4.5 Review

Wafer-Scale Microtensile Testing of Thin Films

Journal

JOURNAL OF MICROELECTROMECHANICAL SYSTEMS
Volume 18, Issue 5, Pages 1062-1076

Publisher

IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/JMEMS.2009.2029210

Keywords

Brittle materials; ductile materials; mechanical properties; microtensile test; thin films; wafer-scale measurements

Ask authors/readers for more resources

This paper reports on the mechanical characterization of thin films using the microtensile technique performed for the first time at the wafer scale. Multiple test structures are processed and sequentially measured on the same silicon substrate, thus eliminating delicate handling of individual samples. The current layout uses 26 test structures evenly distributed over a 4- in wafer, each of them carrying a microtensile specimen that bridges the gap between the fixed and movable parts of the micromachined wafer. A fully automated high- throughput setup makes possible the fast acquisition of data with statistical relevance for the reliable extraction of material properties. The technique was successfully applied to micrometer- and submicrometer- thick films. Two brittle materials, namely, poly-crystalline silicon ( poly- Si) obtained by low- pressure chemical vapor deposition and silicon nitride ( SiNx) produced by plasma-enhanced chemical vapor deposition, and a ductile material, i.e., evaporated aluminum ( Al), were characterized. The extraction of the Young's modulus E, tensile strength sigma(u), mean tensile strength (sigma) over tilde (u), and Weibull modulus m is demonstrated. Young's moduli thus obtained for the poly- Si, SiNx, and Al films were 156.3 +/- 2.6, 112.2 +/- 3.5, and 62.5 +/- 2.5 GPa, respectively. The SiNx layers, which have a mean tensile strength (sigma) over tilde (u) of 2.084(-0.177)(+0.169) GPa and m = 5.9(-1.6)(+1.8) , are the strongest from the fracture point of view when compared to poly- Si with (sigma) over tilde (u) =1.382(-0.026)(+0.023) GPa and m = 17.3(-3.2)(+3.5) and Al with (sigma) over tilde (u) = 0.347 +/- 0.013 GPa. In each case, the best estimate of the mean and the corresponding 90% confidence interval were evaluated using maximum likelihood estimation and the likelihood ratio method, respectively, on the basis of Gaussian and Weibull statistics. [2008-0281]

Authors

I am an author on this paper
Click your name to claim this paper and add it to your profile.

Reviews

Primary Rating

4.5
Not enough ratings

Secondary Ratings

Novelty
-
Significance
-
Scientific rigor
-
Rate this paper

Recommended

No Data Available
No Data Available