Design and characterisation of microfluidic channels fabricated using additive manufacturing techniques

Abstract

Recent advancements in additive manufacturing vis-à-vis 3D printing, have allowed it to make its way into the field of microfluidics. By utilising a low-cost, LCD-based Stereolithography (SLA) 3D printer, a feasibility study was conducted to fabricate different microfluidic platforms. This is done to highlight alternative fabrication techniques aside from conventional methods, such as Polydimethylsiloxane (PDMS) lay-up or subtractive manufacturing, such as milling and etching techniques. Initially, a serpentine mixing module fabricated using PDMS was procured to evaluate the equipment that would facilitate pressure-flow rate experimental testing. Successful preliminary characterisation of the acquired low-volume pressure sensors and thermal flow sensor was conducted. When operating at µL/min flow rate regimes, pressure data correlation with 14 % tolerance was achieved between two pressure sensors, hence verifying their correct usage and data acquisition, with an overall 15 % discrepancy to Computational Fluid Dynamics (CFD) simulations. Various microfluidic platforms were designed in Fusion 360 and successfully printed. These mainly involved open-top tape-capped channel varieties, with semi-circular and rectangular cross-sectional geometries having sub-millimeter dimensions. These platforms were highly influenced by the selection of the adhesive tape that was utilised to cap the surface containing the micro-channels. For these reasons, Gorilla Clear Tape was selected, and experimental testing yielded a maximum operating pressure of 108 kPa or 16 psi, before delamination occurred, causing leakage. Channel widths and depths ranging from 1000 µm down to 200 µm were explored. Platforms containing enclosed channels were also explored, however, such a design proved difficult to yield successful and repeatable printable platforms. Part print failure was mainly attributed to curing of residue resin in the micro-channel thus leading to undesired channel blockage. These designs inherently also required specific part placement and orientation for printing, hence increasing total print time and amount of resin used. Pressure-flow rate experimental results have been compared to analytical calculations and validated CFD simulation results. The unibody open-top tape-capped design yielded good correlation to analytical and simulation results. Optical measurements were also taken using a Cascade microscope, to ensure that the final dimensions of the printed part equal those specified in the Computer Aided Design (CAD) model. Later, a more accurate Sensofar non-contact 3D optical profiler was utilised to analyse some calibration platforms, thus being able to accurately visualise the micro-channel surfaces and cross-sectional profile. Tolerances of on average, ± 10 % in dimensional conformity, were attributed to material deformation such as shrinkage during printing and platform post-processing, and 3D printer limitations such as layer and xy-axis resolution. Experimental comparison of a Polymethyl Methacrylate (PMMA) injection moulded platform containing (1000×200) µm rectangular channels to an SLA 3D printed counterpart was also conducted. The 3D printed platforms employed a similar open-top tape-capped design approach and achieved induced pressures 3 % higher to the PMMA injection moulded counterpart. Based on the obtained results, it was hence concluded that 3D printing can facilitate platform fabrication, that yield similar results to platforms manufactured via a standardised technique. Research into a variety of micro-mixers was also conducted, with a selection of the reviewed topologies being re-designed and fabricated via SLA 3D printing. The micro-mixer designs were able to be printed successfully, with experimental data correlating well with CFD simulations. With a 13 % discrepancy to pressure-flow rate simulations and reasonable dimensional conformity between printed and CAD platforms, it was concluded that 3D printing can be utilised as a suitable fabrication method for more complex micro-channel designs. The profilometry measurements also provided a valuable insight towards an innate attribute of the chosen fabrication method. A sinusoidal ribbed pattern in the floor of the micro-channel having a 1 µm amplitude and 50 µm wavelength were investigated to determine whether these geometrical features affected the induced pressure drop and fluid velocity profile along the channel length. It was found that the ribbed pattern induced similar lengthwise sinusoidal profile in the velocity field, and yielded a small change in the net pressure drop, when compared to a similar yet smooth rectangular channel. These ribbed patterns were attributed to the pixel pitch of 51 µm of the LCD-based 3D printer that was utilised. In this study, successful microfluidic platform fabrication was conducted using a €400 SLA 3D printer that was easy to procure, utilise and maintain. It was concluded that 3D printing can be utilised as a cost-effective alternative to conventional microfluidic platform manufacturing techniques, and can be particularly effective for platform fabrication by researchers conducting preliminary analysis of their designed micro-channels.