I. INTRODUCTION

Section:

ChooseTop of pageABSTRACTI. INTRODUCTION <<II. MATERIALS AND METHODSIII. RESULTS AND DISCUSSI...IV. CONCLUSION AND FUTURE...SUPPLEMENTARY MATERIALREFERENCESMicrofluidics integrates microfabrication techniques with chemistry and biology and has been growing rapidly since the early 1990s.11. B. Weigl, G. Domingo, P. LaBarre, and J. Gerlach, Lab Chip 8(12), 1999–2014 (2008). https://doi.org/10.1039/b811314a Since then, the miniaturization of fluid handling and fluid analysis is a promising area in biological sciences and analytical chemistry resulting in lightweight and tiny microfluidic devices suitable for portable applications such as point-of-care (POC) diagnosis. Among their various components, a micropump is responsible for essential flow function to move the fluid from one place to another.Micropumps are mainly classified into active and passive pumping methods based on their external power requirements.22. L. Xu, A. Wang, X. Li, and K. W. Oh, Biomicrofluidics 14(3), 031503 (2020). https://doi.org/10.1063/5.0002169 Active micropumps include syringe pumps, diaphragm micropumps, magneto-hydrodynamic micropumps,33. S. Derakhshan and K. Yazdani, J. Mech. 32(1), 55–62 (2016). https://doi.org/10.1017/jmech.2015.39 and electroosmotic.44. M. K. Dehghan Manshadi, D. Khojasteh, M. Mohammadi, and R. Kamali, Int. J. Numer. Model.: Electron. Netw. Devices Fields 29(5), 845–858 (2016). https://doi.org/10.1002/jnm.2149 Diaphragm micropumps are classified as check-valve55. J. Kang and G. W. Auner, Sens. Actuators A 167(2), 512–516 (2011). https://doi.org/10.1016/j.sna.2011.01.012 and valveless (diffuser) micropumps.66. C. N. Nhu, L. L. Van, A. N. Ngoc, V. T. Dau, T. T. Bui, and T. C. Duc, Int. J. Nanotechnol. 15(11-12), 1010–1023 (2018). https://doi.org/10.1504/IJNT.2018.099938 However, all these active pumping devices face certain limitations in real world applications. These micropumps often raise the overall cost of the system while also making miniaturization challenging.77. T. M. Squires and S. R. Quake, Rev. Mod. Phys. 77(3), 977–1026 (2005). https://doi.org/10.1103/RevModPhys.77.977,88. S. K. Yoon, G. W. Fichtl, and P. J. A. Kenis, Lab Chip 6(12), 1516–1524 (2006). https://doi.org/10.1039/b609289f Complex circuitry of such pumps, along with their constant external power requirements, makes them unsuited for point-of-care applications. The substrates used in traditional microfluidic devices are still regarded to be expensive, notably for single-use devices. As the market for biodegradable on-chip diagnostic devices grows, there is an increasing demand to replace existing substrates with inexpensive, lightweight, and environmentally benign materials.As an alternative to active micropumps, passive pumping methods are proposed. Passive pumps do not require any external power source. These pumps include capillary,99. A. Olanrewaju, M. Beaugrand, M. Yafia, and D. Juncker, Lab Chip 18(16), 2323–2347 (2018). https://doi.org/10.1039/C8LC00458G surface tension,1010. G. M. Walker and D. J. Beebe, Lab Chip 2(3), 131–134 (2002). https://doi.org/10.1039/b204381e evaporation,1111. C. Nie, A. J. Frijns, R. Mandamparambil, and J. M. J. den Toonder, Biomed. Microdevices 17(2), 47 (2015). https://doi.org/10.1007/s10544-015-9948-7 and osmotic pressure-based pumps.1212. Y. Luo, J. Qin, and B. Lin, Front. Biosci. (Landmark Ed.) 14(10), 3913–3924 (2009). https://doi.org/10.2741/3500 Benefits of these devices include avoidance of an active control system, simple fabrication, easy fluidic manipulation, and no external power requirements. Their limitations are application specific and are usually one-time use devices.The use of paper as a substrate in these passive devices makes them unique among other microfluidic platforms. In everyday living, paper is the most extensively used, cheapest, and environment friendly material. It is lightweight, thin, hydrophilic, and flexible, making it ideal for a range of applications. The microstructure of the paper is made up of randomly interwoven cellulose fibers that provide abundant channels and pores at micro- and nanoscale.1313. B. Yao, J. Zhang, T. Kou, Y. Song, T. Liu, and Y. Li, Adv. Sci. 4(7), 1700107 (2017). https://doi.org/10.1002/advs.201700107,1414. Y. Zhang, L. Zhang, K. Cui, S. Ge, X. Cheng, M. Yan, J. Yu, and H. Liu, Adv. Mater. 30(51), 1801588 (2018). https://doi.org/10.1002/adma.201801588 Capillary force drives fluids within the paper porous structure, and therefore, no external pump is required to control the flow, allowing the device’s size and cost to be reduced to meet the real-world applications.The literature on paper-based microfluidic pumping has reported the capillary-based transport of liquid in rectangular, circular, triangular, rectangular combined with fan shape, and various other shapes of paper pumps.1515. X. Wang, J. A. Hagen, and I. Papautsky, Biomicrofluidics 7(1), 014107 (2013). https://doi.org/10.1063/1.4790819 Moreover, the fluid flow based on direction has been investigated in which a one-dimensional1616. S. Patari and P. S. Mahapatra, ACS Omega 5(36), 22931–22939 (2020). https://doi.org/10.1021/acsomega.0c02407,1717. R. R. Niedl and C. Beta, Lab Chip 15(11), 2452–2459 (2015). https://doi.org/10.1039/C5LC00276A or two-dimensional1818. A. T. Jafry, H. Lim, W.-K. Sung, and J. Lee, Microfluid. Nanofluidics 21(3), 57 (2017). https://doi.org/10.1007/s10404-017-1883-x fluid flow analysis was presented. Experimental models for the 3D fluid flow have been performed by stacking or folding (origami technique) different paper sheets one above the other.1919. J. Park and J.-K. Park, Sens. Actuators, B 246, 1049–1055 (2017). https://doi.org/10.1016/j.snb.2017.02.150Flow patterns that deviate from the Lucas–Washburn model can also be generated by altering the paper shape. Mendez et al. highlighted quasi-linear flow patterns on porous membranes using distinct fan-shaped outlets.2020. S. Mendez, E. M. Fenton, G. R. Gallegos, D. N. Petsev, S. S. Sibbett, H. A. Stone, Y. Zhang, and G. P. López, Langmuir 26(2), 1380–1385 (2010). https://doi.org/10.1021/la902470b Cummins et al. devised a modular “hydraulic battery” that attaches to the outlet of a microchannel and pumps fluid at accurate flow rates.2121. B. M. Cummins, R. Chinthapatla, B. Lenin, F. S. Ligler, and G. M. Walker, Technology 05(01), 21–30 (2017). https://doi.org/10.1142/S2339547817200011 The use of kirigami and quilling to create different fluid channels in paper is another recent development.2222. B. Gao, J. Chi, H. Liu, and Z. Gu, Sci. Rep. 7(1), 7255 (2017). https://doi.org/10.1038/s41598-017-07267-9Attempts to accelerate the flow of fluid in paper channels have involved moving the fluid outside of the paper, where it can flow faster.2323. S. Jahanshahi-Anbuhi, P. Chavan, C. Sicard, V. Leung, S. M. Z. Hossain, R. Pelton, J. D. Brennan, and C. D. M. Filipe, Lab Chip 12(23), 5079–5085 (2012). https://doi.org/10.1039/c2lc41005b Sandwiching the paper between polymer films is another method.2424. E. T. S. G. da Silva, M. Santhiago, F. R. de Souza, W. K. T. Coltro, and L. T. Kubota, Lab Chip 15(7), 1651–1655 (2015). https://doi.org/10.1039/C5LC00022J,2525. H. H. Cho, S. J. Kim, A. T. Jafry, B. Lee, J. H. Heo, S. Yoon, S. H. Jeong, S.-I. Kang, J. H. Lee, and J. Lee, Part. Part. Syst. Charact. 36(6), 1800483 (2019). https://doi.org/10.1002/ppsc.201800483 Alternatively, two paper strips can be sandwiched between a hollow layer of a double-sided tape to create two-ply channels in paper, which substantially lower flow resistance.2626. C. K. Camplisson, K. M. Schilling, W. L. Pedrotti, H. A. Stone, and A. W. Martinez, Lab Chip 15(23), 4461–4466 (2015). https://doi.org/10.1039/C5LC01115A,2727. R. B. Channon, M. P. Nguyen, A. G. Scorzelli, E. M. Henry, J. Volckens, D. S. Dandy, and C. S. Henry, Lab Chip 18(5), 793–802 (2018). https://doi.org/10.1039/C7LC01300K For faster-wicking channels, certain areas of the paper are removed to generate macro-capillaries for the fluid to flow.2828. C. Renault, X. Li, S. E. Fosdick, and R. M. Crooks, Anal. Chem. 85(16), 7976–7979 (2013). https://doi.org/10.1021/ac401786h One method to accelerate the fluid flow is to etch grooves into the paper channels.2929. D. L. Giokas, G. Z. Tsogas, and A. G. Vlessidis, Anal. Chem. 86(13), 6202–6207 (2014). https://doi.org/10.1021/ac501273v Fluid will bypass the paper and flow more rapidly owing to the grooves.3030. J.-H. Shin, G.-J. Lee, W. Kim, and S. Choi, Sens. Actuators, B 230, 380–387 (2016). https://doi.org/10.1016/j.snb.2016.02.085 The use of tape to seal the channels enhances wicking speeds even further.3131. B. Kalish, M. K. Tan, and H. Tsutsui, Micromachines 11(8), 773 (2020). https://doi.org/10.3390/mi11080773 Dosso et al. presented a self-powered 2D paper pump with a wide range of flow rates from 0.07 to 30 μl/min as well as volume of fluid (0.5 up to 150 μl) that can be imbibed.3232. F. Dal Dosso, T. Kokalj, J. Belotserkovsky, D. Spasic, and J. Lammertyn, Biomed. Microdevices 20(2), 44 (2018). https://doi.org/10.1007/s10544-018-0289-1

However, the proposed pumps offer limitations to the flow rate and its control, and only a minute sample volume of liquid can be transported. Most importantly, the increase in the flow rate during fluid imbibition is difficult to achieve in passive paper-based pumps. The reported 1D or 2D paper pumps in literature can only achieve decreasing or consistent flow rates. Therefore, exploration of devices to transport large sample volume of liquid with improved flow rates is still in progress. To achieve an equipment-free and fully functional microfluidic device, it is important to attain a variable flow rate as well as maintain a consistent flow rate on demand. We observed that previous pumping technologies lacked the variable flow rate in a simple and affordable passive pumping device. Due to the high surface to volume area of 2D as well as 3D stacked devices, evaporation problems may occur depending on the surrounding conditions. Furthermore, paper is a material that is prone to tearing and folding, and stability of these devices is a critical factor to consider.

Therefore, in this study, we have developed a three-dimensional paper-based pump to control and program the fluid flow in a microfluidic channel. Desired pumping flow rate can be achieved in the form of a continuous 3D structure unlike stacked pumps. These pumps may enhance the flow rate and its control and allow larger sample volumes of liquid to be absorbed. 3D cylindrical paper structure is compact; hence, it is not susceptible to tearing and folding. Moreover, evaporation problems may be reduced by using 3D pumps as the surface-to-volume area is decreased. Additionally, 3D paper-based microfluidic devices are more desirable than conventional 2D and lateral-flow devices since they allow more assays on small area. The novel 3D passive pumping will allow low-cost, high-performance point-of-care diagnostic devices with enhanced analytical capabilities and offer a promising platform for researchers working toward an efficient miniaturized platform for point-of-care (POC) diagnostic applications. Moreover, the proposed pump allows traditional microfluidic devices to meet the “ASSURED” criteria (i.e., affordable, sensitive, specific, user-friendly, rapid and robust, equipment-free, deliver to the users who need them) set by the World Health Organization.3333. A. Nilghaz, L. Guan, W. Tan, and W. Shen, ACS Sens. 1(12), 1382–1393 (2016). https://doi.org/10.1021/acssensors.6b00578 This will help them achieve commercialization of any miniaturized platform by enhancing user acceptance and deliverability of microfluidic systems.

II. MATERIALS AND METHODS

Section:

ChooseTop of pageABSTRACTI. INTRODUCTIONII. MATERIALS AND METHODS <<III. RESULTS AND DISCUSSI...IV. CONCLUSION AND FUTURE...SUPPLEMENTARY MATERIALREFERENCES

A. Materials

The materials purchased locally for fabricating the microfluidic device are transparent acrylic sheets (5 and 7.5 mm thickness), a double-sided pressure adhesive sheet (180 μm thickness), and a plastic covering (100 μm thickness). Whatman filter paper grade-1 was purchased from Merck Darmstadt, Germany. Distilled water was used as the pumping fluid in the microchannel. The fabrication process consists of two independent units. The microfluidic paper-based setup involved fabrication of a microchannel to provide path for fluid flow and development of a passive pumping device in the form of a paper-based pump. Both these units were integrated together to complete the experimental setup.

B. 3D paper pump

The 3D paper pump was developed in the form of a continuous porous structure. Whatman filter paper grade 1 was cut into smaller pieces and blended with water using Deuron Power Blender (GL-105 350 W). Circular, hollow molds (cartridges) of acrylic were cut with CO2 laser engraving/cutting machine (WR-4040 50 W). Molds were cut initially from the 7.5 mm acrylic sheet in hollow circles of inner diameter of 8, 10, 12, and 14 mm and then stacked one above the other vertically. The bottom sieved plate was made from the same acrylic sheet with cut out holes in it to let the water out. This was joined tightly by the paper tape at the bottom to form a continuous cylindrical column, hence forming a mold assembly (Figs. S1 and S2 in the supplementary material). The solution from the blender was poured into these cylindrical molds with the sieves on one end to filter out excess water. After drying of paper in an oven (Drawell DHG9030A) at 80 °C for 4 h, the molds were removed, and the 3D cylindrical paper-based pumps were ready. The 3D paper pumps with diameters of 8, 10, 12, and 14 mm were developed using the cylindrical molds. All the cylindrical paper pumps had the same height of 15 mm with geometric volume of 754, 1178, 1696, and 2310 mm3. Considering their porosity, the pumps of 0.5 porosity have void volumes of 377, 589, 848, and 1155 mm3, respectively, for 8–14 mm diameters, respectively. Similarly, the pumps of 0.7 porosity have void volumes of 528, 825, 1187, and 1617 mm3, respectively.

C. Variable porosity

Porosity, also known as void fraction, is a measure of a material's empty spaces and is expressed as the fraction of the pore volume to the total volume (dimensionless ratio). Its value ranges from 0 to 1. To change the porosity of the 3D pump, the mass of the paper pulp was weighed and increased during fabrication. The pulp was sieved with a slightly higher pressure from the top of the cylindrical mold to enhance the structural density of the fibers. Imbibition method was used to determine the porosity of the 3D paper pump by using SHIMADZU AX200 Weighing Balance. The mass of dry paper pump was initially determined. Then, the pump was inserted in a flat plate containing water at a small height. After the pump was completely filled with water, the saturated pump was weighed again. Imbibed fluid volume was determined, and as a result, the porosity of the sample was calculated by dividing this imbibed volume, which represented the void volume of the pump to the geometric total volume.

The sequence of steps followed for the development of 3D paper-based pumps is summarized in Fig. 1. Various 3D paper-based pumps of different sizes with two porosities of 0.5 and 0.7 were developed and tested to examine the fluid flow control in the microchannel based on pump diameter, programmability, and porosity.

D. Microfluidic channel

A microfluidic platform was fabricated by first stacking two 180 μm thick double-sided pressure adhesive sheets making a total thickness of 360 μm. The microchannel pattern was then cut with this sheet using a 50 W CO2 laser cutting/engraving machine. The serpentine microfluidic channel had a total length of 800 mm with cross-sectional dimensions of 2 × 0.36 mm2 (width × height). The patterned sheet was sandwiched between an acrylic sheet (5 mm thickness) and a plastic covering (100 μm thickness) to form a complete microfluidic platform. The acrylic sheet had inlet and outlet cylindrical holes cut into it for the fluid flow. An exploded view of the microfluidic platform integrated with a 3D paper-based pump is shown in Fig. 2. The inset shows the 3D flow regime from the outlet hole. This outlet hole is made to a smaller size so that the fluid entering into the cylindrical structure from the center of its base achieves the 3D flow effect in the form of a spherical meniscus up to its diameter. The fabricated microfluidic channel and the 3D paper-based passive pump are shown in Fig. 2(b).

E. Flow rate analysis

Initially, distilled water was injected into the microchannel by a syringe connected to the inlet port (Sec. ) and was filled all the way to the outlet port (Sec. ) as shown in Fig. 2. The syringe was then disconnected, and the 3D paper pump was placed at the outlet of the microchannel at which point the liquid came into contact with the passive pump and started wicking. The outlet of the channel is confined to a smaller cylindrical hole of 3 mm diameter compared to the diameter of the cylindrical pump as shown in subset of Fig. 2. This allowed the water to be absorbed due to capillary action from the central zone at the bottom of the cylindrical structure and spread out radially within the porous structure to flow in a 3D space. The experimental setup was designed in such a way to make sure that fluid imbibing into paper pumps flows in all directions. The meniscus was tracked from the inlet port to determine the fluid velocity and obtain the flow rate. The experiments were recorded using a smartphone camera with a reference scale placed close to the microchannel. The video files were fragmented into images based on appropriate time intervals and these images were analyzed in ImageJ software to determine the length covered by the fluid meniscus.

III. RESULTS AND DISCUSSION

Section:

ChooseTop of pageABSTRACTI. INTRODUCTIONII. MATERIALS AND METHODSIII. RESULTS AND DISCUSSI... <<IV. CONCLUSION AND FUTURE...SUPPLEMENTARY MATERIALREFERENCES

A. Flow rate control based on pump diameter

The 3D paper pumps with diameters of 8, 10, 12, and 14 mm for porosities of 0.5 and 0.7 were developed to investigate the effect of pump diameter on the fluid flow in a microchannel. Figure 3 shows the imbibition length and velocity curves by water in the microchannel with time for a pump porosity of 0.5. The results indicate that initially the flow rate was higher for all diameters, as depicted by the higher gradient of length against time. This is attributed to the three-dimensional flow within the cylindrical structure. As the liquid meniscus rises in the spherical radial direction, the surface area is continuously increasing, leading to a greater number of capillaries and, therefore, higher pumping force. The flow graphs also depict an increasing flow rate with increasing diameter, since a larger number of capillaries would allow more pumping power leading to higher flow velocities.

However, once the spherical liquid meniscus encounters the cylindrical wall, it will enter into a transition zone (3D to 1D flow regime). Here, the flow gradient started to decrease with smaller diameters showing the early sign of going into a transition zone and eventually following the Lucas–Washburn regime of the 1D flow in the cylindrical height direction. At this moment, the liquid front has lesser and constant available surface area compared to its start, which resulted in reduced flow velocity due to viscous effects as the flow progressed. Therefore, with an increase in the diameter of the pump, the flow rate increased, and the microfluidic channel length will be completed in a shorter time.

In a 2D flow, the higher number of capillaries and viscous effects nearly cancel out, causing a constant flow rate. However, for the 3D flow, an additional height factor increased the number of capillaries. Therefore, as shown in our results, it led to an increase in the flow rate not seen before in paper-based microfluidic devices. The maximum flow rate achieved with an 8 mm pump was 2.95 mm3/s and for 10 mm diameter was 3.73 mm3/s.

The available surface area for the fluid flow increased for 12 and 14 mm diameter pumps, leading to higher flow rates. These higher flow rates exhibited 3D fluid flow behavior as described earlier throughout the microchannel length, which caused flow rate vs time curves for both pumps to increase rapidly at the latter part of the curves. Maximum flow rates with 12 and 14 mm diameter pumps were 4.32 and 5.29 mm3/s, respectively.

Similarly, four cylindrical 3D paper pumps of 8, 10, 12, and 14 mm diameters with a porosity of 0.7 were developed and investigated for the effect of pump diameter and porosity on the fluid flow rate. As seen in Fig. 4, the flow results indicated similar trends to those of 0.5 porosity pumps. The flow rate for all pumps was initially higher, as illustrated by the higher gradient of length against time attributed to the 3D flow regime within the cylindrical paper structure. The maximum flow rate achieved with an 8 mm pump was 3.27 mm3/s and with a 10 mm diameter pump was 3.87 mm3/s. Interestingly, the 10 mm diameter had a nearly constant flow rate curve. The maximum flow rates achieved with 12 and 14 mm pumps were 5.86 and 6.97 mm3/s, respectively. Hence, by increasing the porosity of the paper pump, we observed an increase in the flow rate due to larger available pore volume for better fluid imbibition within the regime of the capillary-based flow.

For the total volume of fluid imbibed, the 8 mm diameter pump showed the range of 266 and 403 mm3 for 0.5 and 0.7 porosities, respectively. Larger the diameter, the more the amount of fluid volume pumped. Hence, the 14 mm diameter pump imbibed 565–567 mm3 for 0.5 and 0.7 porosities, respectively.

B. Programmable pump

The flow rate of the fluid moving in a microchannel can be adjusted by changing the geometry of a single, continuous porous structure (paper pump). A change in the geometry can result in the development of a single pump that can adjust the flow rate (program) for desired pumping requirements. Two cylindrical 3D programmable paper-based pumps of variable cross section were developed to investigate the effect of pump geometry on its pumping performance. These pumps had the same top and bottom regions with a narrow neck region of smaller cross section in the middle. The pump diameters were 10-5-10 mm and 14-8-14 mm in the top-middle-bottom regions, respectively. The middle neck region was expected to slow the flow rate of fluid in the microchannel based on its cross section area.

Figure 5(a) compares the length of the microchannel covered by water with respect to time for uniform and variable cross section pumps of fixed porosity of 0.7. A pump with a uniform cross section has a continuous porous structure of 10 mm in diameter, whereas a pump with a variable cross section has a smaller channel in the middle. The results showed that initially similar length-time gradients were observed for both pumps due to their same sized bottom region. However, soon after, both curves drifted apart. The middle narrow region of the programmable pump reduced fluid imbibition, given the smaller surface area for the liquid front. The fluid took longer to complete the microchannel length for the variable cross section programmable pump. Hence, by changing its geometry, the pumping characteristics can be designed.

For the uniform cross section, the flow rate of water through the microfluidic channel increased initially until it reached its maximum value, and then it became constant. The first portion of the curve is attributed to the 3D effect of the paper pump as discussed earlier. The second portion of the curve (a nearly constant value) may be attributed to the 2D fluid flow. This is due to the balancing of forces of capillary (two-dimensional surface area) and the viscous effects, causing a linear trend. The flow rate of a variable cross section pump can also be partitioned into two segments: one for the small initial time interval in which the flow rate increased following 3D fluid flow behavior (similar to a uniform cross section pump) and the other one is the large portion of the curve in which the flow rate was constantly decreasing following the traditional Lucas–Washburn model of the 1D fluid flow. This is since the middle neck region of the programmable pump caused the fluid to pass through a smaller cross-sectional area, hence limiting its 3D effect to only the 1D flow. With the lesser available surface area and higher viscous forces, it caused the fluid to move slowly.

In order to further investigate the effect of pump programmability, Fig. 6(a) compares the length of the microchannel covered by water with respect to time for uniform and variable cross section pumps of fixed porosity of 0.5. A pump with a uniform cross section was 14 mm in diameter, whereas a pump with a variable cross section was 14-8-14 mm in diameter. It was found that initially similar length-to-time gradients were observed for both pumps due to similar geometries. For the uniform cross section pump, a constantly increasing flow rate curve was observed resembling 3D fluid flow behavior.

Interestingly, the flow rate of the variable cross section pump showed a constant flow rate. Initially, for a short time interval, its flow rate increased following 3D fluid flow behavior (similar to a uniform cross section pump); however, it slowly decreased and remained nearly constant throughout. This is because the middle narrow region of the programmable pump caused the fluid to be imbibed slowly given the smaller surface area for the liquid front. This caused the fluid to move slowly. In the latter part of the curve, the flow rate started to increase once again. This happened due to the higher available surface area for the liquid front in the top region. Therefore, it is evident that by changing the geometry of the pump, the flow rate of the fluid moving in a microchannel can be modulated and programmed.

In comparison, a thread-based passive pump relies on the capillary flow and evaporation of liquid to allow a consistent, low volumetric flow rates (0.2–1 μl/min) for long duration tests such as 24 h for cell culture experiments.3434. L. C. Delon, A. Nilghaz, E. Cheah, C. Prestidge, and B. Thierry, Adv. Healthcare Mater. 9(11), 1901784 (2020). https://doi.org/10.1002/adhm.201901784 Compared to this, our 3D pump is designed for short duration, rapid testing where the test requires variable flow velocity (increasing, decreasing, or constant flow) depending upon how much exposure is needed to the reagent when flowing through certain channel locations. In addition to flow rate variation, the volume of the fluid imbibed is in the range of 150–400 μl/min depending upon pump size and working solely on the capillary flow with minimal evaporation phenomenon.

It was shown that the flow rate of a 14-8-14 mm pump is nearly constant in the middle, and then a small increase was observed in the latter part of the flow rate against the time curve. The same behavior cannot be seen for the 10-5-10 mm pump, however, in which the flow rate decreased throughout the curve. The reason behind it was that the available surface area for the 14-8-14 mm pump is larger than the 10-5-10 mm pump. Therefore, when the fluid reached the top segment of a 14-8-14 mm pump, part of its bottom and top segment boundaries was still dry, which provided a stronger capillary pull for the liquid front to flow at a faster rate. In a 10-5-10 mm pump, when the fluid reached the top segment, the bottom and middle segments are already saturated, and liquid already absorbed in the paper pumps reduced the capillary pull for the liquid front that is yet to be imbibed, and hence, a decreased capillary pull caused a decreasing flow rate.

C. Flow rate control based on pump porosity

Figure 7(a) shows the effect of pumping on the flow rate of water for an 8 mm diameter pump with different porosity values of 0.5 and 0.7. For the low porosity pump, the flow rate of the fluid moving through the microchannel decreased with respect to time, whereas the flow rate of the fluid for the higher porosity pump can be partitioned into two segments. For some initial time, the flow rate of water increased due to the 3D fluid flow in the pump but started to decrease afterward, resembling the 1D vertical flow. Available pore volume had increased for the higher porosity pump, causing high capillary force for water to flow at a higher rate. The maximum flow rates achieved with 0.5 and 0.7 porosities were 2.95 and 3.27 mm3/s, respectively. Hence, the fluid flow rate increased with an increase in the pump porosity.

In order to further investigate the effect of porosity on the pumping flow rate, similar experiments were repeated for 10, 12, and 14 mm diameters with porosities of 0.5 and 0.7. For the 10 mm pump with 0.7 porosity value, the flow rate remained nearly constant with respect to time, whereas for the low porosity pump, it increased due to the 3D effect and then decreased afterward. The available pore volume had increased for the higher porosity pump, causing a greater volume of fluid to be imbibed with higher velocity. For the lower porosity, the fluid covered the cylindrical diameter quickly and entered into the 1D flow regime, and hence, the decrease in flow velocity was observed. The maximum flow rates achieved with 0.5 and 0.7 porosities were 3.44 and 3.87 mm3/s.

For the 12 and 14 mm pumps in Figs. 7(c) and 7(d), we observed that as the available surface area had increased further for both, it led to higher flow rates. These higher flow rates exhibited 3D fluid flow behavior throughout the microchannel length. Moreover, the 0.7 porosity pump had a higher flow rate curve than the 0.5 porosity pump, as illustrated. The maximum flow rates achieved with 0.5 and 0.7 porosity pumps of 12 mm diameter were 4.32 and 5.86 mm3/s and for 14 mm diameter were 5.29 mm3/s (317.4 μl/min) and 6.97 mm3/s (418.2 μl/min), respectively. Hence, flow variation as well as the effect on the volume absorbed compared to the available surface area is evident from changing the pump porosity.

The inference of this discussion is that the flow rate of water flowing through the microfluidic serpentine channel increased with an increase in the diameter (geometry) and porosity of the paper pumps. Moreover, these pumps can be programmed to adjust the flow rate of the fluid moving in the microchannel during the pumping process. Hence, the proposed 3D pump overcomes the limitations of 1D or 2D paper pumps by achieving a variable as well as consistent flow rate on demand. More importantly, the same pump can also be used for imbibing small volume of fluid in the microchannel by simply further miniaturizing the microchannel and limiting the length of channel according to the specific protocols required for the bio-assay making it a viable choice for point-of-care devices.

IV. CONCLUSION AND FUTURE OUTLOOK

Section:

ChooseTop of pageABSTRACTI. INTRODUCTIONII. MATERIALS AND METHODSIII. RESULTS AND DISCUSSI...IV. CONCLUSION AND FUTURE... <<SUPPLEMENTARY MATERIALREFERENCES

A cylindrical 3D paper-based pump was developed for a microfluidic channel to experimentally investigate the effect of paper pumping on the flow rate. This effect was investigated based on the pump diameter (geometry), porosity, and programmability. The maximum flow rates achieved for a porosity of 0.5 with 8, 10, 12, and 14 mm pumps were 2.95, 3.44, 4.32, and 5.29 mm3/s, respectively. By changing the porosity to 0.7, the maximum flow rate increased for 8, 10, 12, and 14 mm pumps to 3.27, 3.87, 5.86, and 6.97 mm3/s, respectively. Most importantly, we observed an increase in flow velocity by a 3D fluid flow effect in a porous medium using a cylindrical structure. The limitations to the extent of the 3D flow regime were determined, and it was observed that this effect correlated with an increase in diameter.

Finally, two cylindrical 3D programmable paper-based pumps of variable cross section (10-5-10 and 14-8-14 mm in diameters) were investigated. It was concluded that the flow rate of fluid can be programmed to achieve variable flow velocity by altering the pump geometry. Since these pumps can simultaneously achieve increasing, decreasing, and constant flow velocity using only a simple geometry and different flow regimes, we believe they will be extremely beneficial for point-of-care technology development for resource limited settings as well as provide key components for various microfluidic platforms that otherwise require complex pumping setup for multiplex assays. For future works, the 3D pumps in multiple channels will be investigated to achieve stop and go as well as variation in flow rates to accomplish a complete passive pumping device.