View Article Online / Journal Homepage / Table of Contents for this issue

Analyst Dynamic Article Links < C

Cite this: Analyst, 2012, 137, 5034

www.rsc.org/analyst PAPER
Published on 10 September 2012. Downloaded by Indian Institute of Technology Kanpur on 11/29/2019 6:11:11 AM.

Signal enhancement of protein binding by electrodeposited gold

nanostructures for applications in Kretschmann-type SPR sensors
Noriko Nagase,a Kyohei Terao,*a Nobumitsu Miyanishi,b Kazunori Tamai,c Noboru Uchiyama,c
Takaaki Suzuki,a Hidekuni Takao,a Fusao Shimokawaa and Fumikazu Oohiraa
Received 4th January 2012, Accepted 8th September 2012
DOI: 10.1039/c2an35574d

We propose a novel Kretschmann-type surface plasmon resonance (SPR) sensor chip having a surface
covered with electrodeposited gold nanostructures to enhance the sensitivity of SPR biosensing. The
nanostructure is three-dimensional and has a larger surface area than a conventional flat surface chip,
which increases the amount of protein binding and also induces a large change in the effective dielectric
constant of the sensing area. The gold nanostructures were formed by electrodeposition under
galvanostatic conditions, so their size could be controlled by manipulating the deposition time and
current. The sensing characteristics, including the concentration dependence and noise level, indicated
that the performance of the resulting chip (called a Au-black chip) was equivalent to that of a
conventional sensor chip. We also determined the optimal electrodeposition conditions to obtain a
sharp SPR curve for protein detection assay; the optimal thickness of the gold layer was 50–60 nm.
Enhanced protein sensing was demonstrated by using a binding assay of anti-BSA antibody and BSA
molecules. The protein binding signal was several times higher than that of the conventional assay. The
insights into electrodeposition for SPR sensing presented here will enable improved sensitivity for
detecting low-concentration and small proteins.

Introduction
Surface plasmon resonance (SPR) sensing is a promising method
for the detection of protein molecules because it is a label-free,
real-time technique having high sensitivity.1 Therefore, SPR
sensors are widely used in biochemistry, clinical diagnosis, food
analysis, and environmental biology.2,3 Several types of sensing
techniques using SPR phenomena have been developed,
including the Kretschmann,4 Otto,5 and optical fibre types6 and
localized surface plasmon resonance (LSPR) sensing.7 Among
these techniques, Kretschmann-type SPR sensing has been the
most commonly used because of the simple setup for chip
fabrication and sensing; it is employed in commercially available
equipment from various companies, such as Biacore.8
In the Kretschmann configuration, a sensor chip having a flat
metal layer on a transparent substrate is placed on a prism, and
incident light is totally reflected at the interface between the metal
and the substrate, exciting surface plasmon waves on the metal
surface only at a specific incident angle (the SPR angle). The
intensity of the reflected light is lowest at an angle that depends
a
Department of Intelligent Mechanical Systems Engineering, 2217-20
Hayashi-cho, Takamatsu 761-0396, Japan. E-mail: terao@eng.kagawa-u.
ac.jp; Fax: +81-87-864-2346; Tel: +81-87-864-2346
b
Department of Food Life Sciences, 1-1-1 Izumino, Gunma 374-0193,
Japan
c
Rexxam Co., Ltd., 958 Ikeuchi, Konan-cho, Takamatsu 761-1494, Japan
on the refractive index of the medium in contact with the metal.
For protein detection assays, the metal layer is typically gold
with a flat surface formed by physical vapour deposition, because
gold is resistant to chemical reactions and forms a suitable base
for successive chemical treatments for ligand immobilization.
Analyte proteins in solution bind to the ligands on the sensor
chip, changing the refractive index near the chip surface; these
changes enable the amount of ligand binding to be measured.
For further applications of SPR sensing to clinical diagnosis
and environmental monitoring, the sensitivity must be increased
in order to expand the detection limit toward small and low-
concentration proteins.9 One approach is to increase the amount
of protein binding by increasing the surface area of the gold
layer, as proposed for other surface biosensors, including elec-
trochemical10 and quartz crystal microbalance (QCM)
sensors.11,12 Several fabrication techniques, such as electron
beam lithography13 and nanotemplating,14 are suitable for this
purpose. However, their equipment cost and fabrication time
make them undesirable. To overcome these drawbacks, we
focused on electrodeposition of gold nanostructures15,16 that
allows the facile fabrication of sensor chips having increased
surface areas with a relatively simple setup.
In this paper, we evaluated electrodeposited gold nano-
structures fabricated under various deposition conditions for
applications in protein detection by Kretschmann-type SPR
sensing. Enhanced protein sensing was demonstrated by using a

5034 | Analyst, 2012, 137, 5034–5040 This journal is ª The Royal Society of Chemistry 2012
 View Article Online

binding assay of anti-BSA antibody and BSA; this demonstra-

tion provides insights into the potential and limitations of elec-
trodeposited gold chips in SPR biosensing.

Materials and methods

Published on 10 September 2012. Downloaded by Indian Institute of Technology Kanpur on 11/29/2019 6:11:11 AM.

Principles

In SPR biosensing, ligand molecules are chemically immobilized

onto the surface of a sensor chip, and specific analyte molecules in
solution are bound to them.17 The resulting change in the refrac-
tive index in the sensing field is detected by a shift in the SPR angle.
A conventional SPR chip has a flat gold layer on a glass substrate
(Fig. 1a). In contrast, an SPR chip with gold nanostructures (also
known as Au-black16) such as that proposed here has a larger
surface area than a flat chip (Fig. 1b), which increases the number
of both ligand and analyte molecules binding to the chip. In a Au-
black chip, the refractive index in a sensing field increases more
than for a flat chip, inducing a large shift in the SPR angle. Thus,
the resulting SPR sensor is expected to be more sensitive.
The sensing field of an SPR chip is limited to the range at
which evanescent waves propagate perpendicular to the chip
surface. The distance from the chip surface d is described as
l
d¼ pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi; (1)
2p np sinq  ns 2
2

where l, np, ns, and q are the wavelength of the light, refractive
index of the prism, refractive index of the sample, and incident
angle, respectively.18 In our setup, d is estimated to be approxi-
mately 200 nm, assuming l ¼ 820 nm, ns ¼ 1.333, and ns ¼ 1.5168.
This value determines the limit of the height of the electro-
deposited gold nanostructures because this configuration can
barely detect proteins more than 200 nm away from the interface.
SPR model for a Au-black chip
The SPR curve for a Au-black chip is simulated by using a five-
layer model consisting of glass, chromium, gold, Au-black, and
buffer layers (Fig. 2a). The Au-black layer is a mixture of gold
and the buffer; its average ‘effective’ dielectric constant eAu,b can
be determined by the effective medium theory. We adopt a
Bruggeman equation reported in the literature19 for the Au-black
layer. For a two-component system, it is described as
eAu  eAu;b eb  eAu;b
va þ ð1  va Þ ¼ 0;
eAu þ 2eAu;b eb þ 2eAu;b
where va, eAu and eb are the volume fraction of gold and the
dielectric constants of gold and the buffer solution, respectively.
The SPR curve of the chip was calculated from eqn (2) and the
Fresnel equations (Fig. 2b).
Fig. 2 SPR model of an Au-black chip. (a) Five-layer model for eval-
uation of changes induced by protein binding and (b) calculated SPR
curves for a flat chip and a Au-black chip in protein sensing.
The thickness of the Au-black layer, dAu-b, was determined to
be 43 nm by direct observation of the cross-section, as shown in
the SPR curve measurement section. The thicknesses of the base
metals (Cr and Au) were 3 nm and 12 nm, respectively, which are
also the actual experimental values. The volume fraction va was
taken as 0.75 on the basis of a curve fitting to the experimental
SPR curve. We also calculated the SPR curve for a flat chip
having a gold layer thickness of 47 nm for comparison with that
of the Au-black chip.
Protein binding to a gold surface changes both the effective
dielectric constant and the thickness of the Au-black layer. In this
situation, the layer is considered as a three-component system
(2) governed by the following Bruggeman equation.
0 eAu  eAu;b;p ep  eAu;b;p  0  eb  eAu;b;p
va þ vp þ 1 va  vp ¼ 0;
eAu þ 2eAu;b;p ep þ 2eAu;b;p eb þ 2eAu;b;p
(3)
where va0 , vp, and ep are the volume fractions of gold and protein
in the layer and the dielectric constant of the protein, respec-
tively. The thickness of the layer dAu,b,p can be written as the sum
of dAu,b and the thickness of the bound protein dp, as follows.

dAu,b,p ¼ dAu,b + dp. (4)

Then, the volume fraction, va0 , is written as

Fig. 1 Enhancement of the amount of protein binding to the SPR chip
0 dAu;b
with gold nanostructures. (a) Flat chip and (b) electrodeposited gold chip va ¼ va : (5)
(Au-black chip).
dAu;b þ dp

This journal is ª The Royal Society of Chemistry 2012 Analyst, 2012, 137, 5034–5040 | 5035

When the surface area of the electrodeposited gold is increased
by a factor of n compared with that of the flat base metal, the
volume fraction of protein vp is approximated as follows.
ndp
vp z : (6)
dAu;b þ dp
Published on 10 September 2012. Downloaded by Indian Institute of Technology Kanpur on 11/29/2019 6:11:11 AM.
The values of va0 and vp were calculated to be 0.69 and 0.26,
respectively, from eqn (5) and (6), where we used n ¼ 3 and dp ¼ 4
nm. Then, we substituted these values into eqn (3) and calculated
eAu,b,p, where the dielectric constant eb was set to 2.5, from which
the SPR curve of the protein-bound Au-black chip was
simulated.
The shifts in the SPR angle induced by protein binding were
0.56 in the flat chip and 5.26 in the Au-black chip, showing an
increase of 9.4 times in the signal by the electrodeposited gold
nanostructures. The increase exceeds the increase in the surface
area (n ¼ 3), suggesting that the signal enhancement results from
not only the surface area increase, but also a large dielectric
constant change induced by protein binding in the Au-black
layer.
The shape of the SPR curve for the Au-black chip was broader
than that of the flat chip. A sharp SPR curve, which has a small
half-width, generally allows precise determination of the SPR
angle. Thus, although the Au-black chip produces a large angle
shift, it potentially causes low resolution of the SPR angle.
Chip fabrication
We used glass substrates for SPR sensing and Si substrates for
CV measurements and SEM observations. Glass substrates (10
mm  12 mm  0.3 mm, Matsunami Glass Ind. Ltd.) were
cleaned in a piranha solution (H2SO4 : H2O2 ¼ 3 : 1) at 90  C for
10 min. A 4 in. Si substrate of 300 mm thickness was cut into small
pieces (10 mm  12 mm) and cleaned in a NH4OH : H2O2 (3 : 1)
solution for 15 min. The substrates were rinsed with pure water
and dried. A base gold layer was deposited onto a substrate using
a vacuum deposition system (VPC-1100, ULVAC Inc.). For flat
chip fabrication, chromium and gold were successively deposited
to thicknesses of 3 nm and 47 nm, respectively; the thicknesses
were monitored using a quartz crystal oscillator (CRTM-6000,
ULVAC Inc.) inside the vacuum chamber, and chromium was
used as the adhesive layer between the substrate and the gold
film. Two types of Au-black chip were prepared with gold base
layers of different thicknesses: (Cr 3 nm, Au 12 nm) and (Cr 3
nm, Au 22 nm). After deposition, the chips were rinsed with pure
water and dried. Then, the gold surface was cleaned to remove
organic residues by a 20 min treatment in a UV ozone cleaning
device (144AX, Jelight Company, Inc.).
A polydimethylsiloxane (PDMS) block of 10 mm thickness
was punched to form an 8 mm diameter hole (area: 50.24 mm2)
and adsorbed onto the gold surface of a chip while a 350 ml
electrodeposition solution [83 mM Na(AuCl4), 1.58 mM
Pb(CH3COO)2–H2O] was dispensed. Electrodeposition was
performed as described by Imamura et al.10 under galvanostatic
conditions. We used a conventional three-electrode system with a
Pt counter electrode (002222, BAS Inc.), an Ag/AgCl reference
electrode (RE-1B, BAS Inc.), and a working electrode (gold layer
on a substrate), which were connected to a galvanostat (2090,
Toho Technical Research Co. Ltd.). Electrodeposition was
5036 | Analyst, 2012, 137, 5034–5040
View Article Online
performed at a constant current of 10 mA, 30 mA, or 50 mA
for different times between 5 min and 90 min. The substrate was
subsequently rinsed with pure water and dried. As a control
experiment, we also used substrates without electrodeposition.
Chip verification
SEM observations. The electrodeposited gold surfaces were
observed by field emission scanning electron microscopy (FE-
SEM, JSM-7001F, JEOL Ltd.) at an acceleration voltage of 15
kV and a working distance of 9.5–10 mm for each electrodepo-
sition condition. To observe the cross-section, the electro-
deposited Si substrates were sectioned by hand using their
crystalline anisotropy. Each sample was mounted by colloidal
graphite (12660, Electron Microscopy Sciences) on an observa-
tion stage. The thickness of the gold layer was measured by image
analysis using ImageJ. The ridgelines of the gold nanostructures
were measured, and their average height was calculated from 20
positions selected randomly from a SEM photograph. We
defined this as the thickness of the Au-black layer.
SPR curve measurement. We measured the SPR curves of Au-
black chips to determine the optimal deposition time for SPR
sensing using a Kretschmann-type SPR measuring instrument
(Handy SPR, NTT-AT Corp.). We assumed incident angles of
65 to 75 and a 770 nm wavelength light source whose spot on
the chip has an ellipsoidal shape with a semi-major axis of 0.9
mm and a semi-minor axis of 0.3 mm. In the SPR measurements,
we measured the SPR curves of the two types of Au-black chips
having base gold layers of different thicknesses (12 nm and 22
nm). The measurements were performed in situ during electro-
deposition to evaluate changes in the shape of the curve in real
time. The electrodeposition conditions were as described in
Section 3.2.
Sensor characterization. The concentration dependence of the
SPR angle was characterized by measuring sucrose solutions of
various concentrations with a flat chip (Cr 3 nm, Au 47 nm) and a
Au-black chip (Cr 3 nm, Au 22 nm, Au-black 50 mA, 7.5 min
electrodeposition). We measured 0, 5, 10, 15, and 20% sucrose
(196-00015, Wako Pure Chemical Industries, Ltd.)/H2O solu-
tions at room temperature and plotted the shifts in the SPR
angles (N ¼ 4). To check the background noise, we monitored the
SPR angles of the Au-black and flat chips for 150 s using phos-
phate buffer saline (PBS, pH 7.2, Nacalai Tesque, Inc.) solution,
from which we calculated the standard deviations.
CV measurements. The electrodes described in Section 3.2 were
dipped in 0.5 M H2SO4 (96% H2SO4, Kanto Chemical Co. Inc.)
and connected to a potentiostat. The potential applied to the chip
was swiped from 0.6 V to 1.6 V at a speed of 100 mV s1 using a
function generator (WF1974, NF Corp.), and the data on the
electric current were stored in an oscilloscope (TDS2042B,
Tektronix Inc.). Cyclic voltammetry (CV) measurements were
performed three times for each deposition condition (0 mA per
0 min [flat chip], 50 mA per 7.5 min, 50 mA per 14 min). From
the cyclic voltammograms, we calculated the chips’ surface areas,
as described in the literature.20
This journal is ª The Royal Society of Chemistry 2012
 View Article Online

Protein detection assay

We evaluated the binding of BSA molecules to an anti-BSA

antibody immobilized on the gold surface. We used an assay kit
purchased from Rexxam Co., Ltd. First, an SPR chip was treated
with piranha solution (H2SO4 : H2O2 ¼ 3 : 1) for 30 min. Next,
Published on 10 September 2012. Downloaded by Indian Institute of Technology Kanpur on 11/29/2019 6:11:11 AM.

the substrate was rinsed with pure water and dried. It was treated
by a UV ozone cleaner for 30 min and then dipped in 20 ml of 10
mM 4,40 -dithiodibutyric acid (Dojindo Molecular Technologies,
Inc.)/ethanol solution for 12–15 hours to introduce carboxyl
groups on the surface. The substrate was then rinsed with pure
water and dried. The SPR chips were placed on the prism of an
SPR sensing system (RPK-2010T, Rexxam Co., Ltd.) at a
temperature of 24  0.5  C. A PDMS block of 10 mm thickness
was punched to form an 8 mm diameter hole and adsorbed onto
the gold surface of a chip to form a well. A protein binding assay
was performed in a batch-type solution exchange in the well. First,
250 ml of PBS solution was dispensed into the well, and the SPR
angle was allowed to stabilize for about 60 min. Then, we removed
the 200 ml of PBS and applied 200 ml of 0.1 M N-hydroxy-
succinimide (NHS) and 0.4 M 1-ethyl-3-[3-dimethylaminopropyl]
carbodiimide hydrochloride (EDC) and incubated the chip for 12 Fig. 3 Gold nanostructures formed under various electrodeposition
min to activate the surface. The substrate was rinsed with PBS conditions. Numbers indicate deposition time (min).
solution. Then, 200 ml of 50 mg ml1 anti-BSA antibody–PBS was
SPR curve measurement
introduced into the PDMS well and incubated for 12 min, during
which the molecules were immobilized on the chip by a reaction SPR curve development during electrodeposition was measured
between NHS and EDC. Next, the substrate was rinsed with PBS in situ using a 22 nm thick gold substrate (Fig. 4a), from which we
solution. Then, 200 ml of 100 mg ml1 BSA solution was applied to determined the optimal deposition time for SPR sensing. As
the PDMS well and incubated for 12 min. Again, the substrate was electrodeposition proceeded, the SPR curve changed such that
rinsed with PBS solution. This experiment was performed five the lowest intensity of the reflected light at the SPR angle
times for each deposition condition to monitor the time course of decreased gradually over 7.5 min. The intensity then began to
the shift in the SPR angle (sensorgram). increase. A sharp SPR curve, which has a low intensity at the
SPR angle and a small half-width, generally allows more precise
Results and discussion and stable determination of the SPR angle, and thus is suited to
the detection of protein binding. Therefore, we identified 7.5 min
Structure of electrodeposited gold at 50 mA as the optimal electrodeposition conditions. We refer
A number of studies have shown various shapes of gold nano- to chips produced under these conditions as AB1 chips, where
structures obtained by controlling the potential,21,22 the electric AB stands for Au-black. We also determined that the optimal
current,14,19 or the chemical composition of the solutions.11,23 In conditions for a substrate having a thinner gold base layer (Au 12
this paper, we electrodeposited gold under galvanostatic condi- nm) were 14 min at 50 mA (AB2 chips). The SPR curves of these
tions as described in the literature.19 We set the current to 10 two chips exhibited sharp dips at their respective SPR angles in
mA, 30 mA, or 50 mA and varied the deposition time between 5 our setup; the curves were shaped like that of a conventional flat
min and 90 min. SEM images show spherical gold nanostructures
on the base gold layer (Fig. 3). The diameter of the spheres
ranged from 10 nm to 20 nm at a deposition time of 5 min. The
size increased with the deposition time and was approximately 50
nm at 10 mA and 30 mA and a time of 20 min. The gold
nanostructures deposited at 50 mA for 20 min were relatively
large (50–70 nm). At 90 min, the size increased to 100 nm at 10
mA and 30 mA; at 50 mA, the structures showed a spike-like
shape with a length of 200 nm.
The shape of the electrodeposited gold nanostructures depends
on the balance between nanostructure growth and gold nucle-
ation.23 The results shown here indicate that the growth mode
was dominant at 50 mA, which is consistent with the literature.
Little growth appeared at a current of 10 mA, whereas relatively
high growth speed and shape development appeared at 50 mA, Fig. 4 SPR curves of Au-black chips: (a) in situ measurement of
which is expected to produce a larger surface area. Thus, we development of SPR curve (AB1 chip) and (b) comparison of SPR curves
selected a current of 50 mA for the following experiments. under optimal deposition conditions.

This journal is ª The Royal Society of Chemistry 2012 Analyst, 2012, 137, 5034–5040 | 5037
Published on 10 September 2012. Downloaded by Indian Institute of Technology Kanpur on 11/29/2019 6:11:11 AM. View Article Online

Fig. 5 Cross-sections of Au-black chips: (a) AB1 and AB2 chips and (b) Fig. 6 Concentration dependence of SPR angle.
comparison of metal layer thickness.
Table 1 Experimental and theoretical sensitivities of Au-black chips and
a flat chip in sucrose sensinga
chip, although the SPR angles were different (Fig. 4b). The
difference might be due to the composition of the Au-black layer, Flat AB1 AB2
as predicted in the theoretical SPR model (Fig. 2b).
The cross-sections of the AB1 and AB2 chips were observed by Experiment 0.1425 0.1637 0.1763
Theory 0.1234 0.1637 0.1725
FE-SEM, which demonstrated the thickness of the gold layer
a
(Fig. 5). The gold nanostructures had a hemispherical shape and Sensitivity: angle shift/sucrose concentration [deg. per %]. Theoretical
were stacked vertically, forming a three-dimensional structure, values are calculated from the five-layer model by changing eb.
particularly in the AB2 chip (Fig. 5a). The thickness of the gold
layer was approximately 50 nm, which was almost the same as i.e., relatively high sensitivities in the Au-black chip, suggesting
that of the conventional flat chip (Fig. 5b). This indicates that the that the theoretical model is valid.
optimal thickness in our setup is 50–60 nm, suggesting that a
thinner gold base layer allows the deposition of a thicker gold
nanostructure layer (i.e., more complicated surface structure), Surface area measurement
and thus can be expected to produce a large surface area. In our
experiments, the gold nanostructure layer of the AB2 chip was We also characterized the real surface area of a flat chip, an AB1
thicker than that of the AB1 chip, which was 43 nm. chip, and an AB2 chip by CV measurements (Fig. 7a), which
were based on measurements of the electrochemical reactions in
an oxygen monolayer on a gold surface (Au2O3).11 The estimated
Sensor characteristics surface areas produced under different deposition conditions, as

We measured sucrose solutions having various concentrations in

order to verify the sensing characteristics of the AB1, AB2, and
flat chips (Fig. 6). The SPR angle increased linearly as the
concentration increased for all the chips. For the AB1 and AB2
chips, y ¼ 0.1637x and y ¼ 0.1763x, respectively, where x is the
sucrose concentration and y is the shift in the SPR angle; this
indicates a relatively high sensitivity compared to that of the flat
chip (y ¼ 0.1425x). Regarding the background noise, which
governs the lowest detection limit, the AB1 and flat chips had
noise levels of 0.0023 and 0.0029 , respectively, indicating that
the detection limit of a Au-black chip is comparable to that of a
conventional chip in our setup. These results indicate that a Au-
black chip provides almost the same sensing performance as a
flat chip and does not prevent SPR phenomena, so we conclude
that such a chip is applicable to protein detection by SPR bio-
sensing. In addition, the sensing was carried out by a standard
Kretschmann-type SPR sensor system without modification of
the optical system and sensing conditions, indicating that it is
highly compatible with the conventional system.
The sensitivities of the chips were compared with the theo-
retical values calculated from the five-layer model described in Fig. 7 Surface area measurement by CV: (a) cyclic voltammograms and
the SPR model section (Table 1). They show a similar tendency, (b) comparison of Au-black chips to a flat chip.

5038 | Analyst, 2012, 137, 5034–5040 This journal is ª The Royal Society of Chemistry 2012
Published on 10 September 2012. Downloaded by Indian Institute of Technology Kanpur on 11/29/2019 6:11:11 AM.
Fig. 8 Protein detection using anti-BSA antibody–BSA binding assay. (a) Time course of SPR angle from antibody immobilization to analyte binding
(AB2 chip). SPR angle shift (b) after immobilization of anti-BSA molecules (ligand binding) and (c) after BSA binding (analyte binding).
obtained from the cyclic voltammograms, were 2.32 and 3.28
times higher in the AB1 and AB2 chips, respectively, than in the
flat chip (Fig. 7b). The surface area of the AB2 chip was
apparently larger than that of the AB1 chip. This is attributed to
the three-dimensional structure of the thick electrodeposition
layer, as described in the SPR curve measurement section. These
results also indicate that the electrodeposited structures have a
larger surface area on a thinner base metal layer. In our setup, the
minimum thickness of the base gold layer was empirically
determined to be 15 nm because of difficulty with controlling the
thickness of the layer during thermal deposition.
Protein detection assay
We evaluated the amount of protein binding to a Au-black chip
using anti-BSA antibody and BSA as the ligand and analyte,
respectively. The sensorgram (time course of the SPR angle) of
the AB2 chip shows a curve typical of conventional SPR
measurements (Fig. 8a), indicating successful immobilization of
the ligands and binding of the analyte. It also shows quite a low
drift level, which allows SPR measurements under stable condi-
tions. The angle shift induced by BSA solution was smaller than
that by antibody solution in spite of its high concentration. This
might be due to the fact that the thin sensing field of SPR was
largely occupied by antibody molecules bound to the Au-black
surface. The protein binding signal was evaluated from the SPR
angle shift (Fig. 8b and c). The AB1 and AB2 chips exhibited 1.64
times and 2.38 times the anti-BSA immobilization signal of the
flat chip, respectively. The BSA binding signal was greater by
This journal is ª The Royal Society of Chemistry 2012
View Article Online
1.43 times (AB1) and 2.22 times (AB2). The signal of both ligand
and analyte molecules bound to the surface of the Au-black chips
increased in comparison with those of the flat chip, indicating
that the increase in surface area enhances the amount of protein
binding and also that binding induces large changes in the
dielectric constant of the Au-black layer, as expected.
These results also indicate that the increases in the protein
binding signal agreed well with the increases in the surface area,
although they were much lower than that described in the SPR
model section (an increase of 9.4 times in the theoretical model).
The space between the gold nanostructures was 50 nm, which
was similar in scale to the size of the protein molecules (antibody,
11 nm (ref. 24); BSA, 7 nm (ref. 25)). Thus, the decrease is
attributed to the limited protein diffusion due to the steric
hindrance of gold nanostructures.26 This phenomenon has also
been reported in QCM sensing.11
Conclusions
We demonstrated a novel Kretschmann-type SPR sensor chip
having a nanostructured gold surface in order to enhance the
signal of protein binding to the sensor surface. Three-dimen-
sional gold nanostructures were formed by electrodeposition
under galvanostatic conditions. We found that the size of the
nanostructures can be controlled by manipulating both the
current and the deposition time. The sensing characteristics of
electrodeposited chips having a gold layer with a thickness of 50–
60 nm were equivalent to those of a conventional sensor chip.
The effect on protein binding was evaluated by using anti-BSA
Analyst, 2012, 137, 5034–5040 | 5039

antibody and BSA molecules. The signals of ligand and analyte
binding increased by factors of 2.38 and 2.22, respectively. These
results indicate that the electrodeposited SPR chip will provide
improved sensitivity for detecting low concentrations and small
molecules.
Published on 10 September 2012. Downloaded by Indian Institute of Technology Kanpur on 11/29/2019 6:11:11 AM.
Acknowledgements
This work was supported by the KAKENHI (21651062) of the
Ministry of Education, Culture, Sports, Science and Technology
(MEXT), Japan and the Fund for Kagawa University Young
Scientists 2010 and 2011.
Notes and references
1 J. Homola and G. Gauglitz, Sens. Actuators, B, 1999, 54, 3–5.
2 R. Ince and R. Narayanaswamy, Anal. Chim. Acta, 2006, 569, 1–20.
3 J. Homola, Chem. Rev., 2008, 108, 462–493.
4 K. Kurihara and K. Suzuki, Anal. Chem., 2002, 74, 696–701.
5 I. R. Tamm, P. Dawson, M. P. Connolly, S. H. Raza and
H. S. Gamble, J. Mod. Opt., 1991, 38(8), 1593–1598.
6 R. Slavik, J. Homola and J. Ctyroky, Sens. Actuators, B, 1999, 54, 74–
79.
7 K. A. Willets and R. P. van Duyne, Annu. Rev. Phys. Chem., 2007, 58,
267–297.
8 R. L. Rich and D. G. Myszka, J. Mol. Recognit., 2007, 20, 300–366.
9 X. D. Hoa, A. G. Kirk and M. Tabrizian, Biosens. Bioelectron., 2007,
23, 151–160.
5040 | Analyst, 2012, 137, 5034–5040
View Article Online
10 M. Imamura, T. Haruyama, E. Kobatake, Y. Ikariyama and
M. Aizawa, Sens. Actuators, B, 1995, 24, 113–116.
11 K. Bonroy, J. M. Friedt, F. Frederix, W. Laureyn, S. Langerock,
A. Campitelli, M. Sara, G. Borghs, B. Goddeeris and P. Declerck,
Anal. Chem., 2004, 76, 4299–4306.
12 D. van Noort, R. Rani and C. F. Mandenius, Mikrochim. Acta, 2001,
136, 49–53.
13 W. Rechberger, A. Hohenau, A. Leitner, J. R. Krenn, B. Lamprecht
and F. R. Aussenegg, Opt. Commun., 2003, 220, 137–141.
14 B. M. I. vanderZande, M. R. Bohmer, L. G. J. Fokkink and
C. Schonenberger, J. Phys. Chem. B, 1997, 101, 852–854.
15 C. A. Foss, G. L. Hornyak, J. A. Stockert and C. R. Martin, J. Phys.
Chem., 1992, 96, 7497.
16 S. Toyama, M. Someya, O. Takei, T. Ohtake, R. Usami, K. Horikoshi
and Y. Ikariyama, Chem. Lett., 2001, 30(2), 160–161.
17 W. M. Mullett, E. P. C. Lai and J. M. Yeung, Methods, 2000, 22, 77–
91.
18 W. Knoll, Annu. Rev. Phys. Chem., 1998, 49, 569–638.
19 S. Toyama, O. Takei, M. Tsuge, R. Usami, K. Horikoshi and S. Kato,
Electrochem. Commun., 2002, 4(7), 540–544.
20 S. Trasatti and O. A. Petrii, Pure Appl. Chem., 1991, 63, 711–734.
21 Y. Tian, H. Q. Liu, G. H. Zhao and T. Tatsuma, J. Phys. Chem. B,
2006, 110, 23478–23481.
22 Y. P. Deng, W. Huang, X. Chen and Z. L. Li, Electrochem. Commun.,
2008, 10, 810–813.
23 M. S. El-Deab, T. Sotomura and T. Ohsaka, J. Electrochem. Soc.,
2005, 152, C730–C737.
24 T. Jossang, J. Feder and E. Rosenqvist, J. Protein Chem., 1988, 7,
165–171.
25 H. B. Bohidar, Colloid Polym. Sci., 1989, 267, 292–300.
26 W. M. Deen, AICHE J., 1987, 37, 1497–1510.
This journal is ª The Royal Society of Chemistry 2012