Biosensors and Bioelectronics 23 (2007) 326–331
Controlling the surface density of DNA on gold by
electrically induced desorption
Kenji Arinaga a,b,∗ , Ulrich Rant b,∗∗ , Jelena Knežević b , Erika Pringsheim b ,
Marc Tornow b,c , Shozo Fujita a , Gerhard Abstreiter b , Naoki Yokoyama a
a Fujitsu Laboratories Ltd., 10-1 Morinosato-Wakamiya, Atsugi 243-0197, Japan
Walter Schottky Institut, Technische Universität München, 85748 Garching, Germany
c Institut für Halbleitertechnik, Technische Universität Braunschweig, 38106 Braunschweig, Germany
b
Received 8 January 2007; received in revised form 25 March 2007; accepted 24 April 2007
Available online 29 April 2007
Abstract
We report on a method to control the packing density of sulfur-bound oligonucleotide layers on metal electrodes by electrical means. In a first
step, a dense nucleic acid layer is deposited by self-assembly from solution; in a second step, defined fractions of DNA molecules are released
from the surface by applying a series of negative voltage cycles. Systematic investigations of the influence of the applied electrode potentials
and oligonucleotide length allow us to identify a sharp desorption onset at −0.65 V versus Ag/AgCl, which is independent of the DNA length.
Moreover, our results clearly show the pronounced influence of competitive adsorbents in solution on the desorption behavior, which can prevent the
re-adsorption of released DNA molecules, thereby enhancing the desorption efficiency. The method is fully bio-compatible and can be employed
to improve the functionality of DNA layers. This is demonstrated in hybridization experiments revealing almost perfect yields for electrically
“diluted” DNA layers. The proposed control method is extremely beneficial to the field of DNA-based sensors.
© 2007 Elsevier B.V. All rights reserved.
Keywords: Self-assembled monolayer; DNA; Mercaptohexanol; Desorption; Surface density; Biosensing
1. Introduction
Self-assembled monolayers of nucleic acids on solid substrates have become increasingly important over the last decade
because of their applications in DNA-based sensors (“DNAchips”) and microarrays. (Epstein et al., 2002; Heller, 2002;
Drummond et al., 2003; Nakamura et al., 2003; Tarlov and Steel,
2003; Bang et al., 2005; Spadavecchia et al., 2005). Conductive
materials are particularly interesting substrates since they offer
the possibility to apply electric fields to the immobilized DNA
layers, which, for instance, allows us to conduct electrochemical experiments (Steel et al., 1998; Fan et al., 2003), direct the
adsorption of charged biomolecules, such as target nucleic acid
∗
Corresponding author. Tel.: +81 46 250 8234; fax: +81 46 250 8844.
Corresponding author. Tel.: +49 89 289 12776; fax: +49 89 320 6620.
E-mail addresses: [email protected] (K. Arinaga), [email protected]
(U. Rant).
URLs: http://www.labs.fujitsu.com/ (K. Arinaga), http://www.wsi.tum.de/
(U. Rant), http://www.iht.tu-bs.de/ (M. Tornow).
∗∗
0956-5663/$ – see front matter © 2007 Elsevier B.V. All rights reserved.
doi:10.1016/j.bios.2007.04.012
sequences in solution (Heaton et al., 2001) or realize electrically switchable DNA layers (Rant et al., 2004b, 2006). On gold
substrates, DNA layers are conventionally prepared via selfassembly from solution and tethered to the surface at one end
by a thiol linker (S–Au bond) (Ulman, 1996; Tarlov and Steel,
2003; Love et al., 2005). In a second step, a sub-layer of spacer
molecules (very often mercaptohexanol, MCH) is co-adsorbed
to the surface in order to render a well defined layer structure and
improve the layer stability (Herne and Tarlov, 1997; Arinaga et
al., 2006).
Recently, it has been recognized that the molecular packing
density within the DNA layer crucially determines the functionality of the nucleic acids. For instance, the ability of surface
immobilized probe strands to capture complementary target
sequences from solution is largely suppressed if the layer density is too high (Peterson et al., 2001, 2002). Experiments with
switchable DNA layers gave similar results: the ability to manipulate the molecular orientation of nucleic acids by electric fields
sensitively depends on the spacing of the strands on the surface
(Rant et al., 2004b). In order to obtain DNA layers with opti-
K. Arinaga et al. / Biosensors and Bioelectronics 23 (2007) 326–331
mal packing densities, protocols have been devised which aim at
controlling the density during the self-assembly process, e.g., by
varying the valence or concentration of salt in the buffer solution
(Herne and Tarlov, 1997; Petrovykh et al., 2003), or by varying
the concentration of oligonucleotide in the buffer solution (Steel
et al., 2000). However, these methods merely allow tuning the
surface coverage within a limited range; moreover, the obtained
results are often troublesome when precise reproducibility of the
DNA density is required.
Here we present an approach which relies on the electrically
induced desorption of DNA. Starting from relatively densely
packed monolayers, the surface density is gradually reduced by
releasing nucleic acids from the surface by means of applying
negative potentials to the substrate. We show that this method
allows coarse- and fine-tuning the packing density of nucleic
acids on gold surfaces in situ. The “diluted” layers retain full
biological functionality, which is demonstrated by hybridization
experiments yielding efficiencies of approximately 100%.
The basic principle of reducing the density of oligonucleotides on metal surfaces relies on electrostatic interactions
and electrochemical reduction. Since DNA is highly negatively
charged in solutions of pH > 1, it is repelled from the surface
when negative potentials are applied to the substrate (Kelley et
al., 1998; Grubb et al., 2006; Rant et al., 2006). At the same time,
the sulfur–gold bond, which tethers the DNA to the surface, can
be broken by electrochemical reduction through the application
of negative potentials (Zhong and Porter, 1997; Yang et al., 1997;
Kawaguchi et al., 2000). The electrically induced desorption of
thiolated oligonucleotide layers has first been shown by Wang
et al. (1999) who, however, reported the complete removal of
DNA layers after applying −1.3 V (versus Ag/AgCl) to the supporting gold substrates. Prior desorption studies in our group
addressed the influence of electric screening by the electrolyte
solution (Rant et al., 2003) and linker properties (Takeishi et
al., 2004). Here, we systematically investigate the dependence
of the applied potentials and the oligonucleotide length on the
desorption efficiency and identify electrochemical potentials for
which the DNA surface density can efficiently be reduced without harming the remaining DNA molecules.
2. Materials and methods
2.1. DNA and gold substrate
All chemicals were purchased from general suppliers and
used without further purification. The probe DNA was obtained
from IBA GmbH in Goettingen, Germany, and the sequence of
the 24, 48, 72 and 96mer single stranded (ss) oligonucleotides
were 5 -HS-(CH2 )6 -TAG TCG GAA GCA TCG AAG GCT
GAT-Cy3-3 and 5 -HS-(CH2 )6 -TAG TCG TAA GCT GAT ATG
GCT GAT TAG TCG GAA GCA TCG AAC GCT GAT-Cy33 , 5 -HS-(CH2 )6 -TAG TCG TGA GCA CAT GGA CCT GAT
TAG TCG TAA GCT GAT ATG GCT GAT TAG TCG GAA
GCA TCG AAC GCT GAT-Cy3-3 and 5 -HS-(CH2 )6 -TAG
TCG GAA GCA TCG AAC GCT GAT TAG TCG TGA GCA
CAT GGA CCT GAT TAG TCG TAA GCT GAT ATG GCT
GAT TAG TCG GAA GCA TCG AAC GCT GAT-Cy3-3 . The
327
3 end was labelled with a cyanine dye, Cy3TM (fluorescence
detection), whereas the 5 end was derivatized with a thiol
linker to tether the DNA to Au surfaces. The complementary DNA (cDNA) was used for the preparation of double
stranded DNA and the hybridization efficiency measurements.
Au-electrodes of 2.0 mm diameter were prepared on 3 inch
single crystalline sapphire wafers, by subsequently depositing
Ti(10 nm)/Pt(40 nm)/Au(200 nm) using standard optical lithography and metallization techniques. The average roughness of
the prepared Au surfaces was measured by AFM and found to
be less than 1 nm, that is, insignificant compared to the oligonucleotide length. The substrates were cleaned in piranha solution
(H2 SO4 :H2 O2 (30%) = 7:3) for 15 min (note that piranha solution must be handled with care: it is extremely oxidizing, reacts
violently with organics, and should only be stored in loosely
tightened containers to avoid pressure build up) and prior to
DNA adsorption exposed to HNO3 (60%) for 15 min, followed
by a final rinse with deionized (DI) water.
2.2. Protocol for DNA adsorption and hybridization
Single stranded DNA (ssDNA) was immobilized onto gold
by exposing the surfaces to buffered aqueous DNA solution
([DNA] = 1 ␮M, [Tris] = 10 mM, pH 7.3, [NaCl] = 200 mM) for
1 h. Double stranded DNA (dsDNA) was prepared by hybridizing ssDNA with cDNA in buffer solution ([DNA] = 1 ␮M,
[Tris] = 10 mM, pH 7.3, [NaCl] = 200 mM) for 1 h prior to immobilization. After the adsorption process, the electrodes were
thoroughly rinsed with buffer solution ([Tris] = 10 mM, pH 7.3,
[NaCl] = 50 mM]).
Following the DNA adsorption, the modified Au surfaces
were exposed to mercaptohexanol (MCH, [MCH] = 1 mM, 1 h),
which leads to formation of a mixed DNA/MCH layer. Here,
MCH is used as a spacer molecule which specifically binds to
Au by its sulfur group, thereby removing and replacing loosely
bound nucleic acids, and passivating the surface in-between
DNA molecules physically and electrically (Herne and Tarlov,
1997; Georgiadis et al., 2000; Peterson et al., 2002; Ha et al.,
2004; Arinaga et al., 2006). In addition, the use of MCH allowed
quantifying the surface density of unlabelled DNA by electrochemical means (Steel et al., 1998).
For the hybridization efficiency measurements, ssDNA layers
were hybridized with cDNA in buffer solution ([cDNA] = 1 ␮M,
[Tris] = 10 mM, pH 7.3, [NaCl] = 50 mM]) for 1 h. Afterwards,
the electrodes were thoroughly rinsed with buffer solution
([Tris] = 10 mM, pH 7.3, [NaCl] = 50 mM]).
2.3. Quantification of DNA surface density
The DNA coverage was quantified using electrochemical
methods introduced by Steel et al. (1998). In brief, the DNA
layer is exposed to electrolyte solution of low ionic strength
([Tris] = 10 mM) containing a multivalent redox cation, hexaammineruthenium(III) chloride (RuHex) ([RuHex] = 100 ␮M).
Under these conditions, the DNA compensates the negative
charge of its anionic phosphate groups by electrostatically trap-
328
K. Arinaga et al. / Biosensors and Bioelectronics 23 (2007) 326–331
ping RuHex to its backbone, thereby confining a number of
redox markers to the Au surface that is proportional to the number of DNA molecules in the immobilized monolayer. Upon
application of a potential step from +0.1 V (which is positive
enough to oxidize all RuHex markers on the electrode surface)
to −0.4 V versus Ag/AgCl at the Au electrode, RuHex markers are reduced. The resulting (reductive) current is measured
and the number of RuHex markers is calculated. Accounting for
the number of nucleotides per DNA strand, the number of DNA
molecules on the surface is obtained.
Note that a prerequisite of the described measurement is that
no redox marker adsorption occurs at the surface between DNA
strands, since it would contribute parasitically to the determined
surface excess charge. Therefore, the use of passivation layer
is obligatory and for that reason a MCH adsorption step was
carried out routinely prior to electrochemical measurements.
2.4. Apparatus
Subsequent to preparation, the samples were installed in
an uncapped cell filled with measurement buffer solution
([Tris] = 10 mM, pH 7.3, [NaCl] = 50 mM), continuously purged
with Argon gas, which allowed for optical as well as electrochemical measurements. A potentiostat (Autolab PGSTAT30,
Eco Chemie, The Netherlands) was utilized to monitor and control the voltage of the Au-work-electrodes with respect to a
Ag/AgCl reference electrode, using a Pt-wire counter electrode.
Fluorescence measurements of the immobilized Cy3-labelled
DNA were conducted by positioning an optical fiber mount
above the electrode (Rant et al., 2003). Green light from an Ar+
laser (λ = 514 nm) is guided onto the electrode surface at an angle
of ∼45◦ , whereas fluorescence from Cy3-dyes is collected by a
second fiber oriented normal to the surface plane. Note that the
region of fluorescence detection included not only the electrode
surface but also the electrolyte volume above, defined by the
intersection of the excitation and detection beams. Light from
the detection fiber was coupled into a monochromator (set to
the Cy3-peak-emission wavelength, 565 nm) and detected with
a cooled photomultiplier or an avalanche photo diode operating in single-photon-counting mode. Reference measurements
of unmodified Au surfaces were used for background correction.
3. Results and discussion
3.1. Electrically induced DNA desorption monitored by
optical means
In the following we present a representative desorption experiment in Fig. 1 and describe the optical method used to observe
the electrically induced release of nucleic acids from gold surfaces.
In order to detect the nucleic acids on and above the surface
in situ and in real-time, the DNA molecules are labeled with a
fluorescent marker (Cy3TM ) and the laser induced fluorescence
is monitored. As long as the Cy3-DNA molecules are immobilized on the surface, the fluorescence emission (Fbefore ) is
quenched due to efficient energy transfer which occurs from the
Fig. 1. Representative desorption measurement. The upper panel shows the
applied electrode potential, while the simultaneously recorded Cy3-DNA (48
base pairs) fluorescence intensity is depicted in the lower panel. The transient
fluorescence peak stems from DNA molecules released from the surface. MCH
(1 mM) was present in solution during the measurement. The background signal
is negligible (approximately 50 au).
optically excited dye-label to surface plasmons in the metal surface (Chance et al., 1978; Barnes, 1998). Once the molecules are
released from the surface by applying a negative voltage step, the
observed fluorescence increases. This increase can be attributed
to DNA molecules floating in solution within the optical detection volume: because the energy transfer, which suppresses the
fluorescence emission of Cy3-DNA molecules on the surface, is
short ranged, it virtually does not affect molecules which are further than roughly 100 nm away from the metal surface (Chance
et al., 1978; Barnes, 1998). Thus, strong fluorescence emission is observed from Cy3-DNA molecules floating in solution.
Eventually, the released nucleic acids diffuse out of the detection volume, which causes the fluorescence signal to decrease
again (Rant et al., 2003). Since the diffusion process is slow,
however, the relaxation of the measured fluorescence intensity
takes several ten seconds after the potential has been turned off.
The reduced fluorescence intensity measured after the transient
desorption peak (Fafter ) stems from the residual DNA layer.
In order to prevent re-adsorption of released DNA onto the
gold surface (Yang et al., 1997), the solution contained 1 mM
MCH. MCH has been shown to rapidly adsorb to gold surfaces
and is expected to backfill vacant sites of bare gold immediately;
the importance of this measure will be elucidated in detail later.
In addition, the surface potential before and after the application
of desorption potentials was kept slightly negative with respect
to the potential of zero charge (pzc) (Silva et al., 1990; Kelley
et al., 1998; Rant et al., 2006) in order to electrostatically repel
DNA molecules from the surface.
The electrically induced release of DNA from a gold surface as depicted in Fig. 1 can be understood by the arguments
described in the introduction: under the influence of nega-
K. Arinaga et al. / Biosensors and Bioelectronics 23 (2007) 326–331
Fig. 2. Dependence of the DNA desorption efficiency on the electrode potential, measured for double stranded oligonucleotides of varying length. MCH
(1 mM) was present in solution during all measurements depicted in solid symbols whereas data depicted as open circles were measured in pure buffer solution.
The voltage pulse duration was 30 s. Lines are guides to the eye.
tive substrate potentials, the sulfur-gold bond which tethers
the nucleic acids to the surface is broken by electrochemical
reduction. This process is facilitated by electrostatic repulsion
between the surface and the DNA, which will be further elaborated in the following sections.
3.2. Desorption efficiency
In order to utilize the electrical desorption for the preparation
of DNA layers with defined packing densities, it is necessary to
control the number of strands released during a desorption-cycle.
We characterized the desorption process concerning the following questions: (i) How does the desorption efficiency depend
on the magnitude of the applied potentials? (ii) What is the
influence of the oligonucleotide length on the desorption efficiency? (iii) How does the presence of a competitive adsorbent
in solution interfere with the re-adsorption of desorbed DNA?
To answer these questions, we applied a series of negative potential steps to the electrodes ranging from −0.3 to −1.0 V (versus
Ag/AgCl reference), probed the desorption behavior of four double stranded oligonucleotides of varying length (24, 48, 72, and
96 base pairs (bp)), and tested the influence of the presence of
MCH in solution. The initial DNA surface densities were smaller
than 1 × 1012 molecules cm−2 .
Fig. 2 shows data recorded from consecutive desorptioncycles which have been performed as depicted in Fig. 1.
The X-axis values denote the “desorption potentials”, which
were applied for 30 s each. In case of the “48 bp w/o MCH”
measurement, the electrode potential was cycled between 0 V
(instead of −0.2 V) and the desorption potential, but this is
not expected to be of importance to the presented discussion,
because of the negligible fluorescence difference between 0 and
−0.2 V as can be seen in Fig. 1. The desorption efficiency
ηDE = (Fbefore − Fafter )/Fbefore was determined from the measured fluorescence intensities before (Fbefore ) and after (Fafter )
the desorption-cycle. ηDE = 1 corresponds to the release of all
DNA molecules from the surface. The evaluation of ηDE from
329
the measured fluorescence intensities is based on the assumption
that the fluorescence emitted by the DNA layer is proportional
to the number of molecules on the surface. This assumption
is expected to hold as long as self-quenching effects among
neighboring dyes are not significant, and seems justified for
the DNA densities studied in this work (in prior studies we
found no indications for self-quenching for densities below app.
5 × 1012 molecules cm−2 ) (Rant et al., 2004a). Moreover, orientation effects must be taken into account: as the fluorescence
emission depends on the DNA orientation relative to the metal
(due to the distance-dependent energy transfer) (Rant et al.,
2004b), Fbefore and Fafter were measured at slightly negative
electrode potentials (versus pzc) to ensure that the molecules
were standing up right on the surface.
In the following, we discuss the desorption behavior of 24,
48, 72 and 96 bp oligonucleotides in the presence of MCH in
solution (solid symbols in Fig. 2) and will turn to the MCH-free
measurement later. For all investigated samples, we find a desorption threshold between −0.6 and −0.7 V (versus Ag/AgCl
reference) (Fig. 2). This value is significantly more positive
than the reduction potentials usually reported for sulfur-bound
monolayers on gold (Zhong and Porter, 1997; Yang et al., 1997;
Kawaguchi et al., 2000). We attribute this early onset to the electrostatic repulsion of the negatively charged oligonucleotides
from the surface, which decreases the binding energy of the
molecules on the surface. Thus, the S–Au bond is more likely
to be broken, i.e., electrochemical reduction occurs at positively
shifted electrode potentials.
Noticeably, the desorption behavior of all investigated
oligonucleotides is independent of their length. We propose
that this is a consequence of short-ranged electric interactions.
According to the Gouy-Chapman-Stern (GCS) theory, the electric field emanating from a charged surface into solution decays
rapidly within typically a few nanometers (Bard et al., 1993). In
an earlier publication (Rant et al., 2006), we plotted the evolution of the electric field over the characteristic length scale of a
surface tethered oligonucleotide (cf., a 48 bp oligomer is 16 nm
long) showing that merely a small number of base pairs which are
closest to the surface are exposed to high field strengths for the
used salt condition ([Tris] = 10 mM, pH 7.3, [NaCl] = 50 mM]).
Charged sites on the DNA’s backbone which are further away
than a few nm from the surface are not repelled strongly. For
that reason, the effective electrostatic repulsion is virtually independent of the oligonucleotide length, as long as the molecules
contain roughly 20 bp or more. Note that different salt conditions (concentration and valence) in solution affect the screening
effect. These influences have been described in previous publications (Rant et al., 2003, 2006).
Measurements performed in solutions which did not contain
MCH (cf. open circles in Fig. 2) show significantly reduced
desorption efficiencies. Even after applying a series of strongly
negative potentials, a considerable fraction of DNA molecules
(approximately 50% of the initial coverage) is detected on the
surface when using a MCH-free buffer.
We believe that this behavior reflects the role of MCH as a
competitive adsorbent: regardless of whether the buffer contains MCH or not, DNA molecules may readily be released
330
K. Arinaga et al. / Biosensors and Bioelectronics 23 (2007) 326–331
by the application of negative potentials. However, if dissolved
MCH molecules are present, they can spontaneously adsorb to
exposed parts of the gold surface, since MCH is not charged and
thus, in contrast to DNA, are not repelled from the negatively
charged surface. After the desorption potential is switched off,
re-adsorption of DNA molecules from solution to vacant sites
(holes in the DNA/MCH layer) on the gold surface may occur
(Yang et al., 1997); however, if MCH is present in solution,
it acts as a competitive adsorbent and backfills empty sites in
the layer spontaneously. As a consequence, the gold surface
becomes protected and re-adsorption of DNA is prevented.
Finally we note that the onset of desorption is remarkably
steep. The desorption efficiency is negligible at −0.6 V, but
almost all molecules are released when applying −0.8 V. In order
to release defined fractions of molecules, the applied potentials must be chosen carefully within this regime. One might
expect that the desorption efficiency can be controlled conveniently by adjusting the duration of the applied voltage pulse,
however, the observations made so far indicate that the pulse
width is merely of secondary importance compared to the magnitude of the applied potential. The data obtained up to now
do not suggest a linear dependence of the amount of released
DNA on the pulse duration but point to a more complex desorption behavior which is going to be the subject of future
investigations.
3.3. Adjusting the DNA surface density by electrically
induced desorption
In this section we apply the technique introduced above to
demonstrate that the density of nucleic acid layers can efficiently
be tuned by electrical desorption. Two methods have been used
to evaluate the DNA surface density before and after performing
desorption cycles: fluorescence measurements and electrochemical quantification (see Section 2.3). First, the initially prepared
DNA surface density was measured by electrochemical quantification. Second, the fluorescence intensity from DNA was
monitored during the electrical desorption process (see Fig. 1).
Then, the reduced DNA surface density was measured by
electrochemical quantification. The desorption potentials were
applied to the electrodes for 5 min. The results are depicted in
Fig. 3, which shows the fluorescence intensity versus DNA surface density as determined by the electrochemical method for
24 bp DNA layers.
For potentials more positive than −0.5 V (Nos. 1 and 2 in
Fig. 3), almost no change can be seen in the fluorescence before
and after the electrical desorption process, which is confirmed by
the electrochemically determined DNA density. In this potential
regime, the surface tethered DNA was stable and did not desorb from gold. After applying −0.8 V (No. 3 in Fig. 3), the
fluorescence decreased by a factor of 15 and the electrochemically determined DNA density decreased from 1.0 × 1012 to
1.6 × 1011 molecules cm−2 . After applying −0.8 V to the same
electrode again, the fluorescence decreased by a factor of 3 and
the density decreased to 4 × 1010 molecules cm−2 .
These findings are in good qualitative agreement with the
results of the desorption efficiency measurement in Section 3.2.
Fig. 3. Fluorescence intensity and electrochemically quantified surface density
of 24 base pair DNA layers. The diluted DNA layers () were obtained from
initially prepared layers (䊉) by electrical desorption. MCH (1 mM) was present
in solution during the electrical desorption process. The potentials applied to the
individual electrodes were −0.4 V for No. 1, −0.5 V for No. 2, −0.8 V for No.
3 and −0.9 V for No. 4. The voltage pulse duration was 5 min.
The DNA surface density can be reduced by the proposed desorption technique. This process can be monitored in real-time
by fluorescence measurements and can be applied repeatedly.
3.4. Hybridization efficiency of DNA layers prepared by
electrical desorption
In order to make sure whether DNA layers which have been
“diluted” by the electrical desorption technique retain their biological function, we tested the ability of 48mer single stranded
“probe” DNA layers to bind complementary “target” sequences
in solution. Here, we determined the hybridization efficiency
by electrochemical methods as used before, measuring the
nucleotide density on the surface before and after hybridization (see Section 2.2 for the detailed hybridization protocol). In
case of 100% hybridization efficiency one expects to find twice
as many nucleotides on the surface as for single stranded layers.
Fig. 4 depicts the hybridization efficiency of 48mer targets to
single stranded probe layers of varying surface densities. While
very low hybridization efficiencies (<20%) were observed for
surface densities higher than 2 × 1012 molecules cm−2 , complete hybridization was determined within the experimental
errors for probe layer densities <7 × 1011 molecules cm−2 . The
results are reasonable considering that target strands cannot penetrate densely packed probe layers and are in good agreement
with other reports in literature (Peterson et al., 2001, 2002; Yu
et al., 2004) when taking into account the DNA length and salt
concentration (Tsuruoka et al., 1996). The high hybridization
efficiencies clearly show that the electrical desorption procedure
does not affect the recognition capabilities of single stranded
probe DNA layers. Of course, values greater than 100% for
the determined hybridization efficiencies seem unrealistic. Most
likely, they originate from relatively large errors which occur
during the quantification of extremely low DNA densities where
the detection limit of the electrochemical measurement method
is nearly reached. As the employed desorption potentials are
K. Arinaga et al. / Biosensors and Bioelectronics 23 (2007) 326–331
331
References
Fig. 4. Hybridization efficiency of complementary 48mer targets and single
stranded probe DNA layers as a function of the probe surface density. The densities of the probe layers were adjusted by the electrical desorption technique in
buffer solution containing 1 mM MCH. Error bars are evaluated from statistical
analysis of 16 measured samples. The solid line is a guide to the eye.
moderate, the electrochemical reduction of nucleic acids can
be excluded and the oligonucleotides maintain their full biofunctionality.
4. Conclusions
In conclusion, we have introduced a new protocol to adjust
the density of sulfur-bound nucleic acids on gold surfaces by
electrically induced desorption. We observe a sharp desorption
onset at electrode potentials of approximately −0.65 V versus a Ag/AgCl reference. The desorption behavior is found
to be independent of the oligonucleotide length, which can
be understood in terms of the short ranged repulsive electric
fields. Furthermore, our results indicate a strongly enhanced
desorption efficiency if a competitive adsorbent (MCH) is
present in solution; possibly because MCH prevents the readsorption of released DNA onto the surface. Single stranded
probe DNA layers, diluted by the introduced method, exhibit
virtually 100% hybridization efficiency to complementary target
sequences, demonstrating that the desorption cycles are electrochemically harmless and that the layers maintain their full
bio-functionality. This method is extremely beneficial to the
field of DNA-based sensors because it provides reliable means
to prepare layers of defined packing densities and thus optimized
functionality.
Acknowledgements
We are grateful to Y. Yamaguchi and S. Hirose for preparing
the Au substrates. We acknowledge financial support by Fujitsu
Laboratories of Europe (FLE) and by the Deutsche Forschungsgemeinschaft via SFB 563.
Arinaga, K., Rant, U., Tornow, M., Fujita, S., Abstreiter, G., Yokoyama, N.,
2006. Langmuir 22, 5560–5562.
Bang, G.S., Cho, S., Kim, B.G., 2005. Biosens. Bioelectron. 21, 863–870.
Bard, A.J., Abruna, H.D., Chidsey, C.E., Faulkner, L.R., Feldberg, S.W., Itaya,
K., Majda, D., Murray, R.W., Porter, M.D., Soriaga, M.P., White, H.S., 1993.
J. Phys. Chem. 97, 7147–7173.
Barnes, W.L., 1998. J. Mod. Opt. 45, 661–699.
Chance, R.R., Prock, A., Silbey, R., 1978. Adv. Chem. Phys. 37, 1–65.
Drummond, T.G., Hill, M.G., Barton, J.K., 2003. Nat. Biotechnol. 21,
1192–1199.
Epstein, J.R., Biran, I., Walt, D.R., 2002. Anal. Chim. Acta 469, 3–36.
Fan, C., Plaxco, K.W., Heeger, A.J., 2003. PNAS 100, 9134–9137.
Georgiadis, R., Peterlinz, K.P., Peterson, A.W., 2000. J. Am. Chem. Soc. 122,
3166–3173.
Ha, T.H., Kim, S., Lim, G., Kim, K., 2004. Biosens. Bioelectron. 20, 378–389.
Heaton, R.J., Peterson, A.W., Georgiadis, R.M., 2001. PNAS 98, 3701–3704.
Heller, M.J., 2002. Annu. Rev. Biomed. Eng. 4, 129–153.
Herne, T.M., Tarlov, M.J., 1997. J. Am. Chem. Soc. 119, 8916–8920.
Kawaguchi, T., Yasuda, H., Shimazu, K., 2000. Langmuir 16, 9830–9840.
Kelley, S.O., Barton, J.K., Jackson, N.M., McPherson, L.D., Potter, A.B., Spain,
E.M., Allen, M.J., Hill, M.G., 1998. Langmuir 14, 6781–6784.
Love, J.C., Estroff, L.A., Kriebel, J.K., Nuzzo, R.G., Whitesides, G.M., 2005.
Chem. Rev. 105, 1103–1169.
Grubb, M., Wackerbarth, H., Ulstrup, J., 2006. J. Am. Chem. Soc. 128,
7734–7735.
Nakamura, F., Ito, E., Sakao, Y., Ueno, N., Gatuna, I.N., Ohuchi, F.S., Hara, M.,
2003. Nano Lett. 3, 1083–1086.
Peterson, A.W., Heaton, R.J., Georgiadis, R.M., 2001. Nucl. Acids Res. 29,
5163–5168.
Peterson, A.W., Wolf, L.K., Georgiadis, R.M., 2002. J. Am. Chem. Soc. 124,
14601–14607.
Petrovykh, D.Y., Kimura-Suda, H., Whitman, L.J., Tarlov, M.J., 2003. J. Am.
Chem. Soc. 125, 5219–5226.
Rant, U., Arinaga, K., Fujiwara, T., Fujita, S., Tornow, M., Yokoyama, N.,
Abstreiter, G., 2003. Biophys. J. 85, 3858–3864.
Rant, U., Arinaga, K., Fujita, S., Yokoyama, N., Abstreiter, G., Tornow, M.,
2004a. Langmuir 20, 10086–10092.
Rant, U., Arinaga, K., Fujita, S., Yokoyama, N., Abstreiter, G., Tornow, M.,
2004b. Nano Lett. 4, 2441–2445.
Rant, U., Arinaga, K., Fujita, S., Yokoyama, N., Abstreiter, G., Tornow, M.,
2006. Org. Biomol. Chem. 4, 3448–3455.
Silva, F., Sottomayor, M.J., Hamelin, A., 1990. J. Electroanal. Chem. 294,
239–251.
Spadavecchia, J., Manera, M.G., Quaranta, F., Siciliano, P., Rella, R., 2005.
Biosens. Bioelectron. 21, 894–900.
Steel, A.B., Herne, T.M., Tarlov, M.J., 1998. Anal. Chem. 70, 4670–4677.
Steel, A.B., Levicky, R.L., Herne, T.M., Tarlov, M.J., 2000. Biophys. J. 79,
975–981.
Takeishi, S., Rant, U., Fujiwara, T., Buchholz, K., Usuki, T., Arinaga, K., Takemoto, K., Yamaguchi, Y., Tornow, M., Fujita, S., Abstreiter, G., Yokoyama,
N., 2004. J. Chem. Phys. 120, 5501–5504.
Tarlov, M.J., Steel, A.B., 2003. DNA-based sensors. In: Rusling, J.F. (Ed.),
Biomolecular Films, vol. 111. Marcel Dekker Inc., New York, pp. 545–608.
Tsuruoka, M., Yano, K., Ikebukuro, K., Nakayama, H., Masuda, Y., Karube, I.,
1996. J. Biotechnol. 48, 201–208.
Ulman, A., 1996. Chem. Rev. 96, 1533–1554.
Wang, J., Rivas, G., Jiang, M., Zhang, X., 1999. Langmuir 15, 6541–6545.
Yang, D.F., Wilde, C.P., Morin, M., 1997. Langmuir 13, 243–249.
Yu, F., Yao, D., Knoll, W., 2004. Nucl. Acids Res. 32, e75.
Zhong, C.J., Porter, M.D., 1997. J. Electroanal. Chem. 425, 147–153.