The Applications of Statistical Quantification Techniques in Nanomechanics and Nanoelectronics

Wenjie Mai,Xinwei Deng

DOI: https://doi.org/10.1088/0957-4484/21/40/405704

IF: 3.5

2010-01-01

Nanotechnology

Abstract:Although nanoscience and nanotechnology have been developing for approximately two decades and have achieved numerous breakthroughs, the experimental results from nanomaterials with a higher noise level and poorer repeatability than those from bulk materials still remain as a practical issue, and challenge many techniques of quantification of nanomaterials. This work proposes a physical-statistical modeling approach and a global fitting statistical method to use all the available discrete data or quasi-continuous curves to quantify a few targeted physical parameters, which can provide more accurate, efficient and reliable parameter estimates, and give reasonable physical explanations. In the resonance method for measuring the elastic modulus of ZnO nanowires (Zhou et al 2006 Solid State Commun. 139 222-6), our statistical technique gives E = 128.33 GPa instead of the original E = 108 GPa, and unveils a negative bias adjustment f(0). The causes are suggested by the systematic bias in measuring the length of the nanowires. In the electronic measurement of the resistivity of a Mo nanowire (Zach et al 2000 Science 290 2120-3), the proposed new method automatically identified the importance of accounting for the Ohmic contact resistance in the model of the Ohmic behavior in nanoelectronics experiments. The 95% confidence interval of resistivity in the proposed one-step procedure is determined to be 3.57 +/- 0.0274 x 10(-5) ohm cm, which should be a more reliable and precise estimate. The statistical quantification technique should find wide applications in obtaining better estimations from various systematic errors and biased effects that become more significant at the nanoscale.