Multi-Resin Masked Stereolithography (MSLA) 3D Printing for Rapid and Inexpensive Prototyping of Microfluidic Chips with Integrated Functional Components

Abstract

Stereolithography based 3D printing of microfluidics for prototyping has gained a lot of attention due to several advantages such as fast production, cost-effectiveness, and versatility over traditional photolithography-based microfabrication techniques. However, existing consumer focused SLA 3D printers struggle to fabricate functional microfluidic devices due to several challenges associated with micron-scale 3D printing. Here, we explore the origins and mechanism of the associated failure modes followed by presenting guidelines to overcome these challenges. The prescribed method works completely with existing consumer class inexpensive SLA printers without any modifications to reliably print PDMS cast microfluidic channels with channel sizes as low as ~75 μm and embedded channels with channel sizes as low ~200 μm. We developed a custom multi-resin formulation by incorporating Polyethylene glycol diacrylate (PEGDA) and Ethylene glycol polyether acrylate (EGPEA) as the monomer units to achieve micron sized printed features with tunable mechanical and optical properties. By incorporating multiple resins with different mechanical properties, we were able to achieve spatial control over the stiffness of the cured resin enabling us to incorporate both flexible and rigid components within a single 3D printed microfluidic chip. We demonstrate the utility of this technique by 3D printing an integrated pressure-actuated pneumatic valve (with flexible cured resin) in an otherwise rigid and clear microfluidic device that can be fabricated in a one-step process from a single CAD file. We also demonstrate the utility of this technique by integrating a fully functional finger-actuated microfluidic pump. The versatility and accessibility of the demonstrated fabrication method have the potential to reduce our reliance on expensive and time-consuming photolithographic techniques for microfluidic chip fabrication and thus drastically lowering our barrier to entry in microfluidics research.

Keywords: 3D printing; microfluidic pump; microfluidic valve; microfluidics; resin; stereolithography.

Figure 1

(A). PDMS microfluidic chips developed with 3D printed positive molds. Fabrication steps include (i) developing a 3D CAD model of the positive mold, (ii) stereo-lithographically printing and curing the positive molds, (iii) casting and curing PDMS over the positive molds and (iv) plasma assisted bonding of the cured PDMS with embedded channels onto a glass slide. (B). Incorporation of our custom resin formulation with existing desktop MSLA printers to enable direct fabrication of microfluidic chips with embedded channels without PDMS casting and curing. Our custom resins also enable us to simultaneously incorporate multi-resin composition with tunable mechanical and optical properties in a single chip.

Figure 2

(A). Error between the nominal and printed channel dimensions using the SLA molds for PDMS casts to develop microfluidic devices. (B). A microfluidic droplet generator developed with an SLA printed mold yields channels with a width of 100 μm capable to generating aqueous droplets (C). 3D printed finger twist screw with a corresponding fluid reservoir can be integrated with microfluidic channels to enable pump free fluid transport. (D). Finger-actuated fluid pumping enables precise liquid metering in a microfluidic chiap. (E). Tunable and precise liquid injection into microfluidic chips based on the screw rotation angle in different sized microfluidic channels.

Figure 3

Schematic illustrating (A). trapped liquid resin in an embedded channel and (B). curing of the trapped resin due to multiple UV exposure cycles. (C). Log-linear plot for exposure time (t) vs. final layer thickness (z) for commercially available and custom resin formulations. (D). UV/Vis spectra of commercial and custom resin formulation with 1% NPS. (E). Microscopic images of 300 micron square channel cross-section printed with commercial resin with h_d = 120 μm (blue box, partially clogged), custom resin with low UV blocker composition with h_d = 150 μm (green box, fully clogged) and custom resin with high UV blocker composition with h_d = 60 μm (red box, clear). (F) Intermediate images of the microfluidic channel cross-section (for 300 μm square channel) after curing of each successive layer to seal the channel using both commercial and custom resins.

Figure 4

(A). Cross-sectional profile of embedded square channels) printed from commercial and custom resin (1% NPS) formulation to characterize successful and unsuccessful prints. 9 replicates were performed for each channel cross-section ranging from 50–600 μm. (B). Corresponding p values from a z-test (single tail) were calculated to determine the statistically reliable probability of printing clear unclogged channels [Null-hypothesis: The SLA resin can print microfluidic channels with unclogged embedded hollow channels with reproducibility of >90%; Criterion for rejection of; p value < 0.01].

Figure 5

(A). FTIR spectra of custom resin formulations with varying amounts of EGPEA and PEGDA. (B). Stress-strain profiles, obtained from tensile stress tests performed on custom resin prints. (C). Young’s modulus of custom resin prints as function of varying PEGDA concentration.

Figure 6

Top view (chip valve image) and side view (illustration) of the flexible pneumatic valve sandwiched between rigid flow channel and rigid compressed air chamber in the (A). valve open and (B). valve close sates (scale bar = 300 μm). (C). 3D printed Microfluidic chip incorporating separate inlet port, outlet port, channels for the liquid flow, and compressed air lines. (D). 3D blowout of the pneumatic valve illustrating the location of the flexible membrane (in red). Variation of valve closing pressure as a function of (E). membrane thickness and (F). resin composition.