www.elsevier.com/locate/apsusc

Applied Surface Science 253 (2007) 4226–4231

Investigation of a biofunctional polymeric coating deposited

onto silicon microcantilevers

P. Bergese a,*, E. Bontempi a, M. Chiari b, P. Colombi a, F. Damin b,L.E. Depero a, G. Oliviero a, G. Pirri b, M. Zucca a

a INSTM and Laboratono di Chimica per le Tecnologie, Universita di Brescia, via Branze, 38, 25123 Brescia, Italyb Istituto di Chimica del Riconoscimento Molecolare, CNR, via Mario Bianco, 9, 20131 Milano, Italy

Received 3 August 2006; received in revised form 14 September 2006; accepted 14 September 2006

Available online 1 November 2006

Abstract

The paper deals with an appealing route to activate silicon microcantilevers (90, 110 and 130 mm long, 35 mm wide and 2 mm thick) for specific

binding of biochemical species. The method consists in coating the underivatized microcantilevers with a biofunctional copolymer (based on N,N-

dimethylacrylamide bearing silanating moieties) that was developed for low-density microarray assays on microscope glass slides. Coating

deposition was obtained by dip-coating and its microstructure investigated by analyzing the resonance frequency values of bare and coated

microcantilevers, by SEM and SFM imaging, SFM tip-scratch tests and XRR experiments. Results indicate that the coating is 2.5 nm thick and has

a density of 1.22 g/cm3. The coating surface is nanostructured, displaying nanoblobs, which are from few up to 20 nm wide and, on average, 1.6 nm

high. The diameter of the biggest nanoblobs is of the same order of magnitude of the gyration radius of the copolymer chains, suggesting that

nanoblobs may identify individual macromolecules.

# 2006 Elsevier B.V. All rights reserved.

PACS : 68.37.-d; 87.80.-y

Keywords: Coating; Polymers; Microcantilevers; Biofunctionalization

1. Introduction

Application of microcantilevers (MCs) to biosensing

promises to be a breakthrough in biochemistry, life science

and medicine [1–3]. Several experiments have been success-

fully performed in this direction, revealing DNA hybridization

(including detection of single nucleotide polymorphisms,

SNPs) [4–7] and detecting proteins and antibodies [8–11],

single virus particles [12–14] or bacteria [15,16].

MC biosensor working principle is simple: the MC is

functionalized with a probe that can selectively bind the target

biochemical species. Adsorption and binding site interactions

of the targets change the mechanical response of the MC system

(because of the mass increase and/or the surface stress

generated by changes in Gibbs free energy), providing the

transduction/sensing mechanism.

* Corresponding author. Tel.: +39 0303715802; fax: +39 0303702448.

E-mail address: paolo.bergese@ing.unibs.it (P. Bergese).

0169-4332/$ – see front matter # 2006 Elsevier B.V. All rights reserved.

doi:10.1016/j.apsusc.2006.09.036

Probes immobilization can be obtained through direct

adsorption onto the MC surface [13,14,17] or through specific

functionalization methods. Indeed, accessibility, stability and

efficiency of surface functional groups are of crucial

importance and may constitute decisive feasibility drawbacks

for applications of MCs in biosensing experiments.

The use of thiolated molecules on gold coated MCs is the

most common functionalization method for DNA [4–7], protein

and antibodies [8–11]. Bacteria, viruses [18,19] and in some

cases proteins [20] have been immobilized by activation

techniques based on organosilanization. An alternative

approach to gold coating and organosilanization chemistry is

to coat the underivatized MCs with appropriate functional

polymers. This approach has found wide application in MC

chemical sensing [21–24]. On the contrary, to the best of our

knowledge, it was adopted only once in MC biosensing by

Gunter et al., who used a poly-ethylen-oxide (PEO) coating to

immobilize vaccinia polyclonal antibodies [25].

In the present paper the possibility of activating silicon MCs

by a co-polymeric thin coating based on N,N-dimethylacry-

P. Bergese et al. / Applied Surface Science 253 (2007) 4226–4231 4227

lamide bearing silanating moieties is assessed. This polymer

was developed for DNA and protein microarray assays on

microscope glass slides [26,27]. As showed in the previous

paper [26], the deposition of a thin layer of this polymer

represents a fast, inexpensive and robust method to covalently

bind amino-modified DNA probe strands on silicon dioxide

surfaces with high density. In addition, the film is stable in

aqueous buffers containing various additives, even at water

boiling temperature and, due to its hydrophilic nature and high

homogeneity, it minimizes non-specific interactions. This

feature is highly desirable for surface recognition bioassays,

for which interaction specificity is often an issue. Finally, the

macromolecular chains of the polymer act as spacer arms

keeping the probe molecules away from the surface more

effectively than self assembled monolayers (SAM) obtained by

organosilanization.

The deposition of a thin film of polymer onto the

underivatized MCs (90, 110 and 130 mm long, 35 mm wide

and 2 mm thick) was obtained by dip coating according to the

procedure previously devised for glass slides [26]. It is

worthwhile to notice that, because of the MC size, detailed

insight into the coating microstructure resulted a challenging

task. This goal was achieved by integrating the analysis of the

variation of the resonance frequency of the coated MCs with

respect to the bare MCs with the pieces of information obtained

from the images of scanning electron and scanning force

microscopies (SEM and SFM), SFM tip-scratch tests and from

X-ray reflectivity (XRR) data.

2. Experimental

The ter-polymer (DMA-co-NAS-co-MAPS) contains N,N-

dimethylacrylamide (DMA), N-acryloyloxysuccinimide (NAS)

and 3-(trimethoxysilyl)propyl methacrylate (MAPS), respec-

tively at 97%, 2% and 1% total monomer moles. The

procedures used to prepare and characterize the polymer are

reported in Ref. [26]. Hereafter it will be referred as (DMA-

NAS-MAPS).

We used commercial silicon MC arrays made of three MCs

(NSC12/Tipless/No Al by MikroMasch, Tallinn, Estonia). Each

array consists of three MCs of different lengths, namely 90, 110

and 130 mm, and of the same width and thickness: 35 and 2 mm,

respectively. Resonance frequency measurements were per-

formed in air, at room temperature using a scanning probe

microscope (SPM) head with an integrated laser lever readout

(Jeol, Tokyo, Japan). The Jeol SPM apparatus, equipped with

silicon tips by NT-MDT, Russia, was also used for SFM

experiments. Scanning electron microscopy (SEM) was

performed with a EVO 40 Microscope by LEO.

Polymer coating experiments were organized in two

separate batches. The coating procedure consisted of three

steps. Firstly the arrays were washed with acetone (Carlo Erba,

Milan, Italy) and after drying, they were immersed for 30 min

in a solution of the polymer (1% w/v in a water solution of

ammonium sulphate at 20% saturation level). Finally, they were

accurately washed with demineralized water and dried under

vacuum at 80 8C for 30 min.

The polymer deposition procedure previously assessed

[26,27] requires a pretreatment with HCl 1 M and NaOH

1 M. Such pretreatment is intended to activate the surface

silanols (a thin layer of native silicon dioxide is present on

the silicon surfaces of the MCs), maximizing the number of

hydrogen bonds with the polymer carbonyl groups and

ensuring the physisorption of a higher quantity of polymer.

However, the alkaline pretreatment resulted too aggressive

and changed the mechanical properties of MCs. This was

witnessed by an increase of the MC resonance frequency in

the order of tenth of kHz. In view of this, the MCs were just

carefully washed with acetone to eliminate superficial

contaminations.

In order to analyze the polymeric layer with X-ray

reflectivity (XRR), a silicon wafer slide (approximately

1.5 cm large and 2.5 cm long) was coated following the same

procedure and using the same polymer solution of the second

MCs batch. This allowed to obtain the best matching between

the MCs and the wafer coatings, as it was verified by SFM

analysis. XRR data were collected with a Bruker D8 Advance

reflectometer equipped with a Gobel mirror. The X-ray beam

(Cu Ka radiation, l = 0.154 nm) was collimated by proper slits

in order to set the cross section to 600 mm � 1.5 cm. All XRR

profiles were analyzed with the REFSIM software [28] in order

to extract information on the layer density, thickness and

roughness [29].

3. Results and discussion

The DMA-NAS-MAPS coating on the MCs was imaged by

SEM (Fig. 1) and quantitatively investigated by analyzing the

negative shift of the MC resonance frequency due to the mass of

the adsorbed DMA-NAS-MAPS layer (Fig. 2).

The SEM images of a DMA-NAS-MAPS coated MC

collected by secondary electrons and backscattered electrons

are reported in Fig. 1a and b, respectively. The secondary

electron image, which probes the surface morphology,

evidences that the polymeric layer is continuous, presenting

micron size corrugations localized far from the MC base.

These features disappear in the backscattered electron image.

Since backscattered electrons mainly probe the surface

chemical composition, this result indicates that the men-

tioned features are due to inhomogeneities of the polymeric

layer.

In order to evaluate the deposited mass of DMA-NAS-

MAPS, we used the general equation for the resonance

frequency ( f) of an oscillating cantilever in presence of fluid

damping and added mass [18,30]:

f ¼ 0:325

ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiK

mþ mf þ Dm;

r(1)

where K is the cantilever spring constant, m the cantilever mass,

mf the mass of the fluid film (mainly coming from ambient

moisture) adherent to the cantilever surface during oscillation

and Dm is the added mass. The constant 0.325 is a form factor

that accounts for the cantilever geometry. Eq. (1) is derived

P. Bergese et al. / Applied Surface Science 253 (2007) 4226–42314228

Fig. 1. (a) Secondary and (b) backscattered electrons SEM images of a DMA-

NAS-MAPS coated MC.

assuming that the added mass has no effect on the cantilever

mechanical properties. In general such assumption is valid

when the coating is much thinner than the cantilever or for

localized masses.

The DMA-NAS-MAPS coating on MCs resulted mainly

uniform, as showed by SEM imaging, and less than 5 nm thick

(see later in this section). Eq. (1) thus applies to the coated MCs

and can be simplified further. Actually both Dm and mf are

much smaller than m and can be neglected. Dm� m since the

coating is more than two orders of magnitude thinner than the

Fig. 2. Oscillation frequency profiles of a MC of the second batch before (solid

line) and after (dashed line) the coating with DMA-NAS-MAPS. The resonance

frequency shift, Df, between the two maxima is evidenced; according to Eq. (5),

it corresponds to a deposited mass per surface unit, Dms, of 6.7 fg/mm2.

MC, while the condition mf� m is ensured by the quality

factor of the frequency measurements (Fig. 2), which resulted

higher than 100 for all measurements. By these considerations,

Eq. (1) becomes:

f ¼ 0:325

ffiffiffiffiK

m

r: (2)

The equation that links the resonance frequency shift Df to

the mass variation (or added mass) Dm can be obtained by

differentiating Eq. (2):

Dm ¼ �2D f

fm ¼ �2

D f

frlwt; (3)

where l, w, t and r are the cantilever length, width, thickness,

and mass density, respectively. In the calculations we used the

nominal values of l and w, as they revealed at SEM check to be

reliable. On the contrary, the nominal value of t could not be

used as it resulted affected by a 30% uncertainty. Therefore t

was expressed through the Young modulus, E, using the rela-

tion:

t ¼ fl2

0:162

ffiffiffiffir

E

r; (4)

and taking into account that for the used silicon MCs

E = 169 GPa. In order to compare the coating of MCs of

different sizes it is convenient to use deposited masses per

surface unit (fg/mm2). Thus, by merging Eqs. (3) and (4), the

Fig. 3. Representative AFM image (non-contact mode) collected at the base of

a DMA-NAS-MAPS coated MC and height distribution curve. The vertical line

indicates the mean of the distribution.

P. Bergese et al. / Applied Surface Science 253 (2007) 4226–4231 4229

Fig. 4. SFM tip-scratch test (non-contact AFM image) at the base of a DMA-

NAS-MAPS coated MC. The reported line profile has been taken along the

white line.

Fig. 5. Representative AFM image (non-contact mode) collected on the DMA-

NAS-MAPS coated silicon wafer and height distribution curve. The vertical line

indicates the mean of the distribution.

operative equation we used in calculations reads:

Dms ¼ �12:35D f rl3w

S

ffiffiffiffir

E

r; (5)

where S = 2(lw + lt) + wt.

As described in Section 2, the experiments were organized in

two separate batches. The values of the deposited polymer mass

per surface unit, Dms, were calculated for each MC through

equation (5) and by using the experimental Df values (Fig. 2).

Finally, the mean value of Dms for each batch, Dm, was

obtained and the errors, intended as the standard deviations of

the means, calculated. Dm of the two batches resulted 10 � 2

and 7 � 1 fg/mm2, respectively. These values indicate fairly

good batch-to-batch reproducibility.

In order to gain insight into the microstructure of the DMA-

NAS-MAPS layer, scanning force microscopy (SFM) analysis

were performed on coated MCs. The coated MCs were scanned

in non-contact mode at their base, since it is a region in which

possible artifacts due to resonance interactions between the

observed MC and the SFM probe MC are prevented. A

representative SFM morphology image collected in non-

contact mode at the base of a coated MC and the related height

distribution curve are reported in Fig. 3. The coating surface

presents a nanostructure characterized by nanoblobs, which are

from few up to 20 nm wide with a mean height, according to the

height distribution curve, of 1.6 nm. The full width half

maximum (FWHM) of the height distribution curve is 1.5 nm.

It is worthwhile to note that the diameter of the biggest

nanoblobs is of the same order of magnitude of the gyration

radius of the DMA-NAS-MAPS polymer chains [31] suggest-

ing that each nanoblob may identify a single macromolecule.

We measured the thickness of the DMA-NAS-MAPS

coating by SFM tip-scratch tests [24,32,33]. The technique

consists in scratching the coating with the SFM tip. It is crucial

to set the normal force applied on the tip to a value high enough

to completely remove the polymeric layer but low enough to

avoid significant tip or silicon substrate damaging. After

scratching, the interested region is imaged in a non-contact

mode with the same tip used for scratching. Finally, the scratch

depth, i.e. the layer thickness, is evaluated by image analysis. A

representative scratch image of a scratched MC coating and a

related line profile (taken along the white line reported in the

image) are shown in Fig. 4. The line profile shows that the layer

thickness is about 2.5 nm. The surface area that was not

scratched is characterized by height oscillations of about 1 nm,

in agreement with the nanoblobs mean height evaluated by

SFM imaging. The spherical aggregates (hundreds of nm wide)

localized along the scratches are made of the polymer, which

was removed away from the scratches.

The average density of the DMA-NAS-MAPS coating was

evaluated by X-ray reflectivity (XRR) [29]. The experiment

was performed on a silicon wafer, which was coated by

P. Bergese et al. / Applied Surface Science 253 (2007) 4226–42314230

Fig. 6. SFM tip-scratch test (non-contact AFM image) collected on the DMA-

NAS-MAPS coated silicon wafer. The reported line profile has been taken along

the white line.

Table 1

XRR simulation parameters for the DMA-NAS-MAPS coated silicon wafer

Density

(g/cm3)

Thickness

(nm)

Roughness

(nm)

Wafer (Si) 2.32 1 <0.01

Layer 1 (SiO2) 2.98 1.52 0.12

Layer 2 (DMA-NAS-MAPS) 1.22 3.02 0.56

following the same procedure applied to MCs (see Experi-

mental section for details). The microstructural similarity

between MC and wafer coatings was verified through SFM by

repeating on the coated wafer the analysis performed on coated

MCs. The results, shown in Figs. 5 and 6, confirm a good

Fig. 7. Experimental and simulated XRR profiles of the DMA-NAS-MAPS

coated silicon wafer. The simulation parameters are reported in Table 1.

matching, and evidence some interesting fine differences. With

respect to MCs the DMA-NAS-MAPS forms on the wafer a

slightly thinner layer (about 2 nm) which presents a surface

characterized by less high nanoblobs (mean height of 1.3 nm)

with a narrower height distribution (FWHM = 0.6 nm), e.g.

lower surface roughness, intended as the FWHM of the height

distribution curve. These differences may be attributed to a

different adsorption efficiency of the surfaces due to the

differences of the native silicon oxide layers on the wafer and

MCs, respectively. Further investigation on the influence of the

silicon oxide layer on the DMA-NAS-MAPS adsorption is

ongoing.

The experimental and simulated XRR profiles and the

simulation parameters are reported in Fig. 7 and Table 1,

respectively. The polymeric layer resulted to be 3 nm thick,

with a surface roughness of 0.6 nm and a density of 1.22 g/cm3.

By keeping in mind that XRR overestimates the thickness of

layers of high roughness and that roughness values coincide

[29] it follows that XRR and SFM data are consistent.

The mass of adsorbed DMA-NAS-MAPS (on MCs) per

surface unit, Dms, may be exploited to assess the results

obtained from the different characterization techniques. Dms

can be calculated from the thickness t evaluated through SFM

tip-scratch tests and the density, rp, obtained by XRR,

according to the relation Dms = trp. It results Dms = 3 fg/

mm2, which is lower than the value obtained by analyzing the

MC resonance frequency, Dms ¼ 7� 1 fg=mm2. This differ-

ence was expected. Actually, Dms is related to uniform areas,

while Dms is an overall average that also accounts for the

coating inhomogeneities (Fig. 1). In view of this argument the

two values can be considered in fairly good agreement,

indicating that characterization based on MC resonance

frequency analysis is reliable for a coarse and quick (i.e.

routine) check of the coating.

Finally, by using Dms ¼ 7� 1 fg=mm2 and taking into

account that DMA-NAS-MAPS has an average molecular mass

of 1.5–105 amu [26] it follows that the coating layer (of the

second batch MCs) consists of (2.8 � 0.4) � 104 molecules/

mm2.

4. Conclusions

In this work we presented a method to activate silicon MCs

by a polymeric thin coating based on N,N-dimethylacrylamide

bearing silanating moieties [N,N-dimethylacrylamide (DMA),

N-acryloyloxysuccinimide (NAS) and 3-(trimethoxysilyl)pro-

pyl methacrylate (MAPS)]. The focus was placed on coating

characterization, which was achieved by analysis of the

P. Bergese et al. / Applied Surface Science 253 (2007) 4226–4231 4231

resonance frequency of the bare and coated MCs, SEM and

SFM imaging, SFM tip-scratch tests and XRR data. These

deeply different techniques provided complementary and

consistent results, which allowed to obtain a complete

microstructural picture of the coating.

MCs were coated with DMA-NAS-MAPS in two separate

batches by a dip-coating procedure, and a fairly good batch-to-

batch reproducibility was found. The coating is 2.5 nm thick

and has a density of 1.22 g/cm3. Its surface displays a

discontinuous nanostructure determined by nanoblobs, which

are from few up to 20 nm wide and, on average, 1.6 nm high.

From the average mass of adsorbed polymer per surface unit it

is possible to estimate that the coating consists of

(2.8 � 0.4) � 104 molecules/mm2. The diameter of the biggest

nanoblobs is of the same order of magnitude of the gyration

radius of the copolymer chains, suggesting that nanoblobs may

identify individual macromolecules.

This manuscript discusses the first issues encountered in

making innovative microcantilever-based (bio)sensors: suc-

cessful coating of microcantilevers with a high performance

functional layer and microstructural assessment of the coating,

which poses unique challenges due to the microscopic

dimensions of the microcantilevers. This latter aspect is often

neglected in the literature concerning microcantilever-based

sensors, even if it is mandatory for achieving interpretation of

sensing results, which goes beyond the straight ON/OFF

information.

Experiments to test the DMA-NAS-MAPS coated MCs as

biosensors are underway.

Acknowledgements

The authors acknowledge the funding of the Italian Ministry

of Education, University and Research through the project

PRIN 2005 ‘‘Nano-biosensors with micro-electro-mechanic

technology for genomic and proteomic analysis.’’ They also

wish to thank Alessandro Rivetti for SEM imaging.

References

[1] H.P. Lang, M. Hegner, C. Gerber, Nanotechnology 13 (2002) R29.

[2] R. Bashir, Adv. Drug Deliver. Rev. 56 (2004) 1565.

[3] K.K. Jain, Clin. Chim. Acta 358 (2005) 37.

[4] J. Fritz, M.K. Baller, H.P. Lang, H. Rothuizen, P. Vettiger, E. Meyer, H.J.

Guntherodt, C. Gerber, J.K. Gimzewski, Science 288 (2000) 316.

[5] R. Mukhopadhyay, M. Lorentzen, J. Kjems, F. Besenbacher, Langmuir 21

(2005) 8400.

[6] K.M. Hansen, H.F. Ji, G.H. Wu, R. Datar, R. Cote, A. Majumdar, T.

Thundat, Anal. Chem. 73 (2001) 1567.

[7] R. McKendry, J. Zhang, Y. Arntz, T. Strunz, M. Hegner, H.P. Lang, M.K.

Bailer, U. Certa, E. Meyer, H.J. Guntherodt, C. Gerber, Proc. Natl. Acad.

Sci. U.S.A. 99 (2002) 9783.

[8] G.H. Wu, R.H. Datar, K.M. Hansen, T. Thundat, R.J. Cote, A. Majumdar,

Nat. Biotech. 19 (2001) 856.

[9] P. Dutta, C.A. Tipple, N.V. Lavrik, P.G. Datskos, O. Hofstetter, M.J.

Sepaniak, Anal. Chem. 75 (2003) 2342.

[10] Y. Arntz, J.D. Seelig, H.P. Lang, J. Zhang, P. Hunziker, J.P. Ramseyer, E.

Meyer, M. Hegner, C. Gerber, Nanotechnology 14 (2003) 86.

[11] C. Grogan, R. Raiteri, G.M. O’Connor, T.J. Glynn, V. Cunningham, M.

Kane, M. Charlton, D. Leech, Biosens. Bioelectron. 17 (2002) 201.

[12] B. Ilic, Y. Yang, H.G. Craighead, Appl. Phys. Lett. 85 (2004) 2604.

[13] B. Ilic, D. Czaplewski, M. Zalalutdinov, H.G. Craighead, P. Neuzil, C.

Campagnolo, C. Batt, J. Vac. Sci. Technol. B 19 (2001) 2825.

[14] A. Gupta, D. Akin, R. Bashir, Appl. Phys. Lett. 84 (2004) 2004.

[15] K.Y. Gfeller, N. Nugaeva, M. Hegner, Biosens. Bioelectron. 21 (2005)

528.

[16] B.L. Weeks, J. Camarero, A. Noy, A.E. Miller, L. Stanker, J.J. Yoreo,

Scanning 25 (2003) 297.

[17] R. Marie, S. Schmid, A. Johansson, L.E. Ejsing, M. Nordstrom, D.

Hafliger, C.B.V. Christensen, A. Boisen, M. Dufva, Biosens. Bioelectron.

21 (2006) 1327.

[18] J. Zhang, H.F. Ji, Anal. Sci. 20 (2004) 585.

[19] N. Nugaeva, K.Y. Gfeller, Backmann, N. Lang, H.P. Duggelin, M. Hegner,

Biosens. Biolectron. 21 (2006) 849.

[20] G.A. Campbell, R. Mutharasan, Biosens. Bioelectron. 21 (2005) 597.

[21] F.M. Battiston, J.P. Ramseyer, H.P. Lang, M.K. Bailer, C. Gerber, J.K.

Gimzewski, E. Meyer, H.J. Guntherodt, Sens. Actuators B 77 (2001)

122.

[22] T.A. Betts, C.A. Tipple, M.J. Sepaniak, P.G. Datskos, Anal. Chim. Acta

422 (2000) 89.

[23] M. Maute, S. Raible, F.E. Prins, D.P. Kern, H. Ulmer, U. Weimar, W.

Gopel, Sens. Actuat. B 58 (1999) 505.

[24] G.G. Bumbu, G. Kircher, M. Wolkenhauer, R. Berger, J.S. Gutmann,

Macromol. Chem. Phys. 205 (2004) 1713.

[25] R.L. Gunter, W.G. Delinger, K. Manygoats, A. Kooser, T.L. Porter, Sens.

Actuators A: Phys. 107 (2003) 219.

[26] G. Pirri, F. Damin, M. Chiari, E. Bontempi, L.E. Depero, Anal. Chem. 76

(2004) 1352.

[27] M. Cretich, G. Pirri, F. Damin, I. Solinas, M. Chiari, Anal. Biochem. 332

(2004) 67.

[28] REFSIM Software, Bruker, 1998.

[29] E. Bontempi, Recent Research Development in Chemical Physics, Trans-

world Research Network, 2004, p. 461.

[30] A.N. Cleland, Foundations of Nanomechanics, Springer, 2003.

[31] E.A. Doherty, K.D. Berglund, B.A. Buchholz, I.V. Kourkine, T.M. Przy-

bycien, R.D. Tilton, A.E. Barron, Electrophoresis 23 (2002) 2766.

[32] C. Ton-That, A.G. Shard, R.H. Bradley, Langmuir 16 (2000) 2281.

[33] F. Anariba, S.H. DuVall, R.L. McCreery, Anal. Chem. 75 (2003) 3844.