4526 Langmuir 2002, 18, 4526-4529

Three-Dimensionally Ordered Macroporous upon the change in refractive index of the solution near
Polymer Materials: An Approach for Biosensor the pore surface that occurs during analyte binding.
Applications Interestingly, 3DOM polystyrene films for biosensor
applications not only tap their diffraction properties but
Weiping Qian,†,‡ Zhong-Ze Gu,† Akira Fujishima,§ and also employ other characteristics of the material, including
Osamu Sato*,† high surface area and simplicity of biomolecular im-
mobilization. This approach to the detection of ligand-
Kanagawa Academy of Science and Technology, receptor binding does not require labeling of the analyte,
KSP Bldg. East 412, 3-3-1 Sakado, Takatsu, Kawasaki-shi, can eliminate disturbance due to nonspecific binding, and
Kanagawa 213-0012, Japan, National Laboratory of is sufficiently simple and general that it may find use in
Molecular and Biomolecular Electronics,
Southeast University, Si Pai Lou 2, biochemical assays and immunoassays.
Nanjing 210096, People’s Republic of China, and Recent developments using colloidal crystal templating
Department of Applied Chemistry, School of Engineering, allow the preparation of 3DOM materials8-17 that display
The University of Tokyo, 7-3-1 Hongo Bunkyo-ku, the existence of optical stop bands, in which strong
Tokyo 113-8565, Japan diffraction effects limit the optical transmission of the
films. Under normal incidence, the diffraction peak
Received December 17, 2001. position and the shift of the film can be estimated from
In Final Form: March 12, 2002 Bragg’s law:
See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

Biosensor experiments involve immobilizing one reac-

Downloaded via INDIAN INST OF TECH KANPUR on August 28, 2019 at 11:44:48 (UTC).

tant on a surface and monitoring its interaction with a λ ) 1.663dna (1)

second component in solution. In general, biosensors
consist of two components: a highly specific recognition where λ is the peak wavelength, d represents the center-
element and a transducer that converts the molecular to-center distance, and na is the average reflective index
recognition event into a quantifiable signal. Signal trans- of the film. As in the case of enzyme-linked immunosorbent
duction has been accomplished with electrochemical,1 assays, polystyrene became the choice of a solid substrate
field-effect transistor,2 optical absorption, fluorescence, in our experiments because of the hydrophobic nature of
surface plasmon resonance,3 interferometric,4,5 and other its surface and excellent optical properties. Proteins could
devices.6 In these devices, biomolecules such as antibodies be simply immobilized on the pore surfaces of 3DOM
or oligonucleotides are immobilized on a solid substrate polystyrene substrates by physical adsorption. In our
by numerous steps and used to detect the presence of a biosensor (Figure 1), binding of an analyte to its corre-
target antigen or oligonucleotide. Asher et al. developed sponding recognition partner, immobilized on the 3DOM
a sensing material that reported on analyte concentrations polystyrene substrate, results in a change in the average
via diffraction of visible light.7 Their material was a refractive index of the layer medium and is detected as
mesoscopically periodic crystalline colloidal array (CCA) a corresponding shift in the diffraction peak position.
of polymer spheres polymerized within a hydrogel that Assuming that the average refractive index is determined
swelled and shrank reversibly in the presence of certain from the volume average of the dielectric constants of
analytes. The CCA diffracts light at wavelengths deter- different compositions, we obtain
mined by the lattice spacing, which gives rise to an intense
color. The chemical molecular-recognition events (i.e., λ ) 1.633d ×
crown ethers for metal ions) cause the gel to swell owing
to an increased osmotic pressure, which increases the xVpolystyrenenpolystyrene2 + Vproteinnprotein2 + Vwaternwater2
mean separation between the colloidal spheres and so (2)
shifts the Bragg peak of the diffracted light to longer
wavelengths. We present here a new method for biosensor where npolystyrene, nprotein, and nwater are the reflective indices
fabrication, in which three-dimensionally ordered macro- of polystyrene, protein, and water, respectively, fixed at
porous (3DOM) polystyrene films were used as an im- their typical bulk values npolystyrene ) 1.59, nprotein ) 1.42,
mobilizing and transducing matrix. Proteins were simply and nwater ) 1.33. Vpolystyrene, Vprotein, and Vwater are the
immobilized on the pore surfaces of 3DOM polystyrene volume fractions of polystyrene, protein, and water,
substrates by physical adsorption. The signal transduction respectively. The shift in the diffraction peak position with
was achieved by monitoring the diffraction peak shifts protein binding can be used to quantitatively estimate
the amount of bound protein.
* To whom correspondence should be addressed. E-mail: sato@ We formed colloid crystal templates of silica spheres
fchem.chem.t.u-tokyo.ac.jp. (∼190 nm in diameter; Nissan Chemical Ind., Ltd., Japan)
† Kanagawa Academy of Science and Technology.
‡ Southeast University.
(8) Velev, O. D.; Jede, T. A.; Lobo, R. F.; Lenhoff, A. M. Nature 1997,
§ The University of Tokyo.
389, 447.
(1) Hickman, J. J.; Ofer, D.; Laibinis, P. E.; Whitesides, G. M.; (9) Holland, B. T.; Blanford, C. F.; Stein, A. Science 1998, 281, 538.
Wrighton, M. S. Science 1991, 252, 688. (10) Wijnhoven, J. E. G. J.; Vos, W. L. Science 1998, 281, 802.
(2) McConnell, H. M.; Owicki, J. C.; Parce, J. W.; Miller, D. L.; Baxter, (11) Velev, O. D.; Tessier, P. M.; Lenhoff, A. M.; Kaler, E. W. Nature
G. T.; Wada, H. G.; Pitchford, S. Science 1992, 257, 1906. 1999, 401, 548.
(3) Rich, R. L.; Myszka, D. G. Curr. Opin. Biotechnol. 2000, 11, 54. (12) Jiang, P.; Cizeron, J.; Bertone, J. F.; Colvin, V. L. J. Am. Chem.
(4) Lin, V. S.; Motesharei, K.; Dancil, K. P.; Sailor, M. J.; Ghadiri, Soc. 1999, 121, 7957.
M. R. Science 1997, 278, 840. (13) Park, S. H.; Xia, Y. N. Adv. Mater. 1998, 10, 1045.
(5) Yu, F.; Yao, D. F.; Qian, W. P. Clin. Chem. 2000, 46, 1489. (14) Park, S. H.; Xia, Y. N. Chem. Mater. 1998, 10, 1745.
(6) Sensors Update; Baltes, H., Gopel, W., Hesse, J., Eds.; Wiley- (15) Johnson, S. A.; Ollivier, P. J.; Mallouk, T. E. Science 1999, 283,
VCH: Weinhein, Germany, 1989-1996; Vols. 1-9. 963.
(7) (a) Holtz, J. H.; Asher, S. A. Nature 1997, 389, 829. (b) Holtz, J. (16) Jiang, P.; Hwang, K. S.; Mittleman, D. M.; Bertone, J. F.; Colvin,
H.; Holtz, J. S. W.; Munro, C. H.; Asher, S. A. Anal. Chem. 1998, 70, V. L. J. Am. Chem. Soc. 1999, 121, 11630.
780. (17) Jiang, P.; Bertone, J. F.; Colvin, V. L. Science 2001, 291, 453.

10.1021/la0118199 CCC: $22.00 © 2002 American Chemical Society

Published on Web 04/26/2002
Notes Langmuir, Vol. 18, No. 11, 2002 4527

Figure 1. (A) Schematic diagrams of the solvent evaporation

methods used to fabricate large-area 3DOM polystyrene films
in a free-standing form. (B) and (C) show SEM images, at Figure 2. Schematic representation of a biosensor fabrication.
different magnifications, of 3DOM polystyrene films, prepared (A) An aqueous solution is filled in the pores of a 3DOM
using a ∼190 nm SiO2 colloidal crystal as a template and the polystyrene film. (B) The ligand is immobilized in the pores. (C)
scheme in Figure 1A. Note the excellent ordering in macroporous The places where no ligand immobilization has occurred are
polystyrene films. blocked with BSA. (D) The analyte binds on the pore surfaces.

on microslides using literature procedures18,19 with some biosensor is shown in Figure 2. To immobilize proteins on
modifications and improvements. We simultaneously the pore surfaces in 3DOM polystyrene films, we filled
prepared multiple colloidal crystal templates with same the pores with ethanol first because of the hydrophobic
quality. A glass trough together with a stand was used as nature of the polystyrene surface and then replaced the
the experimental cell. An important feature of our method ethanol solution with an aqueous solution (Figure 2A).
is that all of the microslides were mounted vertically in Ligand (600 µL) at a concentration of 2.0 mg/mL in 50
a stand and kept parallel to each other. Careful control mmol/L PBS at pH 7.2 was added to each sample cell.
over the growth conditions makes it possible for us to After 1 h incubation at room temperature and overnight
obtain high-quality colloidal crystal templates. The de- incubation at 4 °C, the ligand solutions were removed.
tailed description of our method can be found in another After the ligand was immobilized in the pores (Figure
manuscript.20 We infiltrated a toluene solution of poly- 2B), we blocked the places where no protein adsorption
styrene (average molecular weight of 312 000; Wako Pure had occurred with bovine serum albumin (BSA, 10 mg/
Chemical Industries, Ltd., Japan) into the vacant space mL), to prevent nonspecific adsorption of proteins during
between the spheres. After 2-3 days, the polystyrene film subsequent steps (Figure 2C).21 The pores were subse-
together with the template was spontaneously separated quently rinsed and filled with the sample containing the
from the microslide, and then the template was removed putative target analyte (Figure 2D). In situ analyses of
by treatment with 4% hydrofluoric acid, producing large- the ligand immobilization, BSA blocking, and analyte
area polystyrene films with three-dimensional ordering binding on the pore surfaces were performed using a
of pores in a free-standing form. A schematic outline of Shimadzu UV-3101 PC with transmission mode. All
the procedure for producing 3DOM polystyrene films is transmission measurements were performed at normal
shown in Figure 1A. The 3DOM polystyrene films used incidence and constant regions. The instrument resolution
in this study were carefully washed with ethanol. Figure is 0.1 nm, and the spot size of the detecting light was 11
1B,C shows typical SEM images of a 3DOM polystyrene mm × 4 mm. Each peak wavelength maximum was
film with a ∼190 nm diameter. The images exhibit an acquired from the recorded transmission spectrum using
ordered void structure from the silica spheres. These large the software IGOR Pro version 4.0 (WaveMetric, Inc.,
cavities are not isolated but rather are interconnected to USA).
each other. We used polystyrene films immobilized with Staphy-
The experimental cell was assembled from two mi- lococcal protein A (SpA) and goat anti-human IgG (goat
croslides, one silicone rubber spacer and one polystyrene anti-hIgG) to test the validity, sensitivity, and selectivity
film. A “U” type silicone rubber spacer was mounted on of the biosensors. Human IgG (hIgG) was used as an
the macroporous side of a polystyrene film, and then the analyte, because it can be captured by both immobilized
film together with the space was sandwiched between SpA and goat anti-hIgG. SpA is a membrane-bound protein
two microslides. After the fabrication, the experimental from Staphylococcus aureus that binds to the Fc fragment
cell was precisely mounted on the holder supplied by the of IgG.22,23 A whole range of antibodies with different
producer. A schematic for a 3DOM polystyrene film based specificities can also be dissociated from the SpA surface
simply by lowering the pH of the solutions.24 Goat anti-
(18) Dushkin, C. D.; Nagayama, K.; Miwa, T.; Kralchevsky, P. A.
Langmuir 1993, 9, 3695. (21) Qian, W. P.; Yao, D. F.; Yu, F.; Xu, B.; Zhou, R.; Bao, X.; Lu, Z.
(19) Jiang, P.; Bertone, J. F.; Hwang, K. S.; Colvin, V. L. Chem. Mater. Clin. Chem. 2000, 46, 1456.
1999, 11, 2132. (22) Lindmark, R.; Thoren-Tolling, K.; Sjoquist, J. J. Immunol.
(20) Qian, W. P.; Gu, Z.-Z.; Fujishima, A.; Sato, O. Langmuir, Methods 1983, 62, 1.
submitted. (23) Forsgren, A.; Sjoquist, J. J. Immunol. 1966, 97, 822.
4528 Langmuir, Vol. 18, No. 11, 2002 Notes

and a thickness of 4 nm.25,26 If a 12 nm thickness of the

IgG layer is assumed, a shift maximum of 6.6 nm can be
estimated from Bragg’s law. However, it is impossible for
the thickness of the bound IgG layer to reach this level
due to limitations in the number of binding sites and steric
hindrance. So a shift of 4.1 nm for 5 mg/mL of IgG
concentration is reasonable.
Protonation of the binding sites on SpA by decreasing
the pH of the solution releases IgG from SpA.27 Thus,
replacing the solution in the pores with a solution
containing 0.1 M acetic acid (pH 2.78) resulted in an almost
instantaneous blue shift of the diffraction peak close to
a level corresponding to uncomplexed SpA. A subsequent
rinse with water returned the diffraction peak to the
original position measured before the introduction of IgG.
We probed a series of hIgG solutions at various
concentrations (Figure 3B). The shifts induced by hIgG
bound to immobilized SpA increased proportionally with
increasing concentrations of added hIgG from 0.01 to 2.5
mg/mL and then reached a plateau (Figure 3B). For
increasing concentrations from 2.5 to 10 mg/mL, only a
slight change was observed in the peak position. This was
because most of the hIgG binding sites of immobilized
SpA had been saturated by hIgG at the concentration of
2.5 mg/mL. Similar detection had been observed in the
measurements of the immobilized goat anti-hIgG (Figure
3B), but the shifts induced by goat anti-hIgG binding with
hIgG at the same concentrations were smaller than the
shifts induced by SpA-hIgG. This is because goat anti-
hIgG binds to the different regions of hIgG, and also the
binding capacities of SpA and goat anti-hIgG are different.
The possible orientations of hIgG molecules on the
immobilized SpA and goat anti-hIgG surfaces are shown
in the insets of Figure 3B. Our biosensor shows concen-
Figure 3. (A) In situ transmission spectra showing the tration-dependent binding from 0.01 to 2.5 mg/mL and a
immobilization of SpA, the blocking of BSA, and the binding
of hIgG with SpA immobilized on the pore surfaces. (B) sensitivity of 0.01 mg/mL for hIgG.
Correlation between the added hIgG concentrations and the In recent years, several optical sensor techniques have
shifts induced by formed hIgG-SpA complexes and hIgG-goat been developed for the direct monitoring of biomolecular
anti-hIgG complexes; insets show the possible orientations of recognition processes at the surface of a sensor. Among
hIgG molecules. them, surface plasmon resonance (SPR) methods have
made an important contribution to the quantification of
hIgG immobilized in the polystyrene microtiter plates can biomolecular interactions. The sensitivity for the mea-
be used to detect hIgG in an enzyme-linked immunosor- surement of analyte(s) in the optical biosensors is de-
bent assay. pendent on the molecular weight of the analyte. On the
Addition of a protein solution into the pores results in basis of the thickness of the protein corona, we can
a red shift of the diffraction peak. This shift is attributed quantitatively predict the correlation of the shift with the
to two factors. First, replacement of some of the aqueous thickness of binding pairs, as shown in the Supporting
phase with the protein solution will change the mean Information. Considering that our instrument resolution
refractive index of the film and be observed as a shift. The is 0.1 nm, a ∼0.1 nm thickness change of the binding IgG
second factor that contributes to the shift is ascribed to layer can be discriminated in our biosensor. Such a
the adsorption of protein on the pore surfaces. At this detection limit is comparable to that of conventional SPR.28
stage, the film is rinsed thoroughly with the same aqueous A larger concentration range of IgG from 0.01 to 2.5 mg/
phase to ensure that refractive index changes arise only mL could be detected using a simple diffractive optical
from the second factor. In our experiments, addition of a biosensor. Using this approach, we demonstrate proof of
2.0 mg/mL SpA solution with a subsequent water rinsing principle of a label-free optical biosensor to quantify
resulted in a shift of ∼2.1 nm (Figure 3A). Only ∼0.1 nm biomolecular interactions on a surface in a commercially
shift was observed in the following blocking process with available UV-visible spectrophotometer. This is very
BSA (Figure 3A) because of the dense distribution of SpA important for biological analyses.
molecules previously adsorbed on the polystyrene surface. Our biosensor was also found to completely eliminate
Once SPA was immobilized, binding of any additional disturbance due to nonspecific binding. The response of
biomolecules could lead to a shift of the diffraction peak (24) Palmer, D. A.; French, M. T.; Miller, J. N. Analyst 1994, 119,
that directly scales with analyte mass. Addition of a 5.0 2769.
mg/mL IgG solution resulted in a shift of 4.1 nm (Figure (25) Delamarche, E.; Sundarababu, G.; Biebuyck, H.; Michel, B.;
Gerber, C.; Sigrist, H.; Wolf, H.; Ringsdorf, H.; Xanthopoulos, N.;
3A). This detection is consistent with the formation of Mathieu, H. J. Laugmuir 1996, 12, 1997.
molecular packing of hIgG on the pore surface. IgG is a (26) Qian, W. P.; Yao, D. F.; Xu, B.; Yu, F.; Lu, Z. H.; Knoll, W. Chem.
protein composed of two heavy chains (MW 50 kDa) and Mater. 1999, 11, 1399.
two light chain (25 kDa) attached by disulfide bridges (27) Dancil, K. S.; Greiner, D. P.; Sailor, M. J. J. Am. Chem. Soc.
1999, 121, 7925.
that form a Y-like structure; their shape is usually (28) Homola, J.; Yee, S. S.; Gauglitz, G. Sens. Actuators, B 1999, 54,
compared to a disk with an average diameter of 12 nm 3.
Notes Langmuir, Vol. 18, No. 11, 2002 4529

Figure 4. Transmission spectra showing the comparison of the binding of immobilized SpA with whole hIgG molecules, Fc, F(ab′)2,
and Fab fragments. (A) IgG (3.2 nm red shift). (B) Fc fragments (1.1 nm red shift). (C) F(ab′)2 (no shift). (D) Fab (no shift).

the SpA-immobilized polystyrene films was tested with Transduction can be achieved by monitoring the diffraction
whole IgG molecules, Fc, F(ab′)2, and Fab fragments. IgG peak shift, which scales with the mass of analyte bound
is susceptible to proteolytic attack by enzymes such as on the pore surfaces. Consequently, our method is simple
pepsin and papain. Cleavage at the hinge region of IgG and general and can be used to detect a variety of
produces fragments known as Fc and either F(ab′)2 or biomolecular complexes, including oligonucleotides, an-
Fab depending upon the enzyme used. The Fc fragment tibody-antigen interactions, enzyme-substrate inter-
contains the binding domain recognized by SpA. The actions, and lectin-glycoprotein interactions.
F(ab′)2 or Fab fragments contain the antigen binding
regions and do not bind specifically to SpA. Additions of
1.0 mg/mL solutions of the whole IgG molecules and the Supporting Information Available: Correlation be-
Fc fragments resulted in red shifts of 3.2 and 1.1 nm, tween the thickness of the protein corona and the shift in the
respectively (Figure 4A,B). Upon introduction of solutions peak position estimated by Bragg’s law (Supporting Figure 1),
of 1.0 mg/mL of F(ab′)2 and Fab fragments, respectively, and figures showing the experimental cell (Supporting Figure 2)
no wavelength shifts were observed (Figure 4C,D). and the selection of the immobilization concentration of ligands
(Supporting Figure 3). This material is available free of charge
3DOM polystyrene substrate based biosensors measure via the Internet at http://pubs.acs.org.
the change in refractive index of the solution near the
pore surface that occurs during complex formation. LA0118199