Unloading rate and maximum load have been previously shown to affect the response of silicon to sharp indentation, but no such study exists for spherical indentation. In this work, a statistical analysis of over 1900 indentations made with a 13.5-μm radius spherical indenter on a single-crystal silicon wafer over a range of loads (25–700 mN) and loading/unloading rates (1–30 mN/s) is presented. The location of “pop-in” and “pop-out” events, most likely due to pressure-induced phase transformations, is noted, as well as pressures at which they occur. Multiple occurrences of pop-in and pop-out events are reported. Raman micro-spectroscopy shows a higher intensity of metastable silicon phases at some depth under the surface of the residual impression, where the highest shear stresses are present. A stability range for Si-II is demonstrated and compared with previous results for Berkovich indentation.

The influence of stress on the elastic modulus E and hardness H in soda-lime glass was studied in the Vickers residual stress field by nanoindentation. The Oliver–Pharr method of analysis first gave higher values of E and H, but after correcting for the pileup contact areas around the nanoindents, results consistent with literature values were obtained at regions in the stress field where the stresses were either low or close to zero. Determination of the pileup contact areas was made possible by the use of the atomic force microscope, which has facility for generating cross-section images of the indents. The elastic modulus was found to decrease with stress, which is explained with reference to the influence of applied stresses on the Si–O–Si bond angle. The hardness on the other hand did not depend on the stresses except in the region very close to the edge of the Vickers indent where the stresses are high.

This article appeared in the August 2004 issue of Journal of Materials Research. The following corrections are required.

Section II. Experiments p. 2488

The third paragraph in the Experiments section should appear as follows:

One mechanism to explore changes in the shape of an indentation load-displacement response is to normalize the trace by its peak point. It has been demonstrated that the normalized [h/h(PMAX), P/PMAX] experimental responses for bulk polymers indented at constant loading- and unloading rate with the same rise time (but at different peak load levels) are identical. Figure 1(a) shows raw load-displacement data for indentation tests performed at small peak loads in the thickest polymer film (Epon) in the current study. The peak loads, 1 and 2 mN, were chosen to correspond to depths less than 10% of the film thickness in both cases. The responses normalize to the same shape [Fig. 1(b)]. When the 1-mN normalized response is compared with those from much greater load levels (50 and 500 mN), there are clear changes in the shape of the response, both loading and unloading [Fig. 1(c)]. In particular, the loading response shifts from slightly less than quadratic (power law fit with exponent 1.8) for the 1-mN response, as would be expected for a quadratic material with some creep effect; to a response between quadratic and cubic (power law fit with exponent 2.6) for the 500-mN response. The unloading response is also altered in shape, with a steeper unloading tangent at the larger load.