Electrically-Actuated Valves for Woven Fabric Lateral Flow Devices

Subscriber access provided by HACETTEPE UNIVERSITESI KUTUPHANESI

Article

Electrically-Actuated Valves for Woven Fabric Lateral Flow Devices Tanya Narahari, Dhananjaya Dendukuri, and Shashi K. Murthy Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.7b00275 • Publication Date (Web): 24 Mar 2017 Downloaded from http://pubs.acs.org on March 27, 2017

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 21

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

Electrically-Actuated Valves for Woven Fabric Lateral Flow Devices Tanya Narahari1, Dhananjaya Dendukuri2 and Shashi K. Murthy1,3* 1

Department of Chemical Engineering, Northeastern University, Boston, MA, USA

2

Achira Labs Private Limited, Bangalore, India

3

Barnett Institute of Chemical and Biological Analysis, Northeastern University, Boston, MA, USA

*Corresponding author. Address: Northeastern University, 360 Huntington Ave., 313 Snell Engineering, Boston, MA 02115, U.S.A; Email: [email protected]

1 ACS Paragon Plus Environment


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 21

ABSTRACT

_______________________________________________________________________________________ The integration of flow control elements into low-cost biosensors, presents a significant engineering challenge. This article describes the development and integration of active, chemical valves into lateral flow devices, using a scalable, single-step, weaving-based manufacturing approach. The valve was constructed from an electrically conductive polymer, polypyrrole. The polymer switches between wetting and nonwetting states when it is reduced and oxidized via the application of an electrochemical potential. In this work, yarns were first coated with polypyrrole, and integrated into fabric lateral flow sensors. The coated yarns were stimulated in-situ, via integrated electrodes. Coated textiles were characterized for their response to variations in the applied electrical potential, the duration for which the potential is applied, and the chemical composition of the polymer. Among these tuning parameters, the concentration of iron (iii) chloride utilized to catalyze the synthesis of the polymer, was found to be a significant determinant in the wetting range of the polymer. Complete ON/OFF flow control was achieved at applied potentials of 20 V.cm-1, within 120 s of stimulation, using 0.1 M iron (iii) chloride, making the valve fairly easy to incorporate into point-of-care format. The practical utility of the valve was demonstrated by performing a Lowry protein assay in the device, wherein fluid flow was deactivated to allow individual reaction steps to go to completion prior to re-activation. Significant improvements in the sensitivity and linear range of the devices are reported in a simple straight-channel, lateral flow device, with the potential to develop more complex channel geometries

via

the

weaving-based

approach.

_______________________________________________________________________________________


Page 3 of 21

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


INTRODUCTION

Lateral flow assays (LFAs) are microfluidic point of care tests, with a diverse set of applications, ranging from pregnancy and fertility testing1, to drug testing2, infectious disease diagnostics3, food testing4, and metabolite assaying and monitoring5. The primary reason for their success, is that they are constructed from inexpensive wicking materials which imbibe and transport liquids through capillary forces, eliminating the need for pumping hardware6-7. A few drops of serum or urine deposited on one end of the device, reacts with a colorimetric detection reagent pre-dried on a conjugate pad. The analyte-detection complex then flows downstream, where it is captured by a set of immobilized receptors to form colored lines that can be interpreted by an untrained end user. The concentration and intensity of the end product of the assay, or the sensitivity of the device, is predicated on the rate at which the sample carrying the analyte flows past the immobilized receptors1, 8. Flow control is therefore paramount, to the sensitivity of the LFA. Commercial LFAs make use of membranes with tuned liquid absorption and retention properties to achieve this control. However, they lack the ability to

deliver precisely controlled sample-reagent

incubations, limiting their utility to the detection of high concentration analytes (≥ 25 mIU/mL) 9. By comparison, a diverse range of fluidic operations have been demonstrated in paper-based analytical devices1017

. Paper devices making use of geometric modifications to time the flow of multiple reagent streams were

first demonstrated by Yager et. al.13-15. Martinez. et. al. have demonstrated the use of dissolvable, polymer channel segments as passive shut off valves16. These flow control techniques are tremendously useful in broadening the functionality of low-cost tests. However, given the hygroscopic nature of most paper-based materials, these devices can suffer from poor reproducibility outside the laboratory1.

In contrast, the

wettability of a material may be controlled in a robust manner using externally applied forces1,

18-19

. For

instance, a paper-based device with a wax switch activated by heating and cooling the device, was recently demonstrated17. Such ‘actively’ controlled devices offer broader functionality, and better reproducibility, at very low costs. In addition, such devices can be actively reconfigured to suit the requirements of an assay. However, given that this valve is thermally activated, this technique subjects the sample and reagents to undesirable changes in temperature and composition. The devices also require expensive, custom-built manufacturing equipment. There is ample scope to develop a valving technique that overcomes these limitations, and is easy to incorporate into an existing manufacturing process. We recently demonstrated a scalable, weaving-based approach to manufacture fabric-based microfluidic devices20. The primary advantages of this manufacturing



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 21

approach are that complex channel architectures are assembled in a single step, in an automated fashion, using weaving infrastructure that is readily accessible to us through our collaborators in southern India. In the original work, wetting pathways were patterned into silk fabric, by weaving hydrophilic silk yarns into a hydrophobic, metallized yarn background. Further, hydrophilic yarns of varying twist frequency, thickness, and composition, were used to tune fluid flow, thereby affording a degree of passive flow control. We now demonstrate an additional flow control feature: an actively tunable, polypyrrole-based chemical valve, whose wettability may be manipulated in-situ., for precisely timed on/off control. Selecting an actively tunable material is key to the development of such valves. A variety of surface-tunable materials which alter their wetting properties in response to optical21-22, thermal18, and electrochemical stimuli23, have been developed. Electrically activated materials are particularly commercially viable, as electrical components are easily integrated into microfluidic devices at low costs24. We have previously demonstrated that metallized yarn electrodes may be integrated into fabric microfluidic devices by the weaving-based approach, and used for applications such as electrophoresis25. We demonstrated that any resistive heat generated as a result of passing a current through the device, is minimized by using low conductivity buffers and tuning the gap distance between the electrodes. Conductive polymers such as polypyrroles26 and polyanilines27 are examples of electrically activated materials. It is known that thin (< 100 nm) polypyrrole films synthesized in aqueous environments, swell to absorb water when a negative potential is applied to the polymer28-29. Swelling is typically accompanied by electrochemical reduction and neutralization of the polypyrrole, and the release of anionic counterions from the polymer. Oxidation, on the other hand, causes the polymer film to contract and become more hydrophobic. If the dopant molecules were released into a confined volume, they may re-enter the polymer matrix upon oxidation. Doped polypyrroles can therefore be converted reversibly between wetting and nonwetting states (Scheme 1) 23. Scheme 1: Dopant ion exchange facilitated by the electrochemical reduction and oxidation of the polymer

Further, polypyrroles are biocompatible, and polypyrrole films synthesized on flexible substrates such as fabric

30-31

, mimic the porosity and topography of the substrate. In the present work, polypyrrole films were


Page 5 of 21

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


synthesized in textile fabric using a chemical catalytic process30-31 described in Figure 1a. Coated pieces of cotton fabric were first characterized for their wettability switching properties before a polypyrrole-based valve was constructed. The integration of the valve with the device takes place in parallel with the assembly of the device, as shown in Figure 1, with multiple devices being assembled simultaneously. Textile yarns were first coated with the polymer, and wound onto spools. The loom was then prepared for weaving by setting a non-wetting yarn set lengthwise along the loom, and interlacing the device components, which included hydrophilic yarns for the flow channel, polymer coated yarns for the valve, and flexible copper wire electrodes for activation, in a single step (Figure 1b). The weaving process gives rise to an integrated device sheet, shown in Figure 1c. The sheet is cut into individual 4 x 40 mm devices, and utilized to facilitate timed flow and sample-reagent incubation. EXPERIMENTAL METHODS

Reagents The raw materials for polymer synthesis included pyrrole monomer, iron (iii) chloride, and ammonium persulfate. Pyrrole monomer (99%, Acros Organics) was purchased from Fisher Scientific (Pittsburgh, PA). Iron (iii) chloride and ammonium persulfate were purchased from Sigma Aldrich (St. Louis, MO). Reagents for the Lowry assay were purchased from Thermo Scientific (Cambridge, MA). Sample to reagent ratios were optimized for a lateral flow system, and a 25:10:4, v/v/v ratio of the protein sample to Lowry Reagent (LR) to Folin Ciocalteu (FC) reagent from the kit (ThermoScientific, #23240) was found to be optimal. Polypyrrole synthesis Polypyrrole was synthesized in textile fabrics using a chemical oxidative process described previously30, 32. The procedure involved drop coating the fabric with equal volumes of pyrrole monomer and a chemical catalyst (iron (iii) chloride or ammonium persulfate) until the fabric is completely saturated. The fabrics were kept covered in a petri dish for 30 minutes, rinsed with DI water to remove excess reagent, and oven dried at a mild temperature of 40° C. The color of the coated fabric changes slowly from pale yellow to beige, grey, or black, depending on the amount of polymer used. The color change constitutes a visual test for polymerization. Other tests included contact angle and wettability measurements, and have been described below. Wettability switching in polypyrrole-coated fabrics



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 21

The switching protocol was first optimized using a fabric-based assembly mimicking the lateral flow device (supplementary information, Figure S-2). Electrical potentials of 5-20 V.cm-1, alternating between cathodic () and anodic (+), and ranging in duration from 30 to 120 s, were applied to the piece of coated cotton fabric. In this article, a set of successive cathodic and anodic treatment steps comprise a single treatment ‘cycle’. Measuring fabric wettability and polymer coat composition Fabric wettability and chemical composition were measured at each step of the switching process in the characterization experiments. Wettability was measured in terms of the contact angle made by a drop of water placed on the fabric. A sessile droplet, 100 µL in size, was placed on the surface, and imaged using a Proscope HR2 microscope under uniform illumination. The contact angles are obtained from the images using ImageJ software, and were calculated as the angle between the liquid-solid contact line, and the airliquid contact line. In order to deduce the chemical mechanism of the switch, the fabrics were also analyzed for changes in their elemental composition after switching. Compositional analyses were performed using an EDAX ElectronDispersive X-Ray spectroscopy system associated with a Hitachi S-4800 scanning electron microscope located at the Northeastern University Electron Microscopy Core. The X-ray signature of the sample is analyzed for the qualitative composition of the sample as well as the quantity of each component. Valve manufacture Woven lateral flow devices with switchable polymer barriers were manufactured at Achira Labs, Bangalore, India, using commercially available yarns purchased from Silk Touch™, Bangalore. The weaving process is illustrated in Figure 1. Woven fabrics typically consist of two orthogonal sets of intersecting yarns: the ‘warp’ are set along the axis of the loom providing the framework for the interlacement of ‘weft’ yarns, which together make up a woven fabric (Figure 1b). The devices in this work were woven using hydrophobic, raw silk warp yarns and hydrophilic rayon weft yarns. For the valve, rayon weft yarns were coated with polypyrrole. The hydrophobic warp therefore helped prevent seepage across the valve via the warp. Metal wires were also interlaced with the device via the weft, and woven adjacent to the polymer segment. The electrode and the polymer together formed conductive terminals. Multiple devices were produced in a single, integrated sheet within 10-15 minutes (Figure 1c). The sheet was cut into strips resembling individual lateral flow devices (Figure 1d). The flow channel upstream of


Page 7 of 21

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


the valve was ~ 0.4 mm x 3 mm, the valve was 0.4 mm x 2 mm, and the downstream end was ~ 0.4 mm x 3 mm. Valve operation Woven valves are operated as follows: The sample (serum, urine, or other biological fluid) is deposited on the device, upstream of the valve region (Figure 1d). The sample wets the region between the electrodes, closing the electrical circuit. An appropriate potential difference may now be applied for a brief period in order to induce a change in the wettability of the polymer-coated segment. RESULTS AND DISCUSSION

Characterizing polymer deposition on fabric Our first goal was to deposit electronically conductive films of polypyrrole on a non-conductive fabric scaffold, such as cotton. The aqueous synthesis procedure30 used for polymer deposition has been described in Figure 1a, and gives rise to uniform polymer films that permeate the thickness of the hydrophilic fabric substrate. A rudimentary test for coat integrity was performed by rolling the coated fabrics, and the deposits did not flake, peel or deform (Figure 1a, and supporting information, Figure S-1). A selection of two oxidative catalysts, iron (iii) chloride and ammonium persulfate, were tested for their ability to generate conductive polymer films. The former led to the generation of films with low sheet resistance (Figure S-1), potentially due to the fact that chloride ions behave as P-type dopants, imparting semiconducting properties to the polymer32. The latter, however, led to the generation of highly resistant polymer films, that were made more conductive by doping the films with additives such as metallic zinc, graphite, and sodium chloride (Figure S-1). To avoid the use of additional dopants, iron (iii) chloride was selected for use in all future syntheses. The color of the polymer film also serves as a visual confirmatory test for deposition, oxidation, and conductivity. Coated fabrics progressively darkened from pale yellow to pitch black as polymerization proceeded (Figure 1a, Figure S-1). Further, the color of the polymer film deepened as the ratio of iron (iii) chloride to monomer was raised (Figure S-1). The oxidized polymer is known to exist in a charged state, and is more conductive. Therefore, polymer films generated using larger amounts of iron (iii) chloride are expected to be oxidized to a greater degree, have better conductivity, and possess electrostrictive properties which aid in wettability switching.



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 8 of 21

Characterizing the wetting behavior of polymer coated fabrics Contact angle measurements were made to assess the wetting behavior of coated fabrics. Measurements were made at three stages of the switching cycle: prior to switching (termed ‘native’), after a primary cathodic switching treatment, and after a subsequent secondary anodic switching treatment. The polymer films were then categorized as low wetting (water contact angle > 90 °), highly wettable (90 °> contact angle > 5 °), and perfectly wetting (contact angle < 5 °). Native polymer films were hydrophobic, and their hydrophobicity was found to increase with increasing iron (iii) chloride concentration (Figure 2). This observation is in agreement with the fact that oxidation leads to the electro-strictive constriction of the polymer28-29. Another possible explanation for the increase in native hydrophobicity, is that while smaller amounts of catalyst generate a smooth polymer film, larger amounts of catalyst facilitate the nucleation and growth of the film upon the existing smooth polymer layer, resulting in a rough surface topology. A similar increase in hydrophobicity with increasing surface roughness has previously been reported in polypyrrole23 and spiropyran33. An increase in wettability was observed following cathode treatment (Figure 2a, 2b), indicating that cathode treatment led to the electrochemical reduction of the polymer23. The percent change in wettability was found to be directly proportional to the applied potential, as well as the treatment duration (supporting information, Figure S-2). Anodic treatment reverses the effects of cathodic treatment and causes a decrease in wettability via the oxidation of the polymer, with the caveat that less concentrated films (0.01 M and 0.001 M iron (iii) chloride) were more difficult to re-oxidize (Figure 2b). This result is expected, as the catalyst contributes to polymer electrostriction. Though the exact mechanism is as yet unknown, the ability to tune the wettability switching process has been demonstrated (Figures 2, S2). A major advantage of using actively controlled materials, is that a single device may be re-configured in-situ, for different sample and assay types, without the need to re-design the device for each application. The ability to tune the device using the applied voltage and the activation time, has been demonstrated in the Supporting information, Figure S-2, and is indicative of this re-configurability. Further, polymer synthesis by iron (iii) chloride leads to the inclusion of counter-anionic chloride species associated with the positively charged polymer backbone. We hypothesized that the reduction and oxidation of the polymer took place by a reversible electrochemical process (Figure 2c), where cathodic treatment leads to the neutralization of the charge on the polymer via an influx of electrons, resulting in the release of chloride counter-anions to the electrolyte, and vice versa. This hypothesis was validated


Page 9 of 21

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


experimentally via EDS. The results show that cathodic treatments led to a decrease in chloride and an increase in sodium ions in the coated fabric, and vice versa (see supporting information, Figure S-3), lending theoretical support to the results in Figure 2. Similar effects were also reported previously in thin films of polypyrrole doped with ClO4- or KCl 28-29. We may therefore conclude that polypyrrole films synthesized on cotton in aqueous media, and doped with chloride species, exhibit reversible wetting to non-wetting behavior when reduced and oxidized in a conductive liquid. In prior work, a conductivity range of 0.04-17 mS was found to work optimally for a fabric-based electrical device25. Biological samples such as serum, plasma, urine, and oral fluid have conductivities that fall within this range. In the present work, we have verified that common sample types such as plasma, and urine, may be used to elicit wettability changes in polypyrrolecoated fabric (see supporting information, Figure S-3b). We now utilize our switchable polymer to build a valve. Implementing the valve in woven fabric devices On/off control was demonstrated using woven LFAs with polymer barriers (Figure 3a). Rayon yarns were coated by the chemical synthetic procedure described earlier, and used in the valve, while the flow channel was made up of hydrophilic rayon weft yarns against a hydrophobic silk warp. Rayon yarns were selected for their hydrophilic nature and uniform pore size distribution for even polymer coating, while silk yarns were selected for their mechanical strength. Further, the hydrophobic warp yarn set was used to ensure that that no fluid seeps across the polymer valve via the warp. Devices were assembled as described in Figure 1, using a weaving based approach. The primary benefits of the weaving approach are single-step assembly, and the absence of joints or junctions that could cause disturbances to flow. A sample deposited on one end of the device saturates the flow channel, establishing electrical contact between the electrodes, and allowing a potential to be applied in the channel (Figure 3). To maintain uninterrupted fluidic contact, the polymer barrier was located downstream of the electrodes. The application of a cathodic potential, for a period of 120s, led to the conversion of the valve from its native, non-wetting state, to a wetting state. The fluid therefore continues to flow downstream. A control device was maintained alongside the test for contrast, where no potential was applied, and the liquid was held upstream of the valve for an ‘indefinite’ period of time (> 10 minutes), effectively creating a time delay. This feature allows samples to be incubated in the presence of the reagent for extended periods of time, otherwise not possible in valve-less LFAs.



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 21

This concept of active temporal flow control has been illustrated in a plot of wicking velocity against time, shown in Figure 3b, and is an incremental feature over the passive flow control achieved previously. A valve-less strip with no polymer barriers wets faster, and at a rate that is pre-determined by the wetting properties of the device. In contrast, fluid flow may be brought to a complete stop in a strip with a valve, and restarted, per the requirements of the assay. Further, the wettability of the polymer segment may also be reversed by virtue of the hysteresis observed earlier, but only once the device has been dried out. This is a useful feature for applications where devices can be reused. Wicking in fabric is governed by the Washburn equation for capillary flow, which dictates that the velocity of the fluid front decreases as a function of wicked length: ‫ܮ‬ଶ =

ఊ஽௧.ୡ୭ୱ ఏ ସఎ

…(1)

Where L is wicked length (m), D is the average effective pore diameter (m), θ is the contact angle (°), and η is the viscosity of the liquid (Pa.s). Differentiating equation (1) with respect to time gives us the velocity: ௗ௅ ௗ௧

=

ఊ஽.ୡ୭ୱ ఏ ଼ఎ௅

= ‫(… ݒ‬2)

Where v is velocity (m.s-1). Figure 3c illustrates various types of time delays created by manipulating the catalyst concentration in the valve. Plots of velocity against wicked length for valves with varying amounts of catalyst have been provided. In all plots in Figure 3c, a dip in the velocity of the fluid front is noted in the region of the valve. Since the catalyst concentration influences the underlying doping level of the polymer, the extent to which the velocity dips, is a function of iron chloride concentration. Complete on/off control is only achieved when higher concentrations (0.1 M iron (iii) chloride) are used to synthesize the polymer. However, iron (iii) chloride is less expensive, easier to acquire in bulk, and less environmentally toxic than existing dopants used in wettability switching. In addition, since capillary flow is driven by interfacial tension, an increase in velocity is observed when the fluid front crosses the polymer barrier to the flow channel. This post-valve surge in velocity can be detrimental to an assay, as the sample is often transported as a plug. This issue may be resolved by


Page 11 of 21

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


manipulating the dimensions of the valve or by using valves with gradients in polymer concentration. The effects of valve dimensions and position on velocity is illustrated via the following theoretical model: The driving force behind capillary wicking in a horizontal strip of fabric open to the atmosphere, is the capillary pressure, expressed by the Young-Laplace equation: ∆‫݌‬௖ =

ସఊ.ୡ୭ୱ ఏ ୈ

…(3)

Equation (3) may be incorporated into the Washburn equation (Equation 2), to obtain a relationship between pressure and velocity: ∆௣೎ ௅

=

ଷଶ௩ఎ ஽మ

…(4)

This equation is of the same form as the Hagen equation for pipe flow, relating the volumetric flow rate (Q, m3.s-1) to the pressure gradient (∆P/L, Pa.m-1) across the length of the tube: ܳ = ‫ݒ‬. ‫= ܣ‬

∆௉ యమആಽ ಲ.ವమ

=

∆௉ ோ

…(5)

Where A is the cross sectional area of the tube (m2), and R is a lumped term used to denote the hydraulic resistance to flow (Pa.s.m-1). Therefore, the resistance to flow can be computed in terms of the wettability, pore size, and dimensions of the wetting region (Figure 4). The detailed calculations of pore size have been provided in the supporting information, (Figures S-4). Plots of hydraulic resistance in a low concentration coated fabric versus uncoated cotton fabric have been provided in Figure 4a. A detailed derivation of the resistance profiles has been provided in the supporting information, Figure S-5. The plots show that the resistance to flow varies not only with the chemical tuning of the valve, but also with its position in the channel. Hydraulic resistance is therefore a combination of both active and passive factors, which may be used in conjunction with one another to remedy the post-barrier surge in velocity. A demonstration of valve utility: total protein assays Total protein assays are an essential component of many laboratory workflows, and are an important diagnostic indicator of liver and pancreatic function. The modified Lowry protein assay (Thermo ScientificTM) is performed using a two-step procedure, where the sample is first incubated separately with an alkaline copper reagent for 10 minutes, followed by incubation with a second, chromogenic reagent for 30



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 21

minutes (Figure 5). Protein concentration is measured as a function of the absorbance of a blue end product. Protocols such as the above, involving lengthy incubation steps typically require larger, more complex LFA architectures, larger sample volumes, and multiple reagent addition steps. To demonstrate the utility of the valve in overcoming the aforementioned issues, devices with a midchannel valve similar to the devices in Figure 4 were used. The design and operation of the device is illustrated in Figure 5a. The alkaline copper reagent (denoted as Reagent 1, or Lowry Reagent, or LR) is predried upstream of the valve, followed by the chromogenic reagent (denoted as Reagent 2, or Folin-Ciocalteu reagent, or FC) which is pre-dried downstream of the valve. Bovine serum albumin (BSA) samples ranging in concentration from 0 to 1500 µg/ml were tested. The sample was placed on the upstream end, where it is incubated with LR for 8 minutes. This effectively results in a time delay, enabling the reaction to go to completion. The valve is then opened, and the copper protein complexes contact and reduce the FC reagent, generating the blue end product. Continued wicking into an absorbent end zone ensures that there is no backflow, and a uniform blue front is maintained (inset Figure 5b). In the present study, the grayscale intensity of the blue end zone was measured using image analysis, and plotted against protein concentration, to create a calibration curve (Figure 5b). The procedure was benchmarked in test-tubes (Figure 5c) and in devices without valves (Figure 5d). The plots for the valved devices (Figure 5b) and test-tube assays (Figure 5c) are in agreement, with a lower limit of detection of 1 µg/ml. From the end-users’ perspective, the devices may be read visually in reference to a color card. Devices without valves yielded erratic results despite the fact that they were made larger, and wicked slower (Figure 5d). A pictorial comparison of the results in the two types of devices, and additional benchmarking with valve-less devices of a smaller footprint, has been provided in the supporting information, Figure S-6. All results show an improved signal with the valve. We note that a small amount of chloride (1-100 mM) may be released into the sample plug during valve activation. However, we do not foresee any interference with the results of an immunoassay, or a clinical chemistry assay, provided appropriate controls are included. CONCLUSIONS

A conductive polymer-based valve for active flow control has been demonstrated in this work. The valves were easy to incorporate into the device using the weaving based manufacturing approach used to assemble fabric devices. In addition, valving was achieved while still retaining the simple, straight channel 12 ACS Paragon Plus Environment

Page 13 of 21

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


architecture and small footprint of the traditional LFA. The device therefore only uses small sample and reagent volumes. However, the valve is by no means limited to straight channel 2D geometries. Complex channel architectures may be created by programming a pattern into the loom using existing fabric weaving equipment, such as a jacquard attachment compatible with computer-aided design (CAD) software. A sinusoidal channel architecture was created in prior work to demonstrate this capability20. Newer designs for devices with valves will form the subject of future work. Although a minimal amount of additional instrumentation is required for electrical activation, the technique is feasible for adaptation to the point-of-care using a compact, chip-based potentiostat housed in the casing for the device. The potentiostat may be utilized to supply the activation potentials, especially since the valves operate at relatively low potentials (≤ 20 V.cm-1) applied over short time periods (≤ 2 minutes). In addition, the activation protocol can be programmed into the electronics for the device. One of the major advantages of electrically controlled flows is device re-configurability. Further, applications combining electrophoresis (sample pre-processing and pre-concentration) with electroactive valving, may be explored as a means to improving assay sensitivity. To ensure that the polypyrrole valve is robust to fluctuations in ambient temperature and humidity, the wetting behavior of the coated fabrics was tested in two separate locations: one in Bangalore, India, at 29 °C, 60% RH, and at an elevation of 3018 ft, and the other in Boston, MA at 22 °C, 70% RH, and 141 ft. The wetting behavior in the two locations was found to be similar. In contrast, we found that the wetting behavior of uncoated, cotton and paper showed greater variability between the two locations. Coated fabrics stored for an extended period (1-2 months) remained non-wetting in the native state. However, long term variability in wettability was observed in coated fabrics subjected to cathodic or anodic cycling, indicating that the polymer is more vulnerable to oxidation in air after switching. This problem is however rendered irrelevant from a practical standpoint, as assays as performed immediately after switching. Finally, a bill of materials analysis was performed on the devices. The cost of manufacture of devices with valves amounts to as little as $0.07 per device, as compared to $0.0035 per device for devices without valves. The cost increase is minimal, and will be negligible upon scale up. ACKNOWLEDGEMENTS

The authors gratefully acknowledge the help of Tripurari Chowdhary, Manjunath Tahshildar and Anil Modali, for weaving the fabric devices, Fazli Bozal for help with experimental work, and Bill Fowle, for assistance with SEM-EDS data acquisition. Achira Labs gratefully acknowledges support from the



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 21

Biotechnology Industry Research Assistance Council (BIRAC), Government of India, through the funding from grant BT/BIPP/0674/23/12.

CONFLICT OF INTEREST DISCLOSURE

The authors declare a conflict of financial interests. Dr. Dhananjaya Dendukuri is CEO and founder of Achira Labs, and is commercializing fabric-based point-of-care sensor technologies. ASSOCIATED CONTENT

Supporting information Figures S1-S6 in supporting information are referenced in the text. The supporting information (Supporting information for publication.pdf) is available free of charge on the ACS Publications website. REFERENCES

(1)

Mansfield, M. A., Lateral Flow Immunoassay. Springer: 2009.

(2)

Bosker, W. M.; Huestis, M. A. Clin. Chem. 2009, 55, 1910-1931.

(3)

Corstjens, P.; Zuiderwijk, M.; Brink, A.; Li, S.; Feindt, H.; Niedbala, R. S.; Tanke, H. Clin. Chem. 2001, 47, 1885-1893.

(4)

Sharma, S. K.; Eblen, B. S.; Bull, R. L.; Burr, D. H.; Whiting, R. C. Appl. Environ. Microbiol. 2005, 71, 3935-3941.

(5)

Liu, L.; Peng, C.; Jin, Z.; Xu, C. Biomed. Chromatograph. 2007, 21, 861-866.

(6)

Urdea, M.; Penny, L. A.; Olmsted, S. S.; Giovanni, M. Y.; Kaspar, P.; Shepherd, A.; Wilson, P.; Dahl, C. A.; Buchsbaum, S.; Moeller, G.; Hay Burgess, D. C. Nature 2006.

(7)

Fu, E.; Yager, P.; Floriano, P. N.; Christodoulides, N.; McDevitt, J. IEEE pulse 2011, 2, 40-50

(8)

Corporation, E. M. Rapid Lateral Flow Strips: Considerations for Product Development. https://www.emdmillipore.com/US/en/products/ivd-oem-materials-reagents/lateral-flowmembranes/n6mb.qB.L0YAAAE_gut3.Lxi,nav (accessed 23rd May 2016).

(9)

Posthuma-Trumpie, G. A.; Korf, J.; van Amerongen, A. Anal. Bioanal. Chem. 2009, 393, 569-582.

(10)

Martinez, A. W.; Phillips, S. T.; Whitesides, G. M.; Carrilho, E. Anal. Chem. 2010, 82, 3-10.

(11)

Martinez, A. W., Phillips, S.T., Butte, M.J., Whitesides, G.M. Angew. Chem. 2007, 46, 1318-1320

(12)

Carrilho, E.; Martinez, A. W.; Whitesides, G. M. Anal. Chem. 2009, 81, 7091-7095.


Page 15 of 21

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


(13)

Fu, E.; Lutz, B.; Kauffman, P.; Yager, P. Lab Chip 2010, 10, 918-920

(14)

Osborn, J. L.; Lutz, B.; Fu, E.; Kauffman, P.; Stevens, D. Y.; Yager, P. Lab Chip 2010, 10, 26592665.

(15)

Lutz, B. R.; Trinh, P; Ball, C; Fu, E; Yager, P. Lab Chip 2011, 11, 4274-4278.

(16)

Houghtaling, J.; Liang, T.; Thiessen, G.; Fu, E. Anal. Chem. 2013, 85, 11201-11204.

(17)

Phillips, E. A.; Shen, R.; Zhao, S.; Linnes, J. C., Lab Chip 2016, 16, 4230-4236.

(18)

Hamedi, M. M.; Campbell, V. E.; Rothemund, P.; Güder, F.; Christodouleas, D. C.; Bloch, J.; Whitesides, G. M. Adv. Func. Mat. 2016, 26, 2446-2453.

(19)

Rezk, A. R.; Qi, A.; Friend, J. R.; Li, W. H.; Yeo, L. Y. Lab Chip 2012, 12, 773-779.

(20)

Bhandari, P. Narahari, T; Dendukuri, D. Lab Chip 2011, 11, 2469-2638.

(21)

Kenanakis, G.; Stratakis, E.; Vlachou, K.; Vernardou, D.; Koudoumas, E.; Katsarakis, N. Appl. Surf. Sci. 2008, 254, 5695-5699.

(22)

Rosario, R.; Gust, D.; Garcia, A. A.; Hayes, M.; Taraci, J. L.; Clement, T.; Dailey, J. W.; Picraux, S. T. J. Phys. Chem. B 2004, 108, 12640-12642.

(23)

Xu, L. C., W; Mulchandani, A; Yan, Y. Angew. Chem. 2005, 44, 6009-6012.

(24)

Verpoorte, E., de Rooij, N.F., Microfluidics meets MEMS. Proc. IEEE 2003, 91, 930-953.

(25)

Narahari, T.; Dendukuri, D.; Murthy, S. K. Anal. Chem. 2015, 87, 2480-2487.

(26)

Rammelt, U. B. S; El-Dessouki, M; Schulze, R; Plieth, W; Dunsch, L. J. Solid State Electrochem. 1999, 3, 406-411.

(27)

Huang, W.-S.; Humphrey, B. D.; MacDiarmid, A. G. J. Chem. Soc. Faraday Trans. 1. 1986, 82, 2385-2400.

(28)

Hamnett, A.; Christensen, P. A.; Higgins, S. J. Analyst 1994, 119, 735-747.

(29)

J. O. Zerbino, M. G. S., C. Falivene, N. Avaca, A. Maltz. Int. J. Electrochem. 2011, 2011, 1-11.

(30)

Chougule, M. A. P., S.G; Godse, P.R; Mulik, R.N; Sen, S; Patil, V.B. Soft Nanosci. Lett. 2011, 1, 610.

(31)

Zhu, Y.; Li, J.; He, H.; Wan, M.; Jiang, L. Macromolecul. Rapid Commun. 2007, 28, 2230-2236.

(32)

Przyluski, J. Z., M; Pron, A.; Kucharski, Z; Suwalski, J. J. Phys. Chem. Solids 1987, 48, 635-640.

(33)

Gao, L.; McCarthy, T. J. Langmuir 2006, 22, 2966-2967.



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 21

FIGURES

Figure 1: Weaving-based manufacture of fabric devices. (a) Rayon yarns are first selected, and coated with polypyrrole by an in-situ polymerization process catalyzed by iron (iii) chloride. The coat darkens from pale yellow to black, as oxidative polymerization progresses. Coated yarns are wound on a spool to assist in weaving. (b) Coated yarns and electrodes are then integrated into the device via the weaving process, which involves the interlacement of two orthogonal sets of yarn to produce unified fabric devices in single step. (c) The loom outputs a sheet of fabric which can be cut into individual devices. A device includes a flow channel, a pair of metal wire electrodes, a polymer valve (black) in contact with one electrode. The size of the sheet can be scaled up, to accommodate the desired number of devices in a single weaving session. (d) Schematic for the operation of a single device. The device is laminated onto a flat surface and connected to a power supply. A sample is deposited upstream of the valve, but its movement past the valve is stopped as the


Page 17 of 21

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


polymer is initially hydrophobic. The sample closes the circuit between the electrodes, therefore serving as an electrolyte for valve operation. A small voltage is applied for a few seconds to activate the valve, converting it into a wetting state, and activating flow downstream. (The image in Figure 1d was photographed by Anil Modali).

Figure 2: Characterizing the wettability switching process. (a) Native (0) fabrics that were initially hydrophobic transition to a hydrophilic state after cathode treatment (-), and then back to a hydrophobic state after anode treatment (+). (b) Plots of contact angle versus the polarity of the applied potential. The plots show that the hydrophobicity of the native, polymer-coated fabric, increases with increasing iron (iii) chloride concentration, and indicate that oxidation leads to a decrease in wettability. Further, the ability to undergo cyclical changes in wettability is also influenced by iron (iii) chloride concentration, wherein larger amounts of iron chloride (0.01-0.1 M) confer better reversibility. (c) Based on the observations in (a) and (b) above, an electrochemical redox scheme for wettability switching was proposed, in which chloride counterions present within the polymer matrix, leave and re-enter the polymer upon reduction (cathode treatment) and oxidation (anode treatment) respectively. This schematic is corroborated by EDS data shown in the supporting information, Figure S-3.



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 21

Figure 3: A demonstration of the switchable polymer valve. (a) Yellow food coloring dye is stopped upstream of the polymer barrier, as the barrier is initially non-wetting. A cathodic potential (-20 V.cm-1, 2 minutes) is applied to the ‘test’ device, while no potential is applied to the ‘control’. The fluid front resumes flowing in the test device, but remains stopped in the control device for nearly 10 minutes. (b) Plots of wicking velocity against wicking time in devices with valves versus without valves. Without a valve, the entire length of the device is saturated by the fluid in a few seconds, and this time is determined purely by the wicking properties of the fabric. In contrast, when a valve is used, this saturation time can be prolonged based on the flow requirements. In this case, it has been prolonged nearly four-fold, thereby illustrating temporal flow control. (c) Plots of wicking velocity versus wicked length for devices with barriers of varying iron chloride concentration. The shaded region denotes the position of the barrier along the length of the strip. A dip in velocity is observed in all strips, within the shaded region. However, the dip is proportional to the hydrophobicity, or iron chloride concentration of the barrier. Flow is brought to a complete stop in 0.1 M devices where the barrier is completely non wetting, but is slowed in the other strips, demonstrating a degree of tuning. The use of other tuning parameters to demonstrate active re-configurability is illustrated in the supporting information (Figure S-2).


Page 19 of 21

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


Figure 4: A simple Washburn model to calculate the hydraulic resistance through the barrier. (a) Plots of experimental data show that the hydraulic resistance through a polymer coated strip (0.001 M iron chloride) is much higher than the resistance through an uncoated strip. Further, resistance increases with length, per the Washburn law. (b)-(d) are plots of hydraulic resistance against length, for devices where a barrier of higher resistance to flow compared to the rest of the flow channel, has been placed at different distances from the inlet end. Two main observations were noted. First, the resistance to flow increases through the barrier. Second, the resistance to flow increases with wicked length. A polymer barrier of similar composition placed further away from the inlet end therefore has a larger resistance than a barrier placed close to the inlet end.



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 21

Figure 5: The Lowry protein assay in fabric devices with and without polymer valves. (a) An illustration of the assay procedure. A hydrophobic on/off barrier is woven in the center of the device. The alkaline copper reagent (LR) is pre-dried upstream of the barrier, while the chromogenic reagent (FC) is pre-dried downstream of the barrier. A protein sample (BSA) is added upstream of the valve, when flow is still deactivated. The BSA is incubated with LR for 8 minutes, before flow is activated. The copper-protein complexes then contact and reduce the FC into a blue end product. (b)-(d) The greyscale pixel intensity of the blue end zone was obtained via image analysis and plotted. Calibration curves for protein concentration were plotted for (b) devices with valves, (c) the standard test-tube based assay, and (d) larger, slower devices without valves. The plots indicate that a characteristic quadratic calibration curve is obtained in both (b) and (c), while no clear curve is obtained without valves in (d), illustrating the significance of the valve in imparting and the appropriate incubations.


Page 21 of 21

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


Table of contents graphic


Electrically-Actuated Valves for Woven Fabric Lateral Flow Devices

Recommend Documents