Hostname: page-component-cd9895bd7-hc48f Total loading time: 0 Render date: 2024-12-27T01:50:40.653Z Has data issue: false hasContentIssue false
  • English
  • Français

Electromigration Characterization of Damascene Copper Interconnects: Comparison Between CVD Copper and ECD Copper.

Published online by Cambridge University Press:  17 March 2011

Thierry Berger ,
Lucile Arnaud ,
Gérard Tartavel  and
Gérard Lormand
Thierry Berger
Affiliation:
ST MICROELECTRONICS, 38926 Crolles cedex, France
Lucile Arnaud
Affiliation:
Leti (CEA-Grenoble), 17 rue des Martyrs, 38054 Grenoble cedex, France
Gérard Tartavel
Affiliation:
Leti (CEA-Grenoble), 17 rue des Martyrs, 38054 Grenoble cedex, France
Gérard Lormand
Affiliation:
GEMPPM UMR CNRS 5510, INSA, 69621 Villeurbanne cedex, France
Get access

Abstract

We have characterized the electromigration performance of copper damascene interconnects using moderately and highly accelerated lifetime tests respectively at package and wafer level. Two metallizations have been studied: Chemical Vapor Deposition (CVD) copper deposited on CVD TiN (Process A) and electroplated (ECD) copper deposited on CVD TiN using 90 nm of CVD copper as a seed-layer (Process B). All metallizations were passivated with SiO2. Two line widths have been characterized: 0.6μm and 4μm.

For wide lines, we obtained similar activation energies (Ea) for both metallizations (0.63 for process A and 0.65 eV for process B). For narrow lines, the Ea value is 0.8eV for CVD copper whereas it is higher than 1eV for ECD copper. For wide lines of both metallizations, failure analysis performed with a Scanning Electron Microscope (SEM) gave clear evidences that microstructural gradients have a strong impact on voids and extrusions formation (i.e. that grain boundaries are an active diffusion path in spite of low Ea values). For narrow lines, diffusion at the upper interface is believed to be the main diffusion path.

From the reliability point of view, the extrapolated lifetimes of the metallization including ECD copper are much higher (1 to 2 orders of magnitude depending on the line width) than for CVD copper.

Type
Research Article
Information
MRS Online Proceedings Library (OPL) , Volume 714: Symposium – Materials, Technology and Reliability for Advanced Interconnects and Low-k Dielectrics II , 2001 , L6.9.1
Copyright
Copyright © Materials Research Society 2001

Access options

Get access to the full version of this content by using one of the access options below. (Log in options will check for institutional or personal access. Content may require purchase if you do not have access.)

References

REFERENCES

[1] Lloyd, J.R. and Clement, J.J., Snede, R., Microelectronics Reliability 39 (1999) 1595–1602 Google Scholar
[2] Berger, T., Arnaud, L., Gonella, R. et al. , proc. of the 2000 MRS spring meetingGoogle Scholar
[3] Surholt, T., Mishin, Y.M., Herzig, C., Phys Rev. B 50 (1994) 3577 Google Scholar
[4] Gupta, D., Mat. Chem. Phys. 41 (1995) 199 Google Scholar
[5] Burton, B. and Greenwood, G.W., Metal Science Journal 4 (1970) 215–218 Google Scholar
[6] Hu, C-K et al. , Thin Solid Films 308–309 (1997) 443–447 Google Scholar
[7] Berger, T., Arnaud, L. et al. , Microelectronics Reliability 40 (2000) 1311–1316 Google Scholar
[8] Morgan, S. et al. , Proc. IEEE Int. Conf. on Microelec. Test Structures 9 (1996) 283–287 Google Scholar
[9] Giroux, F. et al. , Proc. IEEE Int. Conf. on Microelec. Test Structures (1995) 229 Google Scholar
[10] Liu, X.H. et al. , Mat. Res. Soc. Symp. Proc. 516 (1998) 313–324 Google Scholar
[11] Meier, N.E. et al. , Stress induced phenomena in metallization: 5th Int. workshop (1999) 180 Google Scholar
[12] Glickman, E. and Nathan, M., J. of Appl. Phys. 80 (7) (1996) 3782–3791 Google Scholar
[13] Proost, J. et al. , J. of Appl. Phys. 87 (7) (2000) 2792–2802 Google Scholar