Poster W10 - Pulse-Driven Microfluidic Pumps with Heart Rate-Sensitive and Constant Flow Rates

Introduction: The delivery of liquid pharmaceuticals across the skin currently still relies on powered infusion pumps and syringes, and these instruments of delivery pose many limitations on patients. For instance, all these delivery tools inject the drug subcutaneously with cannulas or needles, which causes injection site pain, infections, bleeding, and erythema. Infusion pumps that facilitate continuous delivery could be bulky and interfere with daily activity, whereas syringes can only deliver pharmaceuticals in discrete boluses. Considering these limitations, we have a critical need to develop slim, painless transdermal patch-pumps for continuous delivery. We developed featherweight, insect-mimetic, and pulse-driven patch-pumps and refined them over eight design generations using polydimethylsiloxane (PDMS). We have evaluated the flow rate performances of these devices at a wide range of physiologically relevant heart rates and blood pressures using a pulse simulator and identified device prototypes with ideal flow rate profiles.

Materials and

Methods: We fabricated three-layer, pulse-driven microfluidic patch pumps using 3D printing and soft lithography techniques. Briefly, we CAD designed master molds with negative channel patterns, which were originally based on design S2 of Chatterjee et al. (2021) and refined over eight generations. We cast molded the pump and actuation layers (~3 mm thick) against their corresponding master molds with PDMS. A thin, elastomeric PDMS membrane (~20 μm thick) was spin-coated onto a silanized silicon wafer and plasma bonded to the device layers. To evaluate the pulse-driven flow rate, we actuated the patch pumps under a custom-built, pressurized air-driven pulse simulator at heart rates from 30 to 150 bpm and blood pressures from 100 to 225 mmHg, and measured the flow rate using video tracking and microfluidic sensors (Elveflow).

Results, Conclusions, and Discussions:

Results: After eight generations of device design refinement, we developed devices that generated flow rates that increased, decreased, or stayed constant with increasing driving heart rate. The device in Fig. 1(a) has a sharply decreasing flow rate with an increasing heart rate. This is advantageous for the delivery of insulin, considering that an increase in heart rate during exercise corresponds to a decrease in need for insulin in diabetic patients. The three devices in Fig. 1(b), on the other hand, exhibit heart rate-independent flow rates at three different magnitudes (0.05, 0.5 and 2.5 μL/s), so that they can be used for different applications (insulin, patient-controlled analgesia, and chemotherapy). Recently, we also characterized the effect of the driving blood pressure on the flow rate generated by the devices and identified devices whose flow rate did not vary with increasing blood pressure (Fig. 1(c)). Despite this, there was still limited inter-device reproducibility (>100% variation) across replicates of an identical design (Fig. 1(d)).
Discussion and

Conclusions: Compared to currently available infusion pumps, we believe that our microfluidic patch-pumps are advantageous because they are featherweight, low-cost, and can be integrated with a painless, minimally invasive microneedle array. We have developed decreasing and constant flow rate devices with increasing heart rate to accommodate various transdermal drug delivery applications. We have recently developed a device with a flow rate independent of the actuating blood pressure to ensure that oscillating blood pressures will not significantly affect the delivery rate. Since infusion pumps are typically expected to have average flow rate variations of ±15%, we still need to make an improvement on the inter-device reproducibility by potentially changing our fabrication techniques or the material of our patch pumps. For future work, we also plan to perform a human subject study (already IRB-approved) to evaluate the performance of our devices and employ a genetic algorithm-driven approach to optimize device design for specific patients.

Acknowledgements (Optional): We thank Mrigank Dhingra, Julia Frederick, Aleksandra Grodzki, Zachary Vaday, Cayla Katz, Jessica Prisbe, Adaliah Dunya, and Vedant Shah for their contributions. This work was supported by the Virginia Tech Student Engineers' Council.