Side-by-Side Comparison of Passive MEMS Strain Test Structures under Residual Compression

Knowledge and control of residual strain is critical for device design in MEMS, and therefore it is important to establish standards for residual strain measurement. In this study, pointer, microring, bent-beam, and fixed-fixed beam test structures are used to evaluate residual strain both theoretically and experimentally. An equation that enables easier evaluation of bentbeam structures is derived. Also, a finite difference model that incorporates the non-idealities of fixed-fixed beams and determines an optimum fit to the measured deflection curve is presented. The model allows accurate residual strain evaluation of each buckled fixed-fixed beam. Experimentally, pointer structures were found to be susceptible to adhesion. Microrings, intended for residual tension assessment, also could not be evaluated because the residual strain was compressive. Bent-beam and fixed-fixed beams could both be evaluated. The main criterion for test structure effectiveness was taken to be the repeatability of residual strain on structures in close proximity; each should exhibit the same value. Using optical microscopy, the residual strain of bent-beams was determined with ± 13 με repeatability based on standard deviation of adjacent structures of similar design. Using optical interferometry, the residual strain of fixed-fixed beams was determined with ± 2 με repeatability based on standard deviation for adjacent beams of different lengths. The strain values obtained from the two structures are in reasonably good agreement. Cantilevers were also evaluated to obtain film curvature values.