Tunable Young's Modulus in Carbon MEMS using Graphene-based Stiffeners

Micro-electromechanical systems (MEMS) development is currently dominated by silicon-based processing, which has been refined extensively by the semiconductor industry. However, the material properties of silicon are not always conducive to the demands placed on high-performance MEMS devices. To alleviate this shortcoming, researchers are increasingly turning to novel MEMS materials. The newly emerging field of carbon-based MEMS (C-MEMS) attempts to leverage the diverse properties of carbon to push the performance of MEMS devices beyond what is currently achievable. Our motivation is to employ a carbon-carbon composite approach, to develop a new class of MEMS accelerometers that is sensitive over a dynamic range from micro-gravitational forces (G) to hundreds of G’s –surpassing the capabilities of currently available commercial MEMS accelerometers. Specifically, pyrolytic carbon, from the pyrolysis of commercial photo-resist, makes up the bulk of the composite of interest. Pyrolytic carbon structures are easy and inexpensive to pattern, and electrically responsive devices can be created with a single lithographic step. The resulting structures exhibit remarkable detail, possessing minimum feature sizes in the single digit micron range. Nano-materials such as graphene, and reduced graphene oxide (RGO) can be blended into the photo-resist prior to pyrolysis to allow for fine tuning of the electromechanical properties and response of the device. As a first step towards developing novel C-MEMS, nano-composite carbon cantilevers were fabricated for testing using 3 different lengths with fixed 10:1 length to width ratios. The devices investigated contained increasing amounts (0wt%, 0.5wt%, 1wt% and 2 wt%) of RGO blended into the novolac photopolymer. An increase in conductivity, from 900 S/cm to 1700 S/cm, was observed due to the conversion of RGO to a conductive graphene-based material during pyrolysis. A laser Doppler Vibrometer (LDV) recorded the response of each beam under base excitation while a piezoelectric actuator randomly excited the substrate and the LDV performed a single point velocity measurement on each beam tip. The first 3 bending modes were recorded for each of the beam variations. The results, illustrate increasing RGO (wt%) increases the natural frequency of the beam. The change in natural frequency indicated an altered stiffness and a model was generated to estimate the change in the Young’s modulus of the material. The Young’s modulus was estimated by matching the frequencies computed by the model to measured values to within 6% error. The Young’s modulus approximations were calculated within the error of the fitted model of 3 bending modes on 4 samples on RGO (wt%). Figure 1, shows the estimation of E for a 200 μm long beam versus nano-composite loading and the error in calculating Young’s modulus is approximately 10%. The unblended compared with 2 wt% RGO; E is estimated to increase from 41GPa to 68GPa, a factor of approximately 1.65, taken from devices in figure 2. This proof of concept study shows graphene-based fillers can substantially alter the frequency response (i.e. Young’s modulus) of carbon-based cantilevers. This research may provide solutions for a variety of MEMS designs that go well beyond our application. A successful demonstration would act as the pioneer of a new class of inexpensive, robust MEMS accelerometers with extreme sensitivity over a wide dynamic range.