Self-assembled monolayers on oxidized metals. 3. Alkylthiol and dialkyl disulfide assembly on gold under electrochemical conditions

Article abstract of DOI:10.1021/ja9823722

Alkylthiol monolayers were assembled in ethanol solutions onto gold surfaces held at positive potentials. The developing monolayer introduces a barrier to electron transfer; hence, measurement of the current corresponding to ethanol oxidation at the applied potential provides a convenient means for real-time monitoring of the self-assembly (SA) process and its completion. Monolayers produced by the new method are formed considerably faster than similar monolayers prepared by the common procedure (no applied voltage). Two other processes which occur under the same applied potential include gold surface oxidation and oxidative desorption of the monolayer, both related to the presence of small amounts of water in the ethanol solution. It is shown that the interplay between the combined processes allows considerable control over the SA process before, during, and after monolayer formation, such as the possibility to perform multiple adsorption-desorption cycles for wettability and surface control. An important consequence of understanding the mechanism of alkylthiol SA onto oxidized gold concerns alkylthiol vs dialkyl disulfide adsorption. While the two types of molecules produce similar monolayers on reduced gold surfaces, a totally different result is obtained with oxidized gold, namely, alkylthiols form compact monolayers whereas dialkyl disulfides do not. This, together with the possibility to determine the extent of gold surface preoxidation, opens the way to rational design of mixed monolayers.

Full text of DOI:10.1021/ja9823722

13444
J. Am. Chem. Soc. 1998, 120, 13444-13452
Self-Assembled Monolayers on Oxidized Metals. 3. Alkylthiol and
Dialkyl Disulfide Assembly on Gold under Electrochemical
Conditions
Hannoch Ron and Israel Rubinstein*
Contribution from the Department of Materials and Interfaces, The Weizmann Institute of Science,
RehoVot 76100, Israel
ReceiVed July 6, 1998
Abstract: Alkylthiol monolayers were assembled in ethanol solutions onto gold surfaces held at positive
potentials. The developing monolayer introduces a barrier to electron transfer; hence, measurement of the
current corresponding to ethanol oxidation at the applied potential provides a convenient means for real-time
monitoring of the self-assembly (SA) process and its completion. Monolayers produced by the new method
are formed considerably faster than similar monolayers prepared by the common procedure (no applied voltage).
Two other processes which occur under the same applied potential include gold surface oxidation and oxidative
desorption of the monolayer, both related to the presence of small amounts of water in the ethanol solution.
It is shown that the interplay between the combined processes allows considerable control over the SA process
before, during, and after monolayer formation, such as the possibility to perform multiple adsorption-desorption
cycles for wettability and surface control. An important consequence of understanding the mechanism of
alkylthiol SA onto oxidized gold concerns alkylthiol vs dialkyl disulfide adsorption. While the two types of
molecules produce similar monolayers on reduced gold surfaces, a totally different result is obtained with
oxidized gold, namely, alkylthiols form compact monolayers whereas dialkyl disulfides do not. This, together
with the possibility to determine the extent of gold surface preoxidation, opens the way to rational design of
mixed monolayers.
Introduction
taneous adsorption). Moreover, the common procedure offers
limited means for manipulating the system at various stages of
SA while monitoring the process in real-time.
The past decade has seen unusual interest in systems based
on organized organic thin films, as part of a prominent trend in
science and technology toward miniaturization and improved
organization. Self-assembled monolayers on solid supports have
played a major role in this field, providing a combination of
molecular dimensions, superior organization, stability, and
In our attempt to control the interface and the SA process
we have concentrated on oxidative pretreatments of Au,
primarily using oxygen plasma or UV/ozone treatment prior to
monolayer SA. We have shown that such Au preoxidation,
followed by oxide removal in ethanol (i.e., oxide reduction by
ethanol), provides highly reproducible Au surfaces for mono-
1
,2
versatility. Monolayers comprising long-chain hydrocarbons
bound to a gold surface via a sulfur-containing headgroup are
perhaps the most widely studied systems of this kind, where
the properties of the gold substrate (electrical conductivity,
chemical inertness) may play an important role in determining
the overall properties and performance of the monolayer
systems.
4
layer SA. In an earlier study we have shown that, if the
preoxidation step is not followed by Au oxide reduction,
alkylthiol SA on the oxidized Au results in densely packed
monolayers where a layer of Au oxide is encapsulated under
5
the monolayer.
An issue nearly untouched, but of substantial importance, is
the electric field at the interface before, during, or after SA,
either at open circuit (OC, i.e., with no externally applied
It is rather self-evident that the nature of the Au-monolayer
interface is a major factor in the SA process and the properties
of the resulting monolayers, hence effective control of various
interfacial properties at all stages of the SA process is highly
desirable. To this end, factors such as the Au surface topography
and roughness, as well as surface pretreatment, have been
6
voltage) or under a certain applied potential. The interfacial
electric field may influence such elements as molecular orienta-
tion, binding of headgroups, and adsorption properties. More-
over, one may envision electrode reactions which serve to
influence or monitor SA processes. Hence, the gold substrate
may be used as an electrode during the SA process, allowing
rational manipulation of basic interfacial variables. Note that
the common SA procedure implies self-assembly at OC, namely,
3
studied. However, a rational approach to controlling the SA
process is still missing, which affects the reproducibility of
monolayer systems obtained by the common procedure (spon-
*
Corresponding author. Phone: +972-8-9342678. Fax: +972-8-
344137. E-mail: cprubin@weizmann.weizmann.ac.il.
1) Finklea, H. O. in Electroanlytical Chemistry; Bard, A. J., Rubinstein,
I., Eds.; Marcel Dekker: New York, 1996; Vol. 19.
2) Ulman, A. Ultrathin Organic Films; Academic Press: New York,
991.
9
(
(4) Ron, H.; Matlis, S.; Rubinstein, I., Langmuir 1998, 14, 1116.
(5) Ron, H.; Rubinstein, I. Langmuir 1994, 10, 4566.
(6) Porter and co-workers demonstrated alkylthiol adsorption in ethanol/
KOH under an applied potential (positive of the alkylthiol reductive
desorption). See: Weisshaar, D. E.; Lamp, B. D.; Porter, M. D. J. Am.
Chem. Soc. 1992, 114, 5860.
(
1
(
3) Guo, L. H.; Facci, J. S.; McLendon, G.; Mosher, R. Langmuir 1994,
1
0, 4588.
1
0.1021/ja9823722 CCC: $15.00 © 1998 American Chemical Society
Published on Web 12/10/1998

Self-Assembled Monolayers on Oxidized Metals
J. Am. Chem. Soc., Vol. 120, No. 51, 1998 13445
9
under rather arbitrary conditions insofar as the interfacial electric
potential is concerned. For example, we have shown that various
surface pretreatment protocols reported for SA of monolayers
on Au result in exceedingly different OC potentials upon
immersion in the adsorption solution; hence, the OC potential
of oxidized Au surfaces (in pH ) 7.0 phosphate buffer as well
as in ethanol) was found to be ∼0.6 V (vs KCl-saturated calomel
reference electrode, SCE), whereas the reduced Au shows a
the latter effecting monolayer desorption. This allows consider-
able control of the system, namely monolayer adsorption,
desorption, partial adsorption, mixed adsorption, repeated
adsorption-desorption, etc. Dynamic control over the wettability
of a gold surface can be achieved by alternating adsorption/
desorption cycles.
3. While alkylthiols and dialkyl disulfides produce similar
1
1
monolayers on reduced Au surfaces, entirely different results
are realized with oxidized Au. This furnishes a new route for
the preparation, on a single substrate, of mixed monolayers with
controlled distributions of the two components, by using
controlled surface oxidation.
5
substantially more negative OC potential.
The stability of n-alkylthiol monolayers on Au electrodes
subjected to various applied potentials was addressed in several
7
studies. Porter and co-workers demonstrated cathodic desorp-
tion of alkylthiol monolayers in KOH solution. Groat and
8
Creager reported that dodecylthiol (C12SH) monolayers are
Experimental Section
stable in aqueous solutions between -0.5 and 0.3 V (vs Ag/
AgCl) and in propylene carbonate between -0.7 and 0.5 V.
The monolayers were damaged upon applying potentials
between 0.5 and 0.8 V but were reconstructed upon adding
C12SH to the solution (propylene carbonate) while performing
a continuous scan between 0.0 V to 0.8 V. Everett and Fritsch-
Gold Substrates. {111} textured gold substrates were prepared by
resistive evaporation of 100 nm gold (99.99%) onto cleaned glass
microscope cover slides, followed by 3 h annealing at 250 °C in air.
Immediately before use the gold substrates were subjected to UV-
ozone treatment followed by 20 min immersion in pure ethanol, a
procedure shown previously to provide clean, reproducible gold
surfaces.4
9
Faules showed that C12SH monolayers are stable between -1.0
Materials. Octadecylthiol (C18SH) (Sigma, AR) was purified by
crystallization from ethanol. Chloroform (Biolab, AR), bicyclohexyl
and 0.5 V (vs Ag/AgCl) both in methylene chloride and in
acetonitrile. They also demonstrated that in the presence of
increasing amounts of water the stability of alkylthiol mono-
layers at positive potentials is decreased. On the basis of the
work of Wagner and Gerischer10 it was suggested that under
positive potentials of 0.5 V the thiol is exchanged with Au oxide.
When a sufficiently positive potential is applied to a gold
electrode immersed in ethanol containing supporting electrolyte
(BCH) (Aldrich, AR), and hexadecane (HD) (Sigma, AR) were passed
through a column of activated basic alumina (Alumina B, Akt. 1, ICN).
Ethanol (Merck, AR) and LiClO
Water was triply distilled.
4
(Merck, AR) were used as received.
Dioctadecyl Disulfide (C18S)
by oxidation of C18SH with iodine as follows: 2.0 g octadecylthiol
C18SH) were dissolved in 150 mL n-hexane. 0.5 g I was dissolved
2 2
Synthesis. (C18S) was synthesized
12
(
2
(
e.g., LiClO4) and a certain amount of water, the following
in 5 mL methanol. The two solutions were gradually added to a flask
with stirring. After 24 h stirring the liquid was left in the flask for
another hour without stirring for efficient phase separation. The hexane
phase was transferred to a separating funnel where it was washed several
times with copious amounts of methanol to remove remaining C18SH.
The product was recrystallyzed from methanol. Melting point: 61-62
processes take place:
Electrochemical Au oxidation upon reaction with traces of water
in the ethanol
-
+
2
Au + 3H O - 6e f Au O + 6H
(1)
2
2
3
°
C.
Monolayer Self-Assembly. Monolayer preparation was carried out
Direct electrochemical oxidation of ethanol
as detailed in the text. After adsorption the samples were rinsed three
times with ethanol and chloroform and dried under a stream of purified
air.
-
+
CH CH OH - 2e f CH CHO + 2H
(2)
3
2
3
Electrochemical Instrumentation. The electrochemical system
comprised a potentiostat (model 303C, Department of Chemistry,
Technion, Haifa, Israel) controlled by a Zenith 486 computer through
a A/D-D/A converter. The program that provided waveform generation,
potentiostatic control and data acquisition was written by the Laboratory
Computers Unit, The Weizmann Institute of Science.
Fast Liquid Exchange ElectroChemical (FLEEC) Cell. A special
cell was designed that allows fast solution exchange under applied
electrical potential, without interruption of the potential or the current
measurement (Figure 1). The FLEEC cell is a three-electrode electro-
chemical cell, constructed as a vertically positioned, two-part cylindrical
tube. The two parts are connected through a Teflon valve, with another
Teflon valve at the bottom end of the tube. The lower cylindrical
compartment is the electrochemical cell, whereas the upper cylindrical
compartment serves as a solution reservoir. In a typical experiment a
gold electrode is placed as a working electrode in the electrochemical
cell containing a counter electrode (Pt wire) and a KCl-saturated calomel
(SCE) reference electrode (in a separate compartment, filled with
Chemical oxidation of ethanol by gold oxide
Au O + 3CH CH OH f 2Au(0) + 3CH CHO + 3H O
2
3
3
2
3
2
(3)
The net result of these reactions is a sustained electrical
current corresponding to anodic oxidation of the solvent
(ethanol), either direct or mediated by Au oxide. These
conditions form the basis for the present work, which concerns
the following issues:
1. SA of n-alkylthiols is demonstrated under potentials where
ethanol is electrochemically oxidized, providing convenient real-
time monitoring of the SA process and its completion by
measurement of the current, assumed to follow the change in
monolayer coverage. Highly oriented alkylthiol monolayers
prepared under these conditions are formed considerably faster
than similar monolayers prepared by the common procedure
4
ethanol + 0.1 M LiClO ). The electrochemical cell is filled with a
(
i.e., no applied potential).
. Under these conditions a competition exists between
alkylthiol SA and Au oxide formation (due to residual water),
solution to the top valve level, which is then closed, and the reservoir
tube is filled with a different solution. A potential is then applied to
the working electrode to initiate the experiment. At a certain instance
the solution in the electrochemical cell is rapidly replaced (within a
2
(
7) (a) Widrig, C. A.; Chung, C.; Porter, M. D. J. Electroanal. Chem.
991, 310, 335. (b) Weisshaar, D. E.; Walczak, M. M.; Porter, M. D.
Langmuir 1993, 9, 323.
1
(11) (a) Nuzzo, R. G.; Fusco, F. A.; Allara, D. L. J. Am. Chem. Soc.
1987, 109, 2358. (b) Bain, C. D.; Biebuyck, H. A.; Whitesides, G. M.
Langmuir 1989, 5, 723.
(12) Capozzi, G.; Modena, G. in The Chemistry of the Thiol Group; Patai,
S., Ed.; Wiley: New York, 1974; pp 794-795.
(
(
(
8) Groat, K. A.; Creager, S. E. Langmuir 1993, 9, 3668.
9) Everett, W. R.; Fritsch-Fauls, I. Anal. Chim. Acta 1995, 307, 253.
10) Wagner, D.; Gerischer, H. J. Electroanal. Chem. 1989, 258, 127.

13446 J. Am. Chem. Soc., Vol. 120, No. 51, 1998
Ron and Rubinstein
Figure 1. Schematic drawing of the Fast Liquid Exchange Electro-
Chemical (FLEEC) cell. (1) Electrochemical cell, (2) solution reservoir,
(
(
3) outlet tube, (4) working electrode, (5) working electrode holder,
6) reference electrode, (7) salt bridge, (8) counter electrode, (9)
magnetic stirrer, (10) stirring bar, (11) and (12) Teflon valves.
Figure 2. Current vs time curves during anodization of a Au electrode
2
(area: 1.7 cm ) in 99.8% ethanol containing 10 mM LiClO
4
at an
fraction of a second) with the solution in the reservoir tube by opening
and closing of the two valves for a short period of time. This allows
fast and efficient solution exchange while applying a desired potential
to the working electrode and stirring the liquids in the cell, without
interrupting the voltage or the current.
applied potential of 1.45 V. (a) The first 10 min, (b) the first 10 s.
Inset in (a): Linear potential scan in 0.1 M H SO for the stripping of
2
4
Au oxide produced by applying 1.45 V (vs SCE) in 99.8% ethanol +
10 mM LiClO (see text), followed by solution exchange to 0.1 M
4
H SO (scan rate: 0.1 V/sec).
2
4
The efficiency of solution exchange in the FLEEC cell was tested
as follows: The concentration of toluene in ethanol was calibrated using
Results and Discussion
1
3
toluene absorbance between 230 and 280 nm. Pure ethanol in the
electrochemical cell was exchanged as described above with 8.3 mM
toluene in ethanol from the reservoir tube. After solution exchange the
concentration of toluene in the electrochemical cell was compared with
its concentration in the mother solution. The procedure was repeated
six times, and the results showed an average decrease of 2.4% in the
toluene concentration compared to the mother solution, which can be
taken as the approximate error in the solution-exchange procedure.
Ellipsometry. Ellipsometric measurements were carried out using
a Rudolph Research Auto EL-IV ellipsometer with a monochromated
tungsten-halogen light source, at an angle of incidence Φ ) 70° and
a wavelength λ ) 632.8 nm. The same three points were measured on
the bare (cleaned) and the monolayer-covered slide. Monolayer
thicknesses were calculated from the change in the ellipsometric ∆ and
Ψ measured before and after monolayer adsorption, using a film
Anodization of Gold in Ethanol. Figure 2 shows a typical
current-time curve (at two time scales) for a gold electrode in
99.8% ethanol + 10 mM LiClO solution, following a potential
4
14
step to 1.45 V. The three processes which take place under
these conditions are given in eqs 1-3 above. The applied
potential effects net oxidation of ethanol (eqs 2 and 3), which
results in a sustained current. The slow current decay is due to
eq 1, resulting from the relatively high concentration of water
in the ethanol (0.2% water in ethanol, ca. 0.1 M). Oxidation of
the Au blocks the surface toward direct ethanol oxidation; the
blocking is slowed due to chemical reduction of Au oxide by
ethanol (eq 3). Still, after prolonged anodization at 1.45 V (ca.
4 h) it is possible to block almost all ethanol oxidation with Au
oxide.
refractive index of n
Contact-Angle (CA) Measurements. Contact angles (advancing and
receding) of H O, BCH, and HD were measured shortly (<10 min)
f
) 1.5, k
f
) 0.
To relate the slow blocking of the electrode (Figure 2a) with
Au oxide coverage, the following experiment was carried out:
We have previously shown that Au oxide is not reduced when
2
after removal of the slides from the adsorption solution. Three
measurements at different spots were carried out with each solvent,
using a Rame-Hart NRL model 100 contact angle goniometer.
Reflection-Absorption Fourier Transform Infrared (RA-FTIR)
Spectroscopy. Grazing incidence FTIR measurements were carried out
using a Bruker IFS 66 spectrometer at an angle of incidence of 80°
5
immersed in aqueous H2SO4. If after prolonged anodization
the electrical circuit is opened and the ethanolic solution in the
(14) The electrochemical measurements reported here were carried out
without iR compensation in order to demonstrate the simplest mode of
operation. This implies, however, that the “true” applied potential is initially
lower than 1.45 V and it increases and approaches the set value as the
current decreases. We verified (by using a range of potentials) that this has
no effect on the system behavior or the conclusions drawn.
-
1
and a resolution of 2 cm
.
(13) Perkampus, H. H. UV-Vis Atlas of Organic Compounds; VCH: New
York, 1985.

Self-Assembled Monolayers on Oxidized Metals
J. Am. Chem. Soc., Vol. 120, No. 51, 1998 13447
Figure 4. Current vs time curves for Au electrodes polarized at 1.45
V applied potential: (a) SA in 3 mM C18SH in 99.8% ethanol + 10
Figure 3. (a) Cyclic voltammogram for the electrochemical oxidation-
4
mM LiClO . Inset: The initial stages, shown in more detail; the initial
4
reduction of a Au electrode in 99.8% ethanol + 10 mM LiClO (scan
2
current increase indicates application of the potential, and the solution
exchange (C18SH introduction) is marked. (b) An electrode treated as
rate: 0.1 V/sec, electrode area: 1.7 cm ). (b) Same as (a), at more
sensitive current and voltage scales.
in (a), polarized in 10 mM LiClO
Inset: After Au oxide reduction at 0.0 V; polarized in 99.8% ethanol
10 mM LiClO . I is the absolute value of the initial current in 4a
or 4b, inset).
4
in 95% ethanol (5% water added).
FLEEC cell is exchanged with aqueous 0.1 M H2SO4, then the
Au oxide formed can be stripped electrochemically by negatively
+
(
4
0
5
scanning the potential from 1.3 V (Figure 2a, inset), showing
the typical reduction of an almost complete monolayer of Au
oxide.15 If at that point the circuit is disconnected, the solution
exchanged back to 99.8% ethanol + 10 mM LiClO4, and a
voltage of 1.45 V applied to the Au electrode, then the original
initial current is restored. This strongly supports the assumption
that the slow current decay is caused by gradual coverage of
the surface by Au oxide.
surfaces treated as in Figure 2 would be covered with a certain
amount of oxide.
Figure 3b shows, in addition, that Au oxide is completely
reduced in ethanol under these conditions when potentials
negative of 0.3 V are applied. It may therefore be assumed that
an OC potential negative of 0.3 V indicates an oxide-free Au
surface (for monolayer-covered Au, this applies to the Au
fraction in contact with the solution).
Further understanding of the anodization of Au in ethanolic
solution containing traces of water is provided by following
the voltammetry of an Au electrode in this solution (Figure 3a).
SA of Alkylthiols under an Applied Positive Potential.
Figure 4a presents a current-time curve for an Au electrode
immersed in 99.8% ethanol + 10 mM LiClO4 (in the FLEEC
cell), where at t ) 0 the potential is stepped from 0.0 to 1.45 V
followed by solution exchange (without voltage or current
interruption) with an identical solution containing 3 mM
octadecylthiol (C18SH). Unlike the slow current decrease in
Figure 2, resulting from Au surface oxidation, a sharp decrease
in the current is immediately observed when the solution is
exchanged, as shown in Figure 4a. This fast current decay is
the result of alkylthiol adsorption, causing effective blocking
of the Au surface by the compact monolayer formed. It is
evident from Figure 4a (inset) that almost all of the Au surface
is coVered within 1-2 s. Since blocking of the surface by Au
oxide formation in the first 10 s is marginal (Figure 2b), the
fast current decay in Figure 4a can essentially be entirely
attributed to C18SH monolayer formation. Hence, SA in ethanol
5
Unlike the voltammogram obtained in aqueous 0.1 M H2SO4,
the positive scan in ethanol shows a continuous current increase
rather than an oxidation wave. Furthermore, the reduction peak
in the negative scan, typical of Au cycling in aqueous solutions,
is barely observed in Figure 3a. Nearly all of the current in the
positive scan is attributed to ethanol oxidation at the Au
electrode. Still, when carefully examined using a more sensitive
current scale (Figure 3b), a small reduction peak is observed
when scanning the potential in the negative direction, attributed
to Au oxide reduction. In Figure 3a the potential was scanned
at 0.1 V/sec, meaning that the Au surface is under oxidative
potential for ca. 10 s; Figure 3b shows that under these
conditions a small amount of Au oxide is already formed. Thus,
it may be assumed that, on the time scale of seconds, Au
(15) Vitus, C. M.; Davenport, A. J. J. Electrochem. Soc. 1994, 141, 1291.

13448 J. Am. Chem. Soc., Vol. 120, No. 51, 1998
Ron and Rubinstein
the alkyl chains with the gold surface due to the electrode charge
1
9
and the local electric field.
. The electric field at the interface may have an additional
3
role, i.e., promoting a perpendicular orientation of the S-C
dipole in the desired orientation for binding to the Au. This
element may be particularly significant at advanced stages of
monolayer formation, where the rate may be controlled by
penetration of alkylthiol molecules to fill small pinholes in a
2
0
largely covered surface.
. Another factor which may improve the monolayer SA is
the continuous “pretreatment” of the Au surface (oxidation/
4
15
reduction) during the SA process. Moreover, as shown below,
anodization in ethanolic solution containing water induces
removal of the alkylthiol monolayer from the Au surface. It is
therefore reasonable to assume that the SA process under these
conditions is accompanied by removal of less ordered monolayer
fractions, allowing only tightly packed monolayer fractions to
remain on the surface.
Figure 5. Grazing incidence reflection-absorption FTIR spectrum of
a C18SH monolayer adsorbed at 1.45 V applied potential as in Figure
The relative contribution of each of these factors is yet to be
determined.
4
a.
The Effect of Alkylthiol Concentration. Figure 6 demon-
strates the effect of C18SH concentration on the adsorption
characteristics (current-time behavior) at 1.45 V applied
potential in 99.8% ethanol + 10 mM LiClO4. The overall
process is presented in Figure 6a, and the initial stages of
adsorption are emphasized in Figure 6b. The current-time
behavior is qualitatively similar for all thiol concentrations
except 0.01 mM, indicating formation of blocking monolayers.
The very slow current decay at long times (Figure 6a) may be
due to the blocking of the remaining pinholes by Au oxide;
this is supported by the OC potential of ca. 0.6 V measured
after long adsorption at 1.45 V, indicating the existence of a
minute fraction of oxidized Au surface not covered with the
under positive applied potentials provides a straightforward
means for real-time monitoring of the process.
Monolayers prepared by the new procedure were character-
ized by ellipsometry, CA measurements, and reflection-
absorption FTIR spectroscopy. They were compared with
C18SH monolayers obtained by the common procedure, i.e.,
adsorption from ethanol (with or without 10 mM LiClO4; the
results are similar) with no applied potential. The FTIR spectra
of monolayers obtained by SA at 1.45 V (as exemplified in
Figure 5) are characteristic of high-quality C18SH monolayers
1
6
on {111} Au. As shown in Table 1, rows 1, 2, 3, and 4, the
monolayer obtained by 18.5 min adsorption at 1.45 V is
indistinguishable (within experimental error) in thickness and
CAs from the one obtained by 24 h regular adsorption, but
distinctly superior to the monolayer obtained by 20 min regular
adsorption. It is therefore clear that alkylthiol adsorption at
positiVe potentials produces high-quality monolayers much
faster (by ca. 2 orders of magnitude) than regular adsorption
5
monolayer and in contact with the solution. The current-time
curve for 0.01 mM C18SH shows that, at the low thiol
concentration, oxide formation overcomes monolayer SA; the
bump around 60-70 s indicates that monolayer removal (see
below) becomes faster than monolayer SA and further current
decrease is dominated by Au oxide formation, as in Figure 2.
As seen in Table 1, rows 4, 5, and 6, similar high-quality
C18SH monolayers are obtained at 1.45 V in the concentration
range between 1 mM and 5 mM, whereas a certain decrease in
quality is seen at 0.2 mM (Table 2, row 2). As shown in Table
1, row 7, at low C18SH concentrations the interference arising
from Au oxidation, effected by residual water, inhibits mono-
layer formation; the ellipsometry and CA results suggest
marginal monolayer coverage with largely scattered results at
different points on the surface.
(i.e., no applied potential).
Several factors may be responsible for the substantial
improvement in the rate of formation of high-quality alkylthiol
monolayer on Au under oxidative electrochemical conditions:
1
. Lowering of the activation energy for the binding process.
It was previously suggested that alkylthiol SA involves forma-
1
1b,17
tion of a Au-S bond,
which in the present case would be
6
strongly favored by the electrochemical oxidation of the gold
+
-
Monolayer Desorption. Figure 4b shows a current-time
curve obtained for an electrode covered with a compact C18SH
monolayer, prepared as in Figure 4a, after exchanging the
adsorption solution to 95% ethanol + 10 mM LiClO4 (5% water,
no alkylthiol) and then applying 1.45 V to the electrode. The
current reaches a maximum of ca. 15% of the initial value
observed with bare Au, followed by a relatively slow current
decay. The solution in the cell was then replaced with 99.8%
ethanol + 10 mM LiClO4 and a potential of 0.0 V applied to
the electrode until the OC potential was negative of 0.3 V (i.e.,
complete removal of any Au oxide). The potential was then
stepped to 1.45 V and the current-time curve in Figure
Au f Au + e
(4)
(5)
+
+
Au + R-S-H f Au-S-R + H
2. During regular adsorption a certain fraction of the alkylthiol
molecules is expected to be in a planar orientation, induced by
the interaction of the alkyl chains with the high-energy gold
surface.18 This would create a kinetic barrier for the formation
of a perpendicularly oriented structure and may introduce
defects. SA at positive potentials minimizes the interaction of
(
16) Laibinis, P. E.; Whitesides, G. M.; Allara, D. L.; Tao, Y.-T.; Parikh,
A. N.; Nuzzo, R. G. J. Am. Chem. Soc. 1991, 113, 7152.
17) (a) Whitesides, G. M.; Laibinis, P. E. Langmuir 1990, 6, 87. (b)
Reference 2, p 288.
18) (a) Bain, C. D.; Troughton, E. B.; Tao, Y.-T.; Evall, J.; Whitesides,
(
(19) Neutral organic molecules adsorb on metal electrodes in a limited
potential range around the potential of zero charge (PZC); see: Gileadi, E.
Electrode kinetics for Chemists, Chemical Engineers and Materials
Scientists; VCH: New York, 1993; pp 257-259, 309-318.
(
G. M.; Nuzzo, R. G. J. Am. Chem. Soc. 1989, 111, 321. (b) Sun, L.; Crooks,
R. M. J. Electrochem. Soc. 1991, 138, L23.
(20) Sabatani, E.; Rubinstein, I. J. Phys. Chem. 1987, 91, 6663.

Self-Assembled Monolayers on Oxidized Metals
J. Am. Chem. Soc., Vol. 120, No. 51, 1998 13449
Table 1. Characterization of C18SH Monolayers on Au Prepared by the Common Procedure or at 1.45 V Applied Potential
contact angle (deg)
H
2
O
BCH
HD
ads. time
(min)
ellips.
thickness (Å)
sample description
1 mM C18SH
adv.
107
rec.
105
adv.
52
rec.
51
adv.
46
rec.
45
1
2
1440
18.9
(
24h)
1 mM C18SH
salt
1440
(24h)
20
18.1
107
105
52
51
47
45
+
3
4
1 mM C18SH
1 mM C18SH
15.1
21.5
102
107
90
104
49
52
47
50
44
45
41
44
18.5
+
salt + pot.
3 mM C18SH
salt + pot.
5 mM C18SH
salt + pot.
0.01 mM C18SH
5
6
7
8
9
7
19.6
19.5
5-13
2.7
107
106
104
103
52
50
45
43
+
6
52
51
45
43
+
18.5
90-104
60-90
30-32
20
∼0-29
30-35
12
∼0-18
+
no. 5
after desorption
1 mM C18SH
salt + pot.
90
60
97
<10
<10
18.5
26.1
109
50
47
44
41
+
salt + pot.,
rd
3
2
adsorption after
adsorption-desorption cycles
monolayer from the Au surface. The ellipsometric and CA
results shown in Table 1, row 8, substantiate this conclusion.
Note that these results are in general agreement with those of
9
Everett and Fritsch-Faules, who demonstrated removal of
alkylthiol monolayers at positive potentials in organic solutions
containing traces of water.
The current-time curves in Figure 4b are instructive in
understanding the process of alkylthiol desorption at positive
potentials in ethanolic solution containing water. The initial
current increase can be attributed to alkylthiol desorption leaving
exposed Au surface, where ethanol can be oxidized. Simulta-
neous (but slower) Au surface oxidation, which inhibits ethanol
oxidation, is responsible for the current maximum and subse-
quent decrease. This suggestion is supported by (i) the observa-
tion that the same qualitative behavior as in Figure 4b is
observed in 99.8% ethanol + 10 mM LiClO4, but on a longer
time scale; and (ii) the initial current in Figure 4b (inset)
following oxide reduction at 0.0 V, which corresponds to bare
Au obtained by complete monolayer desorption and oxide
removal.
The above explanation suggests that the alkylthiol monolayer
desorption occurs in two manners: (i) oxidative thiol desorption,
where the thiol is removed and bare Au surface is exposed,
allowing ethanol oxidation (relatively fast); and (ii) thiol
exchange by Au oxide, induced by dissolved water (relatively
slow). The rather low current maximum in Figure 4b suggests
that most of the monolayer desorption occurs by direct exchange
of the alkylthiol with Au oxide. The latter is also supported by
the rate of current decrease (following the current maximum)
in Figure 4b, which is much faster than the blocking of bare
Au by oxide formation (Figure 2).
It is therefore concluded that the SA of alkylthiols on
anodized Au in “wet” ethanol includes the following three
concomitant processes: (i) Adsorption of the thiol on bare Au,
Figure 6. Current vs time curves for the SA of C18SH in 99.8%
ethanol + 10 mM LiClO at 1.45 V applied potential, carried out with
various C18SH concentrations as indicated in the figure. (a) and (b)
show different time scales of the process. Zero time indicates application
of the potential, while the onset of current decrease indicates solution
4
(
ii) desorption of bound thiol molecules upon replacement with
5
Au oxide, and (iii) adsorption of the thiol at oxidized Au,
formed on bare Au or upon thiol removal. The overall process,
in terms of monolayer formation rate and properties, depends
on the relative concentrations of alkylthiol and water in the
adsorption solution.
0
exchange (C18SH introduction). I is the absolute value of the initial
current in 6a (or 6b).
4
4
b (inset) recorded. It is clear from the initial current in Figure
b (inset) that short anodization at 1.45 V in ethanolic solution
Repeated Adsorption-Desorption of Alkylthiols on Gold.
An issue of substantial basic as well as practical interest is that
containing water causes complete remoVal of the C18SH

13450 J. Am. Chem. Soc., Vol. 120, No. 51, 1998
Ron and Rubinstein
Table 2. Characterization of (C18S) and C18SH Monolayers Assembled at 1.45 V Applied Potential onto Reduced or Preoxidized Au
2
contact angle (deg)
BCH
H
2
O
HD
sample
description
oxide cover.
before SA
ads. time
(min)
ellips.
thickness (Å)
adv.
108
rec.
106
adv.
52
rec.
50
adv.
46
rec.
44
1
2
3
4
5
6
0.1 mM (C18S)
2
low
low
low
high
high
high
10
17.6
18.7
19.6
12
+
salt + pot.
0.2 mM C18SH
salt + pot.
3 mM C18SH
salt + pot.
0.1 mM (C18S)
salt + pot.
0.2 mM C18SH
salt + pot.
3 mM C18SH
salt + pot.
13
108
107
90
104
104
60
49
52
37
49
49
45
50
20
43
35
45
45
25
44
39
40
43
+
18.5
9.5
9.5
6.2
+
2
<10
41
+
20.4
45.6
109
112
102
104
+
33
+
of control over the wettability of surfaces using self-assembled
monolayers. Dynamic electrochemical control of the wettability
of monolayer-covered surfaces was shown previously, either
by changing the oxidation state of a ferrocene-derivatized
2
1
22,23
monolayer or by monolayer desorption-readsorption
based
7
on cathodic alkylthiol desorption. In the former case, where
actual cycling was shown,21 the wettability difference deterio-
rates within several cycles.
The convenient adsorption and desorption of alkylthiol
monolayers at positive potentials in the FLEEC cell, together
with the ability to rapidly exchange solutions at any desired
stage and under potential control, introduce the possibility of
performing multiple C18SH monolayer adsorption-desorption
cycles for wettability and surface control. The CAs and
ellipsometric thickness of a C18SH monolayer adsorbed on a
Au electrode after two successive adsorption-desorption cycles
is presented in Table 1, row 9. Although a certain decrease in
monolayer quality is observed, probably due to surface roughen-
ing, it is clear that a reasonable C18SH monolayer is obtained,
demonstrating the possibility to turn the surface hydrophobic/
hydrophilic upon monolayer adsorption-desorption.
The present approach is different from those of Whitesides
21-23
and co-workers.
Whereas the latter authors demonstrated
wettability control in a solution of a fixed composition by
changing the applied potential, our approach is based on
adsorption and desorption at a fixed (possibly uninterrupted)
applied potential but requires solution exchange. This introduces
an obvious limitation, but also an advantage, in that it allows
the use of different molecules in subsequent adsorptions, thereby
effectively alternating not just the wettability but possibly also
the surface composition.
Figure 7. Current vs time curves for the SA of C18SH (0.2 mM) and
(
C18S)
2
(0.1 mM) onto an Au electrode in 99.8% ethanol + 10 mM
at 1.45 V applied potential. (a) and (b) show different time
SA of Alkylthiols and Dialkyl Disulfides on Oxidized Gold.
Figure 7a presents overall current-time curves for SA of C18SH
and (C18S)2 (dioctadecyl disulfide) onto slightly oxidized Au;
the marginal preoxidation occurs in the short period of time
between application of 1.45 V and introduction (by solution
exchange) of the adsorption solution into the FLEEC cell. Figure
LiClO
4
scales of the process. Zero time indicates application of the potential,
while the onset of current decrease indicates solution exchange. I is
0
the absolute value of the initial current in 7a (or 7b).
(or alkyl chains) is the same in both cases. The results are rather
similar, although the initial stage of SA of (C18S)2 is somewhat
more sluggish. The ellipsometric thickness for the two mono-
layers is similar, whereas a minor difference is observed in the
CAs. The latter is attributed to some encapsulated Au oxide
7b emphasizes the initial stages of SA. The CAs and ellipso-
metric thickness of the two monolayers are shown in Table 2,
rows 1 (0.1 mM (C18S)2) and 2 (0.2 mM C18SH). The
concentrations were chosen such that the number of sulfur atoms
5
under the C18SH monolayer, absent in the case of (C18S)2
(
(
21) Abbott, N. L.; Whitesides, G. M. Langmuir 1994, 10, 1493.
22) Gorman, C. B.; Biebuyck, H. A.; Whitesides, G. M. Langmuir 1995,
(as discussed below). Note that the thickness and CAs of the
(C18S)2 monolayer prepared at 1.45 V (Table 2, row 1) are
1
1, 2342.
very close to those observed for a monolayer adsorbed at 1.45
V from a 3 mM C18SH solution (Table 2, row 3) and for a
(23) Abbott, N. L.; Gorman, C. B.; Whitesides, G. M. Langmuir 1995,
1
1, 16.

Self-Assembled Monolayers on Oxidized Metals
J. Am. Chem. Soc., Vol. 120, No. 51, 1998 13451
The above results can be understood by considering the
mechanism of SA of alkylthiols on oxidized Au surface.
26
Wallace demonstrated that thiols react with oxides of Mn, Co,
Cu, and Fe; the reaction involves reduction of the metal oxide
and oxidation of the thiol to disulfide. Similarly, one can write
for Au oxide
Au O + 6RSH f 2Au(0) + 3RSSR + 3H O
(6)
2
3
2
resulting in bare Au surface. Thiol may then adsorb onto the
exposed Au, probably proceeding according to (at an applied
positive potential)
+
-
Au(0) + RSH f AuSR + H + e
(7)
Alternatively, the oxide reduction may lead directly to adsorbed
5
thiol, as previously suggested by us
Figure 8. Current vs time curves showing the initial stages of SA of
0
+
.1 mM (C18S)
10 mM LiClO
preoxidized Au. Zero time indicates application of the potential, while
2
, 0.2 mM C18SH, and 3 mM C18SH in 99.8% ethanol
Au O + 6RSH f 2AuSR + 2RSSR + 3H O
(8)
at 1.45 V applied potential onto electrochemically
2
3
2
4
In both cases the reaction involves thiol oxidation to disulfide.
Therefore, (C18S)2, which is the reaction product and is not a
reducing agent, cannot form a monolayer on oxidized Au. Note
that the reaction in eq 6 is expected to be faster than the one in
eq 8, since the mechanism of the former requires simultaneous
participation of two thiol molecules, whereas that of the latter
requires simultaneous participation of three thiol molecules.
It should be realized that blocking of the Au surface by oxide
at 1.45 V is a dynamic process where Au oxide is continuously
reduced by ethanol (equation 3) and regenerated by reaction
with water (equation 1). The little (C18S)2 adsorption observed
when the oxidized Au is exposed to the disulfide (Table 2, row
0
the onset of current increase indicates solution exchange. I is the same
as in Figure 4.
C18SH monolayer prepared by regular 24 h SA in 1 mM
solution (Table 1, row 1).
When comparing monolayers of C18SH and (C18S)2, it is
important to examine what are the right adsorption conditions
(i.e., concentrations) for a valid comparison. One may compare
results for, for example, 0.2 mM C18SH and 0.1 mM (C18S)2,
that is, the same number of available alkyl chains. However,
0
0
.1 mM (C18S)2 in ethanol is a supersaturated solution, whereas
.2 mM C18SH is about an order of magnitude below saturation.
4
) is likely to represent those events when the disulfide reacts
Therefore, another valuable comparison would be between, for
example, 0.1 mM (C18S)2 and 3 mM C18SH, both slightly
supersaturated in ethanol. To give a complete picture we chose
in this work to present both.
with reduced Au faster than water during this dynamic process.
As discussed above, C18SH adsorption at 1.45 V onto
oxidized Au may proceed by (i) SA onto bare Au, generated
by reaction with the thiol (eq 6) or with ethanol (eq 3) or (ii)
SA upon reaction with Au oxide (eq 8). The process depicted
in eq 6 is evident in Figure 8, where a current increase is
observed upon introduction of C18SH. Bare Au surface is
assumed to be initially generated (equation 6), providing
additional available surface for ethanol oxidation and thus a
current increase. As expected, this is more pronounced with 3
mM C18SH than with 0.2 mM C18SH and is absent in the case
The SA of C18SH and (C18S)2 monolayers on highly
oxidized Au is shown in Figure 8, emphasizing the initial stages
of the process by displaying the first 10 s. The Au surface was
effectively preoxidized upon applying 1.45 V in ethanol + 10
mM LiClO4 for ca. 4 h (with continuous slow exchange of the
solution to avoid accumulation of acetaldehyde) before intro-
duction of the adsorbing molecules into the cell by solution
exchange. The CAs and ellipsometric characterization of mono-
layers SA at 1.45 V on oxidized Au from 0.1 mM (C18S)2, 0.2
mM C18SH and 3 mM C18SH solutions are shown in Table 2,
rows 4, 5, and 6, respectively.
27
of (C18S)2. The current variations following thiol introduction
reflect a combination of enhanced ethanol oxidation (current
increase) and thiol SA (current decrease). The former is
responsible for the initial current rise; the current maximum
and subsequent decrease seem to indicate that the remaining
(slower) process manifests primarily C18SH adsorption accord-
ing to eqs 7 or 8.
As noted above, the difference in ellipsometric and CA results
between C18SH monolayers adsorbed onto highly oxidized Au
from 0.2 mM solution vs 3 mM solution (Table 2, rows 5 and
6) are indicative of a substantial amount of encapsulated Au
oxide in the latter case and little, if any, encapsulated oxide in
The thickness and CAs for (C18S)2 adsorbed at 1.45 V on
oxidized Au (Table 2, row 4) indicate very poor monolayer
formation. Clearly, Au oxide preVents adsorption of (C18S)2.
On the other hand, the results for C18SH (Table 2, rows 5 and
6
) indicate formation of densely packed monolayers, with a
small amount (row 5) and a large amount (row 6) of encapsu-
5
lated Au oxide. Evidently, this notable difference in the amount
of encapsulated Au oxide results from the faster initial mono-
layer formation in the presence of the much higher C18SH
concentration (as shown in Figure 6b for slightly oxidized Au).
(
25) The work of Hickman et al. (Langmuir 1992, 8, 357) gives the
impression that disulfide molecules adsorb onto oxidized gold. This,
however, is not the case. The authors note that no oxidized gold could be
detected (by XPS) after adsorption; clearly, the oxidized gold is reduced
by the solvent (ethanol), followed by disulfide adsorption onto the reduced
gold.
Interestingly, the observation that disulfides do not self-
assemble onto preoxidized Au was mentioned by Nuzzo and
2
4
Allara, who noted that “plasma oxidation to form gold oxide
surfaces results in no detectable adsorption” (ref 20 in their
paper). However, this remark has gone unnoticed since that
pioneering paper was published.25
(26) Wallace, T. J. J. Org. Chem. 1966, 31, 1217.
(
27) The very small current increase upon introduction of (C18S)2 (Figure
8
) seems to be related to the solution exchange itself; this is evidenced by
the fact that a similar small current increase is observed when the ethanolic
solution in the cell is exchanged with an identical background solution after
prolonged oxidation of the Au electrode.
(24) Nuzzo, R. G.; Allara, D. L. J. Am. Chem. Soc. 1983, 105, 4481.

13452 J. Am. Chem. Soc., Vol. 120, No. 51, 1998
Ron and Rubinstein
5
the former. On the other hand, the results in Figure 8 show
that introduction of 3 mM C18SH causes a larger current
increase (i.e., more oxide removed) than 0.2 mM C18SH. The
two observations combined suggest that the higher C18SH
concentration promotes more efficient initial oxide removal but
the remaining oxide is better preserved than in the case of the
lower C18SH concentration. We suggest that removal of Au
oxide upon exposure to the thiol solution may involve not only
chemical reduction (eq 6) but also detachment of some Au oxide
from the surface upon thiol adsorption. In situ AFM and STM
studies on the anodic formation of Au oxide2 showed an island
growth process, with oxide islands comprising more than a
2. Alkylthiol monolayers are formed by the new technique
considerably faster (by about 2 orders of magnitude) than by
the usual procedure, while monolayers thereby obtained show
exceptional properties in terms of monolayer thickness, wetta-
bility and FTIR results.
3
. Effective removal of alkylthiol monolayers is achieved
upon application of the same potentials in wet ethanol solutions
no thiol), while, as noted above, under the exact same
(
conditions high-quality monolayers are obtained in the presence
of the thiol in solution. These observations provide the basis
for rational manipulation of self-assembled monolayers before,
during, and after adsorption. Hence, effective control over the
wetting properties of a gold surface was demonstrated upon
repeated adsorption-desorption cycles of a C18SH monolayer.
This procedure may, in principle, be applied using different
molecules in subsequent adsorptions, thereby expanding the
possibilities of controlling interfacial properties.
8
4
monolayer of oxide. AFM studies conducted by us show that
thin Au oxide layers coated with a thiol monolayer appear brittle,
with loosely bound material that can be removed by the AFM
tip. Detachment of Au oxide upon massive adsorption of thiol
molecules is feasible only at the beginning of the SA process;
as the SA proceeds, the remaining oxide is protected by the
4. Both alkylthiols and dialkyl disulfides form densely packed
5
thiol coverage. Both processes, namely, initial detachment of
monolayers (with or without the applied potential) on oxide-
free Au surface. However, only alkylthiols (but not dialkyl
disulfides) adsorb onto preoxidized Au surface. This appears
to be due to the mechanism of alkylthiol SA onto oxidized Au,
requiring oxide reduction upon thiol oxidation to disulfide. This
may provide a novel way to form mixed monolayers of
controlled compositions, by subsequent SA of disulfides and
thiols onto partially oxidized gold. The ability to determine the
extent of preoxidation, also related to the oxide distribution on
Au oxide islands and subsequent blocking of the remaining
oxide, are more effective at higher C18SH concentrations. The
slower adsorption in 0.2 mM C18SH would allow more thiol
molecules to contact the oxidized surface and reduce the oxide
before a blocking monolayer is obtained.
Conclusions
Self-assembly of alkylthiol and dialkyl disulfide monolayers
in ethanol solutions onto Au surfaces held at positive applied
potentials where ethanol is oxidized was demonstrated. The new
procedure presents several distinct advantages over the common
4
the surface, introduces the possibility to design systems with
desired relative amounts and distribution of two monolayer
components. This is currently under further investigation.
2
9
procedure.
1
. Changes in the current corresponding to ethanol oxidation
Acknowledgment. We thank Mr. H. Weizman (WIS) for
his help with the synthesis of the disulfide (C18S)2, Mr. V.
Notkin for writing the computer program for the electrochemical
experiments, and the Scientific Services Department of the
Weizmann Institute for skilled technical assistance. Constructive
discussions with Dr. J. R u¨ he (MPI, Mainz) are gratefully
acknowledged. I.R. acknowledges support of this work by the
MINERVA Foundation, Germany, and the Israel Science
Foundation.
can be used for convenient real-time monitoring of the SA
process and its completion. This new method complements
others such as surface plasmon resonance or quartz crystal
microbalance, while being more sensitive and easier to imple-
ment. It presents the possibility to study monolayer adsorption
kinetics, an aspect currently under further investigation.
(
28) Nichols, R. J.; Magnussen, O. M.; Hotlos, J.; Twomey, T.; Behm,
R. J.; Kolb, D. M. J. Electroanal. Chem. 1990, 290, 21.
29) A limitation of the new electrochemical self-assembly procedure is
(
its inapplicability to monolayers containing functional groups that can be
oxidized under the applied potential.
JA9823722