Exchange Dynamics of Redox-Active Self-Assembling Mixed Monolayers

Article abstract of DOI:10.1021/jp952541u

The exchange dynamics as well as the desorption and displacement reactions of redox-active self-assembling monolayers based on transition-metal complexes of osmium and ruthenium have been studied.On the basis of results from the above-mentioned processes, it appears that the exchange dynamics are controlled by the rate of desorption via a dissociative mechanism.We have also made use of wave-shape analysis in terms of shifts in the formal potentials and widths of the voltammeric waves with changes in the surface coverage and monolayer composition.The presence of mixed monolayers (Os/Ru) results in an increased broadening of the voltammeric wave associated with the ruthenium complex relative to the case of a monolayer of the ruthenium complex alone which we ascribe to an enhanced degree of electrostatic repulsion.Also based on an analysis of the voltammeric response, we belive that the mixed monolayers are randomly distributed rather than being composed of discrete and separated phases.

Full text of DOI:10.1021/jp952541u

4556
J. Phys. Chem. 1996, 100, 4556-4563
Exchange Dynamics of Redox-Active Self-Assembling Mixed Monolayers
Jorge D. Tirado and He´ctor D. Abrun˜a*
Department of Chemistry, Baker Laboratory, Cornell UniVersity, Ithaca, New York, 14853-1301
ReceiVed: August 30, 1995; In Final Form: December 14, 1995X
The exchange dynamics as well as the desorption and displacement reactions of redox-active self-assembling
monolayers based on transition-metal complexes of osmium and ruthenium have been studied. On the basis
of results from the above-mentioned processes, it appears that the exchange dynamics are controlled by the
rate of desorption via a dissociative mechanism. We have also made use of wave-shape analysis in terms of
shifts in the formal potentials and widths of the voltammetric waves with changes in the surface coverage
and monolayer composition. The presence of mixed monolayers (Os/Ru) results in an increased broadening
of the voltammetric wave associated with the ruthenium complex relative to the case of a monolayer of the
ruthenium complex alone which we ascribe to an enhanced degree of electrostatic repulsion. Also based on
an analysis of the voltammetric response, we believe that the mixed monolayers are randomly distributed
rather than being composed of discrete and separated phases.
Introduction
Self-assembled monolayers are currently the object of intense
study because their well-defined structural character provides
molecular surfaces of great potential for scientific studies and
technological exploration. Mixed monolayers of self-assembling
molecules have been the subject of several recent publications.1-14
Most of the work of self-assembling monolayers has involved
alkane thiols on gold. Some of these systems involve molecules
with long and short alkane chains which give rise to single-
phase or multiphase monolayers, depending on the conditions
under which the adsorption is carried out.1 Other systems
combine alkane thiols with different end-group functionalities.
For example, electroactive groups (such as ferrocene) mixed
with alkane thiols have been studied.2 In addition, the effects
of different organic functionalities on the properties of mixed
monolayers have been the subject of various studies.3,4 Mixed
self-assembling monolayers have been characterized by a several
techniques including X-ray photoelectron spectroscopy (XPS),5
contact angle,6 ellipsometry,7 and, more recently, time-of-flight
secondary ion mass spectrometry8 and scanning tunneling
microscopy (STM),9 among others.10
As applications of the self-assembly method in electrochem-
istry are explored further, it will become increasingly desirable
to prepare mixed layers for which the relative quantities of two
or more different adsorbates on a surface can be systematically
Figure 1. General structure of the complexes under study and a model
depicting a molecule adsorbed on a platinum electrode surface.
varied and controlled. This will enable complex molecular
superstructures to be prepared with the ultimate goal of
The molecules employed in this work have been previously
accomplishing specific functions such as chemical sensing or
characterized in terms of their thermodynamic parameters,15
various molecular electronic functions.11
dynamics of adsorption,16 and electron-transfer rates17 and
In this paper we discuss the exchange dynamics within a
consist of transition-metal complexes of the type [M(bpy)2LCl]+
where M is either Os or Ru, bpy ) 2,2′-bipyridine, and L )
mixed-monolayer formed by two electroactive self-assembling
molecules using cyclic voltammetry as the electroanalytical tool.
1,3-bis(4-pyridyl)propane which is referred to as dipy. The
Chidsey and co-workers2 previously studied coadsorption of
osmium and ruthenium complexes are referred to as Osdipy
ferrocene-terminated and unsubstituted alkanethiols, and similar
and Rudipy, respectively. The dipy ligand has a pendant
pyridine group which serves as the anchor through which the
complex binds to the surface of a platinum electrode. Figure 1
shows the structure of the molecule and a model of the adsorbed
complex on a platinum surface. Because these two complexes
are virtually identical in size whereas their redox responses are
well separated (by about 400 mV), they provide an ideal system
with which to probe the exchange dynamics of redox-active
experiments, without electroactive groups, have been described
by Allara et al.12 and Whitesides and co-workers.13 Faulkner
and Foster recently demonstrated the use of the difference in
the kinetics of electron transfer of two species with identical
formal potentials in a mixed-monolayer to determine their
surface coverages along with their bulk concentrations.14
X Abstract published in AdVance ACS Abstracts, February 15, 1996.
0022-3654/96/20100-4556$12.00/0 © 1996 American Chemical Society

Redox-Active Self-Assembling Mixed Monolayers
J. Phys. Chem., Vol. 100, No. 11, 1996 4557
self-assembling monolayers which constitutes the objective of
the present work.
polycrystalline platinum electrode was obtained. The electro-
chemically active area of the electrode was obtained by
integration of the hydrogen adsorption waves using a conversion
factor of 210 µC/cm2.19 Roughness factors, defined as the ratio
of electrochemically determined area to geometric area were
typically about 1.2-1.3. The electrode was rinsed with water
and subsequently immersed in an aqueous 0.1 M solution of
KClO4 and cycled between 0.0 and +1.0 V until a steady
voltammogram was obtained. All the cyclic voltammograms
shown in this paper were obtained at 100 V/s. This relatively
fast scan rate was chosen so as to kinetically hinder the oxidation
process of the platinum surface that can take place to a
significant extent at potentials close to +1.0 V and slow scan
rates (e.g. 100 mV/s). It was observed that at a scan rate of
100 V/s the contribution to the background from oxide formation
was small. The complexes were injected as an acetone solution,
typically 0.5 mM, and the solution was subsequently purged
with prepurified nitrogen in order to homogenize the solution
prior to immersing the electrode. The volume injected varied
according to the desired final concentration and concentration
ratios of the complexes but was typically in the microliter range.
The potential of the electrode was held constant during
deposition at 0.0 V. Cyclic voltammetry was used to determine
the coverage of the complexes on the surface of the electrode
by integration of the wave corresponding to the metal-based
oxidation from M(II) to M(III) (where M ) Os, Ru) with formal
potentials at about +0.35 V for Osdipy and +0.75 V for Rudipy,
respectively. Since the solution concentration of the complexes
was typically in the micromolar regime, their contribution to
the measured current (charge) was negligible. Thus, surface-
coverage measurements could be carried out in the deposition
solution with minimal error from the complex in solution. The
coverage was monitored as a function of time for both
complexes at different total concentrations and/or concentration
ratios. Mixed monolayers of Osdipy and Rudipy on platinum
electrodes were prepared by initially coating the surface with
one complex and injecting into the solution the desired volume
of an acetone solution of the other.
In addition, we present data on the kinetics of desorption and
displacement and employ data from all these experiments to
propose a mechanism for the above-mentioned processes.
Moreover, we show how the electrochemical response can be
used to derive (albeit indirectly) information on the micro-
environment within the mixed-monolayer system.
Experimental Section
Reagents. Water was purified with a Hydro purification
system connected in series to a Millipore Milli-Q system.
KClO4 (GFS Chemicals) was recrystallized twice from water
and dried under vacuum at room temperature for 72 h. H2SO4
(Ultrex, J. T. Baker) and 1,3-bis(4-pyridyl)propane; dipy (Al-
drich) and all other reagents were of at least reagent-grade
quality and were used as received unless otherwise specified.
The metal complexes (adsorbates) were synthesized following
procedures described below. The synthesis of Osdipy was
carried out by a procedure previously described17 and which
follows procedures previously reported by Meyer et al.18a,b and
Dwyer et al.18c About 200 mg of [Os(bpy)2Cl2] (0.35 mmol)
was heated at reflux under a nitrogen atmosphere in deaerated
ethylene glycol (ca. 25 mL) for 3 h with 2 equiv of the dipy
ligand. After cooling, an equivalent volume of water was added,
and the solution was subsequently filtered. The complex was
precipitated with a saturated aqueous solution of NH4PF6. The
product was collected, washed with water, and dried with ether.
Purification was effected by chromatography on alumina with
15% methanol in toluene as eluent. The desired complex eluted
as a brownish band. The synthesis and purification of Rudipy
was similar to that described above with the following changes.
About 200 mg of [Ru(bpy)2Cl2‚2H2O] (0.38 mmol) were
refluxed under a nitrogen atmosphere in a deaerated 1:1
ethanol-water mixture (ca. 25 mL) for 3 h with 2 equiv of the
ligand, until a brownish solution was obtained. After reaching
room temperature, an equal volume of water was added and
the solution was filtered. The product was precipitated with a
saturated aqueous solution of NH4PF6. After filtering and
washing with ether, a brownish precipitate was collected.
Purification was by column chromatography on neutral alumina
and 15% methanol/toluene as eluent. The Rudipy complex
eluted as the first brown band.
Instrumentation. Cyclic voltammograms were carried out
with an EI-400 bipotentiostat from Ensman Instrumentation, and
data were collected using a Gateway 2000 PC (486DX) system
equipped with an interface card from National Instruments
Model AT-MIO-16F-5 programmed with LabView. Potentials
were measured against a sodium chloride saturated silver/silver
chloride (Ag/AgCl) reference electrode without regard for the
liquid junction potential. A 15 mL glass vial was used as the
electrochemical cell, with a coiled platinum wire as the counter
electrode.
Procedures. The platinum disk working electrode was made
by sealing, in soft glass, a 3-4 cm piece of annealed platinum
wire (75 µm diameter, Engelhard Ind.). The electrode was
sanded with 400 and 600 grit sand paper (Buehler) and polished
with 1 µm diamond paste (Buehler). Electrical connection to
the electrode was made with a drop of mercury and a piece of
copper wire inserted in the glass tubing. Prior to use the
electrode was polished with 1 µm diamond paste (Buehler) and
washed with water and acetone. The electrode was pretreated
by continuous cycling between the oxide formation and
hydrogen adsorption potentials (about +1.40 and -0.20 V,
respectively) in 1 M H2SO4 until the voltammetry of a clean
Results and Discussion
The adsorption process which governs the formation of self-
assembling monolayers can be viewed as the competition of
ions, solvent, and adsorbate molecules in solution for binding
sites on the substrate surface. The following equilibrium
expression can be used to describe such a situation:
Msol + nSads h Mads + nSsol
(1)
where Msol represents adsorbate molecules in solution, Sads is
the solvent and/or ions adsorbed on the substrate, n of which
are displaced as Ssol into solution by Msol to form Mads, the
adsorbate bound to the surface. The amount of material
adsorbed on the substrate will depend on the concentration of
the adsorbate in solution. At low solution concentrations a
fraction of a monolayer will be formed. As the solution
concentration of the adsorbate increases so does the coverage,
until a concentration value is reached at which the substrate’s
surface is saturated. At concentrations above this value, a
saturation coverage is always obtained. We previously re-
ported15 that for the systems presented in this paper the
minimum concentration to achieve a saturation coverage (which
is about 1.0 × 10-10 mol/cm2) was about 0.5 µM. To maintain
a full monolayer on the surface of the electrode at all times, a
total concentration of 2 µM was used in the experiments
presented in this paper.

4558 J. Phys. Chem., Vol. 100, No. 11, 1996
Tirado and Abrun˜a
Figure 2. Cyclic voltammograms at 100 V/s in aqueous 0.1 M KClO4
of a monolayer of (A) Osdipy (about 8.3 × 10-11 mol/cm2) and (B)
Rudipy (about 8.1 × 10-11 mol/cm2) on a platinum electrode (area )
1.0 × 10-4 cm2).
Let us assume that a monolayer of adsorbate M is formed on
a substrate, as described by eq 1. If this substrate modified
with M is exposed to a solution of a different adsorbate (N), an
additional equilibrium is established which involves the dis-
placement of adsorbate M by N. This equilibrium can be
expressed as:
Figure 3. Cyclic voltammograms at different times for the displace-
ment experiments. (A) An initial monolayer of Rudipy displaced by
Osdipy from a 2 µM solution. Times: (a) 0; (b) 2; (c) 6; (d) 10; (e) 57
min. (B) An initial monolayer of Osdipy displaced by Rudipy from a
2 µM solution. Times: (a) 0; (b) 1; (c) 3; (d) 6; (e) 110 min.
TABLE 1: Solution Concentration and Coverage Ratios for
Osdipy and Rudipy Mixed Monolayers
Mads + Nsol h Msol + Nads
(2)
solution concn ratio
[Ru]/[Os]
coverage ratio
ΓRu/ΓOs
On the other hand, if two different adsorbates (M and N) are
initially present in solution, a competitive equilibrium is
established between these two species for binding sites on the
substrate. This equilibrium can be expressed as
0.33
0.53
1.03
2.09
3.00
0.27
0.41
0.80
1.61
2.26
Msol + Nsol + nSads h Mads + Nads + nSsol
(3)
as described in eq 2. Cyclic voltammograms obtained at
different times after exposure of a monolayer of Rudipy to an
Osdipy solution (Figure 3A) show that this is indeed the case.
Figure 3A shows an initially formed monolayer of Rudipy being
displaced by Osdipy from a 2 µM solution of the latter
containing no dissolved Rudipy. The areas under the voltam-
metric waves are a measure of the amount of each material
adsorbed on the electrode surface. It can be seen that the area
under the voltammetric wave for Rudipy (centered at +0.75
V) decreases to nearly zero while the area under the wave for
Osdipy (centered at +0.35 V) exhibits a concomitant increase.
The same results are obtained for the reverse case, that is a
monolayer of Osdipy initially adsorbed is displaced when put
in contact with a solution of Rudipy. Figure 3B shows the
results obtained from such an experiment.
The effects of different concentration ratios of the complexes
in solution on the composition of the resulting monolayer were
also investigated. If the system is under equilibrium at all times
the composition of the monolayer should reflect the ratio of
the complexes in solution. Table 1 shows the results obtained
in five experiments where the concentration ratio of the
complexes ([Ru]/[Os]) was varied from about 0.3 to 3.0 at a
constant total concentration of 2 µM. The coverage ratio of
the complexes, determined by cyclic voltammetry, showed the
where M and N are different adsorbates competing for the
surface. If there is no selectivity of the surface for a particular
adsorbate, the monolayer composition on the substrate should
reflect the bulk solution composition.
The formation of a mixed monolayer of the Osdipy and
Rudipy complexes was investigated to determine if the equi-
librium situation described above was applicable. These studies
may allow a better understanding of the mechanism of mono-
layer formation as well as the structure of these mixed
monolayers which may provide the basis for further applications
of the concept of mixed monolayers of redox-active self-
assembling molecules.
The cyclic voltammetric signatures of the adsorbed complexes
studied are depicted in Figure 2 with the charge under the waves
in the cyclic voltammograms corresponding to about one
monolayer (1.0 × 10-10 mol/cm2) of Osdipy (A) and Rudipy
(B), respectively. Since the characteristic formal potentials of
the complexes are separated by about 400 mV, identification
and quantification of each material in a mixed monolayer are
easily achieved.
A reversibly adsorbed monolayer in equilibrium with its
complex in solution should be displaced (exchanged) when
contacted with a different solution containing another adsorbate,

Redox-Active Self-Assembling Mixed Monolayers
J. Phys. Chem., Vol. 100, No. 11, 1996 4559
while that for the Rudipy complex exhibits a concomitant
increase. If the charges associated with these peaks are obtained
at different times, the dynamics of exchange can be studied.
Figure 5B shows the results obtained from the data in Figure
5A. The anticipated decrease in the Osdipy charge (coverage)
as well as the increase in that due to Rudipy are observed until
both reach an equilibrium value, determined by their respective
solution concentrations. Also plotted in Figure 5B is the total
charge which, as can be seen, remains virtually constant during
the experiment. This demonstrates that the process being
monitored is the surface exchange dynamics at constant surface
coverage in which no stacking nor multiple layers are formed.
The presence of isopotential points (at +0.45 V in Figure 5A)
is also evident. Since isopotential points arise as a result of an
equilibrium process between two surface species at constant total
coverage,20 this provides compelling evidence that the exchange
process is one that is at equilibrium at all times. This is an
important observation since it implies that the time evolution
of the changes in surface coverage can be employed to ascertain
the kinetics of desorption, displacement, or exchange, depending
on the specific experiment under consideration. These three
processes are pictorially shown in Scheme 1. These processes
are the focus of this paper and are described in detailed below.
From the processes depicted in Scheme 1, adsorption and
desorption rates can be obtained under different conditions
which will help in understanding the mechanism of exchange
between the two species under study. Simple exponential
expressions were used to describe the increase (eq 4) or decrease
Figure 4. Plot of surface coverage ratio (ΓRu/ΓOs) vs solution
concentration ([Ru]/[Os]) ratio.
Qt,a ) Qe,a(1 - exp(-kat))
(4)
(5)
Qt,d ) (Qi,d - Qe,d) exp(-kdt) + Qe,d
(eq 5) of charge due to adsorption or desorption, respectively,
of the complexes. In these equations Qt,a and Qe,a are the charge
at time t and equilibrium charge for the adsorbing species,
respectively; and Qt,d, Qe,d, and Qi,d are the charge at time t,
equilibrium charge, and initial charge for the desorbing species,
respectively. From fits of the experimental points to eqs 4 and
5 the rate constants of adsorption (ka) and desorption (kd) of
the complexes in solution could be obtained. This system of
equations was obtained from an analysis similar to that
previously reported by this laboratory on studies of the
adsorption dynamics of a family of compounds analogous to
Osdipy.16
The solid lines in Figure 5B are fits of the data to eq 5 in the
case of the desorption of Osdipy and to eq 4 for the adsorption
of Rudipy. The dotted line represents the sum of the individual
fits for Osdipy and Rudipy, and the dashed line represents the
average of the total charge. The fits obtained are in excellent
agreement with the data. The rate constants obtained from the
fits to this and other experiments are shown in Table 2. Given
that these two complexes are virtually identical in terms of their
adsorptive behavior and that the total concentration of adsorbate
used (2 µM) was in the saturation coverage regime, one would
a priori anticipate that the rates of adsorption and desorption
would be the same and that the total coverage would remain
constant throughout. The data in Figures 5 and 6 and Table 2
are consistent with this although it should also be noted that
(as can be seen in Table 2) there is significant error in some of
the measurements. Nonetheless, and taking these considerations
into account, the experimental observations are consistent with
expectation. At a microscopic level, the exchange mechanism
could involve a dissociative or an associative process with the
slower of the two controlling the overall exchange dynamics
of the system. We now look at each case in detail.
Figure 5. (A) Cyclic voltammograms at different times (i.e., 0, 3, 10,
20 min) to monitor the exchange dynamics of Osdipy with Rudipy.
(B) Charge variation with time for Osdipy (b), Rudipy (9), and total
charge (2). Solid lines are fit to eqs 4 and 5 for Rudipy and Osdipy,
respectively. The dotted line represents the sum of the two solid lines,
and the dashed line is the average total charge.
same increasing trend although systematically lower (by about
25%) when compared to the solution composition. A plot of
surface coverage ratio vs solution concentration ratio ([Ru]/[Os])
is presented in Figure 4. Although an excellent correlation (r
) 0.999) is obtained, the slope is less than unity (0.75) which
would suggest that there is a slight preference of the surface
for the osmium complex. Given that the free energies of
adsorption of these materials are virtually identical, we are not,
at this time, certain of the origin of this effect. This discrepancy
may be attributed to the charge integration process which would
affect the determination of the coverage by the Ru complex.
Since the platinum oxide formation overlaps somewhat with
the Rudipy signal, it makes it more difficult to subtract its
background current when compared to the Osdipy complex. The
results shown here are comparable to those reported by Faulkner
et al. with a similar system.14
The exchange dynamics were studied by recording cyclic
voltammograms at different times during the equilibration (from
a homogeneous to a mixed monolayer) process. Figure 5A
depicts the formation of mixed monolayer from a solution
containing a 1:1 ratio of Rudipy/Osdipy. In this case the
electrode was initially covered with a monolayer of Osdipy,
and Rudipy was subsequently injected so as to achieve a 1:1
concentration ratio in solution. It is apparent from the figure
that the voltammetric peak due to adsorbed Osdipy decreases

4560 J. Phys. Chem., Vol. 100, No. 11, 1996
Tirado and Abrun˜a
SCHEME 1: Pictorial Representation of the
Experiments Presented in This Papera
a
Desorption experiment: a monolayer is exposed to supporting
electrolyte, and its desorption is monitored as a function of time.
Displacement experiment: a monolayer is exposed to a solution of a
different adsorbate, and the displacement is monitored as a function of
time. Exchange experiment: an initial monolayer in equilibrium with
its adsorbate in solution is exchanged with a different adsorbate injected
into the original solution.
Figure 6. Plot of charge vs time for the desorption of a monolayer of
(A) Osdipy and (B) Rudipy into 0.1 M KClO4 aqueous solution.
TABLE 3: Desorption and Adsorption Rate Constants for
the Displacement of an Initial Monolayer by a Different
Adsorbate from a 2 µM Solution in the Absence of the
Initial Complex
TABLE 2: Desorption and Adsorption Rate Constants for
Osdipy and Rudipy from the Exchange Experiments
initial
monolayer
desorption rate
adsorption rate
initial
monolayer
concn ratio
[Ru]/[Os]
desorption rate
adsorption rate
(kd,i)/10-1 min-1
(ka,i)/10-1 min-1
(kd,i)/10-2 min-1
(ka,i)/10-2 min-1
Osdipy
Rudipy
kd,Os ) 2.6 ( 0.1
kd,Ru ) 1.9 ( 0.1
ka,Ru ) 6.6 ( 0.9
ka,Os ) 1.3 ( 0.1
Osdipy
Rudipy
Osdipy
Osdipy
Rudipy
0.33
0.53
1.03
3.00
3.00
kd,Os ) 4.9 ( 3.5
kd,Ru ) 5.7 ( 0.9
kd,Os ) 7.8 ( 1.4
kd,Os ) 5.0 ( 0.4
kd,Ru ) 3.5 ( 0.7
ka,Ru ) 6.4 ( 1.6
ka,Os ) 8.5 ( 1.2
ka,Ru ) 11.3 ( 1.5
ka,Ru ) 6.6 ( 0.9
ka,Os ) 3.4 ( 0.1
decrease down to the monolayer value. This, however, is clearly
not observed. If an associative mechanism were involved,
desorption and displacement experiments (as shown in Scheme
1) should yield different results. The displacement experiments
would have conditions that would favor the desorption if it takes
place in an associative way when compared to the conditions
present in simple desorption experiments, and this should be
reflected in the rate constants for desorption in each case.
Displacement experiments are shown in Figure 3 and Table 3
summarizes the rate constants obtained after fitting the data to
eqs 4 and eq 5. Desorption experiments were done as follows:
a monolayer of the complex, Osdipy or Rudipy, was adsorbed
on the electrode, which was subsequently placed in a clean
supporting electrolyte solution where the desorption was
monitored with cyclic voltammetry. Figure 6A,B shows the
results obtained for Osdipy and Rudipy, respectively. Solid lines
are fits of the data to eq 5, and the rate constants obtained are
presented in Table 4. All the desorption rate constants obtained
are within the same order of magnitude; slightly larger values
are observed for the displacement experiments which would
suggest that some association might be taking place enhancing
the rate of desorption. If this is taking place in the exchange
experiments, it would have to be to a small extent because there
is no increase in the total charge during the experiment, as
mentioned earlier. On the basis of this analysis, we believe
In a dissociative mechanism where the desorption process is
rate determining, the variation of the total charge (coverage)
with time should be constant. For a molecule in solution to
adsorb, a surface site must be made available. As soon as a
site for adsorption is accessible, the adsorbing species occupies
it, and as a result no change in the total charge is observed at
different times. This is precisely what we observe in our
experiments as shown in Figure 5B. Such behavior is also
consistent with the fact that under the experimental conditions
where the bulk concentration of adsorbate (2 µM) is at the
saturation coverage plateau, the rate of adsorption would be
expected to be rapid.
An associative mechanism involving material in solution
loosely associated with the monolayer, possibly intercalating
within the film and helping adsorbed molecules to desorb could
be a reasonable alternative if the intercalation or weak associa-
tion occurs only to a small extent. The cyclic voltammetric
technique would not detect this interaction if it represents
5-10% of a monolayer since this is the magnitude of the error
in the coverage measurements. If the extent of interaction is
higher than 10%, one would anticipate an increase in the total
charge at the beginning of the experiment with a succeeding

Redox-Active Self-Assembling Mixed Monolayers
J. Phys. Chem., Vol. 100, No. 11, 1996 4561
TABLE 4: Desorption Rate Constants of the Individual
Monolayers into Clean Supporting Electrolyte
initial
monolayer
desorption rate
(kd,i)/10-1 min-1
Osdipy
Rudipy
kd,Os ) 1.5 ( 0.1
kd,Ru ) 0.9 ( 0.2
that a dissociative mechanism and/or an associative mechanism,
to a small extent, can be used to describe the exchange
dynamics.
Previous studies on the adsorption kinetics of this type of
molecule provide additional evidence in support of the proposed
mechanism in the exchange dynamics.16 We have previously
shown that the adsorption process is under kinetic (rather than
diffusion) control and that it obeys the equilibrium described
in eq 1. Clearly, the conditions were such that the adsorption
process dominated over the desorption, resulting in the formation
of a monolayer on a platinum electrode in contact with a solution
of the complex. From those experiments the measured rate
constant for adsorption at a solution concentration similar to
Figure 7. Cyclic voltammograms at different times (i.e., (a) 0, (b) 10,
(c) 30 min) to monitor the exchange dynamics of an initially deposited
monolayer of Rudipy in contact with an aqueous solution of Osdipy
and Rudipy (1:1) at a total concentration of 2 µM in 0.1 M KClO4.
that employed in this work (e.g., 2 µM) was about 0.13 min-1
,
In general, the width of the peak depends on the types of
interactions among the molecules making the monolayer. The
voltammetric peak becomes narrower and sharper if attractive
forces predominate and wider and more rounded if, on the
contrary, repulsive forces prevail.21 In our experiments, the
observed ∆Efwhm is larger than the ideal value indicating
repulsive (destabilizing) interactions. This is expected since the
molecules in the monolayer are positively charged (+1 in the
reduced form, +2 in the oxidized form), and electrostatic
repulsions between the headgroups would take place. Another
interesting observation reported for similar systems is the
dependence of the formal potential (E°′) on the total coverage.17
In a qualitative way one can say that the oxidation becomes
more difficult (shift to more positive potentials) at higher
coverage because the repulsions between molecules increase
with coverage. The ideas described above will be considered
within the context of the experiments performed on mixed-
monolayer systems.
Figure 7 shows the cyclic voltammograms for adsorbed
Rudipy from an experiment in which a Rudipy monolayer was
initially deposited (cyclic voltammogram a), and then Osdipy
was injected to the solution to have an equivalent concentration
of the former and to form the mixed monolayer. Cyclic
voltammogram b was acquired after 10 min, and cyclic
voltammogram c after 30 min. No significant changes were
observed after this time period suggesting that equilibrium was
reached. The expected decrease in the Rudipy signal is observed
together with a concomitant increase in the Osdipy wave. Two
other effects were observed. First, the apparent formal potential
for the Rudipy complex was shifted toward more positive
potentials (by about 40 mV); and second, the wave for the
Rudipy complex became broader; after the presence of the
Osdipy in the mixed monolayer. We believe that the shift in
formal potential and the broadening of the Rudipy wave are
due to the change in the chemical environment around the
Rudipy complex when the Osdipy is present in the mixed
monolayer.
which is significantly faster than all of the desorption rates
presented in Table 2. Desorption experiments shown in Figure
6A,B and their respective rate constants shown in Table 4 were
used for comparison. The magnitude of these rate constants is
comparable to the adsorption rate constant values quoted above.
However, it is important to keep in mind that these desorption
experiments represent a nonequilibrium situation since there is
no complex in solution. Even though the experimental condi-
tions for desorption are different from the adsorption experi-
ments, the rate constants obtained for the former can be used
as a reference point. Under experimental conditions where net
adsorption is observed, the magnitude of the desorption rate
has to be smaller than the rate of adsorption; so the exchange
process is controlled by the desorption step, as is experimentally
observed.
The slope of the solid lines in Figure 5B is a measure of the
rate of the exchange process. From the desorption equation
(eq 5) an expression can be derived to obtain the value of the
rate of the exchange dynamics. From eq 5, eq 6 can be obtained:
dQt /dt ) -kd(Qi,d - Qe,d)exp(-kdt)
(6)
If kd ) 5.4 × 10-2 min-1, the average value of kd from Table
2 is used, and time (t) is set to zero, the initial rate of exchange
can be calculated. For the experiment shown in Figure 5B, an
initial exchange velocity of 5% of a monolayer per minute is
obtained, which indicates a very dynamic system. The rate of
interchange slows down until equilibrium is reached and no
further net exchange is observed.
Wave-Shape Analysis and Near-Neighbor Interactions.
The voltammetric response of electroactive molecules in a self-
assembling monolayer can be used to probe the chemical
environment within the film. In the ideal cyclic voltammogram
20,21
for a surface-confined species,
the current is directly
proportional to the scan rate (ν), the cathodic wave on scan
reversal is the mirror image of the anodic wave reflected across
the potential axis which means that Ep,a ) Ep,c, and the full
width at half-maximum (∆Efwhm) is given by eq 7. Such ideal
As mentioned before, the redox potential for Osdipy is about
+0.35 V and for Rudipy it is about +0.75 V. For a monolayer
containing only Rudipy at potentials prior to its oxidation; say
+0.50 V, all the molecules have a +1 charge. The Rudipy
molecules in the monolayer experience electrostatic repulsions
which is reflected in the value of ∆Efwhm ) 200 mV, larger
than the value of 90.6 mV when there are no interactions
between particles in the monolayer. When Osdipy is introduced
into the film at the same level (e.g., 1:1 ratio within the
∆Efwhm ) 90.6/n mV
(7)
behavior is frequently not observed experimentally. Cyclic
voltammograms of adsorbed species are typically broader than
predicted. Deviations from the ideal behavior can be caused
by the nature of the chemical environment within the monolayer.

4562 J. Phys. Chem., Vol. 100, No. 11, 1996
Tirado and Abrun˜a
Figure 8. Sections of cyclic voltammograms in Figure 7 corresponding
to the oxidation of (A) Rudipy in the monolayer by itself and (B) in
the presence of Osdipy in a 1:1 ratio (c in Figure 7). Symbols are the
experimental points, dashed lines are the background, dotted lines
represent a Gaussian fit to the experimental points after subtraction of
the background, and the solid lines combine the background with the
fit.
Figure 9. (A) Cyclic voltammograms in Figures 8A and 8B normalized
by the total charge under the peak. (B) Overlapped normalized cyclic
voltammograms from Figure 9A.
SCHEME 2: Pictorial Representation of the Two
Possible Mixed States That Could Be Attained after
Reaching Equilibriuma
monolayer) some of the Rudipy molecules will have Osdipy
near neighbors which at a potential of +0.50 V are already
oxidized to +2. This will give rise to larger electrostatic
repulsions so that the Rudipy oxidation wave will be shifted to
more positive potentials since a larger driving force will be
needed to oxidize the Rudipy in this less favorable (for the
oxidized state) environment. In addition, broader peaks are
expected in the presence of Osdipy reflecting the higher
electrostatic repulsion between particles. Figure 8A,B shows
the regions of the cyclic voltammograms in Figure 7 corre-
sponding to the oxidation of Rudipy in the case of a full Rudipy
monolayer and for the 1:1 mixed monolayer (after 30 min),
respectively. The symbols correspond to the data points, the
dashed lines are the background, the dotted line represents a
Gaussian fit to the peak, and the solid line through the symbols
combines the Gaussian fit with the background. To compare
the two fits with respect to their width, the current was
normalized to the total charge in each case. Figure 9A shows
the results obtained after normalization (relative to the total
charge) with the solid line representing the Rudipy monolayer
alone and the dashed line the mixed monolayer, where one can
note the shift in potential in going from the Rudipy monolayer
to the mixed-monolayer. To compare the width of the waves
the Rudipy wave was shifted (by 28 mV) so as to overlap it
with the mixed monolayer wave as shown in Figure 9B. It can
be seen that in the mixed-monolayer case the Rudipy wave is
broader. The ∆Efwhm value increased from 200 mV for the
Rudipy monolayer to 230 mV for the mixed monolayer case.
The changes described above were observed in all the experi-
ments in which the Rudipy monolayer was initially present on
the surface of the electrode. On the other hand, in experiments
where the Osdipy was initially present on the surface of the
electrode, after the introduction of Rudipy to the film, the
Rudipy wave was always broad, as shown for example in Figure
5A.
a
(A) Randomly mixed monolayer. (B) Phase-separated mixed
monolayer.
The electrochemical response of the mixed monolayer can
also provide some information as to the final arrangement of
the monolayer after equilibrium is reached. Scheme 2 depicts
two possible structures of the monolayer after the exchange
experiment has reached equilibrium. One possibility (case A)
is to have a random distribution of the Osdipy and Rudipy
molecules. On the other hand the molecules could segregate

Redox-Active Self-Assembling Mixed Monolayers
J. Phys. Chem., Vol. 100, No. 11, 1996 4563
References and Notes
into two domains as shown in case B. We believe that case A,
the random distribution of particles in the monolayer, better
describes the system. The fact that the cyclic voltammetric wave
of Rudipy shifts to positive potentials and broadens in the
presence of Osdipy indicates the presence of close interactions
between the molecules within the mixed monolayer. The
formation of two phases would allow interactions of this kind
only at the boundaries so that the effects would be less marked.
For the mixed monolayer to segregate, there must be a
preferential interaction between identical molecules. We do not
believe that is the case, since except for the identity of the metal
center, the complexes are virtually identical. A third possibility
would be the result of a combination of the two cases shown.
That is, one could have the surface randomly decorated with
islands of each complex. To differentiate between this case
and that discussed previously, additional studies will be required.
These studies are currently being carried out, and their results
will be reported elsewhere.
(1) For example: (a) Folkers, J. P.; Laibinis, P. E.; Whitesides, G.
M.; Deutch, J. J. Phys. Chem. 1994, 98, 563-571. (b) Folkers, J. P.;
Laibinis, P. E.; Whitesides, G. M Langmuir. 1992, 8, 1330-1341. (c)
Laibinis, P. E.; Fox, M. A.; Folkers, J. P.; Whitesides, G. M. Langmuir.
1991, 7, 3167-3173.
(2) Chidsey, C. E. D.; Bertozzi, C. R.; Putvinski, T. M.; Mujsce, A.
M. J. Am. Chem. Soc. 1990, 112, 4301-4306.
(3) Zhang, L.; Lu, T.; Gokel, G. W.; Kaifer, A. E. Langmuir 1993, 9,
786-791.
(4) Yip, C. M.; Ward, M. D. Langmuir 1994, 10, 549-556.
(5) (a) Bain, C. D.; Whitesides, G. M. J. Phys. Chem. 1989, 93, 1670-
1673. (b) Laibinis, P. E.; Bain, C. D.; Whitesides, G. M. J. Phys. Chem.
1991, 95, 7017-7021.
(6) (a) Ong, T. H.; Ward, R. N.; Davies, P. B.; Bain, C. D. J. Am.
Chem. Soc. 1992, 114, 6243-6245. (b) Dimitrov, A. S.; Kralchevsky, P.
A.; Nikolov, A. D.; Noshi, H.; Matsumoto, M. J. Colloid Interface Sci.
1991, 45, 279-282.
(7) (a) Bain, C. D.; Troughton, E. B.; Tao, Y.-T.; Evall, J.; Whitesides,
G. M.; Nuzzo, R. G. J. Am. Chem. Soc. 1989, 111, 321-335. (b) Porter,
M. D.; Bright, T. B.; Allara, D. L.; Chidsey, C. E. D. J. Am. Chem. Soc.
1987, 109, 3559-3568.
(8) Offord, D. A.; John, C. M.; Linford, M. R.; Griffin, J. H. Langmuir
1994, 10, 883-889 and references therein.
(9) Stranick, S. J.; Parikh, A. N.; Tao, Y.-T.; Allara, D. L. Weiss, P.
S. J. Phys. Chem. 1994, 98, 7636-7646.
Conclusions
The exchange dynamics as well as the desorption and
displacement reactions of redox-active self-assembling mono-
layers of transition-metal complexes of osmium and ruthenium
of the type [M(bpy)2ClL]+ (M ) Os, Ru, bpy ) 2,2′-bipyridine
and L is 1,3-bis(4-pyridyl)propane) have been studied. The
above-mentioned processes appear to be controlled by the rate
of desorption via a dissociative mechanism, although a small
extent of association (less than 10%) is also possible. From
studies of the exchange dynamics, the initial rate of exchange
is estimated to be of the order of 5% of a monolayer per minute.
We have also made use of wave-shape analysis in terms of
shifts in the formal potentials and widths of the voltammetric
waves with changes in the surface coverage and monolayer
composition. The presence of mixed monolayers (Os/Ru)
results in an increased broadening of the voltammetric wave
associated with the ruthenium complex relative to the case of a
monolayer of the ruthenium complex alone which we ascribe
to an enhanced degree of electrostatic repulsion from the
oxidized osmium centers. Also based on an analysis of the
voltammetric response, we believe that the mixed monolayers
are randomly distributed rather than being composed of discrete
and separated phases.
(10) For a summary see ref 8.
(11) (a) Creager, S. E.; Rowe, G. K. J. Electroanal. Chem. 1994, 370,
203-211. (b) Song, S.; Clark, R. A.; Bowden, E. F.; Tarlov, M. J. J. Phys.
Chem. 1993, 97, 6564. (c) Nakashima, N.; Abe, K.; Hirohashi, T.; Hamada,
K.; Kunitake, M.; Manabe, O. Chem. Lett. 1993, 1021.
(12) Allara, D. L.; Nuzzo, R. G. Langmuir 1985, 1, 45-52.
(13) (a) Bain, C. D.; Evall, J.; Whitesides, G. M. J. Am. Chem. Soc.
1989, 111, 7155-7164. (b) Bain, C. D.; Evall, J.; Whitesides, G. M. J.
Am. Chem. Soc. 1989, 111, 7164-7175.
(14) Forster, R. J.; Faulkner, L. R. Anal. Chem. 1995, 67, 1232-1239.
(15) Acevedo, D.; Bretz, R. L.; Tirado, J. D.; Abrun˜a, H. D. Langmuir
1994, 10, 1300-1305.
(16) Tirado, J. D.; Bretz, R. L.; Acevedo, D.; Abrun˜a, H. D. Langmuir
1994, 10, 1971-1979.
(17) Acevedo, D.; Abrun˜a, H. D. J. Phys. Chem. 1991, 95, 9590-9594.
(18) (a) Kober, E. M.; Sullivan, B. P.; Dressick, W. J.; Caspar, J. V.;
Meyer, T. J. J. Am. Chem. Soc. 1980, 102, 7583. (b) Caspar, J. V.; Kober,
E. M.; Sullivan, B. P.; Meyer, T. J. J. Am. Chem. Soc. 1982, 104, 630. (c)
Buckingham, D. A.; Dwyer, F. P.; Goodwin, H. A.; Sargeson, A. M. Aust.
J. Chem. 1964, 17, 325.
(19) Hubbard, A. T.; Ishikawa, R. M.; Katekaru, J. J. Electroanal. Chem.
1978, 86, 271-288.
(20) (a) Wasberg, M. J. Electroanal. Chem. 1994, 379, 541-544. (b)
Untereker, D. F.; Bruckenstein, S. Anal. Chem. 1972, 44, 1009.
(21) Bard, A. J.; Faulkner, L. R. Electrochemical Methods Fundamentals
and Applications; John Wiley & Sons: New York, 1980; p 522.
(22) Laviron, E. J. Electroanal. Chem. Interfacial Electrochem. 1974,
52, 395-402.
Acknowledgment. This work was supported by the National
Science Foundation (DMR 91-07116), and the Office of Naval
Research. J.T. gratefully acknowledges support by a Corning
Foundation Fellowship.
JP952541U