Cyclovoltammetric studies on the dimerization of methyl acrylate by ruthenium complexes

Article abstract of DOI:10.1016/S0022-328X(98)00532-4

Electrochemical measurements are described on transition metal complexes which are or might be catalyst precursors for the tail-to-tail dimerization of methyl acrylate (MA). Thus, RuH₂(PPh₃)₄ and RuHCl(CO)(PPh₃)₃ are studied electrochemically by cyclic voltammetry (CV) in various solvents including neat MA. It is shown that the activation of the catalyst precursor RuH₂(PPh₃)₄ with trifluoromethanesulfonic acid and of RuHCl(CO)(PPh₃)₃ with triethyloxonium tetrafluoroborate generates new complexes that are easy to reduce and rather difficult to oxidize. The electrochemical measurements provide information on the deactivation of the catalyst. CpRuH(PPH₃)₂, which can easily be oxidized but can not be reduced, even at -2.5 V versus SCE, does not act as a catalyst precursor neither in pure form nor in the presence of activating compounds. The ease of reduction or oxidation seems to be connected to a possible catalytic activity. It is, therefore, postulated that cyclic voltammetry can be a useful tool in the study of homogeneous transition metal catalysis.

Full text of DOI:10.1016/S0022-328X(98)00532-4

Journal of Organometallic Chemistry 561 (1998) 181–189
Cyclovoltammetric studies on the dimerization of methyl acrylate by
ruthenium complexes
Claudia Strehblow, Thomas Schupp, Reiner Sustmann *
Institut fu¨r Organische Chemie der Uni6ersita¨t Essen, D-45117 Essen, Germany
Received 20 January 1998
Abstract
Electrochemical measurements are described on transition metal complexes which are or might be catalyst precursors for the
tail-to-tail dimerization of methyl acrylate (MA). Thus, RuH2(PPh3)4 and RuHCl(CO)(PPh3)3 are studied electrochemically by
cyclic voltammetry (CV) in various solvents including neat MA. It is shown that the activation of the catalyst precursor
RuH2(PPh3)4 with trifluoromethanesulfonic acid and of RuHCl(CO)(PPh3)3 with triethyloxonium tetrafluoroborate generates new
complexes that are easy to reduce and rather difficult to oxidize. The electrochemical measurements provide information on the
deactivation of the catalyst. CpRuH(PPH3)2, which can easily be oxidized but can not be reduced, even at −2.5 V versus SCE,
does not act as a catalyst precursor neither in pure form nor in the presence of activating compounds. The ease of reduction or
oxidation seems to be connected to a possible catalytic activity. It is, therefore, postulated that cyclic voltammetry can be a useful
tool in the study of homogeneous transition metal catalysis. © 1998 Elsevier Science S.A. All rights reserved.
Keywords: Ruthenium; Cyclovoltammetry; Dimerization
1. Introduction
tetrafluoroborate at 0°C. An essential feature of this
process is assumed to be a metal–hydride bond into
which MA is inserted. The most effective catalytic
system so far, (p5-1,2,3-trimethylindenyl)-bis(ethene)-
rhodium in the presence of a proton source, was devel-
oped by Brookhart [3]. Prior to our own studies on the
dimerization of MA [4–6] with ruthenium-based cata-
lysts, this metal had been used by McKinney for this
purpose [7]. The catalyst was generated in situ by
reduction of ruthenium trichloride with zinc. Mechanis-
tic studies were not reported.
Amatore et al. [8] have shown that the combination
of electrochemical techniques with NMR spectroscopic
methods can lead to information about individual ele-
mentary steps in transition metal-catalyzed reactions.
Prior to the present studies we reported that the dihy-
Methyl acrylate (MA) can be dimerized by a number
of transition metal complexes as catalyst precursors,
e.g. palladium [1], nickel [2], rhodium [3,4] and ruthe-
nium [5,6], yielding mainly the linear tail-to-tail dimer
trans-D2-hexene-1,6-dioicacid dimethylester (HDM).
Several investigations on the mechanism of the dimer-
ization have been published. Oehme ([1]a) studied the
mechanism of the palladium-catalyzed dimerization of
MA by deuterium labelling and by the analysis of the
thermal decomposition products of model complexes. A
key step in the proposed catalytic cycle is an oxidative
addition of a y-coordinated MA to a hydrido-(i-
methoxycarbonyl)vinyl palladium complex. The latter is
supposed to insert MA into the metal–hydride bond.
Wilke [2] studied the catalytic dimerization of MA in
the presence of p3-allyltrimethylphosphane-nickel
drido
complex
RuH2(PPh3)4
(1)
produces
Ru(0)(MA)2(PPh3)2 (2) on dissolving in MA. Complex
2 was isolated and the structure of a water adduct was
determined by X-ray crystallography [9]. Electrochemi-
* Corresponding author. Tel.: +49 201 1833097.
0022-328X/98/$19.00 © 1998 Elsevier Science S.A. All rights reserved.
PII S0022-328X(98)00532-4

182
C. Strehblow et al. / Journal of Organometallic Chemistry 561 (1998) 181–189
cal studies on this reaction reported in the preceding
contribution [10] revealed that MA undergoes oxidative
addition to Ru(0) via cleavage of a vinylic CꢀH bond,
a fact that could not be taken into account in the
original formulation of the catalytic cycle of the dimer-
ization of MA ([5]a) as it became evident only by the
electrochemical measurements.
Here we describe cyclovoltammetric experiments on
the dimerization of methyl acrylate using ruthenium-
based catalytic systems developed in our laboratory
[4–6]. According to these processes, MA is dimerized to
miscellaneous C-6 units yielding trans-D2-HDM as a
main product (90%) in the presence of RuH2(PPh3)4 and
CF3SO3H as activating acid ([5]a). Turn over numbers
(TON) range from 200 to 600 at ca. 80°C. By CV
experiments we investigate the role of the activating acid
CF3SO3H, which is necessary to initiate the catalytic
cycle. Electrochemical studies are reported for a second
catalytic system, RuHCl(CO)(PPh3)3 (3)/Et3OBF4
(TON: 50–150), where the oxonium salt abstracts a
chloride ion from the complex yielding a free coordina-
tion site. The influence of the oxonium salt on the
electrochemical behavior of the catalyst precursor bear-
ing chloride as a ligand is studied. Electrochemical
electrodes (0.5 mm–10 mm diameter). In this case, a
five-necked flask was used as the electrochemical cell,
which contained 5 ml of solvent. All potentials were
measured against
a
saturated calomel electrode
(Metrohm) using ferrocene as an internal standard [13].
The concentration of the supporting electrolyte tetra-
butylammonium hexafluorophosphate (Merck) was 0.1
M. The concentration of the complex was 1×10−3
M
if not indicated otherwise. The reported potentials are
taken from measurements with a scan rate of 100
mV s−1 if not indicated otherwise.
2.3. Materials
All solvents were thoroughly degassed, dried over
sodium (toluene, methanol and n-hexane), over CaH2
(MA), and distilled. Tetrabutylammonium hex-
afluorophosphate used as the supporting electrolyte in
the cyclovoltammetric studies was obtained from
Merck, used as-supplied by the manufacturer, and kept
under dry argon.
2.4. Preparation of metal complexes
studies
are
further
carried
out
on
Ru(0)(MA)2(PPh3)2 [9], RuH2(PPh3)4 [14], RuH-
Cl(CO)(PPh3)3 [15], Ru(CH2CH2CO2Me)(CO)Cl(PPh3)2
[11], and CpRuH(PPh3)2 [16] were prepared according
to literature procedures.
Ru(CH2CH2CO2Me)(CO)Cl(PPh3)2 (4), the complex
which is the insertion product of MA into the ruthe-
niumꢀhydride bond of 3 [11] and which can be activated
by Et3OBF4, to produce dimers of MA [6]. The catalyt-
ically inactive complex CpRuH(PPh3)2 (5) [6] is included
in our study in order to correlate its electrochemical
behavior with its catalytic potential with respect to the
dimerization of MA.
2.5. Dimerization of MA
All experiments were carried out four times in order
to secure reproducibility. Products were identified by
comparison with authentic samples. 4-HOC6H4OCH3
was added as inhibitor to prevent free radical polymer-
ization. Details on the dimerization of MA in the
presence of RuH2(PPh3)4/CF3SO3H have been described
before [5].
2. Experimental
2.1. Instrumentation
1H-, 1H{31P}-, 31P- and 13C-NMR spectra were
recorded on a Bruker AMX 300 spectrometer. 31P-
NMR spectra chemical shifts were referenced against
external 85% H3PO4. IR spectra were recorded on a
Perkin-Elmer FT-IR 1600.
2.5.1. RuHCl(CO)(PPh3)3/four eq6i6alents Et3OBF4
A solution of 39.9 mg (0.21 mmol) Et3OBF4, 47.6 mg
(0.05 mmol) 3 and 3.3 mg (0.03 mmol) 4-HOC6H4OCH3
in 1.93 g (22.4 mmol) MA was heated to 80°C for 48 h.
GLC analysis: 85.2% trans-D2-HDM, 4.3% trans-D3-
HDM, 1.9% cis-D3-HDM, 1.8% cis-D2-HDM, 5.3%
4-methyl-D2-pentene-1,5-dioicacid dimethylester, 1.8%
cis-D2-hexene-1,6-dioicacid dimethylester, and 0.3% mu-
conicacid dimethylester. TON: 46.
2.2. Electrochemical experiments
Cyclic voltammetry was performed on a modified
polarograph E 310 (Bruker). Data aquisition was real-
ized by a PC with the software package DigiS (GfS,
Aachen) [12]. A platinum button was used as the
working electrode, a platinum foil as auxiliary electrode.
Apart from three electrodes, the electrochemical cell
(Metrohm) was equipped with an argon inlet tube, a
bypass tube and was filled with 15 ml solvent. CVs with
scan rates \1 V s−1 were obtained with platinum disc
2.5.2. Ru(CH2CH2CO2Me)(CO)Cl(PPh3)2/three
equi6alents Et3OBF4
A solution of 30.6 mg (0.16 mmol) Et3OBF4, 40.1 mg
(0.05 mmol) 4, and 4.0 mg (0.03 mmol) 4-
HOC6H4OCH3 in 1.93 9 (22.4 mmol) MA was heated to
90°C for 24 h. GLC analysis: 81.4% trans-D2-HDM,
3.7% trans-D3-HDM, 2.2% cis-D3-HDM, 3.1% cis-D2-

C. Strehblow et al. / Journal of Organometallic Chemistry 561 (1998) 181–189
183
HDM, 5.9% D2-4-methyl-pentene-1,5-dioicacid dim-
3. Results
ethylester, 3.0% 2-methylene-pentanedioicacid dimethyl-
ester, 0.7% muconicacid dimethylester. TON: 90.
3.1. CV studies on RuH2(PPh3)4 (1) in the presence of
CF3SO3H
2.5.3. TONs as a function of acti6ation after
predetermined time inter6als
In the preceding contribution, the electrochemical
behavior of 1 in MA was studied. It was found that MA
undergoes an oxidative addition to 1, producing a
ruthenium–hydride complex (6) which further inserts a
molecule of MA into the ruthenium–hydride bond (7).
Fig. 1 shows the CV of a solution of RuH2(PPh3)4 (1)
(0.01 M) in MA 1 min after dissolving the complex (trace
1). The first oxidation process (O1) at 0.85 V had been
assigned to the oxidation of 6, O2 as due to the oxidation
of 7, and OTPP as corresponding to the oxidation of two
equivalents of triphenylphosphane liberated on dissolv-
ing 1 [10]. Trace 2 in Fig. 1(a) shows the CV after the
addition of 3.2 equivalents CF3SO3H to the solution that
produced trace 1, conditions which had proved to be
optimal for the dimerization of MA. The sulfonic acid
itself displays no electrochemical activity in the accessi-
ble potential range of −2.5 to +2.0 V versus SCE.
Thus, the redox activities in the voltammogram of the
activated catalyst precursor must result from ruthenium
complexes or products derived from them. After the
addition of the activating acid, the current density of O1
decreases. The oxidation of free triphenylphosphane
OTPP can no longer be observed. This is in agreement
with the fact that triphenylphosphane adds to MA under
these conditions and produces a phosphonium salt with
CF3SO3H that had been isolated and characterized
([5]a). The latter is not electroactive within the potential
window of Fig. 1. The oxidation peak O1 disappears
completely within a few minutes after the activation in
favor of the new peak O3. The current density of O3, i.e.
the concentration of the new species, then corresponds
to the initial current density of O1 showing complete
conversion of 6 and 7 to a new complex. Scan rate-de-
pendent measurements reveal that O2 is irreversible up
to 10 V s−1. O3 is irreversible at 100 mV s−1, it becomes
quasi-reversible at scan rates \100 mV s−1, and re-
Solutions of the catalyst precursor
1 and 4-
HOC6H4OCH3 as inhibitor in MA were prepared ac-
cording to the following example and activated by
CF3SO3H after the indicated time intervals. TONs for
the dimers of methylacrylate were determined by GLC.
Example: 0.12 mmol 1 and 0.14 mmol 4-HOC6H4OCH3
were dissolved in 112 mmol (933 equivalents) of MA.
After 1 h, 0.384 mmol (3.2 equivalents) of CF3SO3H
were added and the solution was heated to 80°C for 72
h. All operations were carried out in argon atmosphere.
The following TONs, given in brackets, were obtained
when the solutions were activated after (h): 0.5 (406), 1.0
(418), 3.0 (402), 5.0 (411), 7.0 (409), 24 (401), 72 (411).
2.5.4. Experiments with CpRuH(PPh3)2 (5)
Complex 5 (0.05 mmol) and 2–5 mg 4-HOC6H4OCH3
as inhibitor were dissolved in 2.0 ml MA and heated to
90°C. No dimers had been formed after several hours.
A similar experiment in the presence of four equivalents
CF3SO3H provided identical results.
A solution of 0.074 mmol of 5 in 1.0 ml THF showed
the RuꢀH vibration in the IR at 1950 cm−1. MA (1.10
mmol) was added and the spectrum again recorded after
24 h. The intensity of the RuꢀH vibration had not
changed. The CO-vibration (1734 cm−1), however, ap-
peared in emission when the spectrum of a solution of
1.10 mmol MA in 1.0 ml THF was subtracted. This
indicates coordination of MA to ruthenium. At the same
time 0.07 mmol (0.95 equivalents) of free triphenylphos-
phane could be detected in solution by GLC. Electro-
chemical studies revealed that one equivalent of
triphenylphosphane dissociates from CpRuH(PPh3)2 on
dissolving in MA. This was determined from the current
density of the oxidation of the complex compared with
that of liberated triphenylphosphane. An exchange of
one phosphane ligand against one molecule of MA has
taken place.
versible at scan rates \500 mV s−1
.
In the reductive potential range of the activated
solution (Fig. 1(b)) two barely separated reduction pro-

184
C. Strehblow et al. / Journal of Organometallic Chemistry 561 (1998) 181–189
cesses appear at ca. −0.5 V with a total current density
that corresponds to the initial current density of O1.
Scan rate-dependent measurements at a scan rate of
\1 V s−1 reveal only one reduction process that is
quasi-reversible up to 10 V s−1. The reversibility in-
creases with higher scan rates. We interpret the reduc-
tion as an electron transfer with a chemical follow up
reaction.
In a further experiment, the activation of 1 (c=0.001
M) in MA was carried out with only 1.5 equivalents of
CF3SO3H (Fig. 2). Under these conditions complexes 2
(O0) and 6 (O1) can be identified besides the activated
complex (O3). Apparently this amount of acid does not
completely convert complexes 2 and 6 to the species
oxidized at 1.2 V and reduced at −0.5 V. As the
voltammogram was not recorded up to the potential
where 7 is oxidized its possible presence can not be seen
Fig. 2. CV of RuH2(PPh3)4 (c=0.001 M) in MA in the presence of
1.5 equivalents of CF3SO3H.
in the voltammogram. Due to the smaller absolute
concentration of the catalyst precursor as compared
with the conditions where Fig. 1 had been recorded the
oxidation peaks O0 and O1 can be separated. O0 disap-
pears in favor of O1.
The catalytic dimerization of MA in the presence of
RuH2(PPh3)4/CF3SO3H depends on temperature.
Highest conversions are reached at ca. 80°C. Fig. 3
displays temperature-dependent voltammograms of the
system RuH2(PPh3)4 (0.01 M)/3.2 equivalents CF3SO3H
in MA. These were recorded at a scan rate of 1.0 V s−1
because high temperature causes bubbles at the elec-
trode that lead to CVs of bad quality at a scan rate of
100 mV. The CV recorded at 20°C shows the oxidation
of the ruthenium complex that is formed after the
activation with CF3SO3H (O3) and that of the product
of insertion of MA into the RuꢀH bond of 3, namely 7
Fig. 1. CV of RuH2(PPh3)4 in MA, 1 min after dissolving the
complex (trace 1) and after addition of 3.2 equivalents of CF3SO3H
(trace 2) (a); reductive potential range (b).
Fig. 3. Temperature-dependent CVs of RuH2(PPh3)4 in MA in the
presence of 3.2 equivalents of CF3SO3H.

C. Strehblow et al. / Journal of Organometallic Chemistry 561 (1998) 181–189
185
Fig. 4. CV of RuH2(PPh3)4 in MA in the presence of 3.2 equivalents
of CF3SO3H after heating to 80°C for 18 h.
Fig. 5. CV of RuH2(PPh3)4 in MA after a reaction time of 12 h (trace
1) and after subsequent additions of 3.2 equivalents of CF3SO3H.
(O2) with a low current density. The process O2 is
reversible at a scan rate of 1.0 V s−1. With increasing
temperature the potential of the saturated calomel elec-
trode shifts to lower values and the current density
increases, which is in accordance with general electro-
chemical laws [17] and is not due to a change of the
concentration of the activated catalyst precursor. The
voltammogram obtained at 80°C differs from the CV
obtained at lower temperatures. The oxidation of 7 can
no longer be seen.
a wave for complex 7 (trace 1 in Fig. 5). If this solution
is activated with 3.2 equivalents of CF3SO3H, 7 disap-
pears immediately in favor of the activated complex
(trace 2). It is therefore postulated, that the main
pathway of generating a catalytically active species
proceeds via complex 7, which in turn is formed from 1
via the intermediate complexes 2 and 6.
3.2. CV studies on the system BuH2(PPh3)4/CF3SO3H
Shortly after heating the activated complex to 80°C,
the solution was again cooled to r.t. At a scan rate of
1.0 V s−1 only the activated complex is identified in a
quasi-reversible CV. If this solution is again heated to
80°C, kept at this tempeature for 18 h and cooled to
r.t., the voltammogram shown in Fig. 4 is measured.
The activated complex is present only in small concen-
tration. Instead, the major wave describes a reversible
process at E1/2=0.5 V. It should be noted that gener-
ally the catalytic activity has decreased after 18 h at
80°C ([5]a). Therefore, the main wave must represent a
catalytic non-active ruthenium complex. On further
heating to 80°C for 4 h, the reversible wave at 0.5 V is
recorded only at r.t. It will be demonstrated that such
redox potentials are typical for catalytically non-active
complexes (see below).
after termination of the catalysis
The electrochemical behavior of the catalytic system
RuH2(PPh3)4/CF3SO3H was studied after having
reached the maximum TON. Fig. 6 displays the CV of
a solution at r.t. that had been activated and heated to
80°C for 24 h. The current density of the quasi-re-
Experiments were carried out in order to study the
influence of the time-lag between dissolving 1 in MA
and the activation with CF3SO3H on the catalytic activ-
ity. Thus, we determined the maximum TON of solu-
tions of 1 in MA that were activated after 0.5, 1, 5, 24
and 72 h. These experiments always led to the same
overall TONs. Thus, the point of time of activation has
no influence on the TONs. In view of the mechanistic
interpretation of our results this is of importance. An
electrochemical experiment of similar nature was car-
ried out. A solution of 1 in MA showed after 12 h only
Fig. 6. CV of RuH2(PPh3)4 and 3.2 equivalents of CF3SO3H in MA
after heating to 80°C for 24 h.

186
C. Strehblow et al. / Journal of Organometallic Chemistry 561 (1998) 181–189
versible process after activation with CF3SO3H has
decreased enormously compared with the initial con-
centration of the catalyst precursor. No other redox
processes can be seen. Thus, the concentration of ruthe-
nium complexes that are left in solution seems to be
very low. No reduction processes can be observed in
this voltammogram either. We assume that the remain-
ders of ruthenium complexes that showed redoxactivites
in the freshly activated solution (RuH2(PPh3)4 with 3.2
equivalents CF3SO3H) before heating to 80°C are no
longer dissolved. Indeed, the preparative experiments
proved that the solutions were heterogeneous after the
termination of the catalysis. The complex that remains
in solution in low concentration shows a reversible
oxidation at a low potential (Fig. 6). The CV is identi-
cal to that of the experiment where the catalytic solu-
tion had been heated to 80°C for 18 h and cooled to r.t.
(see Fig. 4 above). This result was interpreted as
demonstrating the generation of a complex that is no
longer catalytically active in the dimerization of MA.
The low current density at the end of the catalysis is
due to traces of oxygen that penetrated the reaction
vessel and led to nonsoluble ruthenium complexes. The
higher current density in Fig. 5 must be due to the fact
that the electrolysis cell was purged permanently with
argon during 18 h, thus excluding oxygen completely.
Intentional contamination of a catalytic solution with
traces of oxygen shows a decrease in the maximum
TON. A voltammogram of such a contaminated solu-
tion displays an enormous decrease in the current
density.
Fig. 7. CV of RuHCl(CO)(PPh3)3 in MA (trace 1) and CV of the
solution after activation with Et3BF4 (trace 2) (oxidative part).
Cl(CO)(PPh3)3 in MA is displayed in Fig. 7 (trace 1).
Two irreversible oxidation steps O4 and O5 are ob-
served. O4 at 0.62 V with a small current density that
disappears within 15 min is assigned to the oxidation of
the hydrido complex 3. This could be confirmed by a
measurement in dimethyl formamide where 3 is stable.
Only one irreversible oxidation wave at 0.6 V is ob-
served. The process O5 can be attributed to the oxida-
tion of the insertion complex Ru(CH2CH2CO2Me)
(CO)Cl(PPh3)2 (4), because a voltammogram of isolated
4 in MA shows the same oxidation peak.
A voltammogram of RuHCl(CO)(PPh3)3 (3) in MA
shows no reduction peak in the negative potential
range. The current density increases at potentials of
B −1 V, however. If the potential range is extended to
more negative values the current density increases enor-
mously, indicating an electrocatalytic step at potentials
B −1 V. Therefore, it is not possible to decide whether
there is a reduction step at low potentials (B −1 V).
The reductive part of the voltammogram of isolated
Ru(CH2CH2CO2Me)(CO)Cl(PPh3)2 (4) shows an irre-
versible reduction wave at −0.88 V that is a reduction
of the dissolved complex and no follow up reduction of
an oxidized complex. Although in a solution of dis-
solved 3 a large portion of the complex has already
been converted to 4, its reduction can not be seen in the
CV due to the electrocatalytic reaction. Thus, the re-
duction behavior of the insertion complex 4 is different
from that of 3.
3.3. Influence of an excess CF3SO3H (ten equi6alents)
It had been found that the addition of more than 3.2
equivalents CF3SO3H to the catalyst precursor 1 influ-
ences the catalytic activity in a negative way. Therefore,
we investigated the electrochemical behavior of an
‘overactivated’ catalytic system. In the CV after the
addition of ten equivalents CF3SO3H we observe an
irreversible oxidation wave at 1.1 V with a very low (10
mA) current density.
This is not the same result as for the catalytic system
after the addition of 3.2 equivalents of activating acid
(Fig. 1). Our interpretation is that only a very small
concentration of the activated complex is left in solu-
tion (O3 in Fig. 1(a)). Other electrochemically active
complexes can not be discovered.
The activation of RuHCl(CO)(PPh3)3 or of
Ru(CH2CH2CO2Me)(CO)Cl(PPh3)2 with Et3OBF4 leads
to identical results. Trace 2 in Fig. 7 shows a CV
shortly after activating 4 with Et3OBF4. The oxidation
of 4 is still visible, disappears, however, within 35 min.
There is an oxidation wave at 1.5 V (O6) that is
assigned to a complex which is produced on addition of
3.4. CV studies on RuHCl(CO)(PPh3)3 (3) and
Ru(CH2CH2CO2Me)(CO)Cl(PPh3)2 (4)
The catalyst precursor 3 which dimerizes MA in the
presence of Et3OBF4 dissolves slowly in MA and inserts
MA into the RuꢀH bond with the loss of one phos-
phane, yielding 4.
A
voltammogram of RuH-

C. Strehblow et al. / Journal of Organometallic Chemistry 561 (1998) 181–189
187
4. Discussion
the oxonium salt. This complex is more difficult to
oxidize than either 3 or 4. The current density of O6 is
ca. 15 mA higher than in the activated solution of 3.Fig.
8 displays voltammograms of RuHCl(CO)(PPh3)3 and
of Ru(CH2CH2CO2Me)(CO)Cl(PPh3)2 after the activa-
tion with oxonium salt starting at ca. 0.5 V. The
addition of Et3OBF4 either to 3 or 4 leads to an easily
reducible complex (Ered= −0.13 V).
The investigation of the mechanisms of transition
metal-catalyzed reactions is difficult because the active
species are often generated in situ from precatalysts,
constitute in many cases labile complexes, and are
normally present in low concentration. In many cases,
indirect methods have to be applied to establish cata-
lytic cycles. However, there remains a deficit which
complements spectroscopic methods like NMR or IR
spectroscopy. Electrochemistry might be such a tool
because the redox behavior of metals depends on their
oxidation state and the ligands surrounding the metal
center. In this investigation it has been shown that
cyclic voltammetry can be used as such an analytical
tool. Current density can be taken as a measure of
concentration and the potentials at which complexes
are reduced or oxidized yield information on the nature
of the complexes in solution.
Which is the knowledge gained by our studies on the
dimerization of methyl acrylate by ruthenium com-
plexes? The preceding contribution confirmed our ear-
lier finding based on the isolation of 2 that two
equivalents of triphenylphosphane are liberated on dis-
solving 1 in MA. The CV studies revealed further that
a new complex, representing Ru(II), is generated by
oxidative addition of MA which then reacts to a third
complex. The latter is presumably the insertion complex
7 of MA into the RuꢀH bond of 6. As no further
triphenylphosphane is liberated during these transfor-
mations, the complexes 6 and 7 should still be coordi-
nated with two molecules of phosphane. No
information is obtained as to the number of coordi-
nated molecules of MA. In case of a 16-electron com-
plex, it should be two molecules, possible is also an
18-electron complex with three y-coordinated
molecules of MA.
3.5. CV studies on CpRuH(PPh3)2 (5)
It was found that 5 has no catalytic acitvity in the
dimerization of MA, neither when heating the complex
in MA nor after the addition of either trifl-
uoromethanesulfonic acid or triethyloxonium te-
trafluoroborate. We studied the electrochemical
behavior of 5 in MA. The CVs obtained when using the
olefin as the solvent were in good agreement with the
results of Tilset [18] in tetrahydrofurane. We found no
reduction down to −2.5 V, however a reversible one-
electron oxidation at 0.22 V, and an irreversible oxida-
tion at 0.77 V. Furthermore, the oxidation of liberated
phosphane appeared at 1.45 V in MA. Quantitative GC
investigations revealed the liberation of 0.95 equivalents
triphenylphosphane. We further subtracted the IR spec-
trum of neat MA from that of 5 in MA. In the
difference spectrum we observe a RuꢀH vibration at
1950 cm−1, and an inverse stretching of the MA car-
bonyl group. Presumably, one liberated phosphane has
been replaced by x-coordinated MA on dissolving 5 in
MA, thus leading to a reduction of the concentration of
MA. No insertion of MA into the RuꢀH bond takes
place.
Addition of trifluoromethanesulfonic acid (3.2 equiv-
alents) produces a complex that is more difficult to
oxidize than either 2 or 3 but can be reduced easily.
Earlier ([5]a) we had shown that three equivalents of
the acid remove three molecules of phosphane. Thus,
the catalytic active complex holds still one phosphane.
If it is formed from 6 or 7, the active species should be
either 8 or 9 (Scheme 1) where one phosphane in 6 or
7 has been replaced by a molecule of MA. The experi-
ments where the activation by CF3SO3H was carried
out at a time when complexes 2 and 6 had disappeared
almost completely in favor of 7 suggests that 9 rather
than 8 describes the catalytic active species (however,
see below). The activation by CF3SO3H then consists in
removing a molecule of phosphane from 6 or 7 and
transforming this phosphane together with the two
molecules that were generated on formation of 2 from
1 to the phosphonium salt 10.
Fig.
8.
CVs
of
RuHCl(CO)(PPh3)3
and
of
Ru(CH2CH2CO2Me)(CO)Cl(PPh3)2 in MA after activation with
Et3OBF4 (reductive part).

188
C. Strehblow et al. / Journal of Organometallic Chemistry 561 (1998) 181–189
On the basis of these findings a plausible catalytic
cycle can be written with 9 as starting point. In contrast
to our earlier proposal, the sulfonic acid is not neces-
sary to produce the catalytically active ruthenium(II)–
hydride complex by protonation. The catalytic cycle
includes both 8 and 9 as necessary intermediates.
How does the deactivation proceed? The electro-
chemical analysis of a ‘dead’ catalysis showed a low
concentration of a complex which can be oxidized
reversibly. No other redox processes can be observed.
Thus, most of the ruthenium ‘disappeared’, i.e. was
deposited as insoluble and, therefore, electrochemically
inactive complexes. The nature of the complex in solu-
tion is not known, except that it should not contain
phosphane because it had been shown earlier that free
phosphane is present in a dead catalysis solution. Ex-
cess of CF3SO3H stops the catalytic dimerization.
Conclusions as to the requirement for a rutheniu-
m(II) complex to be catalytically active in the dimeriza-
tion of MA can also be drawn from the studies on 3
and 4. It could be shown that 3 and 4 have different
redox behavior. Insertion of MA into the RuꢀH bond
of 3 increases the oxidation potential from 0.8 to 1.2 V
versus SCE. Both are irreversible processes under our
conditions. Addition of triethyloxonium tatrafluorobo-
rate leads to the formation of ethylchloride [6], thus
producing a free coordination site for MA. This com-
plex is easily reduced (−0.13 V versus SCE) and seems
to be involved in the catalytic cycle. It is interesting to
note that the catalytic active species produced from 6
and 7 is reducible at an almost identical potential
(−0.5 V versus SCE). The quasi-reversible oxidation
potential of the active complex, however, appears at a
lower potential (+1.2 V versus SCE) than that for
activated 6 or 7 (+1.5 V versus SCE).
Complex 5 looses one triphenylphosphane and adds
one molecule of MA in neat MA. The RuꢀH bond is
Scheme 1. Formation of 10 from either 8 or 9.

C. Strehblow et al. / Journal of Organometallic Chemistry 561 (1998) 181–189
189
[4] M. Bruckmann, Ph.D. Thesis, Universita¨t Essen, Essen, Ger-
many, 1992.
[5] (a) B. Patzke, R. Sustmann, J. Organomet. Chem. 480 (1994) 65.
(b) R. Sustmann, H.J. Hornung, T. Schupp, B. Patzke, J. Mol.
Catal. 85 (1993) 149.
unaffected. The CV of this complex shows a reversible
one-electron oxidation at +0.22 V versus SCE, a po-
tential that is similar to that obtained for the ruthenium
complex in the dead catalysis when 2 had been the
starting ruthenium complex.
[6] Th. Schupp, Ph.D. Thesis, Universita¨t Essen, Essen, Germany,
1994.
[7] R.J. McKinney, M.C. Colton, Organometallics 5 (1986) 1752.
[8] (a) C. Amatore, M. Azzabi, A. Jutand, J. Am. Chem. Soc. 113
(1991) 1670. (b) C. Amatore, M. Azzabi, A. Jutand, J. Am.
Chem. Soc. 113 (1991) 8375. (c) C. Amatore, A. Jutand, F.
Khalil, M.A. M’Barki, L. Mottier, Organometallics 12 (1993)
3168.
[9] R. Sustmann, B. Patzke, R. Boese, J. Organomet. Chem. 470
(1994) 191.
[10] C. Strehblow, R. Sustmann, C. Amatore, preceding publication.
[11] K. Hiraki, N. Ochi, Y. Sasada, H. Hayashida, Y. Fuchita, S.
Yamanaka, J. Chem. Soc. Dalton Trans (1985) 873.
[12] DigiS, User guide, GfS Aachen, 1992.
[13] R.R. Gagne, C.A. Koval, G.C. Lisensky, Inorg. Chem. 19 (1980)
2854.
[14] (a) R. Holm, Inorg. Synth. 12 (1970) 238. (b) R. Young, G.
Wilkinson, S. Komiya, A. Yamamoto, Inorg. Synth. 17 (1977)
75.
5. Conclusion
Besides valuable information on the species formed
in activated or not-activated MA solutions of 1–5, the
conclusion might be that a requirement for a catalytic
active ruthenium(II) complex is an easy reduction at ca.
−0.1 to −0.5 V versus SCE, which is irreversible. If
this reduction is not present then the complexes are
catalyically inactive. This phenomenon might be used
as a heuristic criterion for the screening of ruthenium
complexes to be used in studies on the dimerization of
MA. Studies in this direction are in progress.
[15] N. Ahmed, J.J. Levison, S.D. Robinson, M.F. Uttley, Inorg.
Synth. 46 (1974) 48.
References
[16] J.D. (a), G. Gilbert, J. Wilkinson, A Chem. Soc, 1749. (b) M.I.
Bruce, C. Hameister, A.G. Swinger, R.C. Wellis, Inorg. Synth.
28 (1990) 270. (c) H. Suzuki, O.W. Lee, N. Oshima, Y. Moro-
Oka, Organometallics 6 (1987) (1969) 1569.
[17] A.J. Bard, L. Faulkner, Electrochemicaal Methods, Wiley, New
York, 1980.
[1] (a) G. Oehme, H. Pracejus, J. Organomet. Chem. 320 (1987)
C56. (b) I. Tkatchenko, D. Neibecker, P. Grenouillet,
Organometallics 3 (1984) 1130.
[2] G. Wilke, Angew. Chem. 100 (1988) 189.
[3] (a) M. Brookhart, E. Hauptmann, J. Am. Chem. Soc. 114 (1992)
4437. (b) E. Hauptmann, S. Sabo-Etienne, P.S. White, M.
Brookhart, J.M. Garner, P.J. Fagan, J.C. Calabrese, J. Am.
Chem. Soc. 116 (1994) 8038.
[18] O.B. Ryan, K.-T. Smith, M. Tilset, J. Organomet. Chem. 421
(1991) 315.
.