Technical Note pubs.acs.org/ac
Redox-Magnetohydrodynamics, Flat Flow Profile-Guided Enzyme Assay Detection: Toward Multiple, Parallel Analyses Vishal Sahore Department of Chemistry and Biochemistry, University of Arkansas, Fayetteville, Arkansas 72701, United States Microelectronics and Photonics Graduate Program, University of Arkansas, Fayetteville, Arkansas 72701, United States
Ingrid Fritsch* Department of Chemistry and Biochemistry, University of Arkansas, Fayetteville, Arkansas 72701, United States S Supporting Information *
ABSTRACT: A proof-of-concept superparamagnetic microbead-enzyme complex was integrated with microfluidics pumped by redox-magnetohydrodynamics (MHD) to take advantage of the magnet (0.56 T) beneath the chip and the uniform flat flow profile, as a first step toward developing multiple, parallel chemical analyses on a chip without the need for independent channels. The superparamagnetic beads were derivatized with alkaline phosphatase (a common enzyme label for biochemical assays) and magnetically immobilized at three different locations on the chip with one directly on the path to the detector and the other two locations adjacent to, but off the path, by a distance >5 times the detector diameter. Electroactive p-aminophenol, enzymatically generated at the bead-enzyme complex from its electroinactive precursor p-aminophenyl phosphate in a solution containing a redox species [Ru(NH3)6]3+/2+ for pumping and Tris buffer, was transported by redox-MHD and detected with square wave voltammetry at a 312 μm diameter gold microdisk stationed 2 mm downstream from the beadcomplex on the flow path. Oppositely biased pumping electrodes, consisting of 2.5 cm long gold bands and separated by 5.6 mm, flanked the active flow region containing the bead-enzyme complex and detection site. The signal from adjacent paths was only 20% of that for the direct path and ≤8% when pumping electrodes were inactive.
M
MHD have been achieved previously at the scales suitable for lab-on-a-chip applications but for dimensions that can be much greater than those suitable for electroosmotic flow. To broaden the scope of the redox-MHD flat flow profile toward multiple, parallel analysis, the application to magnetic bead assays is one approach that can utilize the inherent presence of a magnetic field required for redox-MHD pumping to immobilize the capture element. Magnetic bead assays have been integrated with microfluidics systems before using flow cytometry, syringe pump, and electro wetting techniques.9−12 In combination with redox-MHD pumping, assay devices would avoid the requirement of microchannels (which can become clogged with microbeads) and external mechanical parts (such as valves) to generate and direct fluid flow. MHD is governed by the magnetic part of the Lorentz force equation, FB = j × B, with its direction given by the right-hand rule. Here, j is the ionic current density of a fluid element (A/ m2), B (Tesla) is the magnetic flux density, and FB (N/m3) is the MHD body force acting on the fluid element.13−15 In the
icrofluidics from redox-magnetohydrodynamics (MHD) is capable of performing multiple, parallel chemical analyses on a single chip without requiring independent channels. This was demonstrated here in a proof-of-concept study by combining the flat flow profile generated by redoxMHD pumping with the selective placement of a superparamagnetic microbead-enzyme complex in a small solution volume on a silicon chip that contained patterned electrodes and was placed upon a permanent magnet. The ability to measure multiple analytes on a single lab-on-a-chip device is highly desirable because it decreases total analysis time, reduces operational cost, and improves overall device efficiency. The ability to perform simultaneous point-of-care diagnostics on one chip has become an important goal for chemical analysis devices.1−5 Multiple, parallel analysis on a single platform can benefit from undistorted movement of small fluid volumes. A flat flow profile is one way to achieve this.6,7 This microfluidic feature can be attained using electroosmotic flow, but it requires the use of microchannels as well as the application of high voltages, which limit its design options and create problems associated with heat generation, bubble formation, and electrode degradation.8 Uniform, horizontal flat flow profiles using the relatively new pumping method of redox© 2014 American Chemical Society
Received: June 1, 2014 Accepted: August 29, 2014 Published: August 29, 2014 9405
dx.doi.org/10.1021/ac502014t | Anal. Chem. 2014, 86, 9405−9411
Analytical Chemistry
Technical Note
Figure 1. (a) Top-down view of fluid flow and positions P1, P2, and P3 where the bead-enzyme complex was immobilized using a NdFeB magnet beneath the chip. Due to the uniform flow caused by redox-MHD, electroactive species that are enzymatically generated at position P2 will reach the detector, but those produced at positions P1 and P3 bypass the detector along its sides. (b) Lengthwise side view of the enzymatic conversion of PAPP to PAPR by AP-derivatized superparamagnetic beads at position P2. Here, a single bead symbol represents a collection of beads, and it is not to-scale. PAPR species are transported to the detecting electrode by redox-MHD pumping. (c) Crosswise side view illustrating electrochemical conversion and detection of different species at the active electrodes. (For the sake of clarity, inactive electrodes on the chip are omitted.)
■
EXPERIMENTAL SECTION Chemicals and Materials. All chemicals were reagent grade and used as received. Aqueous solutions were prepared using high purity deionized water purchased from Ricca Chemical Company, Arlington, TX. Hexaammineruthenium(II) chloride, sodium azide, and palladium (Pd) catalyst (10 wt % on activated carbon) were purchased from Sigma-Aldrich, St. Louis, MO. The as-received hexaammineruthenium(II) chloride is already about 50% converted to the oxidized form. (Therefore, a 10 mM solution made from this product (using the molecular weight for hexaammineruthenium(II) chloride), resulted in a solution containing approximately 5 mM in [Ru(NH3)6]2+ and 5 mM in [Ru(NH3)6]3+, which we represent as 10 mM [Ru(NH 3 ) 6 ] 2+/3+ .) Tris(hydroxymethyl)aminomethane and magnesium chloride were obtained from J.T. Baker, Phillipsburg, NJ. 4-Nitrophenyl phosphate disodium salt hexahydrate was obtained from Alfa Aeser, Wardhill, MA. Dynabead M-280 Streptavidin for assay immobilization was purchased from Life Technologies, Grand Island, NY, and biotinylated alkaline phosphatase (AP) enzyme was obtained from Vector Laboratories Inc., Burlingame, CA. Polystyrene latex microbeads (10 μm diameter, 2.5% wt dispersion in water) for tracking fluid flow were obtained from Alfa Aeser, Wardhill, MA. A 2 in. long × 2 in. diameter 0.56 T NdFeB permanent magnet was purchased from Amazing Magnets, Irvine, CA. Sylgard 184 silicon elastomer base, Sylgard 184 silicon elastomer curing agent, and the OS-30 solvent (Ellsworth Adhesives, Milwaukee, WI) were used to fabricate the poly(dimethylsiloxane), PDMS, gasket.19 Precleaned microscope glass slides (1.5 × 1.0 × 0.1 in.3) were purchased from VWR. Ultrapure argon gas was obtained from Airgas, Fayetteville, AR. All the chemicals and materials used to fabricate electrode chips were described previously.7
absence of easily available oxidized and reduced species, traditional MHD micropumps require the application of a high voltage, thus creating problems of bubble generation as well as electrode degradation.16−18 However, by using redox species that can sustain faradaic current at low overpotentials, problems associated with traditional MHD in electrolyte solution can be resolved, and this approach is coined “redoxMHD”. In addition, redox-MHD generates a channel-less fluid flow with a simple device that can be programmed by the location, magnitude, and direction of magnetic fields and the careful control of ionic current density through the fluidic cell by activating selected electrode shapes, their positions in threedimensional space, and currents.19−23 Here, two advantageous features of redox-MHD are harnessed, the flat flow profile and the presence of the magnet to immobilize magnetic beads, to demonstrate the capability of redox-MHD for use in multiplex assays and detection on a small scale. The MHD pumping force was generated by applying a constant potential between microband electrodes in contact with a solution containing [Ru(NH3)6]2+/3+ in Tris buffer and in the presence of a NdFeB permanent magnet placed beneath the chip. The electroactive species p-aminophenol (PAPR), generated by the reaction between paminophenyl phosphate (PAPP) and the alkaline phosphatase-derivatized superparamagnetic microbead complex located between the pumping electrodes, was detected at a disk microelectrode lying in the horizontal flow path of the PAPR. Studies were also performed by immobilizing the microbeadenzyme complex at locations off the horizontal path of the detector. A comparison of current from arrival of p-aminophenol at the detector downstream from the different enzyme locations on the chip was made to highlight the importance of the flat flow profile. 9406
dx.doi.org/10.1021/ac502014t | Anal. Chem. 2014, 86, 9405−9411
Analytical Chemistry
Technical Note
microbeads with PIV (Dantec Dynamics) software.7,20 Data analysis began 5 s after the application of a potential step. The existence of a flat flow profile determined in the presence of the superparamagnetic beads is presented in Figure S-4 in the Supporting Information. Preparation of the Superparamagnetic Microbead Assay Complex. The AP-derivatized superparamagnetic microbead complex was made off-chip by reacting biotinylated AP, a common enzyme label for biochemical assays, with streptavidin coated superparamagnetic beads. An 80 μL solution of as-received streptavidin coated superparamagnetic beads was added to 100 μL of the TTL buffer in an Eppendorf tube (VWR International), followed by a wash and decant step in the presence of NdFeB permanent magnet. Two volumes of 100 μL of Tris buffer (pH = 8) were added and removed in separate steps to wash the beads, followed by the addition of 20 μL of Tris (pH = 8), in which the superparamagnetic beads were resuspended. An 8 μL solution containing 7.14 μM biotinylated alkaline phosphatase was diluted with 12 μL of Tris buffer and mixed by a gentle shaking in another Eppendorf tube. This solution was added to and mixed with the solution containing the superparamagnetic beads under gentle vortexing (Vortex Gene-2, Scientific Industries) for 15 min, to form the AP-derivatized superparamagnetic microbead complex. A 100 μL solution of Tris buffer (pH = 8) was added and subtracted three times (while pulling the superparamagnetic beads to the side with an external magnet during this latter step) to wash the unbound AP, thus leaving a 40 μL solution containing microbead-enzyme complex suspended in Tris buffer. Assay Studies Involving the AP-Modified Superparamagnetic Microbeads. A 5 μL solution of the APmodified superparamagnetic beads was used for a single experiment and was dispensed using a micropipette onto an electrode chip which had been placed over a 0.56 T NdFeB permanent magnet and whose electrodes were connected to the bipotentiostat as described above. Three different immobilization locations were chosen across the 5.6 mm electrode gap, with one lying directly on the horizontal path of the detector, whereas the other two were not. See Figure 1a. A 350 μL solution of 10 mM [Ru(NH3)6]2+/3+ and 10 mM PAPP in 0.1 M Tris buffer was dispensed onto the electrode chip with a micropipette and was contained by the PDMS sidewalls and enclosed by a glass microscope slide on top. A concentration of 4 mM PAPP is typical for small volume electrochemical immunoassays involving alkaline phosphatase as an enzyme label.25−27 It was increased here to 10 mM PAPP to ensure that the PAPR signal would be measurable above the background of the pumping species and to account for diffusional dilution. A prior study involving redox-MHD used 8.3 mM PAPP for similar reasons.21 (Note that, to avoid complicating the procedure, fluid flow measurements were not simultaneously made during the assay studies, and therefore, the fluid-tracking microbeads were not added to the solution.) The on-chip solution was then allowed to react with the APmodified superparamagnetic beads for 60 s before a potential step of 0.3 V was applied at the pumping electrodes. The reaction between the AP enzyme and its substrate PAPP generates the electroactive species PAPR. The MHD force produced by activating the pumping electrodes in the presence of the magnetic field moved the solution toward the disk electrode where SWV was performed. The SWV at the detecting electrode yielded signals for both the enzymatically generated PAPR and the pumping redox species [Ru-
PAPP Synthesis. PAPP was synthesized on the basis of procedures described previously.21,24 Briefly, 300 mg of 4nitrophenyl phosphate disodium salt hexahydrate and 89.5 mg of Pd catalyst were dissolved in 75 mL of deionized water, and pH of the solution was adjusted to 6.5 by adding 6 M HCl. The solution was reduced under 1 atm of H2 at room temperature for 2 h, followed by a filtration step to remove the Pd catalyst. The clear filtrate solution was lyophilized under vacuum (20 min before use. Solutions containing [Ru(NH3)6]2+ were kept under Ar for the duration of experiments to minimize further oxidation. The TTL buffer consisted of 100 mM Tris (pH = 8), 0.1% Tween 20, and 1 M LiCl. Experimental Setup and Fluid Flow Evaluation. Chips containing individually addressable electrodes with the design shown in Supporting Information Figure S-1 were microfabricated and electrochemically characterized as described previously.7 The 2.5 cm long microband electrodes (three sets of four microband electrodes each) were used for pumping, and the 312 μm diameter disk electrode was used for detecting. Figure 1a shows their relative positions to each other. An electrical connection from the electrode chip to a CHI 760B bipotentiostat (CHI Instruments, Austin, TX) was made with an edge connector (solder contact, 20/40 position, and 0.05 in. pitch) purchased from Sullins Electronics Corp. (San Marcos, CA). Four microbands in set 1 were connected to the reference lead. Square wave voltammetry (SWV) was applied with working lead 1 at the detecting electrode. The potential step was applied with working lead 2 to activate three shorted microband electrodes in set 3 to serve as anodic pumping electrodes (at 0.3 V). Four shorted microbands in set 2, parallel to and 5.6 mm away from the anodes, served as cathodic pumping electrodes by connecting them to the counter lead. The reference lead was connected to microband electrodes in set 1 that were shorted together and served as the quasireference electrode. A 1 mL volume of a solution containing [Ru(NH3)6]2+/3+, in 0.1 M Tris buffer (pH = 9), was mixed with a 50 μL volume of the polystyrene latex bead solution. The [Ru(NH3)6]2+/3+ served as the MHD pumping species, and the microbeads facilitated particle image velocimetry (PIV).7,20 A 350 μL volume was poured onto the chip and contained by a PDMS gasket having a rectangular opening of 12.5 mm width × 27.5 mm length × 840.0 μm height. See Figure S-2 in the Supporting Information for placement of the PDMS. The chip/PDMS assembly was covered with a glass microscope slide, thus making an enclosed electrochemical cell with dimensions defined by PDMS gasket. The entire assembly was then placed under a microscope (Leica DM 2500M, Nikon) with a magnet underneath the chip, and bead movement was recorded using a Sony HDR-XR 500 V camcorder interfaced with the microscope (See Figure S-3a in the Supporting Information). Velocity data were obtained by processing the recorded movement of polystyrene latex 9407
dx.doi.org/10.1021/ac502014t | Anal. Chem. 2014, 86, 9405−9411
Analytical Chemistry
Technical Note
(NH3)6]2+/3+. See Figure 1b,c for the chemistry involved at different locations in the cell. In a control experiment, the anodic pumping electrodes were disconnected, but the disk remained active while microband electrodes in sets 2 and 1 continued to be used as counter and quasi-reference electrodes, respectively. To carry out the simultaneous pumping and detecting process over an extended period of time, 25 cycles were performed automatically by the potentiostat, back-to-back. The SWV at the detecting electrode was swept from −0.3 V to +0.5 V with increments of 5 mV at a frequency of 25 Hz and amplitude of 50 mV (taking 6.4 s), which was preceded by a 2 s quiet (equilibration) time at −0.3 V and followed by 3.4 s at open circuit for automatic file saving. Simultaneously, the pumping electrodes were stepped to +0.3 V using the chronoamperometery (CA) function, which were held at +0.3 V during the quiet time as well and left at open circuit during the file saving period. Therefore, during each cycle, detecting and pumping electrodes were activated for a total 8.4 s and there were 11.8 s between the start of each cycle. The total time that pumping and detecting electrodes were activated was 210 s, which is relevant in considering movement of chemical species due to MHD pumping. The total time from activating the electrodes at the beginning of the first cycle to inactivating them in the 25th cycle was 292.4 s (this excludes the 3.4 s file saving time in cycle 25), which is of interest in considering the distribution of species due to diffusion. Between experiments, the electrochemical cell was disassembled; the chip was cleaned using deionized water and dried with a flow of Ar, and fresh volumes of superparamagnetic bead enzyme complex solution and pumping/PAPP solution were introduced in preparation for the next experiment. Because the entire cell was rinsed and dried between experiments and placement of AP-modified superparamagnetic beads was performed manually, it was difficult to obtain highly reproducible distribution (spread, depth, and number) of the beads upstream from the detector, resulting in variations in the magnitude of the signal for PAPR and timing of its arrival at the detector from experiment to experiment. Thus, error bars are not provided on the plot in Figure 2b. However, in preliminary studies while conditions and skills were being developed, at least five to six different pumping experiments were performed with beads deposited at each position and for the control (without pumping) that yielded the same trends as those described herein.
Figure 2. (a) SWV at the detector shows an increase in PAPR current signal over time due to its continuous generation at the assay site and arrival at the detector for the fully operational case with the beadenzyme complex at position 2 and redox-MHD pumping having been activated. Only every third SWV response is shown, starting with the first SWV (which follows a 2 s quiet time). Each response, therefore, occurs at 25.2 s intervals over the 210 s duration that electrodes were activated in an experiment. (b) Plots of peak current with time from every response obtained during continuous SWV at the detecting electrode for the fully operational case (closed circles), for the case where pumping electrodes are turned off (open circles), and for the two cases where the bead-enzyme complex is off the path (P1 (open squares) and P3 (closed diamonds)) of the detector. The peak current for PAPR of the first SWV response occurs at 6.64 s (including the 2 s quiet time) and is plotted at that time in the figure. Each point thereafter corresponds to the same time within each subsequent SWV response and therefore appears in the figure at increments of 8.4 s. (This excludes the time required to save files in each cycle, during which pumping is stopped.)
RESULTS AND DISCUSSION Parameters were carefully chosen to provide the most effectual results. Pumping redox species were selected on the basis of previous work in our laboratory,21 where it was shown that hexaammineruthenium(II/III) ([Ru(NH3)6]2+/3+) is compatible with the enzyme activity of AP and the detection of PAPR. The pumping redox species concentration and electrode configuration for achieving the optimum flow rate were determined in separate experiments, and the results are given in Figure S-3 in the Supporting Information. A fluid velocity of 58 μm/s was produced with a current of 57.3 ± 3.0 μA for a solution concentration of 9.5 mM [Ru(NH3)6]2+/3+ in 0.095 M Tris buffer and at 5 s after the potential was applied to the pumping electrodes. (At longer times, the speed drops off, tracking the chronoamperometric response, as shown in Figures S-3 and S-4 in the Supporting Information.) This flow rate was fast enough to transport the enzymatically
generated PAPR electroactive species from the bead-enzyme complex site to the detecting electrode, while the concentration of the pumping species was still low enough to easily distinguish the PAPR signal over that of the pumping species. The presence of superparamagnetic beads between the active pumping electrodes did not cause a noticeable deviation from the typical flat flow profile7,23 that is generated by these electrode and cell geometries. See Figure S-4 in the Supporting Information. The “fully operational” proof-of-concept, redox-MHD-assay experiment is schematically illustrated in Figure 1. It involves a magnet directly below the chip, the bead-enzyme complex placed in position 2 directly upstream from the detecting electrode, a solution of 10 mM [Ru(NH3)6]2+/3+ and 10 mM PAPP in 0.1 M Tris buffer (pH = 8), and both pumping and detecting electrodes turned on (after a 60 s incubation period). An overlay of every third SWV response obtained at the
■
9408
dx.doi.org/10.1021/ac502014t | Anal. Chem. 2014, 86, 9405−9411
Analytical Chemistry
Technical Note
(NH3)6]2+/3+ pumped to a detecting electrode (albeit with a different geometry than that used here) produced a ratio of only 0.5:1 when it should have been 2.6:1 for a bulk solution of the same concentrations. The speed with which the plume of PAPR generated at the AP-modified superparamagnetic beads reached the downstream detector can be quantified by the change of PAPR signal with time. This can be represented by the slope of a regression line in Figure 2b from the 5th to 12th (40.2 to 99.0 s) data point: y = (0.0114 ± 0.0001) (μA/s) x − (0.203 ± 0.010) μA with R2 = 0.999. This region, which contains the inflection point, was chosen because the SWV peak current was measurable above background but had not yet achieved a steady state. We can use this value to compare such “arrival rates” with the control experiment and other bead-enzyme positions. In Figure 2a, it is interesting that the peak current due to the electrochemistry of [Ru(NH3)6]2+/3+ is constant (−8.691 μA ± 0.109 μA) over the course of the experiment (from the 1st to 25th cycle, about 284 s). This implies that the sensitivity of detecting electrodes can be internally calibrated. It also suggests that the flow velocity is fast enough to refresh the solution over the detector in a consistent manner. In addition, compositional changes that occur at the anode and cathode pumping electrodes are swept downstream before they have time to diffuse halfway across (2.6 mm) to reach the edge of the detector. Thus, redox-MHD microfluidics with electrochemical analysis is possible with high reproducibility at the detector and minimal interference from the pumping electrodes. The peak potential for the pumping species did not shift significantly over the course of each experiment. This suggests that compositional changes were inconsequential at the polarizable bare gold “quasi-reference” electrode that we used in our configuration. Nonetheless, the position of the peak potential of the pumping species could be used as a thermodynamic reference for cases where stable reference electrodes are not convenient or possible (e.g., ultrasmall, confined volumes). A control experiment was performed for the situation where the bead-enzyme complex was located at position 2, but where the pumping electrodes were turned off, to determine whether diffusion alone can account for the arrival of PAPR under full operation. Figure S-5 in the Supporting Information shows the schematic of the experiment and raw SWV data from which the plot in Figure 2b was obtained. There was little to no observable signal above background current for PAPR at the detector, although peaks for [Ru(NH3)6]2+/3+ were present. The current at the potential where PAPR was expected did increase, but ever so slowly. Thus, a linear regression was performed but over the entire pumping period, from the first to the 25th data point, giving y = (0.000170 ± 0.000009) (μA/s) x + (0.100 ± 0.001) μA (R2 = 0.936). The slope is only 1.5% of that for full operation. Also the current measured at 210 s of the time that electrodes were active is only 8% of that for full operation. These results are consistent with our estimate of the diffusion length over the entire experiment (60 s incubation period + 210 s when detector is on + 81.6 s for intermittent file saving), which is ∼593 μm, is much smaller than the 2000 μm distance to the bead-enzyme site. To perform parallel assays in a microfluidic environment, the adjacent flow paths need to remain independent of each other to avoid chemical and detection interference that can be caused by overlap. The utility of redox-MHD pumping for parallel assays was investigated by performing experiments with the
detecting electrode is shown in Figure 2a. Each occurs, therefore, at intervals of 25.2 s during the electrode activation time. The peak for the PAPR species that was generated by AP at the enzyme complex site appears at a distinctively different potential (0.28 V) than the peak for [Ru(NH3)6]2+ (0.02 V). Thus, one can follow the arrival of PAPR at the detecting electrode due to pumping by plotting the SWV peak current for PAPR with time, as shown in Figure 2b. As soon as the pumping electrodes are turned on, only electrochemistry of [Ru(NH3)6]2+/3+ can be observed at the detecting electrode. About 35 s later during electrode activation time, a signal for PAPR at the detector starts to become noticeable as it arrives due to transport by redox-MHD from the bead-enzyme complex site. The averaged current at the pumping electrodes is 38.2 ± 9.62 μA over this time period, a factor of 0.67 of the current in the study that measured the speed at 58 μm/s. (The CA current for each cycle was integrated to obtain total coulombs and then divided by the time for that CA; this average single-cycle current for the first four cycles, covering the 35 s, was then averaged and a standard deviation obtained.) Because flow velocity is proportional to the current at the pumping electrodes,22,28 it is possible to estimate the fluid velocity on the basis of the current alone. Thus, we expect the average speed to be ∼39 μm/s. The time for arrival of the first PAPR is faster than expected. Assuming a flow rate of 39 μm/s, the species should travel 1365 μm in 35 s, but the distance between the center of the bead-enzyme complex site and the detector is 2000 μm. This discrepancy is easily explained by molecules arriving from the frontmost edge of the superparamagnetic bead site (∼520 μm in radius, but varies) and spread due to diffusion (i.e., the root-mean-square distance (2Dt)1/2) where molecules will diffuse during the t = 60 s incubation period plus the 35 s transit time and 10.2 s of open circuit time for saving files is ∼324 μm (assuming D = 5 × 10−6 cm2/s). The PAPR signal continues to increase over the course of the entire experiment. However, the change in current with time (the slope) reaches a maximum at the detector at ∼100 s. A possible reason for this behavior is that, during the 60 s incubation time without pumping, a significant amount of PAPR has been generated that is concentrated in a diffusion-limited zone around the bead-enzyme site with the highest concentration at the center. Once the pumping electrodes are turned on, that plume is transported to the detector, presumably without distortion because of the flat flow profile.7 Because AP continues to produce PAPR by reacting with fresh PAPP delivered to the bead-enzyme complex from upstream, PAPR continues to produce a signal at the detector. The experiment did not last long enough to observe a steady state, where the flux at the detector is limited by the flow velocity and shorter time for AP to react with PAPP before it is pumped away. The peak current for PAPR does not exceed 0.2 of that for the [Ru(NH3)6]2+/3+ over the time of our experiment. This ratio is lower than we would have expected, which is 1.4:1, on the basis of a calibration study for SWV of mixtures of PAPR and [Ru(NH3)6]2+/3+ reported by Weston et al.21 and assuming complete conversion of 10 mM PAPP to PAPR and for the 10 mM [Ru(NH3)6]2+/3+ used here, respectively. Suppression of the ratio can be explained by incomplete enzymatic conversion of PAPP and by dilution of the PAPR product by diffusion. The diffusion argument is supported by Weston et al.21 who showed that a plug of known concentration of PAPR and [Ru9409
dx.doi.org/10.1021/ac502014t | Anal. Chem. 2014, 86, 9405−9411
Analytical Chemistry
Technical Note
bead-enzyme complex immobilized at positions 1 and 3, off the central path by >5 times the diameter of the detecting electrode. The schematics of these experiments and raw SWV data used to produce the corresponding plots in Figure 2b are presented in Figures S-6 and S-7 in the Supporting Information, respectively. A small PAPR peak in the SWV response becomes more visible above the background with time for both cases. For position 1, it is not until the eighth SWV response when the signal at the detector pulls away noticeably from the background and then increases slowly. Thus, the linear regression was obtained from the 8th (65 s when electrodes are active) to the 25th SWV response: y = (0.00126 ± 0.00002) (μA/s) x + (0.070 ± 0.003) μA (R2 = 0.997). In the case of position 3, the current pulls away from background at the second SWV response and slowly increases over the entire experiment, and therefore, a linear regression was obtained from the 2nd (15.0 s when electrodes are active) to the 25th SWV response: y = (0.00084 ± 0.00001) (μA/s) x + (0.140 ± 0.001) μA (R2 = 0.997). Compared to the full operation at position 2, the initial rates of arrival, based on the slopes in these equations, are only 11% and 7.4%, respectively. Also, the current measured at 210 s was only 21% and 20%, respectively. These results further confirm that there is only minimal interference between the parallel flow paths while transporting the electroactive species using redox-MHD pumping. The nonzero signal for PAPR can be explained by limitations in our experimentation, rather than by the redox-MHD microfluidics. Because bead placement was performed by hand, a few beads would scatter into adjacent paths. Scattered beads in line with the detector generate PAPR and produce a weak signal at the downstream detector. Another reason is a small overlap in adjacent paths due to diffusion of the electroactive species while transporting; however, as we have shown above, this is less likely because the large distances (1240 μm) between the adjacent paths exceed the maximum diffusion length (593 μm). The data obtained here clearly demonstrate the benefit of the uniform flow profile in transporting the PAPR species with redox-MHD pumping, leading to minimal interference between the parallel flow paths, thus supporting the notion that redox-MHD may perform multiple parallel microfluidics and detection without the need for independent channels. The vertical flow profile, which we have measured previously,7 should be mentioned. It is not uniform from the chip surface to glass lid of the cell. However, the horizontal flow profile at each height remains flat, and thus, adjacent streams should stay isolated, except for the effect of diffusion. The impact of the vertical flow profile on parallel analyses involving redox-MHD microfluidics is a topic for future investigation.
can serve the purpose of an internal standard. The peak potential can be used as a thermodynamic reference, which is especially valuable in miniaturized devices where it is difficult to implement traditional reference electrodes. The fact that the peak current is steady throughout the experiments can be used to calibrate the sensitivity of the electrode. Although not shown here, increases in flow rate and elimination of the redox species from solution are now possible, as well, further suggesting the importance of this new microfluidic pumping approach. The next step is to build devices with multiple detectors in the gap between the pumping electrodes to perform multiplex assay detection. These could be combined, for example, with multiple spots located upstream, each containing different superparamagnetic beads modified with different concentrations or kinds of enzymes, capture antibodies, or capture DNA, to simultaneously obtain an entire calibration curve in one run or quantify different analytes in a single mixture.
CONCLUSIONS A proof-of-concept study toward the development of multiple, parallel chemical analyses on a silicon chip was performed using redox-magnetohydrodynamics (MHD) pumping. This was accomplished using two advantages of redox-MHD, the flat flow profile and the inherent presence of the NdFeB permanent magnet to immobilize a superparamagnetic microbead assay complex. Selective placement of bead-enzyme complex at different locations upstream from an electrochemical detector generated a strong signal when in the flow path of the detector and very weak signals, when outside that path, thus confirming the value of the redox-MHD flat flow profile. The presence of the redox pumping species (e.g., [Ru(NH3)6]2+/3+) in solution
(1) Elnifro, E. M.; Ashshi, A. M.; Cooper, R. J.; Klapper, P. E. Clin. Microbiol. Rev. 2000, 13, 559−570. (2) Goldman, E. R.; Clapp, A. R.; Anderson, G. P.; Uyeda, H. T.; Mauro, J. M.; Medintz, I. L.; Mattoussi, H. Anal. Chem. 2004, 76, 684− 688. (3) Keshishian, H.; Addona, T.; Burgess, M.; Kuhn, E.; Carr, S. A. Mol. Cell. Proteomics 2007, 6, 2212−2229. (4) Vet, J. A. M.; Majithia, A. R.; Marras, S. A. E.; Tyagi, S.; Dube, S.; Poiesz, B. J.; Kramer, F. R. Proc. Natl. Acad. Sci. USA 1999, 96, 6394− 6399. (5) Zheng, G. F.; Patolsky, F.; Cui, Y.; Wang, W. U.; Lieber, C. M. Nat. Biotechnol. 2005, 23, 1294−1301. (6) Whitesides, G. M.; Stroock, A. D. Phys. Today 2001, 54, 42−48. (7) Sahore, V.; Fritsch, I. Anal. Chem. 2013, 85, 11809−11816. (8) Bazant, M. Z.; Ben, Y. X. Lab Chip 2006, 6, 1455−1461.
■
ASSOCIATED CONTENT
S Supporting Information *
The electrode chip design, schematics of PDMS location relative to electrode positions, experimental setup for redoxmagnetohydrodynamics (MHD) experiments, examples of chronoamperometric responses at electrodes during pumping, flat flow profiles both near and distant to superparamagnetic beads, and square wave voltammetric detector responses for when the alkaline-phosphatase-derivatized superparamagnetic bead complex was located in positions 1 and 3 with pumping and in position 2 without pumping (control experiment). This material is available free of charge via the Internet at http:// pubs.acs.org.
■
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected]. Notes
The authors declare the following competing financial interest(s): I. Fritsch has declared a potential conflict of interest due to her equity stake of >5% in SFC Fluidics, Inc. Dr. Fritsch created intellectual property on redox-MHD which is licensed to SFC Fluidics.
■
ACKNOWLEDGMENTS We are grateful for financial support from the National Science Foundation (CHE-0719097 and CBET-1336853) and the Arkansas Biosciences Institute, the major research component of the Arkansas Tobacco Settlement Proceeds Act of 2000.
■
■
9410
REFERENCES
dx.doi.org/10.1021/ac502014t | Anal. Chem. 2014, 86, 9405−9411
Analytical Chemistry
Technical Note
(9) Yang, S. Y.; Lien, K. Y.; Huang, K. J.; Lei, H. Y.; Lee, G. B. Biosens. Bioelectron. 2008, 24, 855−862. (10) Sivagnanam, V.; Song, B.; Vandevyver, C.; Gijs, M. A. M. Anal. Chem. 2009, 81, 6509−6515. (11) Sista, R. S.; Eckhardt, A. E.; Srinivasan, V.; Pollack, M. G.; Palanki, S.; Pamula, V. K. Lab Chip 2008, 8, 2188−2196. (12) Gijs, M. A. M. Microfluid. Nanofluid. 2004, 1, 22−40. (13) Qian, S.; Bau, H. H. Mech. Res. Commun. 2009, 36, 10−21. (14) Leventis, N.; Gao, X. R. Anal. Chem. 2001, 73, 3981−3992. (15) Grant, K. M.; Hemmert, J. W.; White, H. S. J. Am. Chem. Soc. 2002, 124, 462−467. (16) Kang, H. J.; Choi, B. Sens. Actuators, A: Phys. 2011, 165, 439− 445. (17) Pamme, N. Lab Chip 2006, 6, 24−38. (18) Weston, M. C.; Gerner, M. D.; Fritsch, I. Anal. Chem. 2010, 82, 3411−3418. (19) Anderson, E. C.; Weston, M. C.; Fritsch, I. Anal. Chem. 2010, 82, 2643−2651. (20) Scrape, P. G.; Gerner, M. D.; Weston, M. C.; Fritsch, I. J. Electrochem. Soc. 2013, 160, H338−H343. (21) Weston, M. C.; Nash, C. K.; Fritsch, I. Anal. Chem. 2010, 82, 7068−7072. (22) Weston, M. C.; Fritsch, I. Sens. Actuators, B: Chem. 2012, 173, 935−944. (23) Sahore, V.; Fritsch, I. Microfluid. Nanofluid. 2014 (published online); DOI: 10.1007/s10404-014-1427-6. (24) Deriemer, L. H.; Meares, C. F. Biochemistry 1981, 20, 1606− 1612. (25) Niwa, O.; Halsall, H. B.; Heineman, W. R. Anal. Chem. 1993, 65, 1559−1563. (26) Wijayawardhana, C. A.; Halsall, H. B.; Heineman, W. R. Anal. Chim. Acta 1999, 399, 3−11. (27) Wijayawardhana, C. A.; Wittstock, G.; Halsall, H. B.; Heineman, W. R. Anal. Chem. 2000, 72, 333−338. (28) Weston, M. C.; Nash, C. K.; Homesley, J. J.; Fritsch, I. Anal. Chem. 2012, 84, 9402−9409.
9411
dx.doi.org/10.1021/ac502014t | Anal. Chem. 2014, 86, 9405−9411
