Effect of specimen size on Young's modulus and fracture strength of polysilicon

Citation
Wn. Sharpe et al., Effect of specimen size on Young's modulus and fracture strength of polysilicon, J MICROEL S, 10(3), 2001, pp. 317-326
Citations number
37
Categorie Soggetti
Eletrical & Eletronics Engineeing
Journal title
JOURNAL OF MICROELECTROMECHANICAL SYSTEMS
ISSN journal
10577157 → ACNP
Volume
10
Issue
3
Year of publication
2001
Pages
317 - 326
Database
ISI
SICI code
1057-7157(200109)10:3<317:EOSSOY>2.0.ZU;2-Q
Abstract
The microstructure of polysilicon specimens of varying size was examined an d tensile tests were conducted to determine if the measured modulus and str ength depend on the size of the specimen. All specimens were from the same MUMP's 25 run at MCNC, and the thicknesses were 1.5, 2.0, and 3.5 mum. Micr ostructure was examined in specimens as narrow as 2 mum and ranging up to 2 0 mum in width. The tensile specimens tested were 6, 20, or 600 mum wide an d 250, 1000, or 4000 mum long. Nothing in the transmission electron microsc opy (TEM) observations indicates any effect of specimen size on the microst ructure; the columnar grains are fine (0.2-0.5 mum) and uniformly distribut ed. The widths of all specimens were found to differ from the specified mas k values, and a more pronounced variation was measured for the smaller spec imens. Three different approaches are used to measure Young's modulus, and they all give a value of 158 +/- 10 GPa with no evidence of substantial eff ects of specimen size. However, the strength does increase somewhat as the total surface area of the test section decreases-from 1.21 GPa +/-0.08 GPa to 1.65 +/- 0.28 GPa-reflecting the fact that the larger specimens have mor e surface flaws. Test techniques and procedures are briefly presented along with detailed analyzes of the results.