(J-375) Strain-driven power-free pumping via asymmetric microchannel structure for microfluidic wearable applications

Introduction:: Skin mountable strain sensors are studied for a wide range of applications from human-
machine interaction ¹ to physiological monitoring ² and to biomechanical applications ³. In recent
years, there has been considerable interest in the development of microfluidic flow control
mechanisms in skin mountable devices ^4,5 . Human movement actuated pumps are required for
executing complex functions on wearable devices without the requirement of electrical power
and components. Such pumps could expand the applications of wearable microfluidics to
management of physical rehabilitation of patients suffering from musculoskeletal conditions ⁴ and
to drug delivery ⁶ . 
        Microfluidic pumping often requires the additional force of a moving part ⁷ or an electrical
signal that triggers flow ⁸ . We propose a concept where the strain-actuated pump demonstrates
controlled liquid movement due to the dependence of hydraulic resistance on the asymmetric
deformation characteristics of high (and low) aspect ratio (AR) channel geometry. Figure 1a
shows such asymmetric channels on both sides of an actuator region. A close-up schematic of
this channel intersection is included in Figure 1b. The resistances of the two asymmetric
channels are equal (R _Hi = R _Low = R) initially. Under uniaxial perpendicular strain, the resistances
of each side will change asymmetrically (i.e., high AR decreases while low AR increases) as the
channel height and width changes. This asymmetric behavior during periodic strain will result in
a net liquid movement towards the low AR side, demonstrating a novel pumping
method.

Materials and Methods:: Device Fabrication
        The polydimethylsiloxane (PDMS) microfluidic pump is fabricated using a 3D printed (Formlabs), clear resin mold. A 10:1 ratio (A:B) of PDMS is spun onto the mold at 200 rpm, achieving a thickness of about 700 µm. A base layer of 20:1 ratio (A:B) PDMS is then spun on a silicon wafer at 140 rpm, resulting in a 500 µm thickness. The two layers are thermally bonded and filled with 1-butyl-3-methylimidazolium dicyanamide ([BMIM][N(CN)2]) containing blue dye for visibility. Finally, inlets are sealed with NOA 65 and UV cured, and then a layer of Loctite Epoxy is applied to strengthen the sealing. The overall dimensions of the chip were 30 x 30 x 1.2 mm³. Two different pumps were designed, one with a 500 : 100 µm pumping channel ratio (Figure 2), and one with a 500 : 200 µm ratio, and one control chip was designed with symmetrical pumping channels.

Characterization of Fluid Movement
        We characterized the fluid movement on the benchtop using a manual strain application module that allows for controlled amounts of strain on the sensor. The experimental process starts with setting the liquid interface at the desired position and allowing the liquid to settle for 5 minutes, and then applying the identical strain values ten times and allowing the liquid to settle for two minutes before making a measurement. This process was performed up to ten times, resulting in up to ten trials.

Results, Conclusions, and Discussions:: During an actuation cycle, when the strain is applied, the asymmetric change in hydraulic resistance results in greater negative liquid displacement from the high AR side as shown in Figure 1a. This negative liquid displacement is due to the volumetric expansion of our actuator region, which is well-studied in our group ^{2, 4}. When the strain is released, resistances return to initial values of R _Hi = R _Low , resulting in equal and positive liquid displacement in both channels as shown in Figure 1b. Due to the reliance of the operation principle on the asymmetric change of the hydraulic resistance, we have selected diodicity as a performance parameter. We define diodicity as the ratio of the two pumping channels' resistances under strain. Figure 1c shows theoretically calculated diodicity for two ARs; 5:1 and 5:2, demonstrating that as strain increases, diodicity increases. When comparing the 5:1 and 5:2 designs, the 5:1 design demonstrates greater diodicity especially for larger than 5% strain, suggesting that the 5:1 design will be more efficient. Figure 2 shows the image of a fabricated chip at three different stages of a test. Here, the movement from the high AR (right) side to the low AR (left) side is clearly visible.
        Figure 3a shows the experimentally measured liquid movement at the end of each trial and demonstrates greater efficiency of the 5:1 pump in comparison to the 5:2 pump, as it displaces larger liquid volume per trial. The difference in efficiency is qualitatively in agreement with the theoretical calculation in Figure 1c.
        Furthermore, symmetric control chip exhibits little to no movement despite the strain actuation cycle applied in the same manner as shown in Figure 3b, illustrating the significance of asymmetry of the pumping channels.
        Figure 4 shows the dependence of the pumping performance on the strain amplitude for the 5:1 design. Here we observe that an increase in strain results in increased liquid pumping efficiency.
        Overall, these experiments demonstrate that strain driven asymmetric microchannels can be utilized to pump liquids without requiring electrical power, which could be useful for a variety of wearable microfluidic applications.

Acknowledgements (Optional): : This project was supported by the Kuehler grant and NSF program 19-582, grant 2045087. The devices were fabricated in the Center for Nanostructures at Santa Clara University.

References (Optional): : 1. Jin, L., Li, Z., Liu, Z. et al. Flexible unimodal strain sensors for human motion detection and differentiation. npj Flex Electron 6, 74 (2022). DOI: https://doi.org/10.1038/s41528-022-00205-4
2.  Agaoglu, S., Diep, P., Martini, M., Samudhyatha, K. T., Baday, M., & Araci, I. E. (2018). Ultra-sensitive microfluidic wearable strain sensor for intraocular pressure monitoring. Lab on a Chip, 18(22), 3471-3483.
3. C. Votzke, U. Daalkhaijav, Y. Mengüç and M. L. Johnston, "Highly-Stretchable Biomechanical Strain Sensor using Printed Liquid Metal Paste," 2018 IEEE Biomedical Circuits and Systems Conference (BioCAS), Cleveland, OH, USA, 2018,
pp. 1-4, DOI: 10.1109/BIOCAS.2018.8584671.
4. H. Gasvoda, N. Cmager, R. Atlay, J.Y. Lee, I.E. Araci, Integration of capillaric strain sensors toward recognition of human movements, Royal Society of Chemistry, 2022, DOI: 10.1039/d2sd00201a
5. Hnin Yin Yin Nyein, Li-Chia Tai, Quynh Phuong Ngo, Minghan Chao, George B. Zhang, Wei Gao, Mallika Bariya, James Bullock, Hyungjin Kim, Hossain M. Fahad, and Ali Javey. A Wearable Microfluidic Sensing Patch for Dynamic Sweat Secretion Analysis. ACS Sensors 2018 3 (5), 944-952. DOI: 10.1021/acssensors.7b00961
6. Uguz I, Proctor CM, Curto VF, Pappa AM, Donahue MJ, Ferro M, Owens RM, Khodagholy D, Inal S, Malliaras GG. A Microfluidic Ion Pump for In Vivo Drug Delivery. Adv Mater. 2017 Jul;29(27). doi: 10.1002/adma.201701217. Epub 2017 May 15. PMID: 28503731.
7. H. J. Mamin, B. A. Gurney, D. R. Wilhoit, V. S. Speriosu; High sensitivity spin-valve strain sensor. Appl. Phys. Lett. 15 June 1998; 72 (24): 3220–3222. https://doi.org/10.1063/1.121555
8. PW. Yen, SC. Lin, YC. Huang, et al. A Low-Power CMOS Microfluidic Pump Based on Travelling-Wave Electroosmosis for Diluted Serum Pumping. Sci Rep 9, 14794 (2019). DOI: https://doi.org/10.1038/s41598-019-51464-7