Ruthenium-catalyzed transvinylation - New insights

Article abstract of DOI:10.1002/adsc.201300447

The use of ruthenium complexes in transvinylation catalysis has been well established since the 1980s. However, the reaction mechanism and the active catalyst species, which is presumed to contain ruthenium carbonyl carboxylate entities, have so far remained elusive. In this work the synthesis and characterization of three novel ruthenium complexes comprising ruthenium carbonyl carboxylate structural motifs including two single crystal structures as well as the crystal structures of two known ruthenium complexes are reported. These new complexes and four known ruthenium complexes with appropriate structural motifs were applied in transvinylation catalysis. Mechanistic studies including identification and characterization of the active species, isotope labeling experiments and examination of the regioand stereoselectivity of the transvinylation reaction are presented, resulting in the proposal of a probable reaction mechanism, which is supported by DFT calculations on the B3LYP/6-31G* level of theory.

Full text of DOI:10.1002/adsc.201300447

FULL PAPERS
DOI: 10.1002/adsc.201300447
Ruthenium-Catalyzed Transvinylation – New Insights
Jennifer Ziriakus,+a,b Teresa K. Zimmermann.,+a Alexander Pçthig,a
Markus Drees,a Stefan Haslinger,a Dominik Jantke,a and Fritz E. Kꢀhna,*
a
Chair of Inorganic Chemistry/Molecular Catalysis, Catalysis Research Center, Technische Universitꢁt Mꢀnchen,
Ernst-Otto-Fischer Straße 1, 85747 Garching b. Mꢀnchen, Germany
Fax : (+49)-89-289-13473; e-mail: fritz.kuehn@ch.tum.de
WACKER-Institut fꢀr Siliciumchemie, Department of Chemistry, Technische Universitꢁt Mꢀnchen, Ernst-Otto-Fischer
b
Straße 1, 85747 Garching b. Mꢀnchen, Germany
+
J. Z. and T. K. Z. contributed equally to this work.
Received: May 22, 2013; Revised: September 4, 2013; Published online: October 9, 2013
Supporting information for this article is available on the WWW under http://dx.doi.org/10.1002/adsc.201300447.
Abstract: The use of ruthenium complexes in trans- ate structural motifs were applied in transvinylation
vinylation catalysis has been well established since catalysis. Mechanistic studies including identification
the 1980s. However, the reaction mechanism and the and characterization of the active species, isotope la-
active catalyst species, which is presumed to contain beling experiments and examination of the regio-
ruthenium carbonyl carboxylate entities, have so far and stereoselectivity of the transvinylation reaction
remained elusive. In this work the synthesis and are presented, resulting in the proposal of a probable
characterization of three novel ruthenium complexes reaction mechanism, which is supported by DFT cal-
comprising ruthenium carbonyl carboxylate structur- culations on the B3LYP/6-31G* level of theory.
al motifs including two single crystal structures as
well as the crystal structures of two known rutheni- Keywords: carbonyl ligands; carboxylate ligands; re-
um complexes are reported. These new complexes action mechanisms; reactive intermediates; rutheni-
and four known ruthenium complexes with appropri- um
Introduction
ents claiming the use of different ruthenium precursor
species for ruthenium-catalyzed transvinylation,[13a–c]
Vinyl esters are frequently used in industrial applica- to the best of our knowledge mechanistic details have
tions such as paint production,[1] medical products,[2] never been studied. We present here a mechanistic
paper coatings,[3] and construction materials,[4] as well study of the catalytic pathway including isotope-label-
as in organic synthesis and pharmaceutical chemis- ing experiments and examination of regio- and stereo-
try.[5] Thus, a wide variety of methods has been devel- selective aspects. Based on the experimental results,
oped for their preparation. These include oxidative we propose a plausible reaction mechanism. Three
[6]
acetoxylation of olefins using PdACTHNURGTNEUNG(OAc)2 and direct novel and four known ruthenium complexes were syn-
addition of carboxylic acids to terminal alkynes cata- thesized and characterized in order to emulate rele-
lyzed by mercury salts,[7] Ru,[8] Rh[9] or Ir[10] com- vant catalyst structural motifs. The reaction mecha-
plexes. The synthetic approach via transvinylation of nism is supported by DFT calculations on the B3LYP/
carboxylic acids with vinyl donors has been reported 6-31G* level of theory.
using Hg(II)[11] and Pd(II)[12] materials as well as
a series of ruthenium precursors such as ruthenium
carbonyls, ruthenocene or ruthenium trichloride hy- Results and Discussion
drate.[13] The advantages of ruthenium-catalyzed
transvinylation include the accessibility of functional- Structural Motif of the Active Catalyst Species
ized vinyl ester building blocks, lower toxicity com-
pared to mercury-based reactions, accessibility of Patent literature states that a variety of ruthenium
thermally labile vinyl esters and increased stability compounds can be used to generate the catalyst for
compared to Pd(II) systems. Despite a number of pat- transvinylation in situ. The empirical formula
Adv. Synth. Catal. 2013, 355, 2845 – 2859
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2845

Jennifer Ziriakus et al.
FULL PAPERS
[Ru(CO)2ACHTUNGTRENNUNG
(RCO2)] is assigned to this catalyst,[13c]
therefore the most active ruthenium precursors are
believed to be ruthenium carbonyl compounds.[13d]
However, RuCl3 hydrate may also be used as precur-
sor for efficient transvinylation catalysis, which is
preferential due to the significantly lower cost.[13a] In
a typical reaction 0.5–5 mol% of the ruthenium pre-
cursor, and an up to ten-fold excess of vinyl donor rel-
ative to the amount of carboxylic acid are used. If
RuCl3 hydrate is used as ruthenium source, the addi-
tion of an alkali salt such as sodium acetate or sodium
hydroxide is necessary for satisfactory transvinylation
activity.[13] Transvinylations are equilibrium reactions
which typically exhibit equilibrium constants close to
1 (Scheme 1).[13d]
Figure 1. Proposed structure of the Ru(II) (left) and Ru(I)
(right) species in the active reaction mixture. L=H2O, ClÀ
or CH3COOH; R=CH2CH3, CH3.
Based on these observations, a model catalytically
active mixture using 2 mol% RuCl3 hydrate, 2 mol%
NaOH, vinyl acetate and propionic or valeric acid as
model substrates in a molar ratio of 2.7:1 (vinyl ace-
ACHTUNGTRENNUNG
completion of the reaction. IR spectroscopy was used
to analyze the obtained solid residue. Three strong
carbonyl absorption bands were observed. The two
bands at 2131 cmÀ1 and 2059 cmÀ1 indicate the pres-
ence of three CO ligands within a mononuclear
Ru(II) species. The relatively high frequency carbonyl
absorption bands are in a typical range for fac-ruthe-
nium(II) tricarbonyls.[14] 13C NMR spectroscopy shows
a carbonyl signal at 196.2 ppm, further substantiating
the presence of carbonyl ligands. Electrospray
mass spectrometry of the crude reaction mixture
reveals
[Ru2(CO)
a mononuclear [Ru(CO)
the
(RCOO)2(L1)(L2)]À species as well as
3A
(R1COO)(R2COO)]À species
presence
of
a
dinuclear
G
4ACHTUNGTRENNUNG
C
ACHTUNGTRENNUNG
(Figure 1). Fragmentation patterns clearly indicate
the presence of three carbonyl ligands in both cases.
From these results, taking into account the unequal
intensity of the CO vibrational frequencies in the IR
and the presence of dinuclear species as evidenced
via ESI-MS, a mixture of the two species displayed in
Figure 1 is proposed to be present in the reaction mix-
ture formed from alkaline RuCl3 hydrate.
The carbonyl pattern in the IR spectrum is in ac-
cordance with the two proposed active species and
was confirmed by DFT calculations (B3LYP/6-31G*)
(Figure 2). Assuming a mixture of both species, the
IR spectrum exhibits the band structure and absorp-
tion intensities as expected. Differences in the lower
Figure 2. Calculated IR spectra for the model compounds
(top) and experimental spectrum of the dried reaction mix-
ture (bottom).
Scheme 1. Reaction scheme for the model transvinylation reaction with propionic acid and valeric acid as model substrates.
2846
asc.wiley-vch.de
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2013, 355, 2845 – 2859

Ruthenium-Catalyzed Transvinylation – New Insights
Table 1. Crystallographic data for complexes 1 and 3.
frequency region result from organic residuals in the
catalytically active crude mixture.
From the catalytically active crude mixture, differ-
ent ruthenium carbonyl carboxylate species could be
isolated and characterized. The mononuclear salt Na-
1
3
Formula
C15H22NaO15Ru (81%)
C14H22O10Ru2S2
C16H24NaO15Ru (19%)
568.13
P-1
Formula wt
Space group
a (ꢀ)
b (ꢀ)
c (ꢀ)
616.60
P 21/c
15.941(2)
8.5929(11)
16.887(2)
90
100.662(3)
90
2273.2(5)
4
ACHTUNGTRENNUNG[fac-Ru(CO)3ACHTUNGTRENNUNG(CH3COO)3] (1) was obtained after
evaporation of all volatiles, washing with diethyl ether
and recrystallization from THF (see Figure 3). Crys-
tallographic data are given in Table 1. The anionic
complex crystallizes in the non-centrosymmetric
space group P-1 and exhibits a fac configurated octa-
hedral coordination sphere. Within the limits of the
experimental error, equivalent bond lengths are
equal. The octahedral structure is slightly distorted,
resulting in trans-bonding angles of 171.42(14) to
175.55(16) deg. The fac carboxylate ligands are allo-
cated in close proximity due to mutual coordination
by both sodium and ruthenium, while the carbonyl li-
gands spread out. This is reflected in the cis-bonding
angles of 79.78(9), 80.68(10), 82.52(9) deg between
the carboxylate ligands, while the respective angles
between carbonyl ligands range from 89.16(18) to
11.1566(9)
12.327(1)
12.4777(10)
66.021(3)
63.565(3)
77.624(3)
1402.8(2)
2
a (deg)
b (deg)
g (deg)
V (ꢄ3)
Z
Dcalcd. (gcmÀ3)
Nref
1.345
5091
1.802
4009
Npar
351
259
R1 (I>2s(I))
wR2 (all data)
Goodness of fit
0.0386
0.1028
1.144
0.0210
0.0548
1.04
91.61(17) deg. The bonding angles intermittent be-
tween one carboxylate and one carbonyl ligand are
found in the range of 92.09(14) to 96.42(15) deg.
In 19% of the crystallized species, one of the ace-
tate ligands is exchanged for a propionate ligand. This
substitution is always observed at the same carboxyl-
ate ligand position, i.e., overall an exchange of about
0.19ꢃ0.33=0.06 (6%) of acetate ligands in favor of
propionate ligands can be observed in the crystal.
This mixture of carboxylate ligands can be attributed
to the presence of both acetic acid and propionic acid
in the reaction mixture in a ratio of 1:1 after about
1
3 h. It is evident from the integrals in the H NMR
spectrum that the carboxylate ligand positions are oc-
cupied in a 1:1 ratio by propionate and acetate in so-
lution. The observed signal positions appear shifted
slightly upfield compared to the isolated carboxylic
acids, which is attributed to p-backbonding of the
ruthenium metal center. The carbonyl signal in the
13C NMR spectrum is observed at 196.2 ppm. The ob-
servation of only two carbonyl absorption bands in
the IR spectrum of the crystallized species at
2127 cmÀ1 and 2049 cmÀ1 provides evidence for the
predominant symmetrical occupation of the carboxyl-
Figure 3. ORTEP
ACHTUNGTRENNUNG(CH3COO)3] (1) as isolated from the crude reaction mix-
representation
of
NaACTHNGUTERU[NNG fac-Ru(CO)3
ture. Thermal ellipsoids are shown at the 30% probability ate ligand positions with three acetates. The relatively
high frequency absorption band at 2127 cmÀ1 is in
level. Protons are omitted for clarity. Selected bond lengths
a typical range for fac-ruthenium(II) tricarbonyls.[14]
The large difference of 239 cmÀ1 between nsym and
nasymm of the carboxylate vibration band illustrates the
monodentate binding mode of the carboxylate li-
gands.[15]
À
À
(ꢄ) and bond angles (deg): Ru1 C1 1.910(4), Ru1 C2
1.909(4), Ru1 C3 1.905(4), Ru1 O4 2.079(3), Ru1 O6
À
À
À
À
À
À
2.082(2), Ru1 O8 2.085(2), Na1 O10 2.327(4), Na1 O12
2.297(4), Na1 O14 2.326(3), O4 Ru1 C3 171.52(14), O6
Ru1 C2 175.55(16), O8 Ru1 C1 173.21(14), O4 Ru1 O6
79.78(9), O4 Ru1 O8 80.68(10), O6 Ru1 O8 81.52(9),
O6 Ru1 C1 92.49(14), O4 Ru1 C1 95.16(14), O4 Ru1 C2
96.01(15), O8 Ru1 C2 96.42(15), O8 Ru1 C3 92.09(14),
C1 Ru1 C2 89.33(19), C1 Ru1 C3 91.61(17), C2 Ru1 C3
89.16(18).
À
À
À
À
À
À
À
À
À
À
À
À
À
In contrast to the previously reported salt [(n-
À
À
À
À
À
À
(CH3COO)3],[16] the Na+ cation
Pr)4N]ACTHUNRGTENNUG[fac-Ru(CO)3HCATUNGTRENNGUN
À
À
À
À
À
À
À
À
À
À
within the crystal structure interacts directly with
three ruthenium-coordinated carboxylate ligands and
Adv. Synth. Catal. 2013, 355, 2845 – 2859
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
2847

Jennifer Ziriakus et al.
FULL PAPERS
accounts for the co-crystallization of three additional
acetates. Na+ itself is thus coordinated by six carbox-
ylate ligands overall in a distorted octahedral manner.
After addition of water to the reaction mixture, car-
boxylate-bridged ruthenium polymers of the type
[Ru(CO)2ACHTUNGTRENNUNG
(RCOO)]n [R=CH2CH3,[17] (CH2)5CH3 de-
pending on the substrates used] can be isolated and
identified via 1H and 13C NMR spectroscopy
(Figure 4). The valerate bridged polymer has not
Figure 4. [Ru(CO)2ACHTUNGTRENNUNG(RCOO)]n as isolated from the reaction
mixture. R=CH2CH3, (CH2)3CH3 (2).
Figure 5. ORTEP
representation
of
[Ru(CO)2(m-
CH3CH2COO)(dmso)]2 (3) as isolated from the reaction
AHCTUNGTRENNUNG
mixture upon addition of DMSO. Thermal ellipsoids are
shown at the 50% probability level. Protons are omitted for
clarity. Selected bond lengths (ꢄ) and bond angles (deg):
been reported so far. For synthetic purposes, it may
also be synthesized according to the general proce-
À
À
À
Ru1 Ru2 2.6678(4), Ru1 S1 2.4081(7), Ru2 S2 2.4214(8),
dure for [Ru(CO)2ACTHNUGTRNEUNG(RCOO)]n-type polymers published
À
À
À
Ru1 O5 2.1191(18), Ru2 O6 2.1166(18), Ru1 O7
by Lewis et al. in 1969.[17] Upon addition of DMSO,
À
À
À
2.1203(17), Ru2 O8 2.1088(18), Ru1 C1 1.846(3), Ru2 C3
1.838(3), Ru2 Ru1 S1 167.40(2), Ru1 Ru2 S2 165.71(2),
a new dimeric species [Ru(CO)2(m-CH3CH2COO)-
ACHTUNGTRENNUNG(dmso)]2 (3) is formed and can be isolated as yellow
crystals. This complex is a new representative of di-
meric ruthenium complexes of the type [Ru(CO)2
À
À
À
À
À
À
À
À
À
À
O5 Ru1 O7 84.05(7), O6 Ru2 O8 83.13(7), O5 Ru1 C1
À
À
À
À
91.40(10), O6 Ru2 C3 94.23(12), O7 Ru1 C2 94.63(9),
À
À
À
À
À
À
O8 Ru2 C4 93.41(12), C1 Ru1 C2 89.89(12), C3 Ru2 C4
À
À
À
À
ACHTUNGTRENNUNG(RCOO)2(L)]2, which have been known since the
89.20(15), O5 Ru1 C2 177.68(10), O6 Ru2 C4 176.23(12),
1960s.[17,18] The crystal structure is shown in Figure 5,
crystallographic data are given in Table 1.
À
À
À
À
O7 Ru1 C1 175.39(10), O8 Ru2 C3 176.97(12).
The complex crystallizes in a monoclinic system in
the space group P21/c and exhibits a typical sawhorse
structure, with two bridging propionate ligands in the (see above), neither [Ru(CO)
paddlewheel positions. In trans-position to every car- [Ru(CO)2A(RCOO)(L)]2-type complexes have been
boxyl-O atom is a carbonyl ligand. The geometry is employed in transvinylation catalysis so far.
slightly bent, which is reflected in trans-bonding In a more general manner, Ru-carbonyl-carboxyl-
(RCOO)3]À nor
3ACHTUNGTRENNUNG
CTHUNGTRENNUNG
angles of 175.39(10) deg to 177.68(10) deg. The axial ate species have been suggested before by Murray as
DMSO ligands are distorted towards the propionate active catalysts in transvinylation reactions.[13a–c] In his
ligands, which results in bonding angles along the re- patents from the late 1980s/early 1990s, he identifies
spective Ru1 Ru2 S axes below 170 deg. Interesting- “[Ru(CO)2RCO2]” as the active species.[13c] It is stated
ly, the methyl groups are oriented towards the propio- that the choice of ruthenium precursor is not critical
nate ligands, possibly due to interaction of propio- for the reaction to proceed. This indicates that the
nate-O with methyl-H atoms. The DMSO-oxygen same active species is formed regardless of the pro-
atoms are directed towards the carbonyl ligands. vided ruthenium source if the reaction conditions are
Within the limits of the experimental error, equiva- chosen in an appropriate manner.
lent bond lengths are equal.
À
À
1
In the H NMR spectrum, the DMSO signal ap-
pears with a significant downfield shift of 0.38 ppm at Catalytic Activity
3.00 ppm in CDCl3 compared to the free solvent mol-
ecule, illustrating the reduced electron density due to The following comparison of the catalytic activity of
ruthenium coordination. The propionate signals defined ruthenium carbonyl carboxylates in transviny-
appear shifted slightly upfield compared to free pro- lation catalysis aims at confirming the structural
pionic acid, indicating p-backbonding of the rutheni- motifs of the active catalyst species as proposed in the
um metal center to the carboxylate ligands.
previous section (Figure 1). In addition, the reactivity
Even though complexes with structures similar to of mononuclear and dinuclear ruthenium complexes
those that were isolated from the crude transvinyla- is compared in order to elucidate the role of the two
tion reaction mixture are well known in the literature proposed structures of the active species.
2848
asc.wiley-vch.de
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2013, 355, 2845 – 2859

Ruthenium-Catalyzed Transvinylation – New Insights
Synthesis of Reference Compounds
Complex 6 was synthesized from Ru3(CO)12 with
an excess of pivalic acid in toluene, yielding an
Figure 6 provides an overview over the ruthenium orange crystalline product.[19] Single-crystal XRD
catalysts applied. Complexes 5,[14] 6,[19] 7[20] and 8[17] could be carried out in this work for the first time,
were synthesized according to known literature proce- confirming the structure of the complex as assigned
dures; 1, 3, and 4 are new complexes, which have not by Shvo et al. in 1986.[21] Crystallographic data are
been reported so far. For synthetic purposes, complex given in Table 2. Each of the ruthenium atoms is coor-
3 is formed from the [Ru(CO)2(CH3CH2COO)]n poly- dinated in a slightly distorted octahedral manner, with
À
mer 8, which is recrystallized from DMSO to form two t-BuCOOH ligands along the Ru1 Ru2-axis and
two bridging t-BuCOO ligands connecting both ruthe-
nium atoms. The terminal carboxylic acid protons
each form a hydrogen bond to one of the bridging
carboxylate ligands, which results in a distortion of
the axial pivalic acid ligands towards the bridging car-
À
À
boxylate ligands. This is reflected in the O3 Ru1
Ru1#1 bonding angle of 163.13(4), i.e., well below 180
deg. As indicated in Figure 7, the methyl groups of
the bridging pivalate ligands are disordered in the
crystal.
1H NMR spectroscopic investigation reveals two
separate methyl signals for the t-BuCOO and t-
BuCOOH at 1.13 ppm and 1.25 ppm, which are both
shifted slightly upfield compared to the uncoordinat-
ed pivalic acid. The acidic protons are observed at
11.46 ppm.
Compound 7 has been synthesized previously by
Schumann et al.[22] Single crystal structural analysis is
presented here for the first time (Figure 8), crystallo-
graphic data are given in Table 2. The complex crys-
tallizes in an orthorhombic structure and exhibits the
expected structural features of a dinuclear sawhorse
type ruthenium complex. Similar to the dimeric struc-
ture 3, the axial ligands are distorted towards the pro-
Figure 6. Catalyst complexes used for comparison with the
catalytically active crude mixture.
À
À
pionate ligands. This is reflected in the P1 Ru1 Ru2
À
À
and P2 Ru2 Ru1 bonding angles below 180 deg.
the dimeric structure 3. Analytical details are dis-
cussed above. Complex 4 is prepared in a two-step
synthesis from Ru3(CO)12 with propionic acid and
vinyl acetate. The intermediate step is the formation
of [Ru(CO)2(CH3CH2COO)]n, which forms the mono-
meric structure 4 after refluxing with vinyl acetate
and propionic acid. The propionate signals appear at
a slightly upfield position compared to the isolated
carboxylic acid at 1.09 ppm and 2.29 ppm, respective-
ly, in the 1H NMR spectrum (MeOD-d4). The two
binding modes of the propionate ligands result in two
superimposed sets of signals. This is even more evi-
dent in the 13C NMR spectrum, where the carboxylate
C atoms are observed at 178.3 ppm (h2-CH3CH2COO)
and 183.4 ppm (h1-CH3CH2COO), respectively, thus
exhibiting a difference in the chemical shift of more
than 5 ppm. The difference can also be observed at
the aliphatic C atoms, but diminishes with increasing
distance to the metal center. Due to slow relaxation,
only one carbonyl C atom is observed at 198.0 ppm.
Table 2. Crystallographic data of complexes 6 and 7.
6
7
Formula
Formula wt
Space group
a (ꢀ)
b (ꢀ)
c (ꢀ)
C12H19O6Ru
360.34
I 2/a
15.2653(3)
9.8416(2)
21.8388(6)
90
C34H64O8P2Ru2
864.93
P bca
18.0933(10)
17.7207(9)
24.5667(13)
90
a (deg)
b (deg)
106.809(1)
90
3140.77(12)
8
1.524
3877
90
90
g (deg)
V (ꢄ3)
7876.7(7)
8
1.459
Z
Dcalcd. (gcmÀ3)
Nref
7206
Npar
213
435
R1 (I>2s(I))
wR2 (all data)
Goodness of fit
0.0251
0.0576
0.955
0.0202
0.0493
1.056
Adv. Synth. Catal. 2013, 355, 2845 – 2859
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
2849

Jennifer Ziriakus et al.
FULL PAPERS
Catalytic Activity
Figure 9 summarizes the results of the kinetic studies.
The activity of the sample catalyst species decreases
in the order 4>5>6>1>3>8>7. Interestingly, com-
plex 4, which corresponds to the mononuclear active
species as proposed above, exhibits by far the highest
activity, followed by the mononuclear bis(trifluoroa-
cetato) complex 5. The anionic mononuclear species
1 shows only very little activity, just like all investigat-
ed dinuclear complexes. This suggests that uncharged
mononuclear species in general may exhibit the high-
est activities in transvinylation. The reduced activity
of 5 compared to 4 may be attributed to the electron-
Figure 7. ORTEP representation of [Ru(CO)2ACTHUNRGTENNUG(t-BuCOO)ACHTUNGTERN(NUGN t-
BuCOOH)]2 (6).[19] Thermal ellipsoids are shown at the
50% probability level. Protons are omitted for clarity. Se-
À
lected bond lengths (ꢄ) and bond angles (deg): Ru1 Ru1#1
2.6117(3), Ru1 O3 2.2487(13), Ru1 C11 1.841(2), Ru1 C12
À
À
À
À
À
À
1.832(2), Ru1 O1 2.1280(13), O3 Ru1 Ru1#1 163.13(4),
À
À
À
À
À
C12 Ru1 C11 88.78(10), C11 Ru1 Ru1#1 97.13(6), O1
À
À
À
À
À
Ru1 O3 82.59(5), C12 Ru1 O1 174.23(8), C6 O4 H4
106.1(19).
Figure 8. ORTEP
{Ru(CO)2(CH3CH2COO)AHCTNUTGREG[NNUN PACHTUNGTRENNUNG
representation
of
(t-Bu)3]}2 (7).[22] Thermal ellip-
soids are shown at the 50% probability level. Protons are
omitted for clarity. Selected bond lengths (ꢄ) and bond
À
À
À
angles (deg): Ru1 Ru2 2.7085(2), Ru1 P1 2.5962(5), Ru2
P2 2.5957(5), Ru1 O1 2.1251(13), Ru1 O3 2.1330(13),
À
À
À
À
À
Ru2 O2 2.1678(13), Ru2 O4 2.1377(13), Ru1 C7 1.842(2),
Ru1 C8 1.846(2), Ru2 C9 1.842(2), Ru2 C10 1.832(2), P1
Ru1 Ru2 165.721(13), P2 Ru2 Ru1 165.380(13), O1 Ru1
À
À
À
À
À
À
À
À
À
À
À
À
À
O3 85.13(5), O4 Ru2 O2 84.50(5), C7 Ru1 O1 170.87(7),
À
À
À
À
À
À
C8 Ru1 O3 172.76(7), C10 Ru2 O4 166.86(7), C9 Ru2
À
À
À
À
O2 168.44(7), C7 Ru1 C8 86.83(9), C10 Ru2 C9 86.85(9).
Figure 9. Top : Catalytic activity of the model complexes in
transvinylation of propionic acid with vinyl acetate. [Ru]=
0.3 mol% referred to propionic acid, propionic acid:vinyl
acetate=1:2.7, T=1008C. Bottom: Comparison of the cata-
lytic activity of 4 and RuCl3 hydrate at [Ru]=0.3 mol% and
2 mol% in transvinylation of propionic acid with vinyl ace-
tate, propionic acid:vinyl acetate=1:2.7, T=1008C.
Within the limits of the experimental error, equiva-
lent bond lengths are equal.
2850
asc.wiley-vch.de
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2013, 355, 2845 – 2859

Ruthenium-Catalyzed Transvinylation – New Insights
withdrawing trifluoroacetato ligands. The resulting catalyzed transvinylation. Therefore, complex 4 is
electron deficiency seems to slow down the activation chosen as model catalyst for the following mechanistic
of the catalyst. This may be attributed to the fact that investigations.
electron density at the metal center is a crucial pa-
rameter for the addition of the vinyl donor to the
ruthenium center, as will be discussed below. The di- Mechanistic Studies
nuclear complexes 3 and 6 and the polymeric species
8 exhibit mediocre activity during catalysis. Consider- In order to elucidate (i) how the active species is
ably slower conversion is observed compared to 4 and formed from the mixture of RuCl3 hydrate, NaOH,
5. The dinuclear and polynuclear species are assumed RCOOH, and vinyl acetate, and (ii) how the active
to form minor amounts of the mononuclear active species may act in a transvinylation reaction pathway,
species, with the equilibrium shifted clearly towards a series of mechanistic experiments was carried out.
the dinuclear and polynuclear compounds. The bis(tri- These experiments aim to determine the origin of the
tert-butylphosphino) complex 7 is entirely inactive, CO ligands within the active species, the nature of the
corresponding to the previously reported very low ac- transferred functional group, regio- and stereoselec-
tivity of Ru-phosphino complexes.[13a–c] Alkaline tive aspects as well as relevant electronic and steric
RuCl3 hydrate exhibits a long activation period of influences of the vinyl donor.
roughly 30 min, during which the catalytically active
species is formed. After 2 h reaction time, the maxi-
mum conversion (62%) for the mixture prepared Origin of CO Ligands
from RuCl3 hydrate ([Ru]=2 mol% referred to pro-
pionic acid) is observed, after which the transvinyla- It has been reported previously that the same active
tion of the acetic acid formed in situ reaches consider- species is formed in situ from different Ru precur-
able quantities. The fact that ruthenium-catalyzed sors.[13c] An important structural feature of the active
transvinylation is an equilibrium reaction is evidenced species appear to be the carbonyl ligands. However,
by the drop of vinyl propionate content in solution to to date it remains unclear how these may be formed
50% after 4 h reaction time.[13d] A comparison of the from RuCl3 hydrate in transvinylation. By employing
activity of alkaline RuCl3 hydrate with complex 4 is 13C-labeling experiments and analysis of the rutheni-
shown in Figure 9. For induction of significant conver- um products via IR and NMR spectroscopy, the
sion by alkaline RuCl3 hydrate, a higher molar con- origin of the CO ligands was investigated. Four 13C-,
centration of Ru is needed compared to the mononu- respectively, 18O-labeled substrates were used
clear species 4. When [Ru]=0.3 mol%, only very (Scheme 2). The origin of the CO ligands is particu-
little conversion is observed for alkaline RuCl3 hy- larly intriguing because the formation of carbonyl li-
drate, while at [Ru]=2 mol%, the conversion ach- gands from carboxylate groups of the substrate mole-
ieved with both systems is comparable, with the mon- cules could be ruled out. However, when vinyl acetate
onuclear species 4 exhibiting the highest activity in with 13C-labeled vinyl positions was used, 13CO li-
transvinylation for all investigated Ru concentrations. gands could be detected via IR spectroscopy. The iso-
Taking into account that the activity of the neutral topic shift of the carbonyl absorption band positions
mononuclear species 5 is second best within the set of corresponds well to the calculated values within the
catalyst species tested, it may be assumed that mono- approximation for a harmonic oscillator (see the Sup-
nuclear ruthenium complexes of the type [Ru(CO)3 porting Information, Table S1).
ACHTUNGTRENNUNG(RCOO)2] are responsible for activity in ruthenium-
Scheme 2. Investigation of the origin of the CO ligands of the active species.
Adv. Synth. Catal. 2013, 355, 2845 – 2859
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
2851

Jennifer Ziriakus et al.
FULL PAPERS
(cot)][23] and Ru hydride complexes.[24] Path
[RuACTHNUGRTENNUG(cod)ACHTUGNTRENNUGN
Table 3. Composition of NMR mixtures to determine the
mechanism of the origin of the CO ligands. VAM=vinyl
acetate monomer, PA=propionic acid. To every mixture
listed, 7.7 mg of RuCl3 hydrate and 0.1 mL of D2O were
(a) starts from the assumption that the olefinic carbon
atoms may be oxidized to formic acid under the given
reaction conditions. The following steps – formation
of [Ru(CO)2Cl2]n from RuCl3 hydrate and formic
acid[25] as well as the cleavage of the polymeric struc-
ture by donor ligands[18c,25b,c] – are well known in the
literature. Path (b) proceeds via a Wacker-type oxida-
added. VACHTUNGTRENNUNG(VAM)+V(propionic acid)=0.4 mL.
Entry VAM
[equiv.]
PA
NaOH
CO Signal
[equiv.] [mg]
(13C NMR) [ppm]
1
2
3
4
5
6
1.0
1.0
2.7
–
–
2.7
–
–
1.0
1.0
1.0
1.0
–
200.0
202.6, 195.1
202.7, 194.7
–
tion of the resulting olefin at a RuACTHNUTRGNEUNG(III) center to form
1.5
1.5
–
1.5
–
acetaldehyde, which has been reported previously.[26]
Formal oxidative addition of acetaldehyde to the
ruthenium center and reductive elimination of meth-
ane results in the formation of a carbonyl ligand. De-
carbonylation of acetaldehyde under release of meth-
–
202.1, 194.6
ane has been reported for Ir
stage, it remains uncertain along which pathway the
This indicates C=C bond activation, cleavage and reaction proceeds. Further investigations remain nec-
(III) systems.[27] At this
ACHTUNGTRENNUNG
oxidation at the ruthenium center during the activa- essary to understand in detail how the active catalyst
tion period. To elucidate the mechanism and the reac- species is formed from RuCl3 hydrate.
tants involved in this unusual CO ligand formation,
a series of NMR-scale experiments was carried out
(Table 3), mimicking the conditions of the catalytical- Investigation of the Transferred Functional Group
ly active reaction mixture but strategically omitting
single components. The formation of a carbonyl Through application of 18O-labeled acetic acid in
ligand species from only RuCl3 and vinyl acetate a model transvinylation experiment (Scheme 4), we
(Entry 1) indicates that the oxygen is not necessarily found that the vinyl group rather than the vinyloxy
taken from propionic acid, as might be assumed from group is transferred during transvinylation catalysis.
the formation of C18O when 18O-labeled propionic GC-MS measurements provide evidence for the for-
acid is used (Scheme 2).
Based on these results, two possible mechanisms
for the formation of the active species can be drafted
(Scheme 3). Both reaction pathways are initiated by
À
cleavage of the vinyl ester. C O bond cleavage of
vinyl esters at the indicated position has been report-
Scheme 4. Reaction scheme for the determination of the
transferred functional group using 18O-labeled propionic
ed for low-valent ruthenium complexes such as
acid. [Ru]=0.3 mol% 4.
mation of 18O-vinyl acetate, thus demonstrating the
transfer of solely a vinyl entity.
Regioselectivity
Regioselectivity during transvinylation is examined
using 1-substituted vinyl donors such as isopropenyl
acetate as vinyl donor (Scheme 5).
1H and 13C NMR spectroscopic investigations of the
products of the transvinylation cycle reveal isopropen-
yl propionate to be the main reaction product; ethen-
Scheme 3. Possible formation of CO ligands in the active
catalyst species.
Scheme 5. Reaction scheme for the investigation of the re-
gioselectivity of the reaction. [Ru]=0.3 mol% 4.
2852
asc.wiley-vch.de
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2013, 355, 2845 – 2859

Ruthenium-Catalyzed Transvinylation – New Insights
yl acetate is not observed. Thus, bond cleavage and
formation both take place at the same vinyl C atom
during transvinylation catalysis at the ruthenium
center.
Stereoselectivity
The stereoselectivity of the transvinylation reaction
was investigated using a cis/trans-mixture (70:30) of
butenyl acetate for transvinylation of propionic acid
(Scheme 6). During catalysis, both cis-butenyl propio-
Scheme 8. Inactive vinyl ester compounds with sterically de-
manding and electron-withdrawing groups. [Ru]=0.3 mol%
4.
nate and trans-butenyl propionACHTUNTRGNEUNGate are formed, indi-
cating a possibility for configuration inversion (see
Table 7).
This is in accordance with mechanistic investiga- Variation of the Vinyl Donor
tions of Pd(II)-catalyzed transvinylation, where a cis-
insertion of the vinyl donor to the metal complex and Variation of the substituents of the vinyl donor was
subsequent rotation and cis-elimination are proposed used to narrow down the electronic and steric re-
intermediate steps of the reaction mechanism.[12c] In quirements for transvinylation at the ruthenium
addition, the percentages of cis- and trans-butenyl center. None of the electronically or sterically deacti-
acetate indicate a preferential conversion of cis-bu- vated vinyl esters displayed in Scheme 8 reacted
tenyl acetate over trans-butenyl acetate, with a clear under transvinylation conditions. This indicates (i)
possibility for configuration inversion in the product that sufficient electron density at the vinyl group is
mixture. These results can best be summed up in necessary for successful transvinylation and (ii) that
a mechanism (Scheme 7) similar to the results pub- sterically demanding residues can prevent the interac-
lished by Sabel et al. concerning Pd(II)-catalyzed tion of the metal center and the vinyl group entirely.
transvinylation.[12c]
It is interesting to note that vinyl trifluoroacetate
cannot be converted even though complex 5 is an ef-
fective catalyst. This may be attributed to the low
electron density at the vinyl double bond, which is
crucial for the transvinylation reaction to proceed
Scheme 6. Reaction scheme for the investigation of the stereoselectivity of the reaction. [Ru]=0.3 mol% 4.
Scheme 7. Proposed reaction mechanism taking into account the stereoselectivity of the reaction.
Adv. Synth. Catal. 2013, 355, 2845 – 2859
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
2853

Jennifer Ziriakus et al.
FULL PAPERS
(see below). The high activity of 5, on the other hand, (see above). Furthermore, it was attempted to find
is due to its beneficial structural features, which pre- a mechanism including oxidative addition of vinyl
dominate the electron-withdrawing effect of the tri- acetate to the Ru center followed by reductive elimi-
fluoroacetate ligands.
nation of vinyl propionate. However, several attempts
with a separated third carboxyl group failed to con-
verge. Instead, a 7-coordinated Ru intermediate bear-
ing three carboxylic groups and possible transition
states was found. As the free energy of this intermedi-
DFT Calculations
To elucidate the reaction mechanism, a DFT study ate is already above 320 kJmolÀ1 with respect to the
was carried out. According to previous observations starting material, and the transition states are found
made and the catalytic activities of known complexes above 350 kJmolÀ1, this pathway was excluded from
with different structural motifs (see above), a neutral further investigations. This is consistent with the re-
monomeric tris
as active species. Generally, two coordination modes transition state of the outer sphere vinyl exchange
of vinyl acetate to the catalyst are possible: either via mechanism exhibits relative free energy of
the vinyl group or the carboxylic oxygen atom to 270 kJmolÀ1 (see below).
which the vinyl group is bound. Coordination via the Both possible reaction mechanisms, which can be
ACHTUNGTRENNUNG(carbonyl) species has been presumed sults for cycle 1 (Scheme 9, left) where the highest
a
remaining carboxylate oxygen atom was ruled out as drafted from these observations, are presented in
a result of the experimental evidence for transfer of Scheme 9. They were investigated via DFT calcula-
the vinyl group only as opposed to the vinyloxy group tions using the B3LYP/6-31G* level of theory.
Scheme 9. Possible transvinylation reaction mechanisms incorporating coordination of the vinyl donor via carboxylate
oxygen (left, catalytic cycle 1) and via the vinyl group (right, catalytic cycle 2). All energy values are given in kJmolÀ1.
2854
asc.wiley-vch.de
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2013, 355, 2845 – 2859

Ruthenium-Catalyzed Transvinylation – New Insights
In cycle 1, the vinyl donor coordinates with the tion states. Looking at the DG scale, the addition of
oxygen atom adjacent to the vinyl group (TS1a) and the vinyl ester in the first step is the rate-determining
the vinyl transfer takes place directly from one car- step in both cycles, representing the highest single
boxylate to the other. This reaction pathway entails barrier compared to the corresponding other transi-
a four-membered transition state (TS2a) where vinyl tion states of the respective cycle. On the DH scale,
transfer is carried out. On the other hand, the key this is true only for cycle 2, while in cycle 1 the vinyl
transition and intermediate states in cycle 2 (TS2b, transfer via the four-membered cyclic transition state
GS3b, TS3b) comprise a six-membered ring, which is (TS2a) requires the highest activation enthalpy. Con-
formed via addition of the Ru center and one of its sidering the rate-determining steps, the preference of
carboxylate ligands to the vinyl double bond. Experi- cycle 2 is evident already at this stage. The activation
mental evidence indicates that bond cleavage and for- enthalpies in cycle 2 are between 19 kJmolÀ1 and
mation both take place at the same C atom. There- 66 kJmolÀ1 and are thus much lower than in cycle
fore, the carboxylate ligand must bind to the vinyl C 1 (54 kJmolÀ1 to 160 kJmolÀ1). It is evident from the
atom adjacent to the carboxylate group (see above). computed energies that coordination via the vinyl
The respective vinyl carboxylate is formed and group in the first step of the reaction pathway is ener-
cleaved from the Ru center. The graphs displayed in getically much more favorable than coordination via
Figure 10 represent the relative free energies and rel- the carboxylate oxygen atom. Moreover, the vinyl
ative enthalpies of the subsequent ground and transi- transfer runs via the low-energy pathway when the
original coordination takes place at the vinyl group.
This may be attributed to the much lower ring tension
in the six-membered ring transition states TS2b and
TS3b in cycle 2 compared to the four-membered tran-
sition state TS2a in cycle 1. Based on these results, we
propose catalytic cycle 2 as the most likely reaction
mechanism for ruthenium-catalyzed transvinylation.
Conclusions
With this study new insights into the mechanism of
ruthenium-catalyzed transvinylation are provided.
Three new ruthenium complexes including two crystal
structures as well as crystal structures of two previ-
ously known ruthenium complexes are presented. The
most likely catalytically active species is identified
and characterized. A reaction mechanism is proposed
based on analysis of ruthenium species formed from
RuCl3 hydrate, the results of a series of isotope label-
ing experiments, regio- and stereoselective investiga-
tions and DFT calculations (Scheme 10). Based on
these results, it might be possible to develop catalysts
with a more rational design approach than the ruthe-
nium precursors, which are currently in use. Since the
exchange of the carboxylate ligand is crucial for trans-
vinylation catalysis, immobilization might be possible
through adequate replacement of the carbonyl li-
gands.
Experimental Section
General Remarks
Figure 10. Free energies (top) and enthalpies (bottom) of
ground and transition states in catalytic cycle 1 (black
squares, oxo coordination) and catalytic cycle 2 (open
squares, vinyl coordination).
Commercially available solvents and reagents were used as
delivered. Unless otherwise noted, all reactions were carried
out under an argon atmosphere. RuCl3 hydrate was obtained
from Sigma–Aldrich and used without further purification.
Adv. Synth. Catal. 2013, 355, 2845 – 2859
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
2855

Jennifer Ziriakus et al.
FULL PAPERS
paper. These data can be obtained free of charge from The
Cambridge Crystallographic Data Centre via www.ccdc.ca-
m.ac.uk/data_request/cif. Crystallographic details are given
in the supporting information.
Preparation of the Catalytically Active Crude
Mixture
100 mg RuCl3·nH2O (0.48 mmol, 1.0 equiv.) and 20 mg
NaOH (0.50 mmol, 1.3 equiv.) were suspended in 5 mL vinyl
acetate (54 mmol) and 1.5 mL propionic acid (20 mmol) or
2.2 mL valeric acid (20 mmol), respectively, were added to
the mixture. The solution was heated to 1408C in a pressure
tube for 4 h and then slowly cooled to room temperature.
The reaction was performed on air. For further analysis, the
brown suspension was dried under vacuum. For analytical
details and spectra, see the Supporting Information.
Isolation of Na
ACHUTNGTERN[UNG fac-Ru(CO)3ACHTUNGTRNE(NUGN RCOO)3] (R=CH3,
CH2CH3 1:1) (1)
100 mg RuCl3·nH2O (0.48 mmol, 1.0 equiv.) and 20 mg
NaOH (0.50 mmol, 1.0 equiv.) were suspended in 5 mL vinyl
acetate (54 mmol) and 1.5 mL of propionic acid (20 mmol)
were added to the mixture. The solution was heated to
1408C in a pressure tube for 4 h and then slowly cooled to
room temperature. The mixture was filtered and concentrat-
ed under vacuum to give a dark red-brown oil, to which
3 mL of diethyl ether were added. After vigorous stirring,
the resulting solid was washed with pentane (5ꢃ3 mL) and
recrystallized from THF; yield: 102 mg (0.23 mmol, 53%);
C12H15NaO9Ru (427.30)/C9H9NaO9Ru (385.22) crystal: anal.
calcd.: C 32.33%, H 3.29%; found: C 32.37%, H 3.36%; IR
(KBr): n=2127 (s), 2049 (vs), 1614 (s), 1375 (m), 1329 (m),
686 (w), 624 (w), 580 cmÀ1 (w); 1H NMR (400 MHz,
MeOD): d=1.10 (t, 3JH,H =7.6 Hz, 3H, CH2CH3), 1.99 (s,
3H, CH3COO), 2.30 (q, 3JH,H =7.5 Hz, 2H, CH2CH3);
13C NMR (101 MHz, MeOD): d=196.2 (CO), 178.4
(CH3CH2COO), 175.3 (CH3COO), 28.2 (CH3CH2COO),
20.8 (CH3COO), 9.5 (CH3CH2COO).
Isolation of [Ru(CO)2(m-h2-CH3CH2COO)]n
Scheme 10. Proposed mechanism for ruthenium-catalyzed
transvinylation.
Water was added to the filtered and concentrated catalyti-
cally active crude mixture (see above) until an orange solid
precipitated out. This solid was washed with diethyl ether
and dried under vacuum; yield: 4%. Satisfactory analytical
data were obtained.[17]
Ru3(CO)12 was obtained from ABCR and used without fur-
ther purification. Elemental analyses were obtained from
the Microanalytical Laboratory of Technische Universitꢁt
Mꢀnchen. Spectroscopic data were recorded on the follow-
ing instruments: IR spectra: Jasco FT/IR 460 PLUS; NMR
spectra: Bruker DPX-400 and Bruker Avance III 400
(1H NMR 400.13 MHz, 13C NMR 100.53 MHz, 31P NMR
100.61 MHz), T=300 K. Signals were calibrated to the re-
sidual proton resonance or the natural abundance 13C reso-
nance of the solvent, respectively.[26] Signal multiplicities are
abbreviated as: s (singlet), d (doublet), t (triplet), q (quar-
tet), quint (quintet), m (multiplet), br (broad). ESI-mass
spectra were recorded on a ThermoElectron LCQ classic.
GC-mass spectra were recorded on a Hewlett–Packard chro-
matograph HP6890 using chloroform as a solvent. CCDC
940441, CCDC 940442,CCDC 940443 and CCDC 940444
contain the supplementary crystallographic data for this
[Ru(CO)2ACHTNUTRGENNGU 3CAHTUNTGNER(NUGN CH2)3COO)]n (2)
(m-h2-CH
Water was added to the catalytically active crude mixture
(see above) until an orange solid precipitated out. This solid
was washed with diethyl ether and dried under vacuum. For
synthetic purposes, 100 mg Ru3(CO)12 (0.16 mmol,
1.0 equiv.) were suspended in 6 mL valeric acid (5.6 g,
55 mmol) and the mixture was heated under reflux condi-
tions for 12 h. After removal of all volatiles and washing
with diethyl ether (4ꢃ5 mL), the product was obtained as
an orange solid; yield: (113 mg (0.22 mmol, 92%);
(C14H18O8Ru2)n (516.43); anal. calcd.: C 32.56%, H 3.51%;
found: C 32.74%, H 3.50%: IR (KBr): n=2039 (s), 1993 (s),
1969 (s), 1941 (m), 1546 (s), 1412 cmÀ1 (m); 1H NMR
2856
asc.wiley-vch.de
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2013, 355, 2845 – 2859

Ruthenium-Catalyzed Transvinylation – New Insights
(400 MHz, CD3COOD): d=0.95 (t, 3JH,H =7.6 Hz, 3H,
CH3CH2), 1.39 (m, 2H, CH3CH2), 1.63 (m, 2H,
CH2CH2COO), 2.39 (t, 3JH,H =7.2 Hz, 2H, CH2COO);
13C NMR (101 MHz, THF-d8): d=202.7 (CO), 186.4 (COO),
36.9 (CH2COO), 29.1 (CH2CH2COO), 23.1 (CH3CH2), 14.1
(CH3).
Table 5. Overview over catalyzed transvinylation of propion-
ic acid with vinyl acetate. Conversion was determined using
1H NMR spectroscopy in acetone-d6.
Cat.:
1
3
4
5
6
7
8
Time [min]
0
20
40
60
0
2.9
3.8
5.7
9.1
9.1
10.7
13.0
16.0
18.0
0
0
0
0
9.1
9.9
17.1
21.3
24.0
27.5
36.7
44.1
49.0
0
3.8
6.5
9.5
12.3
13.0
16.7
22.5
28.6
31.5
0
0
0
0
0
0
0
0
0
0
0
0
0
2.0
2.0
2.9
3.8
4.8
9.1
10.7
15.3
26.5
32.4
40.1
42.5
47.1
54.3
59.5
61.4
[Ru(CO)2(m-h2-CH3CH2COO)
ACTHUNTRGNE(NUG dmso)]2 (3)
2.9
4.8
5.7
7.4
8.3
11.5
16.0
21.3
30 mg [Ru(CO)2(CH3CH2COO)]n (0.13 mmol, 1.0 equiv.)
were dissolved in 0.2 mL DMSO at 808C. After cooling to
48C overnight, the yellow crystals were filtered off and
washed with pentane; yield: 14 mg (0.02 mmol, 18%);
C14H22O10Ru2S2 (616.87); anal. calcd.: C 27.27%, H 3.60%, S
10.37%; found: C 27.56%, H 3.69%, S 10.65%; IR (KBr):
n=2037 (s), 1997 (m), 1971 (s), 1943 (m), 1558 (s), 1433
80
100
120
160
200
240
1
(m), 1243 (s), 1102 (s), 1022 cmÀ1 (s); H NMR (400 MHz,
CDCl3): d=1.03 (t, 3JH,H =7.5 Hz, 3H, CH2CH3), 2.29 (q,
3JH,H =7.4 Hz, 2H, CH2CH3), 3.00 (s, 6H, SCH3); 13C NMR
(101 MHz, CDCl3): d=200.6 (CO), 189.0 (COOCH2CH3),
Table 6. Overview over catalyzed transvinylation of propion-
ic acid with vinyl acetate using alkaline RuCl3 hydrate (9)
and 4. Conversion was determined using H NMR spectros-
42.4 (OSACHTUNGTRENNUNG(CH3)2), 30.7 (CH2CH3), 10.9 (CH2CH3).
1
copy in acetone-d6.
ACHTUNGTRENNUNG
Cat.:
9[a]
9[b]
4[a]
4[b]
Time [min]
100 mg Ru3(CO)12 (0.16 mmol, 0.33 equiv.) were suspended
in 5 mL propionic acid and heated for 12 h to 1008C. After
evaporating all volatiles under vacuum, 2 mL of propionic
acid and 6 mL of vinyl acetate were added to the yellow
solid and the mixture was heated to 708C for 2 h. Volatiles
were evaporated under vacuum and the resulting colorless
solid washed with pentane (5ꢃ2 mL); yield: 97 mg
(0.29 mmol, 62%); C9H10O7Ru (331.95)·0.5H2O; anal. calcd.
C 31.77%, H 3.26%; found: C 31.62%, H 3.35%; IR: n=
2132 (w), 2055 (s), 1984 (s), 1520 (s), 1413 cmÀ1 (s);
1H NMR (400 MHz, MeOD): d=1.08 (m, 3H, CH2CH3),
0
20
40
60
0
0
0
1.0
0
0
4.7
27.5
0
0
15.3
26.5
32.4
40.1
42.5
47.1
54.3
33.8
53.9
58.8
63.7
64.9
65.4
65.4
80
100
120
160
180
200
240
2.0
3.0
4.7
60.2
56.1
47.9
59.5
61.4
64.1
65.8
3
2.29 (q, JH,H =7.6 Hz, 2H, CH2CH3); 13C NMR (101 MHz,
MeOD): d=198.0 (CO), 183.4 (h1-CH3CH2COO), 178.3 (h2-
[a]
CH3CH2COO),
CH3CH2COO),
CH3CH2COO).
30.0
11.1
(h1-CH3CH2COO),
(h1-CH3CH2COO),
28.1
9.5
(h2-
(h2-
[Ru]=0.3 mol%.
[Ru]=2 mol% referred to propionic acid.
[b]
Catalyses
Table 7. Overview over catalyzed transvinylation of propion-
ic acid with cis/trans-butenyl acetate (70:30) with 0.3 mol%
[fac-Ru(CO)3(h2-CH3CH2COO)(h1-CH3CH2COO)] (4) as
catalyst. Conversion was determined using 1H NMR spec-
troscopy in CDCl3.
See Table 4, Table 5, Table 6, and Table 7.
Table 4. Catalytic transvinylation of propionic acid with iso-
propenyl acetate (1:3) with 0.3 mol% 4 as catalyst. Conver-
sion was determined using 1H NMR spectroscopy in ace-
tone-d6. Propionic anhydride was observed as side-product.
Time trans-bu-
[min] tenyl ace-
tate [%]
cis-buten- trans-butenyl cis-butenyl
yl acetate propionate
propionate
[%]
[%]
[%]
Time [min]
Conversion
0
30
60
120
240
360
570
29.5
30.0
31.0
32.9
31.9
32.6
35.2
70.5
70.0
69.0
67.1
60.2
57.1
47.6
30.9
0
0
0
0
3.0
4.0
7.1
16.0
0
0
0
0
4.8
6.3
10.0
15.2
0
0
1
7
12
18
37
47
15
30
45
70
130
190
1230 38.4
Adv. Synth. Catal. 2013, 355, 2845 – 2859
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
2857

Jennifer Ziriakus et al.
FULL PAPERS
dron 1971, 233–238; c) A. Sabel, J. Smidt, R. Jira, H.
Prigge, Chem. Ber. 1969, 102, 2939–2950; d) J. Smidt,
W. Hafner, R. Jira, R. Sieber, J. Sedlmeier, A. Sabel,
Angew. Chem. 1962, 74, 93–128.
Computational Details
All calculations have been performed with Gaussian03.[28]
The level of theory contains the hybrid DFT functional
B3LYP[29] and the double zeta 6–31G*[30] basis set for all
atoms except Ru. The Stuttgart97-ECP[31] has been applied
for the Ru metal centers. All optimized stationary points
have been checked by frequency calculations if they repre-
sent a transition state (NImag=1) or a ground state
(NImag=0). Enthalpies and Gibbs free energies are given
in kJmolÀ1 with respect to all starting material of a given
catalytic cycle. They have been calculated in gas phase for
T=298.15 K and 1 atm. The simulated IR spectra have been
taken unscaled from the calculated frequencies of the model
compounds.
[13] a) R. E. Murray, U.S. Patent 005155253A, 1991;
b) R. E. Murray, European Patent EP O0506070A2,
1992;
c) R. E.
Murray,
European
Patent
EPO0351603A2, 1989; d) R. E. Murray, D. M. Lincoln,
Catal. Today 1992, 13, 93–102.
[14] T. Funaioli, C. Cavazza, F. Marchetti, G. Fachinetti,
Inorg. Chem. 1999, 38, 3361–3368.
[15] G. B. Deacon, R. J. Phillips, Coord. Chem. Rev. 1980,
33, 227–250.
[16] A. John, G. Wilkinson, Polyhedron 1989, 8, 597–602.
[17] G. R. Crooks, B. F. G. Johnson, J. Lewis, I. G. Williams,
G. Gamlen, J. Chem. Soc. A 1969, 2761–2766.
[18] a) R. W. Hilts, S. J. Sherlock, M. Cowie, E. Singleton,
M. M. de V. Steyn, Inorg. Chem. 1990, 29, 3161–3167;
b) S. J. Sherlock, M. Cowie, Organometallics 1988, 7,
1663–1666; c) B. Therrien, G. Sꢀss-Fink, Coord. Chem.
Rev. 2009, 253, 2639–2664.
Acknowledgements
The authors, particularly J. Z. and T. K. Z. are grateful for
the financial support of this project provided by the Wacker
Chemie AG and appreciate the support of F. Marr and P.
Gigler. J. Z., S. H. and D. J. thank the TUM Graduate
School and NanoCat – an international doctorate college of
the Elite Network Bavaria (ENB). M. D. acknowledges the
provision of computing time by the Leibniz Computing
Centre (LRZ) of the Bavarian Academy of Sciences.
[19] M. Rotem, Y. Shvo, I. Goldberg, U. Shmueli, Organo-
metallics 1984, 3, 1758–1759.
[20] R. Brꢀckner, Organic Mechanisms, Springer, Heidel-
berg, 2010.
[21] M. Rotem, I. Goldberg, U. Shmueli, Y. Shvo, J. Orga-
nomet. Chem. 1986, 314, 185–212.
[22] H. Schumann, J. Opitz, J. Pickardt, J. Organomet.
Chem. 1977, 128, 253–264.
[23] J. G. Planas, T. Marumo, Y. Ichikawa, M. Hirano, S.
Komiya, J. Chem. Soc. Dalton Trans. 2000, 2613–2625.
[24] S. Komiya, A. Yamamoto, J. Organomet. Chem. 1975,
87, 333–339.
References
[1] R. E. Neff, R. G. Ryles, European Patent EP
0311799A1, 1988.
[2] I. M. Nunez, J. G. Linhardt, J. A. McGee, J. Hunt, M.
Alton, D. A. Shipp, J. F. Kunzler, D. M. Ammon, Patent
WO 2010/147864A2, 2010.
[25] a) G. N. Krishnamurthy, N. Shashikala, J. Serb. Chem.
Soc. 2009, 74, 1085–1096; b) D. Mulhern, Y. Lan, S.
Brooker, J. F. Gallagher, H. Gçrls, S. Rau, J. G. Vos,
Inorg. Chim. Acta 2006, 359, 736–744; c) C. M. Kepert,
G. B. Deacon, L. Spiccia, G. D. Fallon, B. W. Skelton,
A. H. White, J. Chem. Soc. Dalton Trans. 2000, 2867–
2873; d) P. A. Anderson, G. B. Deacon, K. H. Haar-
mann, F. R. Keene, T. J. Meyer, D. A. Reitsma, B. W.
Skelton, G. F. Strouse, N. C. Thomas, J. A. Treadway,
A. H. White, Inorg. Chem. 1995, 34, 6145–6157.
[26] M. M. T. Kahn, A. P. Rao, J. Mol. Catal. 1988, 44, 95–
105.
[27] a) A. E. Roa, V. Salazar, J. Lꢅpez-Serrano, E. OÇate,
M. Paneque, M. L. Poveda, Organometallics 2012, 31,
716–721; b) P. J. Alaimo, B. A. Arndtsen, R. G. Berg-
man, Organometallics 2000, 19, 2130–2143; c) M. A.
Garralda, Dalton Trans. 2009, 3635–3645.
[28] M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuse-
ria, M. A. Robb, J. R. Cheeseman, J. A. Montgom-
ery Jr, T. Vreven, K. N. Kudin, J. C. Burant, J. M.
Millam, S. S. Iyengar, J. Tomasi, V. Barone, B. Mennuc-
ci, M. Cossi, G. Scalmani, N. Rega, G. A. Petersson, H.
Nakatsuji, M. Hada, M. Ehara, K. Toyota, R. Fukuda,
J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O.
Kitao, H. Nakai, M. Klene, X. Li, J. E. Knox, H. P.
Hratchian, J. B. Cross, V. Bakken, C. Adamo, J. Jara-
millo, R. Gomperts, R. E. Stratmann, O. Yazyev, A. J.
Austin, R. Cammi, C. Pomelli, J. W. Ochterski, P. Y.
Ayala, K. Morokuma, G. A. Voth, P. Salvador, J. J.
[3] J. Ahlgren, N. Bruun, V. Buutinen, Patent WO 2010/
058087A1, 2010.
[4] C.-B. Ko, Y.-M. Kim, K.-J. Song, K.-H. Lee, U.S. Patent
2010/0048077A1, 2010.
[5] a) W. J. Kerr, M. McLaughlin, P. L. Pauson, S. M. Rob-
ertson, J. Organomet. Chem. 2001, 630, 104–117; b) H.
Yang, E. Henke, U. T. Bornscheuer, J. Org. Chem.
1999, 64, 1709–1712; c) J. J. Snuparek, J. Mleziva,
Angew. Makromol. Chem. 1970, 12, 145–156.
[6] a) W. Kitching, Z. Rapport, S. Winstein, W. G. Young,
J. Am. Chem. Soc. 1966, 88, 2054–2055; b) I. Nicolau,
P. M. Colling, L. R. Johnson, Patent WO 94/08714,
1994; c) X. Lu, G. Zhu, S. Ma, Tetrahedron Lett. 1992,
33, 7205–7206.
[7] P. F. Hudrlik, A. M. Hudrlik, J. Org. Chem. 1973, 38,
4254–4258.
[8] M. Rotem, Y. Shvo, Organometallics 1983, 2, 1689–
1691.
[9] C. Bianchini, A. Meli, M. Peruzzini, F. Zanobini, Orga-
nometallics 1990, 9, 1155–1160.
[10] H. Nakagawa, Y. Okimoto, S. Sakaguchi, Y. Ishii, Tetra-
hedron Lett. 2003, 44, 103–106.
[11] R. Adelman, J. Org. Chem. 1949, 14, 1057–1077.
[12] a) J. E. McKeon, P. Fitton, A. A. Griswold, Tetrahedron
1972, 28, 227–232; b) J. E. McKeon, P. Fitton, Tetrahe-
2858
asc.wiley-vch.de
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2013, 355, 2845 – 2859

Ruthenium-Catalyzed Transvinylation – New Insights
Dannenberg, V. G. Zakrzewski, S. Dapprich, A. D.
Daniels, M. C. Strain, O. Farkas, D. K. Malick, A. D.
Rabuck, K. Raghavachari, J. B. Foresman, J. V. Ortiz,
Q. Cui, A. G. Baboul, S. Clifford, J. Cioslowski, B. B.
Stefanov, G. Liu, A. Liashenko, P. Piskorz, I. Komaro-
mi, R. L. Martin, D. J. Fox, T. Keith, M. A. Al-Laham,
C. Y. Peng, A. Nanayakkara, M. Challacombe, P. M. W.
Gill, B. Johnson, W. Chen, M. W. Wong, C. Gonzalez,
J. A. Pople, Gaussian03, Gaussian, Inc., Wallingford,
CT, 2004.
[29] a) A. D. Becke, J. Chem. Phys. 1993, 98, 5648–5652;
b) C. Lee, Y. W, R. G. Parr, Phys. Rev. B: Condens.
Matter 1988, 37, 785–789.
[30] J. Hehre, R. Ditchfield, J. A. Pople, J. Chem. Phys.
1972, 56, 2257–2261.
[31] M. Dolg, H. Stoll, H. Preuss, R. M. Pitzer, J. Phys.
Chem. 1993, 97, 5852–5859.
Adv. Synth. Catal. 2013, 355, 2845 – 2859
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
2859