High Density Plasma Deposition of SiO2

Why HDP?

High-density plasmas, for the purposes of deposition, are defined as plasmas dense enough to deliver an ion flux exceeding the net deposition flux. Plasma densities of 1E10 to 1E11 /cm3 are required for commercially-interesting deposition rates. Ion bombardment has the fundamental effect of densifying any growing film: pores in the near-surface region are literally squeezed out by knock-on atoms forced into them, as long as the incoming ions are energetic enough to break bonds (50 eV is enough): HDP films deposited at less than 300 C are essentially as dense as thermal oxides, and do not shrink upon anneal even to 900 C, unlike all other low-temperature deposited films. Along with densification comes compressive stress. Since HDP films are dense and compressive, cracking on anneal is not an issue. However, excessive compressive stress can lead to metallization reliability problems. By applying high RF bias power (typically 500-1000 W on a 200 mm wafer), high sputtering rates can also be achieved, giving excellent step coverage and gap fill ability. See our discussion of HDP reactors for details on consequent reactor design issues.

Process Chemistry

Silane, SiH4, is almost universally employed as the precursor for HDP deposition; TEOS has been tried but confers no significant advantages. A mixture of oxygen and argon is added to provide the oxidant and the sputtering agent. The very high density of energetic electrons in the reactor means that the silane is readily cracked at low temperatures, so that in the absence of oxygen, amorphous silicon films will result. Thus, any stoichiometry from hydrogenated silicon to pure silicon dioxide can be deposited. This flexibility is also a process challenge: good control of the gas mixtures is required to deposit pure, hydrogen-free silicon dioxide. Excessive silane leads to Si-H in the films, whereas too much oxygen gives Si-OH groups.

[Lassig, AVS HDP Source Workshop, San Francisco, 1993]

Intentional operation in the silicon-rich regime provides films with good moisture-gettering properties, since the Si-H bonds will react with water molecules to form Si-OH and molecular hydrogen.

Increase of ion bombardment flux, ion energy, or increase of substrate temperature reduce the amount of hydrogen incorporated in the films. It is possible to deposit very low hydrogen films at near room temperature, but at the cost of reducing deposition rate (to increase the ion flux: deposition flux ratio). The hydrogen removal appears to be more subtle than simple sputtering of hydrogen from the surface: increased ion bombardment is much less effective in removing hydrogen from silicon nitride, and kinematics shows that energy transfer from heavy ions such as argon or oxygen to hydrogen atoms is inefficient.

The chemistry of deposition is complex. Ellen Meeks and coworkers at Sandia Laboratories have constructed a partial model, including important gas phase and surface reactions but not stoichiometry (SiO2 is the only allowed solid). The results for dominant gas phase and surface paths are shown below.

SURFACE REACTIONS [G indicates bond to substrate]

["Modeling High-Density-Plasma Deposition of SiO2 in SiH4/O2/Ar", E. Meeks, R. Larson, P. Ho, S. Han, E. Edelberg, E. Aydil and C. Apblett, Sandia Report SAND97-8241-UC 401 March 1997]

[Some other relevant references:
"High Quality High Rate SiO2 and SiN Room Temperature Formation by Utilizing High Excited Ions" T. Fukuda, K. Saito, M. Ohue, K. Shima, and N. Momma [Hitachi] IEDM 1992 p.285
"Role of ions in electron cyclotron resonance plasma-enhanced chemical vapor deposition of silicon dioxide" K. Seaward, J. Turner, K. Nauka, and A. Nel [HP] J. Vac. Sci. Technol. B13 118 (1995)
"Implementation of Electron Cyclotron Resonance Plasma SiO2 into Sub-half Micron Technology" C. Wang, Y. Huang, D. Pirkle, L. Liu, J. Wang, H. Cheng and M. Lin [TSMC/Lam] DUMIC 1995 p. 29
"Low Temperature Chemical Vapor Deposition of High Quality SiO2 Film Using Helicon Plasma Source" Y. Nishimoto, N. Tokumasu and K. Maeda [SPL] Jpn. J. Appl. Phys. 34 1021 (1995)
"Study of surface reactions during plasma enhanced chemical vapor deposition of SiO2 from SiH4, O2 and Ar plasma" S. Han and E. Aydil [UCSB] J. Vac. Sci. Technol. A14 2062 (1996)
"Ion Bombardment Induced Compositional Change in ECR-CVD SiO2 and SiN:H", D. Dobkin, K. Seaward, F. Mertz and K. Nauka, Proc. 13th Int'l Conf on CVD (Electrochemical Society Proc vol 96-5) p. 523 (1996)
"Characterization of High Density Plasma CVD USG Film", W. Lu, J. Zheng, J. Sudijuno, H. Yap, K. Fam, C. Leong, M. Liao, Y. Lin [Chartered], in "Multilevel Interconnect Technology", ed. D. Patel and M. Graef, Austin, TX, October 1-2, 1997, SPIE vol. 3214 p. 94]

 

Films: Table of Contents