Su

Da

b

c

a

ARRA

KIPSNSL

1

niRrcipqaeimte(

fp

n

h0

Sensors and Actuators B 222 (2016) 735–740

Contents lists available at ScienceDirect

Sensors and Actuators B: Chemical

jo ur nal home page: www.elsev ier .com/ locate /snb

ample concentration in a microfluidic paper-based analytical devicesing ion concentration polarization

inh-Tuan Phana, Seyed Ali Mousavi Shaeghb, Chun Yanga,∗, Nam-Trung Nguyenc,∗

School of Mechanical and Aerospace Engineering, Nanyang Technological University, 50 Nanyang Avenue, Singapore 639798, SingaporeHarvard-MIT Division of Health Sciences and Technology, Massachusetts Institute of Technology, Cambridge, MA, USAQueensland Micro- and Nanotechnology Centre, Griffith University, Brisbane 4111, QLD, Australia

r t i c l e i n f o

rticle history:eceived 10 July 2015eceived in revised form 20 August 2015ccepted 29 August 2015

a b s t r a c t

We report a low-cost approach for sample concentration in a microfluidic paper-based analytical deviceusing ion concentration polarization. This device platform can attain liquid sample filling by capillarysuction through the microporous paper and ion selective transport through a nanoporous polymer matrixunder DC electric field. The device demonstrated a fast depletion of fluorescent dye samples and can serveas an effective microfluidic concentrator of fluorescent dye with 60-fold concentration enhancement

eywords:on concentration polarizationaper-based microfluidicsample concentrationanofluidicsample preparation

achieved within 200 s. The device fabrication is simply based on paper cutting and lamination withoutthe need of lithography and printing of hydrophobic material such as wax. The lamination approachpresented in this paper has the potential to be transferred to a large-scale production of paper-basedanalytical devices.

Published by Elsevier B.V.

ab on a chip

. Introduction

Ion concentration polarization (ICP) is an ionic transport phe-omenon that occurs near an ion-selective membrane and is

ncluded the ion-depletion and ion-enrichment processes [1].ecently, ICP has attracted a great attention from the microfluidicsesearch community due to its potential applications in sampleoncentration [2–5], desalination [6,7], mixing [8,9]. Concentrat-ng low-abundance analytes is a critical need for microfluidicoint-of-care (POC) portable analytical devices, because subse-uent analysis or detection is only possible at a concentrationbove the limit of detection (LOD) of the analytical device. Sev-ral research groups have explored various concentration methodsncluding field amplification stacking [10], isoelectric focusing [11],

embrane filtration [12], affinity-based extraction [13], tempera-ure gradient focusing with a combined AC and DC field [14] andlectrokinetic trapping based on ion concentration polarizationICP) [15].

ICP-based electrokinetic trapping method has numerous uniqueeatures. Specifically, concentrating molecules using ICP is inde-endent of the hydrophobicity of binding characteristics of the

∗ Corresponding authors.E-mail addresses: mcyang@ntu.edu.sg (C. Yang),

am-trung.nguyen@griffith.edu.au (N.-T. Nguyen).

ttp://dx.doi.org/10.1016/j.snb.2015.08.127925-4005/Published by Elsevier B.V.

molecules of interest [2]. In addition, an electric field created bysimple electrodes can easily manipulate the concentrating process[2]. This allows ICP-based concentrators to be integrated in a widerange of microfluidic platforms for detection of target moleculesinitially at a low concentration. Conventionally, ICP was observed atthe interface of a permselective nanochannel bridging two adjacentmicrochannels [16]. Once an ICP is established; a depletion zone isformed at one side of the perm-selective nanochannel. In the iondepletion zone, the concentration of both negatively, and positivelycharged species drops sharply. In contrast, on the other side of thenanochannel, an ion enrichment zone is established with a highconcentration of both negatively, and positively charged species[2]. Nanoporous ion-selective materials with non-engineered poreshave been used to make ICP-based microfluidic separators andconcentrators [2]. Nafion is for instance an ion-selective mate-rial with negatively charged sulfonic groups. Kim et al. used aNafion membrane to generate ICP for continuous seawater desali-nation. Nafion-based nanojunction was integrated at the bifurcatedpoint of a microchannel where all charged species, particles andmicroorganisms were repelled into the brine channel and thuspure water was guided to the desalted channel [6]. Kwak et al.adopted a similar approach for continuous-flow concentration of

biomolecules, bacteria, and red blood cells [5]. In both designs, thenano junction had to be in contact with an additional buffer chan-nel, which added complexity to the chip design. Ko et al. reporteda design improvement in a device for protein preconcentration

7 d Actuators B 222 (2016) 735–740

[mrooo5

ftoftavtvb[caomwratst

pmpttTmtml

2

2

dowttalafdrcGt

sssc

Fig. 1. Paper-based analytical device for sample concentration using ion concentra-tion polarization: (a) ICP phenomenon occurred on a straight paper channel underthe influence of an applied electric field across the Nafion junction, which onlyallows cations to pass through its nanopores. The cathode is acted as an unlimited-anions providing source; (b) working concept showing the combination mechanism

36 D.-T. Phan et al. / Sensors an

4]. The Nafion membrane was placed at the bottom of a straighticrochannel. In this design, the buffer channel was eliminated,

esulting in a simplified fabrication process. Both anode and cath-de were located in the same microchannel with one inlet and oneutlet. The preconcentrator could enhance the detection sensitivityf the immunoassay for a sample of C-reactive protein by more than00-fold compared to an immunoassay without preconcentrator.

Over the past few years, paper-based analytical microfluidicsor point-of-care applications has also attracted a great atten-ion [17–19]. The low material, fabrication and operation costsf paper-based devices make the applications extremely suitableor resource-limited regions [20]. Passive capillary wetting (suc-ion) through the porous structure of paper makes fluid handlingn easy task. In addition, the porous structure of paper pro-ides a suitable matrix for effective immobilization of reagentshrough physical adsorption [21]. Having such unique properties,arious analytical assays have been implemented in the paper-ased microfluidic platform; for instance colorimetric detection22], direct electrochemical detection [19] and electrogeneratedhemiluminescence detection [23] are a few to name. The use of

concentrator is inevitable to further improve the detection limitf these analytical methods. Preconcentration process needs flowanipulation, which is challenging since valves should operateithout instrumented actuators and controllers [24]. Gong et al.

ecently integrated a Nafion membrane into a paper strip to maken ICP-based concentrator on paper [25]. The device could concen-rate a fluorescent tracer up to 40-fold in a fully wet paper in atamp-like device with the nanoporous membrane separated fromhe paper-based assay.

This paper presents the development of a paper-based devicelatform using simple integration of microporous and nanoporousaterials for liquid handling and sample concentration. The device

latform functions for liquid sample filling by capillary filling suc-ion through the microporous paper and ion selective transporthrough the nanoporous polymer matrix under DC electric field.he functionality of this platform is demonstrated by ICP-basedicrofluidic sample depletion and concentration. This work is dis-

inct from other published works due to the simpler geometry withaintaining functionality, fast fabrication process thus leading to

ow cost, ease of applicability for practical applications.

. Material and methods

.1. Device concept and fabrication

In order to understand the working concept of the proposedevice, it is necessary to present a detailed description on ICPccurred in the device. Fig. 1(a) describes the movement of ionshen an electric field is applied across the Nafion junction. Due

o the permselectivity of the Nafion, only cations can pass throughhe Nafion membrane, whereas the anions will be stopped rightt the Nafion interface with the cathode side. To maintain theocal electroneutrality condition, charged molecules are trappednd accumulated on the anodic side of the Nafion membrane toorm an ICP boundary. Since our configuration is symmetrical, theepletion zone will develop and expand gradually to two anodiceservoirs over time. On the other hand, on the cathode side, whichan be considered as an unlimited-anions source (i.e. connected toND or a buffer electrode), both ion types are accumulated to form

he enrichment region.Fig. 1(b) shows the concept of the concentration device. A paper

trip forms two reservoirs and a main channel for introducing theample. A thin Nafion membrane is put at the bottom of the papertrip. While the microporous paper represents the microfluidichannels for sample transport, the nanoporous Nafion membrane

of three different forces in the device; (c) device structure containing several layersfabricated with different materials.

represents the nano-channels that are ion selective. Two electrodesare placed at the reservoirs at the two ends of the paper strip forproducing two different DC voltages. A platinum wire acting as theground electrode is placed under the Nafion membrane to activatethe transport process across the nonporous matrix. This simpledevice configuration will demonstrate that sample transport andconcentration can be implemented on a single paper strip. Thisplatform would allow for the integration of a concentrator withany paper-based microfluidic assay for point-of-care applications.

Fig. 1(c) shows the implementation of the concept depictedin Fig. 1(b) through lamination of multiple layers of the differentmaterials. Each layer of the device was cut in appropriate sizes anddimensions using an electronic craft cutter (Silhouette America,Inc., Silhouette Cameo) [25]. Cellulose filter paper (Whatman) ofan average pore size of 11 �m (Grade 1) and an average thicknessof 180 �m was used as the substrate for the paper layer with twocircular reservoirs of 1-cm diameter. The strip connecting the reser-voirs has a width of 1.5 mm and a length of 25 mm. The nanoporousmembrane Nafion (NRE-212, DuPont, USA) with an average thick-

ness of 50 �m was cut in to a 1.5 mm × 5 mm piece and placed underthe paper strip to create the microporous/nanoporous interface.Platinum electrodes with a diameter of 100 �m (Sigma–Aldrich)were connected to the two reservoirs and the Nafion membrane. A

D.-T. Phan et al. / Sensors and Actuators B 222 (2016) 735–740 737

Fig. 2. Experimental setup: (a) actual fabricated device showing connections withexternal electrical wires; (b) first experiment to demonstrate the expansion of ICPzu

dianosten1

2

SspcltzItteiNcNsusM

3

3

itwp

Fig. 3. The expansion of the depletion zone over time. The insets are converted fromcaptured images to black and white format to improve the contrast within the ICP

one under an equal applied voltage of 50 V between two reservoirs; (c) electrical setp in the second experiments for demonstration of diluted sample concentration.

ouble-sided adhesive film with the thickness of 50 �m acts as thensulation layer between the electrodes and the paper strip as wells the Nation membrane. In order to establish the electrical con-ections to the reservoirs and Nafion membrane, the contacts arepened by cutting windows through the adhesive layer. A single-ided 50-�m thick plastics film (Scotch, 3M) was used to laminatehese layers together. Both top and bottom layers allow to form annclosure structure and prevent evaporation from the paper chan-el. The laminated device was clamped and baked at 110 ◦C for0 min. Fig. 2(a) shows the final fabricated device ready for testing.

.2. Experimental setup

Fluorescein sodium salt with negative charges (Acid Yellow 73,igma–Aldrich) in a concentration of 80 mM was used as the sampleolution in our experiments. The diluted sample drops were dis-ensed by an adjustable pipette (Eppendorf Research Plus) into theircular reservoir. The sample is driven into the paper strip by capil-ary wetting and subsequently is transported via electrokinetics. Inhe first experiment for examining the expansion of the depletionone, a DC power supply (Model PS350, Stanford Research System,nc.) provides voltages to the platinum wire electrodes, Fig. 2(b). Inhe second experiment for demonstrating sample concentration,wo independent power sources were used simultaneously to gen-rate the potential difference between the two reservoirs as shownn Fig. 2(c). An inverted fluorescence microscope (Eclipse TE2000-E,ikon) was used to capture the fluorescence images of the con-entration zone. All images were taken by a CMOS camera (D700,ikon) mounted on the front camera port of the microscope via a

pecialized adapter. The software Camera Control Pro 2 (Nikon) wassed to control the capture time and exposure time exactly. Sub-equently, the recorded images were processed by a customizedATLAB program.

. Results and discussions

.1. Expansion of depletion zone

To start with, five 50 �L-drops of the fluorescent tracer were

njected into the left reservoir. Then the solution was sucked intohe paper strip by capillary force. After 2–3 min, the right reservoiras filled with the solution and the entire channel of the device wasre-wetted (inset in Fig. 3). Once a 50 V voltage was applied to two

zone. The bottom and top insets correspond to t = 0 s and t = 30 s, respectively.

reservoirs, an ICP depletion zone was observed after a few secondsand expanded gradually to two reservoirs (inset in Fig. 3). The lightarea in the middle channel region located above the Nafion mem-brane (connected to GND) shows no presence of fluorescence dye(negative charges) in that depletion zone. As the time elapsed, thedepletion zone expanded outward such that the fluorescence inten-sity in the rest of channel increased, indicating that fluorescentdye displaced away from the middle depletion region graduallyaccumulated in the region outside of the depletion zone.

Fig. 3 depicts the ICP depletion length over time. The positionof the ICP boundary, defined as the distance between the bound-ary and the edge of the Nafion membrane, was measured every 10 sfrom the recorded images. The distance representing the size of thedepletion zone was obtained using a customized MATLAB (Math-works) program. Each fluorescent image was first converted to abinary image where the edge between the enriched and depletedzones is detected. Fig. 3 shows that the depletion zone expandsand reaches a steady-state condition after 2 min. The depletionzone initially expands at a rate of 2.7 mm/min but gradually slowsdown to 0.1 mm/min 2 min after switching on the voltage. In a DCelectric field, the three transport mechanisms present are ICP, elec-trophoresis, and electroosmosis. The ICP causes the ion depletion,thereby pushing the fluorescent dye away from the middle region.Meanwhile, the electrophoresis drives the negatively charged fluo-rescent dye to move toward two reservoirs where positive voltagesare applied. Intriguingly, the effect of electroosmosis is relativelycomplicated here. In principle, such DC electric field would gener-ate an electroosmosis flow toward the channel center. Due to thesame amount of voltages applied to the two reservoirs, the symme-try will not allow for generating any net flow. So to counterbalancethe electroosmosis, a back-pressure is induced [26], leading to thezero flow rate and hence no flow induced convective transport ofthe fluorescent dye. Therefore these three mechanisms led to theexpansion of the depletion zone over time. The observed reductionin expansion speed may be caused by two phenomena. First, theICP driven fluorescent dye migration decreased with increasing dis-tance from the Nafion membrane. And second, the fluorescent dyeaccumulation induced a concentration gradient that slows downthe depletion zone expansion. Hypothetically, we believe that the

electrophoresis and depletion force caused by ICP are more dom-inant than the electroosmosis flow. It might be the result of amuch higher fluidic resistance of the fiber matrix as compared to a

7 d Actuators B 222 (2016) 735–740

mtafa

ttinrar

3

dtrttacbbtntmttvectbidwttr1

Fi

Fig. 5. Sample concentration: (a) grayscale images converted from fluorescentimages of the concentrated band at three representative times of 50, 100 and 150 s,

38 D.-T. Phan et al. / Sensors an

icrofluidic channels. The velocity of the flow needs to be highero allow the liquid to penetrate the fiber network. More detailednalytical or numerical models need to be implemented in theuture studies to judge quantitatively the dominant forces in such

complex phenomenon.Moreover, we noticed that the applied voltage was limited by

he resistance of the fiber network of the paper and the conduc-ance of the Nafion membrane. Experimentally, we observed thatf the applied voltage is much lower than 50 V, the electric field isot strong enough to move the ions to penetrate the paper matrixesistance. On the other hand, the membrane may be damaged if thepplied voltage is much higher than 100 V (i.e. breakdown voltage)esulting overlimiting current.

.2. Sample concentration

In contrast to the results shown in Fig. 3 for the expansion ofepletion zone generated by applying the same DC voltages athe two reservoirs, a voltage difference is applied across the twoeservoirs for sample concentration. The left reservoir is connectedo 50 V, whereas the right reservoir is connected to 100 V withhe ground electrode under the Nafion membrane, Fig. 1a. Suchrrangement of DC voltages causes concentration of the fluores-ent dyes, resulting in an enriched band in the paper strip. Theand moves toward the reservoir with a higher potential. It iselieved that such voltage difference would generate a new elec-roosmosis flow toward to the left reservoir (with 50 V). One shouldote here that the electroosmosis flow is in opposite direction tohe direction of ICP and electrophoresis driven motion. Further-

ore, as mentioned early, the ICP effect decreases with increasinghe distance from the Nafion membrane, the combination of thesehree transport mechanisms led to a location where the migrationelocity of fluorescent dye is zero and hence the enrichment. Suchnrichment gets stronger over time, resulting in higher fluores-ent intensity of the concentrated band over time. Fig. 4 depictshe change of the peak fluorescent intensity of the concentratedand over time. Within the first 100 s, the fluorescent intensity

ncreased fast, almost linearly with time. Then, the rate slowlyecreased in the last 100 s. A peak concentration factor of 60 timesas achieved. Compared to the best performance of the device with

he pressed-on nanoporous membrane reported by Gong et al. [25],he maximum value reached 40-fold at 155 s. In this previouslyeported device, the peak concentration factor started to flatten at55 s. Obviously, our device showed a better performance of higher

ig. 4. Maximum fluorescent intensity of the concentrated band versus time. Insetmages are the fluorescent image at 50 (lower) and at 150 s (higher).

respectively; (b) intensity distribution of the concentrated band at three represen-tative instances of 50, 100 and 150 s.

concentration factor. The better performance is presumablyexplained by the purity of the nanoporous Nafion strip, whichwould provide a better selective transport of cations. The uniform inpore size of Nafion membrane leading to an increase in conductancewas also confirmed by an ICP-based concentrator fabricated byintegrating a Nafion strip [27]. In the previously reported in-paperdevice [25], nanoporous membranes were patterned directly intoa chromatography paper (Whatman, No. 1) by manually deposit-ing Nafion 117 resin. The property of the patterned nanoporousmembranes may not be uniform throughout the whole transportedregion leading to the poor performance.

Furthermore, when the sample was enriched in a concentratedband, the band with increasing intensity was observed to movetoward the reservoir with a higher potential. This could be causedby the finite-reservoir size induced back-pressure driven flow thatpush the focused band toward to the opposite direction of the elec-troosmosis flow [26]. Fig. 5(a) shows the representative fluorescentimages of the concentrated band at three representative times of50, 100 and 150 s, respectively. The intensity distribution showsthat not only the intensity (peak) of the band increases but also itssize, Fig. 5(b).

To evaluate the relative increase of the concentration area overtime, we kept a fix threshold value (i.e. reference value) when con-verting the fluorescent images to binary images. Since the recordedimages show different fluorescent intensities over time, the area ofthe enriched band is only evaluated properly by using the samereference value. In details, this threshold value is optimized tominimum the noise of the whole set of the binary images. Then,the background (i.e. channel, outer regions) will be black and theenriched area becomes totally white. Next, we extract the biggestarea of the region of interest (i.e. white area), which consists ofan enriched band. Also, we can convert this area from the numberof pixels into millimeter square using a calibration ratio, which isdetermined based on the given magnification of the imaging sys-

tem. Fig. 6 shows that the area of the enriched band grows almostexponentially over time. The Taylor dispersion may be the majorfactor causing this area expansion over time.

D.-T. Phan et al. / Sensors and Actu

F(b

aaflnbs

3

afwpatwosppaaoaifltipiy

4

abTtfifw

[[[[

[[[[[[[

[

[

[[

[

[[

ig. 6. Area of the concentrated band versus time. The insets are the original imagesleft column) and the converted images (right column) for detecting the enrichedands using a customized MATLAB program.

We noticed that the enriched band became bigger over time,nd the converted shape tended to be more irregular. It might be

result of the delay in bleaching on the paper fiber, although theuorescent tracer was no longer there. This effect introduces moreoise and dispersed areas compared to that of the smaller enrichedand accumulated at the shorter time. The noise will be reducedignificantly by optimizing the threshold value as mentioned above.

.3. Cost and others

In term of the fabrication cost, the previously reported in-paperpproach [25] offers a disposable unit at low cost and potentialor scale-up. Whereas our approach used the off-the-shelf Nafion,hich has a lower material cost than Nafion solution. Besides,lastics films and adhesive tapes are also low-cost and readily avail-ble materials. An electronic cutter is also more easily accessiblehan wax printing. Thus, our proposed devices have potential to beidely used, especially for limited-resource regions. Furthermore,

ur fabrication time required only 20 min (i.e. half of the time waspent for baking process), compared to nearly 60 min reported in in-aper devices [25]. To address the stability issue of the microfluidicaper-based platform, the current device is made from a paper stripnd the plastics films. Indeed, those materials are not considereds mechanically durable and can be deformed easily. The purposef the two plastic films is not only to prevent the evaporation, butlso to protect the paper channel and make it more rigid. Exper-mentally, when performing the tests, we placed the device on aat surface. The device was stable enough to ensure proper liquidransport. Many paper-based analytical kits for point-of-care test-ng such as the pregnancy test stick are used only once. After use, theaper device should be disposed properly. Our sandwiched device

s stable enough for a test without any significant deformations, butet cheap enough for the single use.

. Conclusions

This paper reported a simple fabrication approach for making paper-based microfluidic platform. Off-the-shelf Nafion mem-rane was integrated as a nanoporous junction on the device.he device operated for liquid sample filling by capillary suction

hrough the microporous paper and ion selective transport by DC-eld-driven ICP through the nanoporous Nafion membrane. Two

unctions of fluorescent dye sample depletion and concentrationere successfully demonstrated for our paper-based microfluidic

ators B 222 (2016) 735–740 739

device. It was shown that when a same potential was imposed tothe two reservoirs of the paper strip, the depletion zone expandsoutward from the Nafion membrane and reaches a steady stateafter 2 min. When a potential difference was applied, about 60-foldconcentration enhancement was attained within 200 s. Comparedto the recently reported ICP paper-based device performing theactive concentration and transport with the same negative fluores-cent tracer [25], our device provides a higher concentration factor.Thus more samples can be accumulated for a better visualizationand detection. The enriched band was observed to slowly moveto the reservoir of higher potential with an increasing band size.The physical mechanisms accounting for the sample depletion andenrichment including ICP, electrophoresis and electroosmosis werediscussed. The lamination approach presented in this paper enablesthe integration of the concentrator with other paper-based analyt-ical devices. In addition, the fabrication technique has the potentialto be transferred to a large-scale production with role-to-role tech-nology.

Acknowledgements

D.T.P acknowledges the support from the Nanyang Technologi-cal University PhD Scholarship via Nanyang Environment & WaterResearch Institute (NEWRI). The authors thank to the support onequipment from Tuan Tran (NTU).

References

[1] Q. Pu, Nano Lett. 4 (2004) 1099–1103.[2] S.J. Kim, Y.-A.A. Song, J. Han, Chem. Soc. Rev. 39 (2010) 912–922.[3] S.H. Ko, S.J. Kim, L.F. Cheow, L.D. Li, K.H. Kang, J. Han, Lab Chip 11 (2011)

1351–1358.[4] S.H. Ko, Y.-A. Song, S.J. Kim, M. Kim, J. Han, K.H. Kang, Lab Chip 12 (2012)

4472–4482.[5] R. Kwak, S.J. Kim, J. Han, Anal. Chem. 83 (2011) 7348–7355.[6] S.J. Kim, S.H. Ko, K.H. Kang, J. Han, Nat. Nanotechnol. 5 (2010) 297–301.[7] B.D. MacDonald, M.M. Gong, P. Zhang, D. Sinton, Lab Chip 14 (2014) 681–685.[8] O. Jannig, N.-T. Nguyen, Microfluid. Nanofluid. 10 (2011) 513–519.[9] D. Kim, A. Raj, L. Zhu, R.I. Masel, M.A. Shannon, Lab Chip 8 (2008) 625–628.10] J. Lichtenberg, E. Verpoorte, N.F. de Rooij, Electrophoresis 22 (2001) 258–271.11] H. Cui, K. Horiuchi, P. Dutta, C.F. Ivory, Anal. Chem. 77 (2005) 3451–3457.12] S. Song, A.K. Singh, B.J. Kirby, Anal. Chem. 76 (2004) 4589–4592.13] R. Oleschuk, L. Shultz-Lockyear, Y. Ning, D. Harrison, Anal. Chem. 72 (2000)

585–590.14] Z. Ge, W. Wang, C. Yang, Lab Chip 11 (2011) 1396–1402.15] S.J. Kim, L.D. Li, J. Han, Langmuir 25 (2009) 7759–7765.16] S.J. Kim, Y.-C. Wang, J. Lee, H. Jang, J. Han, Phys. Rev. Lett. 99 (2007) 044501.17] A.K. Yetisen, M.S. Akram, C.R. Lowe, Lab Chip 13 (2013) 2210–2251.18] X. Li, D.R. Ballerini, W. Shen, Biomicrofluidics 6 (2012) 12–14.19] W. Dungchai, O. Chailapakul, C.S. Henry, Anal. Chem. 81 (2009) 5821–5826.20] A.W. Martinez, S.T. Phillips, G.M. Whitesides, E. Carrilho, Anal. Chem. 82

(2010) 3–10.21] N. Chandra, S. Ali, M. Shaegh, S. Huan, L. Ge, S. Ngin, Sens. Actuators: B. Chem.

204 (2014) 414–420.22] A.K. Ellerbee, S.T. Phillips, A.C. Siegel, K.A. Mirica, A.W. Martinez, P. Striehl, N.

Jain, M. Prentiss, G.M. Whitesides, Anal. Chem. 81 (2009) 8447–8452.23] J.L. Delaney, C.F. Hogan, J. Tian, W. Shen, Anal. Chem. 83 (2011) 1300–1306.24] B.J. Toley, J.A. Wang, M. Gupta, J.R. Buser, L.K. Lafleur, B.R. Lutz, E. Fu, P. Yager,

Lab Chip 15 (2015) 1432–1444.25] M.M. Gong, P. Zhang, B.D. MacDonald, D. Sinton, Anal. Chem. 86 (2014)

8090–8097.26] D.G. Yan, C. Yang, X.Y. Huang, Microfluid. Nanofluid. 3 (2007) 333–340.27] D.-T. Phan, C. Yang, N.-T. Nguyen, RSC Adv. 5 (55) (2015) 44336–44341.

Biographies

Dinh-Tuan Phan is a PhD student with School of Mechanical and Aerospace Engi-neering at Nanyang Technological University. He is doing research on the fabricationof nanofluidic devices and nanofluidic phenomena, especially ion concentrationpolarization.

Seyed Ali Mousavi Shaegh completed his PhD in Mechanical Engineering on June2012 at Nanyang Technological University. After his PhD, he joined SingaporeInstitute of Manufacturing Technology (SIMTech, A*STAR), division of MicrofluidicManufacturing Program. For two years, he had contributed and led projects for fabri-cating robust designs of microfluidic functional elements, chips, and electrochemical

7 d Actu

bHoih

CovT

40 D.-T. Phan et al. / Sensors an

iosensors for start-ups and commercial applications. Since April 2014, he joinedarvard-MIT Division of Health Sciences and Technology, Massachusetts Institutef Technology to be involved in Organ-on-a-chip Project with the goal of develop-ng a microfluidic platform with integrated sensors for drug screening on differentuman constructs.

hun Yang obtained his B.S. degree from the Department of Thermal Engineeringf Tsinghua University (Beijing). He then headed his postgraduate studies in Uni-ersity of Science and Technology of China (Hefei) and received his M.S. degree inhermophysics. Before he pursued his Ph.D. degree in Mechanical Engineering from

ators B 222 (2016) 735–740

University of Alberta (Edmonton), he had worked in Shanghai Institute of ElectricPower for six years He is currently an Associate Professor at Nanyang TechnologicalUniversity.

Nam-Trung Nguyen is a professor and the director of Queensland Micro- and Nano-

technology Centre at Griffith University in Australia. He received his Dip-Ing, Dr Ingand Dr Ing Habil degrees from Chemnitz University of Technology, Germany, in 1993,1997 and 2004, respectively. In 1998, he was a postdoctoral research engineer in theBerkeley Sensor and Actuator Center (University of California at Berkeley, USA). ProfNguyen is a Fellow of ASME and a Member of IEEE.