Paper electrodes for bioelectrochemistry: Biosensors and biofuel cells

Biosensors and Bioelectronics ∎ (∎∎∎∎) ∎∎∎–∎∎∎ Contents lists available at ScienceDirect Biosensors and Bioelectronics journal homepage: www.elsevie...

Download PDF

8MB Sizes 5 Downloads 191 Views

Report

Full Text

Biosensors and Bioelectronics ∎ (∎∎∎∎) ∎∎∎–∎∎∎

Contents lists available at ScienceDirect

Biosensors and Bioelectronics journal homepage: www.elsevier.com/locate/bios

Paper electrodes for bioelectrochemistry: Biosensors and biofuel cells Cloé Desmet, Christophe A. Marquette, Loïc J. Blum, Bastien Doumèche n GEMBAS (Génie Enzymatique, Membranes Biomimétiques et Assemblages Supramoléculaires) – ICBMS UMR 5246-Université Lyon 1-CNRS-INSA Lyon-CPE Lyon, 43 bd du 11 Novembre 1918, 69622 Villeurbanne Cedex, France

art ic l e i nf o

a b s t r a c t

Article history: Received 30 April 2015 Received in revised form 21 June 2015 Accepted 22 June 2015

Paper-based analytical devices (PAD) emerge in the scientiﬁc community since 2007 as low-cost, wearable and disposable devices for point-of-care diagnostic due to the widespread availability, longtime knowledge and easy manufacturing of cellulose. Rapidly, electrodes were introduced in PAD for electrochemical measurements. Together with biological components, a new generation of electrochemical biosensors was born. This review aims to take an inventory of existing electrochemical paperbased biosensors and biofuel cells and to identify, at the light of newly acquired data, suitable methodologies and crucial parameters in this ﬁeld. Paper selection, electrode material, hydrophobization of cellulose, dedicated electrochemical devices and electrode conﬁguration in biosensors and biofuel cells will be discussed. & 2015 Elsevier B.V. All rights reserved.

Keywords: Biofuel cells Biosensors Cellulose Electrochemistry ePAD Paper-based analytical devices Screen-printing Wax printing

1. Introduction Paper (and paper-related) is probably one of the oldest manufactured material. From the early China and Egypt to the ﬁrst book impression, paper is a support for transmission of knowledge. During the industrial revolution, paper together with the advances in printing has allowed the access to information to nearly everyone (as long as they were able to read). This has even been more popular with the access to personal printer and personal computers. In the scientiﬁc world, paper publications is the basis of scientiﬁc spreading and, despite most of the resources are nowadays electronically available, articles are still often printed by scientists. Paper itself has also lead to high achievement in science as depicted by the famous Whatman paper #1 found in every lab in the world. The second part of the 20th century was the reign of silicon, when computers and the World Wide Web (internet) replaced progressively newspapers. Even in the lab, paper ﬁlters are replaced by polymeric beads, silica membranes for chromatography or by organic ﬁlters. One could expect that the age of paper should end soon but it is without considering the exhaustion of non-renewable resources and of the availability of silicon of high purity. Therefore paper will probably remain an important material for data storage and dissemination in the next years. n

Corresponding author. E-mail address: [email protected] (B. Doumèche).

Paper could offer some new alternatives and opportunities as material because it beneﬁts for centuries of accumulated knowledge. It is obtained from renewable (inextinguishable) resources and could be manufactured in different form (ﬁlter paper, glossy paper, kraft paper …) to achieve the desired properties. It could be as strong as cardboard or as light as rice paper. One of its main interests is its low price and high availability worldwide. Interestingly, the most important application of paper in the scientiﬁc community during the last ﬁve years was the development of single-use and wearable point-of-care diagnostic devices (POCD). The key publication in this ﬁeld is the assembly in 2007 of a microﬂuidic paper-based analytical device (PAD) by members of the Whitesides' group (Martinez et al., 2007). In this work, microﬂuidic channels and individual test zones were designed in a chromatography paper using a photoresist to allow the simultaneous colorimetric detection of glucose (using glucose oxidase and horseradish peroxidase) and protein (using a tetrabromophenol blue as protein speciﬁc dye) in the same sample. This work has yet unveiled the main interests of what is now called microﬂuidic paper-based analytical devices (mPAD): sample volume is as low as 5 mL, multiplex analysis of single sample is possible in a short time range (about 11 min for all analytes), solid contaminants do not migrate on the test zones and the overall device cost is low. Moreover, it could be used by non-trained person, it does not require any pumping system or speciﬁc reader and it is easily wearable. Therefore, these simple assays are suitable for emerging

http://dx.doi.org/10.1016/j.bios.2015.06.052 0956-5663/& 2015 Elsevier B.V. All rights reserved.

Please cite this article as: Desmet, C., et al., Biosensors and Bioelectronics (2015), http://dx.doi.org/10.1016/j.bios.2015.06.052i

2

Analyte

Electrode

Detection

Paper

Whatman #1 Whatman #1 Whatman #1

Amperometry Ascorbic acid, uric acid, paracetamol Uric acid, ascorbic acid

Pencil drawn electrode

[Fe(CN)6]3

Gold

Direct oxidation

H2O2

SPE

Prussian Blue

Glucose Uric acid Lactic acid Glucose

SPE

Prussian Blue

Iron Glucose

None (colorimetric) Carbon ink

Phenanthrolin [Fe(CN)6]3

Whatman #1

Lactic acid Uric acid Single-strand DNA

Golds NPs on SPE

HRP (direct oxidation)

Glucose

Graphite pencil

Glucose oxidase/4-aminophenol

Carcinoembryonic antigen

Golds NPs on SPE

NADH oxidation

Lactate Glucose

Prussian Blue SPE SPE

Cancer cells (SK BR-3)

Platinum NPs on SPE

Lactate oxidase/H2O2 [Fe(CN)6]3 dried on paper with Glucose oxidase H2O2

Adenosine Lactate Glucose Cholesterol Xanthine Cholesterol

Photo-induced amperometry Carcinoembryonic antigen

SPE Graphene

Polyaniline/poly (vinylpyrrolidone)/Graphene on SPE

Cholesterol oxidase/H2O2

Whatman #1

CdS/ZnO CNTs on SPE

Ascorbic acid

Whatman #1 Whatman #1

ATP

CdS quantum dot and CNTs on SPE Single wave voltammetry or cyclic voltammetry ssDNA Gold nanoparticles on carbon Thrombin Graphene oxide/chitosan Carcinoembryonic antigenalphaon SPE fetoprotein Cancer antigen 125Carbohydrate antigen 153 Glucose SPE Carcinoembryonic antigen Cysteine

Glucose oxidase [Fe(CN)6] concentration difference H2O2

3 /4

Whatman #1 Whatman #1 Whatman #1 Cotton Whatman #1 Whatman #1 Whatman #1

CNTs/Chitosan on screenprinted electrodes Doped pencil

H2O2/Gold

Methylene Blue

Dynamic range

Sensitivity

Stability

3.6 mM

3.6–100 mM

1.3 μA mM 1

210 mM 360 mM 1380 mM

2–100 mM 5–35 mM 2–50 mM 0.5–5 mM

64 μA mM 1 6 μA mM 1 40 μA mM 1 1 mA mM 1

0–20 mM

0.041 mA mM 1

0.35 mM 1.76 mM 0.52 mM 6.3 fM

0–25 mM 0–10 mM 10.0 fM to 100 nM

0.38 mM 0.85 pg mL

Dungchai et al., 2011

0.3 mM 0.22 mM

0.0076 mA mM 0.048 mA mM 1

0.001–1000 ng mL 1 0.1–5 mM 0.43 mA mM 1

Whatman #1

Glucose oxidase /tetrathiafulvalene HRP/phenylene diamine

Japanese paper Whatman #1 Whatman

Zhao et al., 2013

1

0.01–1.5 mM 1

0.3169 mA mM 1

Wang et al., 2014a 5 days at 2–8 °C Santhiago and Kubota, 2013 Wang et al., 2014b Malon et al., 2014 Nie et al., 2010b

0.10 nM–1.0 mM 11.8 mM

0–250 mM

Liu et al., 2014b 0.48 mA mM

1

0.1–15 mM

13.1 nA mM

0.3 mM 0.3 mM 0.3 mM 1 mM

0.1–15 mM 0.1–15 mM 0.1–15 mM 0.05–10 mM

13.3 nA mM 1 29.3 nA mM 1 14.6 nA mM 1 35 mA mM 1 cm ²

4 pg mL 1

0.01–50 ng mL 1

0.2 pM

1–1000 pM

27 nA pM 1

001– 100 ng mL 1(CEA) 1–100 mM

0.01 ng mL 1

0.055 mA mM 1

0.05–50.0 ng mL 1 0.5–10 mM

Liu et al., 2012b

1

0.3 mM

16 nM 0.01 ng mL 1

Reference

Dossi et al., 2013a,, 2013b Carvalhal et al., 2010 Dungchai et al., 2009

64–152 nA mM 1

0.2 mM

30 nM

HRP/tetramethylbenzidineHRP/ phenylene diamine

Co(II) phtalocyanin

LOD

0.92 mA mM 1

Labroo and Cui, 2014

Ruecha et al., 2014

4 weeks at 4 °C

Wang et al., 2013a Ge et al., 2013a

Four weeks under nitrogen

Cunningham et al., 2014

3 weeks at 4 °C

Wu et al., 2014b,, 2013b Shitanda et al., 2013b Wang et al., 2012a Dossi et al., 2014

C. Desmet et al. / Biosensors and Bioelectronics ∎ (∎∎∎∎) ∎∎∎–∎∎∎

Please cite this article as: Desmet, C., et al., Biosensors and Bioelectronics (2015), http://dx.doi.org/10.1016/j.bios.2015.06.052i

Table 1 Analytical characteristics of ePAD used as biosensors.

Gold NPs on SPE

HRP/H2O2/thionine

NADH

SPE

Direct NADH oxidation

Single strand DNA

Gold NPs on SPE

Single strand DNA/Thionine

Cancer antigen 125 Carcinoma antigen 199 Prostate protein antigen Carcinoembryonic antigen

þ

Chitosan and Silver NPs on Ag SPE Cu2 þ MnO2 and Au NPs on SPE Glucose oxidase/ tetramethylbenzidine Gold on SPE Methylene Blue

Whatman #1 Whatman #1 Whatman #1 Whatman #1 Whatman #1 Whatman #1

AFP (alpha-fetoprotein) Human troponin I

PANI on SPE

Ferrocene carboxylic acid [Fe(CN)6]3 /4

Interleukin-2 receptor alpha

PANI on SPE

[Fe(CN)6]3 /4

Electrochemiluminescence 2-(dibutylamino)ethanol

Commercial SPE

Ru(bpy)32 þ

Whatman #4

Gold and graphene oxide/ chitosan on SPE

Tripropylamine/P-acid

Whatman #1

SPE

Ru(bpy)32 þ /Tripropylamine

Whatman #1

Naﬁon/graphene Oxide/ Fe3O4 on SPE Naﬁon/graphene Oxide/ Fe3O4 on SPE

Ru(bpy)32 þ /Tripropylamine

Whatman #1 Whatman #1

NADH Prostate protein antigen Carcinoembryonic antigen Carcinoembryonic antigen

DNA mismatches DNA mismatches

Ru(bpy)32 þ

Trypropylamine ATP

Golds NPs on SPE

Ru(bpy)32 þ /Tripropylamine SnO2 QDs/Luminol

Carcinoembryonic antigen

CNTs/Chitosan on SPE

Ru(bpy)32 þ /trypropylamine

0.004 μg mL 1

0.5–10 mM 1.22 mA mM 1 0.01 to 200.0 μg mL 1

1.8 mM

10–100 mM

0.2 fM 0.02 U mL 0.04 U mL

1

0.5 pg mL 1

Whatman #1

alpha-fetoprotein Cancer antigen 199 Carcinoma antigen 153 Carcinoembryonic antigen

Silver on SPE

QdTe quantum dots/K2S2O8

Whatman #1

Others Human immunoglobulin

CNTs

Resistivity

Filter paper

Carbon nanodots

0.1–100 U mL

1

90% at 4 °C for 30 days 21 days at 4 °C

1

0.1–100 U mL 0.0012 ng mL 1

2.56 mA mL U

1

Metters et al., 2013 Lu et al., 2012 Li et al., 2014c

1

0.91 mA mL U 0.005–100 ng mL 1

0.001–100 ng mL 1

4 week at 4 °C

Li et al., 2014b

10 days at 4 °C

Li et al., 2013a

1

Whatman #1 Whatman #1

Whatman #1

15 mA mM 1

0.0008–500 pM 1

0.8 pg mL

Ge et al., 2013b

1–1000 ng mL 1

5.5 μA ng mL–1 cm 2

0.5–3 ng mL 1

737 μA ng mL–

0.9 mM

3–5000 mM

72 mM 1 pg mL 1

0.2–10 mM 0.003–20 ng mL 1

0.8 pg mL 1 0.001 ng ml 1 (0.008 ng ml 1 in serum) 0.5 nM

0.001–10 ng mL 1 0.005–50 ng mL 1

0.1 nM to 0.5 mM

1 nM

10 nM to 5 mM

5 nM 0.025 pM

10 nM to 10 mM 0.1 pM to 100 nM

4.0 pg mL 1 0.02 ng mL 1 6.0 mU mL 1 5.0 mU mL 1 0.12 pg mL 1

1

Jagadeesan et al., 2012 Kumar et al., 2013

cm 2

Delaney et al., 2011 85% after 1 month at 4 °C

Li et al., 2013c

3 weeks at 4 °C

Yan et al., 2012

Xu et al., 2014 80% remaining activity after 90 days

Xu et al., 2013

5 weeks under ambient conditions

Wang et al., 2014c Wang et al., 2012b

0.5 pg mL 1 to 20 ng mL 1

C. Desmet et al. / Biosensors and Bioelectronics ∎ (∎∎∎∎) ∎∎∎–∎∎∎

Please cite this article as: Desmet, C., et al., Biosensors and Bioelectronics (2015), http://dx.doi.org/10.1016/j.bios.2015.06.052i

#1 Hydrogen peroxide Microcystin-LR

4 days at 4 °C in Li et al., 2013b PBS Pozuelo et al., 2013

CNTs: Carbon nanotubes; SPE: Screen-Printed Electrode; HRP: Horseradish peroxidase.

3

4

C. Desmet et al. / Biosensors and Bioelectronics ∎ (∎∎∎∎) ∎∎∎–∎∎∎

countries (less-industrialised) for detecting disease or monitoring health using easily available biological sample (blood, urine, saliva, tears…). Further thinking about using paper has enlightened other unique advantages of PAD over other types of analytical devices. First, they could be manufactured locally all over the world from renewable sources, at low cost and high-volume (Yetisen et al., 2013). Compared to glass, which could also be produced worldwide, paper is about 1000 times cheaper. Second, paper could be printed or impregnated using polymeric inks, chemicals, biomolecules (Enzymes, antibodies, DNA) or particles using yet mature screen-printing or spotting technologies. Third, paper is biodegradable but, when spoiled by biological material, it could be easily incinerated. It is also considered as a biocompatible material. Finally, the porous structure of paper acts as an autonomous microﬂuidic pumping system. During the last years a tremendous number of articles was published using PAD but, as mentioned in the review of Nery and Kubota (2013), bibliographic studies could be arduous because the term “paper” is also used for scientiﬁc manuscript and the sentence “This paper describes…” is retrieved in nearly any manuscript published. Therefore, abbreviations have emerged such as mPAD, microPAD, oPAD (origami), ePAD (electrochemical) or EmPAD (electrochemical-microﬂuidic paper-based analytical device). More complex denominations are also retrieved such as m-OECI (origami electrochemical immunodevice) or rEPAD (referenced electrochemical paper-based analytical device). In the present review, we will only use the terms mPAD or ePAD for a matter of clarity. Paper-based biofuel cells will also be referred as ePAD to prevent the use of additional abbreviations. Among the different detection systems usually coupled with paper microﬂuidic platforms, the electrochemical detection is one of the most promising due to its easy miniaturisation, speed, simplicity, high sensitivity and quantitative response, making ePAD a powerful tool for different applications ranging from environmental monitoring to healthcare diagnostics. PADs, which can be electrochemical or optical sensors and biosensors were previously extensively reviewed on the basis of one of their characteristic. Point-of-care diagnostics using PAD were reviewed by Yetisen et al. (2013) and Hu et al. (2014) and speciﬁc application in infectious disease by Rozand (2014) or urine analysis by Jeong et al. (2013). Liu et al. (2014a) reviewed paperbased electrochemical biosensors while paper-based batteries were referenced by Nguyen et al. (2014). Detection methods were reviewed by Nery and Kubota (2013). Chemiluminescence (CL)based lab-on-chip, including PAD, were also recently reviewed (Mirasoli et al., 2014). More speciﬁc subjects for PAD manufacturing such as inkjet printing (Komuro et al., 2013), PAD fabrication (Coltro et al., 2014) or biomolecules immobilisation on paper (Kong and Hu, 2012) were also reviewed. The present review will focus on electrochemical paper-based devices (ePAD) including a biological component and more especially on the electrodes themselves. The advantages of paper-based electroanalytical devices over paper-based colorimetric assays will be ﬁrst described. This review takes an inventory of paper origin, electrode material and outline the interest of paper to support tridimensional electrodes and to build hydrophilic channels. Some special features of paper such are chromatographic properties and stabilisation of proteins are also discussed. In the end, this topic includes paper-based biofuel cells, electrochemical and also electrochemiluminescent biosensors. 2. From colorimetric strip tests to paper-based electroanalytical devices Paper-based systems using lateral ﬂow are widely used for

colorimetric rapid assays. Biological samples (blood, urine and sperm) are thus analysed for example for pregnancy, virus and bacteria presence. Colour interferences from the sample (e.g. haemoglobin in the urine) could lead to unusable results. On the mPAD side, colorimetric detection using visual inspection usually lead to non-quantative analyte concentration. To bypass this limitation, mPAD scan or photograph should be preferred in conjunction with image analysis software (on place or mailed by electronic means) (Dungchai et al., 2009). Therefore, the use of an electrochemical measurement could overcome this situation because results are obtained directly as a digit number. Electrochemical devices only require to measure a current intensity or a variation in potential between two electrodes as long as electroactive molecules could be produced locally. Electrochemical methods are usually faster and more sensitive than colorimetric methods, reaching the nanomolar range (corresponding to some pmol of analytes using some mL of sample). Detection is independent of ambient light (except for electrochemiluminescence measurements) and insensitive to dye bleaching. Paper (cellulose) itself is not electroactive and electrochemical interferences should be low. Nevertheless, additional conductive or semi-conductive material should be considered when using electrochemistry compared to colorimetry and this should be included in the cost of the assay.

3. Choice of the paper Paper is the main component of ePAD. Its structure will control i) the ﬂuidic properties of the ePAD, ii) the ability to separate the analytes from the biological matrix and iii) the electrode 3D geometry. Interestingly, most of the ePAD were based on Whatman #1 paper (Tables 1 and 2). This choice is probably primary driven by the availability of this paper in lab and because a dry ﬁlter paper could absorb more than its dry mass of aqueous solution. Electron microscopy image of Whatman #1 paper showing the cellulose ﬁbres structure is presented in Fig. 1-A. Nie et al. (2010b) determined the diffusion coefﬁcient of ferrocene carboxylic acid in ePAD made of Whatman #1 paper and found 4.3 10 6 cm2 s 1 which is fairly close to the reported value of 5.7 10-6 cm2 s 1 in water. Santhiago and Kubota (2013) determined the diffusion coefﬁcient of [Fe(CN)63 /4 ] using a pencil graphite electrode pressed on a Whatman paper. The value (7.3–7.6 10 6 cm2 s 1) is close to the diffusion coefﬁcient determined for the same probe in solution, meaning there is no (or few) diffusion constraint in Whatman paper #1. Using the same paper, Dungchai et al. (2009) showed that the redox signal of Prussian blue particles included in a screen-printed electrode is limited by the diffusion of potassium counter ions, found to be slower than in aqueous solution. Ruthenium bypyridyl ([Ru(Bpy)3]2 þ ) was used by Delaney et al. (2011) to determine diffusion coefﬁcient in Whatman #4, a paper with higher ﬂow rate and porosity but similar diffusion coefﬁcient values (3.9 10 6 cm2 s 1), showing that in all these cases, diffusion of small molecules in hydrated Whatman papers is similar to what can be observed in aqueous solution at the electrode boundary. Nevertheless, some ePAD were constructed using other paper material. Nie et al. (2010a) used Whatman paper #1 or polyester– cellulose blend but, unfortunately, did not discuss of any signiﬁcant difference between the two papers. Japanese paper is claimed to have higher water absorbency than qualitative Whatman paper #2 and was preferred for its theoretical higher ﬂuidic properties (Shitanda et al., 2013a). Thus, Shitanda et al. (2013b) proposed the use of Japanese paper for the construction of a glucose biosensor based on screen-printed carbon electrodes (Fig. 1B). Electrochemical impedance spectroscopy (EIS) was used to

Please cite this article as: Desmet, C., et al., Biosensors and Bioelectronics (2015), http://dx.doi.org/10.1016/j.bios.2015.06.052i

Fraiwan et al., 2013 15 min

Laccase (O2)

[Fe(CN)6]3 /4

Aldose dehydrogenase-(Glucose)

Shewanella oneidensis

CNTs: Carbon nanotubes; HRP: Horseradish peroxidase; TTF: Tetrathiafulvalene; ABTS: 2,2′-Azinobis [3-ethylbenzothiazoline-6-sulphonic acid]-diammonium salt.

0.31 5.5 74

0.45 10 3.5 Whatman #1

0.55 300 Japanese paper 120 Bilirubin oxidase (O2) Glucose oxidase (Glucose)

TTF (anode) on porous carbon ink (Ketjen Black) Carbon inks supplemented by CNTs and mediators: osmium complexes (anode) and ABTS (Cathode) Carbon cloth

0.6 135 Whatman #1 Gold nanoparticles on carbon paste

225

0.59 430 150 Whatman #1

Platinum nanoparticles (O2) Bilirubin oxidase (O2) Glucose oxidase and horseradish peroxidase (Glucose) Glucose dehydrogenase (Glucose)

Gold nanoparticles on carbon paste

0.61 79 34 Cellulose-CNTs Bilirubin oxidase (O2) Fructose dehydrogenase (Fructose)

Filter paper

0.56 38 13.5 Whatman #1 Bilirubin oxidase (O2) Glucose dehydrogenase (Glucose)

CNTs on carbon paste

Toray paper Buckyeye paper Air (O2) Platinum (Half-cell)

Whatman #1

Whatman #1

Ciniciato et al., 2012 2 0.75 h Zhang et al., 2012 OCV drop from 0.6 to Wu et al., 2013b 0.4 V in 2 days One month at 4 °C Wang et al., 2014a n.d. Wang et al., 2014b n.d. Shitanda et al., 2013a 137 h at 0.2 V Jenkins et al., 2012 0.5–0.51 200 730 4757 90

600 min

Voltage (OCV) Electrode material Cathode (Substrate) Anode (Substrate)

Table 2 Electrochemical characteristics of paper-based biofuel cells.

Paper

Power density (mW cm ²)

Current density (mA cm 2)

Stability

Reference

C. Desmet et al. / Biosensors and Bioelectronics ∎ (∎∎∎∎) ∎∎∎–∎∎∎

5

determine the diffusion coefﬁcient of glucose into the sensor; a value of 1.2 10 6 cm2 s 1 was found (vs. 6.0 10 6 cm2 s 1 in aqueous solution). Despite the fact that the sensor performance was not one of the best, this study was one of the few considering intrinsic diffusion coefﬁcient in ePAD. This Japanese paper was later applied for the assembly of a biofuel cell (vide supra). In some sensors (non-biological sensors), paper was used only as chromatographic material, subsequently pressed on electrodes for the chronoamperometric detection of 4-aminophenol and paracetamol (Shiroma et al., 2012). Despite the fact that a nonbiological sensor was used in this study, this work should be mentioned because it is one of the few which compares two Whatman paper materials: Whatman #1 and Whatman P81, an anionic cellulose containing covalently bound phosphate groups. This paper allowed the separation of cationic molecules (here 4-aminophenol and paracetamol). This work suggests that Whatman #1 paper may not be suitable for all applications. The anionic nature of this paper should also be considered for improvement of enzyme immobilization in biosensor/biofuel cell systems or for reduction of electrochemical interferences due to biological material. Unfortunately, Whatman P81 is estimated to be 10 times more expensive than Whatman #1, and is actually discontinued. No examples of mPAD or ePAD were found using cationic paper. Nitrocellulose is another well-known derivative of cellulose obtained by treatment with sulphuric and nitric acid and has usually more hydrophobic properties than cellulose because it is supplied as blend with cellulose acetate. Lankelma et al. (2012) described a paper-based analysis system for the electrochemical detection of glucose in urine. In this work, a urine sample was ﬂowed through a nitrocellulose membrane containing immobilized glucose oxidase. This membrane was in contact with a platinum working electrode, and a continuous ﬂow of buffer solution was driven through the nitrocellulose membrane. The glucose concentration in the urine sample was obtained thanks to the amperometric detection of the hydrogen peroxide produced by the enzymatic glucose oxidation reaction. A more systematic study of different paper material was performed by Metters et al. (2013). Paper-based screen-printed electrodes (SPE) were prepared with either A4 text, graphic inkjet paper (160 g m 2, IP-SPE), A4 lined pad paper (80 g m 2, RP-SPE) or Filter Paper QL 100 (FP-SPE). The IP-SPE gave the best results in term of electrochemical response (compared to RP-SPE and FPSPE) due to the high absorbency of the paper at the connector level. The IP-SPE showed similar behaviour than a commercial screen-printed electrode (using a solid plastic substrate) for the direct oxidation of NADH or nitrite from an environmental sample (canal water solution). Unfortunately, there was no criterion identiﬁed to justify the choice of a paper over another and SPE were used as classical electrodes, e.g. plunged into mL-solution. Therefore, no eventual chromatographic separation properties of the paper were investigated. As cellulose is the main component of paper, Mihranyan et al. (2012) compared 3 nano-cellulose (NC) materials as support for polypyrrole (Fig. 1-Ea, Eb). NC is obtained after cellulose acid hydrolysis. Therefore it could be a model of choice to determine suitable paper for a speciﬁc application or to develop paper-electrodes at the m-metre scale. The formation of polypyrrole increases the pore size as well as the surface area of the electrode but also decreases the mechanical strength of the composite. Despite this study is only dedicated to material sciences, it opens the way of using cellulose material as a support for electrolymerization of organic monomers. Another cellulose-based material readily available is cotton fabric which is related to paper but treated differently to obtain a soft material. In the work of Malon et al. (2014), carbon (containing Prussian blue mediator) and silver chloride electrodes

Please cite this article as: Desmet, C., et al., Biosensors and Bioelectronics (2015), http://dx.doi.org/10.1016/j.bios.2015.06.052i

C. Desmet et al. / Biosensors and Bioelectronics ∎ (∎∎∎∎) ∎∎∎–∎∎∎

6

Fig. 1. Scanning electron microscopy images of (A) Whatman-paper cellulose (Wang et al., 2014a), (B) Japanese paper (Shitanda et al., 2013a), (C) AuNPs on cellulose ﬁbres (Li et al., 2014b), (D) nanoporous silver on cellulose ﬁbre (Li et al., 2013c), (Ea) nanocristalline cellulose and (Eb) nanocristalline cellulose modiﬁed by polypyrrole (Mihranyan et al., 2012), and (F) cellulose ﬁbres after operation in a microbial biofuel cell (Fraiwan et al., 2013). The inset shows a magniﬁcation of a microorganism.

were obtained by patterning the electrodes with an opened plastic foil and delimitating the sensor by ironing a wax impregnated paper. This sensor was used to detect lactate in saliva after adsorption of lactate oxidase onto the carbon electrode. The authors proposed that cellulose could act as material for the transport of lactate to the electrode by passive diffusion, without requiring the manipulation of pipettes. This study is at the frontier between ePAD and wearable textiles or intelligent clothing recently reviewed (Windmiller and Wang, 2013). The term “paper” is sometimes used for non-cellulose based materials and refers to a sheet of thin material. In ePAD, the most confusing material is Toray paper which is a Teﬂon-based material composed of carbon ﬁbres and used as “classical” carbon electrode material in some paper (cellulose)-based biofuel cells. In the ﬁeld of biofuel cells, bucky paper is often used as the electrode material (Strack et al., 2013; Villarrubia et al., 2014). Bucky paper is a compression of CNTs (Carbon Nano-Tubes) suspension (single-walled or multi-walled) onto a porous membrane resulting in a thin ﬁlm of CNT. However, despite the word “paper” is used, bucky paper itself is not made of cellulose and should be considered as a totally different material than a printed electrode on a ﬁlter paper. It was nevertheless used as electrode material once pressed on Whatman paper sheet in some biofuel cell systems (Villarrubia et al., 2014; Ciniciato et al., 2012).

4. Electrode material PAD aims to be cheap, single-used devices. Colorimetric-based PAD only requires visual inspection but in ePAD, electrodes have to be included. For centuries, paper is the support for writing. Therefore, the more logical method to obtain an electrode is to draw it on the paper surface. Commercial 3B grade pencil is basically graphite entrapped in clay or wax. Higher grade (B) correspond to higher graphite content (darker) and thus to higher conductivity. Nevertheless, electrode geometry is fully restricted by the drawer artistic quality. Graphite pencils were used to detect ascorbic acid, dopamine and paracetamol (Dossi et al., 2013a, 2013b). Further work of this group showed that mediators could be included in a paste composed of carbon powder (conductive material), sodium bentonite binding agent and potassium silicate (hardening agent) to

obtain doped pencils. Two mediators were used, decamethylferrocene and cobalt (II) phtalocyanine, for the detection of thiol compounds or hydrogen peroxide, respectively (Dossi et al., 2014). Classical pencil manufacturing requires heating at 800– 1000 °C to remove water and homogenise graphite and binder as a hard component. When organic and thermally unstable mediators are used, this approach is not suitable but hardening could be obtained using potassium silicate as hardening agent. Pencil could be also directly used as electrode, without drawing on the paper surface. Santhiago and Kubota (2013) built a working electrode using a type H cylindrical pencil further glued using epoxy resin in a glass capillary. This electrode was pressed like a pen on the paper prior to the measurement. Gold wires and carbon ﬁbres can be used as electrodes in ePAD once pressed carefully on a paper. A gold microwire modiﬁed by 11-mercapto-undecanoic acid placed perpendicularly to a hydrophilic channel was used as working electrode by Fosdick et al. (2014). This removable electrode could be placed anywhere in the channel according to the need but also required careful handling which limits its use by non-skilled person. Nevertheless, neither pencil based electrodes nor gold wires were used yet in an ePADbased biosensor or biofuel cell. This could be explained by the frequent immobilization of the biological component at the electrode surface in biosensors and biofuel cells. The preferred electrode materials remain inkjet-printed, screen-printed or stencil-printed carbon pastes. Advantages of screen-printed electrodes were recently reviewed (Hayat and Marty, 2014; Komuro et al., 2013; Li et al., 2012; Taleat et al., 2014) but they could be summarised as low cost, compatible with mass production, disposable, portable and permitting mediators' inclusion (with limited stability). Most examples of ePAD biosensors and biofuel cells are based on screen-printed electrodes and, because many biosensors use oxidases, carbon inks including Prussian blue used for the detection of hydrogen peroxide are the most popular (Dungchai et al., 2009, 2011; Malon et al., 2014) (Table 1). Two major advantages of electrode printing for ePAD manufacturing are found in the origami conﬁguration. Working electrodes and counter/reference electrodes could be printed on the same paper (same side or different side) and, after folding, the electrochemical cell would be assembled using the hydrophilic paper as electrolyte support between each electrode (Fig. 2-A).

Please cite this article as: Desmet, C., et al., Biosensors and Bioelectronics (2015), http://dx.doi.org/10.1016/j.bios.2015.06.052i

C. Desmet et al. / Biosensors and Bioelectronics ∎ (∎∎∎∎) ∎∎∎–∎∎∎

7

Fig. 2. (A) Schematic representation of the fabrication and assay procedure for an origami ePAD. Paper is impregnated by wax (green and violet) and electrodes are screenprinted on the hydrophilic zones (Ge et al., 2013a). (a) The hydrophilic cellulose is modiﬁed by metals to obtain nanoporous silver or nanoporous gold and (b) functionalized. (c) The biological recognition element (DNA, antibody) is immobilized before (d) incubation with the sample. In the last step (e), a bioconjuguated containing a secondary antibody or a complementary DNA strand and a reporting molecule is added. The origami ePAD is closed before addition of reagent and electrochemical or light measurement. (B) An origami ePAD in (a and b) open conﬁguration and (c) closed conﬁguration priori to measurement (Li et al., 2013b). (C) Origami ePAD operating in open conformation (Santhiago and Kubota, 2013). (a) Hydrophilic zones of different sizes, (b) folded origami ePAD allowing the diffusion of substrates and products on the three hydrophilic zones and (c) unfolded origami ePAD during the measurement with an external electrode. (For interpretation of the references to colour in this ﬁgure legend, the reader is referred to the web version of this article.)

Counter and reference electrodes could also be printed on two different paper sheets pressed together for the electrochemical measurement resulting in tri-dimensional channels. The second advantage applies more speciﬁcally to biofuel cells. Electrodes could be printed on the two sides of the same paper sheet which act both as a membrane between cathode and anode and as a channel for continuous substrate supply (Shitanda et al., 2013a; Villarrubia et al., 2014) (Figs. 3 and 4). In a more recent example (Labroo and Cui, 2014), graphene nano-platelets covalently modiﬁed with an oxidase (glucose, lactate, cholesterol or xanthine oxidase) and suspended in aqueous solution were directly printed as the electrode material without additional binder. This electrode was used to detect hydrogen peroxide by applying a potential of 0.5 V.

5. Cellulose as support for conductive material Cellulose ﬁbres are organised as a 3D network directly in contact with the electrode material which could be used as a scaffold for polymerisation or deposition of redox or conductive material. This aims to increase the speciﬁc surface and the conductivity of the electrode and consequently to amplify the electrochemical response. In the work of Jagadeesan et al. (2012), electrodes screen-printed on paper were used as support for

polyaniline (PANI) electropolymerization prior to covalent anticardiac Troponin-I immobilization. The PANI-modiﬁed electrodes showed a 5-fold increase of their surface conductivity. This modiﬁed SPE was used to detect 1–1000 ng mL 1 of human troponin in the presence of [Fe(CN)6]3 /4 using a classical three-electrode system and in a relatively short time (150 s). The signal increase was attributed to the attraction of the redox probe by the positively charged human troponin. Further work of this group (Kumar et al., 2013) was focused on the immobilization of anti-sIL2Ra, an antibody directed against a leukaemia biomarker (sIL2Rα). Optimisation of the electropolymerization condition allowed a 14 times conductivity improvement corresponding to a 130-fold improvement of the sensitivity. It is worth to note that PANI could be screen-printed on paper at “high scale” by the roll-to-roll method (Makela et al., 2003). Another electropolymer, polypyrrole, was deposited onto cellulose ﬁbres (Huang et al., 2005) but not used in a biosensor or biofuel cell context (Fig. 1-Eb). In the numerous origami-ePAD described by Jinghua Yu and coworkers, the cellulose scaffold was used for gold or silver deposition (Fig. 1-D). Silver or gold nanoparticles (AgNPs or AuNPs) were adsorbed onto the ﬁbres and acted as seeds for further metal growing onto the cellulose ﬁbres. Silver or gold ions were reduced using ascorbic acid or hydrogen peroxide, respectively, leading to nanoporous silver (NPS) or nanoporous gold (NPG) material (Ge et al., 2013a; Li et al., 2013b; Wang et al., 2014c). NPS is a

Please cite this article as: Desmet, C., et al., Biosensors and Bioelectronics (2015), http://dx.doi.org/10.1016/j.bios.2015.06.052i

C. Desmet et al. / Biosensors and Bioelectronics ∎ (∎∎∎∎) ∎∎∎–∎∎∎

8

Fig. 3. Biofuel cells based on paper material. (A) Anodes and cathodes are separated by a dialysis membrane (Jenkins et al., 2012), (B) electrodes are screen-printed on both side of a Japanese paper (Shitanda et al., 2013a), (C) electrodes are screen-printed on the same side of the paper with a hydrophilic channel acting as membrane (Wu et al., 2013a), (D) air-breathing cathode (Ciniciato et al., 2012) and (Ea) single or (Eb) stacked fan-like biofuel cell painted on two-face of a Whatman paper (Villarrubia et al., 2014). Pictures are adapted from the respective reference.

biocompatible nanomaterial with a large surface area, strong absorption ability, and superior conductivity (Li et al., 2013a). Platinum nanoparticles were also used as seeds for platinum growing onto cellulose ﬁbres after reduction by ascorbic acid. Platinum was further used as an electrocatalyst for the detection of hydrogen peroxide secreted by tumour cells (SK-BR-3 human breast cancer cells) or as a cathode in a biofuel cell (Liu et al., 2014b). Sensitivity of Pt-modiﬁed cellulose is twice higher than with Au-modiﬁed cellulose. Other nanoparticles such as MnO2 nanoparticles were included into NPG (Au-modiﬁed) cellulose in order to catalyse the electro-oxidation of 3,3′,5,5′-tetramethylbenzidine (TMB), a wellknown mediator of H2O2 reduction (Li et al., 2014b). Carbon-based nanomaterials were also proposed for ePAD modiﬁcation. Graphene was dried on the surface of a screenprinted carbon electrode printed on Whatman paper (Lu et al., 2012) carefully polished prior to deposition. Afterwards, graphene was further modiﬁed by AuNPs to enhance conductivity and allow single strand DNA immobilization via gold-thiol chemistry. Suspension of carbon nanotubes (CNTs) in poly(dimethyldiallylammonium chloride) (PDDA) was also used as a positively charged conjugate to immobilise negatively charged ZnO/CdS nanoparticles (Wang et al., 2013a).

6. Hydrophilic channels Cellulose-based paper is a hydrophilic material in which channels have to be designed to favour sample migration to one or more test zones (electrodes in ePAD). The channels also allow the continuity of the electrolyte solution between the electrodes without excessive drying.

The simplest method is to cut the paper to obtain the desired channel geometry by hand or with an automated cutting device. The handling of such small paper strips is nevertheless risky, especially once hydrated and subjected to tear. It also requires a solid substrate for electrode connexion or lamination into a harder material. The most widespread method to design appropriate channel is to draw hydrophobic barriers directly on the paper. The ePAD, which does not only consist of the hydrophilic channel, could be easily manipulated thanks to the hydrophobic surfaces having higher mechanical properties than hydrated paper. Several methods of paper hydrophobization used for the fabrication of ePAD are summarized in the next paragraphs. 6.1. Photolithography Photolithography was used in the ﬁrst ePAD described by Martinez et al. (2007). Photoresist is spread over the paper using a spin-coater before drying at moderate temperature (95 °C). The photomask is applied on the paper and UV cured to polymerise the photoresist and obtain the hydrophobic scheme of the ePAD. After baking and washing with acetone, the paper microﬂuidic is exposed to oxygen plasma. However, this method requires organic solvents, expensive photoresists, and photolithography equipment. An oxygen plasma treatment is also required to create hydrophilic areas. Photolithography was also used to build biosensors (Dungchai et al., 2009; Nie et al., 2010b; Wu et al., 2014b; Xu et al., 2014) or biofuel cells (Fraiwan et al., 2013; Zhang et al., 2012). Dungchai et al. (2011) reported that photoresist residues can interfere during amperometric detection of uric acid and ascorbic acid in ePAD.

Please cite this article as: Desmet, C., et al., Biosensors and Bioelectronics (2015), http://dx.doi.org/10.1016/j.bios.2015.06.052i

C. Desmet et al. / Biosensors and Bioelectronics ∎ (∎∎∎∎) ∎∎∎–∎∎∎

9

Fig. 4. Multiplexed ePAD. (A) Assembly of a 24 electrode ePAD. (a) lower electrodes in columns, (b) sample chambers, (c) common reference and counter-electrodes, (d) upper electrodes in rows and (e) assembled ePAD (Ge et al., 2012b). (B) An ePAD composed of (a) 8 individual electrodes sharing the same (b) counter and reference electrodes (Wu et al., 2014b).

6.2. Chemical polymerization of liquid polymers Alkenyl ketene dimers (AKD) are low-cost materials used as sizing agents in the papermaking industry to provide a certain amount of hydrophobicity to the paper. Their use in paper-based devices is therefore logical. Alkenyl ketene dimer is a four-ring lactone with two hydrocarbon chains (C16–C20). Hydrophobization occurs by the formation of an ester bond between the hydroxyl group of cellulose and carbonyl group of the AKD lactone, leading to a β-Keto-ester (Kumar et al., in press). It was estimated that the cost of liquid AKD for each paper ﬂuidic device is around 2 10 6 euro (Delaney et al., 2011). 6.3. Wax impregnation The wax printing method has become the most frequently used process for quickly producing mPAD. The microﬂuidic layout is printed onto a paper surface using a wax ink printer. The wax printed paper is then heated ( 130 °C) to melt the wax, which allows it to penetrate into the paper, generating hydrophobic barrier within the paper. According to Dungchai et al. (2011), there is less electrochemical background using wax-printing compared to photolithography polymers. Moreover, it does not require any specialized equipment and the layout could be easily modiﬁed using appropriate software. A wax printer cost less than 1000 USD. The wax printing methodology was investigated by the Whiteside's group (Carrilho et al., 2009) to deﬁne the best channel

design (geometry and width) and to identify limitations of this method. In the case of ePAD, electrodes are usually screen-printed after the wax impregnation (Fig. 2-A). Wax printing is the methodology chosen by the origami-PAD developers (Fosdick et al., 2014; Lu et al., 2012; Wang et al., 2012a). In this special case, the electrode is submitted to several washing steps and the hydrophilic part of the paper is open to air on both sides. In the work of Dossi et al. (2013a, 2013b), the back side of the paper was insulated by thermal lamination of a polyethylene (PET) layer (0.1 mm) to prevent any electrolyte leakage during the analysis. The geometry could be either a single round chamber (Fig. 2-A) (Lu et al., 2012), 8 independent chambers (Fig. 5-B) (Zhao et al., 2013) or the construction of a ﬂuidic channel acting as a delay switch in a 3D origami ePAD (Fig. 6-C) (Wang et al., 2013c). Nevertheless, wax could be also simply printed using a screenprinting mask followed by paper deep impregnation on a hot plate (Dungchai et al., 2011) or with metallic block printing with appropriate form, called “movable-type wax printing” (MTWP) (Zhang et al., 2014). This was applied only to colorimetric detection of hydrogen peroxide and not to electrochemical detection, probably because the thermal curing of electrode shall also melt the printed wax, leading to a loss of the wax printed features. Wax dipping is another method used by Noiphung et al. (2013). Brieﬂy, an iron mould was applied on a paper placed on a glass slide. The mould is retained by a magnet on the opposite side of the glass slide and the assembly is dipped for 1 s in melted wax (105–130 °C). After removing the glass slide and the iron mould,

Please cite this article as: Desmet, C., et al., Biosensors and Bioelectronics (2015), http://dx.doi.org/10.1016/j.bios.2015.06.052i

10

C. Desmet et al. / Biosensors and Bioelectronics ∎ (∎∎∎∎) ∎∎∎–∎∎∎

Fig. 5. Wearable electrochemical readers for amperometric measurements with ePAD. (A) (a) ePAD with immobilized [Fe(CN6)]3 (in yellow) adapted for operation with (b) commercial glucometer (Nie et al., 2010a). (B) (a) multiplexed ePAD with eight different screen-printed electrodes for lactate, glucose and uric acid detection. (b) Wearable electrochemical device for the simultaneous measurement of eight electrodes. The device holds in a single hand (Zhao et al., 2013). (For interpretation of the references to colour in this ﬁgure legend,the reader is referred to the web version of this article.)

the hydrophilic features are obtained as an image of the mould shape. Moreover, if different paper sheets were overlapped before wax dipping, they would combine in a ﬁnal single object due to wax penetration. Using this strategy, the authors deﬁned a reaction zone as well as two sample zones.

Other wax type such as wax/grease pencils, wax crayons, candle wax, and lipsticks were also proposed in combination with various paper (printing paper, kitchen towels, napkins, and laboratory paper towels) as microﬂuidic platforms showing that wax melting should be adapted to each wax/paper combination (Zhong

Fig. 6. Self-powered biosensors using (A) capacitor or (B) paper-based supercapacitor. (A) (a) after adsorption of reagents, the ePAD is folded and laminated. (b) Sample migrate in detection zone by two different channel leading to a potential difference due to reaction in one of the channel. The potential difference created is used to charge and discharge a capacitor in a digital multimeter (Liu et al., 2012b). (B) (a) Unfolded biosensor and capacitor prepared on the same paper sheet by screen printing and wax impregnation. (a) Once folded, the two capacitor electrodes are stacked together. (C) Self-powered biosensor including a delay switch (paper channel on the green layer). Current is photo-induced by a chemical reaction (b) and capacitor (d) is charged. Once hydrated, the channel (c) closes the switch and capacitor discharged in the digital multimeter (a) (Ge et al., 2013b). (For interpretation of the references to colour in this ﬁgure legend, the reader is referred to the web version of this article.)

Please cite this article as: Desmet, C., et al., Biosensors and Bioelectronics (2015), http://dx.doi.org/10.1016/j.bios.2015.06.052i

C. Desmet et al. / Biosensors and Bioelectronics ∎ (∎∎∎∎) ∎∎∎–∎∎∎

et al., 2012). Hydrophilic channels could also be drawn in hydrophobic paper by using CO2 laser without cutting the paper itself (Chitnis et al., 2011). Finally, the simplest method is probably to draw the layer with a commercial permanent marker. The ink is usually composed of pigment, solvent (ethanol) and hydrophobic resin. After solvent evaporation, the resin forms the hydrophobic barrier delimiting the hydrophilic channels of the PAD (Nie et al., 2012). A metallic iron pattern could be used to obtain reproducible PAD onto chromatographic paper in about one minute. Moreover, this could be applied for mass production of PAD. Compared to the cutting approach, photolithography and wax or polymer impregnation increase the cost of the ePAD. Moreover, disposal of the ePAD by incineration could lead to environmental hazard that are not yet, but should be evaluated.

7. Electrochemical reader Whereas colorimetric-based mPAD can be read by visual inspection (for semi-quantitative analysis), ePAD usually requires external potentiostats. These devices are usually lab apparatus needing AC power supply or portable system working autonomously. From a cost point of view, these apparatus are not suitable for emerging countries, indeed a single way potentiostat costs at least 1000 USD. Moreover, it limits the number of sample that could be read at a time. Most of the ePAD were characterized electrochemically using lab scale potentiostats but solutions should be found for wearable and low cost applications. Again in the Whitesides' group, a commercial glucometer was diverted from its original function. This apparatus operates as a potentiostat measuring the current generated by the oxidation of [Fe(CN)6]4 formed during glucose oxidation by glucose oxidase at the end of the strip (Fig. 5-A). This enzyme can be replaced by any other oxidase such as lactate oxidase or cholesterol oxidase but also by NAD-dependant dehydrogenase (alcohol dehydrogenase in this work). EPAD were constructed to ﬁt the glucometer geometry and quantitative analysis of the four metabolites in plasma or urine were achieved in a short time ( 10 s). The glucometer should only be calibrated as sensitivity differs for each analyte (Nie et al., 2010a). Zhao et al. (2013) developed a hand-size wearable 8 channels potentiostat able to measure simultaneously 8 electrodes in a 3 electrodes set-up (working, counter and reference electrodes) (Fig. 5-B). The price of the device was estimated to be 90 USD and requires only an external computer. It can be implemented for LCD display of the results. Moreover the apparatus could operate autonomously using a single 3 V battery. This apparatus was used for the simultaneous detection of glucose, uric acid and lactic acid in aqueous solution using the appropriate oxidase and [Fe(CN)6]3 /4 as a redox mediator. More recently, the group of Jinghua Yu developed a strategy based on the use of paper-based supercapacitor and digital multimeter (DMM) (Ge et al., 2013b; Wang et al., 2012b, 2012c, 2014a, 2014c) (Fig. 6-A). In this device, a supercapacitor was formed by soaking graphite electrodes (drawn with a pencil) into hot H2SO4– PVA solution followed by solidiﬁcation at room temperature. Finally, the two H2SO4–PVA impregnated electrodes were ﬁrmly pressed together to obtain the capacitor. This capacitor was connected to the working and counter/reference electrodes of an amperometric biosensor acting in this special case as a biofuel cell. Upon recognition of the analyte, the potential difference created between the working and counter-electrode was used to charge the supercapacitor. The sensor signal was obtained by discharging the supercapacitor into a low-cost DMM (50 USD) using a simple interrupter. The maximal current intensity was correlated to the capacitor loading and therefore to the analyte concentration. Using

11

this strategy, any commercial DMM could be used and no dedicated material would be necessary which would represent an important advantage in low resource countries. Moreover, in the best biosensors described, the signal is about 13 times higher than the one obtained by classical chronoamperometry, which increases the sensitivity of the assay. In the best case, both biosensor and supercapacitor are designed on the same paper sheet reducing the number of electrical connexion and therefore the cost of the assay (Ge et al., 2013a). The only power supply required is the 1.5– 3 V battery of the DMM. In a really recent work, a complete electrochemical biosensor, including battery, digits, display and molecular recognition element was printed onto paper and polyethylene terephthalate sheets together with only two chips (Beni et al., in press). Up to 5 samples could be read simultaneously with removable screenprinted electrodes. The chronoamperometric detection of an oxidase substrate could be easily achieved and this device is wished to replace, in the near future, expensive electrochemical devices in the diagnostic ﬁeld.

8. Separation on paper The probable main interest of mPAD and ePAD is to use cellulose material as microﬂuidic channel for mixing enzymes and samples, controlling the reaction time, delivering the sample in different reaction zone (multiplex analysis) and, because Whatman paper is a chromatographic paper, to separate different components of the sample, without biological sample preparation or sample clariﬁcation. Moreover, paper prevents the formation of air bubbles in the channels. Yetisen et al. (2013) extensively reviewed the use of paper-based devices for microﬂuidic separation. They emphasised that paper-based microﬂuidic devices could succeed in microﬂuidic where PDMS-based devices do not, especially concerning the ease-to-use (no pump or precise pipetting) and elimination after contact with hazardous biological sample (incineration). In the seminal work of Nie et al. (2010b), glucose oxidase was immobilized together with its mediator ([Fe(CN)6]3 ) onto a paper channel thank to homogeneous diffusion of the enzyme solution into the paper followed by simple air-drying. The substrate sample (glucose solution) diffused into the channel and reacted all over the channel leading to the reduction of [Fe(CN)6]3 and to the anodic detection. Artiﬁcial urine was used as an example of biological ﬂuid to evaluate interferences on the linear response of the sensor. Currently, blood or urine samples should be prepared before rapid testing to reach suitable selectivity and sensibility because they are prone to matrix effects. Therefore dilution with buffer is used to reduce sample viscosity and facilitate capillary ﬂow. In order to analyse blood samples without centrifugation, Noiphung et al. (2013) introduced two Whatman blood separation papers at both side of a classical Whatman #1. Papers are combined together during the hydrophobic/hydrophilic surfaces delimitation by waxprinting. The blood sample (2 times 230 mL of unprepared blood) was deposited on the blood separation papers and the analytes reach the detection zone within 4 min (Fig. 7-A). Preparation of blood samples in a laboratory requires at least 10 min, considering the centrifugation step only. Moreover, the use of two deposition zones allows a more homogeneous sample migration to the detection zone. Here, the electrodes are commercial Prussian blue screen-printed electrodes pressed below the detection zone which allows the detection of hydrogen peroxide produced by glucose oxidase spotted beforehand in the detection zone. The electrochemical signal could be obtained by cyclic voltammetry or chronoamperometry. The sensitivity for the detection of hydrogen peroxide was not sufﬁcient using this ePAD. It could

Please cite this article as: Desmet, C., et al., Biosensors and Bioelectronics (2015), http://dx.doi.org/10.1016/j.bios.2015.06.052i

12

C. Desmet et al. / Biosensors and Bioelectronics ∎ (∎∎∎∎) ∎∎∎–∎∎∎

Fig. 7. ePAD using the ﬂuidic properties of paper. (A) Two blood separation papers are prepared as sample zones (indicated by an arrow). After whole blood deposition, serum migrates on the center of the ePAD for electrochemical measurement with a commercial SPE (seen by transparency). The red cells remains trapped in the sample zone (Noiphung et al., 2013). (B) and (C) Three or four different analytes are detected on three or four different test zones after migration from the centre of the ePAD (Dungchai et al., 2009; Labroo and Cui, 2014). (D) Dual strip for colorimetric and electrochemical measurement (Zhu et al., 2014). (For interpretation of the references to colour in this ﬁgure legend, the reader is referred to the web version of this article.)

be nevertheless used with the addition method, complicating the assay. Interferences were not observed with bilirubin (up to 18 mg mL 1) thanks to the dumbbell shape of the ePAD but the presence of haemoglobin (from haemolysed blood cells) led to a decrease of the signal by 14% (at 1 mg mL 1). Blood samples should be handled with care to prevent haemolysis, or haemoglobin should be retained in the separation zone. Zhu et al. (2014) developed an ePAD for the detection of 8-hydroxy-20-deoxyguanosine (8-OHdG), a DNA oxidative damage biomarker. After a ﬁrst recognition step of (8-OHdG) with goldAnti (8-OHdG) conjugates on the sample zone of a paper strip, free gold-Anti (8-OHdG) conjugates were captured by immobilized 8OH-dG-BSA conjugates (Fig. 7-D). Accumulation of gold particles led to a red strip which decreased in colour intensity with the 8-OHdG concentration. Gold-Anti(8OHdG)-8OHdG conjugates are not retained and diffuse to a second test zone where they are retained by pre-immobilized IgG. A second red band could also be observed. This colorimetric assay could be quantiﬁed using appropriate scanner and image analysis software but 8-OHdG could also be directly oxidised on the second test zone if a bucky paper electrode is pressed on. The cost of this colorimetric assay is estimated to be 5 USD and the test requires 10 min. The paper is used as an afﬁnity chromatography for conjugate separation but the authors also point out that the electrochemical signal is inﬂuenced by the urine composition, probably by its ionic strength or pH. The full potentiality of separation using paper was not yet reached in this work, suggesting that colorimetric assays remain better than electrochemical ones. Separation efﬁciency in paper-based microﬂuidic channels is usually improved by increasing the length of separation channels suggesting that strip-shape ePAD should be preferred. Such an increase is however accompanied by a sensitivity decrease, due to

the higher analyte amounts ﬂowing within the paper capillaries which do not meet with the electrode. Therefore, chromatographic properties of paper should probably be used for the separation of the analytes from the biological matrix (particles, insoluble material) using short length channels and for their distribution in different test zones where speciﬁcity is obtained thanks to the biological recognition element (antibody, DNA, enzyme…). One of the ﬁrst example of such multiplexing ePAD was the simultaneous detection of three different analytes (glucose, lactate and uric acid in human serum) by three different oxidases on three test zones (Fig. 7) with a single sample drop (Dungchai et al., 2009). Taking advantage of the paper ﬂuidic properties, simultaneous colorimetric (iron) and electrochemical (glucose) detection could be performed in two different test zones of an ePAD within the same sample (Dungchai et al., 2011). The group of Jinghua Yu has highly contributed to the development of origami ePAD during last years. In most of these papers, various cancer markers (Carcinoembryonic antigen, α-fetoprotein, prostate speciﬁc antigen, Cancer antigen 125, Cancer antigen 158) were detected using different methods (chronoamperometry, electrochemiluminescence, photoinduced amperometry, vide supra) and multiplexed electrodes (Li et al., 2014a, 2014b, 2014c, 2013b, 2013c; Liu et al., 2014b; Wang et al., 2013a, 2012b, 2014b; Yan et al., 2012). Results show that there is no cross-talking between the different test zones. Careful reading of the experimental part reveals that samples were deposited directly onto individual electrodes to allow the biomolecular recognition, e.g. one sample should be used for each electrode. Moreover, the measurements are performed sequentially which limits interferences from one working electrode to the other. The central part of the origami ePAD which is used as sample zone by other groups (Fig. 4-B) is here used as electrolyte chamber for the common reference and

Please cite this article as: Desmet, C., et al., Biosensors and Bioelectronics (2015), http://dx.doi.org/10.1016/j.bios.2015.06.052i

C. Desmet et al. / Biosensors and Bioelectronics ∎ (∎∎∎∎) ∎∎∎–∎∎∎

counter-electrode. Indeed, few interferences are observed with other biological molecules (BSA (100 ng mL 1), glutamic acid (100 ng mL 1), glucose (100 ng mL 1) and haemoglobin (100 ng mL 1)) compared to commercial assays (less than 5% differences). Unfortunately, this speciﬁcity should probably be attributed to a perfect biorecognition rather than to the chromatographic properties of the Whatman #1 paper used. As explained previously, separation between different analytes or between analyte and matrix could also be obtained by ionic interactions using modiﬁed cellulose (Whatman P81, Shiroma et al., 2012) but there is no example yet of such ePAD using enzymes or antibody as recognition element.

9. Stability A question arising is the stability of ePAD upon storage. As paper could play the role of stabilizing agent, enzymes are usually dried on the paper and recover their activity (full or partial) once hydrated. Wu et al. (2014a) showed that antibodies could be stored under dried state for a week at room temperature as long as the relative humidity is maintained at low level or if trehalose or bovine serum albumin (BSA) is added before drying. Nevertheless, stability of paper-based electrochemical immunosensors is mostly mentioned to be at 4 °C for days or weeks (Table 1). Therefore, doubt can be formulated about the transportability and storage of such sensors/biofuel cells in emerging countries or in places where refrigerators are not available. The stability of ePAD including biological components should probably be further studied.

10. Electrode conﬁguration Design of ePAD could be varied indeﬁnitely according to application or imagination. As electrochemical devices, they could include a single reference/counter electrode or have independent counter and references electrodes. Electrodes could be on the same side of the paper, on different faces or even on different sheets. The simplest system is the one described by Nie et al. (2010b) where three electrodes are printed onto a solid plastic support and electrode delimitation is deﬁned by tape (working, reference and counter electrodes). After ink baking, tape protective foil is removed and hydrophilic paper channel is placed perpendicularly to the electrodes. In the 8-ways wearable potentiostat proposed by Zhao et al. (2013), electrodes where also printed on the same side perpendicular to the hydrophilic channel. A similar independent threeelectrode system was used by Dungchai et al. (2009). Using a three electrode system requires the use of 3-ways potentiostats with independent reference way. In the biosensor proposed by Cunningham et al. (2014), a three electrodes system is screen-printed onto wax-impregnated paper. A second paper sheet is also wax-impregnated with a hydrophilic T-shaped. The vertical section of the T is removed to let an open space. Working electrode is modiﬁed with methylene blue-ssDNA or Methylene blue-aptamer using a thiol-gold interaction. The ssDNA forms a loop by intramolecular hybridation. Therefore in both cases, the methylene blue redox probe is at the vicinity of the electrode surface. A background current is measured by adjusting the hydrophilic channel onto the three electrodes and soaking it with an appropriate buffer. The channel is slipped along the vertical T-section axis in such a way that the sample can be introduced on the working electrode. After incubation, the cellulose channel is slipped back and current is measured again. The authors also emphasised that such ePAD could be stored at least 4 weeks at

13

room temperature if dried under nitrogen atmosphere. The drying under air leads to a 40% loss of the biosensor signal. This is not attributed to paper or wax alteration but to oxidation of the thiols and DNA used in the sensor. Nevertheless, as for most of biosensors based on molecular biorecognition, incubation time remains long (3 h) and operation is complex compared to measurements using a single drop of sample. Two kinds of origami ePAD have been already proposed. In the ﬁrst one, the working electrode is printed on a paper sheet and further modiﬁed to enhance its electrochemical conductivity and includes nanoparticles or capture antibodies. After recognition and several washing steps, the electrochemical cell is assembled by stacking a second paper sheet with the counter and reference electrodes. In this case, the paper is used as an electrolyte reservoir for the measurement. The counter and reference electrodes could also be printed on the other side of the paper and the ePAD would be obtained by simple folding of the sheet (Li et al., 2014b, 2014c, 2013b; Lu et al., 2012; Yan et al., 2012) (Fig. 2-A). For multiplexing purposes, this origami conﬁguration could be adapted by using common reference and counter-electrodes. Therefore, examples of four or eight working electrodes using the same counter/reference electrode were described (Ge et al., 2012a; Su et al., 2014; Wu et al., 2014b, 2013b) (Fig. 4-B). One should note that, whatever the number of electrodes, measurements are always performed sequentially on each working electrode, meaning that these assays are more or less single ePAD with reusable counter/reference electrode rather than a true multiplexed assay. Moreover in these assays, the separation properties of paper are not used. On the contrary, the origami ePAD proposed by Santhiago and Kubota (2013) is based on a paper sheet bended on 3 levels with overlapping round ﬂuidic layers (Fig. 2-C). Once stacked, the sample is applied to the upper layer which acts as separation membrane. Sample (here glucose) reaches the second layer where glucose oxidase and 4-aminophenylboronic acid were beforehand absorbed. The product of the enzymatic reaction (e.g. 4-aminophenol (4-AP)) diffuses then to the third layer, having a counter/ reference electrode screen-printed at its outer face. The origami ePAD is then unfolded and, after applying a graphite pencil as working electrode on the paper, the current generated by 4-AP oxidation is recorded within 30 s. This approach was validated with artiﬁcial blood samples, performing the measurement in less than 5 min for the whole assay. This is probably one of the simplest assay as all reagents are extemporaneously immobilized (by adsorption) onto the paper sheet but it has not been applied in multiplex assay yet. Multiplex analysis using ePAD was investigated using different electrode conﬁgurations. A 4 6 electrodes array system was proposed by Ge et al. (2012b) by printing six columns of electrodes on a polyethylene terephthalate sheet and four rows of electrodes on another one (Fig. 4-A). Both layers are separated by 24 electrolyte chambers delimited by wax. Counter and reference electrodes, printed on a fourth-layer, are shared by the 24 electrodes which could be measured individually using only 10 connectors. To the best, 6 individual measurements could be performed simultaneously using appropriate 6-ways electrochemical apparatus. The paper sensing sites were modiﬁed with glutaraldehydechitosan MWCNTs cross-links prior to capture antibody and thionine immobilization. The detection of the antigen was performed using horseradish peroxidase conjugated antibodies-AuNPs/ MWCNTs with bioconjugate in the presence of hydrogen peroxide. The electrochemical performances showed that no interferences occurred between electrodes of two neighbouring columns or rows and that there is no background signal due to nonspeciﬁc bioconjuguate adsorption. The whole measurement requires 30 min of incubation for both antigen and bioconjuguate as well as

Please cite this article as: Desmet, C., et al., Biosensors and Bioelectronics (2015), http://dx.doi.org/10.1016/j.bios.2015.06.052i

14

C. Desmet et al. / Biosensors and Bioelectronics ∎ (∎∎∎∎) ∎∎∎–∎∎∎

careful rinsing between each step. This multiplexed assay was used to detect 4 cancers markers (AFP, CA-125, CA-153 and CEA) using real human samples leading to concentrations similar to the ones measured using an electrochemiluminescent-based referenced method. A multi-chamber ePAD was proposed by Labroo and Cui (2014) in which a blood sample (diluted) is added at the centre of an ePAD and diffuses in 4 channels reaching 4 different oxidase (Glucose oxidase, Lactate oxidase, Cholesterol oxidase, Xanthine oxidase) modiﬁed graphene electrodes (Fig. 7-C). Hydrogen peroxide is directly detected on each electrode through chronoamperometry. The response time is short (about 2 min for the 4 metabolites) but requires a 4 way potentiostat or the use of successive measurements on each electrode. Interestingly, electrodes could be washed by simply adding buffer about 15–20 times with a reasonable loss of activity ( 20%). Together with the work of Dungchai et al. (2009), this multiplexed ePAD biosensor is one of the few which requires the addition of only one sample for the different electrodes. In biosensors application, the reference electrode could be of crucial importance and, in order to obtain a reliable electrochemical response, screen-printed Ag/AgCl pseudo-electrodes are often used. As emphasised by Lan et al. (2013), the applied potential highly depends on the chloride concentration inside the electrolyte. In order to obtain stable reference, these authors printed the reference electrode in another region of the ePAD connected to the sample region by a cellulose hydrophilic channel (Fig. 8). Both sample and reference solutions (KCl solution) were added to their respective zone and diffused in a central zone but do not mix signiﬁcantly. This creates a junction of potential between the reference electrode and the sample solution similar to the one obtained in a classical Ag/AgCl glass reference electrode. The system could be applied in a multiplex system with a single reference electrode and three distinct working/counter electrodes pairs. Such a strategy has not yet been applied to any paper-based electrochemical biosensors and required a three-ways potentiostat which could increase the price of the whole system. Whiteside and co-workers (Lan et al., 2013) showed that a similar ePAD operating without KCl in the electrolyte solution had a potential shift of about 0.15 V. This work demonstrates that, by designing properly

Fig. 8. Referenced ePAD (front and back view). The Ag/AgCl electrode is screenprinted in the reference zone further ﬁlled with KCl solution. Sample is added to the sample zone on the working and counter-electrode. The red dotted line corresponds to the region where KCl and sample do not mix, establishing a junction potential (Lan et al., 2013). (For interpretation of the references to colour in this ﬁgure legend, the reader is referred to the web version of this article.)

the ePAD, a reliable and cheap reference electrode could be obtained. Biofuel cells developed on paper require a special design. Indeed, in a biosensor, the counter-electrode is never considered as long as its surface is large enough. In a biofuel cell, both anode and cathode have equal importance. There are nevertheless fewer examples of biofuel cells than biosensors using paper as the electrode support. Basically, a biofuel cell generates a current from two enzyme-catalysed redox reactions: at the anode a soluble substrate is usually oxidized while molecular oxygen, ideally provided by air, is reduced. Advances in biofuel cells are regularly reviewed, see, among others, Bullen et al., 2006; Cosnier et al., 2014; de Poulpiquet et al., 2014; Le Goff et al., 2015; Leech et al., 2012; Luz et al., 2014; Minteer et al., 2007; Willner et al., 2009. The use of paper in biofuel cell is probably more promising that in electrochemical biosensors because ex-vivo biofuel cells are always intended to be low-cost and are based on easily available enzymes. Moreover, simpliﬁcation of biofuel cell assembly is required to ﬁnd economic viability. In addition, while biosensors requires a high selectivity regarding the analyte, biofuel cells do not require to be highly speciﬁc as their aim is to generate current from biomolecules. Designing electrodes on paper for biofuel cells could offer some advantages over more classical conﬁguration. Thanks to the ﬂuidic properties of paper, biofuel (substrate) could be continuously supplied to the electrode without external pump. Paper is a biocompatible material that could be used in implanted biofuel cells. Biofuel cells usually require expensive membranes, useful for enzymes immobilization; paper could play this role. The assembly of biofuel cells, sometimes considered as complicated, could be simpliﬁed thanks to the electrode printing technologies described above. In the ﬁrst work of Jenkins et al. (2012), both anode and cathode were printed on separate paper sheets and were separated by a dialysis membrane (Separator). The choice of paper as electrode support was not discussed in this work but it seems that it was chosen as good candidate for substrate reservoir. Moreover, the resulting biofuel cell was a ﬂat device easy to handle (Fig. 3-A). In biofuel cells, anodes and cathodes could be printed on the same side of the paper, the hydrophilic channel (or region) between each electrode acting as a membrane (Zhang et al., 2012). In this conﬁguration, the cathode is not immersed in the solution and oxygen from air diffuses to the enzyme (bilirubin oxidase in these cases). This biofuel cell operates with a fuel volume as low as 30 mL (Fig. 3-C). Anode and cathode of a biofuel cell could be screen-printed on each side of a paper sheet. This is the perfect example where the biofuel cell is dipped into the substrate solution for continuous substrate supply (Shitanda et al., 2013a) (Fig. 3-B). To favour the oxygen diffusion at the cathode, air-breathing cathodes should be developed. In the work of Ciniciato et al. (2012) different cathodes using toray paper (current collector) were pressed on a layer of teﬂonized carbon black to create a water repelling cathode. In this half-cell, the paper material is used as the electrolyte support. A complete biofuel cell was constructed recently by this group using the air-breathing cathode and an anode made with pressed buckypaper (Fig. 3-D). Buckypaper is used to achieve high current densities as compared to more traditional screen-printed carbon electrodes. The Whatman paper was designed as a fan, anode and cathode being pressed on the handle of the fan. This geometry allows the improvement of the transport of the fuel to the anode: the handle of the fan is plunged into the substrate solution and reaches the anode by capillary forces but the solution in the fan evaporates. The evaporation become the main driving force resulting in a quasi-steady-state ﬂow (Villarrubia et al., 2014). Power out could be increased by

Please cite this article as: Desmet, C., et al., Biosensors and Bioelectronics (2015), http://dx.doi.org/10.1016/j.bios.2015.06.052i

C. Desmet et al. / Biosensors and Bioelectronics ∎ (∎∎∎∎) ∎∎∎–∎∎∎

stacking three electrodes together using carbon yarn (Fig. 3-D). Wu et al. (2013a) coated two CNT-cellulose electrodes on the same side of a paper sheet to obtain a biofuel cell able to ﬂoat on the substrate solution surface. In this case, the large surface area of paper is used to supply fructose and dissolved oxygen to the cell, meaning that the cathode is not a true air-breathing cathode. Moreover, this biofuel cell could also be stacked to increase its power output. Finally, some biosensors were designed as to create a potential difference between electrodes (open circuit voltage) in order to charge a capacitor. The ﬁrst example is the self-powered biosensor for adenine detection described by Liu et al. (2012b) (Fig. 6-A). Two channels and electrodes were prepared on the same paper sheet and a reaction chamber put into contact with the electrodes by simple folding. The reaction chamber, pre-loaded with glucose and [Fe(CN)6]3 , had the form of a hourglass in order to create a junction potential between the anode and the cathode. The whole ePAD was laminated to prevent evaporation. Upon sample (adenine) addition, a glucose oxidase-DNA conjugate is released and oxidises the pre-loaded glucose together with [Fe(CN)6]3- reduction in the reaction chamber. The other channel does not contain the glucose oxidase. The difference in concentrations of [Fe (CN)6]3 and [Fe(CN)6]4 in the sensing half-cell and control halfcell results in an open circuit voltage that is used to charge a capacitor. The discharge of the capacitor in a digital multimeter (DMM) allows a 15.5-fold increase in the signal compared to direct chronoamperometric measurement. Other examples of this capacitor are presented in the last part of this review.

11. Applied biosensors Paper-based biosensors and biofuel cells, even if more and more described, are still in their early development phase. Despite the general considerations previously mentioned, each ePAD has its own characteristics and was constructed taking totally or partially advantage of paper properties. Table 1 summarizes the analytical ﬁgure of merit of various paper-based biosensors found in the literature while Table 2 summarizes similar data for biofuel cells. This last part attends to present more in detail the operating functions of more complex ePAD and to provide some critical comments. Tan et al. (2012) used glucose oxidase (GOx) immobilized on a paper disc placed upon a commercial screen-printed electrode and clamped in a PDMS shim with a ﬂuidic channel. The principle of the measurement is to measure the inhibition of GOx by silver ions. Here, paper was only used as a disposable immobilization matrix for the enzyme while SPE could be reused. A cholesterol biosensor was developed by screen-printing electrodes on Whatman paper and electrospraying graphene/PVP/PANI nanocomposite onto the electrode surface. Cholesterol oxidase was dried onto the electrode surface (Ruecha et al., 2014). In this study, despite the same sample volume was used, the biological sample was prepared using laboratory centrifuge and accurate dilution. The signal was obtained by chronoamperometry using lab potentiometer. Delaney et al. described ECL detection in ePAD using Ru (bpy)32 þ as a luminophore reagent and inkjet-printed ePADs laminated on commercial screen-printed electrodes. This system was also applied to the detection of NADH, an important biological mediator found in more than 250 biochemical reactions (Delaney et al., 2011). Molecular recognition between antibodies (Ab) and antigens or between single strand DNA (ssDNA), coupled with enzymatic reaction, was used in many paper-based biosensors developed by the group of Jinghua Yu during the last years. These biosensors are multiple variations of the same assay which could be described as

15

follow: i. Hydrophilic zones on Whatman paper were deﬁned by waxprinting for the working electrode and for the counter and reference electrodes. ii. Electrodes were screen-printed on the hydrophilic zones. iii. Cellulose ﬁbres were modiﬁed with silver, gold or platinum nanoparticles to obtain nanoporous metal. Metal could be replaced by CNT-Chitosan bioconjuguate. iv. Capture antibody, ssDNA or quantum dots were immobilized using classical chemistry (EDC/NHS coupling or using glutaraldehyde) prior to electrode surface saturation with BSA for example. v. A biological sample was incubated directly on the working electrode(s) to obtain the biorecognition. vi. The working electrode was extensively washed to remove unbound material. vii. A complicated bioconjugate composed of detection antibody covalently bound to another nanoparticle or carbon nanotubes together with a detection partner such as horseradish peroxidase, Ru(Bpy)32 þ or TMB was incubated, followed by another washing step. viii. The origami ePAD was folded and the substrate for the detection is added on the hydrophilic paper zone. ix. Signal was measured independently on each electrode using differential pulse voltammetry (DPV), electrochemiluminescence or chronoamperometry. In most cases, detection antibodies are directed against cancer markers such as Carcinoembryonic antigen (CEA), prostate speciﬁc antigen (PSA) or α-fetoprotein (AFP). The respective detection methods and bioconjuguate are reported in Table 1. In these biosensors, paper is mostly used for electrode surface ampliﬁcation and for delimitation of the electrochemical chamber. Moreover, the price of the assay is not evaluated but it contains metals, quantum dots (CdTe) and requires the preparation of speciﬁc bioconjuguates. These paper-based ePAD are in fact not designed for single use and regeneration procedures and reusability information was provided by the authors. In some cases, the assay could be performed in about 10 min (Wang et al., 2012a) including the washing steps for the detection of CEA or CA-125 (carbohydrate antigen 125) when paper is modiﬁed with CNT and chitosan using HRP, hydrogen peroxide and o-phenylenediamine as detection system. It could be also as long as 60 min for the detection of PSA when paper is modiﬁed with MnO2@AuNPs using GOx and 3,3′,5,5′-tetramethylbenzidine(TMB) as detection system (Li et al., 2014b). In both cases, the electrochemical detection was performed through DPV using a laboratory apparatus. The eight electrodes origami ePAD described earlier was used for the electrochemical detection of the four cancer markers after modiﬁcation of the paper with chitosan and graphene oxide (Fig. 4-B). Signal ampliﬁcation was obtained with HRP immobilized on SiO2 nanoparticles and reduction of 2,2′-diaminoazobenzene by hydrogen peroxide was detected sequentially on each electrode by DPV (Wu et al., 2013b). This 8-electrodes ePAD could be used for ECL triggering if secondary antibodies are labelled with Ru(bpy)32 þ as catalyst. Interferences due to light emission from another electrode is not considered since electrodes are measured sequentially (Ge et al., 2012a). Moreover, the samples were directly applied to each individual electrode and not from the central zone, losing the interest of multiplexing analysis and separation provided within the paper. Secondary antibodies labelled with Ru(bpy)32 þ were also directly applied to the working electrodes. The common feature observed for these immunoassays is the relatively long time for bio-recognition (about 30 min), as well as the importance of washing steps between each incubation.

Please cite this article as: Desmet, C., et al., Biosensors and Bioelectronics (2015), http://dx.doi.org/10.1016/j.bios.2015.06.052i

16

C. Desmet et al. / Biosensors and Bioelectronics ∎ (∎∎∎∎) ∎∎∎–∎∎∎

DNA detection was also achieved using origami ePAD. Lu et al. (2012) used a screen-printed working electrode modiﬁed with capture ssDNA attached to an AuNP/graphene composite. Similarly to the antibody/antigen based immunoassays, the target ssDNA should be incubated for 2 h and then for 3 h with graphene/DNA/ thionine conjuguate before the DPV detection of thionine. Other origami ePAD include the discrimination of single-nucleotide mismatch in p53-DNA double strand (15 bp) human gene in urine (Xu et al., 2013, 2014) or the detection of two cancer markers on the same ePAD using two different ECL probes (Carbon nanodots and Ru(bpy)32 þ ) and a 3 V battery (Wang et al., 2012b). Other examples are provided in Table 1. The origami ePADs were used to build a nearly autonomous biosensor by coupling an enzymatic paper-based biosensor acting as a biofuel cell and a paper-based supercapacitor (as described earlier) (Fig. 6-A). The principle is based on the oxidation of the analyte at the anode of the biosensor together with oxygen reduction on a paper-based cathode. The potential difference developed between the electrodes is used to charge the supercapacitor. To perform the measurement, the supercapacitor is discharged into a digital multimeter (DMM) and maximal intensity is used as signal. The ﬁrst example of self-powered biosensor was probably described by Liu et al. (2012b) for the detection of adenosine (vide supra). Different conﬁgurations were used for the detection of ssDNA using complementary DNA conjugated with GOx or HRP at the anode and paper modiﬁed by platinum NPs at the cathode (Wang et al., 2014a). The detection of CEA was also achieved through such system, using NADH oxidation (produced by glucose dehydrogenase) and bilirubin oxidase immobilized at the cathode (Wang et al., 2014b). Alternatively, a photocurrent induced biosensor was developed by Wang et al. (2014c). SnO2 quantum dotreduced graphene oxide conjugate, ssDNA immobilized on the electrode and aptamer were used for the detection of ATP. The aptamer was modiﬁed with a bioconjuguate to generate the chemiluminescence reaction (glucose oxidase–H2O2–p-iodophenol– luminol system). The photo-induced current created by the SnO2 quantum dots was used to charge the capacitor made on the same paper sheet. This strategy, despite it requires careful preparation of two bioconjuguates and electrode assembly prior to use, is one of the few wearable ePAD based on biomolecular recognition of small organic molecule with short time response (1 min for the charge of the capacitor and 1 min for the discharge). If such system could be adapted as dipping assay, it would be one of the most promising ePAD. Alternatively, a wheel of 8 different working electrodes could be used with the same reference and counter-electrode (Wang et al., 2013a) for successive analysis. It was also shown that the paper-based supercapacitor could be charged and discharged more than 100-times. Similar assay for the detection of ATP was also obtained using CdS nanoparticles onto cellulose modiﬁed by CNTs (Ge et al., 2013b). The photocurrent was stable for 180 s but the supercapacitior was charged in only 60 s. Despite the time request for sample preparation, hybridisation of aptamer on the ssDNA and washing steps, followed by the assembly and substrate addition, the measurement only required about 1 min. Moreover, using a discharge current instead of using a classical amperometric measurement led to a 13-fold increase of the signal. A ﬂuidic channel was introduced in a similar system in order to have reproducible loading time of the supercapacitor (delay switch) (Fig. 6-B). Once the molecular recognition has occurred with unfolded origami ePAD, the paper sheet was correctly folded and inserted in its support. Substrate-containing buffer is introduced in order to charge the supercapacitor. Simultaneously, buffer ﬁlled a channel and reached dried AuNPs/MWCNT. Upon hydration, electrical contact occurs between the anode and the

cathode of the origami ePAD and permits the discharge of the supercapacitor (Fig. 6-B). The delay for the interrupter to operate (around 10 s) allowed the operator to close the device holder and avoid light interferences. This is one of the few examples using ﬂuidic as interrupters (timer tab) in electrochemical biosensor/ biofuel cell (Wang et al., 2013c). A variation of this system is the use of CdS and ZnO nanoparticles and ascorbic acid as electron donor for the detection of CEA. Paper was used as support for a CNT–PDDA (poly(dimethyldiallylammonium chloride)) conjugate. Photocurrent decreased due to steric hindrances caused by the CEA–antibody interaction. The addition of an electron donor did not require the use of a microﬂuidic timer nor a dark box as it operates under visible light. Experiments were conducted in lab condition e.g. under controlled light intensity (400 μW cm 2) and using a lab potentiostat with 3 electrodes (Wang et al., 2013a). Paper-based cyto-devices offer strong opportunities to realise in situ, simple, miniaturised, and high-throughput cytological or histological researches. Human acute promyelocytic leukaemia cells (HL-60) were detected with gold modiﬁed paper electrodes containing a speciﬁc aptamer (aptamer KH1C12) (Su et al., 2014). A HRP-folic acid conjugate is used to detect and quantify the cells. This ePAD is a simple and cheap macroporous 3D environment for the cells. This system was applied to drug screening by monitoring the apoptosis. Three antileukemic drugs (Cycloheximide, etoposide, and camptothecin) were assayed with the ePAD and the results correlate with ﬂuorescent imaging but required shorter time. In case of apoptosis, the phosphatidylserine of the plasmic membrane ﬂips to outer plasma membrane and is ready to interact with annexin V (HRP-labelled). This does not occur with necrotic cells which allowed the differentiation between apoptosis and necrosis. The paper based electrode showed better performances than its ITO counterpart due to the 3D macrostructure leading to higher cell capture within the electrode. Paper is a cheap matrix for cell culture and it could be easily chemically modiﬁed. In this assay, an ePAD with 4 electrodes was developed but, as with other origami ePAD, measurements are performed sequentially. In another work (Liu et al., 2014b), SKR cancer cells were immobilized onto platinum modiﬁed cellulose with aptamers. This ePAD is a suitable environment for tumour cell proliferation and adhesion for at least 12 h (experiment duration in this work). Hydrogen peroxide secreted by the tumour cells (reactive oxygen species) or released after the cell apoptosis could be easily detected. The time response is relatively short (8 s) after the stimulation of H2O2 secretion and remains constant for 20 s while it decreased on similar system using platinum-modiﬁed glassy carbon electrodes (Liu et al., 2014b). A special ePAD is the chemiresistor proposed by Pozuelo et al. (2013). In this case, paper was used as support for SWCNT adsorption until a resistance of 150–250 kΩ is reached. Glue was used to deﬁne the sensor region and to separate the sensor from the connector. The chemiresitor was obtained by simple adsorption of an antibody (directed against human IgG). The antigen bound to the antibody modiﬁed the sensor resistance that could be measured with a DMM. Such sensor is easy to handle but no measurement in the presence of interfering species or in biological matrix was reported. Electrochemical detection only needs to quantify electrical conductivity or potential changes. However, it requires reactions that produce electroactive molecules. Therefore Davaji and Lee (2014) developed a paper based microcalorimeter to measure enzymatic reaction and ligand interactions without labelling. Since most biochemical reactions or interactions are accompanied by a temperature change, this label free calorimetric detection method extends and enhances the capabilities of existing paper-based microﬂuidic systems to include a wide range of bio-chemical

Please cite this article as: Desmet, C., et al., Biosensors and Bioelectronics (2015), http://dx.doi.org/10.1016/j.bios.2015.06.052i

C. Desmet et al. / Biosensors and Bioelectronics ∎ (∎∎∎∎) ∎∎∎–∎∎∎

sensing and diagnostic applications. The heat generated or absorbed from a bio-chemical reaction causes a temperature change, directly related to the concentration of the reagent, of the samples, and of the enthalpy of the reaction. The paper-based device consisted of a microﬂuidic channel made of a chromatography-grade ﬁlter paper placed directly on top of a resistive temperature detector. The channel was prepared using a knife plotter to avoid chemical contamination by wax or hydrophobic polymers. The enzyme (glucose oxidase) was immobilized by simple adsorption on the centre of the paper channel and the reagent (glucose) was added at the channel entrance. After sample diffusion, the exothermic oxidation of glucose was recorded as a 0.2 °C variation. This assay was also used to quantify DNA oxidation by hydrogen peroxide as well as to detect the interaction between biotin and streptavidin. In this work, the resistance change upon temperature was determined with a lab scale device using 4-points measurement setup with 2% accuracy but, ideally, a resistance could be measured with a DMM and converted into activity, probably with lower accuracy.

12. Speciﬁcity of biofuel cells Biofuel cells are devices for power generation in small, wearable or implantable devices. They are composed of two electrodes (anode and cathode) where oxidoreductases are usually immobilized. In order to reduce the price of biofuel cells as well as to continuously supply substrates to the electrodes, paper microﬂuidic electrochemical devices appear to be promising. Biofuel cells made of paper could be ﬂexible, would not require important volumes of electrolyte and could probably circumvent the cost of existing biofuel cells, especially concerning the use of membranes. The ﬁrst work on biofuel cells using paper based electrode is probably the one of Jenkins et al. (2012). In this work the ink is composed of CNT, enzymes (Laccases or PQQ-dependant aldose dehydrogenase) and mediators (osmium complexes) which were printed onto Whatman paper. This approach allows the immobilization of all partners for each electrode in a single step. Anodes and cathodes were connected to the electrochemical device by direct contact with graphite plates and were separated by a dialysis membrane (Fig. 3-A). The cell could be supplied with 200 mL of glucose solution and could have an operation time of about 90–100 h. This study was mostly dedicated to mediator selections and to power output optimisation, the paper material being used as substrate reservoir only. Further research on biofuel cells on paper seems to indicate that separation membrane (dialysis) is not necessary. Nevertheless, this work shows that numerous ink conditions could be analysed with such a system in a relatively short time. Zhang et al. (2012) showed that the cellulose material could act as an electrolyte support for a membrane-less biofuel cell. In this case, glucose dehydrogenase and bilirubin oxidase were immobilized together with CNT onto screen-printed electrodes. The cell is operating with 30 mL of a glucose solution supplemented by NAD þ . The electron source could be a commercial soft or a coffee solution. The cell was operating for 0.75 h and it could be reﬁlled at least once. Paper was also used as a membrane in the only example of microbial biofuel cell that could be classiﬁed as ePAD (Fraiwan et al., 2013). The paper acts as a proton exchange membrane once modiﬁed by sodium polystyrene sulphonate (PSS). Particles of PSS render the paper more hydrophobic and prevent the mixing of anolyte and catholyte. The open circuit voltage increases 2-times with performances similar to Naﬁon membranes but with lower stability. Shitanda et al. (2013a) used Japanese paper screen-printed on

17

both sides to build a membrane-less biofuel cell. The anode is composed of glucose oxidase and Tetrathiafulvalene (TTF) adsorbed at the surface of Ketjen black/styrene butadiene rubber polymer in order to improve the electrode surface (Fig. 3-B). Styrene butadiene rubber is replaced by PTFE at the cathode in order to create a water repelling electrode. Anode and cathode are on each side of the paper sheet. Glucose is supplied by diffusion from a solution to the anode via the paper ﬂuidic but it is not really clear if oxygen is supplied by air or dissolved in the electrolyte. Performances of this BFC are better than others and it offers one of the ﬁrst opportunity for continuous substrate supply to the electrodes beneﬁting from the capillary force of the paper. The ﬂuidic property of cellulose was also applied by Ciniciato in a half-cell conﬁguration to build an air-breathing cathode (Ciniciato et al., 2012). A carbon ink containing bilirubin oxidase was painted onto nitrocellulose and was further pressed onto a gas diffusion layer (teﬂonized carbon black) and a toray paper as connector (Fig. 3-D). A complete biofuel cell using this cathode was build using an anode based on glucose dehydrogenase together with soluble NAD þ and electro-polymerised methylene green as the mediator. To drive the substrate diffusion to the electrode, the biofuel cell has the form of a fan. A commercial soft solution was used as sugar source but should be supplemented, as previously, by NAD þ . However, the stacking of three of these biofuel cells allows power generation for a minimum of 9 h (Fig. 3-D). Wu et al. (2013a) built a Fructose dehydrogenase–Bilirubin oxidase biofuel cell with a screen-printed electrode on paper overlaying a 2 mL fuel reservoir. Oxygen is supplied from the electrolyte solution but the BFC could be used in rolled form without performance alteration (less than 45°). This system does not required membrane for separation between anode and cathode as paper seems to play this role. Up to now it is not clear if the cathode of these cells is really an air-breathing cathode. Paper-based biofuel cells are still at their infancy despite these encouraging examples. Paper is a non-toxic material that could probably be used in implantable devices. Stability of the paper itself should be investigated once hydrated, as well as microorganism colonisation of the paper in the presence of substrate of the biofuel cell (sugar). Moreover, strategies yet developed to improve electron transfer (electrode surface modiﬁcation) should probably be adapted in order to lower cofactor costs (addition of NAD þ ).

13. Critical conclusion Electrochemical paper-based devices (both biofuel cells and biosensors) are a promising technology, especially when they are conceived from the start as cheap, wearable and storable devices, including the electrochemical reader. Nevertheless, limitations still exist, in particular the number of sample or analytes that could be analysed simultaneously or the lifetime of a biofuel cell. Moreover, ePAD based on antibody–antigen recognition require relatively long incubation times, multi-step surface modiﬁcations, use of complex bioconjuguates for signal ampliﬁcation and an expert manipulator. In such assays, the advantages of ePAD compared to existing assays have still to be demonstrated. EPAD currently described in the literature are mainly developed for the detection of classical targets in diagnostic such as small organic molecules (glucose, lactic acid, uric acid and ascorbic acid) or biomarkers for cancer (CEA, AFP). Several biosensors or methods are already described for these targets. Considering the main argument of using low cost ePAD in emerging countries, a full range of applications in the diagnostic of human virus infection (HIV, SARS, West Nile Virus), in the detection of human and animal pathogen in food (cholera) or water should still be developed for practical

Please cite this article as: Desmet, C., et al., Biosensors and Bioelectronics (2015), http://dx.doi.org/10.1016/j.bios.2015.06.052i

18

C. Desmet et al. / Biosensors and Bioelectronics ∎ (∎∎∎∎) ∎∎∎–∎∎∎

applications. Moreover, manipulation of ePAD still requires an expert manipulator for pipetting, folding or measuring even if great improvements are continuously done. Paper was mostly used as a cheap material for electrode screen-printing, as a 3D support for electrode modiﬁcation and as an electrolyte reservoir. It is worth to note that the separation properties of paper in ePAD biosensors was not strongly investigated leading to assays with numerous washing steps. In ePAD biofuel cells, the ﬂuidic of paper is generally used to continuously supply electrodes with soluble substrates. Paper-based biofuel cells are even younger and, if proof-ofprinciples are shown, real practical applications (as for the other kinds of biofuel cells) should be found. This review has shown that, if paper was a material of the past, it is also a material for the future.

References Beni, V., Nilsson, D., Arven, P., Norberg, P., Gustafsson, G., Turner, A.P.F., 2015. ECS J. Solid State Sci. Technol. (in press), http://dx.doi.org/10.1016/j.arabjc.2012.04.019 Bullen, R.A., Arnot, T.C., Lakeman, J.B., Walsh, F.C., 2006. Biosens. Bioelectron. 21 (11), 2015–2045. Carrilho, E., Martinez, A.W., Whitesides, G.M., 2009. Anal. Chem. 81 (16), 7091–7095. Carvalhal, R.F., Kfouri, M.S., Piazetta, M.H.D., Gobbi, A.L., Kubota, L.T., 2010. Anal. Chem. 82 (3), 1162–1165. Chitnis, G., Ding, Z., Chang, C.-L., Savran, C.A., Ziaie, B., 2011. Lab Chip 11 (6), 1161–1165. Ciniciato, G., Lau, C., Cochrane, A., Sibbett, S.S., Gonzalez, E.R., Atanassov, P., 2012. Electrochim. Acta 82, 208–213. Coltro, W.K.T., Cheng, C.M., Carrilho, E., de Jesus, D.P., 2014. Electrophoresis 35 (16), 2309–2324. Cosnier, S., Le Goff, A., Holzinger, M., 2014. Electrochem. Commun. 38, 19–23. Cunningham, J.C., Brenes, N.J., Crooks, R.M., 2014. Anal. Chem. 86 (12), 6166–6170. Davaji, B., Lee, C.H., 2014. Biosens. Bioelectron. 59, 120–126. de Poulpiquet, A., Ranava, D., Monsalve, K., Giudici-Orticoni, M.-T., Lojou, E., 2014. Chemelectrochem 1 (11), 1724–1750. Delaney, J.L., Hogan, C.F., Tian, J.F., Shen, W., 2011. Anal. Chem. 83 (4), 1300–1306. Dossi, N., Toniolo, R., Impellizzieri, F., Bontempelli, G., 2014. J. Electroanal. Chem. 722, 90–94. Dossi, N., Toniolo, R., Piccin, E., Susmel, S., Pizzariello, A., Bontempelli, G., 2013a. Electroanalysis 25 (11), 2515–2522. Dossi, N., Toniolo, R., Pizzariello, A., Impellizzieri, F., Piccin, E., Bontempelli, G., 2013b. Electrophoresis 34 (14), 2085–2091. Dungchai, W., Chailapakul, O., Henry, C.S., 2009. Anal. Chem. 81 (14), 5821–5826. Dungchai, W., Chailapakul, O., Henry, C.S., 2011. Analyst 136 (1), 77–82. Fosdick, S.E., Anderson, M.J., Renault, C., DeGregory, P.R., Loussaert, J.A., Crooks, R. M., 2014. Anal. Chem. 86 (7), 3659–3666. Fraiwan, A., Mukherjee, S., Sundermier, S., Lee, H.S., Choi, S., 2013. Biosens. Bioelectron. 49, 410–414. Ge, L., Wang, P.P., Ge, S.G., Li, N.Q., Yu, J.H., Yan, M., Huang, J.D., 2013a. Anal. Chem. 85 (8), 3961–3970. Ge, L., Yan, J.X., Song, X.R., Yan, M., Ge, S.G., Yu, J.H., 2012a. Biomaterials 33 (4), 1024–1031. Ge, S.G., Ge, L., Yan, M., Song, X.R., Yu, J.H., Huang, J.D., 2012b. Chem. Commun. 48 (75), 9397–9399. Ge, S.G., Liu, W.Y., Ge, L., Yan, M., Yan, J.X., Huang, J.D., Yu, J.H., 2013b. Biosens. Bioelectron. 49, 111–117. Hayat, A., Marty, J.L., 2014. Sensors 14 (6), 10432–10453. Hu, J., Wang, S.Q., Wang, L., Li, F., Pingguan-Murphy, B., Lu, T.J., Xu, F., 2014. Biosens. Bioelectron. 54, 585–597. Huang, J.G., Ichinose, I., Kunitake, T., 2005. Chem. Commun. 13, 1717–1719. Jagadeesan, K.K., Kumar, S., Sumana, G., 2012. Electrochem. Commun. 20, 71–74. Jenkins, P., Tuurala, S., Vaari, A., Valkiainen, M., Smolander, M., Leech, D., 2012. Enzyme Microb. Technol. 50 (3), 181–187. Jeong, S.G., Kim, J.M., Nam, J.O., Song, Y.S., Lee, C.S., 2013. Int. Neurourol. J. 17 (4), 155–161. Komuro, N., Takaki, S., Suzuki, K., Citterio, D., 2013. Anal. Bioanal. Chem. 405 (17), 5785–5805. Kong, F.Z., Hu, Y.F., 2012. Anal. Bioanal. Chem. 403 (1), 7–13. Kumar, S., Chauhan, V.P., Chakrabarti, S.K., 2012. Arab. J. Chem. (in press) Kumar, S., Jagadeesan, K.K., Joshi, A.G., Sumana, G., 2013. Rsc Adv. 3 (29), 11846–11853. Labroo, P., Cui, Y., 2014. Anal. Chim. Acta 813, 90–96. Lan, W.J., Maxwell, E.J., Parolo, C., Bwambok, D.K., Subramaniam, A.B., Whitesides, G.M., 2013. Lab Chip 13 (20), 4103–4108. Lankelma, J., Nie, Z.H., Carrilho, E., Whitesides, G.M., 2012. Anal. Chem. 84 (9), 4147–4152. Le Goff, A., Holzinger, M., Cosnier, S., 2015. Cell. Mol. Life Sci. 72 (5), 941–952.

Leech, D., Kavanagh, P., Schuhmann, W., 2012. Electrochim. Acta 84, 223–234. Li, L., Li, W.P., Yang, H.M., Ma, C., Yu, J.H., Yan, M., Song, X.R., 2014a. Electrochim. Acta 120, 102–109. Li, L., Xu, J.M., Zheng, X.X., Ma, C., Song, X.R., Ge, S.G., Yu, J.H., Yan, M., 2014b. Biosens. Bioelectron. 61, 76–82. Li, M., Li, Y.T., Li, D.W., Long, Y.T., 2012. Anal. Chim. Acta 734, 31–44. Li, W.P., Li, L., Ge, S.G., Song, X.R., Ge, L., Yan, M., Yu, J.H., 2014c. Biosens. Bioelectron. 56, 167–173. Li, W.P., Li, L., Li, M., Yu, J.H., Ge, S.G., Yan, M., Song, X.R., 2013a. Chem. Commun. 49 (83), 9540–9542. Li, W.P., Li, L., Li, S., Wang, X., Li, M., Wang, S.W., Yu, J.H., 2013b. Sens. Actuators B Chem. 188, 417–424. Li, W.P., Li, M., Ge, S.G., Yan, M., Huang, J.D., Yu, J.H., 2013c. Anal. Chim. Acta 767, 66–74. Liu, B.W., Du, D., Hua, X., Yu, X.Y., Lin, Y.H., 2014a. Electroanal 26 (6), 1214–1223. Liu, F., Ge, S.G., Yu, J.H., Yan, M., Song, X.R., 2014b. Chem. Commun. 50 (71), 10315–10318. Liu, H., Xiang, Y., Lu, Y., Crooks, R.M., 2012b. Angew. Chem. Int. Ed. Engl. 51 (28), 6925–6928. Lu, J.J., Ge, S.G., Ge, L., Yan, M., Yu, J.H., 2012. Electrochim. Acta 80, 334–341. Luz, R.A.S., Pereira, A.R., de Souza, J.C.P., Sales, F.C.P.F., Crespilho, F.N., 2014. ChemElectroChem 1 (11), 1751–1777. Makela, T., Jussila, S., Vilkman, M., Kosonen, H., Korhonen, R., 2003. Synth. Met. 135 (1-3), 41–42. Malon, R.S.P., Chua, K.Y., Wicaksono, D.H.B., Corcoles, E.P., 2014. Analyst 139 (12), 3009–3016. Martinez, A.W., Phillips, S.T., Butte, M.J., Whitesides, G.M., 2007. Angew. Chem. Int. Ed. Engl 46 (8), 1318–1320. Metters, J.P., Houssein, S.M., Kampouris, D.K., Banks, C.E., 2013. Anal. Methods 5 (1), 103–110. Mihranyan, A., Esmaeili, M., Razaq, A., Alexeichik, D., Lindstrom, T., 2012. J. Mater. Sci 47 (10), 4463–4472. Minteer, S.D., Liaw, B.Y., Cooney, M.J., 2007. Curr. Opin. Biotechnol. 18 (3), 228–234. Mirasoli, M., Guardigli, M., Michelini, E., Roda, A., 2014. J. Pharm. Biomed. Anal. 87, 36–52. Nery, E.W., Kubota, L.T., 2013. Anal. Bioanal. Chem. 405 (24), 7573–7595. Nguyen, T.H., Fraiwan, A., Choi, S., 2014. Biosens. Bioelectron. 54, 640–649. Nie, J.F., Zhang, Y., Lin, L.W., Zhou, C.B., Li, S.H., Zhang, L.M., Li, J.P., 2012. Anal. Chem. 84 (15), 6331–6335. Nie, Z.H., Deiss, F., Liu, X.Y., Akbulut, O., Whitesides, G.M., 2010a. Lab Chip 10 (22), 3163–3169. Nie, Z.H., Nijhuis, C.A., Gong, J.L., Chen, X., Kumachev, A., Martinez, A.W., Narovlyansky, M., Whitesides, G.M., 2010b. Lab Chip 10 (4), 477–483. Noiphung, J., Songjaroen, T., Dungchai, W., Henry, C.S., Chailapakul, O., Laiwattanapaisal, W., 2013. Anal. Chim. Acta 788, 39–45. Pozuelo, M., Blondeau, P., Novell, M., Andrade, F.J., Rius, F.X., Riu, J., 2013. Biosens. Bioelectron. 49, 462–465. Rozand, C., 2014. Eur. J. Clin. Microbiol. Infect. Dis. 33 (2), 147–156. Ruecha, N., Rangkupan, R., Rodthongkum, N., Chailapakul, O., 2014. Biosens. Bioelectron. 52, 13–19. Santhiago, M., Kubota, L.T., 2013. Sens. Actuators B Chem. 177, 224–230. Shiroma, L.Y., Santhiago, M., Gobbi, A.L., Kubota, L.T., 2012. Anal. Chim. Acta 725, 44–50. Shitanda, I., Kato, S., Hoshi, Y., Itagaki, M., Tsujimura, S., 2013a. Chem. Commun. 49 (94), 11110–11112. Shitanda, I., Yamaguchi, T., Hoshi, Y., Itagaki, M., 2013b. Chem. Lett. 42 (11), 1369–1370. Strack, G., Babanova, S., Farrington, K.E., Luckarift, H.R., Atanassov, P., Johnson, G.R., 2013. J. Electrochem. Soc. 160 (7), G3178–G3182. Su, M., Ge, L., Ge, S., Li, N., Yu, J., Yan, M., Huang, J., 2014. Anal. Chim. Acta 847, 1–9. Taleat, Z., Khoshroo, A., Mazloum-Ardakani, M., 2014. Microchim. Acta 181 (9–10), 865–891. Tan, S.N., Ge, L., Tan, H.Y., Loke, W.K., Gao, J., Wang, W., 2012. Anal. Chem. 84 (22), 10071–10076. Villarrubia, C.W.N., Lau, C., Ciniciato, G., Garcia, S.O., Sibbett, S.S., Petsev, D.N., Babanova, S., Gupta, G., Atanassov, P., 2014. Electrochem. Commun. 45, 44–47. Wang, P., Ge, L., Yan, M., Song, X., Ge, S., Yu, J., 2012a. Biosens. Bioelectron. 32 (1), 238–243. Wang, P., Sun, G., Ge, L., Ge, S., Song, X., Yan, M., Yu, J., 2013a. Chem. Commun. 49 (88), 10400–10402. Wang, S., Ge, L., Zhang, Y., Song, X., Li, N., Ge, S., Yu, J., 2012b. Lab Chip 12 (21), 4489–4498. Wang, S.W., Dai, W.J., Ge, L., Yan, M., Yu, J.H., Song, X.R., Ge, S.G., Huang, J.D., 2012c. Chem. Commun. 48 (80), 9971–9973. Wang, Y., Ge, L., Ma, C., Kong, Q., Yan, M., Ge, S., Yu, J., 2014a. Chem. Eur. J 20 (39), 12453–12462. Wang, Y., Ge, L., Wang, P., Yan, M., Ge, S., Li, N., Yu, J., Huang, J., 2013c. Lab Chip 13 (19), 3945–3955. Wang, Y., Ge, L., Wang, P., Yan, M., Yu, J., Ge, S., 2014b. Chem. Commun. 50 (16), 1947–1949. Wang, Y., Xu, J., Ma, C., Li, S., Yu, J., Ge, S., Yan, M., 2014c. J. Mater. Chem. B 2 (22), 3462–3468. Willner, I., Yan, Y.M., Willner, B., Tel-Vered, R., 2009. Fuel Cells 9 (1), 7–24. Windmiller, J.R., Wang, J., 2013. Electroanalysis 25 (1), 29–46. Wu, G., Srivastava, J., Zaman, M.H., 2014a. Anal. Biochem. 449, 147–154. Wu, X.E., Guo, Y.Z., Chen, M.Y., Chen, X.D., 2013a. Electrochim. Acta 98, 20–24.

Please cite this article as: Desmet, C., et al., Biosensors and Bioelectronics (2015), http://dx.doi.org/10.1016/j.bios.2015.06.052i

C. Desmet et al. / Biosensors and Bioelectronics ∎ (∎∎∎∎) ∎∎∎–∎∎∎ Wu, Y., Xue, P., Hui, K.M., Rang, Y., 2014b. Biosens. Bioelectron. 52, 180–187. Wu, Y., Xue, P., Kang, Y., Hui, K.M., 2013b. Anal. Chem. 85 (18), 8661–8668. Xu, Y., Lou, B., Lv, Z., Zhou, Z., Zhang, L., Wang, E., 2013. Anal. Chim. Acta 763, 20–27. Xu, Y., Lv, Z., Xia, Y., Han, Y., Lou, B., Wang, E., 2014. Anal. Bioanal. Chem. 405 (11), 3549–3558. Yan, J., Ge, L., Song, X., Yan, M., Ge, S., Yu, J., 2012. Chem. Eur. J. 18 (16), 4938–4945. Yetisen, A.K., Akram, M.S., Lowe, C.R., 2013. Lab Chip 13 (12), 2210–2251. Zhang, L., Zhou, M., Wen, D., Bai, L., Lou, B., Dong, S., 2012. Biosens. Bioelectron. 35

19

(1), 155–159. Zhang, Y., Zhou, C., Nie, J., Le, S., Qin, Q., Liu, F., Li, Y., Li, J., 2014. Anal. Chem. 86 (4), 2005–2012. Zhao, C., Thuo, M.M., Liu, X., 2013. Sci. Technol. Adv. Mater. 14 (5). Zhong, Z.W., Wang, Z.P., Huang, G.X.D., 2012. Microsyst. Technol. 18 (5). Zhu, X., Shah, P., Stoff, S., Liu, H., Li, C.Z., 2014. Analyst 139 (11), 2850–2857.

Please cite this article as: Desmet, C., et al., Biosensors and Bioelectronics (2015), http://dx.doi.org/10.1016/j.bios.2015.06.052i

Paper electrodes for bioelectrochemistry: Biosensors and biofuel cells

Paper electrodes for bioelectrochemistry: Biosensors and biofuel cells

Recommend Documents