Analytica Chimica Acta 815 (2014) 42–50

Contents lists available at ScienceDirect

Analytica Chimica Acta

journal homepage: [Link]/locate/aca

Amperometric carbohydrate antigen 19-9 immunosensor based on

three dimensional ordered macroporous magnetic Au film coupling
direct electrochemistry of horseradish peroxidase
Qi Zhang a , Xiaojun Chen a,b,∗ , Yin Tang c , Lingna Ge a , Buhua Guo a , Cheng Yao a,∗
a
College of Sciences, Nanjing Tech University, Nanjing 211816, PR China
b
State Key Laboratory of Materials-Oriented Chemical Engineering, Nanjing Tech University, Nanjing 210009, PR China
c
Zhangjiagang Hospital of Traditional Chinese Medicine, Zhangjiagang 215600, PR China

h i g h l i g h t s g r a p h i c a l a b s t r a c t

• Three dimensional ordered macropo-

rous magnetic electrode was newly
used in electrochemical immunosen-
sor.
• The large surface area of mac-
roporous magnetic electrode could
improve the immobilized amount of
antibody.
• Au nanoparticles functionalized SBA-
15 was used to immobilize enzyme
labeled Ab2 and enzyme.
• Macroporous magnetic electrode and
Au nanoparticles composite facili-
tated the direct electron transfer of
enzyme.
• The immunoassay avoided adding
electron transfer mediator, simplify-
ing the procedure.

a r t i c l e i n f o a b s t r a c t

Article history: A sandwich-type electrochemical immunosensor for the detection of carbohydrate antigen 19-9 (CA 19-9)
Received 1 December 2013 antigen based on the immobilization of primary antibody (Ab1 ) on three dimensional ordered macro-
Received in revised form 8 January 2014 porous magnetic (3DOMM) electrode, and the direct electrochemistry of horseradish peroxidase (HRP)
Accepted 12 January 2014
that was used as both the label of secondary antibody (Ab2 ) and the blocking reagent. The 3DOMM
Available online 19 January 2014
electrode was fabricated by introducing core–shell Au–SiO2 @Fe3 O4 nanospheres onto the surface of
three dimensional ordered macroporous (3DOM) Au electrode via the application of an external mag-
Keywords:
net. Au nanoparticles functionalized SBA-15 (Au@SBA-15) was conjugated to the HRP labeled secondary
Three dimensional ordered macroporous
magnetic electrode
antibody (HRP-Ab2 ) through the Au–SH or Au–NH3 + interaction, and HRP was also used as the block
Sandwich-type electrochemical reagent. The formation of antigen–antibody complex made the combination of Au@SBA-15 and 3DOMM
immunosensor exhibit remarkable synergistic effects for accelerating direct electron transfer (DET) between HRP and
Carbohydrate antigen 19-9 the electrode. Under the optimal conditions, the DET current signal increased proportionally to CA 19-
Magnetic nanocomposites 9 concentration in the range of 0.05 to 15.65 U mL−1 with a detection limit of 0.01 U mL−1 . Moreover,
Horseradish peroxidase the immunosensor showed high selectivity, good stability, satisfactory reproducibility and regeneration.
Direct electron transfer Importantly, the developed method was used to assay clinical serum specimens, achieving a good relation
with those obtained from the commercialized electrochemiluminescent method.
© 2014 Elsevier B.V. All rights reserved.

∗ Corresponding authors at: Nanjing Tech University, College of Sciences, No. 30 of South Puzhu Road, Nanjing, PR China. Tel.: +86 25 858139527; fax: +86 25 858139527.
E-mail addresses: chenxj njut@[Link] (X. Chen), yaochengnjut@[Link] (C. Yao).

0003-2670/$ – see front matter © 2014 Elsevier B.V. All rights reserved.
[Link]
 Q. Zhang et al. / Analytica Chimica Acta 815 (2014) 42–50 43

1. Introduction and antibodies [24–26]. Here, Au nanoparticles functionalized

SBA-15 (Au@SBA-15) composite with good biocompatibility and
The early cancer diagnosis plays a crucial role in screening, eval- excellent conductivity was employed to construct a labeling
uating and therapeutic treatment efficacy, which requires highly system for the fabrication of electrochemical detection platform.
sensitive methods for the accurate determination of specific pro- In this study, carbohydrate antigen 19-9 (CA 19-9) was cho-
tein biomarkers [1]. Conventional detection methods for cancer sen as a cancer biomarker model to evaluate the electrochemical
biomarkers include enzyme-linked immunosorbent assay (ELISA) immunoassay. CA 19-9 is a preferred label for pancreatic cancer,
[2], radio-immunoassay (RIA) [3] and chemiluminescence enzyme which is a highly lethiferous sarcomata and difficult to be diagnosed
immunoassay (CLEIA) [4]. However, the above methods have limi- early in current clinical medicine. Therefore, the highly sensitive
tations such as environmental pollution, poor reproducibility, high determination of serum CA 19-9 levels is of great significance. In
cost and require time-consuming separations, which cannot meet present clinic, CA 19-9 immunoassay has become a gold standard
the increasing clinical demands for the rapid detections. In addi- for pancreatic cancer diagnosis [27]. For the first time, 3DOMM Au
tion, these techniques do not work well for the detection of cancer electrode was fabricated as the immobilization substrate of pri-
markers with ultra-low biogenic concentrations. Hence, develop- mary antibody (Ab1 ) by the application of an external magnet.
ing a simple, rapid and sensitive method with low cost is a challenge The Au–SiO2 @Fe3 O4 nanospheres with small size of about 70 nm
for cancer diagnosis. could be immobilized into the nanopores of 3DOM Au electrode
Recently, electrochemical immunosensor, especially sandwich- with the pore size of about 400 nm. Both the large active surface
type immunosensor, has exhibited several advantages of simple area of 3DOMM and the streptavidin (SA)–biotin (bio) complex
pre-treatment, rapid detection, low cost and high sensitivity in could enhance the coupled amount of Ab1 on the electrode surface.
the detection of cancer markers [5–7]. Despite many advances Au@SBA-15 microparticles have high conductivity, large surface
in this field, the exploration of new strategies for the improve- area and highly open mesopores and, as a result, show high per-
ment of the simplification and sensitivity of the immunoassays formance in immobilizing enzyme and accelerating direct electron
is still being pursued. In the design and fabrication of elec- transfer (DET) between enzyme and the electrode. The combination
trochemical immunosensors, antibody immobilization and signal of Au@SBA-15 microparticles and 3DOMM exhibited remarkable
amplification are the crucial steps. Furthermore, the sensitivity of synergistic effects for accelerating DET between horseradish perox-
an electrochemical immunosensor for antigen detection at very idase (HRP) and the electrode. With the Au@SBA-15 microparticles
low concentrations can be enhanced by increasing the loading as immobilization matrix of HRP and HRP labeled secondary anti-
amount of antibodies as well as controlling the orientation of body (HRP-Ab2 ), a CA 19-9 electrochemical immunosensor was
immobilized antibodies on electrode surfaces [8]. Various types proposed based on the DET of HRP. Compared with the traditional
of nanostructured materials have been used in the fabrication of immunoassays, the omission of mediator greatly simplified the
immunosensors, due to their unique optical, electrical, catalytic and assay system. Meanwhile, HRP served as both labeling and blocking
magnetic properties [9–12]. reagent, providing the improved signal to detect CA 19-9.
The combination of nanostructured materials with electro-
chemical immunoassay opens new horizons for highly sensitive
2. Experimental
detection of biomarkers [13]. Recently, nanoporous metals have
aroused great attention due to their high surface area, low density
2.1. Reagents
and 3D bicontinuous pore–ligament structure [14]. Nanoporous Au
has been one of the most popular nanoporous metals due to its
Bio-Ab1 , CA 19-9, HRP-Ab2 and pure Ab2 were supplied by
good stability, high conductivity and good biocompatibility [15].
Roche Co., Ltd. (Shanghai, China). Bovine serum albumin (BSA,
In our previous works, we have constructed several electrochem-
96–99%) and SA were purchased from Baoman Bio-tech Co., Ltd.
ical immunosensors based on a 3D ordered macroporous (3DOM)
(Shanghai, China). Monodispersed SiO2 spheres with a diameter of
Au film electrode [16,17]. An ideal method of developing 3DOM
500 nm were purchased from Alfa Aesar. Phosphate buffer saline
Au with controlled size and spatial arrangement of pores is to
(PBS, 0.1 M) with various pH values was prepared by mixing a
use colloidal crystal films as templates for subsequent electro-
stock standard solution of NaH2 PO4 and Na2 HPO4 , which was used
deposition [18]. Magnetic beads (MBs) are recognized as a powerful
as electrolyte for all electrochemistry measurements. The wash-
and versatile tool for the development of immunosensing plat-
ing buffer (PBST) was PBS (0.1 M, pH 6.0) containing 0.05% (w/v)
forms. Their large active surface area enhances the immobilization
Tween 20. EO20 PO70 EO20 (P123), poly(diallyldimethylammonium
of high biomolecules loadings onto the solid phase of the trans-
chloride) (PDDA, 20 wt%) and glutaraldehyde (Glu, 25 wt%) were
ducer through the application of a magnetic field, exhibiting a great
purchased from Sigma-Aldrich. Stock solution of 3 wt% PDDA
convenience for separation [19]. The use of Fe3 O4 nanospheres has
was prepared in double distilled water with Tris and NaCl. The
motivated great interest in bio-applications due to their advantages
supporting electrolyte solution of cyclic voltammetry (CV) and
such as narrow size distributions, controllable size and the pos-
electrochemical impedance spectroscopy (EIS) measurements was
sibility to be functionalized easily [20]. However, the pure Fe3 O4
2 mM [Fe(CN)6 ]4−/3− (1:1) solution containing 0.1 M KCl. All other
nanospheres are very likely to aggregate. As a result, multilay-
chemicals were of analytical reagent grade and used without
ered nanosphere with a magnetic core and biocompatible shell
further purification. Double distilled water was used throughout
is a more attractive composite system in biosensing [21]. To the
the experiments. Human serum samples were obtained from the
best of our knowledge, there is no report focusing on electrochem-
Zhangjiagang Hospital of Traditional Chinese Medicine and used as
ical immunosensing of cancer biomarker by using 3DOM magnetic
received.
(3DOMM) electrode as the substrate.
Silica nanomaterials with ordered structures have not only all
the virtues of inorganic material but also a large specific surface 2.2. Apparatus
area, well-defined tunable pore sizes and good biocompatibil-
ity, which provides great opportunities for the construction of Cyclic voltammetry (CV) and chronoamperometry were per-
immunosensors [22,23]. One of the best-known discovered meso- formed with a CHI 660D electrochemical workstation (Shanghai
porous silica materials is the Santa Barbara Amorphous, SBA-15, CH Instruments Co.). A conventional three-electrode system com-
which has been used for immobilization of enzymes, nanoparticles prised a platinum wire auxiliary electrode, a saturated calomel
44 Q. Zhang et al. / Analytica Chimica Acta 815 (2014) 42–50

reference electrode (SCE) and a 3DOMM Au working electrode. Au@SBA-15 nanocomposite was as follows: 0.01 g SBA-15 was dis-
All potentials herein are referenced to the SCE. Scanning elec- persed in 5 mL 3 wt% PDDA aqueous solution and stirred for 20 min.
tron microscopy (SEM) images were recorded using field-emission After residual PDDA was removed, the positive-charged SBA-15
scanning electron microscopy (FESEM, Hitachi S4800). The ele- nanocomposites then dispersed in 200 mL Au colloid solution and
mental composition was characterized by energy-dispersive X-ray stirred for 8 h. The Au@SBA-15 was collected by centrifugation and
spectroscopy (EDS Falcon 60S, EDAX Inc.). Transmission electron then dried under vacuum at 60 ◦ C for further use.
microscopy (TEM) images were obtained from a JEOL JEM-200CX
microscope. The UV–vis absorption spectra were recorded in the 2.5. Preparation of HRP/HRP-Ab2 /Au@SBA-15 and
range of 250–700 nm using a UV–vis spectrometer (UV–vis 8500). BSA/HRP-Ab2 /Au@SBA-15 labels
The nitrogen sorption and pore diameters were measured with
a Micromeritics ASAP 2020 analyzer. The specific surface area As shown in the inset of Scheme 1, first, 1 mg of Au@SBA-15
was determined using the standard Brunauer–Emmett–Teller (BET) was dispersed in 500 ␮L of PBS (0.1 M, pH 6.0) with sonication to
method, while the pore size distribution was calculated by the obtain a homogeneous dispersion. Then, 40 ␮L HRP-Ab2 and 100 ␮L
Barrett–Joyner–Halenda (BJH) method. HRP (1 mg mL−1 ) were added to the solution, and the mixture was
allowed to react at 4 ◦ C with gentle stirring for 12 h. The reac-
2.3. Synthesis of Au–Fe3 O4 @SiO2 nanocomposite tion was based on the interaction between Au nanoparticles and
amino groups or mercapto groups on the antibody and HRP. Next,
The detailed synthesis of magnetic Fe3 O4 nanoparticles the mixture was washed with PBST, then centrifuged at 5000 rpm
was described earlier [28]. Briefly, 0.5 M aqueous solutions of for 5 min, repeatedly washed and centrifuged three times. The
FeCl2 ·4H2 O and FeCl3 ·6H2 O with the ratio of 1:2 were mixed, then, obtained HRP/HRP-Ab2 /Au@SBA-15 was stored at 4 ◦ C when not
8.0 g of PEG 4000 was added. The mixture was stirred until the reac- in use. The preparation of BSA/HRP-Ab2 /Au@SBA-15 label was the
tants were fully dissolved. Subsequently, a solution of 1.0 M NaOH same as that of HRP/HRP-Ab2 /Au@SBA-15 label, except that BSA
was dropped into the mixture with stirring. After 30 min, the black was used instead of HRP.
suspension was heated at 80 ◦ C for 30 min. The black precipitate
was collected on a permanent magnet and washed several times 2.6. Fabrication of the immunosensor
with ethanol and dried under vacuum at 60 ◦ C for 6 h. The prepara-
tion of Fe3 O4 @SiO2 and Au–Fe3 O4 @SiO2 has been described in our Au substrates were provided by Shanghai Institute of Microsys-
previous paper [29]. tem And Information Technology (Shanghai, China). Before
modification, the Au substrates were cleaned with acetone, ethanol
2.4. Preparation of Au@SBA-15 nanostructures and water in turn, and then dried under a stream of nitrogen.
A 2-mm-thick poly(dimethylsiloxane) (PDMS) film with a circu-
The mesoporous SBA-15 and Au colloid solution were syn- lar opening of 3-mm diameter formed from a predefined mold
thesized following the literatures [30,31]. The synthesis of the was bound to the Au electrode surface to confine the active

Scheme 1. Illustration of the preparation process of the electrochemical immunosensor.

 Q. Zhang et al. / Analytica Chimica Acta 815 (2014) 42–50 45

electrode area. 3DOM Au film modified electrode was prepared 3. Results and discussion
according to the literature [16]. The fabrication process of the elec-
trochemical immunosensor was illustrated in Scheme 1. Prior to 3.1. Characterization of mesoporous SBA-15 and Au@SBA-15
the experiment, the 3DOM Au film electrode was electrochemi-
cally cleaned by cyclic scanning over a potential range of 0 to 1.5 V Textural properties of the mesoporous SBA-15 and Au@SBA-15
in 0.1 M H2 SO4 at a scan rate of 100 mV s−1 until a reproducible CV were obtained from low-temperature (77 K) nitrogen adsorp-
was obtained (Scheme 1a). Subsequently, the pretreated electrode tion/desorption isotherm measurements, which allow calculation
was coated with 5 ␮L of the prepared Au–Fe3 O4 @SiO2 suspension of the specific surface area, specific pore volume and mesoporous
(10 mg mL−1 ) and then dried in air (Scheme 1b). To immobilize size distributions. These isotherms and the corresponding pore
the magnetic nanoparticles onto the electrode surface, a piece of size distributions were calculated using the BJH theory. As shown
NdFeB permanent magnet (8 mm diameter and 5 mm height) was in Fig. 1A, the shape of the obtained isotherm of SBA-15 was
adhered to the other side of the electrode. Thus, the 3DOMM elec- identified as type IV according to IUPAC classification [32] and dis-
trode was fabricated. After immersing the 3DOMM electrode into played a broad H1 type hysteresis loop characteristic of large pore
a 3 mg mL−1 SA solution for 2 h (Scheme 1c), the SA-3DOMM elec- mesoporous solids [33], indicating the presence of cylindrical pore
trode was incubated with a solution of bio-Ab1 for another 12 h at channels with some structure defects in the framework [34]. The
4 ◦ C (Scheme 1d). Following that, the resulting immunosensor was initial increase in adsorption capacity at low relative pressure is
washed with PBST to remove the physically absorbed antibodies due to monolayer adsorption on mesopores. The upward deviation
and was incubated with 1 wt% BSA for 1 h at room temperature in in the range of P/P0 = 0.4–0.8 is associated with progressive meso-
order to block possible remaining active sites and avoid the non- pores filling [35]. As for the Au@SBA-15 (inset of Fig. 1A), type H3
specific adsorption (Scheme 1e). hysteresis exhibited no limiting adsorption at high P/P0 . This behav-
ior could be caused by the existence of non-rigid aggregates of Au
2.7. Electrochemical measurements nanoparticles with SBA-15 [36].
It was noticed from Fig. 1B that the BET surface area of
The electrochemical detection of CA 19-9 was based on the typ- the synthesized SBA-15 was 547 m2 g−1 with a total pore vol-
ical procedure of sandwich-type immunoreactions. The pH 6.0 PBS ume of 0.51 cm3 g−1 and the pore size distribution centered at
was used for all the electrochemical measurements. Chronoam- around 4.8 nm. By combination of Au nanoparticles, the BET sur-
perometry was recorded in PBS at −0.2 V. First, the as-prepared face area decreased notably to 328 m2 g−1 with the pore volume
immunosensor was incubated with various concentrations of CA of 0.35 cm3 g−1 . Because the particle size of Au nanoparticles used
19-9 standards for 1 h at 37 ◦ C (Scheme 1f) and then washed with here was about 15 nm, which was much larger than the pore size of
PBST. After that the immunosensor was immersed into the pre- SBA-15, they could only be adsorbed and piled on the outer sur-
pared multi-labeled nanocomposite (HRP/HRP-Ab2 /Au@SBA-15, face of SBA-15 instead of being incorporated into the channels.
BSA/HRP-Ab2 /Au@SBA-15 or HRP-Ab2 ) buffer solution (2 mg mL−1 ) The Au@SBA-15 composite contained micropores and mesopores,
for 1 h at 37 ◦ C (Scheme 1g), followed by washing with as well as partial pore blockage and some collapsing of the meso-
PBST to remove nonspecifically bound conjugations. After the structure, which could be responsible for the dramatic decrease
background current was stabilized, 3 mM H2 O2 was added in surface area and pore volume values [37]. Also as displayed
to the PBS, and then the stable current change could be in Fig. 1B, the pore size distribution estimated here was ranging
recorded. between 40 and 90 nm, since the determined pores were mainly

Fig. 1. (A) Nitrogen adsorption/desorption isotherm for (a) SBA-15 and (b) Au@SBA-15. (B) Pore diameter distribution curves. SEM images of (C) SBA-15 and (D) Au@SBA-15.
46 Q. Zhang et al. / Analytica Chimica Acta 815 (2014) 42–50

amount of Ab1 owing to the unique physical and chemical prop-

erties of 3D nanostructures. CV was a useful technology to probe
the process of electrode modification. Fig. 3A showed the CVs of
different modified electrodes in 2 mM Fe(CN)6 4−/3− redox couple
in 0.1 M KCl solution at a scan rate of 100 mV s−1 . From the image,
we can see the current of 3DOMM electrode (curve b) increased
compared with the bare 3DOM Au electrode (curve a), owing to
the 3DOMM structure provided larger active surface and exhib-
ited better conductivity [40]. When Ab1 was immobilized on the
electrode, an obvious decline of current response was observed
because the antibody would block the electron transfer (curve c).
After blocking the non-specific adsorption with BSA (curve d), the
peak current decreased again for the reason that BSA was a kind
of protein which hindered the electron transfer. It was interest-
ing to find that the peak currents increased after CA 19-9 was
combined with Ab1 (curve e). Since the isoelectric point (pI) value
of CA 19-9 was 6.5, it was positively charged in the pH 6.0 PBS
dilution, which might help CA 19-9 layer to adsorb the negatively
charged Fe(CN)6 4−/3− , increasing the electron transfer between
Fig. 2. UV–vis spectra of (a) HRP, (b) Ab2 , (c) Au colloid solution and (d) HRP/HRP- redox probe and the electrode surface. Inspiringly, the participation
Ab2 /Au@SBA-15. of HRP/HRP-Ab2 /Au@SBA-15 increased the peak current further
(curve f). The reason might be attributed to the fact that the synthe-
sized HRP/HRP-Ab2 /Au@SBA-15 possessed high conductivity and
come from the piled type rather than the blocked one. Moreover,
good electron transfer efficiency, which facilitated the electron
just as the literature description about the characteristic of type H3
communication between the solution and the base electrode.
hysteresis loop [36], the average pore size of the Au@SBA-15 com-
Especially, the as-prepared immunosensor was used for detec-
posite was regarded as a rough estimation. The HRP and HRP-Ab2
tion of 1.0 U mL−1 CA 19-9. As seen from curve a in Fig. 3B, almost
molecules could be immobilized onto the surface of Au@SBA-15
no redox peaks were observed in pH 6.0 PBS when the as-prepared
micro-particles via the interaction between cysteine or NH3 + -lysine
immunosensor was incubated with 1.0 U mL−1 CA 19-9 for 60 min
residues of the proteins and Au nanoparticles [38]. The large sur-
at 37 ◦ C. Inspiringly, after the labeled HRP/HRP-Ab2 /Au@SBA-15
face area of Au@SBA-15 could increase the capture concentration of
composites were combined onto the electrode surface, a pair of
the Ab2 , as a result, improving the sensitivity of the immunoassay
redox peaks was observed with regard to Fe(III) to Fe(II) con-
[39].
version of the immobilized HRP (curve b). The anodic (Epa ) and
Fig. 1C showed the SEM characterization of the prepared meso-
cathodic (Epc ) peak potentials were located at −0.20 and −0.28 V at
porous SBA-15, which confirmed that it possessed a uniform size of 
100 mV s−1 , respectively (curve b). The formal potential (Eo ), taken
approximately 600 nm. After the introduction of Au nanoparticles,
by averaging potential values of Epa and Epc , was −0.24 V, which
it could be seen that a good deal of Au nanoparticles adhered evenly
was close to the −0.22 V of native HRP in solution, suggesting that
on the outside surface of SBA-15 to form Au@SBA-15 (Fig. 1D). Due
most HRP molecules preserved their native structure after being
to the large surface area and the biocompatibility of Au@SBA-15, 
immobilized onto Au@SBA-15 nanocomposites. The Eo shifted pos-
high loading level of HRP and HRP-Ab2 could be achieved for the
itively at 100 mV as compared with that reported previously in the
immunosensor applications in our work. 
literature (Eo = −0.338 V vs. SCE) [41], which means that the HRP
immobilized in this system may need less activation energy to con-
3.2. UV–vis analysis of HRP/HRP-Ab2 /Au@SBA-15 duct the DET with the electrode [42]. In addition, the large amount
of Au nanoparticles on the Au@SBA-15 could insert into the active
The Soret band of protein UV–vis spectrometry is sensitive to center of HRP and facilitate the electron exchange between HRP
the variation of the microenvironment around the proteins and 
and the 3DOMM electrode. It was even reported that the Eo of HRP
can provide information about the conformational integrity of could be shifted to −0.074 V because of strong interaction between
the proteins. As can be seen from Fig. 2, the Soret absorption of Au@CaCO3 particles with HRP [43].
the native HRP and Ab2 showed the peak at 403 and 275 nm, Upon the addition of H2 O2 in pH 6.0 PBS, an obvious catalytic
respectively, which were the characteristic absorption peaks of characteristic appeared with a dramatic increase of the reduction
proteins. Curve c showed another 521 nm peak attributed to Au current and a sharp decrease of the oxidation current, as shown in
nanoparticles of Au@SBA-15, and the immobilized HRP and Ab2 curve c. Furthermore, the reduction peak current of HRP increased
were located at nearly the same wavelength as that of the native again with the concentration of CA 19-9 increasing to 4.0 U mL−1
state, suggesting that Au@SBA-15 nanocomposite had good bio- (curve d), which provided the preliminary evidence for the quanti-
compatibility and did not induce significant denaturation of the tative detection of CA 19-9 targets within the sandwich-type assay
HRP and Ab2 molecules. The successful preparation of Ab1 -3DOMM format.
and HRP/HRP-Ab2 /Au@SBA-15 could provide a precondition for the
construction of sandwich-type immunosensor. 3.4. Performance of the immunosensor

3.3. Electrochemical characterization of the immunosensor Under the optimal conditions, a sandwich-type immunoassay
format was employed for the detection of CA 19-9 standards. The
Typically, the bioactivity and amount of the immobilized electrochemical detection was based on the DET response of immo-
biomolecules can be largely affected by the surface properties of bilized HRP in HRP/HRP-Ab2 /Au@SBA-15 bio-conjugate to the cat-
the transducer. In this work, we tried to fabricate an improved alytic reduction of H2 O2 . Along with the increase of CA 19-9 concen-
interface using 3DOMM electrode for conjugation of Ab1 molecules. tration, more and more HRP/HRP-Ab2 /Au@SBA-15 could be com-
The use of the 3DOMM is expected to increase the conjugated bined onto the 3DOMM electrode surface via the antigen–antibody
 Q. Zhang et al. / Analytica Chimica Acta 815 (2014) 42–50 47

Fig. 3. (A) CVs of modified electrode recorded in 2 mM Fe(CN)6 4−/3− containing 0.1 M KCl. (a) 3DOM Au electrode, (b) 3DOMM electrode, (c) Ab1 /3DOMM electrode, (d)
BSA/Ab1 /3DOMM electrode, (e) CA 19-9/BSA/Ab1 /3DOMM electrode, (f) HRP/HRP-Ab2 /Au@SBA-15/CA 19-9/BSA/Ab1 /3DOMM electrode. (B) CVs of (a) BSA/Ab1 /3DOMM
electrode after incubation with 1.0 U mL−1 CA 19-9 in pH 6.0 PBS, (b) CA 19-9/BSA/Ab1 /3DOMM electrode after incubation with HRP/HRP-Ab2 /Au@SBA-15 in pH 6.0 PBS, (c)
and (d) catalytic current response to 1.0 and 4.0 U mL−1 CA 19-9 in pH 6.0 PBS containing 3 mM H2 O2 , respectively.

Fig. 4. (A) Chronoamperometric response of the electrochemical immunosensor after incubation with 0, 0.05, 0.15, 0.65, 1.65, 3.65, 7.65 and 15.65 U mL−1 CA 19-9 standards
(from (a) to (h)) in pH 6.0 PBS containing 3 mM H2 O2 , respectively. (B) Calibration curves of the electrochemical immunosensor toward CA 19-9 standards with different
molecular tags of Ab2 (a) HRP/HRP-Ab2 /Au@SBA-15, (b) BSA/HRP-Ab2 /Au@SBA-15 and (c) HRP-Ab2 in pH 6.0 PBS containing 3 mM H2 O2. Error bars show the standard
deviation of five repetitive measurements.

interaction. Chronoamperometry is a more accurate electrochem- electrochemical immunoassay had a wide linear range and a very
ical technique than CV to detect the current signal. The chronoam- low LOD for CA 19-9 standards. Additionally, the analytical prop-
perometric signal, DET response of HRP, was linear proportional to erties of the electrochemical immunoassay were compared with
the amount of CA 19-9 in the solution (Fig. 4A). As shown in lin- other CA 19-9 immunosensors [44–51]. As seen from the results
ear curve a of Fig. 4B, the chronoamperometric response changes listed in Table 1, the sensitivity and linear range of present proposed
(i), calculated by the actual current (i) minus the background electrochemical immunosensor were comparable with those using
response (i0 ), increased with the increasing of CA 19-9 concentra- different chemicals as electron transfer mediators.
tion in the range from 0.05 to 15.65 U mL−1 , with a detection limit of In Refs. [48–51], the CA 19-9 concentration was detected
0.01 U mL−1 at a signal-to-noise ratio of 3 (where  is the standard through the “signal-off” strategy using DET current of HRP or
deviation of the blank, n = 11). A linear relationship between i and GOD, which was different with our “signal-on” mode. Generally,
the logarithm of CA 19-9 concentration, which could be quantified “signal-off” sensors were fabricated based on the target-induced
as the equation of ip (␮A) = 52.53 + 20.51 L−1 g CCA 19-9 (U mL−1 ), reduction of signal strength, which would suffer from limited
(R = 0.992). According to the linear equation, we could detect CA signaling capacity, in which only a maximum of 100% signal
19-9 concentration quantitatively. The results revealed that the suppression can be attained under any experimental conditions

Table 1
Analytical properties of different CA 19-9 electrochemical immunosensors.

Immunosensor Redox species Linear range (U mL−1 ) LOD (U mL−1 ) Reference

Thionine/CA 19-9/SPCE Thionine 0–144 0.2 [44]

Ab1 /DpAu/Au NPs/BSA-CNTs/Au [Fe(CN)6 ]4−/3− 0.15–150 0.06 [45]
Ab1 /HN-Pt/PB-CS/Au Prussian blue 0.5–240 0.13 [46]
HRP-Ab1 /TiO2 sol–gel/GE Catechol 3–20 2.68 [47]
HRP-Ab1 /Au–CPE HRP 2–30 1.37 [48]
GOD/Ab1 /Fe3 O4 @SiO2 –Au@mSiO2 /GOD/CS-MWCNTs/GCE GOD 0.01–476.11 0.004 [49]
HRP-Ab1 /Au NPs/SPCE HRP 0.16–15 0.10 [50]
HRP-Ab1 /Au NPs/sol–gel/SPCE HRP 0.8–190 0.3 [51]
Ab1 /3DOMM HRP 0.05–15.65 0.01 This work
48 Q. Zhang et al. / Analytica Chimica Acta 815 (2014) 42–50

[52]. To circumvent this limitation, “signal-on” sensors have been

developed, in which the signal could be generated directly and
was increased proportionally to the concentration of the target
molecule. Thus, in the “signal-off” mode, a lower precision might
be found for the higher concentration of CA 19-9 tested, and the
drawbacks would be diminished using the “signal-on” mode.

3.5. Comparison of various molecular tags of Ab2

In this work, the DET current signal of immobilized HRP was

observed and amplified based on the Au@SBA-15 nanocompos-
ite. Theoretically, HRP/HRP-Ab2 /Au@SBA-15 has three functions
herein: (1) Au@SBA-15 could provide a large surface area for the
conjugation of HRP-Ab2 . When one HRP-Ab2 on the Au@SBA-15
reacted with one CA 19-9 on the immunosensor, all the con-
jugated HRP-Ab2 could be carried over and participate in the
DET procedure, enhancing the sensitivity of the immunosensor.
(2) The strongly conductive Au@SBA-15 was introduced onto the
electrode surface by the immunoreaction, which could act as an
electron communicator facilitating the DET of HRP. (3) Using HRP Fig. 5. Calibration curves of the electrochemical immunosensor toward CA 19-9
standards based on (a) 3DOMM or (b) 3DOM Au electrode. Error bars show the
as block reagent endowed the immunoassay with higher HRP load- standard deviation of five repetitive measurements.
ing, higher detection sensitivity, wider linear range as well as lower
detection limit, enhancing the DET current signal of HRP.
In the controlled experiment, we fabricated a CA 19-9 Generally, there might be another puzzle whether a strong
immunosensor (BSA/HRP-Ab2 /Au@SBA-15/CA 19-9/Ab1 /3DOMM) signal could be derived from the non-specific absorption of
using BSA as blocking reagent, which exhibited a linear range of Ab1 /3DOMM toward HRP/HRP-Ab2 /Au@SBA-15 during the incuba-
0.5–15.65 U mL−1 with lower sensitivity (curve b, Fig. 4B). Another tion process in the absence of target CA 19-9. To clarify this point,
comparative immunosensor was developed using conventional the newly prepared immunosensor was used for determination of
HRP-Ab2 instead of HRP/HRP-Ab2 /Au@SBA-15. It could be observed 0 and 0.05 U mL−1 of CA 19-9, respectively. The DET current values
that the DET current signal of HRP nearly disappeared, owing to the of HRP corresponding to blank solution and 0.05 U mL−1 of CA 19-9
absence of Au@SBA-15, and thus the electron transfer between HRP were 0.42 and 27.45 ␮A, respectively. Furthermore, HRP/Au@SBA-
and electrode surface might be hindered (curve c, Fig. 4B). 15 was also used to detect the specificity of the immunosensor in a
pH 6.0 PBS containing 0.05 U mL−1 of CA 19-9, only about 0.53 ␮A
3.6. Effect of Au–Fe3 O4 @SiO2 on the immunosensor of the current value was observed. Hence, the change in the DET
current was mainly derived from the reaction between target CA
In this paper, the core–shell Au–Fe3 O4 @SiO2 nanoparticles 19-9 and HRP/HRP-Ab2 /Au@SBA-15, exhibiting high specificity.
were employed to construct the 3DOMM electrode for the immo-
bilization of Ab1 . The beneficial feature of the Au–Fe3 O4 @SiO2
nanoparticles mainly showed in three aspects. First, the using 3.8. Reproducibility, stability and regeneration of the
of core Fe3 O4 greatly simplified the synthesis process of the immunosensor
Au–Fe3 O4 @SiO2 composites, which were collected with a magnet
rather than complex separating steps. Second, Au–Fe3 O4 @SiO2 pos- Reproducibility of the immunosensor of CA 19-9 was assessed
sessed large active surface area for high loading of Ab1 molecules, with intra- and inter-assay precision. Intra-assay precision of the
and excellent conductivity for DET of HRP. Third, the bioconjugates immunosensor was evaluated by measuring one level of CA 19-9 for
of Au–Fe3 O4 @SiO2 with Ab1 molecules could be firmly immobi- five replicate measurements. Inter-assay precision was estimated
lized on the 3DOM Au electrode surface by an external permanent by determining one CA 19-9 level with five immunosensors. The
magnet, improving the stability of the proposed immunosensor. relative standard deviation (RSD) of intra- and inter-assay obtained
In addition, as seen from curve b in Fig. 5, a control experiment from 2.0 U mL−1 CA 19-9 was 4.2% and 5.1%, respectively, suggest-
based on 3DOM Au electrode without Au–Fe3 O4 @SiO2 displayed a ing the reproducibility of the immunosensor was acceptable.
detection linear range of 0.5–50 U mL−1 , showing a lower sensitiv- Stability is a key factor in the application and develop-
ity. On the whole, the magnetic Au–Fe3 O4 @SiO2 nanocomposites ment of the proposed immunosensor. The stability of the
could significantly amplify the response signal. HRP/HRP-Ab2 /Au@SBA-15 nanoparticles was examined. When
HRP/HRP-Ab2 /Au@SBA-15 was not in use, it was stored in 0.1 M
3.7. Selectivity and specificity of the immunosensor pH 6.0 PBS at 4 ◦ C. After a storage period of 30 days, the DET cur-
rent of the immunosensor using these HRP/HRP-Ab2 /Au@SBA-15
The selectivity of the immunosensor played an important role in nanoparticles as labels did not show an obvious decline of its initial
analyzing biological samples in situ without separation. To evalu- value. The excellent stability of HRP/HRP-Ab2 /Au@SBA-15 might
ate the selectivity of the electrochemical CA 19-9 immunosenor, we be attributed to the fact that the Au@SBA-15 provided a biocom-
challenged the system with other low-abundance proteins which patible microenvironment for stabilizing the biological activity of
might exist in many gastrointestinal malignancies, such as carci- biomolecules.
noembryonic antigen (CEA) and CA 72-4. The 1.5 U mL−1 of CA 19-9 The regeneration of the immunosensor was also performed.
solution containing 6.0 U mL−1 of interfering substances was mea- After each sandwich immunoassay, the immunosensor was
sured by the immunosensor and the current variation due to the immersed in glycine–hydrochloric acid buffer solution (pH 2.2)
interfering substances was less than 4.7% of that without interfer- for 10 min to interrupt the antigen–antibody immunocomplex. The
ences (Fig. 6A), indicating the selectivity of the immunosensor was consecutive measurements could be repeated five times with a RSD
acceptable. of 5.4%.
 Q. Zhang et al. / Analytica Chimica Acta 815 (2014) 42–50 49

Fig. 6. (A) Amperometric response of the immunosensor to (a) 1.5 U mL−1 of CA 19-9, (b) 1.5 U mL−1 of CA 19-9 + 6.0 U mL−1 CEA, and (c) 1.5 U mL−1 of CA 19-9 + 6.0 U mL−1
CA72-4. (B) Amperometric response of the immunosensor after incubation of (a) blank analyte, and (b and c) 0.05 U mL−1 of CA 19-9 using (b) HRP/Au@SBA-15 and (c)
HRP/HRP-Ab2 /Au@SBA-15 as molecular tags, respectively. Error bars show the standard deviation of five repetitive measurements.

Table 2 Acknowledgments
Comparison results of two methods obtained in serum sample.

Sample Proposed immunosensor (U mL−1 ) ECL (U mL−1 ) Bias (%) We greatly appreciate the support from the National Natural
1 4.08 4.01 1.75 Science Foundation of China (20905035) and State Key Labora-
2 7.99 8.02 −0.37 tory of Materials-Oriented Chemical Engineering (KL10-15). This
3 11.98 12.03 −0.42 work is also supported by Qing Lan Project and Overseas Research
4 14.10 14.03 4.99 & Training Program of Education Department of Jiangsu Province.

Appendix A. Supplementary data

3.9. Real sample analysis
Supplementary data associated with this article can be found, in
The analytical reliability and applicable potential of the elec- the online version, at [Link]
trochemical immunoassay were evaluated by testing four clinical
serum specimens, which were gifted from the Zhangjiagang Hos-
pital of Traditional Chinese Medicine. The results were compared References
with the referenced values obtained from the commercialized
electrochemiluminescent (ECL) automatic analyzer. As shown in [1] J.F. Rusling, C.V. Kumar, J.S. Gutkinde, V. Patel, Analyst 135 (2010) 2496.
[2] L.G. Ye, S.L. Guan, C. Zhang, K.H. Lee, S.L. Sun, J. Wei, B.G. Liu, Tumor Biol. 34
Table 2, the bias of the proposed immunosensor ranged from (2013) 1873.
−0.42% to 4.99%, indicating an acceptable agreement between the [3] R. Falzarano, V. Viggiani, S. Michienzi, B. Colaprisca, F. Longo, L. Frati, E. Anastasi,
two methods. Furthermore, it is known that the cutoff of CA 19-9 Tumor Biol. 34 (2013) 387.
[4] R.R. Liu, J.Y. Gong, J. Chen, Q. Li, C.J. Song, J. Zhang, Y.M. Li, Z.J. Liu, Y. Dong, L.H.
in diagnosing pancreatic cancer is 37 U mL−1 [53], which is higher
Chen, B.Q. Jin, Cancer Immunol. Immun. 61 (2012) 855.
than the upper limit of the linear detection range. It means that [5] S.G. Ge, L. Ge, M. Yan, X.R. Song, J.H. Yu, S.S. Liu, Biosens. Bioelectron. 43 (2013)
the serum samples of CA 19-9 with high concentrations in cancer- 425.
[6] S. Abdollah, K. Begard, F. Farden, H. Rahman, Biosens. Bioelectron. 42 (2013)
ous patients are needed to dilute for detection. Thus, the proposed
439.
immunosensor is sensitive enough for the diagnosis of pancreatic [7] X.M. Pei, B. Zhang, J. Tang, B.Q. Liu, W.Q. Lai, D.P. Tang, Anal. Chim. Acta 758
cancer at early stage. (2013) 1.
[8] F. Darain, S.P.Y. Shim, Biosens. Bioelectron. 18 (2003) 733.
[9] F. Liu, Y. Zhang, S.G. Ge, J.J. Lu, J.H. Yu, X.R. Song, S. Liu, Talanta 99 (2012) 512.
[10] Y. Xiang, Y.Y. Zhang, B.Y. Jiang, Y.Q. Chai, R. Yuan, Sens. Actuators, B 155 (2012)
4. Conclusions 317.
[11] M. Larguinho, P.V. Baptista, J. Proteomics 75 (2012) 2811.
[12] Z. Altintas, S.S. Kallempudi, U. Sezerman, Y. Gurbuz, Sens. Actuators, B 174
In summary, we have devised a new sandwich-type CA 19-9 (2012) 187–194.
immunosensor based on DET of HRP. Core–shell nanostruc- [13] M.C. Daniel, D. Astruc, Chem. Rev. 104 (2004) 293.
tured Au–Fe3 O4 @SiO2 magnetic beads were immobilized onto the [14] C.X. Xu, Q. Li, Y.Q. Liu, J.P. Wang, H.R. Geng, Langmuir 28 (2012) 1886.
[15] M. Li, M. Zhang, S.G. Ge, M. Yan, J.H. Yu, J.D. Huang, S. Liu, Sens. Actuators, B 181
surface of 3DOM Au film electrode via an external magnet. The con- (2013) 50.
structed 3DOMM electrode possessed many advantages including [16] X.J. Chen, Y.Y. Wang, J.J. Zhou, W. Yan, X.H. Li, J.J. Zhu, Anal. Chem. 80 (2008)
high magnetization, large surface area and excellent conductivity, 2133.
[17] X.J. Chen, J.J. Zhou, J. Xuan, W. Yan, L.P. Jiang, J.J. Zhu, Analyst 135 (2010) 2629.
forming an ideal platform for the high loading of primary antibod- [18] J.E.G.J. Wijnhoven, S.J.M. Zevenhuizen, M.A. Hendriks, D. Vanmaekelbergh, J.J.
ies. Additionally, Au@SBA-15 was used as immobilization matrix, Kelly, W.L. Vos, Adv. Mater. 12 (2000) 888.
which could not only simultaneously upload a high concentration [19] M. Eguílaz, M. Moreno-Guzmán, S. Campuzano, A. González-Cortés, P. Yáñez-
Sedeño, J.M. Pingarrón, Biosens. Bioelectron. 26 (2010) 517.
of HRP and HRP-Ab2 , but also effectively transfer electrons from
[20] X.D. Wang, L.J. Chen, X.R. Su, S.Y. Ai, Biosens. Bioelectron. 47 (2013) 171.
electrode surface to the redox center of HRP. This strategy dis- [21] Y. Zhuo, P.X. Yuan, R. Yuan, Y.Q. Chai, C.L. Hong, Biomaterials 29 (2008) 1501.
pensed with the need of mediators, which decreased the analytical [22] M. Hasanzadeh, N. Shadjou, M. Eskandani, M. de la Guardia, TrAC, Trends Anal.
Chem. 40 (2012) 106.
time and analytical expense. Meanwhile, the proposed electro-
[23] Z. Zhou, M. Hartmann, Chem. Soc. Rev. 42 (2013) 3894.
chemical immunosensor also exhibited good precision, acceptable [24] S. Muthukoori, N. Narayanan, M.S.S. Chanadra, S. Sethuraman, U.M. Krishnan,
stability and reproducibility, and could be used for the detection of J. Biomed. Nanotechnol. 9 (2013) 907.
CA 19-9 in real samples. This work might be extended to the deter- [25] J.S. Gao, X.Y. Zhang, Y. Yang, J. Ke, X.Y. Li, Y.B. Zhang, F. Tan, J.W. Chen, X. Quan,
Chem. Asian J. 8 (2013) 934.
mination of other biomarkers, and provide a promising potential in [26] D. Wu, Y. Zhang, L. Shi, Y.Y. Cai, H.M. Ma, B. Du, Q. Wei, Electroanalysis 25 (2013)
clinical applications. 427.
50 Q. Zhang et al. / Analytica Chimica Acta 815 (2014) 42–50

[27] D.V. Gold, D.E. Modrak, Z.L. Ying, T.M. Cardillo, R.M. Sharkey, D.M. Goldenberg, [38] D.P. Tang, R. Yuan, Y.Q. Chai, Electroanalysis 18 (2006) 259.
J. Clin. Oncol. 24 (2006) 252. [39] D. Matschulat, A. Deng, R. Niessner, D. Knopp, Analyst 130 (2005) 1078.
[28] J.P. Li, X.P. Wei, Y.H. Yuan, Sens. Actuators, B 139 (2009) 400. [40] A. Kaushik, P.R. Solanki, A.A. Ansari, S. Ahmad, B.D. Malhotra, Electrochem.
[29] X.J. Chen, J.W. Zhu, Z.X. Chen, C.B. Xu, Y. Wang, C. Yao, Sens. Actuators, B 159 Commun. 10 (2008) 1364.
(2011) 220. [41] X.X. Xu, S.Q. Liu, H.X. Ju, Sensors 3 (2003) 350.
[30] D.Y. Zhao, Q.S. Huo, J.L. Feng, B.F. Chmelka, G.D. Stucky, J. Am. Chem. Soc. 120 [42] X.H. Kang, J. Wang, Z.W. Tang, H. Wu, Y.H. Lin, Talanta 78 (2009) 120.
(1998) 6024. [43] W.Y. Cai, Q. Xu, X.N. Zhao, J.J. Zhu, H.Y. Chen, Chem. Mater. 18 (2006) 279.
[31] X.J. Chen, Z.X. Chen, J.W. Zhu, C.B. Xu, Y. Wei, C. Yao, Bioelectrochemistry 82 [44] J. Wu, Z.J. Zhang, Z.F. Fu, H.X. Ju, Biosens. Bioelectron. 23 (2007) 114.
(2011) 87. [45] Y. Zhuo, R. Yuan, Y.Q. Chai, C.L. Hong, Analyst 135 (2010) 2036.
[32] K.S.W. Sing, D.H. Everett, R.A.W. Haul, L. Moscou, R.A. Pierotti, J. Rouquerol, T. [46] H. Liu, R.J. Yu, K.F. Peng, H.W. Zhao, L. Li, X.F. Wu, Electroanalysis 22 (2010)
Siemieniewska, Pure Appl. Chem. 57 (1985) 603. 2577.
[33] J. Lynch, Physico-Chemical Analysis of Industrial Catalysts: A Practical Guide to [47] D. Du, F. Yan, S.L. Liu, H.X. Ju, J. Immunol. Methods 283 (2003) 67.
Characterisation, Ed. Technip, Paris, 2003. [48] D. Du, X.X. Xu, S.F. Wang, A.D. Zhang, Talanta 71 (2007) 1257.
[34] J. Liu, J. Yang, Q.H. Yang, G. Wang, Y. Li, Adv. Funct. Mater. 15 (2005) 1297. [49] F.X. Hu, S.H. Chen, R. Yuan, Sens. Actuators, B 176 (2013) 713.
[35] L.F. Gutierrez, S. Hamoudia, K. Belkacemi, Appl. Catal., A 425–426 (2012) [50] J. Wu, Y.T. Yan, F. Yan, H.X. Ju, Anal. Chem. 80 (2008) 6072.
213. [51] J. Wu, F. Yan, X.Y. Zhang, Y.T. Yan, J.H. Tang, H.X. Ju, Clin. Chem. 54 (2008) 1481.
[36] M. Thommes, Chem. Ing. Tech. 82 (2010) 1059. [52] Z.G. Yu, R.Y. Lai, Chem. Commun. 48 (2012) 10523.
[37] E. Rombi, M.G. Cutrufello, C. Cannas, M. Occhiuzzi, B. Onidac, I. Ferinoa, Phys. [53] X.G. Ni, X.F. Bai, Y.L. Mao, Y.F. Shao, J.X. Wu, Y. Shan, C.F. Wang, J. Wang, Y.T.
Chem. Chem. Phys. 14 (2012) 6889. Tian, Q. Liu, D.K. Xu, P. Zhao, EJSO—Eur. J. Surg. Oncol. 31 (2005) 164.