Organometallics
Article
Scheme 1. Reaction Mechanism Reported by Lin21
catalysis; however, they have been widely used in electro-
chemical, photophysical, and luminescence studies.10 Other
types of dinuclear gold(I) complexes are complexes based on
bridging diphosphine ligands, which require handling under
inert conditions.11 Air-stable dinuclear gold(I) complexes
bearing a mono-NHC with a bridging (μ-OH) group have
also been extensively studied by Nolan and co-workers.8b,12
These dinuclear gold species were shown to be highly active
catalysts for silver-free and acid-free gold-catalyzed trans-
formations.8b,12 Meanwhile, only a few examples of dinuclear
gold(I) complexes bearing an alkyl bridging two NHC ligands
([Au2(L)X2] where X = Cl, Br and L = bridged bis(NHC))
have been reported in the literature.13 However, the synthetic
procedures leading to their formation present several draw-
backs: (a) the use of a carbene transfer agent via trans-
metalation between bis(Ag-NHC) and a gold precursor,
typically [Au(DMS)Cl] (DMS = dimethyl sulfide), generating
silver waste, (b) the treatment of the highly sensitive
methylene analogue complex [Au2(L)Me2] with a strong
acid (e.g., HCl),13b and (c) the use of a lithium halide additive
(LiBr).13a While developing stable bis(NHC)-modified gold
nanoparticles,14 Crudden and co-workers demonstrated a more
viable methodology using a weak base approach which is
closely related to our previous work on accessing gold-,15
copper-,16 and palladium-based7a,b NHC complexes. Never-
theless, the described protocol was limited to a few examples
and was not optimized. In the pursuit of developing a new
architecture for dinuclear gold(I) complexes bearing alkyl-
bridged bis(NHC) ligands and testing their efficiency in
catalysis, we now report on the straightforward synthesis of
new [Au2(L)X2] complexes. The activity of the synthesized
complexes was investigated in the carboxylative cyclization of
propargylamine. This reaction is of great interest to the
scientific community, as it involves the insertion and fixation of
CO2, which is an abundant greenhouse gas, is inexpensive, and
is a readily available C1 feedstock.17 The use of CO2 as a C1
building block has attracted much attention in recent
years,17,18 as it is one of the most promising alternatives to
phosgene (in certain applications), which is highly toxic.19
During the past decade, several reports dealing with the
capture of CO2 by propargylamines to access valuable
heterocycles have been described using various catalytic
systems.9,20 Studies by Ikariya and co-workers9 have shown
that NHC−mononuclear gold(I) complexes were very
effective catalysts in the carboxylative cyclization of propargyl-
amines. Detailed computational studies of the reaction
mechanism have been reported. The overall mechanism
shown in Scheme 1 below is adapted from the work of Yuan
and Lin.21
complexes experimentally are essential to understand and
design an improved catalyst. This is particularly critical for the
design of dinuclear complexes which are derived from
mononuclear species, for which a mechanistic understanding
is essential. Therefore, in this article we present the first steps
in work aimed at more complete delineation of the catalytic
mechanism for both mononuclear and dinuclear complex
systems.
Seminal work on the mechanism of gold(I)-catalyzed
hydroalkoxylation of alkynes reported by Zhdanko and
Maier22 has highlighted the importance of determining the
resting state of the catalyst and, where possible, isolating and
determining isolated rates of the reaction. The premise that the
catalytic cycle starts as shown in Scheme 1 with binding of
PPA to 9 may be valid in silico but does not correspond to the
experimental conditions, where the catalytic process is itself
much slower than carbonylation of PPA. Even if CO2 is added
after mixing of all reagents, the rate of reaction of CO2 and
PPA producing a mixture of CA and CS in an equilibrium
mixture is much more rapid than the rate of catalysis. In
keeping with the results of Kortunov and co-workers,23 the
concentration of free PPA can be taken as negligible at room
temperature under a CO2 pressure of 1 atm or more. It is also
surprising that the potential role of carbamate complexes in
this catalytic cycle has not received more attention. Toste,
Bergman, and co-workers have shown that [Au(IPr)NR1R2]
rapidly adds CO2 to produce a [Au(IPr)O(OC)NR1R2]
complex,24 and Calderazzo and co-workers25 have determined
the crystal structure of [Au(PPh3)O(OC)NR1R2]. While
stable intermediate complexes of alkynes have been isolated in
several cases and also structurally characterized, these are
always with “noncoordinating” counterions.26 There are a wide
range of carbamate complexes for many metals indicating that
the counterion is not a “noncoordinating” anion.27
A breakthrough achievement in this mechanism was the
isolation and structural characterization, by the Ikariya group,9a
of the IPrAu-vinyl complex labeled IKa in Scheme 1. This was
also shown to be a key intermediate in the catalytic cycle by
both its cleavage by acid to products and its use as a precatalyst
capable of achieving the same rate of reaction as for 9. Key
unanswered questions in the mechanism of this reaction
include why the empirical rate law dP/dt = k [9][PPA]0.75 is
followed, as well as the identity of the resting state of the
catalyst. This experimentally derived reaction order is most
likely the average dependence on temperature of a complex
rate law comprised of a more complex set of elementary steps.
Elucidation of those elementary steps experimentally as well as
identification of the possible resting state and intermediate
RESULTS AND DISCUSSION
■
Synthesis and Characterization of Dinuclear Gold(I)
Complexes. As shown in Scheme 2, the reaction of the ligand
(L·2HX) with [Au(DMS)Cl] (DMS = dimethyl sulfide) in the
presence of K2CO3 at 60 °C in acetone afforded the desired
complexes [Au2(L)X2] (1−8) in 40−120 min. The products
B
Organometallics XXXX, XXX, XXX−XXX