Spatiotemporally Defining Biomolecule Preconcentration by Merging

Dec 7, 2015 - ... are always preconcentrated at the designated location at all lengths of ... Aemi S. Abdul Keyon , Hong Heng See , Mohamed Dawod , Jo...
0 downloads 0 Views 2MB Size
Subscriber access provided by Mount Allison University | Libraries and Archives

Article

Spatiotemporally Defining Biomolecule Preconcentration by Merging Ion Concentration Polarization Rhokyun Kwak, Ji Yoon Kang, and Tae Song Kim Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.5b03855 • Publication Date (Web): 07 Dec 2015 Downloaded from http://pubs.acs.org on December 15, 2015

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Analytical Chemistry is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 26

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Spatiotemporally Defining Biomolecule Preconcentration by Merging Ion Concentration Polarization

Rhokyun Kwak1*, Ji Yoon Kang1, and Tae Song Kim1

1

Center for BioMicrosystems, Korea Institute of Science and Technology (KIST), Seoul, 136791, Republic of Korea

* Correspondence should be addressed to Rhokyun Kwak Email: [email protected]; phone:+82-2-958-6975; fax:+82-2-958-6910

KEYWORDS: Ion concentration polarization, preconcentration, biomolecule, ion exchange membrane

1 ACS Paragon Plus Environment

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 26

ABSTRACT

The ion concentration polarization (ICP) phenomenon at micro-nanofluidic interfaces has been extensively

utilized

to

preconcentrate

low-abundance

biological

samples.

Although

preconcentration by ICP is robust, its multiphysics phenomenon does not permit the clear prediction of the preconcentration conditions and sites. Here, we present a new method for spatiotemporally defining preconcentration, which can generate target-condensed plugs in a very specific region (< 100 µm) regardless of operating conditions (time, applied voltage, ionic strength, and pH). In contrast to previous devices that only use ion depletion, this device uses merged ICP zones with opposite polarity, i.e., ion depletion and ion enrichment. In this regard, ICP is initiated between two line-patterned cation exchange membranes. When voltage is applied across two membranes, an ion depletion (enrichment) zone occurs on the anodic (cathodic) side of the membranes. Two ICP zones are then merged and confined between the membranes. Consequently, the preconcentration action is also confined between the membranes. We demonstrate that fluorescent dyes are always preconcentrated at the designated location at all operating times, at a broad voltage (0.5-100 V), ionic strength (1-100 mM KCl), and pH (3.710.3) range. This device successfully condenses proteins up to 10,000-fold in a specific region of the channel (100 × 50 × 10 µm3) in 10 min. This work not only characterizes the unique scientific phenomenon of ICP overlapping, but it also opens the possibility of integrating ICP preconcentrators into commercial analysis equipment, which requires a known, stationary preconcentration site.

2 ACS Paragon Plus Environment

Page 3 of 26

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

INTRODUCTION Ion concentration polarization (ICP) has been frequently observed in various electrochemical systems (e.g., electrodialysis, fuel cells, and batteries) with ion exchange membranes1,2. As the membrane transfers only cations or anions, the ion concentration at one side of the membrane decreases (ion depletion), whereas it increases at the other side (ion enrichment). This polarization effect reduces the system’s efficiency because it induces an additional voltage drop (a.k.a. concentration overpotential)1,3. Additionally, ion transport is limited when the lowest ion concentration in the ion depletion zone reaches zero (diffusionlimited transport). Indeed, ICP is not desirable in conventional applications; however, researchers have found a way to use this phenomenon for sample pretreatment in microfluidics through filtration4-7 and preconcentration8-10 applications. In particular, ICP preconcentrators show very robust and high performance for any charged target sample. Therefore, enormous modulations of the membrane materials (e.g., Nafion, charged hydrogel, and nanochannels)8,9,1113

, geometry (batch-types with single/dual channels14-17, continuous-flow types4,18, and

multiplexed systems19,20), and integration (droplet microfluidics18,21 and sensors22) of ICP preconcentrators have been possible. Despite extensive research in the use of ICP preconcentrators, there is no clear operating condition for the development of stationary preconcentrated plugs. Conventionally, to preconcentrate targets, electro-osmotic flow (EOF) or pressure-driven flow toward the ion depletion zone is required23. This flow prevents the propagation of the depletion zone and delivers the targets to be condensed4,9. Zangle et al.24 clarified the condition of non-propagating ICP and propagating ICP, but it depends on all given parameters (channel geometry, samples’ mobility, and ionic strength). Therefore, determining operating condition is almost impossible 3 ACS Paragon Plus Environment

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 26

with this condition in real experiments. Even though the operating condition is designated, it would be valid only for one target in one solution. Otherwise, Yossifon et al.25 successfully demonstrated that the propagation of depletion zone was slowed down by focusing electric field geometrically; however, this method cannot suppress the propagation perfectly (the length of depletion zone increases as higher voltage applied). To escape this issue with the transient depletion zone, Senapati et al. and Slouka et al. developed ion selective membrane sensors by using ion enrichment and the membrane itself as a sensor26,27. Although the preconcentration of nucleic acids are clearly demonstrated, this method should hold the membrane materials and the sensing point on the membrane. Accordingly, there is geometric / material limits to enjoy enormous modulations listed above. In this paper, we demonstrate a novel ICP preconcentrator that can condense targets in a spatiotemporally defined region through various samples and operating conditions. Departing from a previous strategy (pushing flows), the ion enrichment zone is merged to stop the propagation of the ion depletion zone between two identical cation exchange membranes. We characterize the merging dynamics of the ICP and prove that the overlapped ICP zone is stabilized and stationary. The preconcentration plug is always in a predicted location, i.e., between the membranes, even when operating conditions (voltage and time), samples (dye and protein), and solutions (ionic strength and pH) are changed. This system successfully maintains the plug location until a protein is condensed up to 10,000-fold (within ~10 min).

4 ACS Paragon Plus Environment

Page 5 of 26

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

THEORETICAL CONCEPT When a relatively low voltage or current is applied, ion enrichment and depletion zones have linear ion concentration profiles by balancing diffusion and electromigration1,8,9,28. Weak linear ICP does not change the electric resistance of the system; then, we obtain the resistor-like linear current-voltage response (a.k.a. Ohmic regime). At a higher voltage, the slope of a linear concentration profile becomes steeper, and the lowest ion concentration in the depletion zone will reach zero at the interface of the electrolyte and the membrane. At this moment, the slope is maximized, and the current is saturated even when we apply a higher voltage (a.k.a. limiting regime with limiting current1,3). This linear ICP model explains two current regimes: Ohmic and limiting. In microfluidic channels, the surface charges of the channel/membrane are significant. This surface charge and strong electric field can lead to nonlinear ion transport in the depletion zone. Dydek et al.29 categorized three transport mechanisms outside of diffusion and electromigration:

surface

conduction,

electro-osmotic

flow

at

the

dead

end,

and

electroconvection (EC)30,31. All of these mechanisms drive two distinct characteristics: a flat ion depletion zone with significantly low conductance and an overlimiting current (higher current by overcoming diffusion-limited ion flux). This flat depletion zone is impenetrable to any charged particle because there is not enough ions to screen their surface charges9. For preconcentration applications, EOF and/or pressure-driven flow push the particles toward this zone. Next, the particles are condensed on the depletion zone boundary. In this scenario, if the flow is too weak, the ion depletion zone expands (propagating ICP)24, whereas if the flow is too strong, it bursts (bursting ICP)32. The preconcentrated plug will stay in the field of view for at least ten minutes only when the two are balanced (nonpropagating ICP)24,32. Although Zangle et al.24 and Kim et 5 ACS Paragon Plus Environment

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 26

al.32 indicated that when the ICP is propagating and bursting, it is hard to predict the ICP mode prior to experimentation. For a single preconcentration test, we still need to find a proper set of applied voltage and flow velocity. This is why previous ICP preconcentrators cannot escape the use of microscopes and fluorescent tracers to monitor preconcentration action4,9,16-18,20-22. To overcome this issue, we suggest a novel spatiotemporally defined ICP platform that can preconcentrate targets at a designated region under any condition, including a variety of targets, applied voltage or current, ionic strength, and pH (Figure 1). For this purpose, we initiate an ion enrichment zone in response to the ion depletion zone by patterning two cation exchange membranes (Nafion). These thin Nafions (thickness < 1 µm) are positioned on the bottom of a microchannel (see Experimental Methods). The ends of the Nafion lines are placed in buffer reservoirs to apply the electric field. When voltage is applied through the two Nafions, ion depletion (enrichment) zones occur on the anodic (cathodic) side of the membranes. Then, two ICP zones are merged and remain between the membranes (Figure 1b-c). This is because that cations can circulate from the anode to the cathode along two Nafions and merged ICP zone. In previous systems, one Nafion keep removes cations in the channel, so ion depletion zone should expand by sending anions toward the anode9,24 (those anions are then neutralized by Faradic reaction on the anode). However, here we supply the same amount of the removed cations through one more Nafion, which is connected to the reservoir and the anode. Therefore, the anions between two Nafions do not need to be away. The anions (and cations also) are just redistributed to make steady concentration gradient and constant cation flux in merged ion enrichment-depletion zone with linear slope (Ohmic and limiting regime, Figure 1b) or nonlinear slope (overlimiting regime, Figure 1c). In the overlimiting case, the steeper concentration gradient drives the larger ion flux by diffusion and electromigration in the enrichment part, and 6 ACS Paragon Plus Environment

Page 7 of 26

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

this larger flux is matched by convection with vortices in the depletion part (we will address about the source of vortices in the next section). Consequently, the location of the preconcentration plug is also confined. As shown in Figure 1a, the charged targets (hollow circles) are preconcentrated on the interface of the ion depletion and enrichment zones (curved dotted line). The channel wall is negatively charged, and this generates the EOF between the Nafions by the electric field. This EOF delivers charged particles toward the interface. Although the preconcentration process is similar to that of previous systems9, in this ICP platform, preconcentration always occurs over a wide range of applied voltages at a predicted point between two Nafions; i) even under a weak EOF, the ion depletion zone does not need to be expanded as described above (no propagating ICP), and ii) even under a strong EOF, its velocity cannot be larger than that of the recirculating flow (vortices) by EOF mismatching (no bursting ICP)33. In detail, the EOF velocity can be defined by the Helmholtz-Smoluchowski equation1, uEOF = −εζE/µ (ε: permittivity, ζ: zeta potential of the wall, E: electric field, and µ: dynamic viscosity). Because the ion concentration in the depletion zone is significantly low, most of the voltage V drops, and higher electric field is generated than that in the enrichment zone. Hence, EOF speed jumps across the interface of the depletionenrichment zones. As a result, this induces recirculating back flow, i.e. vortices, to maintain fluid mass33 (Figure 1d). In the similar theoretical analogy in Minerick et al.33, the EOF flow rate in enrichement zone (QEOF1) is equal to the total flow rate in depletion zone by EOF (QEOF2) and pressure-driven recirculating flow (QHP):

εζ E1 εζ E2  dP  w 3h = +−  ,  dx  12 µ µ µ

(1)

7 ACS Paragon Plus Environment

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

 dP  12 µζ ( E1 − E2 ) ,  −  = dx w2

Page 8 of 26

(2)

where E1 and E2 are the electric field in the enrichment and depletion zones, P is pressure, w is the channel width, h is the channel height. With stronger electric field in depletion zone (E1 < E2), the direction of pressure gradient (−dP/dx) and corresponding flow are reversed as shown in Figure 1d. Here, the depletion zone will not burst, if the EOF velocity in the enrichment zone (uEOF1) is smaller than the averaged back flow velocity in the depletion zone (average(uHP)):

εζ E1  dP  w 2 h . < −   dx  12 µ µ

(3)

By substituting the equation (2), this no bursting condition is simplified by the ratio of electric fields in depletion and enrichment zones:

E2 / E1 > 2 .

(4)

Typically, the ion concentration in depletion zone is dropped less than 1% of initial concentration6,7. Approximately, the electric field is inversely proportional to the ion concentration, resulting E2/E1~100. This value is much larger than the threshold ratio 2. In this regard, the new ICP platform has no propagating or bursting modes of ICP. The merged ICP zones between the two Nafions are unconditionally stationary; thus, we can preconcentrate any charged target under any given condition.

8 ACS Paragon Plus Environment

Page 9 of 26

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

EXPERIMENTAL METHODS The spatiotemporally defined preconcentrator described in this study was fabricated with two Nafion lines on a glass substrate and one poly(dimethylsiloxane) (PDMS) channel (Figure 1e). Using the micro-flow patterning method19, Nafion perfluorinated resin (20 wt %, SigmaAldrich, St. Louis, MO) was patterned along a PDMS mold, which has two parallel channels (width: 100 µm; height: 50 µm; inter-channel distance: 100-500 µm) resulting in two Nafion lines on a glass substrate (white lines, Figure 1e). This Nafion-patterned substrate was bonded with the PDMS microchannel (width: 50-100 µm; height: 10 µm) by an oxygen plasma treatment. In this device, two buffer reservoirs were positioned on the Nafions to apply the electric field, and two sample reservoirs were on the ends of the channel. In the buffer reservoirs, a 1 M NaCl solution was filled to compensate for the ICP effect in these reservoirs. After filling the sample solutions in the channel, we released one large droplet to connect the sample reservoirs. This process caused the pressure gradient to be zero across the channel. In this way, we were able to eliminate the undesirable pressure-driven flows. PDMS channels were fabricated by conventional molding and photolithography. To visualize the ICP phenomenon and preconcentration action, a negatively charged fluorescent dye (1.55 µM Alexa Fluor 488, Invitrogen, Carlsbad, CA) was added in various test solutions: 1-100 mM NaCl solution (pH~7), the mixture of 1 mM NaCl and 0.2 mM HCl (pH~3.7), and the mixture of 1 mM NaCl and 0.2 mM NaOH (pH~10.3). The dye’s fluorescent intensity follows the ion concentration qualitatively31,34,35. The fluorescent images were obtained using an inverted epifluorescence microscope (Olympus, IX-71) and a charged-coupled device (CCD) camera (Hamamatsu Co., Japan). The obtained images were analyzed using ImageJ software. DC voltage was applied with Ag/AgCl electrodes, and the current responses were 9 ACS Paragon Plus Environment

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 26

measured with a source measurement unit (Keithley 2635A, Keithley Instruments, Inc., Cleveland, OH).

RESULTS & DISCUSSION ICP phenomenon between two Nafions. First, we visualize the ICP phenomenon between two Nafions represented by the current-voltage response, fluorescent images, and intensity profiles (Figure 2). The ion enrichment (depletion) zone occurs from the left (right) Nafion, which has the higher (lower) fluorescent intensity than the reference value at 0 V. The two ICP zones smoothly overlap between the Nafions. For a similar ICP with a single membrane9,29,30, the merged ICP has three current regimes (Ohmic, limiting, and overlimiting) with distinct characteristics (Figure 2a). The Ohmic and limiting regimes (0.5-1 V) have linear concentration profiles, and their slopes indicate (approximately) the current levels (Figure 2b-c). The lowest intensity reaches the background signal in the limiting regime (1 V, red line at 150 µm), which indicates that the lowest ion concentration is nearly zero1,30. With the linear transport mechanisms (diffusion and electromigration), the current fluctuation is negligible (Figure 2a). In contrast, the overlimiting regime (5 V) shows a nonlinear intensity profile with a flat depletion zone by rejecting most ions (and dyes) (at 100-150 µm, Figure 2c). In this flat zone, a vortex pair generates a curved depletion boundary (red arrows, Figure 2b) and current fluctuation (> 2 V, Figure 2a)9,29. We directly visualize this vortex pair confined between two Nafions (Supporting Figure S1 and Video 1). Merging Dynamics of ICP. From a more microscopic perspective, Figure 3 shows transient dynamics during the overlapping ion depletion and ion enrichment zones. In Ohmic and limiting 10 ACS Paragon Plus Environment

Page 11 of 26

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

regimes (0.5-1 V), linear concentration gradients evolve from both Nafions (< 1 sec), and they are then connected together (> 1 sec) (Figure 3a-b). In the overlimiting regime (5 V), vortices drive strong ion depletion zone and the preconcentration shock on the boundary by pushing the dyes (black arrow at 0.2 sec, Figure 3c). The two ICP zones are also overlapped, but this happens more quickly (< 0.6 sec). The current-time responses in all three regimes show an initial drop in the current, which represents developing of the depletion zone with a low concentration (and corresponding high electric resistance) (Figure 3d-f). Similar to previous work, the weak linear ICP in the Ohmic/limiting regime (0.5-1 V) does not significantly reduce the current level. However, the merged ICP in the overlimiting regime (5 V) shows a unique current-time response, i.e., current recovery (Figure 3f). Previous studies presented exponential current drops in this regime35,36. This current recovery is due to i) the electric field, vortices, and corresponding flat depletion zones, which are confined together in a very small region between the Nafions, and ii) convective transport by vortices that can be strong enough to compensate for the reduction of diffusion and electromigration in the low ion concentration zone. If we increase the inter-Nafion distance, from 100 µm to 300-500 µm, the vortices are not well-confined, and the current is not fully recovered (Supporting Figure S2). As described in the Theoretical Concept section, currents are finally saturated, which indicates steady and merged concentration profiles (Figure 3d-f). Spatiotemporally Defined Preconcentration. After the ICP zones are merged with the stationary concentration profile and a constant current, the fluorescent dyes are stably condensed (Figure 4a). As time passes and the EOF delivers the dyes, the fluorescent intensity increases continuously (Figure 4b-c). The peak location is fixed near ~60 µm, and the preconcentrated dyes are diffused out toward the left (< 60 µm). It is because that the electric field is biased through the short cut, from the right edge of the left Nafion (at ~50 µm) to the left edge of the 11 ACS Paragon Plus Environment

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 26

right Nafion (at ~150 µm), and because there is no electric barrier on the left side, while the barrier (the depletion zone) prevent the diffusion on the right side. All of the tendencies described above are the same at the higher voltages in the overlimiting regime (10-50 V, Figure 5), whose values have been typically used for previous preconcentrators9: i) fixed peak locations weighted toward the left Nafion (ion enrichment part), ii) diffusion out of the preconcentrated plug toward the left side (Figure 5b), and iii) current recovery up to the initial level (Figure 5c). As a higher voltage generates a stronger EOF, the peak location is shifted to the right (Figure 5b) and a faster preconcentration can be achieved (Figure 5d). Notably, the preconcentration plug also becomes sharper because a strong EOF and recirculating vortices flowing in opposite directions can compress the plugs better. From the peak intensity-time curves in Figure 5d, we define the preconcentration speed as the slope of the linear regime (Figure 5e). As observed, the speed is linearly proportional to the applied voltage (5-30 V). However, at 50 V, the peak fluorescent intensity is different from the tendency; as Green et al.37 indicated, dyes are mostly trapped at the stagnation points −the corners of the wall and the Nafion (Supporting Figure S3)− with violent vortices, resulting in stronger confinement of the preconcentration plug and higher peak intensity. The preconcentration speed is gradually reduced, presumably due to unfavorably fast diffusion by the strong concentration gradient (< 350 µm, Figure 5b). In addition to the spatiotemporally fixed preconcentration over a wide voltage range (0.550 V), we verify that the preconcentration plug is also stationary between Nafions, even at various ionic strengths (1-100 mM NaCl), pH values (3.7-10.3) and under 10-100 V (Figure 6). Generally, the high ionic strength weakens the ICP phenomenon because there are many ions available to compensate for the mismatched cations and anions near the membrane1,3,8,9. The zeta 12 ACS Paragon Plus Environment

Page 13 of 26

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

potential of the channel wall and EOF are also weakened for the same reason38. As a result, the preconcentration fold of the peak intensity (after 100 sec) becomes lower with higher ionic strength solutions (10-100 mM, Figure 6b). High or low pH induces co-ion leakage on Nafions39, and low pH also induces a decrease of the PDMS’s zeta potential (and EOF)38. As a result, we obtain the weakened ICP preconcentration with an acidic or basic solution (Figure 6a and 6c). Although ICP and preconcentration performances are diminished in these unfavorable conditions, it is important that the preconcentration plugs remain between the two Nafions. Enhancing the EOF and Nafions’ conductance permits restoration of this weakened ICP through various modulations, including higher applied voltage, shorter inter-Nafion distance, wider crosssectional area of the Nafion, narrower channel width (see Figure 7), and divided channels as like EOF micropumps40. Demonstration of Protein Preconcentration. Finally, protein preconcentration is demonstrated in this new ICP preconcentrator (Figure 7). To preconcentrate fluorescein isothiocyanateconjugated albumin (FITC-albumin, Sigma-Aldrich, St. Louis, MO), which is a high ionic strength solution (1x phosphate buffer saline (PBS)), we use wider Nafions (width: 200 µm) and a narrower channel (width: 50 µm). In this way, the ICP phenomenon can be facilitated. Wider Nafions can transfer a larger portion of the cations in a smaller liquid volume between the Nafions. Under 100 V, the preconcentration plug occurs in a designated region between the Nafions (white dotted box, Figure 7a) and achieves a 10,000-fold concentration of the protein in 10 min. The fluctuation of the peak fluorescence intensity is due to the perturbation of the vortices41,42. However, the current maintains a constant level and represents the robust and stable operation of this platform.

13 ACS Paragon Plus Environment

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 26

Comparative Discussion of Different Platforms utilizing ICP between Two Identical Membranes. Previously, ICP between two identical cation exchange membranes was used to desalinate salty water43,44 and to preconcentrated biomolecules in paper-based platforms45. Although the device we proposed here shares the similar structure with those previous works, there are two major distinctions: i) the existence of vortices and/or its confinement, and ii) the existence of steady lateral flow over the membranes. First, inter-membrane distances in Kwak et al.

44,45

and Gong et al.45 were few millimeters (2~5 mm); therefore, the ICP zones are not

merged spontaneously. Even the depletion and enrichment zones were merged after a long time (~420 sec)45, the merged ICP was not confined in a small region. In addition, recirculating back flow, which is the core phenomenon dealt in this work, did not recognized in bulky channels (with negligible EOF)44,45 and highly porous papers45 (few micron pores prevent the vortex evolution). As a result, no current recovery with highly confined vortices and merged ICP zones were not shown in previous works. Secondly, no normal flow over the membranes exists in the previous platforms. As addressed in the Introduction, a normal flow against the depletion zone is required to suppress its propagation and/or delivers the targets to be condensed. Without such normal flow, in Gong et al.45, the depletion zone and preconcentrated plug propagate away from one Nafion until they touch the physical end of the channel (the other Nafion). Moreover, with the limited number of analytes in the “reservoir between the membranes”, the preconcentration factor was significantly reduced (~40-fold)45 than that of previous ICP preconcentrators (and the presented device here).

14 ACS Paragon Plus Environment

Page 15 of 26

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

CONCLUSION In this paper, we demonstrate a spatiotemporally defined preconcentrator by merging the ICP between two identical Nafion membranes. When the current flows through two juxtaposed Nafions; the ion depletion zone and ion enrichment zone face each other on the anodic and cathodic sides of the Nafions. If they are close enough, two ICP zones merge and stabilize their location even if operating conditions such as the voltage (0.5-100 V), ionic strength (1-100 mM), and pH (3.7-10.3) are changed. This spatiotemporally defined ICP allows us to preconcentrate dyes and proteins up to 10,000-fold in 10 min with the pinpointed location of the preconcentration plug, i.e., between the Nafions. Therefore, we no longer require visualization instruments and tracers, which were previously used to specify the preconcentration site. This unique spatiotemporal controllability presents a commercial opportunity for integrating ICP preconcentrators into generic benchtop analysis such as the real-time polymerase chain reaction (PCR) and enzyme-linked immunosorbent assay (ELISA).

15 ACS Paragon Plus Environment

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 26

FIGURES and FIGURE LEGENDS

Figure 1. Schematic of spatiotemporally defined preconcentrator. a) Targets (black hollow circles) are preconcentrated between two Nafion strips, which are line-patterned on the bottom of the PDMS channel. b-c) Schematic ion concentration profiles between the Nafions by merging ion enrichment and depletion zones. Linear concentration profile occurs in Ohmic and limiting regime (Figure 1b), whereas nonlinear concentration profile occurs in overlimiting regime with vortices (circular arrow, Figure 1c). C0 indicates the initial ion concentration. d) Schematic profiles of flow velocity in enrichment and depletion zones. e) The fabricated device.

16 ACS Paragon Plus Environment

Page 17 of 26

a)

b)

−7

x 10

t = 10 sec

0V 1.6

overlimiting

current (A)

1.4

A

A’

limiting

1.2 1

Nafion

Ohmic

0.8

0.5 V

0.6 0.4 0.2 0 0

c) 16

fluorescent intensity (A.U.)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

1

2

3

4

5

6

voltage (V)

x103

1V 5V 1V 0.5V 0V back

14 12 10

5V

8 6 4 2

Nafion 0 0

20

40

Nafion 60

80 100 120 140 160 180 200

distance (µm)

50 µm

Figure 2. Merged ICP phenomenon between two Nafions. a) Current-voltage curve shows three distinct regimes (Ohmic, limiting, and overlimiting). The current-voltage response is measured by ramping up the voltage by discrete voltage jumps of 0.25V in every 40 sec, which is repeated three times. b-c) In three regimes, fluorescent images (Figure 2b) and intensity profiles along AA’ at the middle of the channel (Figure 2c) are obtained. Yellow dotted boxes indicates the location of Nafions. We used the device with a PDMS microchannel (width: 100 µm; height: 10 µm) and two patterned Nafions (width: 100 µm; inter-Nafion distance: 100 µm; height: ~1 µm). 1mM KCl solution with 1.55 µM (1µg/mL) Alexa Fluor 488 is used.

17 ACS Paragon Plus Environment

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 26

Figure 3. Dynamics of merging ion depletion and enrichment zones. a-c) concatenated concentration profiles are measured in fluorescent images, after the constant voltage is applied. The intensity curves are with the different scales: 1000 A.U. (0.5V, Figure 3a), 1250 A.U. (1V, Figure 3b), and 2500 A.U. (5V, Figure 3c). Initially, freestanding ion enrichment and ion depletion are visualized. d-f) During this operation, current-time responses are also measured during 115 sec in every 0.1 sec.

18 ACS Paragon Plus Environment

Page 19 of 26

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Figure 4. Spatiotemporally fixed preconcentration in three current regimes. a) Time-lapse fluorescent images after merging ICP zones at 5V. White dotted boxes indicate the location of Nafions. b) Along the middle of the channel (A-A’), time-lapse intensity profiles are stacked during 100 sec. Although the peak intensity is increasing, as dyes are preconcentrated, the peak locations are not changed. The initial intensity profile forms evenly at ~6000 A.U.. Supporting Video 2 is also available with real-time fluorescent images and corresponding intensity profiles. c) This peak fluorescent intensity also presented in time. Initially at 5V, the spike is observed by the preconcentration shock (black arrow, Figure 3c). The time-lapse profiles and peak intensities are measured in every 0.1 sec.

19 ACS Paragon Plus Environment

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 26

Figure 5. Spatiotemporally defined preconcentration in various operating voltages (5-50V). a) Fluorescent images after 100 sec operation, and b) fluorescent intensity profiles along A-A’. Yellow dotted boxes indicate the location of Nafions. c) In these high voltage tests, currents are also fully recovered the initial values. d) The peak fluorescent intensities are traced during 100 sec with the reference lines, which indicate the actual concentration of fluorescent dyes (1-100 µg/mL of Alexa Fluor 488 in 1mM KCl solution). The same device is used in Figure 2-4. e) In addition, preconcentration speeds are calculated from the linear regime of the peak intensity-time curves in Figure 5d. Under 30V, the preconcentration speed is linearly proportional to the applied voltage (the dotted line is the linear fitting curve), while the speed is much faster at 50V.

20 ACS Paragon Plus Environment

Page 21 of 26

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Figure 6. Spatiotemporally defined preconcentration in various ionic strength (1-100mM NaCl) and pH (3.7-10.3). a) Fluorescent images obtained after 100 sec operation at 50V. As can be seen, the location of preconcentration plugs are still between two Nafions (yellow dotted boxes), even its’ intensity are weakened under high ionic strength and strong acidic or basic solutions. bc) the location of peak intensity and its intensity fold (i.e. how many times over the initial intensity) are mapped under 10, 20, 50, and 100 V. 100V is not tested in 1mM NaCl (pH7), because the peak intensity already touches the highest values (due to the saturation of the CCD camera) at 50V. From the peak intensity profile (as like Figure 5b), the peak region is also identified with 1% below of the peak intensity, which represented by errorbars in Figure 6b-c. In common, higher voltage and stronger EOF shift the peak location to the right, with higher intensity fold and sharper preconcentration plug. Gray boxes indicate the location of Nafions. The origin of the distance is the right edge of the left Nafion.

21 ACS Paragon Plus Environment

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 26

Figure 7. Demonstration of spatiotemporally fixed protein preconcentration. 1 µg/mL FITCalbumin in 1x PBS is used. 0.1% tween 20 is also added to prevent non-specific binding. a) With applied voltage 100V, the peak and averaged fluorescent intensities are traced in the white dotted box, which is the region between Nafions. In 10 min operation, the proteins are preconcentrated up to 10 mg/mL (peak) and ~0.1 mg/mL (averaged), indicating 10,000-fold and 100-fold preconcentration, respectively. The inset fluorescent images are obtained at 0, 10, 20 min. b) During the proteins are highly condensed, the current maintains the nearly constant value (10 µA).

22 ACS Paragon Plus Environment

Page 23 of 26

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

ASSOCIATED CONTENTS Supporting Information includes supporting video 1-2 and supporting figure S1-S3. This material is available free of charge via the Internet at http://pubs.acs.org.

ACKNOWLEDGMENT This work was supported by the KIST institutional program 2E25590 and 2N40800.

23 ACS Paragon Plus Environment

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 24 of 26

REFERENCES (1) Probstein, R. F. Physicochemical Hydrodynamics: An Introduction, 2nd ed.; WileyInterscience: New York, 2003. (2) Strathmann, H. Ion-exchange membrane separation processes; Elsevier, 2004; Vol. 9. (3) Newman, J.; Thomas-Alyea, K. E. Electrochemical Systems, 3rd ed.; Wiley-Interscience: New York, 2004. (4) Kwak, R.; Kim, S. J.; Han, J. Anal Chem 2011, 83, 7348-7355. (5) Jeon, H.; Lee, H.; Kang, K. H.; Lim, G. Sci Rep-Uk 2014, 4. (6) Kim, S. J.; Ko, S. H.; Kang, K. H.; Han, J. Nat Nanotechnol 2010, 5, 297-301. (7) MacDonald, B. D.; Gong, M. M.; Zhang, P.; Sinton, D. Lab Chip 2014, 14, 681-685. (8) Schoch, R. B.; Han, J. Y.; Renaud, P. Rev Mod Phys 2008, 80, 839-883. (9) Kim, S. J.; Song, Y. A.; Han, J. Chem Soc Rev 2010, 39, 912-922. (10) Scarff, B.; Escobedo, C.; Sinton, D. Lab Chip 2011, 11, 1102-1109. (11) Pu, Q. S.; Yun, J. S.; Temkin, H.; Liu, S. R. Nano Lett 2004, 4, 1099-1103. (12) Mai, J. Y.; Miller, H.; Hatch, A. V. Acs Nano 2012, 6, 10206-10215. (13) Kim, B.; Heo, J.; Kwon, H. J.; Cho, S. J.; Han, J.; Kim, S. J.; Lim, G. Acs Nano 2013, 7, 740-747. (14) Wang, Y. C.; Stevens, A. L.; Han, J. Y. Anal Chem 2005, 77, 4293-4299. (15) Lee, J. H.; Cosgrove, B. D.; Lauffenburger, D. A.; Han, J. J Am Chem Soc 2009, 131, 10340-+. (16) Cheow, L. F.; Han, J. Y. Anal Chem 2011, 83, 7086-7093. (17) Ko, S. H.; Song, Y. A.; Kim, S. J.; Kim, M.; Han, J.; Kang, K. H. Lab Chip 2012, 12, 44724482. (18) Phan, D. T.; Chun, Y.; Nguyen, N. T. Rsc Adv 2015, 5, 44336-44341. (19) Lee, J. H.; Song, Y. A.; Han, J. Y. Lab Chip 2008, 8, 596-601. (20) Ko, S. H.; Kim, S. J.; Cheow, L. F.; Li, L. D.; Kang, K. H.; Han, J. Lab Chip 2011, 11, 1351-1358. (21) Chen, C. H.; Sarkar, A.; Song, Y. A.; Miller, M. A.; Kim, S. J.; Griffith, L. G.; Lauffenburger, D. A.; Han, J. J Am Chem Soc 2011, 133, 10368-10371. (22) Dextras, P.; Payer, K. R.; Burg, T. P.; Shen, W. J.; Wang, Y. C.; Han, J. Y.; Manalis, S. R. J Microelectromech S 2011, 20, 221-230. (23) Slouka, Z.; Senapati, S.; Chang, H. C. Annu Rev Anal Chem 2014, 7, 317-335. (24) Zangle, T. A.; Mani, A.; Santiago, J. G. Chem Soc Rev 2010, 39, 1014-1035. (25) Yossifon, G.; Mushenheim, P.; Chang, Y. C.; Chang, H. C. Phys Rev E 2010, 81. (26) Senapati, S.; Slouka, Z.; Shah, S. S.; Behura, S. K.; Shi, Z. G.; Stack, M. S.; Severson, D. W.; Chang, H. C. Biosens Bioelectron 2014, 60, 92-100. (27) Slouka, Z.; Senapati, S.; Shah, S.; Lawler, R.; Shi, Z. G.; Stack, M. S.; Chang, H. C. Talanta 2015, 145, 35-42. (28) Sparreboom, W.; van den Berg, A.; Eijkel, J. C. T. Nat Nanotechnol 2009, 4, 713-720. (29) Dydek, E. V.; Zaltzman, B.; Rubinstein, I.; Deng, D. S.; Mani, A.; Bazant, M. Z. Phys Rev Lett 2011, 107. (30) Rubinstein, I.; Zaltzman, B. Phys Rev E 2000, 62, 2238-2251. (31) Kwak, R.; Pham, V. S.; Lim, K. M.; Han, J. Y. Phys Rev Lett 2013, 110. (32) Kim, S. J.; Li, L. D.; Han, J. Langmuir 2009, 25, 7759-7765. (33) Minerick, A. R.; Ostafin, A. E.; Chang, H. C. Electrophoresis 2002, 23, 2165-2173. 24 ACS Paragon Plus Environment

Page 25 of 26

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

(34) Chambers, R. D.; Santiago, J. G. Anal Chem 2009, 81, 3022-3028. (35) Kwak, R.; Guan, G. F.; Peng, W. K.; Han, J. Y. Desalination 2013, 308, 138-146. (36) Rubinstein, S. M.; Manukyan, G.; Staicu, A.; Rubinstein, I.; Zaltzman, B.; Lammertink, R. G. H.; Mugele, F.; Wessling, M. Phys Rev Lett 2008, 101. (37) Green, Y.; Yossifon, G. Phys Rev E 2013, 88. (38) Kirby, B. J.; Hasselbrink, E. F. Electrophoresis 2004, 25, 187-202. (39) Strathmann, H. Desalination 2010, 264, 268-288. (40) Wang, X. Y.; Cheng, C.; Wang, S. L.; Liu, S. R. Microfluid Nanofluid 2009, 6, 145-162. (41) Druzgalski, C. L.; Andersen, M. B.; Mani, A. Phys Fluids 2013, 25. (42) Rubinstein, I.; Zaltzman, B. Phys Rev Lett 2015, 114. (43) Kwak, R. Nonlinear ion concentration polarization : fundamentals and applications. Massachusetts Institute of Technology2013. (44) USA, US 2014/0374274 A1, 2014. (45) Gong, M. M.; Zhang, P.; MacDonald, B. D.; Sinton, D. Anal Chem 2014, 86, 8090-8097.

25 ACS Paragon Plus Environment

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 26 of 26

TABLE OF CONTENTS

26 ACS Paragon Plus Environment