Electronic Supplementary Information

Electronic Supplementary Material (ESI) for Chemical Communications This journal is © The Royal Society of Chemistry 2013

Electronic Supplementary Information Paper-based Photoelectrochemical Immunoassay for Low-cost and Multiplexed Point-of-care Testing Panpan Wang, Lei Ge, Shenguang Ge, Jinghua Yu,* Mei Yan, Jiadong Huang

Reagents and apparatus The All reagents were of analytical grade and directly used for the following experiments as supplied. Ultra-pure water obtained from a Millipore water purification system (resistivity ≥ 18.2 MΩcm) was used in all assays and solutions. All mouse monoclonal α-fetoprotein (AFP), carcinoma antigen 153 (CA153), carcinoma antigen 199 (CA199), and carcinoembryonic antigen (CEA) capture antibodies (Ab1), unlabeled AFP, CA153, CA199, and CEA signal antibodies (Ab2), and standard AFP, CA153, CA199, and CEA solutions were all obtained from Linc-Bio Science Co. Ltd. (Shanghai, China). Blocking buffer for the residual reactive sites on the Ab1 immobilized paper microzones was phosphate buffer solution (PBS) containing 0.5% bovine serum albumin (BSA) and 0.5% casein. Tween-20 (0.05%) was spiked into 0.01 M PBS (pH 7.4) as a washing buffer to minimize unspecific adsorption. The clinical serum samples were from Shandong Tumor Hospital. Whatman chromatography paper #1 was purchased from GE Healthcare Worldwide and used with further adjustment of size (A4 size).


Polyvinyl alcohol (PVA) (98-99% hydrolyzed, medium molecular weight), poly(dimethyldiallylammonium chloride) (PDDA) (20%, w/w in water, molecular weight = 200 000-350 000), and PIP were purchased from Alfa Aesar. Multi-walled carbon nanotubes (CNTs, diameter, 30-50 nm) were purchased from Nanoport. Co. Ltd. (Shenzhen, China). ABEI and N-Hydroxysuccinimide (NHS) were obtained from Sigma

(St.

Louis,

MO).

1-Ethyl-3-(3-dimethylaminopropyl)

carbodiimide

hydrochloride (EDC) was purchased from Pierce (Rockford, IL). ABEI-AuNPs and the labeled Ab2 (Ab2-AuNPs-ABEI) were synthesized according to previous work with modification 1. CdS NPs (5.5 nm) were prepared in water 2. The H2SO4-PVA gel electrolyte was prepared by mixing 6.0 g PVA powder with 6.0 g H2SO4 in 60 ml water and subsequent heating under stirring to ~85°C until the solution becomes clear 3

.

Design and fabrication of this μ-PPECD The shape for printing μ-PPECD was designed using Adobe illustrator CS4. As shown in Scheme S1, the wax patterns of this μ-PPECD was comprised of a PEC reaction disc (Scheme S1A,B: red-circle with a diameter of 40.0 mm) and a PEC collection pad (Scheme S1C,D) containing a blue isosceles trapezoidal support (length of the upper side: 12.0 mm, length of the under side: 40.0 mm, height: 17.0 mm) and a blue rectangle (40.0 mm×11.0 mm) with two rectangular paper legs (10.0 mm×30.0 mm). On the PEC reaction disc, there were eight unprinted paper sample zones with the same diameter of 6.0 mm (Scheme S1A). On the PEC collection pad,


an unprinted paper auxiliary zone with a diameter of 6.0 mm was designed. Furthermore, the two rectangular paper legs, comprised of unprinted rectangular area (10.0 mm × 20.0 mm) and green square area (10.0 mm × 10.0 mm) with a diagonal fold-line (1 mm in width), were designed as the electrode substrates to integrate PSC into the μ-PPECD.

Scheme S1. The schematic representation, size and shape of this μ-PPECD. (A) PEC reaction disc; (B) The reverse side of (A); (C) PEC collection pad; (D) The reverse side of (C); (E, F) Pictures of this μ-PPECD after assembling.


Wax-printing in bulk on A4 paper sheets was performed using a wax printer (Xerox Phaser 8560N color printer) set to the default parameters in a high-resolution printing mode (Figure S1 and S2). Then the wax-printed paper sheet was baked at 130 ℃ for 150 s to let the printed wax melt and penetrate through the paper to form the hydrophobic patterns (Figure S3 and S4). After the curing process, the unprinted area still maintained good hydrophilicity, flexibility, and porous structure and will not affect the further modifications 4. The as-prepared wax-penetrated paper sheets were then ready for screen-printing of circuit components (Scheme S1B,C,D) after cooling to room temperature (within 1 min), including carbon electrodes (Figure S5 and S8) on its corresponding paper zone and silver wires (Figure S6 and S7). The carbon electrodes and silver wires could firmly attach to the paper surface due to the surface roughness of paper and binding reagents promoted transport or penetration of ink into the macroporous paper substrate 5.


Figure S1. Wax-printed PEC reaction discs on a paper sheet (A4) before baking.


Figure S2. Wax-printed PEC collection pads on a paper sheet (A4) before baking.


Figure S3. Wax-printed PEC reaction discs on a paper sheet (A4) after baking.


Figure S4. Wax-printed PEC collection pads on a paper sheet (A4) after baking.


Figure S5. PEC reaction discs on a paper sheet (A4) after screen-printing of SPCWEs.


Figure S6. PEC collection pads on a paper sheet (A4) after screen-printing of silver wires.


Figure S7. PEC collection pads on a paper sheet (A4) after screen-printing of silver wires (The reverse sides of Figure S6).


Figure S8. PEC collection pads on a paper sheet (A4) after screen-printing of SPCCE.


Integration of all-solid-state PSC on μ-PPECD In a typical fabrication process, first, the two paper legs with graphite electrodes for the PSC were fabricated through pencil drawing (manual mechanical abrasion) on the surface of two paper legs as indicated in Figure S9 and Figure S10. This drawing was repeated for three times to form uniform coating on the surface of .two paper legs. Thereafter, the resulted paper sheet was cut into individual PEC collection pad by scissors. Subsequently, the two unprinted paper legs were immersed into the hot solution of H2SO4-PVA gel electrolyte (the two green squares were kept out) for 10 min and picked out. The great suitability of the H2SO4-PVA gel electrolyte for flexible and all-solid-state energy storage devices has been well demonstrated 3. The reason for the immersion in the hot solution was that, after it cooled down, the solution would become a little viscous, which would not be in favor of the thoroughly soakage of H2SO4-PVA gel electrolyte into the unprinted paper legs.


Figure S9. PEC collection pads on a paper sheet (A4) after drawing of graphite electrodes.


Figure S10. PEC collection pads on a paper sheet (A4) after drawing of graphite electrodes (Reverse side of Figure S9)..


Figure S11. (A) After immersing into the hot solution of H2SO4-PVA gel electrolyte, silver conductive adhesive (a) was spread onto the green triangle followed by folding along the diagonal fold-line (b); (B) Schematic structure (cross section) of this PSC on the μ-PPECD.

Then, silver conductive adhesive was spread onto the green triangles (Figure S11Aa) followed by folding along the diagonal fold-line (Figure S11Ab), the silver wire could glue closely to the graphite coating with the aids of the silver conductive adhesive after folding. Then, the two paper legs were stacked quickly to form a structure of graphite-film/paper/paper/graphite-film as illustrated in Figure S11B followed by performing a even pressure of ~10 MPa (a weight of ~400 g) for 10 min to glue the two paper legs together through the adhesive polymer electrolyte. In addition, the two unprinted rectangular area of the paper legs soaked with H2SO4-PVA gel electrolyte functioned as both separator and electrolyte, which can minimize the total thickness and simplify the fabrication process. Finally, the all-solid-state PSC was obtained after the evaporation of the excess water. The typical thickness of the all-solid-state PSC was ~0.50 mm. The scanning electron microscopy (SEM) images of the graphite electrodes were recorded on a QUANTA FEG-250


scanning electron microscope. Electrochemical characterizations were performed at room temperature using a CHI 660d workstation (CH Instruments Inc., Austin, TX). Figure S12 showed that the graphite layer formed a continuous and uniform coating (Figure S12A), comprised of multilayer graphenes (Figure S12B) with a large portion of edge structures (Figure S12C), on the rough surface of paper, which have been proven to be beneficial for electrochemical capacitance 6.

Figure S12. SEM images of the resulted graphitic film electrode on paper under different magnification.

Figure S13 (A) GCD curves of this PSC; (B) Durability test of this PSC by measuring 100 charge-discharge cycles.


Construction of Ab1/CdS/PDDA-CNTs/PWE Firstly, PDDA-CNTs were prepared according to the following procedure, as reported previously with some modifications 7. CNTs were first sonicated with 3:1 (v/v) H2SO4/HNO3 for 4 h to obtain carboxylic group-functionalized CNTs. The resulting dispersion was filtered and washed repeatedly with water until the pH was about 7.0. Then 0.5 mg·mL-1 CNTs were dispersed into a 0.25% PDDA aqueous solution containing 0.5 M NaCl and the resulting dispersion was sonicated for 30 min to give a homogeneous black suspension. Residual PDDA was removed by centrifugation, and the complex was rinsed with water. The collected complex was redispersed in water with mild sonicating. The PWEs on the PEC reaction disc were firstly modified through sequentially assembling positively charged PDDA-CNTs and negative charged CdS NPs onto the surfaces of interwoven cellulose fibers in paper sample zones. Briefly, 20 µL solution of PDDA-CNTs and 20 µL solution of CdS NPs was alternately dropped into the bare paper sample zone (Scheme S2A,B) and kept it for 10 min. After each dropping step, the paper sample zone was rinsed thoroughly according to the method demonstrated in

our

previous

work

8

.

Second,

the

Ab1

was

immobilized

into

the

CdS/PDDA-CNTs/PWEs (Scheme S2C) through the classic EDC coupling reactions. Typically, a fresh prepared 15 µL solution of EDC (10 mg·mL-1) and of NHS (20 mg·mL-1) were added into the CdS/PDDA-CNTs/PWE to active the COOH groups on thioglycolic acid-capped CdS NPs for 30 min followed by washing thoroughly with water, then 20 µL of 20 µg·mL-1 Ab1 solution was dropped into the activated PWE


and the PEC reaction disc was then incubated at 4 ℃ for 4 h. Subsequently, physically absorbed excess Ab1 were rinsed with phosphate buffer solution (PBS, pH 7.4). Finally, the Ab1/CdS/PDDA-CNTs/PWE was blocked by the blocking buffer for 0.5 h to cover the possible remaining active sites. Electrochemical impedance spectroscopy (EIS) was performed on an IM6x electrochemical working station (Zahner Co., Germany).

Scheme S2. Schematic representation of the fabrication procedures for Ab1/CdS/PDDA-CNTs/PWE: (A) bare PWE: (a) wax-penetrated paper, (b) SPCWE, (c) cellulose fibers in paper sample zone (unprinted macroporous paper); (B) Sequential assembling of PDDA-CNTs and CdS NPs onto surfaces of cellulose fibers in bare PWE; (C) After the immobilization of Ab1 on the surfaces of PDDA-CNTs/CdS modified cellulose fibers in PWE and the incubation with antigen and Ab2 to form Ab1/antigen/ Ab2 sandwich complexes.


Figure S14. SEM images of PDDA-CNTs entangled cellulose fibers in paper sample zone.

Characterization of Ab1/CdS/PDDA-CNTs/PWE The morphologies of PDDA-CNTs modified paper sample zones were characterized by SEM. Figure S14 shows that the nano-scale interconnected PDDA-CNTs were firmly entangled on micro-scale cellulose fibers in the form of small bundles or single tubes. The formation of PDDA-CNTs/CdS/Ab1 in PWE was further confirmed by the EIS measurements, which is an effective tool for characterizing the interface properties of electrodes. EIS were carried out in a background solution of 5 mM [Fe(CN)6]3-/4- at a bias potential of 170 mV (versus Ag/AgCl). As shown in Figure S15, the bare PWE showed a relatively small electron-transfer resistance, Ret (curve a). After PDDA-CNTs was formed on the cellulose fiber surface in paper sample zone, a much smaller Ret was observed (curve b) while the anchoring of CdS NPs showed a much larger Ret (curve c). The reason may be that the PDDA-CNTs immobilized on the cellulose fiber surfaces played an


important role similar to a conducting wire, which provided a fast electron transfer path to the SPCWE, while the assembly of CdS NPs into the PDDA-CNTs/PWE partially blocked the electron transfer of the redox probe. Remarkable increase in the Ret value was observed after both the immobilization of Ab1 (curve d) and blocking with BSA (curve e) on the surfaces of PDDA-CNTs/CdS modified cellulose fibers in PWE, indicating that the electron-transfer kinetics of the redox probe was slow down, which testified the successful immobilization of Ab1 and BSA blocking.

Figure S15. EIS of the PWE under different condition in 10.0 mM [Fe(CN)6]3-/4solution containing 0.5 M KCl: (a) bare PWE; (b) PDDA-CNTs/PWE; (c) CdS/PDDA-CNTs/PWE; (d) Ab1/CdS/PDDA-CNTs/PWE; (e) Ab1/CdS/PDDA-CNTs/PWE after blocking.

Influencing factors on PEC response For

the

optimizations

below,

the

photocurrent

generated

from

this

Ab2/antigen/Ab1/CdS/PDDA-CNTs/PWE with internal CL light source was measured directly on the electrochemical workstation at an applied potential of 0 V (vs. Ag/AgCl) and recorded as the average photocurrent between the measurements from 80 s to 240 s (Fig. 1).


The incubation process was performed at room temperature. The successful development of the multiplex immunoassay required that the common incubation time must be suitable for all analytes. All the PEC responses for increased with the increasing incubation time and then leveled off, which indicated maximum formation of Ab2/antigen/Ab1 sandwich immuno-complexes in the modified PWE. The optimal total incubation time of 10.0 ng·mL-1 AFP, 10.0 U·mL-1 CA153, 10.0 U·mL-1 CA199, and 10.0 ng·mL-1 CEA immuno-complexes was 250 s, 240 s, 260 s and 240 s, respectively. Hence, considering the optimal analytical performance and development of this method to high sample throughput, the hybridization time of 240 s was selected. The incubation process in this macroporous PWE needed shorter time compared with 40 min for the traditional electrochemical immunoassay on flat working electrode 9. This was mainly due to the high surface-to-volume ratio, incompact macroporous structure, and the small volume of the PWE (0.18 mm thickness and 6 mm in diameter). Thus, the Ab2 and antigen should diffuse only short transport distances to react with the Ab1 10. Furthermore, as the solution dried in paper sample zone of PWE, the concentration of each reagent increased; this concentration maybe further enhance the binding kinetics of Ab2/antigen/Ab1 sandwich immuno-complexes 10. The short hybridization time could be favorable to high sample throughput and rapid POC diagnosis.


Figure S16. Effects of H2O2 concentration on photocurrent response at 20.0 ng·mL-1 CEA concentration, where n = 11 for each point.

In this work, as shown in Scheme 1, H2O2 was not only used as the co-reactant in the ABEI-AuNPs-H2O2-PIP CL system, but also as the electron donors to suppress the corrosion (lattice dissolution) of CdS NPs under illumination as well as to facilitate the generation of stable photocurrent. Thus, the generated photocurrent was optimized using CEA as a model by changing the concentration of H2O2 (Figure S16) in the range of 2.0 to 5.0 mM. When the concentration of H2O2 was lower than 3.5 mM, the photocurrent increased with increasing H2O2 concentration; when the concentration of H2O2 was further elevated, the photocurrent would decrease with increasing H2O2 concentration. This may be attributed to the oxidation of the CdS NPs by excess H2O2 that yielded surface defects and traps

11

. These resulted traps

could enhance the recombination of the generated electron-hole pairs in the CdS NPs, thus inhibiting the electron injection to the PWE. Therefore, 3.5 mM H2O2 was used in this work.


Scheme S3. The schematic representation of the light use-efficiency (A) in the 3D space among the interwoven cellulose fibers in macroporous paper sample zone of PWE and (B) on relatively uniform and even surface of the ITO.

Figure S17. Enlarged photograph of the screen and buttons on DMM.

PEC operation and assay procedures of this μ-PPECD First, sample solution (10 µL) containing different concentrations of AFP, CA153, CA125, CEA were added to the corresponding modified paper sample zones of PWE and allowed to incubate for 240 s at room temperature, followed by washing


with washing buffer. Then the labeled Ab2 PBS solution (10 µL, 20 µg·mL-1) was added to corresponding PWEs, and allowed to incubate for 240 s at room temperature to form the Ab2/antigen/Ab1 complex (Scheme S2C) followed by washing thoroughly with PBS for preventing the nonspecific binding and achieving the best possible signal-to-background ratio. Thereafter, as shown in Scheme S1E,F and Figure S18A, the PEC reaction disc was reversibly installed onto the PEC collection pad by piercing through the pivot points (indicated in Scheme S1) on the PEC reaction disc and PEC collection pad using a thumbtack. Thus, the eight paper sample zones on the PEC reaction disc could exactly and successively align to the paper auxiliary zone on PEC collection pad through the rotation of the PEC reaction disc around the thumbtack to lap the two green marked lines (indicated in Scheme S1) on the PEC reaction disc and PEC collection pad, respectively. Then, the μ-PPECD was put into a model cassette (demonstrated in our previous work

12

) and clamped between two compatibly

designed circuit boards (Figure S18B) to fix and connect this μ-PPECD to the DMM. The hydrophilic paper auxiliary zone and paper sample zones constituted the reservoirs of the paper PEC cells (~50 μL) after being clamped between the two compatibly designed circuit boards. The two screen-printed electrodes (SPCWE and SPCCE) will be connected once the paper PEC cell was filled with buffer solution. The detail circuit diagram of the entire system was shown in Figure S18C. For CL excited PEC assay, 0.1 M Tris-HCl buffer solution (pH 8.5, 50 µL) containing 3.5 mM H2O2 and 0.5 mM PIP was added into the paper PEC cell using a micro-syringe through the hole on the cassette. As shown in Figure S18C, the generated


photocurrent was collected between the modified PWE on reaction tab and the carbon counter electrode on collection tab to charge the PSC for 60 s. A high instantaneous current through the DMM was obtained once the switch was closed. Finally, the Max/Min button (Figure S17) on the DMM was pressed once to display the maximum value of measurements.

Figure S18. Operation and PEC assay procedures of this μ-PPECD. (A) Configuration of this entire system: (a) the two compatibly designed circuit boards, (b) copper conductor for the connection between alligator clip and silver wire, (c) alligator clip, (d) micro-syringe, (e) DMM, (f) switch, (g) cassette, (h) black metallic cover. The centre hole on the circuit board was used for the addition of H2O2-PIP into PWE. (B) After clamping the μ-PPECD between the two circuit boards and connecting with the DMM. (C)Detail circuit diagram of the entire system.


Figure S19. The relationship between current response and concentration of AFP with (red) and without (blue) the amplification of PSC.

As shown in Figure S19, the results showed that both the photocurrent generated directly from modified PWE (IPWE) and the amplified current released from PSC (IPSC; charge time: 60 s) had a fairly good logarithmical relationship with the concentration of the AFP ranging from 0.01 ng·mL-1 to 65 ng·mL-1. The regression equation was expressed as IPWE (nA) = 423.64lg[AFP(ng·mL-1)] + 861.91 (R = 0.9962) and IPSC (nA) = 6353.83lg[AFP(ng·mL-1)] + 12931.3 (R = 0.9970) with the same limit of detection of 2.3 pg·mL-1 (3σ).


Figure S20. The relationship between current response and concentration of CA153 with (red) and without (blue) the amplification of PSC.

As shown in Figure S20, the results showed that both the photocurrent generated directly from modified PWE (IPWE) and the amplified current released from PSC (IPSC; charge time: 60 s) had a fairly good logarithmical relationship with the concentration of the CA153 ranging from 0.02 U·mL-1 to 70 U·mL-1. The regression equation was expressed as IPWE (nA) = 479.35lg[CA153 (U·mL-1)] + 816.26 (R = 0.9965) and IPSC (nA) = 7190.78lg[CA153 (U·mL-1)] + 12246.09 (R = 0.9971) with the same limit of detection of 7.1 mU·mL-1 (3σ).


Figure S21. The relationship between current response and concentration of CA199 with (red) and without (blue) the amplification of PSC.

As shown in Figure S21, the results showed that both the photocurrent generated directly from modified PWE (IPWE) and the amplified current released from PSC (IPSC; charge time: 60 s) had a fairly good logarithmical relationship with the concentration of the CA199 ranging from 0.05 U·mL-1 to 80 U·mL-1. The regression equation was expressed as IPWE (nA) = 566.16lg[CA199 (U·mL-1)] + 748.48 (R = 0.9961) and IPSC (nA) = 8491.26lg[CA199 (U·mL-1)] + 11224.21 (R = 0.9974) with the same limit of detection of 16.3 mU·mL-1 (3σ).


Assay results of real human serum Table S1. Assay results of real human serum by the proposed and reference method Samples

AFP concentration (ng·mL-1)

CA153 concentration (U·mL-1)

Proposed method

Reference method

Relative error (%)

Proposed method

Reference method

Relative error (%)

Sample-1

9.8

10.3

-5.1

10.2

9.8

3.9

Sample-2

56.5

54.7

3.1

11.9

11.4

4.3

Sample-3

60.2

57.5

4.5

16.4

15.8

3.7

Sample-4

18.5

19.3

-4.3

86.1

90.4

-5.0

CA199 concentration (U·mL-1) Samples

CEA concentration (ng·mL-1)

Proposed method

Reference method

Relative error (%)

Proposed method

Reference method

Relative error (%)

Sample-1

75.3

79.0

-5.0

30.7

31.9

-3.9

Sample-2

20.6

21.5

-4.4

48.4

51.1

-5.6

Sample-3

53.2

51.7

2.8

53.4

51.5

3.6

Sample-4

31.3

29.5

5.8

55.7

58.6

-5.2

According to the criteria suggested by Li

13

, the cut-off values of the four tumor markers

(AFP, CA153, CA199 and CEA) in human serum are 25 ng·mL-1, 35 U·mL-1, 35 U·mL-1, and 5.0 ng·mL-1, respectively. Therefore, the blood donor of sample-1 may be at the highest intestinal cancer and pancreatic cancer risk; the blood donor of sample-2 may be placed in the class of the highest ovarian cancer risk; the blood donor of sample-3 can be placed in the class of the highest cholangiocarcinoma risk; and the blood donor of sample-4 can be at the highest breast cancer risk.

Specificity of this μ-PPECD Specificity of this μ-PPECD was evaluated by comparing the current responses of a specific analyte in the absence and presence of three other non-cognate analytes as interferences. When the concentrations of the interferents changed in the range of 50∼200 ng·mL-1 or 50∼200 U·mL-1, the maximum changes of current responses for 5.0 ng·mL-1 AFP, 5.0 U·mL-1 CA153, 5.0 U·mL-1


CA199, and 5.0 ng·mL-1 CEA were 2.1, 1.7, 1.9, and 2.4%, respectively, which essentially indicated that the cross reactivity between these antibodies and their non-cognate antigens in the PWE was negligible.

Reproducibility and stability of this μ-PPECD The reproducibility of this μ-PPECD for CEA (as a model) was investigated with inter-assay precision. The RSD for the parallel detections of 0 ng·mL-1, 1.0 ng·mL-1, and 25.0 ng·mL-1 CEA with ten μ-PPECDs was 2.87%, 3.24%, and 2.76%, respectively. When this μ-PPECD was stored in a valve bag at 4 °C and measured at intervals of three days, no obvious change was observed after 4 weeks under ambient conditions. These results indicated that this μ-PPECD has better reproducibility, stability, and precision during manufacture, storage or long-distance transport to remote regions and developing countries.

Reference 1 (a) D. Tian, H. Zhang, Y. Chai and H. Cui Chem. Commun., 2011, 47, 4959-4961; (b) W. Tu, W. Wang, J. Lei, S. Deng and H. Ju Chem. Commun., 2012, 48, 6535-6537. 2 G.-L. Wang, P.-P. Yu, J.-J. Xu and H.-Y. Chen J. Phys. Chem. C, 2009, 113, 11142-11148. 3 (a) L. Yuan, X. Xiao, T. Ding, J. Zhong, X. Zhang, Y. Shen, B. Hu, Y. Huang, J. Zhou and Z. L. Wang Angew. Chem. Int. Ed., 2012, 51, 4934-4938; (b) L. Yuan, X.-H. Lu, X. Xiao, T. Zhai, J. Dai, F. Zhang, B. Hu, X. Wang, L. Gong, J. Chen, C. Hu, Y. Tong, J. Zhou and Z. L. Wang ACS Nano, 2011, 6, 656-661. 4 Y. Lu, W. Shi, J. Qin and B. Lin Anal. Chem., 2009, 82, 329-335. 5 Z. Nie, F. Deiss, X. Liu, O. Akbulut and G. M. Whitesides Lab Chip, 2010, 10, 3163-3169. 6 G. Zheng, L. Hu, H. Wu, X. Xie and Y. Cui Energy Environ. Sci., 2011, 4, 3368-3373.


7 L. Wang, S. Guo, L. Huang and S. Dong Electrochemistry Communications, 2007, 9, 827-832. 8 J. Yan, L. Ge, X. Song, M. Yan, S. Ge and J. Yu Chem. Eur. J., 2012, 18, 4938-4945. 9 Y.-J. Li, M.-J. Ma and J.-J. Zhu Anal. Chem., 2012, 84, 10492–10499. 10 C.-M. Cheng, A. W. Martinez, J. Gong, C. R. Mace, S. T. Phillips, E. Carrilho, K. A. Mirica and G. M. Whitesides Angew. Chem. Int. Ed., 2010, 49, 4771-4774. 11 X. Li, Y. Zhou, Z. Zheng, X. Yue, Z. Dai, S. Liu and Z. Tang Langmuir, 2009, 25, 6580-6586. 12 S. Wang, L. Ge, Y. Zhang, X. Song, N. Li, S. Ge and J. Yu Lab Chip, 2012, 12, 4489-4498. 13 Li TX Modern clinical immunoassay. Beijing: Military Medical Science Press; 2001. pp. 178-209.