Phosphine- and pyridine-functionalized N-heterocyclic carbene methyl and allyl complexes of palladium. Unexpected regiospecificity of the protonation reaction of the dimethyl complexes

Article abstract of DOI:10.1021/om0608408

Square planar neutral dimethyl and cationic allyl complexes of palladium with the electronically nonsymmetric diphenylphosphinomethyl- and pyridyl-N-heterocyclic carbene ligands have been synthesized and characterized. The products from the protonation of the dimethyl complexes with 1 equiv of acid at low temperatures are monomethyl cations, the exact nature of which is dependent on the type of ligand; in pyridine-carbene complexes the Pd-Me bond cleaved is trans to the carbene, while for the phosphino-carbene complexes it is trans to the phosphine. Density functional calculations suggest that protonation in these complexes occurs directly at the methyl ligands and that the site of protonation determines the selectivity of Pd-Me cleavage. For the pyridine-carbene complexes there is a clear preference for protonation trans to the carbene. For phosphino-carbene complexes, however, the site of protonation depends on the steric bulk of the N-heterocyclic carbene ligand. Protonation trans to carbene is favored with small substituents (H, Me), but the bulky 2,6-Prⁱ₂C₆H₃ susbstituent induces protonation trans to the phosphine, as is seen experimentally.

Full text of DOI:10.1021/om0608408

Organometallics 2007, 26, 253-263
253
Articles
Phosphine- and Pyridine-Functionalized N-Heterocyclic Carbene
Methyl and Allyl Complexes of Palladium. Unexpected
Regiospecificity of the Protonation Reaction of the Dimethyl
Complexes
Andreas A. Danopoulos,*,† Nikolaos Tsoureas,† Stuart A. Macgregor,*,‡ and
Christopher Smith‡
School of Chemistry, UniVersity of Southampton, Highfield, Southampton, U.K. SO17 1BJ, and School of
Engineering and Physical Sciences, William Perkin Building, Heriot-Watt UniVersity,
Edinburgh, EH14 4AS, U.K.
ReceiVed September 14, 2006
Square planar neutral dimethyl and cationic allyl complexes of palladium with the electronically
nonsymmetric diphenylphosphinomethyl- and pyridyl-N-heterocyclic carbene ligands have been synthesized
and characterized. The products from the protonation of the dimethyl complexes with 1 equiv of acid at
low temperatures are monomethyl cations, the exact nature of which is dependent on the type of ligand;
in pyridine-carbene complexes the Pd-Me bond cleaved is trans to the carbene, while for the phosphino-
carbene complexes it is trans to the phosphine. Density functional calculations suggest that protonation
in these complexes occurs directly at the methyl ligands and that the site of protonation determines the
selectivity of Pd-Me cleavage. For the pyridine-carbene complexes there is a clear preference for
protonation trans to the carbene. For phosphino-carbene complexes, however, the site of protonation
depends on the steric bulk of the N-heterocyclic carbene ligand. Protonation trans to carbene is favored
with small substituents (H, Me), but the bulky 2,6-Pri2C6H3 susbstituent induces protonation trans to the
phosphine, as is seen experimentally.
spectroscopic and chemical behavior of other co-ligands. For
example, due to their strong σ-donation (comparable to that of
the trialkyphosphines),5 the NHCs should exert a strong trans
influence (and trans effect) weakening the M-L bonds trans
to them. Although this view has been explicitly spelled out in
the literature,6 it has not been studied in detail.
Introduction
The study of N-heterocyclic carbene (NHC) ligands tethered
to a classical heteroatom donor via an organic linker of variable
length and rigidity has attracted intense interest.1 These ligands
of various architectures and denticities2 provide opportunities
for the design of new homogeneous catalysts and the study of
elementary organometallic reactions in precisely tailored elec-
tronic and steric environments. The NHC ligand forming strong
M-CNHC bonds was originally introduced in order to circumvent
problems of lability and ligand decomposition associated with
the electronically similar and ubiquitous phosphines.3 However,
NHC ligands also exhibit unusual reactivity (insertion, reductive
elimination, substitution),4 which must be well understood for
successful catalyst design. The coordinated NHC should affect
We have recently reported7 the use of the diphenylphosphino-
functionalized NHC ligand 3-aryl-1-[â-(diphenylphosphino)-
ethyl]imidazol-2-ylidene (C∼P) (aryl ) DiPP, 2,6-Pri2C6H3;
Mes, 2,4,6-trimethylphenyl) for the preparation of a range of
Pd(II) dimethyl complexes of the type (κ2-C∼P)Pd(CH3)2. We
have found that their protonation with 1 equiv of Bro¨nsted acids
(4) (a) McGuinness, D. S.; Saendig, N.; Yates, B. F.; Cavell, K. J. J.
Am. Chem. Soc. 2001, 123, 4029. (b) McGuiness, D. S.; Cavell, K. J.; Yates,
B. F.; Skelton, B. W.; White, A. H. J. Am. Chem. Soc. 2001, 123, 8317. (c)
Gru¨ndemann, S.; Albrecht, M.; Kovacevic, A.; Faller, J. W.; Crabtree, R.
H. J. Chem. Soc., Dalton Trans. 2002, 2163. (d) Danopoulos, A. A.;
Tsoureas, N.; Green, J. C.; Hursthouse, M. B. Chem. Commun. 2003, 756.
(e) Clement, N. D.; Cavell, K. J.; Jones, C.; Elsevier, C. Angew. Chem.,
Int. Ed. 2004, 43, 1277. (f) Dorta, R.; Stevens, E. D.; Nolan, S. P. J. Am.
Chem. Soc. 2004, 126, 5054. (g) Gru¨ndemann, S.; Kovacevic, A.; Albrecht,
M.; Faller, J. W.; Crabtree, R. H. J. Am. Chem. Soc. 2002, 124, 10473. (f)
Galan, B. R.; Gembicky, M.; Dominiak, P. M.; Jerome, B.; Keister, J. B.;
Diver, S. T. J. Am. Chem. Soc. 2005, 127, 15702.
(5) (a) Green, J. C.; Scurr, R. G.; Arnold, P. L.; Cloke, F. G. N. Chem.
Commun. 1997, 1963. (b) Green, J. C.; Herbert, B. J. Dalton Trans. 2005,
1214.
(6) Gru¨ndemann, S.; Albrecht, M.; Loch, J. A.; Faller, J. W.; Crabtree,
R. H. Organometallics 2001, 20, 5485.
* Corresponding
s.a.macgregor@hw.ac.uk.
authors.
E-mail:
ad1@soton.ac.uk;
† University of Southampton.
‡ Heriot-Watt University.
(1) (a) Herrmann, W. A. Angew. Chem., Int. Ed. 2002, 41, 1290. (b)
Bourisou, D.; Guerret, O.; Gabbai, F. P; Bertrand, G. Chem. ReV. 2000,
100, 39.
(2) See for example: (a) Herrmann, W. A.; Ko¨cher, C.; Goossen, L. J.;
Artus, G. R. J. Chem.-Eur. J. 1996, 2, 1627. (b) Danopoulos, A. A.;
Motherwell, W. B.; Winston, S. Chem. Commun. 2002, 1376. (c) Frankel,
R.; Kniczek, J.; Ponikwar, W.; Noth, H.; Polborn, K.; Fehlhammer, W. P.
Inorg. Chim. Acta 2001, 312, 23. (d) Hu, X.; Castro-Rodriguez, I.; Meyer,
K. J. Am. Chem. Soc. 2004, 126, 13464.
(3) van Leeuwen, P. W. Homogeneous Catalysis. Understanding the Art;
Kluwer Academic Publishers: Dortrecht, 2004; p 295.
(7) Tsoureas, N.; Danopoulos, A. A.; Tulloch, A. A. D.; Light, M. E.
Organometallics 2003, 22, 4750.
10.1021/om0608408 CCC: $37.00 © 2007 American Chemical Society
Publication on Web 12/07/2006

254 Organometallics, Vol. 26, No. 2, 2007
Danopoulos et al.
Scheme 1. Synthesis of Neutral Palladium Dimethyl
Complexesa
Results and Discussion
Synthesis of the Neutral Dimethyl Complexes. The syn-
thesis of the new complex (κ2-C∧P)Pd(CH3)2, 1, was carried
out as before,7 i.e., by reacting Pd(tmed)(CH3)2 with the
phosphine-functionalized N-heterocyclic carbene (C∧P) gener-
ated in situ from the imidazolium salt and KN(SiMe3)2 in THF.
The new complex was isolated as a yellow, air-stable powder
in good yields. It was characterized by analytical and spectro-
scopic methods. The 31P{1H} NMR consists of a singlet at 32.7
ppm [cf. 12.45 and 12.24 for the analogous (C∼P) complexes].7
Scheme 3. Synthesis of Cationic Palladium Methyl
Complexesa
a
Complexes in bold numbers have been structurally characterized.
Ar ) 2,6-diisopropylphenyl. Reagents and conditions: (i) 1 equiv
[(C∧P)H]Br, 1.1 equiv KN(SiMe3)2, THF, -78 °C, then 1 equiv
Pd(tmeda)Me2, THF, -78 °C to RT. (ii) 1.4 equiv (C-N) (R ) H), 1
equiv (C-N)* (R ) Me), THF, -78 °C to RT. (iii) 2 equiv (C-N),
THF, RT.
having noncoordinating anions at low temperatures in organic
noncoordinating solvents, followed by trapping of the organo-
metallic intermediate with a ligand L (L ) pyridine, CH3CN,
PMe3, etc.), proceeds with high regioselectivity, leading to the
formal substitution by L of the methyl trans to the P atom and
not trans to the NHC as expected.
a Complex in bold number has been structurally characterized. Ar
) 2,6-diisopropylphenyl. (i) 1 equiv [H(OEt2)2]+[BArF ]-, CH2Cl2, -78
4
°C to -40 °C, then 1.1 equiv 4-Butpyridine, -40 °C to RT; (ii) 1 equiv
[H(OEt2)2]+[BArF ]-, CH2Cl2, -78 to -40 °C, then 1.1 equiv pyridine,
4
In this paper we report (i) the use of the new, smaller bite
angle ligand 3-DiPP-1-[(diphenylphosphino)methyl]imidazol-
2-ylidene (C∧P) for the synthesis of the analogous complex
(κ2-C∧P)Pd(CH3)2, (ii) the use of the 3-DiPP-1-(2-pyridyl)-
imidazol-2-ylidene (C-N) and 3-DiPP-1-(2-(3-methyl)pyridylim-
idazol-2-ylidene (C-N*) for the synthesis of complexes of type
(κ2-C-N)Pd(CH3)2 and (κ2-C-N*)Pd(CH3)2; (iii) the prepara-
tion of cationic methallyl complexes of the type [(C∼P)Pd-
(methallyl)]+ and [(κ2-C-N)Pd(methallyl)]+; (iv) the regiose-
lective protonation of the new dimethyl complexes with
Bro¨nsted acids having noncoordinating anions; and (iv) attempts
to rationalize the regiospecificity of the protonation reaction with
the aid of density functional calculations.
-40 °C to RT.
The inequivalent Pd-methyls appear as doublets (3JPH ) 8.1
3
Hz, JPH ) 8.2 Hz) almost isochronous to the methyls of Pd-
8
9
(tmed)(CH3)2 but more shielded relative to cis-Pd(P)2(CH3)2
(P ) tertiary phosphine) and [(NHC)2-CH2]Pd(CH3)2.10 The
low molecular symmetry of 1 was also manifested by the
appearance of the Pri groups of DiPP as two doublets. This
observation supports a nonplanar five-membered chelate ring.
The carbene NHC carbon is found at δ 190.4. The structure of
1 was determined crystallographically and is in agreement with
the spectroscopic data (Figure 1).
As expected, the Pd adopts a square planar coordination
geometry with a bidentate C∧P ligand. The five-membered
chelate ring is puckered, and the ligand bite angle is 80.39(14)°.
All new complexes and the chemical transformations leading
to them are summarized in Schemes 1-3.
Scheme 2. Synthesis of Cationic Methallyl Palladium Complexesa
a Complexes in bold numbers have been structurally characterized. Ar ) 2,6-diisopropylphenyl.

N-Heterocyclic Carbene Complexes of Palladium
Organometallics, Vol. 26, No. 2, 2007 255
Figure 1. ORTEP representation of the structure of 1 showing
50% probability ellipsoids. H atoms are omitted for clarity. Selected
bond lengths (Å) and angles (deg) with estimated standard
deviations: C(2)-Pd(1) ) 2.068(5), P(1)-Pd(1) ) 2.293(3),
C(30)-Pd(1) ) 2.067(5), C(29)-Pd(1) ) 2.090(5); C(2)-Pd(1)-
P(1) ) 80.39(14), C(30)-Pd(1)-P(1) ) 98.61(14), C(29)-Pd(1)-
P(1) ) 176.56(14), C(2)-Pd(1)-C(29) ) 96.17(19), C(30)-
Pd(1)-C(2) ) 176.0(2), C(30)-Pd(1)-C(29) ) 84.82(19).
Figure 2. ORTEP representation of the structure of 2b showing
50% probability ellipsoids. H atoms are omitted for clarity. Selected
bond lengths (Å) and angles (deg): C(2)-Pd(1) ) 2.038, N(1)-
Pd(1) ) 2.130, C(22)-Pd(1) ) 2.017, C(23)-Pd(1) ) 2.082;
C(22)-Pd(1)-C(2) ) 100.7, C(22)-Pd(1)-C(23) ) 85.1, C(2)-
Pd(1)-C(23) ) 172.2, C(22)-Pd(1)-N(1) ) 174.5, C(2)-Pd(1)-
N(1) ) 76.8, C(23)-Pd(1)-N(1) ) 97.8.
The Pd-CH3 [trans to the NHC, 2.067(5) Å, trans to the
phosphine, 2.090(5) Å] and the Pd-C(NHC) [2.068(5) Å]
distances are all equal within the observed esd’s. The interplanar
angle between the heterocyclic ring plane and the coordination
plane is 23.5°.
In order to gain a better insight into the structures and
reactivity of the palladium cis-dimethyl complexes supported
by heterobifunctional NHC ligands, we prepared the Pd(κ2-C-
N)(CH3)2 and Pd(κ2-C-N*)(CH3)2 complexes, 2a and 2b,
respectively, by the reaction of the isolated (C-N) or (C-N*)11
with Pd(tmed)(CH3)2. cis-Pd(κ2-C-N*)(CH3)2, 2b, was obtained
in good yields, by careful addition of the ligand to a cold (-78
°C) solution of the organometallic precursor, followed by slow
warming; the same procedure was applied for the synthesis of
2a, but the isolated yields were lower. Mixtures of 2a and cis-
Pd(κ1-C-N)2(CH3)2, 3, were obtained when an excess of the
ligand was used and/or by mixing the reagents at room
temperature. Optimized conditions for the selective preparation
of 2a and 3 are given in the Experimental Section. The
complexes 2a, 2b, and 3 were characterized by analytical and
spectroscopic methods. The low symmetry in 2a and 2b is
manifested by the presence of a pair of two inequivalent Pd-
CH3 resonances at δ 0.12 and 0.01 (for 2a) and 0.22 and 0.12
(for 2b), respectively; the downfield peaks were assigned by
NOESY experiments to the methyls trans to the NHC groups.
Compared to the corresponding methyls of the (κ2-C∼P)Pd-
(CH3)2 and (κ2-C∧P)Pd(CH3)2, the methyls of 2a and 2b are
deshielded. The 1H NMR of 3 implies a C2 symmetric structure
(one singlet for the Pd-CH3, four doublets and two septets for
the o-(CH3)2CH of DiPP). The CNHC was observed only for 3
at 200.1 ppm.
Figure 3. ORTEP representation of the structure of 3 showing
50% probability ellipsoids. H atoms are omitted for clarity. Selected
bond lengths (Å) and angles (deg) with estimated standard
deviations: C(1)-Pd(1) ) 2.058(3), C(2)-Pd(1) ) 2.064(5), Pd-
(1)-C(7) ) 2.101(1), Pd(1)-C(8) ) 2.103(5); C(8)-Pd(1)-C(2)
) 170.87(11), C(7)-Pd(1)-C(1) ) 171.08, C(1)-Pd(1)-C(8) )
87.96(15), C(2)-Pd(1)-C(7) ) 87.73(17), C(1)-Pd(1)-C(2) )
101.16(16).
adopts an almost coplanar conformation with the pyridine plane
slightly tilted, in order to accommodate the steric requirements
of the methyl substituent of the pyridine heterocycle. The ligand
bite angle of 76.8° is smaller than the ones observed for
complexes (κ2-C∼P)Pd(CH3)2 and 1.
7
Complex 3 adopts a C2 symmetrical conformation, with the
geometry around the metal center being square planar. The C2
rotation axis lies on the coordination plane and bisects the Me-
Pd-Me angle. The Pd-CNHC bond distances are equal to and
in the same range (within esd’s) as the ones observed for the
bidenate complexes 1 and 2b. The same trend is also observed
for the Pd-CH3 bond lengths. The planes of the carbene
moieties form a dihedral angle of approximately 75.5° with the
coordination plane. The C2 conformation is possibly adopted
due to steric reasons, a fact that is also supported by the large
C(1)-Pd(1)-C(2) angle [101.16(16)°].
The structures of 2b and 3 were confirmed crystallographi-
cally and are shown in Figures 2 and 3, respectively.
Even though the quality of the data obtained for 2b is not
suitable for the estimation of bond lengths and angles at the
desired accuracy, they establish unambiguously the connectivity
in the complex and allow the extraction of trends in the metrical
data. The geometry around the metal center is square planar,
with the (C-N*) ligand occupying two cis positions. The ligand
Synthesis of the Cationic Methallyl Complexes. Heterobi-
functional bidentate co-ligands have been used on Pd allyl
complexes in order to promote selective allyl substitution at
one end of the coordinated allyl.12 The directing effect is
believed to be the electronic differentiation due to the donors
(8) de Graaf, W.; Boersma, J.; Smeets, W. J. J.; Spek, A. L.; van Koten,
G. Organometallics 1989, 8, 2907.
(9) Tooze, R. P.; Chiu, K. W.; Wilkinson, G. Polyhedron 1984, 3, 1025.
(10) Douthwaite, R. E.; Green, M. L. H.; Silcock, P. J.; Gomes, P. T. J.
Chem. Soc., Dalton Trans. 2002, 1386.
(11) Winston, S.; Stylianides, N.; Tulloch, A. A. D.; Wright, J. A.;
Danopoulos, A. A. Polyhedron 2004, 23, 2813.
(12) Heumann, A. In Transition Metals for Organc Synthesis; Beller,
M., Bolm, C., Eds.; Wiley-VCH: Weinheim, 2004; Vol. 1, p 307.

256 Organometallics, Vol. 26, No. 2, 2007
Danopoulos et al.
Figure 5. ORTEP representation of the cation in 5 showing 50%
probability ellipsoids. H atoms are omitted for clarity. Selected bond
lengths (Å) and angles (deg) with estimated standard deviations:
C(13)-Pd(1) ) 2.039(5); P(1)-Pd(1) ) 2.2825(14); C(30)-Pd-
(1) ) 2.157(5); C(31)-Pd(1) ) 2.202(5); C(32)-Pd(1) ) 2.164(6);
C(13)-Pd(1)-P(1) ) 92.56(14); C(13)-Pd(1)-C(30) 168.0(2).
Figure 4. ORTEP representation of the cation in 4 showing 50%
probability ellipsoids. H atoms and one molecule of ether are
omitted for clarity. Selected bond lengths (Å) and angles (deg) with
estimated standard deviations: C(2)-Pd(1) ) 2.014(3), C(21)-
Pd(1) ) 2.165(4), C(22)-Pd(1) ) 2.160(4), C(23)-Pd(1) )
2.096(5), N(1)-Pd(1) ) 2.113(4); C(2)-Pd(1)-C(23) ) 106.53(18),
C(2)-Pd(1)-N(1) ) 78.44(16), C(23)-Pd(1)-N(1) ) 173.28(15),
C(2)-Pd(1)-C(22) ) 139.74(18) C(23)-Pd(1)-C(22) ) 38.73(17),
N(1)-Pd(1)-C(22) ) 138.40(15), C(2)-Pd(1)-C(21) ) 172.87-
(18), C(23)-Pd(1)-C(21) ) 68.06(18), N(1)-Pd(1)-C(21) )
106.59(16), C(22)-Pd(1)-C(21) ) 37.66(17), C(21)-C(22)-
C(23) ) 116.3(4), C(21)-C(22)-C(24) ) 120.2(4), C(23)-C(22)-
C(24) ) 121.8(4).
those observed with the neutral dimethyl complexes described
previously and may be related to differences in the trans
influences of donor groups of the bidentate ligand. The allyl
plane forms an angle of ca. 54.6° (in 4) and ca. 53.5° (in 5)
with the coordination plane.
Protonation of the Complexes. The reaction of the com-
plexes (κ2-C∼P)Pd(CH3)2 with 1 equiv of acids H+A-, A-
)
noncoordinating anion, at low temperatures, followed by trap-
ping of the organometallic intermediate with L (L ) py, PMe3,
MeCN, etc.), has already been reported.7 The unexpected
regioselectivity of the reaction (100% substitution of the methyl
that is trans to the phosphine) prompted us to explore it in detail
with the other related Pd complexes that were described in the
previous section.
of the electronically nonsymmetric bidentate ligand. [Pd(NHC)-
(allyl)Cl] complexes with monodenate NHCs have been reported
and used for the generation of reactive “monoligated” [Pd(0)-
(NHC)] centers.13 In order to probe the electronic differentiation
exerted by the phosphine- and pyridine-functionalized NHC
ligands described above on the allyl group coordinated on Pd,
we undertook the synthesis of the complexes [Pd(C-N)-
(methallyl)]+ and [Pd(C∼P)(methallyl)]+. Complexes [Pd(C-
N)(methallyl)](BPh4), 4, and [Pd(C∼P)(methallyl)](BPh4), 5,
were prepared by the reaction of [Pd(methallyl)Cl]2 with (C-
N) or (C∼P). The crude [Pd(C-N)(methallyl)]Cl and [Pd-
(C∼P)(methallyl)]Cl, which were not isolated, were transformed
to 4 and 5 by anion exchange with NaBPh4. The NMR spectra
of the new products manifested the lack of any symmetry
element in the cations and evidenced the presence of one isomer
in solution. The Pri of the DiPP groups as well as the
diastereotopic protons of the termini of the methallyl group
appear as multiplets. We could not observe the CNHC signal in
the 13C{1H} NMR spectra in any of these complexes. The
structures of 4 and 5 were confirmed crystallographically and
are shown in Figures 4 and 5.
Both 4 and 5 feature bidentate (C-N) and (P∼C) ligands,
the former adopting a virtually planar conformation (excluding
the DiPP substituent), while in the latter the conformation of
the six-membered chelate ring is best described as half-chair.
The Pd-CNHC bonds in 4 and 5 are similar (within the measured
esd’s) to the corresponding bond lengths in 1 and 2. The bonding
of the methallyl group to the metal in both complexes is
nonsymmetrical, with significantly different Pd-C bonds. In 4
the Pd-C allyl carbon that is trans to the CNHC is 2.165(4) Å,
and the one trans to the pyridine is 2.096(5) Å. In 5 the Pd-C
allyl carbon that is trans to the CNHC is 2.154(4) Å, and the one
trans to the P donor is 2.168(5) Å. The Pd-C (central allyl
carbon) is 2.160(4) Å in 4 and 2.202(5) Å in 5. The trends in
the magnitute of these Pd-C bond lengths are analogous to
The reaction of 1 with 1 equiv of (CF3)2CHOH or
[H(Et2O)]+[BArF ]- in CH2Cl2 or CH2Cl2/ether at -78 °C,
4
followed by addition of 1 equiv of 4-But-pyridine at -40 °C
and warming to room temperature, gave on workup a white,
air-stable solid, 6. The appearance of one Pd-CH3 peak in the
1H NMR spectrum (δ 0.04, broad singlet) in addition to
resonances due to pyridine, the (C∧P) ligand, and the anion,
and one signal in the 31P{1H}NMR spectrum (δ 55.3), deshield-
ed relative to 1, support the presence of one isomer in solution.
Even though crystals of this complex suitable for X-ray
diffraction could not be grown, the proposed structure (Scheme
3) in which the But-pyridine is trans to the phosphine is based
on comparison of the NMR data with the already structurally
characterized [(κ2-C∼P)Pd(CH3)(py)]+ analogue.7 It is interest-
ing to notice that monitoring the reaction by 1H NMR
spectroscopy at low temperatures confirmed that a single species
was formed at lower temperatures exhibiting the same spectrum
as the isolated complex, eliminating the possibility of isomer-
ization in solution prior to the isolation of 6.
Analogous reaction of 2b with 1 equiv of [H(Et2O)]+[BArF ]-
4
followed by addition of pyridine gave the cationic complex 7
as a single isomer, which was characterized analytically,
spectroscopically, and crystallographically. In the 1H NMR
spectrum, the single Pd-CH3 resonance (δ -0.09) of the
cationic 7 is shielded relative to 2b. The configuration of 7 in
solution was established by NOESY NMR spectroscopy and
showed that substitution trans to the NHC donor has taken place.
Monitoring the reaction at lower temperatures (-40 °C) by 1H
NMR shows that the product formed originally at lower
temperatures was identical to that of the isolated 7. The structure
of 7 was determined crystallographically and is in agreement
with the spectroscopic assignments (Figure 6).
The geometry around the Pd center is square planar, with
the pyridine ligand being trans to the NHC and the remaining
(13) Marion, N.; Navarro, O.; Mei, J.; Stevens, E. D.; Scott N. M.; Nolan,
S. P. J. Am. Chem. Soc. 2006, 128, 4101.

N-Heterocyclic Carbene Complexes of Palladium
Organometallics, Vol. 26, No. 2, 2007 257
Table 1. Crystallographic Data for the Compounds Described in the Paper
1
2b
3
4
5
7
chemical formula
fw
cryst syst
space group
a/Å
b/Å
c/Å
R/deg
â/deg
γ/deg
Z
C30H37N2PPd
562.99
orthorhombic
P212121
10.031(13)
11.34(2)
24.499(8)
90
C23H31N3Pd
455.9
monoclinic
P21/c
10.408(6)
12.139(7)
17.357(11)
90
105.97(4)
90
C42H52N6Pd
747.30
monoclinic
P21/n
10.936(14)
27.06(3)
13.47(4)
90
C48H50BN3Pd
786.12
triclinic
P1h
13.149(2)
14.181(2)
15.6998(17)
115.110(12)
90.856(13)
114.148(16)
2
C57H60BN2PPd
921.25
monoclinic
Cc
16.070(3)
16.422(3)
19.758(4)
90
C59H45BF24N4Pd
1383.20
monoclinic
P21/n
24.5559(8)
19.7453(7)
25.285(8)
90
90
90
4
107.64(19)
90
4
112.77(3)
90
4
97.3400(10)
90
8
4
T/K
µ/mm-1
120(2)
0.743
120(2)
0.892
120(2)
0.526
120(2)
0.425
120(2)
0.458
120(2)
0.420
no. of data collected
no. of unique data
goodness of fit on F2
Rint
24 799
6316
1.028
0.1060
0.0483
0.0785
22 013
3714
1.188
0.1620
0.1037
0.1204
33 787
6436
1.047
0.0674
0.0363
0.0507
40 241
10 833
1.056
0.0936
0.0680
0.0989
13 505
5433
0.997
0.0648
0.0417
0.0618
120 423
27 861
1.031
0.1433
0.0786
final R(|F|) for Fo > 2σ(Fo)
final R(F2) for all data
0.1526
Table 2. Computed Pd-Ligand Bond Distances (Å) in
(K2-C∼P)Pd(CH3)2 and (K2-C-N*)Pd(CH3)2 Species Using
Model 1 and Model 2a
Me group trans to the pyridine. The angle formed between the
pyridine (L) plane and the coordination plane is 74.9°. The Pd-
CNHC and Pd-CH3 bonds in 7 are shorter than in the corre-
sponding [(κ2-C∼P)Pd(CH3)(py)]+ previously reported,7 while
the Pd-N(pyridine) bonds in the two complexes are equal.
Computational Studies. We have employed density func-
tional calculations to rationalize the selectivities of the proto-
nation/substitution reactions and have focused on the (κ2-
C∼P)Pd(CH3)27 and (κ2-C-N)Pd(CH3)2 systems. In particular,
the observation of substitution trans to phosphorus upon
protonation of (κ2-C∼P)Pd(CH3)2 was initially surprising, as
the NHC arm of the C∼P ligand might have been expected to
exert a higher trans influence and trans effect than the phosphine
moiety.6
Two model systems were used in the calculations. In model
1 all ligand substituents were replaced by hydrogens, while in
model 2 the full (κ2-C∼P)Pd(CH3)2 and (κ2-C-N)Pd(CH3)2
molecules were considered. Computed Pd-ligand distances are
compared with those determined crystallographically in Table
2.
Model 1 generally provides good agreement with experiment,
although the Pd-N bond length is somewhat overestimated
compared to the crystal data for (κ2-C-N*)Pd(CH3)2. In model
2 inclusion of the full ligand causes the Pd-P bond in (κ2-
a Corresponding distances determined crystallographically are included
for comparison.
C∼P)Pd(CH3)2 to lengthen by 0.04 Å, but otherwise the Pd-
ligand bond distances are not significantly affected, suggesting
minimal electronic effects arising from the ligand substituents.
Of prime interest are the relative lengths of the Pd-CH3 bonds,
and for (κ2-C∼P)Pd(CH3)2 the calculations find these to be
slightly longer when trans to carbene, by 0.005 Å (model 1)
and 0.013 Å (model 2). Experimentally, it is the Pd-CH3 bond
trans to carbene that is slightly shorter, although the difference
is not significant when the experimental errors are considered.
Both models of (κ2-C-N)Pd(CH3)2 correctly reproduce the
longer Pd-CH3 bond trans to carbene seen experimentally, this
being longer by 0.032 Å (model 1) and 0.025 Å (model 2),
although these differences are somewhat smaller than those
determined experimentally.
The experimental and computed Pd-CH3 distances provide
a basis for estimating the trans influence of the donor atoms in
the C-N and C∼P ligands. For the C-N ligand both approaches
indicate that the NHC arm has the significantly higher trans
influence. In contrast, for the C∼P ligand the trans influences
of the NHC and the alkyldiarylphosphine moieties appear to
be very similar, although the calculations do suggest a slightly
higher trans influence for the NHC donor in the C∼P ligand.
Figure 6. ORTEP representation of the structure of the cation in
7 showing 50% probability ellipsoids. The asymmetric unit consists
of two cations and two anions. The [BArF ]- anion and H atoms
4
are omitted for clarity. Selected bond lengths (Å) and angles (deg)
with estimated standard deviations: C(2)-Pd(1) ) 1.964(5), N(4)-
Pd(1) ) 2.084(4), N(1)-Pd(1) ) 2.154(4), C(27)-Pd(1) ) 2.021-
(5), C(2)-Pd(1)-C(27) ) 96.9(2), C(2)-Pd(1)-N(4) ) 172.2(2),
C(27)-Pd(1)-N(4) ) 87.96(19), C(2)-Pd(1)-N(1) ) 78.16(18),
C(27)-Pd(1)-N(1) ) 172.1(2), N(4)-Pd(1)-N(1) ) 97.65(16).

258 Organometallics, Vol. 26, No. 2, 2007
Danopoulos et al.
Figure 7. Key stationary points computed along the protonation/
substitution reaction of (κ2-C-N*)Pd(CH3)2, with model 1 (R )
R′ ) H). Energies are relative to trans-N [(κ2-C-N*)Pd(CH3)-
(CH4)]+ set to zero. No local minimum corresponding to [(κ2-C-
N*)Pd(CH3)2(H)]+ could be located.
Figure 8. Relative energies (kcal/mol) of species along the
protonation/substitution reaction of (κ2-C-P)Pd(CH3)2 with model
1 (R ) R′ ) H) and (in parenthesis) model 2 (R ) DiPP; R′ )
Ph). Energies are relative to trans-P [(κ2-C∼P)Pd(CH3)(CH4)]+ set
to zero. No local minimum corresponding to [(κ2-C∼P)Pd(CH3)2-
(H)]+ could be located with model 1.
If this is the case, however, then it would appear to be
inconsistent with the experimental observation of substitution
trans to phosphine upon protonation.
We then considered the protonation/substitution reaction of
(κ2-C∼P)Pd(CH3)2, again initially with model 1 (see Figure 8).
As with (κ2-C-N)Pd(CH3)2 we were unable to locate a stable
species resulting from direct protonation at Pd, and uncon-
strained hydride structures corresponding to [(κ2-C∼P)Pd(CH3)2-
(H)]+ all relaxed to trans-C [(κ2-C∼P)Pd(CH3)(CH4)]+. This
species was found to be 2.7 kcal/mol more stable than the
trans-P isomer, and this preference was retained in the final
product, [(κ2-C∼P)Pd(CH3)(py)]+, where the trans-C isomer
was more stable by 3.3 kcal/mol. CH4/py substitution is again
a strongly exothermic process. Most importantly, with model 1
both the initial site of protonation and the energy of the final
product indicate a small preference for reaction trans to the
carbene arm of the C∼P ligand. This is at odds with the
experimental observation of protonation/substitution trans to the
phosphine arm of the C∼P ligand in (κ2-C∼P)Pd(CH3)2. We
therefore extended our study to include the full ligand substit-
uents through use of model 2.
The energies of the species calculated with model 2 are
included in parentheses in Figure 8. Unlike model 1, in this
case we were able to locate a local minimum corresponding to
[(κ2-C∼P)Pd(CH3)2(H)]+, although this remains a very high-
energy structure, which is at least 30 kcal/mol above both
isomers of [(κ2-C∼P)Pd(CH3)(CH4)]+. Direct protonation at the
methyl ligands is therefore much more likely, and with model
2 the trans-P isomer of [(κ2-C∼P)Pd(CH3)(CH4)]+ in fact
becomes the more favored species, being 2.3 kcal/mol more
stable than the trans-C isomer. Upon CH4/pyridine substitution,
however, the preference swaps again and the trans-C product
becomes more stable, albeit by only 0.3 kcal/mol. The calcula-
tions therefore suggest that the introduction of the full steric
bulk of the C∼P ligand causes a switch in the preferred site of
protonation, from trans to carbene in model 1 to trans to
phosphine in model 2. The energy differences involved are
small, however, and so to ensure that the computed preferred
site of protonation does indeed equate to the same selectivity
in the final observed products, we have studied the mechanism
of CH4/py substitution in both [(κ2-C∼P)Pd(CH3)(CH4)]+
intermediates.
We therefore turned to consider the protonation and subse-
quent CH4/py substitution reactions, and our approach is
illustrated in Figure 7 for (κ2-C-N)Pd(CH3)2. Protonation of
this species could occur at the Pd center, to give the five-
coordinate intermediate [(κ2-C-N)Pd(CH3)2(H)]+. From this
species C-H bond coupling would yield one of two isomers
of the CH4 σ-complex, [(κ2-C-N)Pd(CH3)(CH4)]+, with CH4
trans to either carbene (trans-C) or pyridine (trans-N). Alter-
natively, protonation could occur initially at the methyl groups
to form either of these two CH4 σ-complexes directly. CH4/py
substitution would then give the trans-C or trans-N forms of
the final product, [(κ2-C-N)Pd(CH3)(py)]+. We shall first
describe our results for the (κ2-C-N)Pd(CH3)2 system for which
the trans-C product is observed experimentally. An analogous
scheme will then be considered for (κ2-C∼P)Pd(CH3)2.
With model 1, all attempts to locate the putative five-
coordinate intermediate [(κ2-C-N)Pd(CH3)2(H)]+ as a local
minimum failed. An approximate geometry and energy for this
species were generated by assuming a square pyramidal structure
with an axial hydride and with all H-Pd-ligand angles
constrained to 90°. Upon releasing these angular constraints,
however, the axial hydride transferred directly onto the CH3
ligand trans to carbene to form the trans-C CH4 σ-complex [Pd-
(κ2-C-N)(CH3)(CH4)]+. This σ-complex was estimated to be
at least 30 kcal/mol more stable than the constrained-angle
structure of [Pd(κ2-C-N)(CH3)2(H)]+. Moreover, the trans-C
form of [Pd(κ2-C-N)(CH3)(CH4)]+ was 13.0 kcal/mol more
stable than the alternative trans-N isomer (see Figure 7). This
preference was retained in the final product, the trans-C form
of [(κ2-C-N)Pd(CH3)(py)]+ being 9.4 kcal/mol more stable than
the trans-N alternative. Thus the calculations employing model
1 indicate a strong preference for direct protonation of the
methyl group trans to carbene, and this is consistent with the
experimental observation of substitution trans to carbene upon
protonation. This assertion assumes facile CH4/py substitution
in [Pd(κ2-C-N)(CH3)(CH4)]+, and this process is certainly
extremely exothermic (by more than 30 kcal/mol). Subsequent
calculations on the [(κ2-C∼P)Pd(CH3)(CH4)]+ system also
indicated that the barriers associated with CH4/py substitution
in these systems should be small (see below). The relative
energies of the isomers of [Pd(κ2-C-N)(CH3)(CH4)]+ and [(κ2-
C-N)Pd(CH3)(py)]+ both reflect the higher trans influence of
the carbene arm of the C-N ligand, with the more stable form
in each case having the more weakly bound ligand (CH4 and
py, respectively) trans to the carbene.
The two limiting mechanisms for CH4/py substitution in [(κ2-
C∼P)Pd(CH3)(CH4)]+ involve either dissociative or associative
displacement, and computed reaction profiles for both isomers
are shown in Figure 9. Dissociation of CH4 leads to the T-shaped
14e intermediates, TC or TP [(κ2-C∼P)Pd(CH3)]+, of which the
TP isomer is significantly more stable by 9 kcal/mol. Moreover,
we were able to locate a transition state for CH4 loss from

N-Heterocyclic Carbene Complexes of Palladium
Organometallics, Vol. 26, No. 2, 2007 259
Figure 9. Computed reaction profiles (kcal/mol) for CH4/py substitution in [(κ2-C∼P)Pd(CH3)(CH4)]+ species to give trans-P [(κ2-C∼P)-
Pd(CH3)(py)]+ (right) or trans-C [(κ2-C∼P)Pd(CH3)(py)]+ (left). Key species for dissociative (blue) and associative substitution (red) are
indicated. See text for details.
trans-P [(κ2-C∼P)Pd(CH3)(CH4)]+, which corresponds to a very
low activation barrier of only 2.2 kcal/mol. In contrast, a reaction
profile probing CH4 loss from trans-C [(κ2-C∼P)Pd(CH3)-
(CH4)]+ showed the energy to rise steadily to the level of the
TC 14e intermediate, and the energy of this species (plus free
CH4) therefore provides a lower limit for the CH4 dissociation
barrier of 6.4 kcal/mol, relative to trans-C [(κ2-C∼P)Pd(CH3)-
(CH4)]+. Overall the calculations indicate that a dissociative
mechanism for CH4/py substitution in [(κ2-C∼P)Pd(CH3)-
(CH4)]+ species will be more accessible from the trans-P rather
than the trans-C isomer. The lower energy of the TP form of
[(κ2-C∼P)Pd(CH3)]+ appears to be due to an agostic interaction
between Pd and a C-H bond from one of the isopropyl
substituents (Pd‚‚‚H ) 2.38 Å; C-H ) 1.12 Å). In contrast
the closest Pd‚‚‚H contact in the TC isomer is 2.67 Å (again to
an isopropyl hydrogen), although in this case the C-H bond is
directed toward an axial position and so is poorly oriented to
participate in agostic bonding.
However, of the two possibilities, CH4/py substitution trans to
phosphine is more accessible than that trans to carbene.
The key factor in determining the site of protonation appears
to be the relief of steric strain that occurs upon protonation cis
to the bulky DiPP group. With model 1swhich may be
considered as giving the ideal structure in the absence of steric
effectssthe computed structure of (κ2-C∼P)Pd(CH3)2 is close
to square planar with trans-C-Pd-P and trans-C-Pd-C angles
of 175.3° and 178.5°, respectively, and the carbene ring is close
to coplanar with the Pd coordination plane (N-C-Pd-P )
16°). With model 2, however, a more distorted geometry is seen
(trans-C-Pd-C ) 173.6°, trans-C-Pd-P ) 167.1°) and the
carbene ring is forced to rotate out of the Pd coordination plane
(N-C-Pd-P ) 29°). Similar distortions are maintained in the
structure of trans-C [(κ2-C∼P)Pd(CH3)(CH4)]+ (trans-C-Pd-P
) 168.6°; N-C-Pd-P ) 37°), where DiPP is still cis to a
rigid Pd-Me bond. In contrast, in trans-P [(κ2-C∼P)Pd(CH3)-
(CH4)]+ the trans-C-Pd-C angle increases to 176.9° while the
carbene is able to rotate closer to the “ideal” orientation of model
1 (N-C-Pd-P ) 20°, cf. 16°). This is presumably because
the DiPP group is now cis to a more accommodating CH4 ligand.
For the associative displacement mechanism the very large
size of the [(κ2-C∼P)Pd(CH3)(CH4)]+/py system hampered the
location and full characterization of transition state structures.
Instead we have employed reaction profiles based on Pd‚‚‚N(py)
distances to estimate the barriers involved. This gives a value
of 1.0 kcal/mol for CH4/py substitution in trans-P [(κ2-C∼P)-
Pd(CH3)(CH4)]+ while the transition state is estimated to be
about 4 kcal/mol higher in the trans-C isomer (see Supporting
Information). Subsequently, we were able to locate a transition
state for the trans-P pathway, and this corresponded to an
activation barrier of 0.9 kcal/mol, in good agreement with the
estimated value. As with the dissociative process modeled above,
therefore, CH4/py substitution in [(κ2-C∼P)Pd(CH3)(CH4)]+ via
an associative process is also kinetically more accessible from
the trans-P rather than the trans-C isomer.
It follows from the above that the use of a smaller substituent
on the carbene should tend to favor reaction trans to the carbene.
We have tested this computationally by replacing the DiPP
group in model 2 with Me, and as predicted, the isomers of
[(κ2-C∼P)Pd(CH3)(CH4)]+ do become closer in energy, although
the trans-P form is still slightly more stable by 0.5 kcal/mol.
Unfortunately, we have been unable to test this idea experi-
mentally, as the reaction of Pd(tmed)(CH3)2 with the analogue
of C∼P having a methyl-substituted NHC donor gave mixtures
of complexes that were difficult to purify, thus hampering further
characterization. We were, however, able to test the link between
trans influence and the nature of the carbene substituent by
computing different models for the structure of [(κ2-C∼P)Pd-
(2-methylallyl)]+ (see Table 3). For model 1, where all
substituents were represented by hydrogen, the carbene moiety
does exert the greater trans influence, with the Pd-C bond trans
to carbene being longer by 0.025 Å. This difference is reduced
to 0.017 Å upon introduction of a Me substituent on the carbene
The calculations on the full system using model 2 show that
the preferred site of protonation at (κ2-C∼P)Pd(CH3)2 is the
methyl group trans to the phosphine arm. The site of protonation
will also dictate the regioselectivity of CH4/py substitution, as
the subsequent steps in this process (by either a dissociative or
an associative mechanism) have minimal activation barriers.

260 Organometallics, Vol. 26, No. 2, 2007
Danopoulos et al.
Table 3. Computed Pd-Ligand Bond Distances (Å) in
[(K2-C∼P)Pd(η-allyl)]+ Complexesa
Experimental Section
General Methods. Elemental analyses were carried out by the
London Metropolitan University microanalytical laboratory. All
manipulations were performed under nitrogen in a Braun glovebox
or using standard Schlenk techniques, unless stated otherwise.
Solvents were dried using standard methods and distilled under
nitrogen prior use. The light petroleum used throughout had a bp
of 40-60 °C.
The starting materials Pd(tmed)Me2,8 [H(OEt2)2]+[BArF ]-,15 and
4
(C-N)11 (R ) H, R ) Me) ligands were prepared according to
literature procedures. NMR data were recorded on Bruker AV-300
and DPX-400 spectrometers, operating at 300 and 400 MHz (1H),
respectively. The spectra were referenced internally using the signal
from the residual protiosolvent (1H) or the signals of the solvent
(13C). 31P{1H} NMR spectra were recorded on a Bruker AV-300,
at 121.44 MHz, and referenced externally relative to 85% H3PO4
in D2O. 19F NMR spectra were recorded on a Bruker AV-300, at
282.36 MHz, and were referenced externally relative to CFCl3.
Synthesis of Proligand (P∧CH)+Br-. A 3.0 g (10 mmol)
amount of Ph2P(O)CH2Br16 and 3.5 g (15 mmol) of 3-DiPP-
imidazole were charged in an ampule, which was sealed under
vacuum. The ampule and its contents were heated at 140 °C for
one week in a thermostated oil bath. After the completion of the
reaction, the contents of the ampule were dissolved in dichlo-
romethane. Removal of the volatiles under reduced pressure
produced a brown, sticky solid. This was redissolved in the
minimum amount of dichloromethane and precipitated by addition
of a 4:1 (v/v) mixture of ether/petroleum ether. The hygroscopic
solid was immediately filtered and washed with ether (3 × 60 mL).
It can be further purified by dissolving in the minimum amount of
THF and precipitating with ether. The white solid thus isolated was
pure for the next step. Crystalline material was obtained by cooling
a Corresponding distances determined crystallographically for 5 are
included for comparison.
(model 2). The trend is reversed in the computed structure of
the full molecule (model 3), with the Pd-C bond trans to
phosphine now being longer by 0.018 Å. Thus the relative trans
influences of the carbene and phosphine arms in these mixed
C∼P ligands are clearly a function of the steric bulk of the
carbene substituent.
Conclusions
The synthesis of cis dimethyl complexes of palladium
supported by mixed donor N-heterocyclic carbene-phosphine
or -pyridine bidentate ligands allowed a detailed comparative
study of the ligand effects on structural and reactivity differences
of the methyl co-ligands trans to the NHC and classical donor,
respectively. The experimental data show that the N-heterocyclic
carbene exerts a stronger trans influence than the pyridine and
promotes selective substitution of the methyl trans to it. These
findings are confirmed by density functional calculations, which
show the preferred site of protonation is at the methyl group
trans to the NHC.
In contrast, the trans influence of the NHC compared to the
phosphine is not clear-cut experimentally due to the esd’s of
the crystal structure determinations. However, in this case
substitution occurs selectively at the position trans to phosphine.
The calculations suggest a slightly higher trans influence
compared to phosphine in the NHC-phosphine ligands. How-
ever, the selectivity of protonation depends on the steric bulk
of the carbene substituent: protonation is favored trans to the
NHC for small substituents (H, Me) and swaps to trans to
phosphine for the DiPP substituent used experimentally. Thus
agreement with experiment could only be obtained once the
full ligand substituents were incorporated into the computational
model. The computed results suggest that the selectivity of the
protonation/substitution reaction can be controlled by judicious
choice of ligand substituent. Such subtle effects may arise from
the unique shape of the NHC ligand and must be taken into
consideration when using them as phosphine substitutes in
homogeneous catalysis.
1
the filtrate of the last precipitation at 5 °C. H NMR δ (CDCl3):
3
3
1.01 [6H, d, JHH ) 6.8 Hz, CH(CH3)2], 1.12 [6H, d, JHH ) 6.8
Hz, CH(CH3)2], 1.94 [2H, septet, 3JHH ) 6.8 Hz, CH(CH3)2], 6.25
[2H, d, JPH ) 6.6 Hz, Ph2P(O)CH2-imidazolium], 6.99 (1H, dd,
2
JHH ) 1.7 Hz, JHH ) 1.9 Hz, imidazole backbone), 7.21 (2H, d,
3JHH ) 7.7 Hz, aromatic), 7.32 (1H, br s, imidazole backbone),
7.50 (1H, t, 3JHH ) 7.7 Hz, aromatic), 7.55-7.63 (6H, m, aromatic),
8.22-8.31 (5H, m, aromatic), 10.2 (1H, br d, JHH ) 1.13 Hz,
NCHN). 13C{1H} NMR δ (CDCl3): 24.5 [s, CH(CH3)2], 24.6 [s,
1
CH(CH3)2], 28.9 [s, CH(CH3)2], 49.5 (d, JPC ) 64.5 Hz), 124.0
(s, imidazolium backbone), 124.6 (s, imidazolium backbone), 125.1
(s, aromatic), 128.3 (d, 1JPC ) 19.0 Hz, aromatic), 129.6 (d, JPC
)
12.6 Hz, aromatic), 130.1 (s, aromatic), 131.9 (d, JPC ) 10.3 Hz,
aromatic), 132.4 (s, aromatic), 133.5 (d, JPC ) 2.3 Hz, aromatic),
138.3 (s, ipso to the imidazole moiety), 145.5 [s, (s, NC(H)N]. 31P-
{1H} NMR δ (CDCl3): 36.6 (s, Ph2P(O)CH2-imidazolium). High-
resolution ES+: 443.2242. Calcd for [C28H32N2PO]+: 443.2247.
The phosphine oxide (2.0 g, 4 mmol) was placed in a three-
neck, 1 L flask (tap adapter, double jacket condenser, and a
pressure-equalizing dropping funnel) and was dissolved in 110 mL
of dried-degassed chlorobenzene by heating to the reflux temper-
ature under a nitrogen atmosphere. The temperature was then
maintained at 120 °C, and excess trichlorosilane was added in small
potrions (10 mL, 72 mmol, 19-fold excess). After completion of
the addition the reaction mixture was heated at 130 °C for 3 h and
then cooled to room temperature. After addition of 100 mL of
dichloromethane, the excess SiHCl3 was quenched by careful
dropwise addition of degassed 10% (w/v) NaOH(aq) at 0 °C. The
Finally, cationic Pd-allyl complexes with the same hetero-
bifunctional bidentate ligands show Pd-Callyl bond trends in
agreement with the above classification of the NHC, phosphine,
and pyridine donors. Computation of the Pd-Callyl bond lengths
shows that they are dependent on the substitution of the P and
NHC donors in line with the dimethyl complexes described
above.
Further work involving extensions to Pt and other transition
metals and aiming at the understanding of the metal-NHC
bonding in relation to observed reactivity is in progress.
(14) Brookhart, M.; Grant, B.; Volpe, A. F. Organometallics 1992, 11,
3920.
(15) Tsevtkov, E. N.; Tkachenko, S. E.; Yarkevitc, A. N. Bull. Soc. Chim.
Fr. 1988, 2, 339.
(16) Hooft, R. COLLECT; Nonius BV: The Netherlands, 1997-2000.
Otwinoski, Z.; Minor, W. SCALEPACK, DENZO. Methods Enzymol. 1997,
276, 307.

N-Heterocyclic Carbene Complexes of Palladium
Organometallics, Vol. 26, No. 2, 2007 261
organic phase was separated, and the aqueous phase was washed
with dichloromethane (3 × 50 mL). The combined organic extracts
were dried over MgSO4 and filtered, and the volatiles were removed
under reduced pressure. The resulting white solid was washed with
ether (2 × 150 mL) and dried under vacuum overnight. The
imidazolium salt was further dried azeotropically with toluene and
aromatic), 147.2 (s, ipso carbon of the aryl group). Anal. Found:
C, 59.55; H, 6.59; N, 9.31. Calcd for C22H29N3Pd: C, 59.79; H,
6.61; N, 9.51.
(C-N*)Pd(CH3)2, 2b. This was prepared by a method analogous
to 2a from 0.080 g (0.317 mmol) of Pd(tmeda)Me2 and 0.101 g
1
(0.317 mmol) of free carbene. Yield: 0.100 mg (81%). H NMR
1
was stored in the glovebox. Yield: 1.0 g (51.6%). H NMR δ
δ (CD2Cl2): 0.12 (3H, s, PdCH3), 0.22 (3H, s, PdCH3), 1.12 (6H,
3
3
3
(CDCl3): 1.03 [6H, d, JHH ) 6.8 Hz, CH(CH3)2], 1.15 [6H, d,
d, JHH ) 6.9 Hz, CH(CH3)2), 1.31 (6H, d, JHH ) 6.9 Hz, CH-
3
3
3JHH ) 6.8 Hz, CH(CH3)2], 2.01 [2H, septet, JHH ) 6.8 Hz, CH-
(CH3)2), 2.71 (2H, sept, JHH ) 6.9 Hz, CH(CH3)2), 2.91 (3H, s,
3-CH3-pyridine), 6.92 (1H, s, ylidene backbone), 7.31 (3H, m,
2
(CH3)2], 5.77 (2H, d, JPH ) 6.4 Hz, Ph2PCH2-imidazole), 6.91
3
3
(1H, t, JHH ) 1.7 Hz, aromatic), 7.08 (2H, d, JHH ) 7.7 Hz,
aromatic), 7.41 (1H, br t, aromatic), 7.72 (1H, br dd, aromatic),
3
3
3
aromatic), 7.31-7.38 (6H, m, aromatic), 7.41 (1H, t, JHH ) 7.7
8.07 (1H, d, JHH ) 2.5 Hz, aromatic), 8.61 (1H, d, JHH ) 5 Hz,
aromatic). 13C{1H} NMR δ(CD2Cl2): -0.2 (s, PdCH3), 0.0 (s,
PdCH3), 19.8 (s, 3-CH3-pyridine), 21.7 (s, CH(CH3)2), 22.8 (s, CH-
(CH3)2), 26.9 (s, CH(CH3)2), 115.9 (s, ylidene backbone), 120.4
(s, ylidene backbone), 120.8 (s, aromatic), 122.1 (s, aromatic), 122.5
(s, aromatic), 128.3 (s, aromatic), 141.5 (s, aromatic), 143.8 (s,
aromatic), 144.1 (s, aromatic), 144.5 (s, aromatic), 147.2 (s, ipso
carbon of the aryl group). Anal. Found, C, 60.33; H, 6.56; N, 9.13.
Calcd for C23H31N3Pd: C, 60.59; H, 6.85; N, 9.22. Yellow crystals
were grown by layering a toluene solution of the complex with
light petroleum ether.
3
Hz, aromatic), 7.51 (1H, t, JHH ) 1.7 Hz, aromatic), 7.58-7.77
(4H, m, aromatic), 10.5 (1H, t, JHH ) 1.5 Hz, NCHN). 13C{1H}
3
NMR δ (CDCl3): 23.1 [s, CH(CH3)2], 23.3 [s, CH(CH3)2], 27.6
[s, CH(CH3)2], 47.7 (d, 1JPC ) 19.52 Hz, Ph2PCH2-imidazolium),
121.44 (d, JPC ) 5.1 Hz, aromatic), 122.6 (s, imidazolium,
backbone) 123.6 (s, imidazolium backbone), 128.2 (d, JPC ) 7.6
Hz, aromatic), 129.1 (s, aromatic), 129.3 (s, aromatic), 130.8 (s,
aromatic), 131.6 (d, JPC ) 11.5 Hz, aromatic), 132.6 (d, JPC ) 20.3
Hz, aromatic), 138.2 (s, ipso to the imidazolium moiety), 144.3 [s,
NC(H)N]. 31P{1H} NMR: δ (CDCl3): -10.0 (s, Ph2PCH2-
imidazolium).
(C-N)2Pd(CH3)2, 3. In the glovebox 80 mg (0.317 mmol) of
Pd(tmeda)Me2 was placed in a Schlenk. A second Schlenk was
charged with 194 mg (0.636 mmol) of free carbene. The contents
of the two Schlenks were dissolved in dry-degassed THF, and the
solution of Pd(tmeda)Me2 was transferred at room temperature via
cannula to the solution of the free carbene. The reaction mixture
was stirred overnight, volatiles were removed, and the pale yellow
solid was washed with light petroleum ether. The compound is air
stable and soluble in dichloromethane, toluene, and ether. Yield:
(P∧C)Pd(CH3)2, 1. This was prepared following a previously
reported procedure7 from 0.080 g (0.317 mmol) of Pd(tmeda)Me2,
0.161 g of [(P∧C)H]+Br- (1 equiv), and 0.070 g of KN(SiMe3)2
1
(1.1 equiv). Yield: 0.130 g (73%). H NMR δ (CD2Cl2): -0.53
3
3
(3H, d, JPH ) 8.1 Hz, PdCH3), -0.32 (3H, d, JPH ) 8.2 Hz,
PdCH3), 1.06 [6H, d, 3JHH ) 6.9 Hz, CH(CH3)2], 1.23 [6H, d, 3JHH
) 6.9 Hz, CH(CH3)2], 2.55 [2H, sept, 3JHH ) 6.9 Hz, CH(CH3)2],
2
4.21 [2H, d, JPH ) 3.3 Hz, (PPh2CH2-ylid)PdMe2], 6.84 (1H, s,
1
ylidene backbone), 7.32 (4H, br d, aromatic), 7.44 (6H, br s,
aromatic), 7.62 (4H, m, aromatic). 13C{1H} NMR δ (CD2Cl2): -0.4
(d, 2JPC ) 18.2 Hz, PdCH3), 0.2 (d, 2JPC ) 19.2 Hz, PdCH3), 22.4
[s, CH(CH3)2], 23.7 [s, CH(CH3)2], 27.5 [s, CH(CH3)2], 51.9 (d,
1JPC ) 24.5 Hz, [PPh2CH2-ylid)PdMe2], 117.9 (s, ylidene back-
bone), 122.6 (s, ylidene backbone), 123.8 (s, aromatic), 127.8 (d,
JPC ) 8.85 Hz, aromatic), 128.1 (s, aromatic), 128.3 (s, aromatic),
129.4 (s, aromatic), 131.7 (d, JPC ) 26.1 Hz, aromatic), 132.3 (d,
JPC ) 13.5 Hz, aromatic), 144.9 (s, ipso carbon of the aryl group),
190.4 (s, NCN). 31P{1H} NMR δ (CD2Cl2): 32.7 [s, (PPh2CH2-
ylid)PdMe2]. Anal. Found: C, 63.71; H, 6.38; N, 4.75. Calcd for
C30H37N2PPd: C, 63.99; H, 6.62; N, 4.98. X-ray-quality crystals
were grown by slow diffusion of petroleum ether into a THF
solution of the complex.
201 mg (85%). H NMR δ (d8-toluene): -0.11 (6H, s, PdCH3),
3
3
0.82 (6H, d, JHH ) 6.9 Hz, CH(CH3)2), 0.98 (6H, d, JHH ) 6.7
Hz, CH(CH3)2), 1.09 (6H, d, 3JHH ) 6.7 Hz, CH(CH3)2), 1.42 (6H,
3
3
d, JHH ) 6.7 Hz, CH(CH3)2) 2.62 (2H, sept, JHH ) 6.7 Hz,
3
CH(CH3)2), 3.12 (2H, sept, JHH ) 6.7 Hz, CH(CH3)2), 6.45 (1H,
3
d, JHH ) 1.9 Hz, ylidene backbone), 6.51 (1H, d, JHH ) 1.6 Hz,
aromatic), 6.55 (1H, d, JHH ) 1.4 Hz), 6.63 (1H, ddd, JHH ) 1.1,
3
4.9, and 7.1 Hz), 6.83 (2H, t, JHH ) 7.8 Hz, aromatic) 6.85 (1H,
d, 3JHH ) 1.9 Hz, ylidene backbone), 6.92 (3H, m, aromatic), 7.03
(1H, br s, aromatic), 7.09 (1H, d, JHH ) 1.6 Hz, aromatic), 7.11
(1H, m, aromatic), 8.14 (2H, dd, JHH ) 1.1 Hz, 3.8 Hz, aromatic),
8.42 (2H, d, JHH ) 2.2 Hz, aromatic), 9.33 (2H, d, JHH ) 8.2 Hz,
aromatic). 13C{1H} NMR δ (d8-toluene): -0.4 (s, PdCH3), 22.1
(s, CH(CH3)2), 24.1 (s, CH(CH3)2), 26.5 (s, CH(CH3)2), 27.3 (s,
CH(CH3)2), 28.6 (s, CH(CH3)2), 28.8 (s, CH(CH3)2), 117.1 (s,
ylidene backbone), 119.1 (ylidene backbone), 121.3 (s aromatic),
123.8 (s, aromatic), 124.7 (s, aromatic), 125.2 (s, aromatic), 130.7
(s, aromatic), 138 (s, aromatic), 138.9 (s, aromatic), 145.7 (s,
aromatic), 147 (s, aromatic), 148.6 (s, aromatic), 152.2 (s, aromatic),
200.1 (s, NCN). Anal. Found: C, 67.61; H, 7.16; N, 11.01. Calcd
for C42H52N6Pd: C, 67.50; H, 7.01; N, 11.25. Colorless crystals
were grown, by cooling a saturated solution of the compound in
toluene to 5 °C.
(κ2-C-N)Pd(CH3)2, 2a. To a solution of 0.080 g (0.317 mmol)
of Pd(tmeda)Me2 in 30 mL of THF at -78 °C was added a
precooled solution of 0.135 g (0.443 mmol, 1.4 equiv) of the free
carbene in the same solvent (10 mL). The reaction mixture was
allowed to reach room temperature slowly overnight. After removal
of the volatiles under reduced pressure the pale yellow solid residue
was washed with light petroleum (10 mL) and cold ether (-30 °C,
5 mL) to give the product. Yield: 0.050 g (42%). The product
deposits Pd black after prolonged exposure to toluene or dichlo-
1
romethane. H NMR δ (CD2Cl2): 0.01 (3H, s, PdCH3), 0.12 (3H,
[(C-N)Pd(methallyl)](BPh4), 4. A precooled (-78 °C) solution
of (C-N) (61 mg, 0.20 mmol) in THF was added to a cold (-78
°C) solution of 40 mg (0.10 mmol) of Pd[(η3-methallyl)Cl]2 in the
same solvent. The mixture was allowed to reach room temperature
and stirred for 20 min, and then solid NaBPh4 (68 mg, 0.20 mmol)
was added in one portion. After stirring for an additional 15 min
the volatiles were removed under reduced pressure and the solid
residue was dissolved in CH2Cl2 and filtered through Celite. The
organic filtrate was evaporated to dryness. Yield: 110 mg (69%).
Colorless X-ray-quality crystals were obtained by layering a CH2-
3
s, PdCH3), 1.09 (6H, d, JHH ) 6.9 Hz, CH(CH3)2), 1.21 (6H, d,
3
3JHH ) 6.9 Hz, CH(CH3)2), 2.73 (2H, sept, JHH ) 6.9 Hz,
3
CH(CH3)2), 6.91 (1H, d, JHH ) 1.5 Hz, ylidene backbone), 7.22
(2H, distorted ddd, JHH ) 5.7, 5.4, and 7.6 Hz, pyridine), 7.31 (2H,
d, 3JHH ) 7.9 Hz, aromatic), 7.52 (1H, t, 3JHH ) 7.9 Hz, aromatic),
7.92 (1H, distorted ddd, JHH ) 1, 0.7, and 8.2 Hz, pyridine), 8.06
(1H, d, 3JHH ) 1.5 Hz, ylidene backbone), 9.18 (br dt, 1H, pyridine).
13C{1H} NMR δ (CD2Cl2): -0.1 (s, PdCH3), 0.0 (s, PdCH3), 21.7
(s, CH(CH3)2), 22.2 (s, CH(CH3)2), 26.9 (s, CH(CH3)2), 118.9 (s,
ylidene backbone), 120.4 (s, ylidene backbone), 121.4 (s, aromatic),
122.1 (s, aromatic), 123.7 (s, aromatic), 128.3 (s, aromatic), 141.5
(s, aromatic), 143.8 (s, aromatic), 144.1 (s, aromatic), 145.5 (s,
1
Cl2 solution with light petroleum. H NMR δ (CD2Cl2): 1.11-
1.23 (12H, m, CH(CH3)2), 1.82 (3H, s, CH2CCH3CH2), 2.01-2.41
(4H, br m, CH(CH3)2 and CH2CCH3CH2), 3.25 (1H, s, CH2CCH3-

262 Organometallics, Vol. 26, No. 2, 2007
Danopoulos et al.
CHH), 4.12 (1H, s, CH2CCH3CHH), 6.84 (2H, tt, JHH ) 1.3 Hz,
8.5 Hz, aromatic), 7.09 (4H, two coinciding d, JHH ) 7.4 Hz, 7.5
Hz, aromatic), 7.12 (1H, d, 3JHH ) 2.1 Hz, ylidene backbone), 7.35
(12H, br m, aromatic), 7.55 (1H, br t, aromatic), 8.17 (4H, m,
H(OEt2){B[(3,5-CF3)2C6H2]4} (92 mg, 1 equiv), and 4-tert-bu-
1
tylpyridine (9 µL, 1 equiv). Yield: 110 mg (68%). H NMR δ
(CD2Cl2): -0.09 (3H, d, 3JPH ) 2.6 Hz, PdCH3), 0.92 (6H, d, 3JHH
3
) 6.8 Hz, CH(CH3)2), 1.01 (6H, d, JHH ) 6.8 Hz, CH(CH3)2),
3
1.22 (9H, s, 4-tBu-pyridine), 2.30 (2H, sept, JHH ) 6.7 Hz,
3
aromatic), 8.45 (1H, d, JHH ) 2.1 Hz, ylidene backbone), 8.51
CH(CH3)2), 4.93 (2H, d, 2JPH ) 6.6 Hz, [PPh2CH2-ylidene]PdMe-
(1H, ddd, JHH ) 0.8 Hz, 1.7 Hz, 5.4 Hz, aromatic). 13C{1H} NMR
δ (CD2Cl2): 22.? (s, CH(CH3)2), 22.1 (s, CH(CH3)2), 22.4 (s,
CH2CCH3CH2), 23.8 (s, CH(CH3)2), 24.1 (s, CH(CH3)2), 27.2 (s,
CH(CH3)2), 27.3 (s, CH(CH3)2), 48.1 (s, CH2CCH3CH2), 52.5 (s,
CH2CCH3CH2), 118.1 (s, ylidene backbone), 119 (ylidene back-
bone), 121.3 (s aromatic), 123.7 (s, aromatic), 124.8 (s, aromatic),
125.3 (s, aromatic), 125.7 (s, aromatic), 126.3 (br s, aromatic), 127.2
(s, aromatic), 127.9 (s, CH2CCH3CH2), 128.8 (s, aromatic), 132.5
(s, aromatic) 138.1 (s, aromatic), 139.2 (s, aromatic), 145.7 (s,
aromatic), 146.6 (s, aromatic), 149.4 (s, ipso aromatic to aryl group).
Anal. Found: C, 73.11; H, 6.19; N, 5.21. Calcd for C48H50N3BPd:
C, 73.33; H, 6.41; N, 5.35.
(4-tBu-py)+), 7.01 (4H, m, aromatic), 7.33 (2H, m, aromatics), 7.46
(10H, m, aromatics), 7.79 (11H, m, aromatics), 8.01 (2H, d, JHH
)
4.5 Hz, aromatics). 13C{1H} NMR δ (CD2Cl2): 1.9 (s, PdCH3),
23.2 (s, CH(CH3)2), 25.1 (s, CH(CH3)2), 28.6 (s, 4-C(CH3)3-
pyridine), 30.2 (s, CH(CH3)2), 53.6 (d, 1JPC ) 48.9 Hz, [PPh2CH2-
ylidene]PdMe(4-tBu-py)+), 117.8 (s, ylidene backbone), 122.7 (s,
ylidene backbone), 123.1 (s, aromatic), 124.3 (s, aromatic), 126.3
(s, aromatic), 126.8 (br s, aromatic), 128.9 (s, aromatic), 129.2 (s,
aromatic), 129.5 (s, aromatic), 129.8 (d, JPC ) 15.2 Hz, aromatic),
130.3 (s, aromatic), 130.9 (s, aromatic), 132.8 (s, aromatic), 133.5
(d, JPC ) 20.0 Hz, aromatic), 135.2 (s, aromatic), 145.6 (s,
1
aromatic), 150.1 (s, aromatic), 163.1 (q, JCF ) 49.9 Hz, CF3 s).
[(C∼P)Pd(methallyl)](BPh4), 5. A precooled solution of the
ligand P∧C generated from 212 mg (0.41mmol) of [(P∧C)H]+Br-
and 89 mg (0.45 mmol) of KN(SiMe3)2 as described above was
added to a solution of 80 mg (0.20 mmol) of Pd[(η3-methallyl)-
Cl]2 in THF. After completion of the addition the reaction mixture
was allowed to reach room temperature and stirred for 1 h. Addition
of solid NaBPh4 (137 mg, 0.40 mmol) was followed by removal
of the volatiles under reduced pressure extraction of the residue in
CH2Cl2, filtration through Celite, and evaporation of the solvents
from the organic filtrates. Yield: 90 mg (90%). Colorless X-ray-
quality crystals were grown by layering a CH2Cl2 solution with
ether. 1H NMR δ (CD2Cl2): 0.71-0.89 (12H, m, CH(CH3)2), 1.35
(3H, s, CH2CCH3CH2), 2.28-2.37 (2H, m, CHHCCH3CH2 and
CH(CH3)2), 2.43 (1H, sept, 3JHH ) 6.8 Hz, CH(CH3)2), 2.91-3.09
(3H, m, CHHCCH3CH2 and [(PPh2CH2CH2ylid)Pd(η3-methal-
lyl)]+), 3.12 and 3.73 (each 1H, br s, CH2CCH3CH2), 4.82 (2H, m,
[(PPh2CH2CH2ylid)Pd(η3-methallyl)]+), 6.91 (1H, s, ylidene back-
bone), 7.11-7.33 (6H, m, aromatic), 7.35 (5H, br d, aromatic),
7.52 (12H, br s, aromatic), 7.61-7.72 (10H, m, aromatic), 8.17
(1H, s, ylidene backbone). 13C{1H} NMR δ (CD2Cl2): 21.9 (s, CH-
(CH3)2), 22.5 (s, CH(CH3)2), 22.3 (s, CH2CCH3CH2), 23.9 (s, CH-
(CH3)2), 24.2 (s, CH(CH3)2), 26.5 (d, 1JPC ) 27.9 Hz, [(PPh2CH2-
CH2-ylid)Pd(η3-methallyl)]Cl), 27.4 (s, CH(CH3)2), 27.6 (s,
CH(CH3)2), 47.? (s, CH2CCH3CH2), 62.5 (d, 2JPC ) 9.1 Hz, [(PPh2-
CH2CH2-ylid)Pd(η3-methallyl)]+), 68.4 (d, JPC ) 30.3 Hz, CH2-
CCH3CH2), 117.6 (s, ylidene backbone), 122.8 (s, ylidene back-
bone), 123.1 (s, aromatic), 124.2 (d, JPC ) 3.6 Hz, aromatic), 125.8
(s, aromatic), 127.7 (s, CH2CCH3CH2), 127.9 (s, aromatic), 128.2
(d, JPC ) 11.7 Hz, aromatic), 128.6 (br s., aromatic), 129.9 (s,
aromatic), 130.3 (d, JPC ) 8.6 Hz), 131.2 (br s, aromatic), 133.2
(s, aromatic), 134.4 (br s, aromatic), 134.9 (s, aromatic), 135.2 (s,
aromatic), 149.6 (s, ipso aromatic to aryl group). 31P{1H} NMR δ
(CD2Cl2): 22.3 (s, [(PPh2CH2CH2-ylid)Pd(η3-methallyl)]+). Anal.
Found: C, 74.05; H, 6.28; N, 2.89. Calcd for C57H60N2BPPd: C,
74.31; H, 6.56; N, 3.04.
31P{1H} NMR δ (CD2Cl2): 53.8 (s, [PPh2CH2-ylidene]PdMe(4-
tBu-py)+). 19F{1H} NMR δ (CD2Cl2): -63.1 (s, CF3). Anal.
Found: C, 54.25; H, 3.22; N, 2.50. Calcd for C70H55N3F24BPPd:
C, 54.51; H, 3.59; N, 2.72.
[(C-N*)Pd(CH3)(py)]+[BArF ]-, 7. This was prepared follow-
4
ing the general method starting from 0.080 g (0.18 mmol) of 2b,
0.165 g (1 equiv) of [H(OEt2)]+{B[(3,5-CF3)2C6H2]4}-, and 15 µL
1
(1.1 equiv) of pyridine. Yield: 0.200 g (82%). H NMR δ (CD2-
3
Cl2): -0.09 (3H, s, PdCH3), 1.01 (6H, d, JHH ) 6.9 Hz, CH-
3
(CH3)2), 1.09 (6H, d, JHH ) 6.9 Hz, CH(CH3)2), 2.61 (2H, sept,
3JHH ) 6.9 Hz, CH(CH3)2), 2.74 (3H, s, 3-CH3 pyridine backbone),
3
7.13 (1H, d, JHH ) 2.2 Hz, ylidene backbone), 7.20 (1H, m,
aromatics), 7.33 (3H, m, aromatics), 7.53 (6H, br s, aromatics),
7.71 (9H, br s, aromatics), 7.86 (1H, dd, JHH ) 1.0 Hz, 7.7 Hz,
3
aromatic), 7.91 (1H, br s, aromatic), 8.07 (1H, d, JHH ) 2.2 Hz,
ylidene backbone), 8.54 (2H, dd, JHH ) 1.5 Hz, 6.2 Hz, aromatic).
13C{1H} NMR δ (CD2Cl2): 7.7 (s, PdCH3), 21.5 (s, CH(CH3)2),
23.8 (s, 3-CH3 pyridine backbone), 25.2 (s, CH(CH3)2), 29.4 (s,
CH(CH3)2), 118.2 (s, ylidene backbone), 119.8 (s, ylidene back-
bone), 121.3 (s, aromatic), 124.0 (s, aromatic), 124.2 (s, aromatic),
124.6 (s, aromatic), 125.0 (s, aromatic), 126.7 (s, aromatic), 127.2
(s, aromatic), 129.8 (q, 2JCF ) 30.2 Hz), 131.7 (s, aromatic), 135.1
(s, aromatic), 135.6 (s, aromatic), 139.8 (s, aromatic), 145.5 (s,
aromatic), 145.8 (s, aromatic), 146.1 (s, aromatic), 151.8 (s,
1
aromatic), 162.3 (quart, JCF ) 49.7 Hz, CF3 s). 19F{1H} NMR δ
(CD2Cl2): -62.8. Anal. Found: C, 51.02; H, 3.12; N, 3.92. Calcd
for C59H45N4F24BPd: C, 51.23; H, 3.28; N, 4.05. Colorless crystals
of 8 were grown by slow diffusion of petroleum ether in an ether
solution.
X-ray Crystallography. A summary of the crystal data, data
collection, and refinement for compounds 1, 2b, 3, 4, 5, and 7 are
given in Table 1.
All data sets were collected on a Enraf-Nonius Kappa CCD area
detector diffractometer with an FR591 rotating anode (Mo/Ka
radiation) and an Oxford Cryosystems low-temperature device
operating in ω scanning mode with ψ and ω scans to fill the Ewald
sphere. The programs used for control and integration were Collect,
Scalepack, and Denzo.17 The crystals were mounted on a glass fiber
with silicon grease, from Fomblin vacuum oil. All solutions and
refinements were performed using the WinGX package18 and all
software packages within. All non-hydrogen atoms were refined
using anisotropic thermal parameters, and hydrogens were added
using a riding model.
General Method for the Synthesis of the Cations 6 and 7.
Two separate Schleck tubes were charged under N2 with 1 equiv
of the dimethyl complex and the acid [H(OEt2)]+{B[(3,5-
CF3)2C6H2]4}-. The solids were dissolved in CH2Cl2, the solutions
were cooled to -78 °C, and the acid was added dropwise to the
dimethyl complex. Upon addition, the color changed from yellow
to almost colorless. The reaction mixture was allowed to warm to
-40 °C, when 1.1 equiv of dry-degassed pyridine derivative was
added by means of a microsyringe. Then the mixture was allowed
to reach room temperature slowly and stirred for 30 min. Removal
of the volatiles under reduced pressure and washing of the solid
residue with petroleum ether gave the pure cation, which was dried
under vacuum.
Computational Details. All calculations employed the Gaussian
98 program18 and used DFT calculations with the BP86 functional
throughout. The same approach was used for both models 1 and 2.
(17) Farrugia, L. J. J. Appl. Crystallogr. 1999, 32, 83.
(18) Frisch M. J.; et al. Gaussian 98, Revision A.11.4; Gaussian, Inc.:
Pittsburgh, PA, 2001.
[(C∧P)Pd(CH3)(But-py)]+[BArF ]-, 6. This was prepared fol-
4
lowing the general method starting from 1 (55 mg, 0.098 mmol),

N-Heterocyclic Carbene Complexes of Palladium
Organometallics, Vol. 26, No. 2, 2007 263
Pd and P centers were described using the Stuttgart RECPs and
the associated basis sets.19 For P an extra set of d-orbital polarization
functions was added (úP ) 0.387).20 6-31G** basis sets were used
for C, N, and H atoms.21 For model 1 all stationary points were
characterized as minima by computation of the Hessian matrix (all
positive eigenvalues). The size of the species in model 2 precluded
the routine computation of frequency analyses at the full DFT level,
and so for consistency the energies reported in the text do not
include zero-point energy corrections. Optimizations were based
on the experimental structures of (C∼P)Pd(CH3)2 and trans-P
[C∼P)Pd(CH3)(py)]+. Transition states located with model 2 were,
however, fully characterized and show one unique imaginary
frequency.
Acknowledgment. We thank Johnson Matthey Catalysts, the
University of Southampton, and Heriot-Watt University for
financial support (to N.T. and C.S., respectively). We also thank
Johnson Matthey Catalysts for a loan of palladium salts.
Supporting Information Available: Full details of the X-ray
crystal structures, including complete tables of crystal data, atomic
coordinates, bond lengths and angles, and positional and anisotropic
thermal parameters. Cartesian coordinates and energies of all
computed species. Full ref 18. This material is available free of
charge via the Internet at http://pubs.acs.org.
(19) Andrae, D.; Ha¨usserman, U.; Dolg, M.; Stoll, H. Preuss Theor. Chim.
Acta 1990, 77, 123.
(20) Ho¨llwarth, A.; Bo¨hme, M.; Dapprich, S.; Ehlers, A. W.; Gobbi,
A.; Jonas, V.; Ko¨hler, K. F.; Stegmann, R.; Veldkamp, A.; Frenking, G.
Chem. Phys. Lett. 1993, 208, 237.
(21) (a) Hehre, W. J.; Ditchfield, R.; Pople, J. A. J. Chem. Phys. 1972,
56, 2257. (b) Hariharan, P. C.; Pople, J. A. Theor. Chim. Acta 1973, 28,
213.
OM0608408