2000 Fiscal Year Final Research Report Summary
Advanced ULSI multilevel metallization using non-steady state by Flow Modulation technique
| Project/Area Number |
11450269
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (B)
|
| Allocation Type | Single-year Grants |
|---|
| Section | 一般 |
|---|
| Research Field |
Material processing/treatments
|
|---|
| Research Institution | The University of Tokyo |
|---|
Principal Investigator |
SHIMOGAKI Yukihiro Graduate School of Engineering, Associate Prof, 大学院・工学系研究科, 助教授 (60192613)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
OHBA Takayuki Semiconductor Leading Edge Technologies, Advanced Technology Research Dept., Research Fellow, 基盤技術研究部, 主任研究員
HAMAMURA Hirotaka Japan Society for the Promotion of Science, Research Fellow, 特別研究員
|
|---|
| Project Period (FY) |
1999 – 2000
|
| Keywords | CVD / Barrier Metal / Sequential Process / Nucleation / Thin Film Growth / Crystallization / Impurity Control / Resistivity |
|---|
| Research Abstract |
The device integration requires narrow wiring technique which results in high resistivity. This will lead to employ multilevel metallization scheme, in which the width of conductive metal line keeps the same level. Thus, the number of metal line layers now exceeds 8 and it will go up to 12, in the near future. The so-called back end process to fabricate multilevel metal layers became the dominant process for ULSI process. These structures require CVD process to form metal lines, because it has superior step coverage profile in high aspect ratio trenches/holes. The ultra thin films with a thickness of less than 10nm are also important, because of the small dimension of the ULSI devices. However, the integration of novel material will suppress the formation of continuous thin layers due to the nucleation issues.
The normal CVD employed steady process conditions, such as gas flow rate, temperature, and pressure. If we introduce sequential change of these parameters, it will enhance the ads
… More
orption and desorption of the precursor. That may contribute to control the initial step of thin film growth. When we started this research project, there were not so much related research, but recently ALD (Atomic Layer Deposition) process was proposed to improve ULSI device fabrication. Those methods are using sequential flow of source precursors. For example, in our study, TiN deposition from TiCl_4/NH_3 chemistry was investigated. NH_3 was supplied to the reactor all the time of deposition. TiCl_4 supply was switched on/off to make TiN deposition with NH_3 or just to make reduction by NH_3. The step coverage of TiN films from TiCl_4/NH_3 becomes excellent when the concentration of TiCl_4 becomes high. The reason for this behavior can be explained by following mechanisms. When the TiCl_4 concentration is high, the growth rate becomes independent of TiCl_4 concentration due to saturated surface adsorbates. This is an ideal situation for uniform step coverage, however, the high concentration of surface adsorbates turns into high residual chlorine concentration. Thus the FMCVD sequence is proposed. The deposition will be made with high TiCl_4 concentration range to realize excellent step coverage. In the next step, only NH_3 will be supplied to reactor to make chlorine reduction. We have proved that FMCVD sequence can realize the excellent step coverage and low resistivity at the same time. We also worked on how the residual chlorine is removed from TiN film using TDS, TEM amd EDX. The results clearly showed that residual chlorine mainly exists in the grain boundaries of TiN poly crystals. The FMCVD with many cycle times also contributed to remove chlorine which was incorporated in TiN grain. These investigations contributed to design the optimum process conditions for FMCVD and ALD. Less
|
