Hostname: page-component-788cddb947-m6qld Total loading time: 0 Render date: 2024-10-15T19:19:18.886Z Has data issue: false hasContentIssue false

Effect of Top Dielectric Morphology and Gate Material on the Performance of Nitride-based FLASH Memory Cells

Published online by Cambridge University Press:  01 February 2011

Antonio Cacciato
Affiliation:
cacciato@imec.be, IMEC, CMOSDR, Kapeldreef 75, Leuven, B-3001, Belgium
Laurent Breuil
Affiliation:
laurent.breuil@imec.be, IMEC, Kapeldreef 75, Leuven, B-3001, Belgium
Geert Van den bosch
Affiliation:
vandenbosch@imec.be, IMEC, Kapeldreef 75, Leuven, B-3001, Belgium
Olivier Richard
Affiliation:
richardo@imec.be, IMEC, Kapeldreef 75, Leuven, B-3001, Belgium
Aude Rothschild
Affiliation:
aude.rothschild@imec.be, IMEC, Kapeldreef 75, Leuven, B-3001, Belgium
Arnaud Furnémont
Affiliation:
Furnémont@imec.be, IMEC, Kapeldreef 75, Leuven, B-3001, Belgium
Hugo Bender
Affiliation:
bender@imec.be, IMEC, Kapeldreef 75, Leuven, B-3001, Belgium
Jorge A. Kittl
Affiliation:
kittlj@imec.be, IMEC, Kapeldreef 75, Leuven, B-3001, Belgium
Jan Van Houdt
Affiliation:
Jan.vanhoudt@imec.be, IMEC, Kapeldreef 75, Leuven, B-3001, Belgium
Get access

Abstract

The nitride-based SONOS cell, for its excellent scalability and process simplicity, is the candidate to push the scaling roadmap for FLASH memories beyond the limit imposed on floating-gate memories by the electrostatic interference between adjacent cells. The traditional SONOS cell consists of a nitride layer (the storage element) encapsulated by two SiO2 layers which isolate the nitride layer from the Si substrate and the poly-Si gate (Poly-Si/SiO2/Si3N4/SiO2/c-Si). However, the thick tunnel oxide necessary to meet the retention requirements imposes a severe limit on the erase performance because of the erase saturation phenomenon. One possibility to guarantee both the erase and the retention performance is the replacement of the top SiO2 layer with materials of higher dielectric constant (high-k dielectric). The presence of a high-k dielectric reduces the electric field across the top dielectric, thus decreasing the unwanted parasitic electron injection from the gate during the erase operation. This will allow the cell to erase deep so to meet a basic requirement for Gigabit multilevel NAND memories. The introduction of high-k materials in the SONOS stack is unfortunately not straightforward. One problem is the Fermi-level pinning at the poly-Si/high-k interface. Another problem is the morphological changes the high-k material undergoes during the device fabrication thermal budget. These changes can modify the k-value and affect the band offset between gate and high-k material. The results may, in both cases, be the decrease of the barrier for electron injection from the gate and, as a consequence, the deterioration of the erase performance. In this paper we study the effect of gate material and of the morphological transformation associated with the high-k post deposition anneal on the erase and the retention behaviour of nitride-based cells. Two different high-k dielectrics are investigated: Al2O3 (which has already been found to be able to significantly improve the erase operation, guaranteeing at the same time excellent endurance and sufficient bake retention) and HfAlO. We show that both for Al2O3 and HfAlO a trade-off exists between erase and retention, higher PDA temperatures being beneficial for erase but detrimental for retention. We also discuss the effect of Fermi level pinning and poly-Si depletion on the erase behaviour and compare the erase performances of several PVD- and AVD-deposited metal gates.

Type
Research Article
Copyright
Copyright © Materials Research Society 2008

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. White, M. IEEE Circuits and Devices, 16, 22 (2000).Google Scholar
2. Van, J. Houdt Proceedings of the ICICDT, p. 43 (2006).Google Scholar
3. Hobbs, C.C. et al. IEEE Trans. Electron Devices, 51, 971 (2004).Google Scholar
4. Lee, C.H. et al. , Proceedings of the SSDM, p. 162 (2002).Google Scholar
5. Lee, C. H. et al. – IEDM Tech. Dig., p. 26.5.1 (2003).Google Scholar
6. Shin, Y. et al. , Proceedings of the IEEE Non-Volatile Memory workshop, p. 58 (2003).Google Scholar
7. Park, K. et al. , Proceedings of the IEEE Non-Volatile Memory workshop, p. 54 (2006).Google Scholar
8. Govoreanu, B. et al. presentation AM G1.4 to the MRS Spring Meeting (2006).Google Scholar
9. Cacciato, A. et al. , Proceedings of the ICMTD proceedings, p. 217 (2007).Google Scholar
10. Afanas, V.V.'ev, Stesmans, A. Mrstik, B.J. and Zhao, C. Appl. Phys.Lett. 81, 1678 (2002).Google Scholar
11. Pierret, R. F. in Semiconductors fundamental- Modular series on solid state devices (Addison-Wesley Publishing Company, 1989) p. 54.Google Scholar
12. Furnémont, A., PhD Thesis, University of Leuven, 2007, p. 166.Google Scholar
13. Hou, Y.T. et al. IEDM Tech. Dig., (2005),Google Scholar
14. Pan, J. et al. IEEE Trans. Electron Devices, 51, 581 (2004).Google Scholar