The η3-furfuryl ligand: Plausible catalytic intermediates and heterocyclic η3-benzyl analogues with superior binding ability

Article abstract of DOI:10.1021/om100786h

Five-membered heterocyclic analogues of the well-known η³- allyl and η³-benzyl ligands, e.g., η³-thienyl and η³-furfuryl, have been proposed as catalytic intermediates, yet are poorly understood and have never been structurally confirmed. Herein we characterize an η³-furfuryl complex by multinuclear NMR spectroscopy and crystallography and offer a computational survey of allyl and allyl-like ligands. Crystallographically and computationally, we show that the η³-furfuryl ligand may bind more strongly than corresponding η³-benzyl ligands and is plausible as a catalytic intermediate.

Full text of DOI:10.1021/om100786h

Organometallics 2010, 29, 4431–4433 4431
DOI: 10.1021/om100786h
The η3-Furfuryl Ligand: Plausible Catalytic Intermediates and
Heterocyclic η3-Benzyl Analogues with Superior Binding Ability
€
€
Rian D. Dewhurst,* Robert Muller, Martin Kaupp,* Krzysztof Radacki, and Kathrin Gotz
Institut fu€r Anorganische Chemie and Institut fu€r Physikalische und Theoretische Chemie,
€
€
€
Universitat Wurzburg, Am Hubland, 97074 Wurzburg, Germany
Received August 12, 2010
Scheme 1. (Top) Well-Known η3-Allyl (I) and η3-Benzyl (II)
Ligands and Their Five-Membered Heterocyclic Analogues
(III and IV); (Bottom) Synthesis of 1
Summary: Five-membered heterocyclic analogues of the
well-known η3-allyl and η3-benzyl ligands, e.g., η3-thienyl and
η3-furfuryl, have been proposed as catalytic intermediates, yet are
poorly understood and have never been structurally confirmed.
Herein we characterize an η3-furfuryl complex by multinuclear
NMR spectroscopy and crystallography and offer a computational
survey of allyl and allyl-like ligands. Crystallographically and
computationally, we show that the η3-furfuryl ligand may bind
more strongly than corresponding η3-benzyl ligands and is
plausible as a catalytic intermediate.
η3-Allyl complexes of transition metals (I, Scheme 1) are
an indispensable class of organometallic species, as they feature
one of the simplest possible hydrocarbon π-ligands.1 They are
known to be intermediates in a number of catalytic processes,
particularly the nucleophilic substitution of allylic esters, known
as the Tsuji-Trost reaction.2 In 1966, analogous η3-benzyl com-
plexes (II, Scheme 1) were prepared3 and crystallographically
characterized.4 Like its allylic cousin, the η3-benzyl ligand has
since become the linchpin in a number of catalytic processes,
particularly the Tsuji-Trost-style nucleophilic substitution of
benzylic esters5 and the catalytic polymerization,6 hydrobora-
tion,7 hydrosilylation,8 and hydroamination9 of vinylarenes.
Intermediate η3-benzyl complexes also influence the rates of
aryl vs benzyl C-H activation of alkyl benzenes by Rh and Pt
complexes.10
monometallic η3-heterocyclic chemistry and total absence of
structural data are surprising. A report from 1969 disclosed the
synthesis of two η3-thienyl (IV, Scheme 1) complexes, while a
paper from 1981 reported two η3-furfuryl (III) complexes.11
However, in all cases structural assignment was made primar-
1
ily via H NMR and IR spectroscopy, while X-ray crystal-
lographic data for heterocyclic analogues of η3-benzyl com-
plexes exist only for chelating, multinuclear systems.12
A
limited number of palladium-catalyzed coupling reactions
with heterocyclic “allylic” esters exist in the literature, impli-
cating the structurally unconfirmed η3-coordination mode of
the precursor heterocycles.13
Given the importance of η3-allyl and η3-benzyl ligands in
organometallic transformations, the scarcity of corresponding
Our objective was to target stable complexes containing
heterocyclic analogues of the η3-benzyl ligand, such as
η3-furfuryl derivatives (III, Scheme 1), in order to confirm their
validity as catalytic intermediates or explain the infrequency of
heterocyclic Tsuji-Trost reactions in the literature. Treatment
of commercially available methyl 5-(chloromethyl)-2-furancar-
boxylate with [Pd(PPh3)4], followed by silver tetrafluoroborate,
led to a yellow solid formulated as the η3-furfuryl complex
1 (Scheme 1). The compound displayed two doublets in the
31P{1H} NMR spectrum (δP 30.6, 20.6; JPP = 44.5), matching
well with the comparable η3-benzyl complex [Pd(η3-benzyl)-
(PPh3)2][PF6] (δP 32.2, 22.7; JPP = 41)14 and confirming the cis
arrangement of the two phosphorus nuclei in solution.
*Corresponding author. E-mail: r.dewhurst@uni-wuerzburg.de.
(1) Collman, J. P.; Hegedus, L. S.; Norton, J. R.; Finke, R. G.
Principles and Applications of Organotransition Metal Chemistry; Oxford
University Press: Oxford, 1987.
(2) Trost, B. M.; Crawley, M. L. Chem. Rev. 2003, 103, 2921.
(3) King, R. B.; Fronzaglia, A. J. Am. Chem. Soc. 1966, 88, 709.
(4) Cotton, F. A.; LaPrade, M. D. J. Am. Chem. Soc. 1968, 90, 5418.
(5) (a) Kuwano, R.; Kondo, Y.; Matsuyama, Y. J. Am. Chem. Soc.
2003, 125, 12104. (b) Kuwano, R. Synthesis 2009, 1049.
(6) (a) Rix, F. C.; Brookhart, M.; White, P. S. J. Am. Chem. Soc. 1996,
118, 2436. (b) Nozaki, K.; Komaki, H.; Kawashima, Y.; Hiyama, T.;
Matsubara, T. J. Am. Chem. Soc. 2001, 123, 534.
(7) Hayashi, T.; Matsumoto, Y.; Ito, Y. Tetrahedron: Asymmetry
1991, 2, 601.
The 1H NMR spectrum of 1 at 298 K showed two doublet-
of-doublet signals (C3H, C4H) and two broad singlets (CH2).
(8) LaPointe, A. M.; Rix, F. C.; Brookhart, M. J. Am. Chem. Soc.
1997, 119, 906.
(9) (a) Kawatsura, M.; Hartwig, J. F. J. Am. Chem. Soc. 2000, 122,
9546. (b) Nettekoven, U.; Hartwig, J. F. J. Am. Chem. Soc. 2002, 124, 1166.
(c) Johns, A. M.; Utsunomiya, M.; Incarvito, C. D.; Hartwig, J. F. J. Am.
Chem. Soc. 2006, 128, 1828. (d) Johns, A. M.; Tye, J. W.; Hartwig, J. F.
J. Am. Chem. Soc. 2006, 128, 16010.
(10) (a) Shimada, S.; Batsanov, A. S.; Howard, J. A. K.; Marder, T. B.
Angew. Chem., Int. Ed. 2001, 40, 2168. (b) Lam, W. H.; Lam, K. C.; Lin, Z.;
Shimada, S.; Perutz, R. N.; Marder, T. B. Dalton Trans. 2004, 1556. (c)
Heyduk, A. F.; Driver, T. G.; Labinger, J. A.; Bercaw, J. E. J. Am. Chem. Soc.
2004, 126, 15034. (d) Williams, T. J.; Caffyn, A. J. M.; Hazari, N.; Oblad,
P. F.; Labinger, J. A.; Bercaw, J. E. J. Am. Chem. Soc. 2008, 130, 2418.
(11) (a) King, R. B.; Kapoor, R. N. Inorg. Chem. 1969, 8, 2535. (b)
Onishi, M.; Ito, T.; Hiraki, K. J. Organomet. Chem. 1981, 209, 123.
(12) (a) Wong, W.-Y.; Ting, F.-L.; Lam, W.-L. J. Chem. Soc., Dalton
Trans. 2001, 2981. (b) Tunik, S. P.; Khripoun, V. D.; Balova, I. A.; Borovitov,
M. E.; Domnin, I. N.; Nordlander, E.; Haukka, M.; Pakkanen, T. A.; Farrar,
D. H. Organometallics 2003, 22, 3455.
(13) (a) Urata, H.; Suzuki, H.; Moro-oka, Y.; Ikawa, T. Bull. Chem.
Soc. Jpn. 1984, 57, 607. (b) Lindsey, C. C.; O'Boyle, B. M.; Mercede, S. J.;
Pettus, T. R. R. Tetrahedron Lett. 2004, 45, 867.
(14) Lin, Y.-S.; Yamamoto, A. Bull. Chem. Soc. Jpn. 1998, 71, 723.
r
2010 American Chemical Society
Published on Web 09/24/2010
pubs.acs.org/Organometallics

4432 Organometallics, Vol. 29, No. 20, 2010
Dewhurst et al.
Figure 2. Selected structural parameters and NPA charges
of model complexes 10-40 (RI-BP86/def2-TZVPP//RI-BP86/
def2-TZVP). Atomic charges are printed in italics; values in
parentheses correspond to charges in the free ligand. Charge
transfer between the coordinated ligand and the metal fragment
is in bold.
complexes) of 1.06.5b Thus, the η3-furfuryl ligand of 1 is roughly
equivalent to the most symmetrical examples of known
η3-benzyl ligands on palladium (with ratios ranging from 1.05 to
1.29). A further feature of the structure of 1 is the increase of
˚
the furan C2-C3 double bond length (1.402(4) A) compared to
˚
the unbound C4-C5 bond (1.342(4) A). This lengthening is
approximately 5% of the shortest distance, presumably reflect-
ing back-donation from palladium into the C2-C3 π* anti-
bonding orbital. Overall, the crystallographic data of complex 1
show that the η3-furfuryl ligand can match even the most “allyl-
like” of known η3-benzyl ligands in terms of symmetry and
bond lengths.
To obtain a consistent comparison of the structural and
electronic features of allyl-type ligands I-IV (Scheme 1), we
performed calculations based on density functional theory
(DFT) at RI-BP86/def2-TZVPP//RI-BP86/def2-TZVP levels
with a quasirelativistic pseudopotential on palladium (Figure 2).
The calculations were undertaken with a simplified model of
1 (10, class III), the analogous thienyl complex 20 (class IV),
parent allyl complex 30 (class I), and benzyl complex 40 (class II).
The calculations explicitly included BF4- counterions.
1
Figure 1. (Top) Section of the variable-temperature H NMR
spectrum of 1 showing the methylene proton signals. (Bottom)
Molecular structure of the cation of 1. Thermal ellipsoids are set
˚
at 50% probability. Selected bond lengths [A] and angles [deg]
for 1: Pd1-C6 2.163(3), Pd1-C2 2.236(2), Pd1-C3 2.289(3),
C2-C3 1.402(4), C3-C4 1.438(4), C4-C5 1.342(4); C6-C2-O1
121.6(2), O1-C2-C3 109.4(2), C3-C2-C6 126.8(3).
A variable-temperature 1H NMR experiment showed resolution
of the broad methylene proton signals into doublet-of-doublet
signals at approximately 243 K. However, no dynamic beha-
vior was detected in the 31P NMR spectrum of 1 (CD2Cl2) in
the range 193-303 K. This room-temperature fluxionality can
be attributed to fast η3-η1-η3 rearrangement to a T-shaped
intermediate (thus broadening the methylene proton signals but
leaving the phosphorus nuclei inequivalent), but without accom-
panied flipping through a Y-shaped transition state, for which
we would expect a single broad signal for the methylene protons
Computed energies of ligand exchange reactions show the
heterocyclic complexes 10 and 20 to be of similar stability. Both
are about 41-43 kJ mol-1 more stable than benzyl complex 40
and about the same amount (46-48 kJ mol-1) less stable than
the parent allyl complex 30. The energy required to computa-
tionally force the ligands from η3 to η1 binding modes by
imposing BF4- coordination to Pd (this mode represents a local
minimum on the potential energy surface in all three cases)
is smallest for the benzyl ligand and again similar for the furfuryl
and thienyl ligands, i.e., 40 (25.0 kJ/mol) < 20 (41.5 kJ/mol) ≈ 10
(45.3 kJ/mol). Trends in the bonding may be inferred from key
optimized structural parameters (Figure 2). The aforementioned
d(Pd-C3)/d(Pd-C1) ratio becomes more symmetrical going
from the benzyl to the allyl system: 40 (1.18) > 20 (1.15) > 10
(1.12) > 30 (1.00). The overestimated Pd-C3 bond lengths reflect
the neglect of dispersion interactions in the RI-BP86/def2-TZVP
optimizations. Inclusion of semiempirical dispersion corrections
shortens the Pd-C3 bonds for all systems with aromatic ligands
but does not alter any of the main conclusions drawn.
1
in the H NMR spectrum and broadening of the 31P NMR
signals.3,5b,9c,15
The solid-state structure of 1 derived from crystallographic
data confirmed the formulation (Figure 1). The furfuryl ligand
is unsymmetrically bound, with a d(Pd-C3)/d(Pd-C6) ratio
(a measure of symmetry defined by Kuwano for η3-benzyl
(15) (a) Cotton, F. A.; Marks, T. J. J. Am. Chem. Soc. 1969, 91, 1339.
€
(b) Werner, H.; Schafer, M.; N€urnberg, O.; Wolf, J. Chem. Ber. 1994, 127, 27.

Communication
Organometallics, Vol. 29, No. 20, 2010 4433
Computed natural atomic (NPA) charges provide insight
into the two main aspects affecting the symmetry and
strength of the η3 coordination: the competition between
aromaticity and metal coordination at the C3 atom, and the
electronegativity of the heteroatom, which partly determines
the charge on the exocyclic methylene carbon. Thereby, the
benzyl complex 40, clearly the most aromatic, exhibits the
least negative charge on C3 and thus the weakest and longest
C3-Pd contact. The heterocyclic ligands in 10 and 20 have
diminished cyclic delocalization and thus allow a more
negative charge on C3 and consequently stronger coordina-
tion of this site. On the other hand, the electronegativity
of the oxygen heteroatom in 10 leads to a pronounced
O-C2-C3 charge alternation (Figure 2), giving the largest
negative charge on the methylene carbon in the free ligand.
As a consequence, the furfuryl complex 10 (followed closely
by 20) exhibits the most symmetrical η3 coordination of the
three ligands, leading to an overall stronger bond.
heterocycles,13 has been synthesized, crystallographically
characterized, and studied computationally. The metrics gar-
nered from the crystallographic and computational study of 1
strongly suggest that when bound to suitable palladium(II)
fragments, η3-furfuryl (and -thienyl) ligands may lead to signi-
ficantly more stable complexes than η3-benzyl ligands, an effect
partly due to the reduced aromaticity of the heterocycles.
Further studies on the synthesis, reactivity, and electronic
structure of η3-furfuryl (and related heterocyclic) complexes
are in progress.
Acknowledgment. Financial support was provided
€
by the University of Wurzburg and the Deutsche
Forschungsgemeinschaft (DFG). We thank Prof. H.
Braunschweig for continued support and N. Wiechert
for some initial calculations.
Supporting Information Available: Experimental, spectro-
scopic, and computational details. This information is available
free of charge via the Internet at http://pubs.acs.org
In summary, a stable mononuclear η3-furfuryl complex, long
assumed to be an intermediate in catalytic transformations of