Anal. Chem.

2008, 80, 2133-2140

Electrochemical Impedance Immunosensor Based

on Three-Dimensionally Ordered Macroporous Gold
Film
Xiaojun Chen,†,‡ Yuanyuan Wang,† Jinjun Zhou,† Wei Yan,† Xinghua Li,† and Jun-Jie Zhu*,†
Key Laboratory of Analytical Chemistry for Life Science, Ministry of Education of China, School of Chemistry and Chemical
Engineering, Nanjing University, Nanjing 210093, People’s Republic of China, and College of Sciences, Nanjing University of
Technology, Nanjing, People’s Republic of China

A novel label-free immunosensor for the detection of Electrodes modified with 3DOM materials can provide large active
C-reactive protein (CRP) was developed based on a three- surface, which is promising to increase functional density and
dimensional ordered macroporous (3DOM) gold film facilitate electron exchange. Xia and co-workers reported the use
modified electrode by using the electrochemical imped- of a 3DOM gold film modified electrode for the direct electron
ance spectroscopy (EIS) technique. The electrode was transfer of hemoglobin10 and the fabrication of nonenzymatic
electrochemically fabricated with an inverted opal tem- glucose and methanol sensors.11,12 Recently, macroporous ultra-
plate, making the surface area of the 3DOM gold film up microelectrodes (UMEs)13 and 3DOM Prussian blue (PB) film
to 14.4 times higher than that of a classical bare flat one, electrodes14 have also been prepared for glucose detection. In
characterized by the cyclic voltammetric (CV) technique. these systems, the film efficiency or sensing enhancement was
The 3DOM gold film which was composed of intercon- significantly higher than that of respective nonporous films.
nected gold nanoparticles not only has a good biocom- Due to the highly sensitive and selective nature of the
patible microenvironment but also promotes the increase recognition between antigen (Ag) and antibody (Ab), immunoas-
of conductivity and stability. The CRP immunosensor was says are very useful in widespread applications such as medical
developed by covalently conjugating CRP antibodies with detection, processing quality control, and environmental monitor-
3-mercaptopropionic acid (MPA) on the 3DOM gold film ing.15 Traditional methods used in immunoassays involve radio-
electrode. The CRP concentration was measured through immunoassay (RIA) and enzyme-linked immunosorbent assay
the increase of impedance values in the corresponding (ELISA). Although they are sensitive, RIA exposes laboratory
specific binding of CRP antigen and CRP antibody. The workers to a significant safety hazard, and ELISA is tedious and
increased electron-transfer resistance (Ret) values were time-consuming. New techniques, such as electrochemistry,16
chemiluminescence,17 piezoelectricity,18 and surface plasmon
proportional to the logarithmic value of CRP concentra-
resonance,19 have attracted extensive interest in immunoassays
tions in the range of 0.1 to 20 ng mL-1. The detection of
due to their simple and specific characteristics. Among these
CRP levels in three sera obtained from hospital showed
techniques, electrochemical immunoassay has received much
acceptable accuracy. attention for its high sensitivity and low cost. As most antibodies
and antigens are electrochemically inert, the label-free technique
Since the development of the first glucose biosensor by Clark of electrochemical impedance spectroscopy (EIS) is developed
and Lyons in 1962,1 electrochemical biosensors have been of high
(6) Xu, L.; Zhou, W. L.; Frommen, C.; Baughman, R. H.; Zakhidov, A. A.;
interest due to their rapid and sensitive response as well as the Malkinski, L.; Wang, J. Q.; Wiley, J. B. Chem. Commun. 2000, 997-998.
simple and convenient operation. The emergence of nanomaterials (7) Braun, P. V.; Wiltzius, P. Adv. Mater. 2001, 13, 482-485.
has opened new opportunities for electrochemical biosensors.2 (8) Braun, P. V.; Wiltzius, P. Nature 1999, 402, 603-604.
Some particular nanomaterials, such as gold and semiconductor (9) Juárez, B. H.; Golmayo, D.; Postigo, P. A.; López, C. Adv. Mater. 2004, 16,
1732-1736.
quantum-dot nanoparticles, have already been widely used due (10) Wang, C. H.; Yang, C.; Song, Y. Y.; Gao, W.; Xia, X. H. Adv. Funct. Mater.
to their good biocompatibility.3,4 Recently, three-dimensional 2005, 15, 1267-1275.
ordered macroporous (3DOM) films have attracted increasing (11) Song, Y. Y.; Zhang, D.; Gao, W.; Xia, X. H. Chem. Eur. J. 2005, 11, 2177-
2182.
attention due to their fascinating properties, and several methods
(12) Zhang, D.; Gao, W.; Xia, X. H.; Chen, H. Y. Chin. Sci. Bull. 2006, 51, 19-
concerning the synthesis of 3DOM films have been developed.5-9 24.
(13) Szamocki, R.; Velichko, A.; Holzapfel, C.; Mücklich, F.; Ravaine, S.; Garrigue,
* To whom correspondence should be addressed. E-mail: [email protected]. P.; Sojic, N.; Hempelmann, R.; Kuhn, A. Anal. Chem. 2007, 79, 533-599.
†
Key Laboratory of Analytical Chemistry for Life Science, Ministry of (14) Qiu, J. D.; Peng, H. Z.; Liang, R. P.; Xiong, M. Electroanalysis 2007, 19,
Education of China, School of Chemistry and Chemical Engineering, Nanjing 1201-1206.
University. (15) Tang, T. C.; Deng, A.; Huang, H. J. Anal. Chem. 2002, 74, 2617-2621.
‡
College of Sciences, Nanjing University of Technology. (16) Wilson, M. S. Anal. Chem. 2005, 77, 1496-1502.
(1) Clark, L. C.; Lyons, C. Ann. N.Y. Acad. Sci. 1962, 102, 29-45. (17) Konry, T.; Novoa, A.; Shemer-Avni, Y.; Hanuka, N.; Cosnier, S.; Lepellec,
(2) Alivisatos, P. Nat. Biotechnol. 2004, 22, 47-52. A.; Marks, R. S. Anal. Chem. 2005, 77, 1771-1779.
(3) Rosi, N. L.; Mirkin, C. A. Chem. Rev. 2005, 105, 1547-1562. (18) Zuo, B. L.; Li, S. M.; Guo, Z.; Zhang, J. F.; Chen, C. Z. Anal. Chem. 2004,
(4) Katz, E.; Willner, I. Angew. Chem., Int. Ed. 2004, 43, 6042-6108. 76, 3536-3540.
(5) Velev, O. D.; Tessier, P. M.; Lenhoff, A. M.; Kaler, W. W. Nature 1999, (19) Kurita, R.; Yokota, Y.; Sato, Y.; Mizutani, F.; Niwa, O. Anal. Chem. 2006,
401, 548. 78, 5525-5531.

10.1021/ac7021376 CCC: $40.75 © 2008 American Chemical Society Analytical Chemistry, Vol. 80, No. 6, March 15, 2008 2133
Published on Web 02/23/2008
to provide a direct detection of immunospecies by measuring the grade. Ultrapure fresh water obtained from a Millipore water
change of impedance. In addition to its convenience, EIS provides purification system (MilliQ, specific resistivity >18 MΩ cm, S.A.,
a nondestructive means for the characterization of the electrical Molsheim, France) was used in all runs.
properties in biological interfaces.20 Apparatus. All electrochemical experiments were carried out
C-reactive protein (CRP) can be synthesized by the liver and on a CHI660B electrochemical workstation (Shanghai CH Instru-
is well-known as one of the classical acute-phase reactants and a ments Co.) using a traditional three-electrode system. A platinum
marker of cardiac and inflammation.21,22 Generally, the CRP level foil and a saturated calomel electrode (SCE) were used as counter
is less than 2 mg L-1 for healthy individuals. However, during an electrode and reference electrode, respectively, and the 3DOM
acute phase of inflammation, CRP levels can rapidly increase up gold film modified electrode was used as working electrode. All
to 1000-fold over normal trace amounts, which will make label- potentials herein are referenced to the SCE. The geometric area
free detection an easy task.23 CRP testing is currently performed of the working electrode was controlled by insulating tape covering
with turbidimetric and nephelometric homogeneous immunoassay the edges of SiO2 layers and determined to be 0.07 cm2. The size
technologies.24 However, these methods are not sensitive enough of the prepared monodispersed SiO2 spheres was determined from
and need expensive clinical analyzers.25 Therefore, it is important transmission electron microscopy (TEM, JEOLJEM-200CX) and
to explore some new routes to detect CRP. a Brookhaven BI-9000AT laser dynamic light scattering (DLS)
Herein, a highly sensitive label-free immunosensor for the system (Brookhaven Instruments Corporation, U.S.A.). The
detection of CRP was developed, based on EIS and a 3DOM gold morphology of the 3DOM gold films was verified by SEM (Philips
film modified electrode. The electrode was electrochemically XL30 series ESEM, using an accelerating voltage of 20 kV) and a
synthesized by an inverted opal template and characterized by Philip-X’Pert X-ray diffractometer taken with a Cu KR X-ray source.
cyclic voltammetry (CV), scanning electron microscopy (SEM), Preparation of Monodisperse Silica Spheres. Monodis-
and X-ray diffraction (XRD). The 3DOM gold film was first perse SiO2 spheres with diameter of ∼500 nm (relative standard
covalently assembled with 3-mercaptopropionic acid (MPA), and deviations 3.8%) were synthesized by using a sonochemical
then CRP antibodies were combined to the MPA-modified technique varied from the Stöber method.26 Three solutions were
electrode to yield the sensing interface. After that, the unreacted prepared in the experiments: solution A was composed of 0.44
covalent-active surface groups were passivated by bovine serum mL of TEOS and 11.2 mL of EtOH, solution B included 2.82 mL
albumin (BSA). The resulting 3DOM-modified electrode was of ammonia and 11.2 mL of EtOH, and solution C was composed
tested as an EIS immunosensor for CRP detection. To the best of of 2.2 mL of TEOS and 22.2 mL of EtOH. First, solutions A and
our knowledge, we have not found reports about the EIS B were mixed quickly under ultrasonic stirring at 25 °C for 10
immunosensor based on a 3DOM gold film electrode. This min. Then 2.78 mL of ammonia was added. After 5 min, solution
method is cost-effective, versatile, and highly sensitive for immu- C was introduced into the system. A volume of 2.78 mL of
noassays. The detailed optimization and attractive performance ammonia was successively added every 5 min, then, solution C
characteristics of the developed immunosensor are reported in was added again after another 5 min. The addition process
the following sections. repeated five times. Finally, the total system was left for 2 h after
being stirred for 5 min. The as-synthesized silica alcosol was
EXPERIMENTAL SECTION washed with anhydrous EtOH by six centrifugation/dispersion
Chemicals and Materials. CRP-Ag (10.2 mg dL-1) and cycles in order to remove impurities, such as ammonia, water,
polyclonal affinity-purified CRP-Ab (4.89 mg mL-1) were gifts from and unreacted TEOS. After characterizations, these particles were
the Pointe Biotech. Co. Ltd. Human serum samples were obtained sintered at 200 °C for 2 h, and then redispersed in EtOH.
from Nanjing Gulou Hospital and used as received. N-Hydrox- Preparation of 3DOM Gold Films. Gold substrates were
ysuccinimide (NHS) and 3-mercaptopropionic acid (MPA) were provided by the 55th Institute of China Electronic Group (Nanjing,
purchased from Aldrich (St. Louis, MO). Bovine serum albumin China). They were prepared by sputtering a 200 nm thick gold
(BSA, 96-99%) was obtained from Sigma (St. Louis, MO). 1-Ethyl- top layer onto the quartz wafers, which was previously coated with
3-(3-dimethylaminopropyl) carbodiimide hydrochloride (EDC) was a few nanometers of Cr adhesion layer under vacuum. Before use,
purchased from Pierce (Rockford, IL). Phosphate-buffered saline gold substrates were cleaned with acetone in an ultrasonic bath
(PBS buffer, 10 mM, pH 7.4) were prepared by varying the ratio for 10 min and dried under nitrogen flow, followed by immersion
of NaH2PO4 and Na2HPO4. The standard CRP-Ag solutions were in a piranha solution (7:3 v/v, H2SO4/H2O2) for 1 min in order to
prepared daily in the PBS buffer solution, and the CRP-Ab was get rid of inorganic and organic contaminants on the substrate
stored at 4 °C. All other chemicals, such as tetraethoxysilane surface. The gold substrates were subsequently rinsed thoroughly
(TEOS, 98%), anhydrous ethanol (EtOH), acetone, H2SO4, H2O2, with absolute EtOH and finally dried under nitrogen flow. Scheme
NaOH, and HAuCl4‚4H2O, NH3‚H2O (25%) were of analytical 1 shows the fabrication procedure of the preparation of the 3DOM
(20) Chen, H.; Jiang, J. H.; Huang, Y.; Deng, T.; Li, J. S.; Shen, G. L.; Yu, R. Q. gold electrode. First, the vertical deposition technique was used
Sens. Actuators, B 2006, 117, 211-218. to assemble the silica spheres on the gold substrates, forming
(21) Whicher, J.; Biasucci, L.; Rifai, N. Clin. Chem. Lab. Med. 1999, 37, 495- (111) close-packed crystals (Scheme 1b). Before metal was
503.
(22) Pentikanen, M. O.; Oorni, K.; Ala-Korpela, M.; Kovanen, P. T. J. Intern. Med. deposited, the silica colloidal crystals were sintered at 200 °C
2000, 247, 359-370. under nitrogen atmosphere for 2 h. This sintering step ensured
(23) Vikholm-Lundin, I. Biosens. Bioelectron. 2006, 21, 1141-1148. the mechanical strength of the template and the formation of small
(24) Roberts, W. L.; Moulton, L.; Law, T. C.; Farrow, G.; Cooper-Anderson, M.;
Savory, J.; Rifai, N. Clin. Chem. 2001, 47, 418-425. necks between neighboring spheres. Those small necks were
(25) Husebekk, A.; Hansson, L. O. In C-reactive Protein in Clinical Practice;
Nycomed Pharma AS: Oslo, Norway, 1999; pp 1-49. (26) Stöber, W.; Fink, A.; Bohn, E. J. Colloid Interface Sci. 1968, 26, 62-69.

2134 Analytical Chemistry, Vol. 80, No. 6, March 15, 2008

Scheme 1. Procedure for Preparation of 3DOM Gold Film Electrodes

Scheme 2. Schematic Illustration of the Stepwise Immunosensor Fabrication Process

essential for the following etching off the silica colloid template magnification shows that the 3DOM gold film was actually
steps and create interconnections among the pores (Scheme 1c). composed of interconnected nanoparticles (Scheme 1f).
The electrode area was controlled by an apertured insulating tape Fabrication of the Immunosensor. After the preparation of
covering the edge of SiO2 layers and was determined to be 0.07 the 3DOM gold film electrode, it was electrochemically cleaned
cm2 (Scheme 1d). Then, it was immersed into a mixture of 0.1% by cyclic scanning with a potential range of 0 to 1.5 V in 0.1 M
(w/w) HAuCl4 and 0.1 M HClO4 solution for 1 h prior to H2SO4 at a scan rate of 100 mV s-1. The 3DOM gold electrode
electrolysis. Gold was then electrodeposited into the interspaces was first dipped in 2 mM MPA aqueous solution for 24 h at 4 °C.
of the silica-crystal template at a potential of 0.5 V. To ensure that After thoroughly rinsed with ultrapure water to remove physically
the electrons could only be used to reduce AuIII ions, the adsorbed MPA, it was immersed in a solution with 20 mg mL-1
electrolyte solution was deaerated with N2 and N2 flowed over of EDC and 10 mg mL-1 of NHS for 1 h. After the activated 3DOM
the solution throughout the whole process (Scheme 1e). After gold film/MPA was thoroughly rinsed with PBS, it was soaked
electrodeposition, an ordered through-pore array was formed by in 489 µg mL-1 of CRP-Ab PBS solution overnight to yield sensing
dissolving the template in aqueous HF (5%) for about 5 min. The interfaces. Then, unbound antibodies were washed away with PBS.
Analytical Chemistry, Vol. 80, No. 6, March 15, 2008 2135
Figure 1. (a) SEM image of the silica template. (b) SEM image of the 3DOM gold film after removal of the silica template. (c) Cyclic
voltammograms of the macroporous gold film electrode in 0.1 M H2SO4 at a scan rate of 100 mV s-1 (solid curve). For comparison, the CV of
a flat gold electrode is also shown (dotted curve). (d) X-ray diffraction pattern of the macroporous gold film.

The unreacted covalent-active surface groups were subsequently film was obtained. As shown in Figure 1b (0.1 C charge was
passivated by reaction with 3% (w/w) BSA at room temperature passed during electrodeposition), the electrodeposited gold film
for 3 h, followed by washing carefully 3 times with PBS. The consists of an interconnected, periodic hexagonal array of mono-
fabricated CRP immunosensor was stored at 4 °C when not in dispersed pores. The gold film was electrochemically character-
use. The fabrication process is shown in Scheme 2. ized in 0.1 M H2SO4 at a scan rate of 100 mV s-1 as shown in
Electrochemical Measurements. The prepared CRP immu- Figure 1c, solid curve. For comparison, the voltammogram of a
nosensor was incubated in 0.5 mL of incubation solution con- bare flat gold electrode with the same geometric surface area is
taining different concentrations of CRP-Ag at 37 °C for 60 min also presented in Figure 1c, dotted curve. Gold oxidation started
and washed carefully with PBS. The electrochemical measure- at about 1.0 V, showing two anodic current peaks. The formed
ments including CV and EIS were performed in a degassed gold oxide was then electrochemically reduced in the negative
PBS solution containing 0.1 M KCl and 2 mM Fe(CN)63-/ potential sweep. By integrating the charge required for reducing
Fe(CN)64-. the gold oxide formed in the positive sweep, the real surface area
The real serum samples were diluted to the appropriate of the gold film electrode was determined to be 1.01 cm2.
concentrations (1 to 10 ng mL-1) with PBS buffer, respectively. Assuming that the reduction of a monolayer of gold oxide requires
The data of condition, optimization, and calibration curve were 386 µC cm-2,27 whereas the geometrical area of the bare flat gold
the average of three measurements. electrode was only 0.07 cm2, the roughness factor (Rf) of the
3DOM gold film was calculated to be 14.4. Obviously, the
RESULTS AND DISCUSSION interconnected, three-dimensionally ordered gold film electrode
Preparation of the Macroporous Gold Film Electrode. has a much larger area than the real surface, in addition to a
Monodispersed SiO2 spheres were first assembled on gold slides microenvironment conductive to the efficient alignment of protein
to form a highly ordered colloidal crystal template using the molecules due to the biocompatible affinity of Au in the 3DOM
vertical deposition technique. The SEM image showed that the structure. Figure 1d shows the XRD patterns of the 3DOM gold
silica template consisted of close-packed silica spheres with film. The position and intensity of the four diffraction peaks match
diameter of ∼500 nm (Figure 1a). In comparison with the well with (111), (200), (220), (311) crystal faces of cubic phase
traditional Stöber method, the ultrasonic-assisted preparation used Au (JCPDF04-0784). The diffraction peaks are broadening, which
here is much more facile and time-saving. Gold was then indicates that the gold product is composed of small particles with
electrochemically deposited into the interspaces of the silica-crystal average size of 32.2 nm calculated by the Debye-Scherrer
template at a potential of 0.5 V. The amount of gold deposited in equation.
the template can be determined and controlled by the charge
passed through the cell. After chemical removal of the silica (27) Kozlowska, H. A.; Conway, B. E.; Hamelin, A.; Stoicoviciu, L. J. Electroanal.
template with aqueous HF, a highly ordered macroporous gold Chem. 1987, 278, 429-453.

2136 Analytical Chemistry, Vol. 80, No. 6, March 15, 2008

Figure 2. Cyclic voltammograms of a modified 3DOM gold electrode
recorded in PBS (10 mM, pH 7.4) solution containing 0.1 M KCl and
2 mM Fe(CN)63-/Fe(CN)64-: (a) 3DOM gold electrode; (b) MPA/
3DOM gold electrode; (c) CRP-Ab/MPA/3DOM gold electrode; (d)
BSA/CRP-Ab/MPA/3DOM gold electrode; (e) CRP-Ag/BSA/CRP-Ab/
MPA/3DOM gold electrode. The scan rate was 100 mV s-1.

Electrochemical Characteristics of the Electrode Surface.

All electrochemical measurements were performed in PBS solu-
tion containing 0.1 M KCl and 2 mM Fe(CN)63-/Fe(CN)64-. The
CV of ferricyanide was chosen as a marker to investigate the
changes of the electrode behavior after each assembly step. Figure
2 shows the CV of Fe(CN)63-/4- at the 3DOM gold electrode
(curve a), MPA-combined electrode (curve b), CRP-Ab-modified
electrode (curve c), CRP-Ab-modified electrode blocked with BSA
(curve d), and CRP antigen combined electrode (curve e),
respectively. As shown in Figure 2, stepwise modification on the
3DOM gold electrode was accompanied by a decrease in the
Figure 3. (A) Nyquist plots of a modified 3DOM gold electrode
amperometric response and an increase in the peak-to-peak recorded in PBS (10 mM, pH 7.4) solution containing 0.1 M KCl and
separation between the cathodic and anodic waves of the redox 2 mM Fe(CN)63-/Fe(CN)64-: (a) 3DOM gold electrode; (b) MPA/
probe, showing that the electron-transfer kinetics of Fe(CN)63-/4- 3DOM gold electrode; (c) CRP-Ab/MPA/3DOM gold electrode; (d)
is obstructed. After the 3DOM gold film electrode was function- BSA/CRP-Ab/MPA/3DOM gold electrode; (e) CRP-Ag/BSA/CRP-Ab/
MPA/3DOM gold electrode. The frequency range is from 0.01 Hz to
alized with MPA, the electron transfer between the electrochemi-
100 kHz with a signal amplitude of 5 mV, and the scan rate was 100
cal probe and electrode surface was inhibited, owing to the mV s-1. The inset is the plots of (a) and (b). (B) Fitted (solid line) and
electrostatic repulsion of MPA with negative charges and the experimental (scattered line) Nyquist plots of impedance spectra. The
negatively charged electrochemical probe. When CRP-Ab and BSA inset is the equivalent circuit applied to fit the impedance spectroscopy
were immobilized on the electrode surface, the peak currents of using a constant phase element (CPE) instead of capacitance owing
to the rough surface in the presence of the redox couple of
the redox couple of Fe(CN)63-/Fe(CN)64- decreased again. In
Fe(CN)63-/4-.
particular, after CRP antigen molecules were combined with the
antibody molecules, an obvious disappearance of the anodic peak
and cathodic peak was observed (curve e). Subsequently, CRP antibodies were covalently combined on the
EIS can also give detailed information on the impedance MPA-modified electrode and the Ret increased again (curve c),
changes in the modification process. The impedance spectra after the electrode was blocked with BSA, the Ret increased greatly
include a semicircle portion and a linear portion. The semicircle (curve d), and Ret increased in the same way when the immun-
portion at higher frequencies corresponds to the electron-transfer- osensor was used to detect the CRP-Ag (curve e). The reason is
limited process, and the linear portion at lower frequencies that the protein layer on the electrode acted as the inert electron
represents the diffusion-limited process. The semicircle diameter and mass-transfer blocking layer, and they hindered the diffusion
equals the electron-transfer resistance, Ret. Figure 3A shows the of ferricyanide toward the electrode surface significantly. Ad-
Faradaic impedance spectra observed upon the stepwise modifica- ditionally, since the isoelectric points (pI) of CRP-Ab, CRP, and
tion process. The 3DOM gold film modified electrode reveals a BSA are 7.6, 5.1, and 4.7, respectively,29,30 at pH 7.4, CRP and BSA
very small semicircle domain (curve a), implying a very low are negatively charged, whereas CRP-Ab is positively charged,
electron-transfer resistance of the redox probe. After the electrode (28) Li, Z. P.; Liu, C. H.; Su, Y. Q.; Duan, X. R. Chem. J. Internet [Online] 2005,
was conjugated with MPA, the Ret was slightly bigger (curve b), 7, 83. http://www.chemistrymag.org.
showing that the self-assembled layers of COO- terminal groups (29) Pei, R.; Cheng, Z.; Wang, E.; Yang, X. Biosens. Bioelectron. 2001, 16, 355-
361.
on the electrode surface generated a negatively charged surface (30) Laurent, P.; Potempa, L. A.; Gewurz, G.; Fiedel, B.A.; Allen, A. C.
that reduced the ability of the redox probe to access the layer.28 Electrophoresis 2005, 4, 316-317.

Analytical Chemistry, Vol. 80, No. 6, March 15, 2008 2137

so BSA and CRP could block the electron transmission transfer Table 1. Electrochemical Impedance Results for
much more severely than CRP-Ab. In the modification process, it Stepwise Assembled Electrodes Obtained from
was obvious that the impedance values of curves d and e were Figure 3
much larger than those of previous reports,31-34 indicating that electrode Ret(Ω)
the 3DOM gold film could greatly enhance the amount of
bare gold film electrode (Au film) 68
immobilized proteins because of its larger active surface area. The MPA/Au film 152
results were consistent with the CV curves shown in Figure 2. In Ab/MPA/Au film 368
comparison with CV, the results of EIS presented more apparent BSA/Ab/MPA/Au film 1498
Ag/BSA/Ab/MPA/Au film 3354
differences to multilayers deposited on the 3DOM gold film
electrode, indicating better sensitivity.
The impedance data were fitted with commercial software
Zview2. A modified Randle’s equivalent circuit and the fitting of antibody, an obvious semicircle part of the impedance spectrum
one measured spectrum to the equivalent circuit (solid line) are was observed, and Ret increased to 368 Ω (Figure 3A, curve c). A
both shown in Figure 3B, indicating good agreement with the remarkable increase in Ret to 1498 Ω (Figure 3A, curve d) was
circuit model and the measurement system over the entire shown in the successive step of BSA obturation. The increase of
measurement frequency range. The circuit, which is often used Ret value was associated with the blocking behavior of the
to model interfacial phenomena, includes the following four assembled layer on the electrode surface for the redox probe
elements: (i) the ohmic resistance of the electrolyte solution, Rs; Fe(CN)63-/4-, reflecting in the impedance spectroscopy as the
(ii) the Warburg impedance, Zw, resulting from the diffusion of increase in the diameter of the semicircle at high frequencies.
ions from the bulk electrolyte to the electrode interface; (iii) the These results were consistent with those obtained by CV mea-
interfacial double layer capacitance (Cdl) between an electrode and surements. After the immobilization of CRP-Ag on the electrode,
a solution, relating to the surface condition of the electrode; since Ret became very high, 3354 Ω (Figure 3A, curve e). The reason
the surface of the 3DOM gold film electrode was very rough, it was that CRP antibody, antigen, and BSA are all insulated, and
had a larger real surface area; therefore, we used a constant phase they hindered the diffusion of the electrochemical probe toward
element (CPE) instead of the classical capacitance to fit the the electrode surface.
impedance data, because the electrolyte side of the interface To evaluate the reaction between Ab and Ag, we exposed the
dominated the impedance of the interface;35-37 (iv) the electron- BSA/Ab/MPA/Au film electrode to various concentrations of CRP
transfer resistance, Ret, which exists if a redox probe is present antigen CAg. The corresponding Nyquist plots of impedance
in the electrolyte solution. The parallel elements (CPE and Zw + spectra are shown in Figure 4A, and the fitting values of Ret are
Ret) of the equivalent circuit were introduced since the total current presented in Table 2. It was found that the diameter of the Nyquist
through the working interface was the sum of respective contribu- circle increased with the adding of antigen. This may be because
tions from the Faradaic process and the double layer charging. more antigen molecules bind to the immobilized antibodies in
Ideally, Zw and Rs represented the bulk properties of the electrolyte higher concentrations of antigens, which acts as a definite kinetic
solution and diffusion features of the redox probe in solution and, barrier for the electron transfer. The electrochemical CRP immu-
thus, are not affected by modifications on the electrode surface. nosensor displayed well-defined concentration dependence. As
A negligible change in Rs was observed during the modification shown in Figure 4B, a linear relation between the relative Ret
process. As shown in Figure 3A, further results demonstrate that responses and the logarithmic value of antigen concentrations was
the ohmic resistance of the solution was not affected by the observed in a range from 0.1 to 20.0 ng mL-1. The linear
modification of electrode. At the same time, as can also be seen regression equation is ∆Ret(Ω) ) 4215.8 + 2651.7 log CAg (ng
in Figure 3A, the changes in Ret were much larger than those in mL-1) with a correlation coefficient 0.992 (n ) 6). The lowest
other impedance components. Thus, Ret was a suitable signal for detection limit of CRP antigen concentration was 0.1 ng mL-1.
sensing the interfacial properties of the prepared immunosensor According to the linear equation, we could detect CRP concentra-
during all these assembly procedures. tion quantitatively. Higher serum CRP levels could be detected
Detection of CRP Antigen. The fitting values for the stepwise by an appropriate dilution with pH 7.4 PBS.
assembled layers on the electrode are presented in Table 1. For In Table 2, the change in Ret (∆Ret) is calculated by the
the macroporous gold film electrode, the value of Ret is 68 Ω, following equation:
exhibiting a nearly straight line in the Nyquist plot of impedance ∆Ret ) Ret(Ag-Ab) - Ret(BSA) (1)
spectroscopy (Figure 3A, curve a), showing a diffusion-limited
electron-transfer process. After the assembly of MPA, the value
of Ret was 152 Ω (Figure 3A, curve b), slightly bigger than that of where Ret(Ag-Ab) is the value of the electron-transfer resistance after
the bare macroporous electrode. After immobilization of CRP CRP-Ag coupling to the immobilized Ab on the gold film electrode.
Ret(BSA) represents the value of the impedance after blocking the
(31) Yuan, R.; Tang, D.; Chai, Y.; Zhong, X.; Liu, Y.; Dai, J. Langmuir 2004, 20, remaining adsorption-reactive sites by BSA. As can be seen, ∆Ret
7240-7245. increased with increasing antigen concentrations within the
(32) Huang, H. Z.; Liu, Z. G.; Yang, X. R. Anal. Biochem. 2006, 356, 208-214.
(33) Dong, H.; Li, C. M.; Chen, W.; Zhou, Q.; Zeng, Z. X.; Luong, J. H. T. Anal. detection range. However, the increases in ∆Ret were not obvious
Chem. 2006, 78, 7424-7431. at higher antigen concentrations due to steric hindrance or
(34) Yang, L.; Li, Y.; Erf, G. F. Anal. Chem. 2004, 76, 1107-1113. saturation of coupled antigen molecules.
(35) Pajkossy, T. J. Electroanal. Chem. 1994, 364, 111-124.
(36) De Levie, R. Electrochim. Acta 1965, 10, 113-130. Optimization of Experimental Conditions. The combination
(37) Scheider, W. J. Phys. Chem. 1975, 79, 127-136. of proteins on the 3DOM gold film electrode could change the
2138 Analytical Chemistry, Vol. 80, No. 6, March 15, 2008
Figure 4. (A) Faradaic impendence spectra that corresponded to the 3DOM gold electrode before and after incubating with different
concentrations of CRP antigen in PBS (10 mM, pH 7.4) solution containing 0.1 M KCl and 2 mM Fe(CN)63-/4-: (a) blank solution; curves b-g
represent 0.1, 0.5, 1.0, 5.0, 10.0, and 20.0 ng mL-1 CRP antigen, respectively. (B) Calibration curve for the CRP immunosensor.

Table 2. Electrochemical Impedance Results for to the resistance change (∆Ret) with different concentrations of
Antigen-Antibody Interaction from Figure 4A the target analyte (CRP-Ag) and contrast analyte (LDL). The
concn of antigen changes of resistance are calculated following the equations:
(ng/mL) Ret(Ω) ∆Ret(Ω)
∆Ret ) Ret(Ab-CRP) - Ret(immunosensor) (2)
0.1 3354 1856
0.5 4612 3114
1.0 5760 4262 ∆Ret ) Ret(Ab-LDL) - Ret(immunosensor) (3)
5.0 7039 5541
10.0 8603 7105
20.0 9419 7921
As can be seen in Figure 6, there is only a slight variation on
the impedance with the increase of LDL. Such small changes of
the electron-transfer resistance of the nonspecific adsorption are
interface properties of electrodes, resulting in a change of Ret. acceptable. This indicated that the observed changes of the
Several factors, such as Ag incubation temperature and incubation electron-transfer resistance with CRP antigen were due to specific
time, were investigated. antibody-antigen interaction. Without much interference from
The effect of incubation temperature on the EIS value for the nonspecific adsorption, the CRP immunosensor has good selectiv-
antibody-antigen reaction was studied in a temperature range of ity.
20-50 °C. The CRP-Ab and BSA modified 3DOM gold film Precision, Reproducibility, and Stability of the CRP
electrode was immersed in 0.1 ng/mL CRP-Ag solution at different Immunosensor. The reproducibility of the biosensor for CRP
temperatures for 60 min. The electron-transfer resistance, Ret, was was investigated with intra- and interassay precision. The intra-
recorded by Faradic impedance measurement. The maximum assay precision of the immunosensor was evaluated by assaying
response occurred at a reaction temperature of 37 °C, as shown one CRP level for three reduplicate measurements, whereas the
in Figure 5A. interassay precision was estimated by measuring one CRP level
Adsorption time also greatly affected the Ag-Ab combination. with three immunosensors prepared independently at the same
The CRP-Ab and BSA modified 3DOM gold film electrode was experimental conditions. The intra- and interassay variation
immersed in 0.1 ng/mL CRP-Ag solution at 37 °C for different coefficients obtained from 1 ng mL-1 CRP were 4.8% and 6.5%,
time periods. The effect of reaction time on the EIS response is respectively, indicating acceptable precision and fabrication re-
shown in Figure 5B. With the increase of reaction time, the producibility. The immunosensor could retain its EIS and CV
electrochemical response of the immunological reaction increased response after a storage period of 60 days in PBS (10 mM, pH
and then reached a plateau when the reaction time was longer 7.4) at 4 °C, without obvious decline. Evidently, it indicated that
than 60 min. the 3DOM gold film electrode prepared by the inverse-opal
As a result, an incubation temperature of 37 °C and reaction technique can provide large active surface, which is efficient to
time of 60 min were selected for the immunoassay of CRP antigen retain the bioactivity of antibodies. On the other hand, because
combining the immobilized antibody. of the covalent interaction between MPA-modified gold macroporous
Nonspecific Interactions. Usually, nonspecific adsorption is structure and primary amine groups in biomolecules, it could also
a major problem in label-free immunosensing, since it cannot be prevent the biomolecules from leaking out. The regeneration of
distinguished from specific adsorption in unlabeled electrochemi- the immunosensor was developed by rinsing with stripping buffer
cal sensing antigens. To confirm that the above-observed imped- of pH 2.8 Gly-HCl solution to dissociate the Ag-Ab complex.
ance changes arise from specific interaction between CRP-Ag and The as-renewed immunosensor could restore 95% of the initial
the Ab, at the same time to reveal the selectivity of the binding, value after five assay runs, showing high reusability and stability.
contrast experiments were performed. After the immobilization In addition, after five assay cycles, if the immunosensor was
of CRP-Ab and blocking with BSA, the electrode was exposed to further treated with piranha solution to violently peel all the
various concentrations of an unrelated protein LDL (low-density
lipoproteins). LDL is another important index associated with (38) Havel, R. J.; Kane, J. P. In The Metabolic Basis of Inherited Disease, 6th ed.;
heart disease.38 Figure 6 shows calibration plots that correspond McGraw-Hill: New York, 1989; Chapter 44A.

Analytical Chemistry, Vol. 80, No. 6, March 15, 2008 2139

Figure 5. (A) Effect of incubation temperature on Ag-Ab interaction. (B) The effect of Ag-Ab reaction time on the response of EIS.

performed in Nanjing Gulou Hospital for clinical diagnosis. These

results indicated the presented method is in good agreement with
the traditional clinical method. Thus, this method could satisfy
the clinical need for immunoassays of CRP levels.

CONCLUSIONS
A label-free EIS immunosensor was developed upon a 3DOM
gold film modified electrode, which was synthesized via the
inverted crystal template technique. Due to the large active
surface, the special 3DOM gold film electrode possessed superior
conductivity, activity, and adsorption capacity. Composed of gold
nanoparticles, the film also provided a good microenvironment
for the immobilization of proteins while retaining biological activity
for the immunosensor. C-reactive protein, a cardiac and inflam-
Figure 6. Calibration plots show ∆Ret of the different concentrations
mation marker, was studied in our impedance immunosensor with
of (a) CRP antigen and (b) unrelated analyte for the immunosensor.
both synthetic and real samples using this electrode. This method
Table 3. Comparison of Serum CRP Levels Determined was stable, versatile, and highly sensitive and could be applied
Using the Two Methods for other immunoassays.
turbidimetric
or nephelometric ACKNOWLEDGMENT
proposed homogeneous relative The work was supported by the National Natural Science
method immunoassay deviation (%) Foundation of China for Distinguished Young Scholars (20325516),
sample (ng/mL) (ng/mL) (ng/mL)
the Key Program (20635020), the General Program (20575026,
1 1.0 1.1 -9.1 90606016), and the Creative Research Group (20521503). We also
2 5.4 5.2 +3.8
3 17.8 18.7 -4.8 thank the support of the National Basic Research Program of
China (2006CB933201). The authors thank Professor Zhen-Lin
Wang and Professor Xing-Hua Xia, from the Department of
Physics and Department of Chemistry, Nanjing University, re-
adsorbed MPA and biomolecules from the 3DOM gold substrate
spectively, for their kind help. We also thank Professor Wei-Ping
surface, complete renewal of the modified electrode could be
Qian from the School of Biological and Medical Engineering,
achieved, with reuse lifetime of more than 20 assay runs.
Southeast University, for his warm help.
Application of the Immunosensor in Human CRP Levels.
The CRP levels in three serum samples obtained using the
proposed immunosensor are shown in Table 3, which compares Received for review October 17, 2007. Accepted January
the results with those obtained from the commercial turbidi- 4, 2008.
metric or nephelometric homogeneous immunoassay technology AC7021376

2140 Analytical Chemistry, Vol. 80, No. 6, March 15, 2008