Design of a micro-Functional Testing System for process characterization of a hot micro-embossing machine

Abstract

Growth in industrial, commercial, and medical applications for micro-fluidic devices has fueled heightened research and development into micro-fluidic design, materials, and increasingly manufacturing. Polymers (Poly(methyl methacrylate)-PMMA in particular) are the current material of choice given their low cost, wide range of material properties, and biocompatibility. Given most fabrication processes have focused on hard materials for the semiconductor industry, an alternate set of processes such as hot micro-embossing (HME) have received increased attention as manufacturing processes for high-volume polymer-based micro-fluidic production. An understanding of the equipment, process physics, control strategy, and metrology for part fabrication are required when moving from the lab to production level. An initial statistical analysis of PMMA parts fabricated on the first generation HME system showed the need to: (1) design a new HME system; and (2) establish alternative methods for characterizing micro-fluidic parts.

(cont.) A second generation HME system was constructed with fellow Manufacturing and Process Control Laboratory (MPCL) graduate students and a FTS (Functional Testing System) was developed to test whether HME parts from the new HME system were capable of flowing fluid and establish output metrics for process control based on fluid pressure and flow rate. The new characterization method was shown to have re-registration error as low as + 1.03% (overall RMS uncertainty of ±1.51%). The experimental data from tests run on the FTS fit a fluid model developed to the expected accuracy of --± 10% for all but the lowest aspect ratio micro-channel. Moreover, the FTS results were consistent with optical scans of a series of parts made with varying HME parameters. The FTS was able to detect differences that a few isolated optical scans could not. The FTS provided a bulk quantity to assess the geometry of the channel rather than at a specified location. These results and the deficiencies in existing metrology techniques warrant further exploration into functional-based testing for micro-fluidic devices to parallel well established testing methods in place in the IC industry. Functional testing does not have the capacity to replace traditional metrology; however, it can add an important output metric-a quantitative measure of the output parts fluid flow.