The influence of electron delocalization upon the stability and structure of potential N-heterocyclic carbene precursors with 1,3-diaryl-imidazolidine-4, 5-dione skeletons

Article abstract of DOI:10.1039/b9nj00771g

Targeting N-heterocyclic carbenes (NHCs) with increased π-acceptor character featuring N-fluorophenyl substituents, the molecular 2-chloro-1,3-bis(fluorophenyl)imidazolidine-4,5-diones (1a-c) were isolated from the condensation of the corresponding formamidine with oxalyl chloride. These formal adducts of NHCs with hydrogen chloride demonstrated reactivity akin to that of alkyl halides: 1,3,1′,3′-tetrakis(2,6-dimethylphenyl)-[2, 2′]diimidazolidinyl-4,5,4′,5′-tetraone (2b) was formed via the reductive coupling of 1b, while 1,3-bis(2,6-diisopropylphenyl)-4,5- dioxoimidazolidin-2-yl acetate (3c) was formed as the result of a metathesis reaction with mercury(ii) acetate. Chloride abstraction resulted in the formation of imidazolium-4,5-dione salts (4a-c) that decomposed rapidly, except in the case of the kinetically-stabilized 1,3-bis(2,6-diisopropylphenyl) imidazolium-4,5-dione hexafluorophosphate 4c. All imidazolium-4,5-dione hexafluorophosphate salts decomposed to neutral 2-fluoro-1,3-bis(aryl) imidazolidine-4,5-diones (5a-c) via fluoride abstraction. 2-Methoxy-1,3-di(aryl) imidazolidine-4,5-diones (6a-c) were also prepared and they failed to undergo thermolysis and yield the free NHCs. Computational analyses revealed that the instability of NHCs with an oxalamide skeleton, as well as that of imidazolium-4,5-diones, results from a π-framework which extends over both carbonyl moieties and gives rise to a very low energy LUMO, rendering the compounds in question highly electrophilic.

Full text of DOI:10.1039/b9nj00771g

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/njc | New Journal of Chemistry
The influence of electron delocalization upon the stability
and structure of potential N-heterocyclic carbene precursors
with 1,3-diaryl-imidazolidine-4,5-dione skeletonsw
a
Matthew G. Hobbs,a Taryn D. Forster,a Javier Borau-Garcia,a Chrissy J. Knapp,z
Heikki M. Tuononenb and Roland Roesler*a
Received (in Montpellier, France) 17th December 2009, Accepted 9th February 2010
First published as an Advance Article on the web 26th April 2010
DOI: 10.1039/b9nj00771g
Targeting N-heterocyclic carbenes (NHCs) with increased p-acceptor character featuring
N-fluorophenyl substituents, the molecular 2-chloro-1,3-bis(fluorophenyl)imidazolidine-4,5-diones
(1a–c) were isolated from the condensation of the corresponding formamidine with oxalyl
chloride. These formal adducts of NHCs with hydrogen chloride demonstrated reactivity akin to
that of alkyl halides: 1,3,10,30-tetrakis(2,6-dimethylphenyl)-[2,20]diimidazolidinyl-4,5,40,50-tetraone
(2b) was formed via the reductive coupling of 1b, while 1,3-bis(2,6-diisopropylphenyl)-4,5-
dioxoimidazolidin-2-yl acetate (3c) was formed as the result of a metathesis reaction with
mercury(II) acetate. Chloride abstraction resulted in the formation of imidazolium-4,5-dione
salts (4a–c) that decomposed rapidly, except in the case of the kinetically-stabilized
1,3-bis(2,6-diisopropylphenyl)imidazolium-4,5-dione hexafluorophosphate 4c. All imidazolium-4,5-
dione hexafluorophosphate salts decomposed to neutral 2-fluoro-1,3-bis(aryl)imidazolidine-4,5-
diones (5a–c) via fluoride abstraction. 2-Methoxy-1,3-di(aryl)imidazolidine-4,5-diones (6a–c) were
also prepared and they failed to undergo thermolysis and yield the free NHCs. Computational
analyses revealed that the instability of NHCs with an oxalamide skeleton, as well as that of
imidazolium-4,5-diones, results from a p-framework which extends over both carbonyl moieties
and gives rise to a very low energy LUMO, rendering the compounds in question highly
electrophilic.
Arduengo’s discovery, the chemistry of NHCs has developed
Introduction
into a vast area of research with extensive applications.5
The s-symmetric non-bonding electron pair allows for
strong s-coordination of NHCs to transition metal centers.
As a result of their excellent two-electron donor ability, stable
NHCs have become an important class of ancillary ligand in
homogeneous catalysis.6 While they are typically known
for their exceptional s-donor ability,7 recent studies have
demonstrated that p-interactions may also play a significant
role in the bonding of these ligands to transition metals.8 With
the aim of improving the p-acceptor character of NHCs in
order to study the importance of p-interactions between
carbenes and transition metals, we targeted the synthesis of
NHCs with substitution patterns that would result in a
lowering of the energy of the LUMO. Substitution with
highly-electronegative fluorine atoms has been used to this
end with phosphines.9 Recently, Bielawski and co-workers have
reported the isolation of complexes of 1,3-dimethylimidazolin-2-
ylidene A, possessing s-withdrawing trifluoromethyl substituents
on its backbone.10 While A and its derivatives offer an intriguing
opportunity to study the role of p-back-bonding in metal–
carbene coordination, the reported synthetic protocol does
not allow for the easy variation of the exocyclic N-substituents.
The role of fluorinated N-substituents in tuning the electronics
of NHCs therefore merits attention. NHCs bearing fluoroalkyl
and fluorobenzyl groups at the nitrogen centers have also
been reported,11 however, the hydrocarbon bridges required
Carbenes are compounds featuring a neutral divalent carbon
center bearing a non-bonding electron pair.1 Arduengo reported
the isolation of the first stable N-heterocyclic carbene (NHC)
in 1991, achieving the goal of a long series of investigations
initiated by Wanzlick,2 and proving that this class of compounds
was not restricted to transient species.3 Arduengo’s carbene
was based on an imidazole framework, in which the carbene
center benefits from the stabilization associated with the
s-electron-withdrawing, and p-electron-donating character
of the nitrogen centers.4 The combined effect of these electronic
interactions lowers the energy of the heterocycle’s HOMO,
and raises that of its LUMO. The resulting increase in the
HOMO–LUMO gap is largely responsible for the observed
stability of N-heterocyclic carbenes (NHCs). In the years following
a Department of Chemistry, University of Calgary,
2500 University Dr NW, Calgary, Alberta T2N 1N4, Canada.
E-mail: roesler@ucalgary.ca
b Department of Chemistry, University of Jyvaskyla, P.O. Box 35,
¨
¨
F1-40014 Jyvaskyla, Finland
¨
¨
w Electronic supplementary information (ESI) available: Complete
NMR spectra for all compounds. CCDC reference numbers
743953–743957. For ESI and crystallographic data in CIF or other
electronic format see DOI: 10.1039/b9nj00771g
z Currently at Department of Chemistry, University of Alberta, 11227
Saskatchewan Dr., Edmonton, Alberta, T6G 2G2, Canada.
ꢀc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2010 New J. Chem., 2010, 34, 1295–1308 | 1295

View Article Online
to tether the fluorinated substituents to the heterocyclic
moieties in these compounds diminish their electronic influence
on the carbene center itself. Fluorinated N-substituents that
are formally conjugated to the NHC ring on the other hand,
are expected to exert a more significant electronic influence. To
date, several complexes of such carbenes bearing 4-fluorophenyl
(B)12 and 2,6-difluorophenyl N-substituents (C)13 respectively,
have been prepared. The most highly-fluorinated such species
reported to date is the unsymmetrically N-substituted D,14
with one pentafluorophenyl, and one mesityl N-substituent.
Despite the fact that D possesses five fluorines, while C bears
only four, the less highly-fluorinated C exerts a greater influence
on the nCO stretching vibrations observed in the corresponding
[(NHC)Rh(CO)2Cl] complexes, suggesting that D is a better
electron donor than C. This demonstrates that the electronic
influence of the pentafluorophenyl substituent in D has been
somewhat tempered by that of the N-mesityl substituent.
Interestingly, the observed trends in the nCO stretches of these
complexes are attributed to reduced s-basicity of the fluorinated
ligands, while consideration is not given to the possibility of
increased p-acidity. It is most likely that the observed trends
result from a synergy of the two effects. With the aim of
disambiguating these competing effects, and preparing NHCs
with superior p-acceptor character, we sought to prepare a
highly-fluorinated NHC bearing two N-pentafluorophenyl
groups.
halides, oxalyl chloride appeared to be well-suited to serve as
the required two-carbon backbone fragment.
The reaction between oxalyl chloride and one equivalent of
N,N0-bis(pentafluorophenyl)formamidine (Scheme 1) proceeded
fast and quantitatively according to 1H and 19F NMR spectro-
scopy. The 19F NMR spectrum, however, featured five signals
indicating that the reaction product was less-symmetrical than
the anticipated product 1,3-bis(pentafluorophenyl)-4,5-dioxo-
4,5-dihydro-3H-imidazolium chloride F. Structure 1a, in
which the two faces of the heterocycle are diastereotopic,
was proposed to account for the observed 19F NMR data.
At room temperature, the signals assigned to the ortho- and
meta-fluorines of 1a were broad, suggesting that the rotation
of the perfluorophenyl N-substituents was competitive with
the NMR timescale. Upon cooling below 248 K, the five
signals became sharp and the fine structure of the spectrum
could be resolved. Upon heating, the five signals observed in
the 19F NMR spectrum coalesced into three signals (Fig. 1).
Using the coalescence temperature of 336.0 K for the signals
corresponding to the ortho-fluorine substituents, the energy barrier
for this rotation was determined to be 60.3 ꢁ 0.8 kJ molꢂ1 18
.
The 13C NMR spectrum of 1a displayed the expected
resonances for the N2C(H)Cl and CO fragments of the hetero-
cycle at 83.7 ppm and 153.6 ppm, respectively.
Results and discussion
Scheme 1 Synthesis of 2-chloro-1,3-di(aryl)imidazolidine-4,5-diones
1a–c.
The classical synthesis of imidazolium salts suitable for use as
NHC precursors most often involves the [4 + 1] ring-closing
reaction of an N,N0-disubstituted ethylenediamine, or 1,4-
diazabutadiene moiety with a suitable carbon electrophile such
as triethylorthoformate, or chloromethyl ethyl ether.15 This
was also the synthetic route applied by Grubbs and co-workers in
their preparation of the 1,3-bis(2,6-difluorophenyl)imidazolium
chloride parent to carbene C.13 Bis(pentafluorophenyl)-
imidazolium salts however, could not be prepared in this
manner owing to the much poorer nucleophilicity of the
highly-fluorinated ethylenediamine, or 1,4-diazabutadiene
precursors.16 As a result, we sought to investigate new methods
suitable for the preparation of highly-fluorinated imidazolium
salts such as 1,3-bis(pentafluorophenyl)imidazolium, E. Given
the relative ease with which N,N0-bis(pentafluorophenyl)-
formamidine can be prepared,17 we sought highly-electrophilic
two-carbon backbone fragments which would be capable of
undergoing [3 + 2] ring-closing reactions with the poorly
nucleophilic formamidines to generate imidazolium-type
precursors F suitable for the preparation of highly-fluorinated
NHCs. As a result of the extreme electrophilicity of acyl
Fig. 1 19F NMR spectra of 1a at 336.0 K, 318.0 K, 288.0 K, and
248.0 K (from top to bottom, respectively).
ꢀc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2010
1296 | New J. Chem., 2010, 34, 1295–1308

View Article Online
Table 1 Selected data and structure refinement details for 1a, 2b, 3c, 5a, and 6b
1a
2b
3c
5a
6b
Empirical formula
Formula weight
Crystal system
Space group
a/A
b/A
c/A
a/1
C15HClF10N2O2
466.63
Orthorhombic
Cmc21
25.723(2)
10.273(1)
6.031(1)
90
C84H92N8O10
686.83
Monoclinic
C12/c1
21.883(1)
14.616(1)
21.925(1)
90
C29H38N2O4
478.61
Orthorhombic
Pna21
19.300(2)
12.239(2)
11.804(1)
90
C15HF11N2O2
450.18
Orthorhombic
Cmc21
25.917(2)
9.789(1)
6.040(1)
90
C20H22N2O3
338.40
Monoclinic
P21/c
13.992(1)
10.152(1)
12.991(1)
90
b/1
90
90
1593.7(4)
4
1.945
96.910(1)
90
6961.6(7)
8
1.311
90
90
2788.3(4)
4
1.140
90
90
1532.4(3)
4
1.951
102.895(6)
90
1798.8(3)
4
1.250
g/1
V/A3
Z
dc/g cmꢂ3
2ymax/1
55.00
0.369
54.96
0.086
44.96
0.076
55.12
0.220
48.04
0.085
m(Mo-Ka)/mmꢂ1
Independent reflections
Data/restraints/parameters 1016/1/139
1016 (Rint = 0.0562) 7937 (Rint = 0.1077) 1896 (Rint = 0.1252) 969 (Rint = 0.1430) 2811 (Rint = 0.1112)
7937/0/461
1.018
0.0579
1896/1/325
1.078
0.0551
969/1/139
1.185
0.0587
0.1080
2811/0/230
1.081
0.0697
GOF on F2
1.098
0.0374
0.0835
R1(F) [I 4 2s(I)]
wR2(F2) [all data]
0.1570
0.1275
0.1944
p-bonding interaction between the nitrogen lone-pairs and the
carbonyl carbons of the backbone. The C–C bond between
the two sp2-hybridized backbone carbons is relatively long
(1.540(6) A), consistent with other oxalamide-type systems. For
comparison the C–C bond length in oxalyl chloride is 1.474 A.23
The C–O bond length in 1a (1.198(4) A) is typical of other
carboxyl derivatives (1.17–1.25 A). All together, this data
indicates a certain degree of p-delocalization across the N–CCO
bonds of 1a. The dihedral angle between the pentaflurophenyl
N-substituents and the heterocyclic core of the compound is
66.3(2)1. The intramolecular distance between the imidazolium
proton of 1a (geometrically positioned) and the two nearest
ortho-fluorines on the N-substituents is 2.74(3) A, comparable
to the sum of the van der Waals radii for these two atoms.
Molecules of 1a crystallize in a linear head-to-tail arrangement,
allowing the ring proton to undergo an interaction with the
carbonyl oxygens of an adjacent molecule (HꢃꢃꢃO = 2.55(5) A).
The IR spectra of compounds 1 demonstrate that the
carbonyl moieties present on the heterocyclic backbone are
sensitive to substitution at the nitrogen centers. The electron-
poor N-pentafluorophenyl-substituted derivative 1a features
the highest energy CO stretch (n = 1784 cmꢂ1), while the CO
stretching vibrations in the more electron-rich derivatives 1b
Fig. 2 Molecular structure of 1a with thermal ellipsoids drawn at
50% probability. Atoms related by the mirror symmetry (ꢂx, y, z) are
marked with prime.
A molecular structure similar to that proposed for 1a has
been proposed for G (the N-phenyl substituted derivative)
however, this structure has not been confirmed crystallo-
graphically.19 X-Ray structural analysis of single-crystals of
1a (Table 1), which were obtained by slow evaporation of the
solvent from a CH2Cl2 solution of 1a, proved that the
compound was indeed molecular 2-chloro-1,3-bis(2,3,4,5,6-
pentafluorophenyl)-4,5-imidazolidinedione 1a, rather than
the originally expected ionic imidazolium-4,5-dione (F) chloride
salt. Derivative 1a (Fig. 2) is a planar five-membered hetero-
cycle with a mean deviation from the plane of the heterocycle
of 0.020(3) A, featuring crystallographically imposed mirror
symmetry. The sum of the bond angles around the nitrogen
and carbonyl centers of the heterocycle are 359.1(3)1 and
360.0(3)1 respectively, clearly indicating an sp2-hybridization
of these centers. Interestingly, in the absence of the carbonyl
moieties present on the backbone of 1a, the geometry of the
nitrogen centers becomes significantly more pyramidalized
(sum of the bond angles 351.3–352.61) in other imidazolidines
containing a four-coordinate diaminocarbon center.15,20 The
ring proton was not located, however, the sum of the bond
angles at the diaminocarbon centre C(1) is 324.5(3)1 (Table 2),
reflecting its distorted tetrahedral geometry. The intraannular
N–CCO bond distance in 1a is relatively short (1.373(5) A),
and comparable to the N–C bond distances of pyrroles,21
imidazoliums,15 and oxalamides,22 suggesting the presence of a
and 1c require less energy (n = 1773 and 1772 cmꢂ1
,
respectively). These observations may be rationalized using
resonance forms H and H0. Zwitterionic resonance form H0,
which bears a formal positive charge at the nitrogen center,
will be more favoured in the presence of electron donating
N-substituents, accounting for the lower energy CO stretches
observed in compounds 1b and 1c relative to the electron-poor
derivative 1a. Bielawski and co-workers have recently demon-
strated that a naphthoquinone moiety, as well as cyano groups
that are formally conjugated to the nitrogen centers of NHCs
may be used to probe the extent of metal–carbene p-back-
bonding in NHC complexes.8b,d The sensitivity of the backbone
carbonyl groups of heterocycles 1 to substitution at nitrogen
suggests that NHCs bearing an oxalamide framework may
also be efficient probes for metal–NHC p-backbonding.
ꢀc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2010 New J. Chem., 2010, 34, 1295–1308 | 1297

View Article Online
Table 2 Selected bond lengths (A) and angles (1) for 1a, 2b, 3c, 5a, and 6b
1a
2b
3c
5a
6b
Csp3–Na
N–CCO
1.441(4)
1.373(5)
1.540(6)
1.417(4)
1.198(4)
1.793(4)
102.5(4)
111.2(2)
121.2(3)
124.6(3)
128.7(3)
104.9(2)
126.4(2)
539.1
1.462(3)–1.476(2)
1.357(3)–1.381(2)
1.523(3), 1.524(3)
1.447(3)–1.465(3)
1.208(3)–1.215(2)
1.577(3)
103.2(2), 103.5(2)
109.9(2)–119.2(2)
122.5(2), 124.8(2)
115.9(2)–125.0(2)
126.7(2)–128.5(2)
105.7(2)–107.7(2)
125.5(2)–126.0(2)
540.0, 540.0
1.454(7), 1.442(8)
1.368(8), 1.359(8)
1.510(9)
1.440(7), 1.445(7)
1.212(7), 1.217(7)
1.430(7)
1.448(6)
1.377(7)
1.544(9)
1.418(6)
1.201(6)
1.374(8)
102.3(5)
109.4(4)
122.5(4)
123.8(4)
128.3(4)
105.0(3)
126.6(3)
538.6
1.457(5), 1.445(5)
1.348(5), 1.356(5)
1.515(6)
1.439(5), 1.426(5)
1.220(5), 1.220(5)
1.407(5)
b
CCO–CCOc
N–Cextra
C–O
C–Xd
N–Csp3–N
N–Csp3–X
Cextra–N–Csp3
Cextra–N–CCO
N–CCO–O
N–CCO–CCO
103.2(5)
102.9(3)
108.6(5), 109.5(5)
125.1(5), 122.3(5)
123.3(5), 125.1(5)
127.3(6), 127.4(6)
106.6(5), 105.8(5)
126.0(6), 126.8(6)
539.5
112.7(3), 108.3(3)
123.9(3), 123.1(3)
123.4(3), 124.7(3)
127.5(4), 127.7(4)
106.1(3), 106.3(3)
126.3(4), 126.0(4)
539.9
O–CCO–CCO
P
pentagon angles
a
b
Csp3 indicates the four-coordinate diaminocarbon. CCO indicates the backbone carbonyl carbons. Cextra indicates the extraannular carbon
d
directly attached to the heterocycle. X = Cl, C, OAc, F, and OMe for 1a, 2b, 3c, 5a, and 6b, respectively.
c
Attempts to deprotonate 1a with strong non-nucleophilic
bases such as potassium hexamethyldisilazide (KHMDS),
lithium N,N,N0,N0-tetramethylpiperidide (LiTMP), and potassium
tert-butoxide (KOtBu) to generate the free N-perfluorophenyl-
substituted carbene resulted in the formation of intractable
mixtures that could not be further characterized. The decom-
position of 1a in the presence of strong bases was accompanied
by an immediate colour change from colourless to bright
orange/red in all cases. This observation suggested that redox
chemistry might be playing a role in the decomposition
pathway. The bis(trimethylsilyl)aminyl radical is well-known,24
and is a potential product of the oxidation of KHMDS with 1a.
Attempts to generate a free carbene by elimination of
hydrogen chloride from 1a with weaker bases such as 1,8-
diazabicycloundec-7-ene (DBU), triethylamine, and pyridine
also resulted in the formation of complex mixtures. While the
reaction mixtures could not be further characterized, X-ray
quality crystals isolated from the reaction of 1a with triethyl-
amine were determined to be triethylamine hydrochloride, the
presumed by-product of NHC formation. This suggested the
transient formation of an unstable NHC derived from 1a upon
deprotonation with triethylamine. Attempts to trap the
proposed carbene with classical carbene trapping agents such
as elemental sulfur and rhodium(I) 1,5-cyclooctadiene chloride
dimer, [Rh(cod)Cl]2, failed and in all cases deprotonation in
the presence of a trapping agent resulted in the formation of
complex mixtures. The formation of silver(I)–NHC complexes
by direct metallation of imidazolium precursors is a ubiquitous
means by which a variety of NHC complexes may be
obtained.25 However, 1a failed to react with silver(I) oxide,
Ag2O, under classical reaction conditions.
non-fluorinated, sterically-demanding N-substituents were
prepared in order to assess the suitability of this endocyclic
substitution pattern for the preparation and isolation of
stable carbenes. Reaction of the appropriately-substituted
formamidines with oxalyl chloride resulted in the formation
of compounds 1b and 1c bearing 2,6-dimethylphenyl- and
2,6-diisopropylphenyl (Dipp) N-substituents, respectively
1
(Scheme 1), in high yields (80–94%). The H and 13C NMR
spectra of the products were consistent with the formation of
molecular heterocycles analogous to 1a, suggesting that the
oxalamide-derived skeleton favours molecular heterocycles
over the initially anticipated imidazolium-4,5-dione (F) chloride
salts. 1b exhibits two signals corresponding to the methyl
groups in both the 1H and 13C NMR spectra, while four
1
methyl signals are observed in the H NMR spectrum of 1c.
The above observations indicate that like 1a, the heterocyclic
rings 1b and 1c have two diastereotopic faces due to the
presence of an sp3-hybridized N2C(H)Cl center in the hetero-
cycle. Unlike the broad resonances observed for the
N-substituents at room temperature in the 19F NMR spectrum
of 1a, the N-subsituent signals observed in the 1H NMR
spectra of 1b and 1c are sharp, and the fine structure is
fully-resolved indicating that the free-rotation of these
sterically bulkier substituents does not occur on the NMR
timescale. The resonance corresponding to the heterocyclic
ring proton in compounds 1b (7.66 ppm) and 1c (7.18 ppm)
is shifted upfield relative to the same resonance in the
perfluorophenyl-substituted 1a (7.78 ppm), consistent with
the inductive effect of the N-substituents. From the above
results, it is clear that the molecular, heterocyclic ring structure
of compounds 1a–c, featuring an sp3-carbon, arises from the
influence of the oxalamide framework, and is not significantly
affected by the nature of the exocyclic N-substituents.
Despite the steric protection offered by the 2,6-dimethyl-
phenyl, and 2,6-diisopropylphenyl N-substituents in 1b and 1c,
the compounds undergo decomposition reactions similar to
those observed for 1a in the presence of a variety of
bases. Once again, clear solutions containing compounds 1b
and 1c undergo dramatic color changes upon addition of
strong bases, forming complex mixtures (according to NMR
Since NHCs with an oxalamide framework have not
been reported, analogues of 1a with more conventional,
ꢀc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2010
1298 | New J. Chem., 2010, 34, 1295–1308

View Article Online
1,3-bis(2,6-diisopropylphenyl)imidazolium chloride (IPrꢃHCl),
a common NHC precursor, cannot be reduced in the electro-
chemical window of THF. The values also demonstrate that,
as expected, the most electron-rich compound of the series,
N-diisopropylphenyl-substituted 1c, is more difficult to reduce
than the more electron-poor compounds 1a and 1b. In the case
of the most electron-poor N-pentafluorophenyl-substituted
derivative 1a, a second irreversible reduction is also observed
(Ec = ꢂ2.568 V vs. Fc/Fc+). These observations clearly
demonstrate that molecular structures 1a–c are more prone
to reduction than conventional cationic imidazolium precursors
such as IPrꢃHCl.
As observed with perfluorophenyl-substituted 1a, hydrocarbyl-
substituted 1b and 1c did not react with silver(I) oxide to form
silver(I) NHC complexes. Similar to silver(I) oxide, mercury(II)
acetate is another common reagent used for the in situ
deprotonation of imidazolium salts, and subsequent metallation
of the resulting NHCs.26 Unlike silver(I) oxide, mercury(II)
acetate reacts with the diisopropylphenyl-substituted 1c. The
1H NMR spectrum of the product 3c of this reaction however,
indicates that 1c does not undergo deprotonation in the
presence of mercury(II) acetate. Instead, the resonance corres-
ponding to the heterocyclic proton in 3c (7.44 ppm) is shifted
downfield relative to that of the precursor 1c (7.18 ppm).
Additionally, a new methyl signal corresponding to an acetate
group is observed in the 1H NMR spectrum of 3c. The
presence of a highly downfield CO signal at 169.8 ppm, and
an additional methyl signal at d = 20.8 ppm in the 13C NMR
spectrum of 3c can also be attributed to the presence of an
Fig. 3 Molecular structure of 2b with thermal ellipsoids drawn at
50% probability. The ortho-methyl groups and the hydrogen atoms on
the phenyl substituents have been omitted for clarity.
spectroscopy) of a bright orange/red colour. In one instance,
X-ray quality crystals obtained as a minor product from the
small-scale reaction of 1b with KHMDS were determined to
be 1,3,10,30-tetrakis(2,6-dimethylphenyl)-[2,20]diimidazoli-dinyl-
4,5,40,50-tetraone 2b (Fig. 3). Compound 2b is the product
of the reductive coupling of two 1b units, supporting the
hypothesis that redox processes are involved in the decom-
position of compounds 1a–c in the presence of bases. Un-
fortunately, no reliable synthetic procedure for the
preparation of 2b could be devised. The structure of this
derivative (Tables 1 and 2) is composed of two units of 1b
coupled via the sp3-hybridized carbon center of the hetero-
cycle. The relatively long length of this newly-formed C–C
bond (1.579(2) A) may be due to the presence of the sterically-
demanding 2,6-dimethylphenyl substituents on the hetero-
cyclic rings. The distorted tetrahedral geometries about the
coupled carbon centers, as evidenced by the average sum of the
bond angles of 332.29(13)1, as well as the length of the C–C
bond, clearly demonstrate that the dimer 2b arises as a result
of the reductive coupling of two units of 1b, and not the
dimerization of two intermediate carbene centers. The metric
parameters of the two heterocyclic units of 2b are not
significantly different from those found in the analogous
chloro-substituted monomer 1a. The two heterocyclic units
of 2b are twisted relative to one another (dihedral angle
between the planes containing the heterocyclic rings:
63.01(9)1), likely as a means of alleviating the steric strain
between the bulky N-substituents. The observation of reductive
coupling product 2b clearly supports the fact that redox
processes occur during the decompositions observed upon
reaction of compounds 1 with strong bases such as KHMDS.
The likely oxidation by-product of this reaction is the
bis(trimethylsilyl)aminyl radical.24 The presence of radical
species in these reactions probably accounts for the bright
colors observed during the reactions, as well as the complex
mixtures of products obtained from them.
1
acetate group in this species. Based on the H and 13C NMR
spectra of 3c, the overall symmetry of the compound does not
appear to be different than that observed in 1c. The above
results indicate that 1c undergoes a simple metathesis reaction
in the presence of mercury(II) acetate to give 1,3-bis(2,6-
diisopropylphenyl)-4,5-dioxoimidazolidin-2-yl acetate 3c
(Scheme 2). The earlier reported N-phenyl-substituted chloro
compound G is known to react with acetic acid to give a
1
similar acetate ester whose H and 13C NMR spectra match
the spectra obtained for compound 3c.19
Crystals of 3c suitable for X-ray diffraction were obtained
by slow evaporation of the solvent from a CH2Cl2 solution of
3c, and allowed for unambiguous confirmation of its structure
(Tables 1 and 2). The structure of 3c (Fig. 4) clearly shows that
1c undergoes a metathesis reaction in which the 2-chloro
substituent is replaced by an acetate group with a C–O single
The proposed reductive decomposition of compounds 1a–c
in the presence of strong bases is supported by the
cyclic voltammetry (CV) of these compounds. The cyclic
voltammograms reveal that compounds 1a–c are prone
to irreversible reduction (1a, b, c: Ec
=
ꢂ1.992,
Scheme 2 Attempted in situ deprotonation and metallation of 1c with
ꢂ1.987, and ꢂ2.293 V vs. Fc/Fc+ in THF). For comparison,
mercury(II) acetate.
ꢀc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2010 New J. Chem., 2010, 34, 1295–1308 | 1299

View Article Online
Scheme 3 Chloride abstraction from neutral 2-chloro-1,3-bis(aryl)-
imidazolidine-4,5-diones 1.
Fig. 4 Molecular structure of 3c with thermal ellipsoids drawn at
50% probability. The methyl groups and the hydrogen atoms on the
2,6-diisopropylphenyl substituents have been omitted for clarity.
19F NMR spectrum of 5a closely resemble those of their
precursors 1a–c indicating that these products have similar
symmetry. While the heterocyclic ring protons in the 1H NMR
spectra of compounds 1a–c display singlet resonances,
doublets signals are observed in the 1H NMR spectra of
5a (6.89 ppm) and 5b (6.64 ppm). New doublets are also
observed in the 19F NMR spectra of these compounds
(d(5a) = ꢂ113.57 ppm and d(5b) = ꢂ116.95 ppm). Once again,
the electron-withdrawing nature of the pentafluorophenyl sub-
stituents in 5a results in deshielded signals in the NMR spectra
of 5a relative to those observed for 5b. The observed coupling
bond length of 1.430(7) A. This substitution does not signifi-
cantly alter the interatomic bond distances observed within the
heterocycle in 3c relative to the structurally characterized
2-chloro species 1a. The marginal lengthening of the exocyclic
N–Cipso bonds on going from the chloro-substituted species 1a
(N–C = 1.417(4) A) to the acetate-substituted species 3c
(average N–C = 1.443(7) A) is likely due to the additional
steric strain caused by the much bulkier diisopropylphenyl
substituents. The dihedral angles between the heterocycle and
the aryl substituents also increase significantly in 3c (average
dihedral angle = 81.6(2)1) relative to 1a (dihedral angle =
66.3(2)1), once again allowing for a more efficient relief of the
steric strain associated with the bulky diisopropylphenyl
N-substituents.
1
constants in both species suggest that the doublets in the H
and 19F NMR spectra arise as the result of the replacement of
the chloro-substituent in 1a and 1b with a fluoro-substituent in
the products 5a and 5b. Additionally, closer inspection of
the new signal observed at ꢂ113.57 ppm in the 19F NMR
spectrum of 5a reveals it to be a doublet of pentets, owing to
coupling with the one geminal proton (2J(F,H) = 80.4 Hz),
and the four ortho-fluorine nuclei of the extraannular
perfluorophenyl substituents (5J(F,F) = 4.7 Hz). Both 5a and
5b are very poorly soluble, even in polar organic solvents such
as THF and CH2Cl2. The poor solubility of these compounds,
and the presence of fluorine substituents prevented the acquisi-
tion of satisfactory 13C NMR spectra of 5a and 5b. However,
peaks corresponding to the molecular ions were observed in the
mass spectra of both species. Additionally, all other signals
NHCs and their most common precursors, imidazolium-
type salts, display an sp2-hybridized diaminocarbon center
which benefits from the stabilizing p-interaction with the lone
pairs on the neighbouring nitrogen centers.4 Given the
observed structures of heterocycles 1a–c, and their reactivity
in the presence of bases and trapping agents such as mercury(II)
acetate, it seems clear that the diaminocarbon centers in
compounds 1a–c prefer to adopt an sp3-hybridization in which
p-interactions of the type described above are not possible.
The prospect of generating imidazolium salts with an oxalamide
framework via abstraction of the chloride substituent in
compounds 1a–c was also investigated, in the hope that
imidazolium salts isolated in this manner would display more
typical reactivity in the presence of bases, and allow for the
isolation of the desired family of carbenes.
Using silver(I) salts of weakly-coordinating anions,27 we
sought to exploit the high lattice energy of AgCl to drive the
abstraction of the chloro-substituent from compounds 1a–c
and isolate salts of imidazolium-4,5-dione cations 4a–c. Given
the observation that chloro-substituted molecular compounds
1a–c were formed preferentially to imidazolium-4,5-dione (F)
chloride salts in the reaction of oxalyl chloride with formamidines,
weakly coordinating anions were employed to favor forma-
tion of the desired salts, rather than simple metathesis
reactions as already observed in the formation of acetate-
substituted 3c. A white precipitate presumed to be AgCl was
observed instantaneously upon addition of silver(I) hexa-
fluorophosphate (AgPF6) to solutions of 1a and 1b in CH2Cl2
with concomitant formation of 5a and 5b (Scheme 3). The
1H NMR spectra of products 5a and 5b, as well as the
1
observed in the H and 19F NMR spectra of these compounds
are consistent with the proposed structures 5a–c. These com-
pounds also appear to have low stabilities. White solid 5a for
instance, becomes brown upon standing overnight, even under
an inert argon atmosphere.
Once isolated as solids, both compounds fail to completely
redissolve in CH2Cl2, possibly due to partial decomposition
upon isolation. By redissolving 5a in CH2Cl2, and removing
the insoluble solid by filtration, crystals suitable for X-ray
structural analysis could be grown by slow evaporation of the
filtrate. The obtained structural data (Table 1) clearly demon-
strates that the chloride substituent of 1a has been replaced by
a fluoride in 5a (C–F = 1.374(8) A) (Fig. 5). All of the other
metrical parameters of heterocycle 5a (Table 2) are essentially
identical to those observed in the isomorphous 1a. Like 1a, 5a
is also arranged in a linear supramolecular head-to-tail
arrangement which allows for an intermolecular interaction
between the heterocyclic proton and both backbone oxygens
of the neighboring molecule (Hꢃ ꢃ ꢃO = 2.61(5) A). The
observed interactions may be at least partially responsible
for the poor solubility observed for compounds 5a–c.
ꢀc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2010
1300 | New J. Chem., 2010, 34, 1295–1308

View Article Online
the fluorinated decomposition product 5c can be observed in
the NMR spectra of 4c. Upon standing overnight, complete
conversion to 5c occurs according to NMR spectroscopy.
Thus far, attempts to isolate the imidazolium-4,5-dione 4c
have failed, resulting instead in the isolation of mixtures
containing both 4c and its decomposition product 5c in varying
proportions. While these results indicate that imidazolium-
4,5-diones may be observable species given sufficient kinetic
stabilization, they are clearly highly electrophillic and thus,
unstable systems. Based on the obtained results, the oxalamide
framework imposes a clear preference for an sp3-hybridization
about the diaminocarbon center of the heterocycles examined.
Given that novel carbenes could not be synthesized simply
by deprotonating 1a–c with bases and that chloride abstraction
from 1a–c failed to generate stable imidazolium precursors, the
electronic structures of the targeted NHCs with an oxalamide
backbone were examined using computational methods. The
results of density functional calculations performed for the
parent system substantiate the conclusions drawn from experi-
mental analyses. The novel carbene has a p-type LUMO and a
HOMO–LUMO gap of 3.97 eV which is considerably smaller
than that calculated for the parent imidazolin-2-ylidene
(7.14 eV). In addition, the energy of the LUMO is significantly
negative (ꢂ2.42 eV) and the orbital has a major contribution
from the carbene carbon as well as from the two carbonyl
functionalities (Fig. 6). We note that in the parent imidazolin-
2-ylidene, an orbital with similar p-bonding characteristics is
the LUMO + 2, residing 7.5 eV above the HOMO. Further-
more, a bonding analysis using electron localization function
reveals that in the systems examined herein, the N-lone pairs
seem to be somewhat more ‘‘localized’’ than in the common
imidazolium cations in which they are involved in stabilizing
p-interactions with the sp2-hybridized diaminocarbon center.
Taken together, the computational analyses clearly confirm
that the carbene carbon center in oxalamide-based systems
is considerably more electrophilic than the diaminocarbon
center in more conventional NHCs.
Fig. 5 Molecular structure of 5a with thermal ellipsoids drawn at
50% probability. Atoms related by the mirror symmetry (ꢂx, y, z) are
marked with prime.
The most likely mechanism for the formation of fluoro-
compounds 5a–c would proceed by initial abstraction of the
chloride, and concomitant formation of an electron-poor
imidazolium-4,5-dione hexafluorophosphate intermediate,
4a–c. The unstable imidazolium-4,5-dione would then abstract
a fluoride from the non-coordinating counterion to generate
the observed fluoro-compounds 5a–c. The abstraction of
fluoride from weakly-coordinating anions such as hexafluoro-
phosphate in the presence of electrophillic species is well-known,28
and results in the formation of by-products such as PF5 and
PF3.29 31P NMR spectroscopy of the reaction mixtures reveals
PF3 to be the major by-product of this reaction. If the chloride
abstraction is performed in THF rather than CH2Cl2, AgCl
precipitation is still observed, however, solvent polymerization
occurs a short time later. It is not known whether this
THF-polymerization is due to the presence of an electrophillic
imidazolium-4,5-diones 4a–c or the presence of the fluorine-
containing by-products of fluoride abstraction, although the
polymerization of THF by PF5 has been reported previously.28
Similar results are obtained when silver(I) tetrafluoroborate is
used as the chloride abstracting agent, while complex mixtures
are generated when silver(I) tetraphenylborate is used.
Most interestingly, when chloride abstraction with AgPF6 is
attempted with the more sterically-shielded N-diisopropyl-
phenyl-substituted 1c, in addition to the immediate precipitation
of AgCl, the typically clear solution becomes bright orange.
1H NMR characterization of the reaction mixtures immediately
following the addition of AgPF6 reveals a highly deshielded
peak at 10.58 ppm in the 1H NMR spectrum. This was
assigned to the heterocyclic ring proton of 4c, an intermediate
highly electron-poor imidazolium-4,5-dione, kinetically protected
by the presence of bulky N-diisopropylphenyl substituents.
The highly-symmetric nature of this heterocycle is evidenced
by the presence of only one septet in the methine region of the
1H NMR spectrum of 4c. This observation indicates that upon
chloride abstraction the tetrahedral geometry about the diamino-
carbon center of 1c rearranges to give an sp2-hybridized
carbon center consistent with that expected for an imidazolium-
4,5-dione such as 4c. As a result, the two diastereotopic faces
of 1c become equivalent in 4c. The change in geometry about
the diaminocarbon center allows for a more complete p-
delocalization in the heterocycle, and the observed orange
color is likely due to a lowering of the heterocycle’s
HOMO–LUMO gap into the visible region. Imidazolium salts
with an oxalamide framework appear to be highly unstable,
and even after only 5 min reaction time at room temperature,
The identity of the amino substituents does not seem to have
a considerable influence on the key molecular properties of
oxalamide-based carbenes. For example, the calculated
Fig. 6 Frontier molecular orbitals of imidazolin-2-ylidene (left) and
imidazolidin-4,5-dioxo-2-ylidene (right).
ꢀc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2010 New J. Chem., 2010, 34, 1295–1308 | 1301

View Article Online
HOMO–LUMO gaps for –NPh and –NC6F5 substituted
species do not differ markedly from the value reported for
the parent system, and in all cases the energy of the LUMO
remains considerably negative. Consequently, NHCs bearing
an oxalamide backbone are extremely electron-poor and thus
the possibility of their isolation is questionable. This is rather
unfortunate given that their p-system gives rise to a LUMO
whose spatial characteristics would render the carbenes superior
p-acceptors. The above is also true of imidazolium-4,5-diones
as they share a common oxalamide framework. The cations
are almost as p-electrophilic as the carbenes they give rise to
via deprotonation. For example, the parent imidazolium-4,5-
dione with hydrogens as N-substituents has a HOMO–LUMO
gap of 4.76 eV as well as a low-lying p-LUMO whose
morphology matches that of the corresponding carbene.
The enhanced reactivity of imidazolium-4,5-diones is nicely
demonstrated experimentally by the ability of the observed
transient 4c to abstract fluoride from weakly-coordinating
anions such as hexafluorophosphate and tetrafluoroborate.
Such behaviour is not observed with traditional imidazolium
salts.
methoxy-substituent. As observed in the fluoro-compound 5c,
the signals coresponding to the diaminocarbon in the
13C NMR spectra of the methoxy derivatives (6a–c: d = 92.0,
96.9, and 99.7 ppm, respectively) are downfield shifted relative
to the chloro-precursors 1 (1a–c: d = 83.7, 86.9, and 88.9 ppm,
respectively) due to the relative electronegativities of the
neighboring oxygen and chlorine centers. Consistent with the
relative electronegativities of fluorine and oxygen, these
downfield shifts are of a smaller magnitude for the methoxy
derivatives 6a–c than those observed in the fluoro-system 5c
(d = 101.9 ppm). Similar to compounds 1b and 1c, the
1H NMR spectra of derivatives 6b, and 6c display two and
four methyl resonances, respectively. The 19F NMR spectrum
of 6a is also remarkably similar to that of its precursor 1a, with
only slight shifts observed. All-together, these observations
indicate that both the chloro- and the methoxy derivatives
1a–c and 6a–c possess similar geometries featuring a four-
coordinate carbon center within the heterocyclic ring. The
tetracoordinate nature of the diaminocarbon center in methoxy-
substituted 6b was unambiguously comfirmed by X-ray
crystallographic characterization (Table 1, Fig. 7). Like the
N-diisopropylphenyl-substituted acetate derivative 3c, the
dihedral angles between the N-dimethylphenyl substitutents
of 6b and the heterocyclic skeleton (71.4(2)1 and 79.7(2)1) are
greater than those observed for the less bulky pentafluorophenyl-
substituted compound 1a (66.3(2)1). The dihedral angles
observed for 6b, however, are smaller than those observed in
the more sterically encumbered 3c (81.6(2)1), reflecting the
smaller nature of the 2,6-dimethylphenyl N-substituents relative
to the 2,6-diisopropylphenyl substituents in 3c. Examination
of the skeletal bond distances in 6b (Table 2) reveals that the
intraannular N–CCO bond is very slightly shortened (average
N–C = 1.352(5) A), while the carbonyl C–O bond distance
is marginally lengthened (C–O = 1.216(4) A) relative to
those observed in the chloro-substituted derivative 1a
(N–C = 1.372(4) A and C–O = 1.200(4) A). This observation
is consistent with the more electron-donating nature of
the methoxy-substituent relative to the chloro-substituent.
Consequently, the more electron-donating methoxy-group
favours resonance form H0 to a greater degree than the
chloro-derivative. This trend is correlated to the IR spectra
of the methoxy-substituted derivatives 6a–c. The CO stretch in
While the preparation of NHCs from precursors containing
an sp2-hybridized diaminocarbon center (deprotonation of
imidazolium salts, reduction of thiourea derivatives etc.) is
common, their preparation from systems containing a four-
coordinate tetrahedral diaminocarbon center is also possible.30
Four-coordinate NHC-alcohol adducts are known to behave
as ‘‘protected carbenes’’, releasing the alcohol by thermolysis
to generate the free unprotected carbenes in situ.31 Since the
preparation of 2-methoxy-1,3-diphenylimidazolidine-4,5-
dione was reported,19 we sought to prepare the methoxy
adducts 6a–c in order to investigate their utility as potential
carbene sources. After all, the computational analyses
showed that the target carbenes are stable species on the
potential energy hypersurface, though their chemical stability
is compromised by high electrophilicity.
Refluxing THF solutions of 1a–c overnight with excess
methanol (10 equivalents) gave high yields (84–95%) of the
desired methoxy-substituted derivatives 6a–c (Scheme 4).
Upon methoxy substitution, the heterocyclic ring proton is
shifted upfield in the 1H NMR spectra of these compounds
(6a–c: d = 6.83, 6.12, and 5.73 ppm, respectively) relative to
the chloro-substituted precursors 1a–c, consistent with the
relative inductive effects of methoxy- and chloro-substituents.
As observed in the derivatives discussed above, 6a bearing the
most electron-withdrawing N-substituents exhibits the most
deshielded signal, while 6c bearing the most electron-donating
N-substituents exhibits the most shielded signal. An additional
methyl signal is also observed in the 1H NMR and 13C NMR
spectra of compounds 6a–c, establishing the presence of the
these compounds (6a–c: n = 1781, 1748, and 1755 cmꢂ1
)
Fig. 7 Molecular structure of 6b with thermal ellipsoids drawn at
50% probability. All hydrogen atoms on the 2,6-dimethylphenyl
substituents have been omitted for clarity.
Scheme
4 Synthesis of 2-methoxy-1,3-bis(aryl)imidazolidine-
4,5-diones 6.
ꢀc
1302 | New J. Chem., 2010, 34, 1295–1308 This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2010

View Article Online
occurs at a lower energy than observed for chloro-compounds
1a–c (n = 1784, 1773, and 1772 cmꢂ1), indicating that
resonance form H0 is indeed a greater contributor to the
structures of compounds 6a–c, than it is to compounds 1a–c.
Despite their resemblance to other four-coordinate alcohol
adducts of NHCs, methoxy adducts 6a–c were stable under a
variety of thermolytic conditions. In the solid state, they could
be heated to 200 1C under dynamic vacuum (1 ꢄ 10ꢂ3 Torr)
overnight with no sign of methanol release, or decomposition.
The compounds were also stable to thermolysis upon heating
up to 80 1C for several days in solution. The thermolysis of
compounds 6a–c was also attempted in the presence of several
transition metal complexes known to form NHC–metal
complexes in the presence of carbene sources. It was hoped
that the formation of a metal–carbene bond would provide an
additional enthalpic driving force, allowing for the loss of
methanol from adducts 6a–c, and the formation of the desired
NHCs trapped by metal complexes. Grubbs and co-workers
have reported the trapping of NHCs generated by thermolysis
of methoxy-adducts with their first-generation catalyst system,31b
however, the methoxy adducts described in the current manuscript
fail to react under the conditions reported (80 1C), even upon
heating for several days. They are equally stable under similar
conditions in the presence of [Rh(cod)Cl]2, another common
NHC trapping agent.31a Since all methoxy-adducts 6a–c,
regardless of the nature of the N-substituents, fail to undergo
thermolysis to release methanol and the desired NHCs, we
conclude that five-membered heterocycles bearing an oxalamide
framework, and a four-coordinate diaminocarbon center are
ill-suited to serve as NHC precursors regardless of the chemical
methodology employed.
Scheme 5 Attempted reduction of 2-chloro-1,3-bis(2,6-diisopropyl-
phenyl)imidazolidine-4,5-diones 1.
displays only two signals in the methyl region, and one
methine signal indicating that the faces of the heterocycle
are no longer diastereotopic. The reduction of the N2C(H)–Cl
bond to form an N2CH2 center would account for the
increased symmetry of 7c relative to its precursor 1c. This
hypothesis is supported by the emergence of a signal assigned
to the new N2CH2 center (d = 4.27 ppm, 2 H), and the
disappearance of the N2C(H)Cl signal at 7.18 ppm (1 H). No
additional peaks are observed in the hydroxyl, or aliphatic
region of the 1H NMR spectrum of 7c suggesting that reduction
has not occurred at the carbonyl moieties. Longer reaction
times, and greater excesses of trichlorosilane only resulted in
decomposition to complex mixtures. These results demonstrate
the stability of the oxalamide framework present in this
imidazolidine-4,5-dione system, and indicate that the C–Cl
bond at the four-coordinate diaminocarbon is likely the most
reactive site.
Conclusions
The [3 + 2] ring-closing reaction of formamidines with oxalyl
chloride was investigated in an attempt to prepare NHCs
bearing N-pentafluorophenyl substituents. The resulting novel
products were found to display unanticipated, molecular
2-chloro-substituted heterocyclic structures 1a–c, containing
a four-coordinate diaminocarbon center instead of the planar
tricoordinate center expected for imidazolium-4,5-diones 4a–c.
Abstraction of the chloride substituent from 1a–c appears to
result in the transient formation of unstable imidazolium-
4,5-diones 4a–c. In the presence of sterically-demanding
N-diisopropylphenyl substituents, the more stable intermediate
imidazolium-4,5-dione 4c could be observed spectroscopically,
however, this intermediate featuring a highly electron-poor
sp2-hybridized diaminocarbon center decomposed rapidly.
The high electrophilicity of the diaminocarbon center in
intermediates 4a–c, which is responsible for their low stability,
results from a p-framework which extends to the two carbonyl
functionalities of the heterocyclic backbone, and gives rise to a
very low-lying LUMO with a major contribution from the
diaminocarbon center. This enhanced electrophilicity accounts
for the observed reactivity of compounds 1a–c, which failed to
form stable carbenes in the presence of a variety of bases, and
trapping agents. Indeed, metathetical reactions of 1a–c involving
mercury(II) acetate or methanol demonstrated that these
compounds exhibit a very strong preference for a four-coordinate
geometry about the diaminocarbon center, displaying reactivity
typical of alkyl halides. While four-coordinate NHC methoxy
adducts have been used as carbene sources in certain instances,
the methoxy adducts reported herein are very stable with
regard to thermolysis, even in the presence of metal centers.
Given that the heterocyclic systems 1a–c and 6a–c discussed
above appear to be poorly-suited to serve as NHC-precursors
as a result of the carbonyl moieties that they possess, methods
by which these moieties could be reduced were investigated. In
fact, appropriately-substituted oxalamides are frequently
reduced using borane reagents in the synthesis of ethyl-
enediamines that are subsequently used in the [4 + 1] ring-
closing step of imidazolium preparation.32 To this end we
investigated the reaction of pentafluorophenyl-substituted
chloro-compound 1a with a variety of reducing agents, both
in stoichiometric quantities and in excess. No reaction was
observed in the presence of reducing agents such as BH3ꢃTHF,
and lithium metal, whereas complete decomposition was
observed in the presence of stronger reducing agents such as
lithium aluminium hydride. Similar reactivity was observed
for both 1b and 1c which do no bear pentafluorphenyl
N-substituents, suggesting once again that the reactivity of
these compounds is dominated by the oxalamide backbone,
and not the substitution at nitrogen. Attempted reductions
with trichlorosilane required the use of toluene as a solvent to
prevent polymerization of THF. This limited the scope of this
reduction to N-diisopropylphenyl-substituted 1c since 1a and
1b were not sufficiently soluble in toluene. Refluxing 1c and
trichlorosilane in toluene overnight appears to result in the
reduction of the N2C–Cl bond, rather than of the carbonyl
groups of the backbone to give 1,3-bis(2,6-diisopropylphenyl)-
imidazolidine-4,5-dione 7c (Scheme 5). Like imidazolium-4,5-
1
dione 4c, the H NMR spectrum of the reduction product 7c
ꢀc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2010 New J. Chem., 2010, 34, 1295–1308 | 1303

View Article Online
The similar reactivity observed for the above-discussed
compounds with a variety of N-substituents (pentafluorophenyl,
2,6-dimethylphenyl, and 2,6-diisopropylphenyl) demonstrates
that the oxalamide backbone fragment exercises a much greater
influence on the stability and reactivity of these compounds
than do the N-substituents. Computational analyses reveal
that the oxalamide backbone lowers the energy of the LUMO,
increasing the p-acceptor character of the resulting carbenes
and accomplishing part of the goal of the present study; the
magnitude of this enhancement, however, proved too great to
allow for the isolation of these compounds.
Non-fluorinated formamidines were prepared according to
reported procedures.36 N,N0-bis(pentafluorophenyl)formamidine
was prepared using a slight modification of a reported
procedure,17 which was very effective for the synthesis of
formamidines with a higher degree of fluorination but not
for their non-fluorinated analogues. In a typical trial, one drop
of concentrated hydrochloric acid was added to a neat
stoichiometric mixture of the aniline and triethylorthoformate.
Upon stirring for several minutes, the mixture solidified,
yielding the desired formamidines in near quantitative yields.
Oxalyl chloride was purchased from Sigma Aldrich and used
without further purification. All other reagents were purchased
from commercial suppliers. NMR spectra were collected on a
Bruker Advance DRY-400 spectrometer and calibrated with
respect to THF-d7 (1H, 3.58 ppm), THF-d8 (13C 67.57 ppm),
CDHCl2 (1H, 5.32 ppm), CD2Cl2 (13C, 54.00 ppm), toluene-d7
(1H, 2.09 ppm), and toluene-d8 (13C, 20.40 ppm). 19F NMR
spectra were collected on a Bruker Advance UGI-400 spectro-
meter and calibrated with respect to C6F6 (19F, ꢂ164.9 ppm).
Mass spectra and elemental analyses were performed by the
Analytical Instrumentation Laboratory, Department of
Chemistry, University of Calgary.
An interesting comparison is provided by the anionic
carbenes I described by Cesar et al., which are formally
´
obtained via insertion of a CRꢂ fragment between the backbone
carbon atoms of the 1,3-bis(diaryl)-4,5-dioxoimidazolidin-2-
ylidenes targeted in the present study.33 Given the competitive
electron delocalization over the anionic backbone in these
derivatives, which encompasses the carbonyl moieties, the
influence of the latter functionalities on the electronic properties
of the diaminocarbon centre is reduced. This is reflected by the
longer skeletal N–CCO bonds (generally 1.41–1.47 A) indicating a
reduced p-component. By comparison, the respective metric
parameters in the derivatives described herein measure
1.35–1.38 A, as expected for a more significant p-interaction
over the entire heterocyclic skeleton. The closely related
neutral carbene J, very recently reported by Hudnall and
Bielawski, as well as its amidinium precursor, also feature
long intraannular N–CCO bonds (1.44–1.46 A).34a The transient
N,N0-dimesityl substituted analog to J, as well as its metal
complexes, have also been reported, and the structure of its
chloride precursor featured an sp3-hybridized N2C(H)Cl
center akin to those observed in derivatives 1a–c described
Synthesis
2-Chloro-1,3-bis(pentafluorophenyl)imidazolidine-4,5-dione 1a.
Oxalyl chloride (1.62 g, 12.8 mmol) was added dropwise to a
stirred solution of N,N0-bis(pentafluorophenyl)formamidine
(4.80 g, 12.8 mmol) in THF (50 mL). After stirring for
30 min at room temperature, the volatiles were removed in
vacuo. The resulting solid was washed with hexane, and dried
to give 2-chloro-1,3-bis(pentafluorophenyl)imidazolidine-4,5-
dione 1a as an air-stable white powder. X-Ray quality crystals
were grown by slow evaporation from a CH2Cl2 solution.
Total yield: 5.01 g, 10.7 mmol, 85%. 1H NMR (400 MHz,
THF-d8): d 7.79 (1 H, s, N2ClCH). 19F NMR (377 MHz,
THF-d8): d ꢂ141.4 (2 F, br s, o-Ar–F), ꢂ146.8 (2 F, br s,
o-Ar–F), ꢂ152.0 (2 F, tt, 3J(F,F) = 20.7 Hz, 4J(F,F) =
3.8 Hz, p-Ar–F), ꢂ162.4 (2 F, br s, m-Ar–F), ꢂ162.7 (2 F,
br s, m-Ar–F). 13C{1H} (101 MHz, THF-d8): d 153.6 (s, CO),
145.7 (dm, 1J(C,F) = 256.8 Hz, o/m-C-Ar), 144.1 (dtt, 1J(C,F) =
257.0 Hz, 2J(C,F) = 18.0 Hz, 4J(C,F) = 4.5 Hz, p-C-Ar),
139.5 (dm, 1J(C,F) = 250.5 Hz, o/m-C-Ar), 108.7 (tm, 2J(C,F) =
14.8 Hz, NC-Ar), 83.7 (s, N2C(Cl)H). IR (KBr pellet)
n = 1784 cmꢂ1, vs., CQO). Anal. calc. for C15HN2O2F10Cl:
C, 38.61; H, 0.21; N, 6.00. Found: C, 38.66; H, 0.40; N, 5.97.
MS (ESI): m/z = 489 [M + Na]+.
herein.34b Lavigne, Ce
´
sar and Glorius and co-workers showed
that five-membered monoamidocarbenes K can be incorporated
in transition metal complexes.35 These derivatives feature
shorter intraannular N–CCO bonds (1.393(4) A) and longer
intraannular N–CC bonds (1.458(4) A). It appears clear that
the destabilizing influence of the amido functionalities on the
diamidocarbene moiety in the five-membered cyclic ligands
targeted in our study exceeds that observed in the diamido-
and aminoamidocarbene ligands I, J and K. While the
stabilization of NHCs with oxalamide backbone in the
coordination sphere of transition metals is not out of question,
the isolation of the highly electrophilic free ligands is highly
doubtful.
2-Chloro-1,3-bis(3,5-trifluoromethylphenyl)imidazolidine-4,5-
dione. Oxalyl chloride (136 mg, 1.07 mmol) was added dropwise
to a stirred solution of N,N0-bis(3,5-trifluoromethylphenyl)-
formamidine (500 mg, 1.07 mmol) in THF (10 mL). After stirring
for 30 min at room temperature, the volatiles were removed in
vacuo. The resulting solid was washed with hexane and dried to
give 2-chloro-1,3-bis(pentafluorophenyl)imidazolidine-4,5-dione
as an air-stable white powder. Total yield: 458 mg, 0.875 mmol,
The tuning of the p-accepting properties of NHCs with
six-membered frameworks of type I using N-fluorinated aryl
substituents is currently under investigation in our group.
Experimental section
1
82%. H NMR (400 MHz, THF-d8): d 8.61 (1 H, s, N2ClCH),
Unless otherwise noted, all operations were performed under
an argon atmosphere using standard Schlenk and glove box
techniques. Solvents were dried and deoxygenated prior to use.
8.47 (4 H, s, o-Ar–H), 8.13 (2H, s, p-Ar–H). 19F NMR
(377 MHz, THF-d8): d ꢂ65.67 (12 F, s, Ar–CF3). 13C{1H}
(101 MHz, THF-d8): d 154.9 (s, CO), 137.0 (s, ipso-C-Ar),
ꢀc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2010
1304 | New J. Chem., 2010, 34, 1295–1308

View Article Online
133.8 (q, 1J(C,F) = 34.2 Hz, CF3), 125.6 (s, 1J(C, o-C-Ar), 122.9
by slow cooling of a warm toluene solution did not result in
improved product purity. Total yield: 282 mg, 0.589 mmol,
89%. 1H NMR (400 MHz, THF-d8): d 7.44–7.41 (2 H, m,
Ar–H), 7.44 (1 H, s, N2C(OC(O)CH3)H), 7.32–7.29 (4 H, m,
3
(m, m-C-Ar), 121.9 (septet, J(C,F) = 3.5 Hz, p-C-Ar), 82.2 (s,
N2C(Cl)H). IR (KBr pellet) n = 1747 cmꢂ1, vs., CQO). MS
(EI): m/z = 558 [M]+.
3
Ar–H), 3.17 (2 H, septet, J(H,H) = 6.8 Hz, Ar–CH(CH3)2),
3.00 (2 H, septet, 3J(H,H) = 6.8 Hz, Ar–CH(CH3)2), 1.84
(3 H, s, N2C(OC(O)CH3)H), 1.37 (6 H, d, 3J(H,H) = 6.8 Hz,
2-Chloro-1,3,-bis(2,6-dimethylphenyl)imidazolidine-4,5-dione
1b. Oxalyl chloride (503 mg, 3.96 mmol) was added dropwise
to a stirred solution of N,N0-bis(2,6-dimethylphenyl)formami-
dine (1.0 g, 3.96 mmol) in THF (10 mL). After stirring for
30 min at room temperature, the volatiles were removed
in vacuo. The resulting solid was washed with hexane and dried
to give 2-chloro-1,3-bis(2,6-dimethylphenyl)imidazolidine-4,5-
dione 1b as an air-stable white powder. Total yield: 1.09 g, 3.18
Ar–CH(CH3)2), 1.30 (6 H, d, 3J(H,H)
Ar–CH(CH3)2), 1.23 (12 H, d, 3J(H,H)
=
6.8 Hz,
6.8 Hz,
=
Ar–CH(CH3)2). 13C{1H} (101 MHz, THF-d8):
d 169.7
(s, C(O)CH3), 157.6 (s, CO), 149.7 (s, Ar–C), 147.7
(s, Ar–C), 131.4 (s, Ar–C), 128.69 (s, Ar–C), 125.4
(s, Ar–C), 125.34 (s, Ar–C), 88.8 (s, N2C(OAc)H), 30.6
(s, Ar–CH(CH3)2), 30.2 (s, Ar–CH(CH3)2), 25.5
(s, Ar–CH(CH3)2), 25.4 (s, Ar–CH(CH3)2), 24.3 (s,
Ar–CH(CH3)2), 24.1 (s, Ar–CH(CH3)2), 20.8 (s, OC(O)CH3).
IR (KBr pellet) n = 1759 cmꢂ1, vs., CQO). MS (MALDI-TOF,
a-HCCA matrix): m/z = 501 [M + Na]+.
1
mmol, 80%. H NMR (400 MHz, THF-d8): d 7.66 (1 H, s,
N2ClCH), 7.30–7.27 (2 H, m, Ar–H), 7.24–7.20 (4 H, m,
Ar–H), 2.39 (6 H, s, Ar–CH3), 2.38 (6 H, s, Ar–CH3).
13C{1H} (101 MHz, THF-d8):
d 155.6 (s, CO), 139.3
(s, Ar–C), 136.9 (s, Ar–C), 132.1 (s, Ar–C), 130.6 (s, Ar–C),
130.1 (s, Ar–C), 129.9 (s, Ar–C), 86.9 (s, N2C(H)Cl), 19.6
(s, Ar–CH3), 18.3 (s, Ar–CH3). IR (KBr pellet) n = 1773, vs.,
CQO) cmꢂ1. Anal. calc. for C19H19N2O2Cl: C, 66.57; H, 5.59;
N, 8.17. Found: C, 66.47; H, 5.90; N, 8.10. MS (ESI): m/z =
307 [M ꢂ Cl]+.
2-Fluoro-1,3-bis(pentafluorophenyl)imidazolidine-4,5-dione 5a.
A
solution of silver(I) hexafluorophosphate (500 mg,
1.98 mmol) in CH2Cl2 (10 mL) was added to a stirred CH2Cl2
(10 mL) solution of 2-chloro-1,3-bis(pentafluorophenyl)imid-
azolidine-4,5-dione 1a (923 mg, 1.98 mmol). A white precipitate
crashed out immediately, and the reaction mixture was stirred
for an additional hour at room temperature before filtering off
the silver(I) chloride precipitate. The volatiles were then
removed from the filtrate under high vacuum to give the
product 5a as an unstable white solid. X-Ray quality crystals
of the product were grown by slow evaporation of a freshly
filtered CH2Cl2 solution of crude 5a. Upon standing overnight,
the product decomposes to an insoluble brown solid. Total
yield: 357 mg, 0.793 mmol, 40%. 1H NMR (400 MHz,
CD2Cl2): d 6.89 (1 H, d, 2J(H,F) = 80.4 Hz, N2CH(F)).
19F NMR (377 MHz, CD2Cl2): d ꢂ113.6 (1 F, dp, 2J(F,H)
2-Chloro-1,3,-bis(2,6-diisopropylphenyl)imidazolidine-4,5-dione
1c. Oxalyl chloride (696 mg, 5.49 mmol) was added dropwise
to
a
stirred solution of N,N0-bis(2,6-diisopropylphenyl)-
formamidine (2.0 g, 5.49 mmol) in THF (20 mL). After stirring
for 30 min at room temperature, the volatiles were removed
in vacuo. The resulting solid was washed with hexane, and dried
to give 2-chloro-1,3-bis(2,6-diisopropylphenyl)imidazolidine-
4,5-dione 1c as an air-stable white powder. Total yield:
2.35 g, 5.38 mmol, 94%. 1H NMR (400 MHz, THF-d8):
d 7.50–7.46 (2 H, m, Ar–H), 7.41–7.33 (4 H, m, Ar–H),
7.18 (1 H, s, N2ClCH), 3.23 (2 H, septet, 3J(H,H) =
6.8 Hz, Ar–CH(CH3)2), 2.89 (2 H, septet, 3J(H,H) =
6.8 Hz, Ar–CH(CH3)2), 1.38 (6 H, d, 3J(H,H) = 6.8 Hz,
5
= 80.4 Hz, J(F,F) = 4.7 Hz, N2CF(H), ꢂ143.0 (4 F, br s,
o-Ar–F), ꢂ149.2 (2 F, tt, 3J(F,F) = 20.7 Hz, p-Ar–F), ꢂ160.0
(4 F, br s, m-Ar–F). IR (KBr pellet) n = 1790 cmꢂ1, vs., CQO.
MS (EI): m/z = 450 [M]+.
Ar–CH(CH3)2), 1.31 (6 H, d, 3J(H,H)
Ar–CH(CH3)2), 1.26 (6 H, d, 3J(H,H)
Ar–CH(CH3)2), 1.18 (6 H, d, 3J(H,H)
=
=
=
6.8 Hz,
6.8 Hz,
6.8 Hz,
Ar–CH(CH3)2). 13C{1H} (101 MHz, THF-d8): d 157.1 (s,
CO), 150.1 (s, Ar–C), 147.3 (s, Ar–C), 131.6 (s, Ar–C), 129.0
(s, Ar–C), 126.1 (s, Ar–C), 125.2 (s, Ar–C), 88.9 (s, N2C(H)Cl),
30.6 (s, Ar–CH(CH3)2), 30.4 (s, Ar–CH(CH3)2), 25.7
(s, Ar–CH(CH3)2), 23.9 (s, Ar–CH(CH3)2), 23.8
(s, Ar–CH(CH3)2). IR (KBr pellet) n = 1772 cmꢂ1, vs.,
CQO. Anal. calc. for C27H35N2O2Cl: C, 71.27; H, 7.75; N,
6.15. Found: C, 70.22; H, 7.72; N, 6.02. MS (ESI): m/z = 455
[M+H]+.
2-Fluoro-1,3-bis(2,6-dimethylphenyl)imidazolidine-4,5-dione 5b.
NMR scale: To a solution of 2-chloro-1,3-bis(2,6-dimethyl-
phenyl)imidazolidine-4,5-dione 1b (30 mg, 0.088 mmol) in
CD2Cl2 (B1 mL) silver(I) hexafluorophosphate (22 mg,
0.088 mmol) was added. Upon shaking, a white-yellow precipitate
formed. After 20 min reaction time at room temperature, the
solids in the NMR tube were separated by centrifugation and the
product in solution was characterized by NMR. 1H NMR
(400 MHz, CD2Cl2): d 7.37–7.33 (2 H, m, p-Ar–H), 7.27–7.23
(4 H, m, m-Ar–H), 6.64 (1 H, d, 2J(H,F) = 79.2 Hz, N2CH(F)),
2.32 (6 H, s, Ar–CH3), 2.29 (6 H, s, Ar–CH3). 19F NMR
(377 MHz, CD2Cl2): d ꢂ116.9 (1 F, d, 2J(F,H) = 79.2 Hz,
N2CF(H)). MS(EI): m/z = 326 [M]+.
1,3-Bis(2,6-diisopropylphenyl)-4,5-dioxo-imidazolidin-2-yl
acetate 3c. Mercury(II) acetate (105 mg, 0.330 mmol) was
added to
a solution of 2-chloro-1,3-bis(2,6-diisopropyl-
phenyl)-4,5-imidazolidinedione 1c (300 mg, 0.660 mmol) in
THF (20 mL). The solution was stirred for an hour at room
temperature before the volatiles were removed under high
vacuum. The crude product was recrystallized from CH2Cl2
and washed with pentane, allowing for the isolation of the
product 3c as an air-stable white powder containing traces of
the mercury(II) chloride by-product. Further recrystallization
2-Fluoro-1,3-bis(2,6-diisopropylphenyl)imidazolidine-4,5-dione
5c and 1,3-bis(2,6-diisopropylphenyl)imidazolium-4,5-dione
hexafluorophosphate 4c. NMR-scale: To
a solution of
2-chloro-1,3-bis(diisopropylphenyl)imidazolidine-4,5-dione 1c
(30 mg, 0.066 mmol) in CD2Cl2 (B1 mL) silver(I) hexa-
fluorophosphate (17 mg, 0.066 mmol) was added. Upon
ꢀc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2010 New J. Chem., 2010, 34, 1295–1308 | 1305

View Article Online
shaking, the clear solution became bright orange due to
the formation of transient 1,3-bis-(2,6-diisopropylphenyl)-
imidazolium-4,5-dione hexafluorophosphate 4c, and a white
precipitate of silver(I) chloride was formed. The precipitate
was separated by centrifugation and the 1H NMR spectrum of
the sample was taken immediately. Upon standing overnight,
the sample decomposed quantitatively to 2-fluoro-1,3-bis(2,6-
dimethylphenyl)imidazolidine-4,5-dione 5c with a concomitant
4J(F,F) = 1.9 Hz, p-Ar–F), ꢂ162.4 (2 F, br s, m-Ar–F), ꢂ162.0
(2 F, br s, m-Ar–F). 13C{1H} (101 MHz, THF-d8): d 154.9
(s, CO), 145.2 (dm, 1J(C,F) = 259.5 Hz, o/m-C-Ar), 143.2
(dtt, 1J(C,F) = 255.5 Hz, 2J(C,F) = 13.6 Hz, 4J(C,F) =
1
5.0 Hz, p-C-Ar), 139.5 (dm, J(C,F) = 250.5 Hz, o/m-C-Ar),
109.6 (tm, 2J(C,F) = 14.8 Hz, NC-Ar), 92.0 (s, N2C(OCH3)H),
50.3 (s, N2C(OCH3)H). IR (KBr pellet) n = 1781 cmꢂ1, vs.,
CQO). Anal. calc. for C16H4N2O3: C, 41.58; H, 0.87; N,
6.06. Found: C, 41.43; H, 1.03; N, 5.98. MS (EI): m/z = 480
[M+NH4]+.
1
loss of its orange color. 4c: H NMR (400 MHz, CD2Cl2): d
10.58 (1 H, s, N2CH), 7.72–7.70 (2 H, m, p-Ar–H), 7.38–7.36
(4 H, m, m-Ar–H), 2.62 (4 H, septet, 3J(H,H) = 6.7 Hz,
Ar–CH(CH3)2), 1.32 (12 H, d, 3J(H,H)
Ar–CH(CH3)2), 1.22 (12 H, d, 3J(H,H)
=
6.7 Hz,
2-Methoxy-1,3-bis(2,6-dimethylphenyl)imidazolidine-4,5-dione
6b. Methanol (467 mg, 14.6 mmol) was added to a solution of
2-chloro-1,3-bis(2,6-dimethylphenyl)-4,5-imidazolidinedione 1b
(500 mg, 1.46 mmol) in THF (10 mL). The stirred solution was
refluxed in a Schlenk tube for 12 h and the volatiles were
subsequently removed under high vacuum. The obtained solid
was washed with pentane and isolated as an air-stable white
powder. X-Ray quality crystals of 6b were grown by slow
evaporation of the product in CH2Cl2. Total yield: 468 mg,
=
6.7 Hz,
Ar–CH(CH3)2). 19F NMR (377 MHz, CD2Cl2): ꢂ73.0
(6 F, br s, PF6ꢂ).
Isolated yield: A solution of silver(I) hexafluorophosphate
(500 mg, 1.98 mmol) in CH2Cl2 (10 mL) was added to a stirred
CH2Cl2 (10 mL) solution of 2-chloro-1,3-bis-(pentafluoro-
phenyl)imidazolidine-4,5-dione 1c (900 mg, 1.98 mmol). A
white precipitate crashed out immediately, and the solution
became bright orange due to the presence of transient 1,3-
bis(2,6-diisopropylphenyl)imidazolium-4,5-dione hexa-fluoro-
phosphate 4c. The reaction mixture was stirred overnight at
room temperature until the orange color had disappeared. At
this point, the silver(I) chloride precipitate was filtered off. The
volatiles were then removed from the filtrate under high
vacuum to give a light-orange crude product 5c. The crude
product was purified by extraction with toluene, producing a
white solid after removal of the solvent. X-Ray quality crystals
of the product 5c were grown by slow evaporation of a
CH2Cl2 solution. Total yield: 176 mg, 0.40 mmol, 20%.
1
1.38 mmol, 95%. H NMR (400 MHz, THF-d8): d 7.26–7.22
(2 H, m, Ar–H), 7.19–7.17 (4 H, m, Ar–H), 6.12 (1 H, s,
N2(OCH3)CH), 2.99 (3 H, s, OCH3), 2.35 (6 H, s, Ar–CH3),
2.33 (6 H, s, Ar–CH3). 13C{1H} (101 MHz, THF-d8): d 156.3
(s, CO), 139.7 (s, Ar–C), 136.9 (s, Ar–C), 133.4 (s, Ar–C), 130.1
(s, Ar–C), 129.6 (s, Ar–C), 129.5 (s, Ar–C), 96.9 (s, N2C(OCH3)H),
58.5 (s, N2C(OCH3)H), 18.9 (s, Ar–CH3), 18.3 (s, Ar–CH3). IR
(KBr pellet) n = 1748 cmꢂ1, vs., CQO). Anal. calc. for
C20H22N2O3: C, 70.99; H, 6.55; N, 8.28. Found: C, 70.72; H,
6.61; N, 8.26. MS (MALDI-TOF): m/z = 338 [M]+.
2-Methoxy-1,3-bis(2,6-diisopropylphenyl)imidazolidine-4,5-
1H NMR (400 MHz, toluene-d8):
d 7.21–7.17 (2 H,
dione 6c. Methanol (352 mg, 11.0 mmol) was added to a
m, p-Ar–H), 7.10–6.96 (4 H, m, m-Ar–H), 6.20 (1 H,
d, 2J(H,F) = 81.2 Hz, N2CH(F)), 3.30 (2 H, septet,
3J(H,H) = 6.8 Hz, Ar–CH(CH3)2), 2.83 (2 H, septet,
solution
of
2-chloro-1,3,-bis(2,6-diisopropylphenyl)-4,5-
imidazolidinedione 1c (500 mg, 1.10 mmol) in THF (10 mL).
The stirred solution was refluxed in a Schlenk tube for 12 h
and the volatiles were subsequently removed under high
vacuum. The resulting solid was washed with pentane yielding
6c that was isolated as an air-stable white powder. Total yield:
3
3J(H,H) = 6.8 Hz, Ar–CH(CH3)2), 1.26 (6 H, d, J(H,H) =
6.8 Hz, Ar–CH(CH3)2), 1.21 (6 H, d, 3J(H,H)
=
6.8 Hz, Ar–CH(CH3)2), 1.10 (6 H, d, 3J(H,H) = 6.8 Hz,
Ar–CH(CH3)2), 1.06 (6 H, d, 3J(H,H)
=
6.8 Hz,
1
458 mg, 1.02 mmol, 92%. H NMR (400 MHz, THF-d8): d
7.45–7.42 (2 H, m, Ar–H), 7.34–7.30 (4 H, m, Ar–H), 5.73
Ar–CH(CH3)2). 19F NMR (377 MHz, toluene-d8): d ꢂ114.8
2
(1 F, d, J(F,H) = 81.2 Hz, N2CF(H)). 13C{1H} (101 MHz,
3
(1 H, s, N2C(OCH3)H), 3.20 (2 H, septet, J(H,H) = 6.8 Hz,
toluene-d8): d 156.8 (d, 3J(C,F) = 3.0 Hz, CO), 149.4
(s, Ar–C), 146.4 (s, Ar–C), 131.0 (s, Ar–C), 127.9 (s, Ar–C),
125.1 (s, Ar–C), 124.4 (s, Ar–C), 101.9 (s, N2C(H)F), 29.8
(s, Ar–CH(CH3)2), 29.7 (s, Ar–CH(CH3)2), 24.9 (s, Ar–
CH(CH3)2), 24.7 (s, Ar–CH(CH3)2), 23.6 (s, Ar–CH(CH3)2).
IR (KBr pellet): n = 1772 cmꢂ1. MS (EI): m/z = 438 [M]+.
Ar–CH(CH3)2), 2.99 (2 H, septet, 3J(H,H)
= 6.8 Hz,
Ar–CH(CH3)2), 2.95 (3 H, s, N2C(OCH3)H), 1.36 (6 H, d,
3
3J(H,H) = 6.8 Hz, Ar–CH(CH3)2), 1.25 (6 H, d, J(H,H) =
6.8 Hz, Ar–CH(CH3)2), 1.23 (6 H, d, 3J(H,H) = 6.8 Hz,
Ar–CH(CH3)2), 1.20 (6 H, d, 3J(H,H) = 6.8 Hz, Ar–CH(CH3)2).
13C{1H} (101 MHz, THF-d8): d 157.4 (s, CO), 150.3 (s, Ar–C),
147.40 (s, Ar–C), 131.0 (s, Ar–C), 129.9 (s, Ar–C), 125.5
(s, Ar–C), 125.0 (s, Ar–C), 99.7 (s, N2C(OCH3)H), 59.8
(s, N2C(OCH3)H), 30.4 (s, Ar–CH(CH3)2), 30.4 (s, Ar–CH(CH3)2),
25.4 (s, Ar–CH(CH3)2), 25.3 (s, Ar–CH(CH3)2), 24.6
(s, Ar–CH(CH3)2), 23.9 (s, Ar–CH(CH3)2). IR (KBr pellet)
n = 1755 cmꢂ1, vs., CQO). Anal. calc. for C28H38N2O3: C,
74.63; H, 8.50; N, 6.22. Found: C, 74.20; H, 8.51; N, 6.12. MS
(MALDI-TOF): m/z = 473 [M + Na]+.
2-Methoxy-1,3-bis(pentafluorophenyl)imidazolidine-4,5-dione
6a. Methanol (172 mg, 5.38 mmol) was added to a solution of
2-chloro-1,3-bis(pentafluorophenyl)imidazolidine-4,5-dione 1a
(502 mg, 1.08 mmol) in THF (10 mL). The stirred solution
was refluxed for 12 h and the volatiles were subsequently
removed under high vacuum. The obtained solid was washed
with pentane yielding 6a as an air-stable white powder. Total
yield: 419 mg, 0.907 mmol, 84%. 1H NMR (400 MHz,
THF-d8): d 6.83 (1 H, s, N2(OCH3)CH), 3.30 (3 H, s, OCH3).
19F NMR (377 MHz, THF-d8): d ꢂ142.2 (2 F, br s, o-Ar–F),
ꢂ147.2 (2 F, br s, o-Ar–F), ꢂ154.2 (2 F, tt, 3J(F,F) = 20.7 Hz,
1,3-Bis(2,6-diisopropylphenyl)imidazolidine-4,5-dione 7c. NMR-
scale: To a solution of 2-chloro-1,3-bis(diisopropyl-phenyl)imid-
azolidine-4,5-dione 1c (20 mg, 0.044 mmol) in toluene-d8 (B1 mL),
ꢀc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2010
1306 | New J. Chem., 2010, 34, 1295–1308

View Article Online
trichlorosilane was added via syringe (100 mg, 0.74 mmol). The
solution was refluxed for 72 h, at which point it was characterized
by 1H NMR spectroscopy. 1H NMR (400 MHz, toluene-d8):
d 7.21–7.17 (2 H, m, Ar–H), 7.05–7.03 (4 H, m, Ar–H), 4.27 (2 H,
s, N2CH2), 2.90 (4 H, septet, 3J(H,H) = 6.8 Hz, Ar–CH(CH3)2),
1.17 (12 H, d, 3J(H,H) = 6.8 Hz, Ar–CH(CH3)2), 1.15 (12 H, d,
3J(H,H) = 6.8 Hz, Ar–CH(CH3)2).
J. N. Harvey, F. Paolini and J. Schutz, Angew. Chem., Int. Ed.,
¨
2004, 43, 5896–5911; (e) V. Nair, S. Bindu and V. Sreekumar,
Angew. Chem., Int. Ed., 2004, 43, 5130–5135; (f) D. Enders,
O. Niemeier and A. Henseler, Chem. Rev., 2007, 107, 5606–5655;
(g) O. Kuhl, Chem. Soc. Rev., 2007, 36, 592–607; (h) D. Tapu,
¨
D. A. Dixon and C. Roe, Chem. Rev., 2009, 109, 3385–3407;
(i) O. Schuster, L. Yang, H. G. Raubenheimer and M. Albrecht,
Chem. Rev., 2009, 109, 3445–3478.
6 (a) L. Jafarpour and S. P. Nolan, J. Organomet. Chem., 2001,
617–618, 17–27; (b) D. Enders and H. Gielen, J. Organomet.
Chem., 2001, 617–618, 70–80; (c) A. C. Hillier, G. A. Grasa,
M. S. Viciu, H. M. Lee, C. Yang and S. P. Nolan, J. Organomet.
Chem., 2002, 653, 69–82; (d) E. Peris and R. H. Crabtree, Coord.
Chem. Rev., 2004, 248, 2239–2246; (e) C. M. Crudden and
D. P. Allen, Coord. Chem. Rev., 2004, 248, 2247–2273;
Crystal structure determinations
Data were collected at a temperature of 173(2) K using the o
and j scans on a Nonius Kappa CCD diffractometer with
graphite monochromated Mo-Ka radiation with l = 0.71073 A
(Table 1), and corrected for Lorentz and polarization
effects and for absorption using the multi-scan method.37 The
structures were solved by the direct methods38 and expanded
using Fourier techniques.39 Because there were no atoms with
significant anomalous dispersion effects in 1a, 3c, and 5a, the
absolute structure could not be determined and the Friedel
reflections were merged before the final refinement.
Non-hydrogen atoms were refined anisotropically using
SHELXL97.40 Hydrogen atoms were included at geometrically
calculated positions during the refinement using the riding
model.
(f) V. Cesar, S. Bellemin-Laponnaz and L. H. Gade, Chem. Soc.
´
Rev., 2004, 33, 619–636; (g) N. M. Scott and S. P. Nolan, Eur. J.
Inorg. Chem., 2005, 1815–1828; (h) N-Heterocyclic Carbenes in
Synthesis, ed. S. P. Nolan, Wiley-VCH, Toronto, 2006; (i) N-
Heterocyclic Carbenes in Transition Metal Catalysis, ed.
F. Glorius, Topics in Organometallic Chemistry, vol. 21, Springer,
New York, 2007; (j) E. A. B. Kantchev, C. J. O’Brien and
M. G. Organ, Angew. Chem., Int. Ed., 2007, 46, 2768–2813;
(k) N. Marion and S. P. Nolan, Acc. Chem. Res., 2008, 41,
1440–1449; (l) S. Wuertz and F. Glorius, Acc. Chem. Res., 2008,
41, 1523–1533; (m) J. M. Praetorius and C. M. Crudden, Dalton
Trans., 2008, 4079–4094; (n) J. C. Y. Lin, R. T. W. Huang,
C. S. Lee, A. Bhattacharyya, W. S. Hwang and I. J. B. Lin, Chem.
Rev., 2009, 109, 3561–3598.
7 J. Huang, H.-J. Schanz, E. D. Stevens and S. P. Nolan, Organo-
metallics, 1999, 18, 2370–2375.
Computational details
8 (a) D. Nemcsok, K. Wichmann and G. Frenking, Organometallics,
2004, 23, 3640–3646; (b) M. D. Sanderson, J. W. Camplain and
C. W. Bielawski, J. Am. Chem. Soc., 2006, 128, 16514–16515;
(c) S. Diez-Gonzalez and S. P. Nolan, Coord. Chem. Rev., 2007,
251, 874–883; (d) D. M. Khramov, V. M. Lynch and
C. W. Bielawski, Organometallics, 2007, 26, 6042–6049;
(e) A. Kausamo, H. M. Tuononen, K. E. Krahulic and
R. Roesler, Inorg. Chem., 2008, 47, 1145–1154; (f) M. Srebro and
A. Michalak, Inorg. Chem., 2009, 48, 5361–5369.
9 C. A. Tolman, J. Am. Chem. Soc., 1970, 92, 2953–2956.
10 D. M. Khramov, E. L. Rosen, J. A. V. Er, P. D. Vu, V. M. Lynch
and C. W. Bielawski, Tetrahedron, 2008, 64, 6853–6862.
11 (a) L. Xu, W. Chen, J. F. Bickley, A. Steiner and J. Xiao,
J. Organomet. Chem., 2000, 598, 409–416; (b) S. Burling,
M. F. Mahon, S. P. Reade and M. K. Whittlesey, Organometallics,
2006, 25, 3761–3767.
12 K. Randell, M. J. Stanford, G. J. Clarkson and J. P. Rourke,
J. Organomet. Chem., 2006, 691, 3411–3415.
13 T. Ritter, M. W. Day and R. H. Grubbs, J. Am. Chem. Soc., 2006,
128, 11768–11769.
14 G. C. Vougioukalakis and R. H. Grubbs, Chem.–Eur. J., 2008, 14,
7545–7556.
15 A. J. Arduengo III, R. Krafczyk, R. Schmutzler, H. A. Craig,
J. R. Goerlich, W. J. Marshall and M. Unverzagt, Tetrahedron,
1999, 55, 14523–14534.
The theoretical calculations were performed using the Gaussian
program package.41 The geometry optimizations were carried
out at the PBE1PBE level.42 The Gaussian basis sets were
those reported by the Ahlrichs group (TZVP).43 The nature of
stationary points found was verified by performing frequency
calculations Orbital plots were obtained by the program
gOpenMol.44
Acknowledgements
This work was supported by the Natural Sciences and
Engineering Research Council of Canada, the Canada
Foundation for Innovation, the Academy of Finland and
the Alberta Science and Research Investments Program.
TDF thanks the NSERC for a CGS-D scholarship, JBG
thanks CONACYT for a PhD scholarship, and CJK thanks
the NSERC for two USRA scholarships.
Notes and references
16 M. G. Hobbs and R. Roesler, unpublished results. The
unsymmetrically N-substituted imidazolium salt containing one
pentafluoropheynl substituent and one mesityl substituent used
as a precursor to D is prepared via the [4 + 1] ring closing reaction
of an ethylenediamine hydrochloride and triethylorthoformate.
The presence of one alkylphenyl-substituted N-substituent,
however, appears to be required in order for the ring-closing
1 (a) D. Bourissou, O. Guerret, F. P. Gabbaı and G. Bertrand,
¨
Chem. Rev., 2000, 100, 39–92; (b) Carbene Chemistry, ed.
G. Bertrand, Marcel Dekker, New York, 2002; (c) F. Ekkehardt
Hahn and M. C. Jahnke, Angew. Chem., Int. Ed., 2008, 47,
3122–3172; (d) K. Hirai, T. Itoh and H. Tomioka, Chem. Rev.,
2009, 109, 3275–3332; (e) J. Vignolle, X. Cattoe
D. Bourissou, Chem. Rev., 2009, 109, 3333–3384.
¨
n and
reaction described to take place14
17 H. E. Abdou, A. A. Mohamed, J. M. Lopez-de-Luzuriaga and
´
.
2 (a) H.-W. Wanzlick and ESchikora, Angew. Chem., 1960, 72, 494;
(b) H.-W. Wanzlick and E. Schikora, Chem. Ber., 1961, 94,
2389–2393.
3 A. J. Arduengo III, R. L. Harlow and M. Kline, J. Am. Chem.
Soc., 1991, 113, 361–363.
J. P. Fackler, Jr., J. Cluster Sci., 2004, 15, 397–411.
18 H. Friebolin, Basic One- and Two-Dimensional NMR Spectroscopy,
Wiley-VCH, Toronto, 3rd edn, 1998, pp. 305–311.
19 E. A. Barsa and R. Richter, J. Org. Chem., 1986, 51,
4483–4485.
20 S. Csihony, D. A. Culkin, A. C. Sentman, A. P. Dove,
R. M. Waymouth and J. L. Hedrick, J. Am. Chem. Soc., 2005,
127, 9079–9084.
4 D. A. Dixon and A. J. Arduengo III, J. Phys. Chem., 1991, 95,
4180–4182.
5 (a) W. A. Herrmann and C. Kocher, Angew. Chem., Int. Ed. Engl.,
¨
1997, 36, 2162–2187; (b) A. J. Arduengo III, Acc. Chem. Res., 1999,
32, 913–921; (c) W. A. Herrmann, Angew. Chem., Int. Ed., 2002,
41, 1290–1309; (d) R. W. Alder, M. E. Blake, L. Chaker,
21 R. Goddard, O. Heinemann and C. Kruger, Acta Crystallogr.,
¨
Sect. C: Cryst. Struct. Commun., 1997, 53, 1846–1850.
ꢀc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2010 New J. Chem., 2010, 34, 1295–1308 | 1307

View Article Online
22 K. Yamaguchi, G. Matsumura, N. Haga and K. Shudo, Acta
Crystallogr., Sect. C: Cryst. Struct. Commun., 1992, 48, 558–559.
23 P. Groth and O. Hassel, Acta Chem. Scand., 1962, 16, 2311.
24 J. C. Brand, B. P. Roberts and J. N. Winter, J. Chem. Soc., Perkin
Trans. 2, 1983, 261–270.
39 P. T. Beurskens, G. Admiraal, G. Beurskens, W. P. Bosman, R. de
Gelder, R. Israel and J. M. M. Smits, The DIRDIF-94 Program
System, Technical Report of the Crystallography Laboratory,
University of Nijmegen, The Netherlands, 1994.
40 G. M. Sheldrick, Acta Crystallogr., Sect. A: Found. Crystallogr.,
2008, 64, 112–122.
25 J. C. Garrison and W. J. Youngs, Chem. Rev., 2005, 105,
3978–4008.
41 M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria,
M. A. Robb, J. R. Cheeseman, J. A. Montgomery, Jr.,
T. Vreven, K. N. Kudin, J. C. Burant, J. M. Millam,
S. S. Iyengar, J. Tomasi, V. Barone, B. Mennucci, 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. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev,
A. J. Austin, R. Cammi, C. Pomelli, J. Ochterski, P. Y. Ayala,
K. Morokuma, G. A. Voth, P. Salvador, J. J. 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. Komaromi, R. L. Martin, D. J. Fox, T. Keith, M. A. Al-Laham,
C. Y. Peng, A. Nanayakkara, M. Challacombe, P. M. W. Gill,
B. G. Johnson, W. Chen, M. W. Wong, C. Gonzalez and
J. A. Pople, GAUSSIAN 03 (Revision C.02), Gaussian, Inc.,
Wallingford, CT, 2004.
26 H. W. Wanzlick and H.-J. Schonherr, Angew. Chem., Int. Ed.
¨
Engl., 1968, 7, 141–142.
27 (a) W. Beck and K. Sunkel, Chem. Rev., 1988, 88, 1405–1421;
¨
(b) S. H. Strauss, Chem. Rev., 1993, 93, 927–942; (c) I. Krossing
and I. Raabe, Angew. Chem., Int. Ed., 2004, 43, 2066–2090.
28 R. F. Jordan, W. E. Dasher and S. F Echols, J. Am. Chem. Soc.,
1986, 108, 1718–1719.
29 H. Ohtani, S. Ishimura and M. Kumai, Anal. Sci., 2008, 24, 1335–1340.
30 E. Peris, Top. Organomet. Chem., 2007, 21, 83–116.
31 (a) D. Enders, K. Breuer, G. Raabe, J. Runsink, J. H. Teles,
J.-P. Melder, K. Ebel and St. Brode, Angew. Chem., Int. Ed. Engl.,
1995, 34, 1021–1023; (b) M. Scholl, S. Ding, C. W. Lee and
R. H. Grubbs, Org. Lett., 1999, 1, 953–956.
32 G. Xu and S. R. Gilbertson, Org. Lett., 2005, 7, 4605–4608.
33 V. Cesar, N. Lugan and G. Lavigne, J. Am. Chem. Soc., 2008, 130,
´
11286–11287.
34 (a) T. W. Hudnall and C. W. Bielawski, J. Am. Chem. Soc., 2009,
131, 16039–16041; (b) V. Ce
Inorg. Chem., 2010, 361–365.
35 (a) L. Benhamou, V. Ce
´
sar, N. Lugan and G. Lavigne, Eur. J.
´
sar, H. Gornitzka, N. Lugan and
42 (a) J. P. Perdew, K. Burke and M. Ernzerhof, Phys. Rev. Lett.,
1996, 77, 3865–3868; (b) J. P. Perdew, K. Burke and M. Ernzerhof,
Phys. Rev. Lett., 1997, 78, 1396; (c) J. P. Perdew, M. Ernzerhof and
K. Burke, J. Chem. Phys., 1996, 105, 9982–9989; (d) M. Ernzerhof
and G. E. Scuseria, J. Chem. Phys., 1999, 110, 5029–5036.
G. Lavigne, Chem. Commun., 2009, 4720–4722; (b) A. T. Biju,
K. Hirano, R. Frohlich and F. Glorius, Chem.–Asian J., 2009, 4,
¨
1786–1789.
36 K. E. Krahulic, G. D. Enright, M. Parvez and R. Roesler, J. Am.
Chem. Soc., 2005, 127, 4142–4143.
37 Z. Otwinowski and W. Minor, Methods Enzymol., 1997, 276,
307–326.
43 (a) A. Schafer, C. Huber and R. Ahlrichs, J. Chem. Phys., 1994,
¨
100, 5829–5835.
44 (a) L. J. Laaksonen, J. Mol. Graphics, 1992, 10, 33–34;
(b) D. L. Bergman, L. Laaksonen and A. Laaksonen, J. Mol.
Graphics Modell., 1997, 15, 301–306.
38 A. Altomare, M. Cascarano, C. Giacovazzo and A. Guagliardi,
J. Appl. Crystallogr., 1993, 26, 343–350.
ꢀc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2010
1308 | New J. Chem., 2010, 34, 1295–1308