92
F. Ding et al. / Journal of Molecular Catalysis A: Chemical 386 (2014) 86–94
(instead of 1 equiv. as in Fig. 4) were initially introduced no carbene
signals were noticed, at any time. Possibly, several carbene species
are implied: at low EDA concentration the benzylidene carbene is
gradually replaced by a second carbene fragment (ı = 17.128 ppm),
whereas at high EDA concentrations a distinct, very reactive car-
benoid species is formed. Therefore, deviations observed in FTIR
measurements at initial stages of EDA decomposition (Fig. 1) could
be explained by a change in the active species when a large excess
of EDA is applied.
We assume that the catalytic cycle involves first the decompo-
sition of the diazoester, with extrusion of nitrogen gas, to afford the
reactive [Ru] = CHCOOEt intermediate which is playing a role in the
various channels for C C coupling (Scheme 4). Though not detected
is known to be quite unstable, even in solution at low tempera-
metathesis promoter [19].
study could be accounted for by two main mechanistic path-
ways [4c,6b]. They imply essentially the carbene transfer from the
[Ru] = CHCOOEt carbene to styrene via either a carbenoid pathway
(A) or/and a coordination pathway (B) (Scheme 4).
As inferred from previous data [4c], in the carbenoid pathway
formation of the cyclopropane moiety involves a late, unsymmetri-
cal transition state A-I with build-up of a positive charge at the most
distant carbon atom of styrene. Taking into account the 16-electron
configuration of our phosphine Ru complexes the carbenoid mech-
anism, triggered immediately after dissociation of one phosphine,
seems to be the favoured pathway. Obviously, for sterical rea-
sons, C C coupling in the transition state A-I occurs predominantly
towards the trans diastereoisomer.
the new catalysts, 4 and 10, suggest generation of this unstable
metalcarbene in both the carbenoid and the coordination mecha-
3. Experimental
Catalysts 1, 2, 5, 6 are commercially available and were used
without further purification. The ruthenium complexes 3 [10], 7
[13], 8 [13] and 9 (in situ) [2c] were synthesized according to liter-
ature procedures as shown in Scheme 3.
Synthesis of the new ruthenium promoters 4 and 10 was carried
out in this work. For the synthesis of complex 4 a THF solution of
the thalium salt of the Schiff base ligand [20] was added the equiv-
alent amount of Cl2Ru(PPh3)3 and the mixture stirred overnight
at room temperature. After work-up [11b,20] the solid residue was
dissolved in a minimal amount of toluene, reprecipitated with pen-
tane, filtered off, briefly washed on the funnel with pentane and
dried in vacuo to afford an orange-brown powder (78% yield) which
was stored under inert atmosphere. 1H NMR (300 MHz, CDCl3): ı
2.36 [s, 3H, CH3]; 7.10–7.80 [m, 34H, aryl-CH]; 9.95 ppm (s, 1H,
aldimine ligand).
Catalyst 10 was obtained by adding to1,3-bis(2,6-
diisopropylphenyl)imidazolinylidene (in situ prepared in THF)
[3c] one equivalent of Cl2Ru(PPh3)3. The reaction mixture was
refluxed in THF with stirring for 1 h, then cooled down and the
solid materials filtered off. The filtrate was concentrated in vacuo,
the residue solved in a minimal amount of toluene, reprecipitated
with pentane and then handled as above for complex 4 to give
a dark-brown powder. Isolated yield: 81%. 1H NMR (300 MHz,
CDCl3): ı 1,25 [d, 24H, CH(CH3)2]; 2,93 [m, 4H, CH(CH3)2], 3.65 ppm
(m, 4H, CH2).
On the other hand, in the coordination mechanism (path B),
both the olefin and the CHCOOEt carbene coordinate at the metal.
Therefore, the catalyst must provide two coordination sites, easily
accessible through ligand dissociation within our 18- or 16-electron
Ru complexes which contain either two displaceable phosphines (1,
4, 5, 10) or a p-cymene (2, 3, 9). Thus, highly reactive, coordinatively
unsaturated 14- or 12-electron Ru species may arise. However,
the steric configuration imposed by the Schiff-base and NHC lig-
ands in our Ru complexes might sometimes impede simultaneous
coordination of both the olefin and the carbene moiety favouring
instead the alternative carbenoid pathway (A), hence cyclopropa-
nation. As soon as the complex B-I is formed, it rearranges to the
ruthenacyclobutane complex B-II which, by reductive elimination
of the Ru fragment provides substituted cyclopropanes (trans- and
cis-1-carbethoxy-2-phenylcyclopropane) or, by [2 + 2] cyclorever-
sion gives the metathesis products (trans- and cis-ethylcinnamate).
Because in our experiments the cis- and trans-cinnamic esters were
detected by GC only in trace amounts, pathway B-b2 seems dis-
favoured. Indirect support for formation of distinct carbene species
a study with the Grubbs I catalyzed tandem enyne metathesis-
cyclopropanation where the Ru-complex 5 is modified in situ by
the diazoester to form a cyclopropanation active catalyst that no
longer promotes metathesis [6b]. Competing in the overall mecha-
nistic scheme, the carbenoid pathway (C, involving a highly reactive
transition state C-I and the intermediate complex C-II) and the
tion of C C coupling products (maleic and fumaric esters) and
metathesis products (cis- and trans-stilbene). The commonality
in this ensemble is the intermediacy of the ß-carbonyl-carbene
species [Ru] = CHCOOEt. Although also our 1H NMR experiments
(Fig. 4) could not detect the real occurrence of the carbene
Ru = CHCOOEt, the high cyclopropanation yields obtained with
4. FTIR measurements
Kinetic measurements on the decomposition of diazoacetate
in the presence of the Ru catalysts 1–10 were performed using a
Brucker FT-Raman/FT-IR spectrometer with a Nernst glowing ele-
ment as the IR source and a DTGS detector. The flow-through IR-cell
consists of KBr windows with a 2 mm inner separation. The fol-
lowing operations were carried out in a typical FT-IR experiment.
From a preliminary prepared solution, with precise concentration,
of the catalyst in freshly distilled toluene, 3 mol of catalyst were
transferred under continuous Ar-flow to an empty 15 ml vessel. The
solvent was evaporated under vacuum giving a small amount of
solid catalyst. Then 12 ml of extra dry toluene was added to adjust
concentration at a predetermined value. For experiments also using
an olefin substrate, 4 mmol of olefin were usually added. The ves-
sel was sealed with a septum and stored at −18 ◦C. Just before the
experiment the vessel was introduced into a heating block to reach
the desired temperature and 400 mol EDA were added all at once
by a syringe. A needle was inserted through the septum to prevent
pressure building. The reaction mixture was circulated through the
IR cell in a closed circuit by use of a peristaltic pump. All measure-
ments were performed by following the (N N) stretch vibration
at 2110 cm−1. Each experiment was repeated at least 5 times. All
spectra were normalized to the solvent, integrated and correlated
to the concentration by means of a calibration curve.
5. GC measurements
To determine the yield and chemoselectivity in cyclopropana-
tion, capillary GC measurements were performed using a Varian
CDS 401 with a Supelco 5DB-24030 capillary column and a FID
detection system.