Typography-Like 3D-Printed Templates for the Lithography-Free Fabrication of Microfluidic Chips

Materials

Fluorescent dyes (Fluka, Shanghai, China), phosphate-buffered saline (PBS; pH 7.4; Radnor, PA), and CCK-8 (Dojindo, Tokyo, Japan) were used in this study. Cisplatin (TopScience, Shanghai, China), as a common chemotherapeutic used to treat a number of cancers, was involved as an illustration. Formlab standard clear resin and PLA were complimentary obtained along with the purchase of the FDM 3D printer (PrintRit, Guangzhou, China). PDMS and its curing agent were purchased from Dow Corning GmbH (Midland, MI).

The tumor cells (A549) were cultured at 37 °C and 5% CO₂ in Dulbecco’s modified Eagle’s medium (DMEM) with 10% fetal bovine serum and 15 mL/L antibiotic solution (penicillin–streptomycin solution). The culture media was changed every other day. Before experiments, the tumor cells were taken from flasks and diluted with DMEM to the required concentrations after centrifugation.

FDM 3D-Printed Templates for the Fabrication of Drug Screening Microfluidic Chips

A commercial FDM 3D printer (SmartPrint2.0) ( Fig. 1a ) was used in the study. The designed drug screening microfluidic chip was first split into basic units and plotted with CAD software ( Fig. 1b ). To print the additive material (PLA), the printing temperature was set above 155 °C. The units were printed onto a glass pane and were easily cut off from the glass pane after cooling. A coordinate paper was stuck under a glass slide or a petri dish as spatial reference to align accurately. An electronic hot plate and a welding pen were used to fix and connect the basic units together ( Fig. 1c ). The completed template was obtained after connecting all the modular units as designed and cooling in the air.

OPEN IN VIEWER

Figure 1.

(a) Photograph of FDM 3D printer (SmartPrint 2.0) used in this study. (b) The PLA template units were involved to fabricate a drug screening microfluidic chip. (c) Photograph of 3D-printed PLA template on a glass slide. (d) A large serpentine template was employed to explore the optimal conditions in this study.

PDMS and its curing agent were mixed at a weight ratio of 10:119 under stirring. After degassing for 20 min and removing bubbles, the PDMS mixture was poured over the completed template and cured in an oven at 80 °C for 1 h to form molds. PDMS molds should be stripped from the 3D-printed PLA templates under the optimal conditions. Finally, the PDMS molds were treated with oxygen plasma (Harrick Plasma Inc., Ithaca, NY) for 1 min to activate the silicon–oxygen bonds on their surface and were bonded together. To strengthen bonding, the developed microfluidic chips were stored at 80 °C for another 10 min.

The heating temperature and time greatly influenced the PLA template deformation. Therefore, the optimal conditions of template bonding had been explored based on a large serpentine template unit of 2.5 cm in width and 7.5 cm in length ( Fig. 1d ). The large template unit was placed on a glass slide and heated by an electronic hot plate from 140 to 220 °C for 1 min at each temperature. The pictures taken every 5 s were collected in order to find the variation tendency of the PLA template attached to the glass substrate.

Fluorescent dyes diluted with PBS were used to characterize the drug screening microfluidic chip. They were injected into the microfluidic chip at 20 μL/h with a syringe pump (LSP04-1A, LongerPump, Baoding, China) to generate a gradient concentration. Before each injection, all microchannels in the chip were filled with PBS buffer.

Cell Viability Study in the Drug Screening Microfluidic Chip

Before tumor cell loading, the microfluidic chip device was autoclaved at 135 °C for 30 min and sterilized by UV light overnight. Cell culture medium was continuously injected into the chip at 20 μL/h until all the microchannels were filled. To seed tumor cells into each culture chamber, 200 μL of DMEM solution containing cells at a density of 1 million/mL was directly added into the open chamber. The microfluidic chip was kept at 37 °C and observed every 12 h until the cells became confluent in the chambers. Finally, 0.1 μg/mL cisplatin in DMEM was injected at 20 μL/h from the other inlet, while cell culture medium was continuously injected into the microfluidic chip at 20 μL/h. Twenty-four hours later, cell images were taken and live cell numbers were compared between each chamber.

Bonding Efficiency and Deformation of PLA Templates

After the pouring and molding of PDMS at 80 °C, a critical step is to peel the PDMS molds off of the PLA-attached glass substrate without destroying the templates, which depends on the delicate balance between the PLA–substrate bonding efficiency and the PLA module deformation under an optimal heating temperature of the glass pane and the time.

Therefore, a parameter (ΔS) is defined to calibrate the PLA module deformation:

Δ S = ( C 0 - C t ) / C 0 × 100 %

C = b / a

where b is the length of the template and a is the width of the peak structure as shown in Figure 1d . C₀ is the initial b/a ratio when no heating is applied, and C_t is the b/a ratio at a certain temperature after heating for t seconds.

As we observed, the commercial PLA printing materials started to melt at 150 °C. Figure 2 showed the variation tendency of ΔS when the temperature was increased from 140 to 220 °C and the heating time ranged from 0 s to 60 s. When the temperature of the glass pane was kept around 180 °C, the PLA templates of the design shape were well attached to the glass substrate within 30 s. When heating was stopped, the PLA templates were easily removed from the cold glass pane. However, at even higher temperatures or longer times, deformation of the PLA templates occurred, which obviously not only impacted the microscopic shape of the microfluidic channels but also destroyed the reliability and reusability of the PLA templates. The yellow curve in Figure 2 exhibits the optimal conditions to keep the PLA templates attached to the glass pane without deformation (ΔS < 5%). Therefore, in this study, the PDMS molds were peeled off from their templates as soon as the PLA templates were bonded to the glass at 180 °C for 25 s.

OPEN IN VIEWER

Figure 2.

Color distinction graph of the PLA module deformation. ΔS was calculated to describe the deformation degree of the PLA template on a glass slide at a heating temperature ranging from 140 to 220 °C and a time from 0 to 60 s. The region lower than the yellow curve shows that the PLA modules could not firmly adhere on the glass substrates. Therefore, the red dot was chosen as the optimal conditions in this study. The PDMS molds were peeled off from their templates as soon as the PLA templates on glass were heated at 180 °C for 25 s.

Viability of Tumor Cell and Drug Screening on Microfluidic Chips

Under the optimal conditions mentioned above, several 3D-printed PLA modules were prepared and assembled together to produce a complex template for concentration gradient generators ( Fig. 3a ) with open chambers as the outlets. The serpentine units were 0.75 cm in width and 1.75 cm in length. The diameter and height of the outlet chambers were each 0.5 cm. The SEM image shows that the diameter of the microchannel after PDMS replication was measured as 355 μm ( Fig. 3c ).

OPEN IN VIEWER

Figure 3.

(a) Top view of the PDMS multichannel mixer mold. (b) Concentration gradient generated from yellow and red dye solutions. Yellow and red flows were injected from two separate inlets. The flow was divided equally in each junction so that the yellow and red flows were mixed in the four outlets at four different ratios. When tumor cells were incubated in the four chambers where cisplatin and cell culture medium were automatically mixed into four ratios, the cell images before (d,f,h) and after (e,g,i) dosing in each chamber were exhibited. (c) SEM image of the PDMS channel produced with the resulting PLA templates; the diameter of the channel was 355 μm.

After plunging two separate streams of dyes at 20 μL/h for 3 h, Figure 3b shows an obvious progressively diluted effect in the microfluidic chip generating four mixing ratios at the outlet chambers. The further from the inlets, the lower the concentration of the dyes in the microchannel. A more diluted and mixed solution is attainable by constructing more linear and serpentine units. The gradient profile depends on several factors, such as the concentration of solutions, the design of the channel, and the flow rate. The side-by-side laminar flow is dominant inside a microchannel where mixing across the streamlines is achieved by molecular diffusion only and predicted accurately. ¹⁹ Concentration gradients automatically produced in microfluidic chips have great benefits in high-throughput drug screening. ²⁰

After successfully loading tumor cells in the open chambers, 0.1 μg/mL cisplatin in DMEM and cell culture medium was injected from the two inlets separately. During the 3 h injection, excess liquid leaking from each chamber was carefully removed with micropipettes. The drug screening microfluidic chip was then kept in an incubator at 37 °C for 24 h. The cell images from each chamber were taken and compared with the live cell number before dosing.

Taking cisplatin as an example, the drug screening capability of the developed microfluidic chip was evaluated by monitoring the cell growth before and after dosing. To have a better grasp, the pseudocolor operation was run by a computer-assisted program to distinguish live and dead cells. The tumor cells in the first chamber containing mostly cell culture medium grew commonly ( Fig. 3d,e ) after 1 day. The last chamber containing cisplatin of the highest concentration had a scant amount of live cells ( Fig. 3h,i ) after 24 h of incubation. Cell viability in the second chamber was in between the two chambers mentioned above due to the drug concentration gradient generated automatically in the designed microfluidic chip.

Besides the planar structure, the 3D-printed templates also easily enable the alignment of several layers and accomplish more complicated 3D structures. Unlike other molds developed by technologies such as photoresist, stereolithography, and computer numerical control machines, the 3D-printed modules are reusable and reassembled within 5 min, with an average cost of only $0.08 USD. The flexible 3D-printed templates create a modularized fabrication process and promote broader access to establish the integrated microfluidic systems with minimal requirements and improved efficiency.