electrode can vary from 1.4 to 241 by changing pressure from 0.1 to 100 Torr. The tendency of pressure dependence of CNTs deposition is conformed to the results in the literature, where the SWNTs can be produced with a pressure > 50 Torr by CO2 laser ablation. In other words, to deposit the CNTs-assisted LiCoOx electrodes with Li/Co ratio of 1.4, multiple-steps process is required to alternately deposit CNTs and LiCoOx under different conditions. Such selective ablation efficiency of different elements in the target is due to the fact that element with higher melting temperature and atomic size is more difficult to be excited by laser beam. Considering effect of substrate temperature on electrode deposition, the results indicate that at temperatures > 300oC there are no pure Co-oxides can be detected by Raman spectroscopy. In other words, a higher heating temperature can cause compositions and crystallinity of the deposited electrode to approach LiCoO2 crystals at 600oC. This is due to a greater chance to form the thermodynamically stable phases by increasing the elements mobility at higher temperature. On the other hand, a higher substrate temperature may substantially increase surface roughness of the electrode from Ra 2.7 nm (RT) to 20.6 nm(600oC) due to grain coarsening effect, which is a disadvantage from application point of view to fabricate a small sized battery. By combining the pressure and temperature effects, the CNTs-assisted composite electrodes made of alternating LiCoOx–rich and CNTs-rich layers were fabricated by alternately changing deposition pressure at the same deposition temperature. The XRD results show that the intensity of LiCoO2(003) peak (2θ=18.7o) increases with increasing temperature, and the layer-turbostatic (001) peak (2θ=42.7o), which signifies the existence of CNTs, is not detectable at all deposition temperatures, except at 600oC. This may be due to more crystallization of CNTs at higher temperature and a greater X-ray penetration depth by burning off more amorphous materials in the electrode. The temperature effect on LiCoOx electrode deposition is in the same track for electrodes with no CNTs or composite electrodes with CNTs. For both electrode with no CNTs and composite electrode with CNTs, the cyclic voltammetry(CV) analyses indicate that the loop area in the CV curve, which is an index of power density (mA*V*cm-2) variation for each cycle, increases with substrate temperature (0.028 ~ 0.078 and 0.01 ~ 0.130 mA*V*cm-2 from 300 to 600oC for electrodes without and with CNTs, respectively). At lower temperatures, the peak intensities of Li-related oxidation-reduction reactions are relative weak due to poor crystallization of LiCoOx and less CNTs formation to act as conductive channels for the reactions. As the temperature reaches 600oC, where both CNTs and LiCoOx in composite electrodes crystallize more completely, the CV peaks for electrodes with no CNTs are located around 3.3 ~ 3.4 V, in contrast to 2.2 V, 3.2 ~ 3.3 V and 3.7 ~ 3.9 V for CNTs-assisted composite electrode. In other words, the power density for the CNTs-assisted composite electrode can be improved 1.7 times greater then the corresponding electrode without CNTs at 600oC. The percent CNTs in the electrode may be able to adjust to make a better improvement. Considering the formation mechanisms of MWNTs by laser ablation, there are two types of CNTs morphologies: MWNTs encapsulated with and without catalyst. The corresponding tube diameters are 40 ~ 50 nm and 10 ~ 15 nm for CNTs, respectively. Where the yarmulke model and the lip-lip interaction bridging model may be applied to explain their grow mechanisms, respectively. The sizes of the catalyst-assisted CNTs are determined by the catalyst sizes, i.e. Co is the catalyst in this case.