Process development to synthesize single-walled carbon nanotubes arrays by ECR-CVD

碩士 === 國立交通大學 === 工學院半導體材料與製程產業專班 === 96 === Processes to synthesize the catalyst-assisted single walled carbon nanotubes (SWNTs) on Si wafer were developed by oxygen-plasma pretreatment and electron cyclotron resonance chemical vapor deposition (ECR-CVD), using CH4 and H2 as source gases, Co, Ni an...

Full description

Bibliographic Details
Main Author: 黃碩傑
Other Authors: 郭正次
Format: Others
Language:zh-TW
Published: 2008
Online Access:http://ndltd.ncl.edu.tw/handle/44088480506875861144
Description
Summary:碩士 === 國立交通大學 === 工學院半導體材料與製程產業專班 === 96 === Processes to synthesize the catalyst-assisted single walled carbon nanotubes (SWNTs) on Si wafer were developed by oxygen-plasma pretreatment and electron cyclotron resonance chemical vapor deposition (ECR-CVD), using CH4 and H2 as source gases, Co, Ni and Fe as catalysts. The catalysts were first sputtered on Si wafer by physical vapor deposition (PVD) method with pure metal (Ni 99.99%, Co 99.99% and Fe 99.99%) as targets under Ar atmosphere, and then followed by O-Plasma pretreatment to obtain the well-distributed catalyst particles. The carbon nanostructures were then deposited on the pretreated specimens by ECR-CVD. One pattern made of Ni-oxide-catalyst and SiO2/Si substrate was also conducted to examine selective deposition of CNTs. Effects of stacking sequence of specimens to vary the specimen protection from direct plasma bombardment on carbon nanotubes (CNTs) growth were examined. The structures and properties after each step were characterized by SEM, HRTEM, XPS and Raman spectroscopy. From the experimental results, the following conclusions can be drawn. To grow SWNTs, the results show that there are existence of optimum deposition time (~ 50 min) and optimum bias (-150 V) to obtain the SWNTs with the highest tube number density (diameter ~ 1.13 nm for Co catalyst, and 0.87 ~ 1.21 nm for Ni catalyst), where the pretreated particle sizes are much greater than the tube diameters. The results also indicate that the tubes are mainly the rooy-growth SWNTs net works. In terms of effects on CNTs deposition, the O-plasma pretreatment is basically to provide the oxidized surfaces with many smaller extrusion sites to grow SWNTs at each site. The SWNTs deposition can be imagined that each extrusion site is first reduced by H-plasma to become nano-sized pure metal particles during initial CNTs deposition period, and then carbon atoms will diffuse in the pure metal catalysts from the higher temperature side and to precipitate to the lower temperature side of the catalyst to form SWNTs at the original extrusion sites of the oxidized particles. In other words, the optimum time is basically to have enough time to activate all extrusion sites to grow SWNTs, and the optimum substrate bias is basically to give enough energy to activate the reduction of all oxide extrusion sites. Furthermore, the order of tube diameters for different catalysts is Fe and Co > Ni. The results also indicate that the pattern made of Ni-oxide and SiO2/Si substrate can be successfully used to grow CNTs pattern for potential applications in device fabrication, where SWNTs can merely grow on the positions with Ni-oxide.