High-Capacity Redox Polymer Electrodes: Applications in Molecular and Cellular Processing

Casting PEDOT:PSS Electrode Materials

High-capacity PEDOT:PSS electrodes were fabricated using two methods. In the first method (i.e., method 1), PEDOT:PSS emulsion (poly(3,4-ethylenedioxythiophene)-poly(styrenesulfonate), 1.3 weight percent dispersion in water, Sigma-Aldrich 483095) was cast across sheets of cellulose paper (Whatman 31 ETF). The cellulose paper was first cut into square sheets (6.35 cm sides), and the squares were mounted in an aluminum frame and positioned on a heating surface. The heating surface was maintained at 130 °C, and 2 mL of the PEDOT:PSS solution was evenly applied to the surface of the paper and left to dry for 15 min. The frame was then turned over, and another 2 mL of the polymer solution was cast on the opposite side and dried. This process was repeated three times for each side of the 31 ETF sheets. Once dried, the paper was covered in 2 mL of 80% EtOH solution and heated for an additional 30 min at 130 °C. The dried PEDOT:PSS-impregnated paper was then removed from the heating surface and frame and incubated in 0.25 M MgSO₄ for 1 h for additional cross-linking. This step was repeated for an additional hour to provide two separate cross-linking incubation periods. The PEDOT:PSS paper samples were then placed into an oven for final drying before further processing or packaging within the electrokinetic devices.

A second method (i.e., method 2) to produce high-capacity PEDOT:PSS electrodes used a slurry of cellulose fiber/PEDOT:PSS in place of the premade cellulose sheet. In this method, cellulose fibers were first created using a ball milling process, in which pieces of Whatman 31 ETF were placed in a 4 L glass jar. The jar was filled with distilled water and glass beads (diameter = 0.5 mm). The jar was sealed, placed on a lab roller, and mixed for 24 to 48 h, allowing the cellulose paper to break down into fibers. The fibers were then isolated through a fritted glass filter by vacuum filtration, washed in distilled water, and air dried. Following the wash process, 1.5 g of dry cellulose fibers was added to 15 g of Clevios PH 100 PEDOT:PSS solution and mixed for 15 min. A glass ring (diameter = 6.4 cm) was then placed on a clean surface, and 11.1 g of the fiber/PEDOT:PSS mixture was poured into the ring. The cast solution was incubated for 3 d to allow for water evaporation and formation of a thick PEDOT:PSS/cellulose film. The film was then removed from the glass ring and cross-linked using the MgSO₄ step described above. The final cross-linked films were incubated in an oven at 60 °C for 1 h and stored in sealed air-tight glass vessels before further processing or packaging within the electrokinetic devices.

Assembly of PEDOT:PSS/Cellulose Materials into Electrodes

Two different methods were used to assemble the PEDOT:PSS/cellulose materials into electrodes (each being used for specific device assemblies and electrokinetic applications discussed below). As shown in Figure 1A and B , the intact PEDOT:PSS cellulose sheets (method 1) were cut in 10 mm disks for lamination with thin copper contacts (shaped in a circle to provide electrical contact to the polymer disk around the entire perimeter). The adhesive Kapton film used to laminated these two parts together (i.e. the PEDOT:PSS/cellulose disc and metal contact) was compressed, leaving only the interior (polymer only) section exposed to liquid during use within electrokinetic devices. The copper contact enabled attachment to the leads of constant voltage or current power supplies, electronic conduction into the polymer disk, and finally ion conduction from the center of the polymer disk to the liquid within the device. Figure 1B shows an example of use in assembly of an electroosmotic pump, in which two PEDOT:PSS electrodes are placed on either side of an anodized alumina nanoporous membrane (Anodisc; GE Healthcare, Chicago, IL) and laminated into an EOP assembly using the adhesive Kapton film. The design is simple and allows integration of the pump into standard fluid flow cells (as shown below).

OPEN IN VIEWER

Figure 1.

PEDOT:PSS components and device fabrication methods. (A) An image showing a cellulose sheet within the aluminum frame over the hot plate after coating with the PEDOT:PSS solution and a punched electrode (before application of the copper connector). (B) A schematic of the single-layer electroosmotic pump (EOP) assembly, made by laminating a nanoporous alumina disk in between two opposing punched PEDOT:PSS electrodes. The image on the right shows the laminated pump and copper connectors used to make the electrical connection to the PEDOT:PSS disk electrodes. (C) An image of a disc of the cast PEDOT:PSS/cellulose fiber material after punching the 3 mm discs and placing them into two vials. The vial on the left contains the MgSO₄ cross-linking solution (and shows intact punched disc within the vial), whereas the vial on the right contains water (in which the discs without cross-linking have disintegrated). The image on the right shows the simple fluidic connector from Upchurch Scientific that was used to house the compacted PEDOT:PSS/cellulose material to make a threaded electrode. (D) A computer-aided design image of the rapid prototyped device that both accepted the threaded polymer electrodes and placed the tips of those electrodes (with exposed PEDOT:PSS material) in a central fluidic channel. The image on the right shows the device with both a cathode and anode threaded into the fluidic chamber and fluidic tubes attached to inject liquid into the fluidic structure.

The second assembly method for electrodes used commercially available threaded fluidic connectors (Upchurch flangeless, PEEK fittings) as a mold for the PEDOT:PSS/cellulose slurry described above (method 2). As shown in Figure 1C , the internal opening within the fluidic connector was filled with the dried PEDOT:PSS/cellulose slurry, and a small metal rod was used to press the polymer/cellulose material into the connector, resulting in a tightly packed polymer electrode core. A copper wire was placed within the internal opening of the fluidic connector at the distal end, which served as the contact to connect to the leads of the external power supplies. As shown in Figure 1D , the threaded fitting was assembled into electrokinetic devices by fabricating threaded ports into the polymer fluidic device. The surface of the screw-in electrodes was designed to reach the fluid channel within the device (with the electrodes wetting during priming and filling of the devices with water or buffer). Like the laminated electrode assemblies described above, the dry copper/PEDOT interface within the fitting served to enable electronic conduction into the polymer, whereas the liquid/PEDOT interface within the flow channel enabled ion conduction from electrode to liquid/buffer. These two assembly techniques described enabled electrode integration into many different types of fluid flow channels for use in bioprocess and bioassay applications, including rapid prototyped or injection-molded plastic flow cells, laminated microfluidic components, and static fluid reservoirs (application in these devices is discussed below).

Microfluidic Devices for Electrical Cell Lysis Experiments

Figure 2A shows the threaded polymer electrodes interfaced within a microfluidic flow cell designed for high-throughput cell lysis. The microfluidic channel was fabricated using a three-layer lamination-based technique. The base and top layer were constructed from a 0.005 inch thick polyimide sheet (Kapton HN; DuPont Electronic Technologies, Circleville, OH). The inner layer of Dupont Pyralux LF0111 Bondply was machined using a 355 nm ultraviolet wavelength laser (ESI 5200; Electro Scientific Industries, Portland, OR). The laser-cut profile included a microfluidic channel located between inlet and outlet ports for injection of mammalian cell suspensions ( Fig. 2B, C ). Two other ports were built into the microfluidic channel to allow access to two threaded polymer electrodes. The microchannel was machined to include a narrow fluidic constriction at the center of the device, in order to provide a high-resistance element that would isolate most of the voltage drop applied from electrode 1 to electrode 2 in an 80 µm segment of the channel. Cells that traverse the fluidic channel are guided into this area of high electric field strength for rapid electrical lysis. For the experiment presented herein, a rapid prototyped part (Protolabs, Maple Plain, MN) was designed with a needle inlet and outlet port for the introduction and collection of the cell suspensions and built-in threaded ports that accepted the PEDOT:PSS electrodes built into the Upchurch connectors (as described above). Prior to introducing the threaded electrodes, a 4 mm Ultracel 3Kda regenerated cellulose ultrafiltration disc (catalog No. PLBC04310; Millipore Corp., Billerica, MA) was placed into the electrode port. This disc provided an ion-permeable (but large-biomolecule-impermeable) barrier between the fluidic circuit and the electrode ports. A syringe pump was used to control flow rates through the electrical lysis device, where the flow rate controlled both the rate of cells entering the high electrical field constriction and the width of the electric field pulse experienced by the cell ( Fig. 2D ). Prior to use, the device was primed with low-conductivity buffer, consisting of 10 mM KH₂PO₄, 1 mM MgCl₂, and 250 mM sucrose. This priming step ensured the elimination of any bubbles or air pockets within both the microchannel and the threaded electrode ports and enabled passage of ionic current across the regenerated cellulose barrier.

OPEN IN VIEWER

Figure 2.

Electrical cell lysis device. (A) The top image shows the fluidic device used to perform electrical cell lysis with the threaded PEDOT:PSS electrodes shown above. As shown, the rapid prototyped part was altered to accept a thin-film microfluidic device (that was fabricated using laser machining polyimide films) and could be adhered to the bottom of the device using pressure-sensitive adhesive. The zoom-in images show the center of the fluidic channel through which the cells passed and the electric field concentrating constriction at the center of the channel. The threaded ports were designed so that the PEDOT:PSS anode and cathode were screwed down into holes machined into the top of the microfluidic channels. (B) A schematic image showing the electric field that the cell is exposed to as it enters the bottom fluidic inlet port (outside of the horizontal electric field produced by the threaded polymer electrodes), enters the fluidic constriction (which contains the highest electric field gradient due to its small geometry and higher fluidic/electrical resistance), and then exits the device for characterization. (C) As shown, the act of traversing the fluidic circuit provides an electric field pulse that may act to electroporate or lyse the cell (depending on the magnitude of the DC field applied across the electrodes).

Fluidic Chambers for Rapid DNA Extraction from Nucleic Acid Storage Cards

Figure 3 shows a second device that was built using the threaded polymer electrode interface. In this device, a rapid prototyped part was constructed that enabled loading of an FTA® nucleic acid storage card and placement of PEDOT:PSS electrodes directly above or below the card. Once loaded, the fluidic channel contained within the rapid prototyped device (containing the FTA card) was loaded with a DNA extraction buffer (Tris-HCl). An electric field was then applied across the card to provide an electrophoretic force on nucleic acid within the card. The regenerated cellulose barrier was again used between the fluidic channel and the electrode port, providing a surface for nucleic acid accumulation (without loss into the electrode material). After application of the electric field, additional DNA extraction buffer was pulsed through the device using a syringe pump, and the electrophoretic extracted DNA was collected. The collection times and efficiencies were compared with the standard extraction technique, which is to simply incubate the FTA in extraction buffer for 0.5 to 1 h and allow passive diffusion of the nucleic acids off of the storage card.

OPEN IN VIEWER

Figure 3.

Device for rapid elution of DNA. (A) A computer-aided design image of another fluidic device that accepts the threaded PEDOT:PSS electrodes (for introduction into a fluidic channel) and also accepts a plastic card that contains a 5 mm circle of FTA DNA storage paper at its center. Insertion of the FTA card into the device places the FTA paper sample in the center of the fluidic channel with the anode and cathode electrodes on either side of the FTA containing fluidic channel. (B) An image of an FTA card (from which the 5 mm sample is taken) that shows DNA within the cellulose fiber lattice. Typically, DNA is removed from the cellulose (after drying and storage) by incubating in an aqueous buffer for a half-hour or longer. The device shown enables electrical elution by providing an electrophoretic force to the DNA for a rapid elution from the card. The image shows a completed device with the FTA paper circle loaded into the center fluidic channel.

Alternating Layer Electroosmotic Pumps for Low-Voltage, High-Pressure Operation

Figure 4A depicts a new type of multilayered EOP that was fabricated using the PEDOT:PSS cellulose sheet electrodes described above. In a single-layer pump, a pair of the electrodes was used on either side of a nanoporous anodized alumina membrane (Whatman Anodisc, 25 mm, 0.02 mm pores; GE Healthcare), one as an anode and one as a cathode (as shown above). This type of electrode/Anodisc/electrode assembly was placed in a fluidic channel (integrated EOPs shown in Fig. 4B ) and primed with deionized water or buffer. As with previous EOPs, application of an electric field across the pores within the Anodisc produced an electrokinetic force on the charged ions adjacent to the charged alumina surface, which produced a fluid flow in the direction of the applied field. Within the nanopores, the viscosity of the liquid is sufficient to cause fluid movement, and fluid flow in each nanopore within the disc contributes to bulk movement of liquid within the larger fluidic channel. As the Anodisc is thin (100 µm), a large electric field is applied across the nanopores using a relatively low applied voltage between the PEDOT:PSS electrodes. However, in traditional EOPs, any back-pressure applied to the fluidic circuit would act in opposition to the electroosmotic flow and act to stall the pump. Previous studies have shown that increased pumping pressure can be realized by increasing the overall thickness of the nanoporous material (i.e., increasing the internal fluidic resistance of the pump using a ceramic frit or other material). However, this increase in thickness typically requires an increase in the applied electric field (as the spacing between the electrodes must be increased to accommodate the thicker nanoporous material).^30,31 For these reasons, pumps with practical pumping pressure typically require application of electric fields greater than 500 V, maintained by electrolytic reactions between two metal electrodes producing chemical and gaseous by-products.

OPEN IN VIEWER

Figure 4.

Electroosmotic pump (EOP) assembly. A schematic of the electrode and pumping layers within a novel multilayer EOP. In this concept, the pumping pressure is increased by adding additional nanoporous electroosmotic discs to the pump (compared with the single-layer pump shown in Fig. 1 ). As shown, the PEDOT:PSS electrodes are layered in between each layer of the electroosmotic disc; this allows continued use of low applied voltages (as the electrode spacing remains the same as the single-layer EOP). However, if only the Anodisc (positive zeta potential electroosmotic disc) was used, then electroosmotic flow would occur in reverse directions in each successive layer (i.e., always moving toward the anode). Instead, the team has used a negative zeta potential polyvinylidene fluoride (PVDF) membrane in each successive layer. Using this arrangement, the electric field direction is reversed in each layer of the pump, but the direction of flow remains the same (i.e., toward the cathode in the PVDF layer and toward the anode in the Anodisc layers). This novel configuration allows increased pumping pressure without the usual need to increase applied voltage on the pump (in order to pump across increasingly thick layers of electroosmotic material).

The new PEDOT:PSS cellulose sheet electrodes were used to redesign high-pressure pumps and enable operation at high back pressure with low operating voltages. Figure 4A shows the design of these novel EOPs in which the polymer electrodes were alternatively placed between the positive zeta charge Anodisc and the negative zeta charge polyvinylidene fluoride (PVDF) nanoporous membrane (durapore membranes; Millipore). In this configuration, ions within the Anodisc nanopores were driven toward the cathode (positioned at the output of that pumping layer), whereas ions within PVDF membrane nanopores were driven toward the anode during application of an electric field (using the alternating cathodic/anodic PEDOT:PSS electrodes). Thus, each successive layer pumped the fluid using positive- or negative-charged ions successively, and additional layers could be added to the pump by simply placing on another successive layer of Anodisc or PVDF membrane. With this novel alternating design, additional layers (and higher pumping pressure) do not require higher applied voltages (as each pump layer has its own anode and cathode placed directly in contact).

Flow Batteries: Precharging Polymer Electrodes in EOPs for Flow upon Discharge

In addition to use as anode and cathode materials in driving electrokinetic devices, PEDOT:PSS has been used in fabrication of type I capacitors (i.e., the anode and cathode are made from the same material). The capacitance formed in the PEDOT:PSS electrolyte interface can be understood by envisioning the PEDOT:PSS electrode as a degenerately doped p-type semiconductor (i.e., holes on the PEDOT are compensated by sulfonate anions within the PSS). When a voltage is applied across the material, de-doping occurs, in which cations are injected into the polymer electrode and holes are extracted through the metal contact on the other side of the polymer. The overall capacitance of any PEDOT:PSS electrode is also dependent on the overall volume of the polymer material within the electrode (as the electrode contains a finite number of sulfonate anion/hole pairs, and injected cations can replace these sites only up to the material’s hole density). Previously, these properties of PEDOT:PSS have been used to temporarily store electric charge and discharge in the form of an electrical current. Figure 5 shows a schematic diagram of a new use of this property in precharging an EOP and using the stored energy to discharge in the form of electroosmotic flow, thus producing a flow battery, in which fluid flow is generated in the absence of a traditional battery or applied electric field.

OPEN IN VIEWER

Figure 5.

Electroosmotic pump (EOP) actuation sequence. A schematic outlining the steps required to run a flow battery using a single-layer EOP. Step 1: The pump is first run in the forward direction using a battery that acts to provide electroosmotic flow (EOF) and in doing so charges the PEDOT:PSS electrodes. Step 2: The electrical contact to the battery is removed and the charge built in step 1 is stored in the electrode of the pump. Step 3: The anode and cathode within the pump are shorted using a metal wire (across the contacts of each electrode), which causes the electrochemical charge stored in the pump to discharge. This original charge on the electrodes acts as an applied electric field across the pump, and the discharge causes EOF in the opposite direction (of the original EOF during the initial charging step).

Integrated EOPs for Remote Operation of Fluidic Valves and Controls

Finally, the high-pressure, low-voltage EOPs were interfaced with simple fluidic check valves and used to generate the pressure required to actuate the control structures. Currently, a variety of external pressure sources are used to control fluidic valves and control structures in bioprocess and bioassay applications.^32,33 These control components are important to time and manage the injection of reagents into a variety of biological and chemical reactions. However, these external controls make the systems complex (requiring tubing connections and wires for every valve built into the system). Application of the new low-voltage, high–back-pressure EOPs to directly actuate valves within fluidic systems eliminates the need for tubing interfaces to external pressure sources and provides operating parameters that may be driven using common batteries (and low-voltage power sources). A one-way flow-check valve was designed by constructing a flexible flap above the flow port. The flap was laser machined from 0.001 inch thick Kapton HN polyimide, the port below was ablated in 0.005 inch Kapton HN material, and the two layers were bonded through a lamination process with the use of an intermediary adhesive layer of Pyralux LF0100. In use, fluid flow through the port structure would deflect the check valve flap and allow flow through the valve, whereas reverse flow would force the flap back in plane, sealing the port and restricting flow in the opposite direction.

Cell Culture, Quantification, and Analysis

CHO cells were cultured using standard methods and expanded to achieve cell densities of 1 million cells/mL prior to use in the electrical lysis experiments. Cell density was measured using a Nucleocounter NC-100 automated cell counter (Chemometic, Allerod, Denmark), and electrical lysis was measured using the viability standard provided in the commercial system (and measuring the percentage of viable cells before and after flow through the electrical lysis device). Lysis efficiency was reported as the percentage of total viable cells that were lysed into exposed nuclei.

DNA Quantification and Analysis

Commercially available genomic DNA (Thermo Fisher Scientific, Waltham, MA) was applied to FTA cards (GE Healthcare) at known concentrations for all DNA extraction studies. The quantity of DNA added to the card and then extracted using the electrical extraction procedure was measured using a picogreen DNA quantification kit (Thermo Fisher Scientific) and a standard multiwell spectrophotometer. DNA extraction yield was reported as the mass of DNA applied onto the FTA cards versus that which was recovered in the eluent after the electrical extraction procedure described above.

Flow Measurement and Pump Characterization

In all EOP and pump experiments, a flow sensor (optical flow sensor; Sensirion, Westlake Village, CA) was attached to the outlet of the EOP and filled during the priming procedure for the pump described above. After fluid priming, the EOP sensor system was allowed 10 min to ensure pressure equilibration across the system, and the sensor was then turned on to monitor the fluid flow produced by applying a voltage to the EOP. The sensor was calibrated (per the manufacturer’s instructions) to measure fluid flow between 0 and 500 µL/min. The pumping pressure of the EOP assemblies was measured by placing a positive-pressure source at the output of the pump (i.e., a measured height of water head) to control the pressure-resisting fluid movement in the EOP outlet tube. The pump stall pressure was measured as the applied back pressure that returned the EOP to zero flow (using a ruler to measure head height), despite an actively applied electric field.

Applications in High-Throughput Cell Electrolysis

Figure 6A depicts the results of operating the electrical lysis device across a range of operating voltages, compared with a standard chemical lysis procedure. In each experiment, 400 µL of sample was pumped through the device at 200 µL/min at a cell density of 1.5 × 10⁶ cells/mL. The results show that the percentage of lysed cells increases with applied voltage and that the PEDOT:PSS electrode assembly maintains adequate capacity to drive the DC field for operation times that enable passage of millions of cells through the device. In addition to increasing applied voltage to achieve higher lysis efficiency, the high flow speeds of the system can be leveraged (as shown in Fig. 6B), and the cell suspension can simply be recirculated through the device multiple times to increase lysis efficiency. The figure shows that operating the device at 200 V provides a lysis efficiency of only 30% with one pass, but this efficiency increases to 50% and 80% after passing through the device 5 or 10 times, respectively. After 10 passes through the device, the cell lysis efficiency matches that of standard chemical lysis techniques. Furthermore, Figure 6C shows that the DNA collected after electrical lysis is compatible with standard analytical techniques, as both DNA from the device and a control gDNA sample amplified an expected 1.1 Kbp amplicon using a standard DNA purification kit.

OPEN IN VIEWER

Figure 6.

Electrical cell lysis. (A) Data showing the percentage of cells that are lysed within the electrical lysis device (shown in Fig. 2 ) at different field strengths applied to the microfluidic channel (the field strength shown is that calculated for the fluidic constriction) compared with standard chemical lysis. (B) Data showing that simply increasing the number of times the cells passed across the fluidic constriction increased the lysis efficiency within the device (without requiring an increase in field strength). Lysis efficiency approaches that of standard chemical lysis after 10 passes across the constriction. (C) Gel electrophoresis data showing that DNA collected from cells lysed in the device is of similar size and quantity (compared with human gDNA run through the same DNA purification kit) as a control sample.

Applications in Rapid DNA Extraction from Paper Storage Cards

Figure 7 depicts the results of operating the DNA extraction device on FTA cards that were preloaded with 1 µg of genomic DNA and fully dried (as per the manufacturer’s suggestion for storing DNA on the card). Electrical extraction was performed by applying 25 V across the FTA card for 60 s, starting flow at 100 µL/min (for a total of 100 µL), and collecting the eluate in a centrifuge tube. During fluid flow, two electrical pulses were applied in the reverse direction (–25 V), to drive DNA off the regenerated cellulose barrier between the fluid compartment and the electrode port. Figure 7B shows that an equivalent amount of DNA was recovered from the storage card using the 2 min electrical lysis procedure and the standard 1 h buffer incubation process recommended by the manufacturer, ²⁹ providing a new rapid method of extraction for downstream processing. Figure 7C quantifies this effect by running extraction buffer through the device with and without application of the electric field. The regenerated cellulose barrier between the fluid and electrode chambers in these experiments underwent an additional coating procedure using a Nafion solution and then drying. Downstream DNA amplification was used to evaluate the extent of DNA damage due to oxidation caused by contact with the electrodes. Figure 7D shows that when run at field strengths of 500 V/m, the use of standard platinum electrodes in the system (to replace the treated PEDOT:PSS/cellulose electrodes) resulted in significant oxidative damage to the extracted DNA and PCR failure, whereas use of the PEDOT:PSS electrode resulted in quality extracted DNA. However, this DNA quality was dependent of the PEDOT:PSS electrode undergoing the Mg cross-linking and Nafion coating treatment discussed above. The figure shows that while operating at these high electric field strengths, only the cross-linked and coated PEDOT:PSS electrodes enable postextraction DNA amplification, but at the lower operation voltage (100 V/m), both the platinum and cross-linked PEDOT:PSS electrodes enable downstream DNA amplification.

OPEN IN VIEWER

Figure 7.

Electrical elution of DNA. (A) A schematic of the steps involved with electrical elution of DNA from the FTA card within the device. (B) A comparison of the amount of DNA collected from the 5 mm FTA punch after electrical extraction in the device versus a standard 1 h incubation in buffer. (C) The amount of DNA extracted from each 20 µL sample of buffer run through the fluidic chamber with (red) or without (blue) a previous electrical elution step. The data shown in (B) are the total of DNA collected in all 160 µL of sample run through the device. These data show that the majority of the DNA is removed from the device within the first 100 µL. (D) Gel electrophoresis data showing that at low voltages (i.e., 25 V), positive PCR amplicons from DNA are collected from the device using platinum, cross-linked (x-link), and Nafion-coated PEDOT:PSS electrodes, cross-linked (but not Nafion coated) PEDOT:PSS electrodes, but at high applied voltages (200 V). Positive PCR amplicons are collected only when using the cross-linked and Nafion-coated PEDOT:PSS electrodes.

Applications in Low-Voltage, High-Pressure EOPs

Figure 8A shows that the flow rates from single-layer EOPs (assembled using one 8 mm Anodisc and a pair of PEDOT:PSS cellulose sheet electrodes) were adequately measured using the flow sensors previously described and that flow was rapidly established upon applying a voltage to the pump assembly. Figure 8B demonstrates that use of the PEDOT:PSS electrodes enabled flow rates greater than 100 µL/min at applied voltages of less than 50 V and maintenance of a linear applied voltage versus flow curve over the entire measured range. Figure 8C shows that replacing the PEDOT:PSS electrodes with a platinum alternative (20 µm pore platinum mesh) resulted in much lower flow rates at a given applied voltage and a nonlinear relationship between voltage and flow at voltage greater than 30 V. Figure 8D shows that the addition of a second/alternating EOP layer (using the PVDF membrane and an additional PEDOT:PSS electrode as described above) doubles the pumping pressure of the EOP, and building the pump with four alternating layers provides the expected 4× increase in pumping pressure. The pumping pressure was measured by running the EOP into a fluid channel connected to a water reservoir and measuring the output of the pump against a known head of water (the zero-flow position is equivalent to the output pressure for each pump).

OPEN IN VIEWER

Figure 8.

Flow rates of electroosmotic pumps (EOPs). (A) Flow rates obtained from a single-layer EOP at applied voltages of 1 to 5 V during a pumping time of ~45 s. (B) A plot of the maximum voltage obtained when driving the EOP at voltages from 5 to 50 V. (C) Flow rates obtained from a single-layer EOP built with the same dimensions (as B) but assembled using platinum mesh electrodes. The flow rates obtained were lower for each applied voltage (compared with the PEDOT:PSS EOPs), and the relationship between applied voltage and flow rate became nonlinear at ~40 V. (D) A plot showing flow rate versus applied back-pressure for one-, two-, and four-layer EOPs (showing increasing stalling pressure for the thicker EOPs).

Applications in Flow Batteries: Electroosmotic Flow through Electrochemical Discharge

Figure 9A shows the flow generated by first charging a single-layer Anodisc EOP at 10 V for 10 min versus 20 s, removing the power source, waiting for 10 min with the electrode contacts in an electrically open state, and then shorting the anode and cathode contacts together to allow electrochemical discharge. As shown, the energy stored within the PEDOT:PSS electrodes was capable of driving fluid flow within the EOP for more than 50 min (after the 10 min charge) with a peak flow rate of ~5 µL/min and an average flow rate of 0.5 µL/min. However, the shorter 30 s charge enabled fluid flow for a much shorter time (less than 5 min). To further demonstrate that the measured fluid flow (after shorting the anode and cathode) is due to electrochemical discharge through the EOP structure, the diameter of the electrode/Anodisc pump was changed from 10 mm to 5 mm, and the peak flow rate was measured for each size. As shown in Figure 9B , the flow rate dropped for the smaller pump, demonstrating that the lower capacity/lower surface area electrode provided less of an electrochemical potential for electroosmotic flow. Finally, the ability of the EOP to function under electrochemical discharge should be dependent on the electrical path (i.e., short) between the precharged anode and cathode. To test this, the electrical connection between the anode and cathode of a precharged (10 min at 10 V) EOP was broken at the ~17 min time point during discharge. As shown in Figure 9C , upon opening the electrical circuit, the fluid movement is discontinued, and reclosing the circuit (i.e., shorting the anode and cathode leads) results in reestablishment of electroosmotic flow, thus demonstrating that the EOP is being primarily driven by discharge from the precharged PEDOT:PSS electrodes.

OPEN IN VIEWER

Figure 9.

Characterization of electroosmotic pump (EOP) flow rates. (A) Measured discharge flow from single-layer flow batteries after charging for 10 min versus 30 s at 10 V; a longer initial charging period corresponded with a longer duration of fluid flow. (B) Peak flow rate was measured for single-layer charged EOPs (charge for 10 min at 10 V) in pumps with 5 or 10 mm diameter EOP surfaces. Peak flow rate was observed to correlate positively with pump diameter/surface area. (C) Instantaneous flow rate during discharge of charged single-layer EOP after opening of the circuit (flow subsides but is restored upon subsequent reestablishment of cathode and anode connection).

Applications in Low-Voltage, Small-Footprint Actuation of Microfluidic Valves

Figure 10 shows the fluid flow output from a simple fluid channel containing the check valve shown. It was assembled and operated using the flow output from an alternating five-layer EOP. The EOP output pressure was monitored without the check valve and achieved output pressure greater than 350 cmH₂O (i.e., 5 PSI). The check valve was then placed in series with the EOP, and the output pressure required to close the check valve was monitored by measuring the fluid flow through the check valve in both the open and closed flow direction (using the flow sensor described above). The simplest designed check valve (horizontal flap valve structure) was completely closed at pressures exceeding 4 PSI, demonstrating that the low-voltage, high-pressure EOPs can be applied for fluidic valving and control in the absence of external pressure or high-voltage power supplies.

OPEN IN VIEWER

Figure 10.

Design of fluidic check valve and performance evaluation. (A) Top-down image of the one-way check valve. The top flap, seen as a U-shaped structure here, was not bonded to the layer below it. This allowed for the flap to deflect under positive pressure from the port below. Reverse flow was impeded, as the flap would reseat on the port, sealing it off. The top flap and bottom port layers were constructed from laser-profiled 1 mil thick Kapton HN. These layers were bonded with use of Pyralux LF0100 acrylic adhesive. (B) An evaluation of the performance of the check valve was conducted to determine the optimal pressure range of device operation. Forward and backward flow rates were determined at specific test pressures. Although some back flow is observed at low pressures, pressures higher than 3 PSI result in less than 0.5 µL/min of back flow, thus defining the operational pressure range of the device.