in-situ ATR-FTIR 기법을 이용한 반도체소자용 팔라듐치환반응 및 구리 무전해도금 공정의 표면반응기구에 대한 연구 : The in-situ ATR-FTIR analysis on surface reaction mechanism of Pd displacement reaction and Cu electroless deposition in semiconductor device fabrication

다국어 초록 (Multilingual Abstract)

Recently, as feature sizes in integrated circuits continue to decrease, copper interconnect technology has emerged as a leading candidate to replace conventional aluminum technology. Copper process has been successfully adapted in interconnect fabrication of logic device using electroplaiung technique for several years. Copper has lower resistivity than aluminum, higher maximum current density, and high resistivity for electromigration, properties that are critical to improve device performance and reliability. These properties are sufficient to provide the necessary increase in performance and reliability for the next generation of ultra large scale integration (ULSI) devices. In addition, with dual-damascene processing, vias and interconnects can be fabricated at the same time, thereby decreasing the number of process steps.
Electroless deposition process is the simplest technique because it does not need any external power or vacuum process. Furthermore, it does not need the copper seed layer, which used in current electrochemical technique. However, it has several difficulties, such as unstable electrolytes, impurity, high resistivity, and low adhesion, for applying to current interconnection fabrication. It is hard to find the basic research for surface chemical reaction according to process conditions, even though it has been researched for many decades. Therefore, this thesis is focused on the demonstration of chemical mechanism between electrolyte and substrate to resolve the difficulties of ELD in CHAPTER ?W?wThe Study on Mechanism of Electroless Copper Deposition in Formaldehyde Based Electrolyte?x
Also, I developed selective bottom-up growth by modulating Pd pre-treatment to adapt it to current dual damascene process. I have successfully controlled the selective copper growing only in vias and trenches of TiN/SiO2/Si and Ta/SiO2/Si substrates by placing palladium nanoparticles only in the patterned vias and trenches. We found that the post-cleaning conditions of de-ionized (DI) water rinsing and N2 blowing after Pd activation also strongly affects the topographical selectivity of copper growing in our electroless deposition. I explained this selective growth in CHAPTER ?U ?wThe Control of Topographical Selectivity in Palladium-Activated Electroless Copper Metallization?x
ELD consists of two process steps. Firstly, the substrate has to be catalyzed during the Pd activation, which is also belonged to electroless deposition. Then, copper is deposited on the Pd catalyzed surface. I found that as-grown Pd nanoparices are affected by temperature and the composition of electrolyte. Pd pre-treatment has an effect on following Cu ELD. Therefore, my mechanism studies were executed for both of the Pd activation and Cu ELD process in CHAPTER ?V ?wThe Study on Palladium Displacement Reaction on Hydrogen-Terminated Silicon Surface. In the study of Pd activation, I have studied on the mechanism of Pd autocatalytic displacement reaction on silicon surface in both of PdCl2-HCl-HF-NH4OH and PdCl2-HCl-HF baths. Using the in-situ and ex-situ ATR-FTIR and XPS experiments, it has been monitored that the Pd(NH3)2Cl2 in the electrolyte is reduced to metallic Pd particles by displacement reaction and the (NH4)2SiF6 compound is generated during the Pd activation in PdCl2-HCl-HF-NH4OH electrolyte. From these results, I have concluded that different mechanisms of displacement reaction in each electrolyte result in the different characteristics of Pd growth. Hydrofluoric acid, which is generated in Si etching reaction, is directly dissolved to SiF62- during the Pd displacement reaction. The SiF62- in the PdCl2-HCl-HF electrolyte plays a role as Si etchant, whereas SiF62- bonds with NH4+ ion in PdCl2-HCl-HF-NH4OH electrolyte. These different mechanisms result in the difference of Pd growth speed and characteristics.
Finally, I have investigated the reaction mechanism of Cu-ELD in various electrolyte conditions by in-situ and ex-situ FTIR analysis using ATR-liquid cell in CHAPTER ?W. In the analysis on surface reaction, I have monitored that gem-diol (CH2(OH)O), which is hydrated form of formaldehyde, is oxidized to HCOO− by Pd catalyst generating H2O at room temperature. In this case, the copper nuclei has been generated around Pd catalyst and developed to dense Cu film, which may result in "bottom-up growth" in via filling process. At 60??, however, the spontaneous oxidation of gem-diol has been observed even without Pd catalyst in my FTIR analysis. Therefore, Cu-EDTA is spontaneously reduced in electrolyte and, then, precipitated onto Si surface. In this case, the surface morphology of as-grown copper film is very coarse and it has a bad adhesion to surface. When the Cu ELD is executed on Pd/Si at 60??, furthermore, the surface reaction catalyzed by the Pd particles and the spontaneous reaction in electrolyte take place simultaneously together. From my experiment, I concluded that to optimize the Cu ELD process, therefore, this spontaneous reaction has to be prevented by modification of electrolyte or using additives.

목차 (Table of Contents)

• CHAPTER 1: The Basic Knowledge of Cu Electroless Deposition in Semiconductor Device and The mechanism of in Situ Attenuated Total Reflectance Fourier-Transform Infrared Spectroscopy = 1
• 1. Introduction = 2
• 2. Literature Survey = 3
• 2.1 Roadmap of Interconnection Process in Semiconductor device = 3
• 2.2 Copper Metallization Process = 9
• 2.3 Electroplating = 12
• 2.4 Electroless and Displacement Deposition =14
• 2.5. The Mechanism of Copper Electroless Deposition = 21
• 2.6. in-situ Total Attenuated Reflection Fourier-Transform Infrared Spectroscopy (in-situ ATR FTIR) = 23
• References = 27
• CHAPTER 2: The Control of Topographical Selectivity in Palladium-Activated Electroless Copper Metallization = 30
• Abstract = 31
• 1. Introduction = 32
• 2. Experiments = 33
• 3. Results and Discussions = 36
• 3.1. Surface Etching after Cleaning Process = 36
• 3.2. The Different Characteristics of Palladium Growth = 39
• 3.3. The Selective Copper Growth on Patterned TiN Surface = 40
• 3.4. The Selective Copper Growth on Patterned Ta Surface = 48
• 4. Conclusions = 49
• References = 54
• CHAPTER 3 : The Study on Palladium Displacement Reaction on Hydrogen-Terminated Silicon Surface = 56
• Abstract = 57
• 1. Introduction = 58
• 2. Experiments = 61
• 2.1. ATR-liquid cell = 61
• 2.2. FTIR Experiment Conditions = 62
• 2.3. Experiment Sequence = 62
• 2.4. X-ray Photon Spectroscopy (XPS) = 66
• 2.5. Scanning Electron Microscope (SEM) and X-ray diffractometer = 66
• 3. Results and Discussions = 69
• 3.1. Surface Chemical State of Si surface in HF solutions = 69
• 3.2. The Fourier-transform-infrared spectroscope study = 69
• 3.3. The x-ray photo-electron spectroscopy stu dy for the Pd-ELD on Si surface = 82
• 3.4. The mechanisms of Pd ELD on Si surface in PdCl2-HCl-HF and PdCl2-HCl-HF-NH4OH electrolyte = 90
• 4. Conclusions = 97
• References = 99
• CHAPTER 4: The Study on Mechanism of Electroless Copper Deposition in Formaldehyde Based Electrolyte = 101
• Abstract = 102
• 1. Introduction = 103
• 2. Experiments = 105
• 3. Results and Discussions = 108
• 3.1. The in-situ and ex-situ Fourier-transform infrared spectroscope study for surface oxidation and reduction reaction on Si and Pd surface = 108
• 3.2. The in-situ and Fourier-transform infrared spectroscope study for the Cu-ELD on Si and Pd surface = 114
• 3.3. The mechanisms of Cu ELD on Si and Pd activated surface in different temperature conditions = 122
• 3.4 The effects and surface reaction of TMAH and additive on the characteristics of copper growth = 123
• 4. Conclusions = 140
• References = 141
• APPENDIX = 143
• A.1 Kinetics of Cu Electroless Deposition = 144
• A.2. Total Internal Reflection = 150
• ACCOMPLISHMENTS = 164
• 초록 = 169
• 감사의 글 = 173