Homogeneous catalysis and selectivity in electrochemistry

Article abstract of DOI:10.1021/om401225r

The relationship between homogeneous catalysis and electrochemistry is examined in light of two examples based on our work concerning (a) catalyst activation, regarding selective electrochemical C-H oxidation with Ru^III/Ru^IV mediation, and (b) catalyst suppression, regarding controlling selectivity in electrochemical aromatic chlorination. The first example (a) deals with the role of catalysis at the working electrode. The electrochemical (EC) oxidation of specific hydrocarbons such as tetralin and indane is performed using tris(acetonitrile)ruthenium trichloride (Ru(CH₃CN)₃Cl₃) as a mediator. The role of this mediator in the oxidation of tetralin has been reported. This homogeneous C-H activation by electron transfer (ET) is accompanied by the redox transitions of the mediator in the course of the catalytic oxidation, and these are the main points of interest here. Additional studies with a rotating ring disk electrode (RRDE) provided a follow-up of creation and recovery of Ru^III/Ru^II and Ru^III/Ru^IV species in the process. Using electrochemistry linked with electrospray ionization mass spectrometry (EC/ESI-MS) gave additional information on the structure of the reduced and oxidized forms of Ru(CH₃CN)₃Cl₃ and the effect of water in the solvent on their lifetimes. The second example (b) of electrochlorination has been reported elsewhere and is brought up as complementary remarks. Aromatic electrophilic chlorination of 1,4-dimethoxy-2-tertbutylbenzene is autocatalyzed and unselective. The EC procedure provides a simple means to inhibit the catalytic runaway reaction. This example shows how the counter electrode affects catalysis and selectivity. (Figure Presented)

Full text of DOI:10.1021/om401225r

Article
pubs.acs.org/Organometallics
Homogeneous Catalysis and Selectivity in Electrochemistry
†
‡
Michael Michman,* Lina Appelbaum, Jenny Gun, Alexander D. Modestov, and Ovadia Lev
The Institute of Chemistry, The Hebrew University, Campus Safra, 91904 Jerusalem, Israel
ABSTRACT: The relationship between homogeneous catalysis
and electrochemistry is examined in light of two examples based
on our work concerning (a) catalyst activation, regarding selective
III
IV
electrochemical C−H oxidation with Ru /Ru mediation, and
(
b) catalyst suppression, regarding controlling selectivity in
electrochemical aromatic chlorination. The first example (a)
deals with the role of catalysis at the working electrode. The
electrochemical (EC) oxidation of specific hydrocarbons such as
tetralin and indane is performed using tris(acetonitrile)ruthenium
trichloride (Ru(CH CN) Cl ) as a mediator. The role of this mediator in the oxidation of tetralin has been reported. This
3
3
3
homogeneous C−H activation by electron transfer (ET) is accompanied by the redox transitions of the mediator in the course of
the catalytic oxidation, and these are the main points of interest here. Additional studies with a rotating ring disk electrode
III
II
III
IV
(
RRDE) provided a follow-up of creation and recovery of Ru /Ru and Ru /Ru species in the process. Using electrochemistry
linked with electrospray ionization mass spectrometry (EC/ESI-MS) gave additional information on the structure of the reduced
and oxidized forms of Ru(CH CN) Cl and the effect of water in the solvent on their lifetimes. The second example (b) of
3
3
3
electrochlorination has been reported elsewhere and is brought up as complementary remarks. Aromatic electrophilic
chlorination of 1,4-dimethoxy-2-tertbutylbenzene is autocatalyzed and unselective. The EC procedure provides a simple means to
inhibit the catalytic runaway reaction. This example shows how the counter electrode affects catalysis and selectivity.
5
INTRODUCTION
mediators and electrogenerated reagents as intermediates.
■
Electrochemical catalysis in solution may therefore be
considered as a “hybrid” case between heterogeneous and
homogeneous catalysis. A critical analysis by Saveant et al.
́
described electrocatalysis and turnover values for several
Homogeneous catalysis in organometallic chemistry deals with
aspects of structural interactions, steric control, ligand
exchange, addition−elimination steps, and (not least) redox
reactions and electron transfer (ET). A recent review covered
these reactions from the perspective of catalytic C−H
systems where the catalyst is a defined solution component
6
1
rather than a modified electrode. Jutand et al. have elucidated
activation. The specialized methods and procedures of
7
the mechanisms for many catalytic cycles using EC techniques.
electrochemistry are largely concerned with the last two
categories. Organometallic homogeneous catalysis in electro-
chemical reactions is aimed at the control of ET and selectivity.
Recent advances were featured in a special issue of Chemical
Our intent is to consider experimental aspects of the
relationship between homogeneous catalysis and electro-
chemistry as they are reflected in two of our works and in
further results. Both examples were chosen as representative
illustrations of this relationship. The first case deals with an
2
Reviews, and the possibilities in this discipline are far from
exhausted. The ET process is anything but simple, as detailed in
a status report by a leading team, yet the attraction of ET as a
electrogenerated ruthenium mediator based on Ru-
3
8
(
CH CN) Cl , for selective C−H activation. The second
3
3
3
clean and manageable method, in comparison with the
employment of hazardous oxidation reagents, is clear. Catalysis
in electrochemistry is often concerned with solid electrode
materials and modification or adsorption of catalysts on the
electrode and, in many respects, emulates heterogeneous
catalysis. Homogeneous catalysis of electrochemical reactions
in solution is intricate, even if the working electrode is
considered as inert. Unlike chemically homogeneous catalysts
that act throughout the entire solution, electrochemical
reactions and primary electron transfers take place at the
electrode surface and are confined within the solution to the
narrow domain of the capacitor-like double layer. The reactions
are further regulated within the Nernst diffusion layer, by
diffusion-controlled concentrations of reactants that are very
case describes selectivity in an aromatic chlorination where
9
inhibition of autocatalysis is affected by the counter electrode.
Catalysis by ruthenium compounds is a considerably large
subject. We studied electrochemistry with RuCl3 in
acetonitrile solutions, where the actual identity of the catalyst
is not always clear. Indeed, Duff and Heath found that RuCl3
10
11
is strongly solvated in acetonitrile to give several complexes,
12
including Ru(CH CN) Cl . Reduction of RuCl in acetoni-
3
3
3
3
−
trile with Zn, for example, yields Ru(CH CN) X (X = BF ,
3
6
2
4
−
13
OTf ). Ru(CH CN) Cl was also selectively obtained by
controlled heating in acetonitrile. Reduction of RuCl in
3
3
3
14
3
Special Issue: Organometallic Electrochemistry
4
different from those in bulk solution. In addition to this
solution regimen, electrochemical reactions often involve
Received: December 23, 2013
Published: July 1, 2014
©
2014 American Chemical Society
4729
dx.doi.org/10.1021/om401225r | Organometallics 2014, 33, 4729−4737

Organometallics
Article
acetonitrile with hydrogen yields a mixture of at least four
ruthenium acetonitrile compounds, among these Ru-
(
CH CN) Cl , Ru(CH CN) Cl , and Ru(CH CN) Cl .
3 4 2 3 3 3 3 2 4
Screening by cyclic voltammetry (CV) proved effective in
following the solvation of RuCl3 in acetonitrile and for
detection of these various complexes and observing the specific
catalytic reactivity of Ru(CH CN) Cl , which Ru(CH CN) Cl
2
3
3
3
3
4
and Ru(CH CN) Cl do not possess.
3
2
4
RESULTS AND DISCUSSION
■
Selective Electrochemical Oxidation with a Ru(III)/
Ru(IV) Catalyst. Tris(acetonitrile)ruthenium trichloride (Ru-
CH CN) Cl ) is in fact a precatalyst. When oxidized, it acts as
(
3
3
3
an electrogenerated mediator for the oxidation of cyclohexene,
methylcyclohexene, indane, tetralin, and intermediates such as
Figure 1. Cyclic voltammetry of 0.1 mM Ru(CH CN) Cl (black),
3
3
3
oxidation of 1 mM indane alone (violet), and catalyzed oxidation of 1
mM (green) and 2 mM (red) indane. The solution background is
shown by a dotted line. Conditions: acetonitrile solution with 1.0 M
1
-tetralol. This selectivity toward these hydrocarbons is due to
their oxidation potentials being only slightly higher than that of
the mediator, as observed with cyclic voltammetry. Preparative
electrolysis of tetralin and indane shows that subsequent
reactions of the oxidized hydrocarbon are consistent with
oxidation at the benzyl position. Tetralin yields 1-tetralol, 1-
tetralone, 1-dehydrotetralin, and 1-acetamidotetralin, whereas
indane gives 1-indanol, 1-indanone, and 1-indene.
−1
TBAP; scan rate 20 mV s ; Pt electrode vs SCE; anode Pt tip ∼1
2
2
mm ; counter electrode Pt plate 1 cm .
II
III
between the Ru and Ru states occurs at ∼0.0−0.1 V, while
III
IV
the conversion between the Ru and Ru states occurs at ∼1.7
V (vs SCE). Both of these redox steps are single-electron
III
IV
Three reaction zones can be defined for Ru(CH CN) Cl , as
3
3
3
transfers. A bias in the shape of the Ru /Ru transition at 1.7
V is noted, especially at high scan rates above 20 mV/s.
The hydrocarbons of interest all have oxidation potentials of
around 1.85 V. The short trace for indane is shown at 1.82−
illustrated in Scheme 1 for the oxidation of tetralin: reversible
Scheme 1
1
.84 V (violet, Figure 1). The oxidation potential for indane is
about 130 mV higher than that for Ru(CH CN) Cl . When
3
3
3
III
IV
indane is present, the Ru /Ru transition shows a catalytic
oxidation current at 1.6 V (1 mM indane (green) and 2 mM
indane (red); Figure 1). The catalytic traces show the expected
sigmoidal shape, Nernstian narrowing of the return curve, and
1
6a
II
the absence of reduction current. At the same time the Ru /
III
Ru step remains unaffected by the presence of indane, and
this step can be regarded as an internal standard.
Electrolysis at the catalysis potential yields 1-indanol, 1-
indanone, and 1-indene, identified by GC-MS using the GC
retention times and the mass spectral details. For prolonged
electrolysis carbon counter electrodes are better than Pt, since
Pt gets passivated. Acetonitrile used for these tests had a water
content of up to 0.03% (17 × 10− M). This is the source of
oxygen in the hydrocarbon products. However, water is
destructive to the catalyst, as shown below, and increasing
the concentration of water or the electrolysis time is
counterproductive.
We have not studied indane further, as behavior analogous to
that observed previously for tetralin electrolysis is evident. All
subsequent experiments described below refer to reactions with
tetralin.
electrogenerated mediation (red), reaction with hydrocarbons
(
black), and loss of the oxidized ruthenium complex through
reaction with water or reduction to the nonreactive Ru-
CH CN) Cl (blue). Different experimental methods are
3
(
3
4
2
III
suitable for inspecting each of these zones. The primary Ru /
IV
Ru step and subsequent electron transfers (ET) can be
followed using fast CV and rotating ring disk electrode (RRDE)
tests. The hydrocarbons and the products of their organic
reactions, which voltammetry cannot detect, have been
established by electrolysis and product studies. Accumulation
of Ru(CH CN) Cl can be monitored by collecting CV data for
3
4
2
the solution during electrolysis. Further insight into both the
ET and the ruthenium complexes involved is given by
combined electrochemistry−electrospray ionization mass spec-
Voltammetry with Rotating Ring Disk Electrode
(RRDE). The catalytic step (red, Scheme 1) was examined
with reference to the oxidation of tetralin.
15
17
troscopy (EC/ESI-MS), which also shows the destruction of
oxidized Ru(CH CN) Cl by water in the course of catalysis.
Unlike CV, RRDE, termed hydrodynamic voltammetry, is a
fast-flow system, conducted under fast movement of the
solution. Fast rotation causes the solution to encounter the
central disk in fast vertical flow and be driven sideways in a
radial course to the ring. The ring response depends on the
collection efficiency of the ring. Potentials of the central disk
and of the surrounding ring are operated independently by a
double potentiostat.
3
3
3
Cyclic Voltammetry (CV) and Electrolysis. Cyclic
voltammetry scans are sensitive to the mediator and the
primary ET step. The reaction of Ru(CH CN) Cl with tetralin
3
3
3
8
has been described previously in detail. Here, Figure 1 shows
CV scans for indane oxidation in the range −0.3 to +1.8 V, with
the baseline of the solution scanned to 2.2 V (dotted line). The
voltammetry trace for Ru(CH CN) Cl alone in solution (black
In the present example the disk electrode was scanned from
−0.05 to 2.5 V to record the current−potential traces shown in
3
3
3
line) shows two reversible redox conversions. The conversion
4
730
dx.doi.org/10.1021/om401225r | Organometallics 2014, 33, 4729−4737

Organometallics
Article
II
III
Figure 1. The ring electrode was kept at 1.5 V in order to
the disk, Ru that reaches the ring is oxidized to Ru on the
II
IV
III
IV
oxidize any Ru or reduce any Ru compounds produced on
the disk. At 1.5 V it is indifferent to tetralin. Experiments were
run with rotation frequencies of 700, 1500, 2000, and 2500
rpm. The results for 700 rpm are shown in Figure 2. The
ring. At 1.75 V Ru that is oxidized to Ru on the disk is
reduced to Ru on the ring. These are reversible single-
electron transfers (eq 1); therefore, equal current intensities are
observed on the ring as well.
III
−
[
Ru(CH CH) Cl ]
3
3
3
−
−
e
Ho oo I Ru(CH CN) Cl
3
3
3
−
−
+
e
e
−
+
Ho oo I [Ru(CH CH) Cl ]
3
3
3
−
+
e
(1)
When the solution contains both Ru(CH CN) Cl and
3
3
3
tetralin, the disk and the ring current intensities at 0.0−0.15 V
remain unchanged, whereas at 1.75 V the disk current surges,
showing a catalytic current shape. Concurrently, the ring
IV
current at 1.75 V falls off. Less Ru reaches the ring because
some is consumed by tetralin. This takes place at ∼250 mV
below the oxidation potential of tetralin itself.
Figure 2. RRDE tests at 700 rpm: (red) 0.1 mM, Ru(CH CN) Cl
3
3
3
The collection efficiency of the ring (ring/disk current
intensities) for this specific RRDE electrode set was determined
by using standard reversible reactions (see the Experimental
Section) and found to be 0.236. It is important to note that for
a single ET process the collection efficiency does not depend
on rotation rate (see p 87 of ref 4).
alone (disk bright red, ring thin dark red); (blue) Ru(CH CN) Cl 0.1
3
3
3
mM + tetralin 0.5 mM. (disk dark blue, ring light blue); (gray) 0.5
mM tetralin alone (oxidized on the disk above 2.1 V; ring current
remains unchanged, a straight line). Conditions: Pt electrodes in
acetonitrile with 1.0 M TBAP; potential sweep of disk from −0.05 to
+
2.5 V; ring potential fixed at 1.5 V (Ag/AgCl/KCl (3 M) reference).
When Ru(CH CN) Cl is tested alone, the collection
3
3
3
II
III
responses of the disk scans from −0.05 to +2.5 V are the traces
starting from the bottom left and increasing toward the right,
whereas the ring responses are the traces starting from the top
left and decreasing toward the right. Disc currents are marked
on the left-side ordinate and ring currents on the right-side
ordinate.
efficiency for the oxidation from Ru to Ru around 0.0 V is
within 0.230 ± 0.005 at all rotation rates, which is consistent
with a reversible single-electron transfer. At 1.75 V, where Ru
III
IV
is recovered on the ring from the oxidized Ru , the collection
efficiency is only 0.210 ± 0.005 at all rotation rates, which
IV
means 90−95% recovery. This small loss of Ru may be due to
−3
Figure 2 shows disk scans at 700 rpm with a constant ring
potential of 1.5 V, with the gray line indicating 0.5 mM tetralin
alone, the red line indicating 0.1 mM Ru(CH CN) Cl alone,
reaction with water (∼17 × 10 M in the solvent) or with
acetonitrile, forming Ru(CH CN) Cl (blue trace in Scheme
3
4
2
1). This loss is slow in comparison with the reaction with
tetralin and is not easily detected by CV analysis but is clearly
observed by EC/ESI-MS, which allows for more time (see
below).
3
3
3
and the blue line indicating 0.1 mM Ru(CH CN) Cl with
3
3
3
added 0.5 mM tetralin. Tetralin is inert until it is oxidized
irreversibly on the disk at potentials above 2 V; as the ring is at
a lower potential, there is no reaction of tetralin at the ring. The
disk scan for the solution of Ru(CH CN) Cl alone shows the
In the presence of tetralin, the collection efficiency around
III/II
0.0 V remains unaffected, 0.230 ± 0.005, and thus the Ru
3
3
3
II
III
low-potential transition (Ru to Ru ) from 0.0 to 0.15 V and
the high-potential transition (Ru to Ru ) at 1.8 V (red lines)
with equal current intensities. With potentials up to 0.15 V on
transition is indifferent to tetralin and can be considered as an
III
IV
internal standard for both the disk and the ring measurements.
III
IV
However, at 1.75 V (Ru recovery from Ru ), the collection
Figure 3. Plot of the ratio i /ω1/2C versus log ω (rad s ) for rotation rates ω of (left to right) 700, 1500, 2000, and 2500 rpm: (white ○)
[
−1
d
Ru(CH CN) Cl ] alone, C = 0.1 mM; (blue ◇) [Ru(CH CN) Cl ] alone, C = 1.0 mM; (red △) [Ru(CH CN) Cl ] 1.0 mM + 0.5 mM tetralin;
3 3 3 3 3 3 3 3 3
(
blue □) [Ru(CH CN) Cl ] 0.1 mM + 1.0 mM tetralin. C = concentration of Ru(CH CN) Cl .
3
3
3
3
3
3
4
731
dx.doi.org/10.1021/om401225r | Organometallics 2014, 33, 4729−4737

Organometallics
Article
Figure 4. EC-ESI-MS study of the reduction of Ru(CH CN) Cl (a) trace of current potential scan in the EC cell, of 1 mM Ru(CH CN) Cl in 50
3
3
3
3
3
3
mM lithium triflate/CH CN from 0.1 to −0.8 V and back; (b) intensities of all ions obtained at 0.1 V; (c) intensities of all ions obtained at −0.8 V.
3
efficiency falls dramatically to 0.136 (0.65 intensity) at 2500
rpm and to 0.11 (0.52 intensity) at 700 rpm (Figure 2). This
current deficiency therefore indicates that some of the oxidized
Ru has been regenerated to Ru by reaction with tetralin
even before reaching the ring electrode.
mass transport to the disk and prolong contact are a critical
factor for turnover frequency (TOF).
Electrochemistry Linked with Electrospray Ionization
Mass Spectrometry, EC/ESI-MS. CV and RRDE tests show
reversible reduction and oxidation of Ru(CH CN) Cl within
IV
III
3
3
3
The following summary of runs at different rotation rates is
illustrative. Tests that were run at different rotation rates gave
the same pattern, but the intensities of the response for catalysis
were smaller at higher rotation rates. The current obtained on
the disk depends not only on concentration but also in a
complex way on factors of mass transport in solution, on
convection, and on viscosity. Altogether, for the single-electron
the time scale of the experiments, implying preservation of the
coordination sphere (eq 1). Whereas reduction of Ru-
(CH CN) Cl is reversible and is not involved in catalysis,
the situation is less clear for the oxidized complex in the course
of catalytic oxidation.
3
3
3
Direct spray ionization MS detects ionized forms of
+
Ru(CH CN) Cl . The cluster [Ru (CH CN) Cl Na] was
3
3
3
2
3
6
6
+
transfers the disk current (i ) is given by the Levich equation:
detected (m/z 683−685, [M + Na] ) along with ions at m/z
42 and 602, consistent with the loss of one and two CH CN
d
6
3
id = 0.62nFD /3ν−1/6
2
ω
1/2C
8
units from this cluster, respectively.
The combined EC/ESI-MS follows the redox cycle of
where ω is given in rad s−1 and ν is the kinematic viscosity and
Ru(CH CN) Cl and can be used to detect the associated
3
3
3
2
−1
D the diffusion coefficient (both in cm s ).
molecular products. A radial thin-layer flow-through cell with
three electrodes is used for the EC step. The product stream is
introduced into the mass spectrometer at atmospheric pressure.
The transfer times from the electrochemical cell to the
ionization port are 1−70 s and are controlled by the flow
rates. This time scale for detection in EC/ESI-MS is long
It is possible to extricate the effect of catalytic current in the
1
/2
following way. The plot of i /ω C against log ω, where C is
d
the concentration of [Ru(CH CN) Cl ], is shown in Figure 3.
3
3
3
1/2
The term i /ω C is the current output in terms of electrons/
d
mole per mass transport. It remains unchanged between 700
−
1
4
and 2500 rpm (73.3−262 rad s ), when Ru(CH CN) Cl
compared with the time scale of RRDE, which is <50 ms.
3
3
3
alone is present, whether C = 0.1 or 1.0 mM, reflecting a single
electron transfer process.
Survival of products or secondary reactions may therefore
become important factors in the interpretation of EC/ESI-MS
data. The observed clusters are compared with calculated MS
clusters determined using the “Sheffield ChemPuter Isotope
When C = 1.0 mM and 0.5 mM tetralin is added (red △,
Figure 3), a slight increase is seen at ω = 700 rpm. The
ruthenium compound is in excess and only a small part of it
interacts with tetralin. However, when tetralin is in excess (C =
18
Pattern Calculator”.
II/III
Reduction of Ru(CH CN) Cl . The Ru
transition (eq 1,
3
3
3
0.1 mM, tetralin 1.0 mM) the value increases at all rotation
left) is used as a test case. The cell potential is scanned from 0.1
rates, and most dramatically at 700 rpm, showing a 3.5-fold
to −0.8 V (vs Ag/AgNO ), and the forward and reverse
3
increase. This reflects the current produced in excess due to
voltammetry traces are shown in Figure 4a. The mass spectra
collected at 0.1 V and at −0.8 V are shown in Figures 4b,c,
respectively. The prominent ion cluster observed at 0.1 V
IV
catalysis: i.e., to recovery and repeated use of the mediator Ru
(red arrows in Scheme 1). Slower rotation rates that slow the
4
732
dx.doi.org/10.1021/om401225r | Organometallics 2014, 33, 4729−4737

Organometallics
Article
Figure 5. (top) Decay and rise of intensity of cluster m/z 661−675 in the course of the potential sweep between 0.1 V and −0.8 V at 1 mV/s.
+
+
Proposed assignments are m/z 669 [Ru(CH CN) Cl ] Li or m/z 668{[Ru(CH CN) Cl ][Ru(CH CN) Cl ]} . (bottom) Intensity of the ion m/z
3
3
3
2
3
3
3
3
4
2
8
80, increase and decay, along the same potential scan between 0.1 V and −0.80 V. The proposed structure for this cluster is
+
[
RuCl (CH CN) Li] Li with attached CF SO Li and CH CN. The insets show the clusters in detail.
3 3 3 2 3 3 3
occurs at m/z 660−670, whereas at −0.8 V, after reduction, the
prominent ion cluster is at m/z 880.
solution were new clusters observed. For these tests,
acetonitrile was dried over molecular sieves and the electrolyte
salts were vacuum-treated before use. In the very dry solvent,
new clusters of ruthenium complexes appeared in the
electrospray when the cell potential was 1.4 V and disappeared
when potentials rose above 1.8 V, with the maximum intensity
obtained under 1.5 V. These clusters correspond to oxidation of
the ruthenium complex. Ion intensities and the corresponding
cluster structure consistent with the observed m/z values are
shown in Figure 6b.
In the course of the potential scans (1 mV/s) between 0.1 V
and −0.8 V and back to 0.1 V, mass spectra were recorded
continuously at a rate of 8−12 spectra per minute, at selected
m/z ranges, to follow the ion output. The cluster at m/z 668−
6
69 that is the dominant species between 0.0 and 1.4 V (vs Ag/
AgNO ) decays below 0.0 V.
3
Figure 5 shows the decay and rise in abundance of m/z 668−
69 that is accompanied by a corresponding rise and decay of
6
the m/z 880 species during the potential scan.
It should be noted that potentials of 1.5 V for these EC/ESI-
MS measurements roughly correspond to 1.8 V (vs Ag/AgCl)
in CV and RRDE experiments. In Figure 6, 1.5 V corresponds
to the range in which the catalytic reaction is the most intense.
The absence of observable oxidized ions in the EC/ESI-MS
measurements in standard acetonitrile (water content ∼0.03%
III
The cluster at m/z 668−669 is assigned as a Ru state. This
cluster can be interpreted as [Ru(CH CN) Cl ][Ru-
3
3
3
+
+
(
(
CH CN) Cl ] (m/z 668) or as [Ru(CH CN) Cl ) Li]
3 4 2 3 3 3 2
+
m/z 669) in analogy to [(Ru(CH CN) Cl ) Na] mentioned
3
3
3 2
above. The ion proposed for the cluster at m/z 880 is
+
II
−3
IV
[
[Ru(CH CN) Cl Li] Li , CH CN, CF SO Li], which is a Ru
H O, 17 × 10 M), implies a reaction of the Ru state with
3
3
3
2
3
3
3
2
IV
state. The insets in Figure 5 show the clusters in detail. These
assignments for the m/z 669 and 880 clusters reflect
preservation of the ligand arrangement of Ru(CH CN) Cl
3
water. In the RRDE tests, the Ru complex formed on the disk
survives transfer to the ring with only partial loss of about 5−
10%. Detection in RRDE experiments occurs on a time scale
that is 3−4 orders of magnitude faster than detection in EC/
ESI-MS, and during the transfer time of the EC/ESI-MS
experiment, reactions with water can completely consume the
oxidized ions (blue route, Scheme 1).
3
3
during reduction; however, other ions such as [Ru-
+
(
CH CN) Cl ][Ru(CH CN) Cl ] (m/z 668) cannot be
3 3 3 3 4 2
ruled out. Indeed, the presence of the compound Ru-
CH CN) Cl in the solution is discussed below.
(
3
4
2
The reduction can be written as in eq 2.
Catalyzed Oxidation of Tetralin with Ru(CH CN) Cl as a
3
3
3
Mediator. Potential sweep tests (Figure 7, top left) for the
−
+
2
[Ru(CH CN) Cl ] + 2e + 3Li
3
3
3
catalyzed oxidation of tetralin, with 0.1 mM Ru(CH CN) Cl
3
3
3
+
and excess tetralin (2 mM), must be run in standard acetonitrile
⇌
[Ru(CH CN) Cl Li] + Li
(2)
3
3
3
2
−3
(
∼0.03% H O, 17 × 10 M) to provide oxygen to tetralin
2
The reduction step appears to be reversible with preservation
of the acetonitrile complex structure.
Oxidation of Ru(CH CN) Cl . Figure 6a traces the intensities
of the cluster at m/z 668−669 and the clusters at m/z 331−346
derived from the monomer [Ru(CH CN) Cl Li] ), in the
course of a potential sweep from 0.8 to 1.8 V and back to 0.8 V,
at 1 mV/s. The intensities of these clusters decrease at the
oxidation potential of Ru(CH CN) Cl . No new cluster
signifying ruthenium oxidation is observed in the high-potential
range: namely, no oxidized form appropriate for the spray
ionization route survived the transfer to the MS analyzer. Only
when the experiment was repeated with a very dry acetonitrile
(black route, Scheme 1). Under these conditions, no oxidized
ruthenium clusters are detected, as explained in the preceding
paragraph. With tetralin the catalytic process consumes the
oxidized clusters even more quickly. However, another tracer is
observed here. In the absence of tetralin, a cluster at m/z 330−
3
3
3
+
(
3 3 3
III
+
340 corresponding to a Ru state, [Ru(CH CN) Cl ] , is
3 4 2
observed in dry as well as in standard acetonitrile at potentials
between 1.4 and 2.0 V (Figure 7, bottom left). In the presence
3
3
3
+
of tetralin [Ru(CH CN) Cl ] is detected, but its intensity is
3
4
2
strikingly deficient within the potential range of the catalytic
reaction (Figure 7, bottom right). Above 1.65 V, outside the
range of the catalytic reaction, where the noncatalyzed
4
733
dx.doi.org/10.1021/om401225r | Organometallics 2014, 33, 4729−4737

Organometallics
Article
Figure 6. Oxidation of Ru(CH CN) Cl in specially dried acetonitrile: (a) ion intensities of the m/z 668.5 cluster (left) and intensities of the
3
3
3
overlapping clusters of m/z 330−350 (right); Ion intensities of oxidized forms and corresponding experimental isotopic patterns. The potential was
+
swept between 0.8 and 1.8 V. (c) m/z 850−864 for {[Ru(CH CN) Cl ][Ru(CH CN) Cl (CF SO )]} + 2CH CN (left), m/z 850−880 for
3
3
3
3
3
2
3
3
3
+
+
{
[Ru(CH CN) Cl ][Ru(CH CN) Cl Li]} + CH CN + CF SO Li (middle), and m/z 560−580 for [Ru(CH CN) Cl ] + 2CH CN + CF SO Li
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
(
right).
+
+
oxidation of tetralin takes over, [Ru(CH CN) Cl ] is again
[Ru(CH CN) Cl ] , hence the declining intensity of [Ru-
3 3 3
3
4
2
+
observed with significant intensity (Figure 7, bottom right).
Higher potentials are destructive to the ruthenium acetonitrile
complexes, and above 2.2 V the MS signals for these complexes
decline sharply (Figure 7, bottom right).
These observations in the course of the catalyzed reaction
with tetralin are remarkable. [Ru(CH CN) Cl ] is possibly
(CH CN) Cl ] in the catalytic range.
3 4 2
EC/ESI-MS was initially intended to follow structural
features of the different oxidation states, yet it also exposes
the competition among water, acetonitrile, and hydrocarbon in
+
the reaction with [Ru(CH CN) Cl ] (blue and black routes,
+
3
3
3
3
4
2
Scheme 1).
formed by nucleophilic attack of CH CN (present in large
3
+
Controlling Selectivity with the Counter Electrode. In
contrast to the notion of catalyst activation on the working
electrode, a case of aromatic electrophilic chlorination shows
that occasionally simply removing the catalyst can increase
selectivity.
excess) on the oxidized and charged [Ru(CH CN) Cl ] rather
3
3
3
than the neutral Ru(CH CN) Cl . Indeed, gradual accumu-
lation of Ru(CH CN) Cl during electrolysis has been
recorded in our earlier work. The much faster reaction of
tetralin in the mediated ET depletes the solution of
3
3
3
3
4
2
4
734
dx.doi.org/10.1021/om401225r | Organometallics 2014, 33, 4729−4737

Organometallics
Article
Figure 7. (top left) CV scan from 1.0 to 2.0 V and back of 0.1 mM Ru(CH CN) Cl in 50 mM CF SO Li in standard CH CN. (bottom left) Ion
3
3
3
3
3
3
intensity of m/z 330−350. (bottom right) Ion intensity of m/z 330−350, during scan from 0.8 to 2.5 V of 0.1 mM Ru(CH CN) Cl in presence of 2
3
3
3
mM tetralin. (top right) Corresponding experimental isotopic pattern, which fits [Ru(CH CN) Cl ].
3
4
2
Figure 8. Reaction profile for chlorination of 2-tert-butyl-1,4-dimethoxybenzene (1): (a) chlorination with Cl ; (b) anodic chlorination. Products: 2-
2
tert-butyl-6-chloro-1,4-dimethoxybenzene (2); 2-tert-butyl-5-chloro-1,4-dimethoxybenzene (3); 2,5-dichloro-1,4-dimethoxybenzene (4); 2-tert-butyl-
3,6-dichloro-1,4-dimethoxybenzene (5). Compounds also formed in small amounts: 3-chloro-1,4-dimethoxybenzene (6); 2,3,5-trichloro-1,4-
dimethoxybenzene (7); 2-tert-butyl-3,5,6-trichloro-1,4-dimethoxybenzene (8). Reproduced with permission from ref 9.
C−H bond activation by halogenation is well known. Indeed,
transition metals have been used to gain selectivity in
electrochemical chlorination or bromination, as in the
an undivided cell, protons are reduced on the cathode and the
solution pH is kept at ∼7. With the catalyst removed from the
9
solution, selectivity is obtained. A detailed report is not
19
chlorination of benzoquinoline aided by PdCl2. In this
reiterated here, but in fact this electrochemical procedure in an
undivided cell yielded just two monochlorinated products and
the tert-butyl group is preserved, even at 80% completion of the
reaction (Figure 8). The possibility of such simple comple-
mentary control using a counter electrode is an exclusive asset
of the electrochemical method.
reaction PdCl does not act as a mediator but as a directing or
2
protecting group in a proper reactive manifold that is frequently
20
observed with palladium compounds.
,4-Dimethoxy-2-tert-butylbenzene is very reactive toward
1
electrophilic aromatic substitution, in this case chlorination.
The reaction with chlorine is autocatalyzed by the released
protons, and the reaction accelerates as the pH falls
significantly. Seven products are obtained from the chlorination
reaction, some of which are chlorinated with greater ease than
the starting material and others correspond to loss of the tert-
butyl group.
Electrochemistry plays a double role in this chlorination
reaction. The first role is the production of chlorine in situ from
harmless materials in a controlled manner, instead of using free
chlorine with the demanding technicalities and hazards
involved. The second role is that of the counter electrode; in
CONCLUSION
■
The two examples illustrate various parameters concerning
links of electrochemistry with homogeneous catalysis. The parts
played by the working electrode and the counter electrode, as
concerns catalysis, are illustrated with two very different cases.
Ru(CH CN) Cl is in a sense an artless example, being
3 3 3
molded in solution of RuCl3 in CH CN with no other
3
especially tailored ligand attached. RuCl is known to catalyze
3
oxidations similar to those of cyclohexene and cyclohexanol,
4
735
dx.doi.org/10.1021/om401225r | Organometallics 2014, 33, 4729−4737

Organometallics
Article
when the transition metal is supported by strong oxidizers such
was electrolyzed at a potential halfway up the catalytic current, at 1.9−
2
1
22
2
.1 V, ∼1.2 mA, 2 h (∼8.5 C) in an undivided cell, with a Pt-wire
as NaOCl and NaIO4 possibly involving RuO /RuO4
2
anode and a graphite cathode (vs Ag/AgCl in 3 M KCl). A Pt cathode
could not be used due to passivation.
GC-MS identified products (retention time/min (m/z), com-
pound): 8:06 (118), indane; 8:18 (116), indene; 11:25 (134), 1-
indanol (m/z (intensity) 134 (18), 133 (100), 115 (46), 105 (16), 91
intermediates, in nonpolar solvent or in phase transfer.
Replacing use of such reagents by electrode activation enables
identification of active ruthenium species in solution and
follow-up of the catalyzed oxidation, as delineated in Scheme 1.
Several techniques such as CV, constant-potential electrolysis,
RRDE, and ECSI-MS provide different perspectives of the time
scale and material transport regime, oxidation states, and
background reactions. That reduced and oxidized ruthenium
complexes retain their basic coordination sphere is probably
also due to the solvent and main ligand being the same.
Evidence for RuO /RuO is not observed in this system.
(16); 12:00 (132)), 1-indanone (m/z (intensity) 132 (100), 104 (60),
103 (48), 102 (9), 78 (95), 77 (25), 50 (20)).
A peak at 15:77 (167) for N-cyclohexylpyrrolidone was found from
the processed graphite cathode. It was absent when the cathode was
replaced.
RRDE was carried out under Ar with a double potentiostat using a
Pt electrode with disk radius 4.57 mm (r ), ring internal radius 4.927
1
2
4
mm (r ), and outer radius 5.38 mm (r ). The disk−ring distance was
2
3
The intimate details of contact between the ruthenium
mediator and hydrocarbon are not known. Possibly, the
aromatic HOMO level of tetralin or indane can act as a
0
.357 mm, and the ring width was 0.45 mm.
The collection efficiency n of the electrode (n = −i /i (i sign
r
d
r
opposite from i )) was determined for a 5 mM solution of 1,4-
d
+
nucleophile toward [Ru(CH CN) Cl ] . A consequent tetralin
dimethoxy-2,5-ditertbutylbenzene, which is a reversible oxidation, and
was found to be 0.236 ± 0.003 for rotation rates in the range of 700−
2500 rpm.
3
3
3
(
or indane) cation radical is expected to be located on the
aromatic ring, but the radical, after loss of proton, is positioned
RRDE tests were run at 700, 1500, 2000, and 2500 rpm. The ratio
of ring current to disk current is constant at 0.235 ± 0.005 for these
rotation rates and is close to the collection efficiency with the standard.
Electrospray ionization mass spectrometry was carried out on a
quadrupole ion trap mass spectrometer. The details of the online
concentric coaxial EC cell and experimental setup have been described
at the benzyl carbon, as are the final products. Such C−H
IV
activation is in a way indirect if oxidized Ru does not attack
the actual CH bond that is eventually activated but initially
triggers the adjacent aromatic segment or the double bond in
cyclohexene.
In contrast, the case of chlorination focuses on the role of the
counter electrode. Whereas the working electrode activates the
chlorination process, the counter electrode suppresses the
secondary effect of autocatalysis. This double operation is a
special feature of electrochemistry.
2
3
elsewhere. Briefly, the working electrode is a Pt disk of 1.6 mm
diameter placed at the bottom of the cell. Compartments with counter
and reference electrodes were separated from the main flow of
electrolyte by tube filters. The spray ionization was operated in the
positive ion mode with a spray voltage of 2.0 kV (once at 3 kV), a
capillary voltage of 25 V, and a source temperature of 120 °C. A
syringe pump drove the solution through the outer tube to the
electrode surface. The analyte jet contacts the surface of the working
electrode through the outer tube of the double coaxial system, and the
products of electrode reactions were transferred to the MS through the
inner tube. The inner tube had an inner diameter of 125 μm and an
outer diameter of 1.6 mm and was 15 cm long. The reagent flow rate
EXPERIMENTAL SECTION
■
Materials. Solvents were degassed, and all procedures were carried
out under Ar. Commercially dry acetonitrile used routinely contained
−3
up to ∼0.03% (17 × 10 M) water and was used as received unless
additional drying was specified. This drying was carried out by
prolonged storage over molecular sieves A. Lithium trifluoromethane-
sulfonate (LiOTf) was used as supplied. RuCl (CH CN) (1) was
−1
was set to 3 μL min . With the help of auxiliary acetonitrile, the
sample stream was delivered to the electrospray interface. The
acetonitrile flow in the auxiliary tubing was 30 μL min− for potential
sweep experiments.
3
3
3
1
prepared as in ref 8.
Instrumentation and Methods. The electrode potential was
controlled with a potentiostat/galvanostat. Cyclic voltammetry was
performed under Ar in an undivided three-electrode cell with a Pt
Mass spectra were obtained by scanning the mass analyzer from m/
z 50 to 2000 with five total micro scans. The maximum injection time
into the ion trap was 50 ms. The analyzer was operated at a
2
2
working electrode (1.7 mm ) and a Pt counter electrode (0.5 cm ) vs
−5
background pressure of 2 × 10 Torr.
Ag/AgBF (0.1 M in acetonitrile). Constant-potential electrolysis was
4
carried out under Ar in an undivided three-electrode cell with a Pt
working electrode (2 cm ) and a counter electrode of Pt or graphite vs
Ag/AgCl (3 M in KCl). Reported potentials are related to the SCE for
CV (Figure 1) and to Ag/AgCl (3 M in KCl) for electrolysis and
RRDE tests (Figure 2). Potentials for the EC/ESI-MS experiments are
given relative to a Ag/AgNO3 (0.1 M in acetonitrile) reference
2
AUTHOR INFORMATION
Corresponding Author
M.M.: e-mail, michman@mail.huji.ac.il; fax, 972 (0)35245559;
tel, 972 (0)547482388.
Present Addresses
Makhteshim Chemical Works Ltd., Saadia Mallal 3, 84100
Beer Sheva, Israel. E-mail: linaapelbaum@gmail.com; Lina.
Apelbaum@adama.com.
Russian Academy of Sciences, A. N. Frumkin Institute of
■
*
electrode. Potential values of −0.25 and 1.45 V (vs Ag/AgNO )
3
†
correspond to 0.14 and 1.85 V (vs Ag/AgCl). The electrolytes were
tetrabutylammonium perchlorate (TBAP) for CV and RRDE, LiClO4
for electrolysis, and LiOTf for EC/ESI-MS.
Catalyzed electrolysis of indane was done under Ar in acetonitrile
‡
with LiClO as the electrolyte. To the resulting solution was added
4
Physical Chemistry and Electrochemistry, Leninski Prospect 31,
19991 Moscow, Russia. E-mail: modestov@elchem.ac.ru.
water, followed by extraction with CDCl . This same procedure was
3
1
followed for electrolyzed solutions and for control solutions. Analysis
was by GC-MS (injector 250 °C, split 1:50, temperature gradient 60
Notes
°
C/min for 3 min, then 10 °C/min to 330 °C). The column was fused
The authors declare no competing financial interest.
silica (Rxi-5Sil MS) 30 m × 0.25 mm. First mass was at m/z 40. The
solution after electrolysis was compared with the control, as well as
with 1-indanol and 1-indanone.
ACKNOWLEDGMENTS
■
This paper is based on a presentation in Molecular Electro-
chemistry in Organometallic Science, 46th Heyrovsky Dis-
cussion, Trest, Czech Republic, August 2013. We thank Dr.
Omer Yehezkeli (Vilner’s group) and Dr. Liang Liu (Mandler’s
A sample electrolysis is as follows. A 20 mL CH CN solution was
3
prepared of 2.1 mg (0.003 mM) of Ru(CH CN) Cl , 40 mg (0.34
3
3
3
mM) of indane, and 0.2 mL of H O, doped with benzonitrile as a
2
marker. A 10 mL portion was kept as the control solution, and 10 mL
4
736
dx.doi.org/10.1021/om401225r | Organometallics 2014, 33, 4729−4737

Organometallics
Article
group) of the Hebrew University for help in experiments with
indane. We thank Michael Mogilnitsky and Dr. Natalia
Movshovich, from Makhteshim Chemical Works, for the GC-
MS analysis.
REFERENCES
1) Shul’pin, G. B. Dalton Trans. 2013, 42, 12794.
2) Molecular and Biomolecular Electrochemistry, J. M. Saveant,
́
Guest Ed.: Chem. Rev., 2008, 108, 2111−2391.
3) For a status report, see: Bard, A.; Abruna, H. D.; Chidsey, C. E.;
Faulkner, L. R.; Feldberg, S. W.; Itaya, K.; Majda, M.; Melroy, O.;
Murray, R. W.; Porter, M. D.; Soriaga, M. P.; White, H. S. J. Phys.
Chem. 1993, 97, 7147.
■
(
(
(
̃
(
4) Gileadi, E. Electrode Kinetics for Chemists, Chemical Engineering
and Materials Scientists; VCH: Weinheim, Germany, 1993; pp 82−92.
5) Simonet, J. Organic Electrochemistry, Baizer, M. M., .Lund, H.,
Eds.; Marcel Dekker: New York, 1983; p 843.
6) (a) Saveant, J. M. Chem. Rev. 2008, 108, 2348. (b) Costentin, C.;
Drouet, S.; Robert, M.; Saveant, J. M. J. Am. Chem. Soc. 2012, 134,
1235.
(
(
́
́
1
(
7) Jutand, A. Chem. Rev. 2008, 108, 2300.
(
8) Appelbaum, L.; Heinrichs, C.; Dehmtschuk, J.; Michman, M.;
Oron, M.; Schaefer, H. J.; Schumann, H. J. Organomet. Chem. 1999,
92/2, 240.
9) Appelbaum, L.; Danovich, D.; Lazanes, G.; Michman, M.; Oron,
M. J. Electroanal. Chem. 2001, 499, 39.
10) (a) Griffith, W. P. Ruthenium Oxidation Complexes: Their Uses as
5
(
(
Homogenous Organic Catalysts; Bianchini, C., Cole-Hamilton, D. J., van
Leeuven, P. W. N. M., Eds.; Springer: New York. 2011; Catalysis by
Metal Complexes 34. (b) Ruthenium Catalysts and Fine Chemistry;
Bruneau, C., Dixneuf. P. H., Eds.; Springer: Heidelberg, Germany,
2004; Topics in Organometallic Chemistry11.
(
11) Chocron, S.; Michman, M. J. Mol. Catal. 1991, 66, 85.
(
12) (a) Duff, C. M.; Heath, G. A. Inorg. Chem. 1991, 30, 2528.
(
b) Duff, C. M.; Heath, G. A. J. Chem. Soc., Dalton Trans. 1991, 2401.
13) Uderwood, C. C.; Stadelman, B. S.; Sleeper, M. L.; Brumaghim,
J. L. Inorg. Chim. Acta 2013, 405, 470.
14) Deglmann, P.; Ember, E.; Hofmann, P.; Pitter, S.; Walter, O.
Chem. Eur. J. 2007, 13, 2864.
15) Gun, J.; Bharathi, S.; Gutkin, V.; Rizkov, D.; Voloshenko, A.;
(
(
(
Shelkov, R.; Sladkevich, S.; Kyi, N.; Rona, M.; Wolanov, Y.; Koch, M.;
Mizrachi, S.; Pridkhochenko, P. V.; Modestov, A. D.; Lev, O. Isr. J.
Chem. 2010, 50, 360.
(
16) (a) Greef, R.; Peat, R.; Peter, L. M.; Pletcher, D.; Robinson, J. In
Instrumental Methods in Electrochemistry; Kemp, T. J., Ed.; Ellis
Horwood: Chichester, U.K., 1985; Southampton Electrochemistry
Group, Ellis Horwood Series in Physical Chemistry. (b) Reference
1
6a, p 198. (c) Reference 16a, p 136.
17) Bard, A. J.; Faulkner, L. R. Electrochemical Methods:
Fundamentals and Applications, 2nd ed..; Wiley: New York, 2001; p
39.
18) http://www.chem.shef.ac.uk/Web.Elements. Cross checks were
(
3
(
also made with Isotope Distribution Calculator, Scientific Instruments
Services, www.sisweb.com.
(
19) Kakiuchi, F.; Kochi, T.; Mutsutani, H.; Kobayashi, N.; Urano, S.;
Sato, M.; Nishiyama, S.; Tanabe, T. J. Am. Chem. Soc. 2009, 131,
1310.
1
(
20) Lyons, T. W.; Sanford, M. S. Chem. Rev. 2010, 110, 1147.
(
21) Wolfe, S.; Hasan, S. K.; Campbell, J. K. J. Chem. Soc. D 1970,
1
(
420.
22) Shing, T. K. M.; Tai, V. W.-F.; Tam, E. K. W. Angew. Chem., Int.
Ed. 1994, 33, 2312.
23) (a) Modestov, A. D.; Gun, J.; Mudrov, A.; Lev, O. Electroanalysis
004, 16, 367. (b) Modestov, A. D.; Gun, J.; Savotin, I.; Lev, O. J.
Electronal. Chem. 2004, 567, 7.
(
2
4
737
dx.doi.org/10.1021/om401225r | Organometallics 2014, 33, 4729−4737