Facile access to key reactive intermediates in the Pd/PR 3-catalyzed telomerization of 1,3-butadiene

Article abstract of DOI:10.1002/anie.201003090

Found at last: A simple and efficient one-pot synthesis procedure leads to several cationic [Pd(1,2,3,7,8-⁵-octa-2,7-dien-1-yl)(PR ₃)]⁺ complexes (see structure; Pd pink, O red, P orange), which are the long sought-after k

Full text of DOI:10.1002/anie.201003090

Communications
DOI: 10.1002/anie.201003090
Nucleophilic Addition
Facile Access to Key Reactive Intermediates in the Pd/PR3-Catalyzed
Telomerization of 1,3-Butadiene**
Peter J. C. Hausoul, Andrei N. Parvulescu, Martin Lutz, Anthony L. Spek,
Pieter C. A. Bruijnincx, Bert M. Weckhuysen,* and Robertus J. M. Klein Gebbink*
The Pd-catalyzed telomerization of 1,3-dienes is an important
atom-efficient transformation, which effectively adds nucle-
À
ophiles (NuH, for example, H2O, MeOH, NH3) over two C C
coupled dienes in a 1,6- or 3,6-fashion (see Scheme 1).[1] As
such, telomerization provides an economically attractive
route for the production of C8 bulk chemicals, such as 1-
octanol[2] and 1-octene.[3] Pd-catalyzed telomerization is also
increasingly explored as a potential route for the valorization
of biomass-derived feedstock.[1a]
Although the catalytic cycle (Scheme 1) is generally
regarded as well understood, only a limited number of studies
have been reported that involve the preparation of reactive
intermediates.[4–6] Clearly cumbersome preparative methods
and the poor stability of [Pd(1,2,3,8-h4-octadien-1,8-diyl)-
(PR3)] (B) are limiting further study. As a result the
preparation of reactive intermediates derived from commer-
cially relevant ligands (e.g. TPPTS; TPPTS = 3,3’,3’’-phosphi-
nidynetris(benzenesulfonic acid) trisodium salt) remains
elusive.[7] B mainly acts as a base on NuH during catalysis
and is readily converted into [Pd(1,2,3,7,8-h5-octa-2,7-dien-1-
yl)(PR3)]+ (C)[4–5] and its derivatives. Both from a synthetic
and catalytic point of view, C is the key intermediate as it lies
at the focal point of the catalytic cycle and leads via several
mechanistic pathways to species such as D, E, and F.
Scheme 1. Overall reaction and catalytic cycle of the Pd/PR3-catalyzed
Previously, we have reported on the catalytic activity of
telomerization of 1,3-butadiene.
Pd/TOMPP
(TOMPP = tris(2-methoxyphenyl)phosphine)
with biomass-derived substrates such as glycerol, glycols,
sugars, and sugar alcohols.[8–10] Mechanistic studies on this
catalyst system resulted in a general method for the direct
preparation of complexes of type C and could be successfully
applied to a series of phosphine ligands including PPh3,
TOMPP, and TPPTS. Herein, we report a simple and efficient
method for the preparation of cationic complexes of type C
(1b–5b), a detailed description of their solid- and solution-
state structures, as well as a comparison of their reactivity
towards the common methoxide nucleophile.
[*] P. J. C. Hausoul, Prof. Dr. R. J. M. Klein Gebbink
Organic Chemistry and Catalysis
Debye Institute for Nanomaterials Science, Utrecht University
Padualaan 8, 3584 CH, Utrecht (The Netherlands)
Fax: (+31)30-252-3615
E-mail: r.j.m.kleingebbink@uu.nl
Complexes 1b[5]–4b (Scheme 2) were prepared by a
simple one-pot procedure involving the dropwise addition
of [HPR3]BF4 salt (1a–4a; 1 equiv) to a mixture of [Pd(dba)2]
(1 equiv; dba = trans,trans-dibenzylideneacetone) and 2,7-
octadienol (2 equiv) in CH2Cl2 at room temperature. Isolation
by precipitation gave the air-stable solids 1b–4b as pure
(off)white powders in good to excellent yields (78–94%). For
the preparation of TPPTS complex 5b, a CH2Cl2/MeOH (1:1,
v/v) solvent mixture was used to improve solubility. Precip-
itation from MeOH/acetone gave 5b as the disodium salt in
85% yield.
P. J. C. Hausoul, Dr. A. N. Parvulescu, Dr. P. C. A. Bruijnincx,
Prof. Dr. B. M. Weckhuysen
Inorganic Chemistry and Catalysis
Debye Institute for Nanomaterials Science, Utrecht University
Sorbonnelaan 16, 3584 CA, Utrecht (The Netherlands)
Dr. M. Lutz, Prof. Dr. A. L. Spek[+]
Crystal and Structural Chemistry
Bijvoet Insitute for Biomolecular Research, Utrecht University
Padualaan 8, 3584 CH, Utrecht (The Netherlands)
[+] Correspondence pertaining to the crystallographic studies should
be addressed to a.l.spek@uu.nl.
[**] The authors kindly thank the ACTS-aspect program for financial
support.
Structural information on complexes 1b and 3b was
obtained by means of single-crystal X-ray crystallography
(Figure 1, Table 1; see the Supporting Information for 1b). To
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201003090.
7972
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2010, 49, 7972 –7975

Angewandte
Chemie
Interestingly, the TOMPP ligand displays an exo-2 configu-
ration (i.e. only two methoxy groups point outward from the
pyramidal phosphine structure).[11] This result shows that the
large steric demand of the octadienyl ligand, exemplified by a
C1-Pd-C8 angle close to 1808, results in a decrease of the PR3
cone angle from 2058 (exo-3) for the free ligand to 1768 (exo-
2) for the complex.[12] In both complexes the olefin moiety
(C7, C8) is oriented in-line with the coordination plane of
palladium (i.e. defined by Pd1, C1, C3, and P1). This
preference is expected for trigonal/square-planar ML3 com-
plexes[13] and has been observed previously for complexes of
the type [M(h3-allyl)(h2-olefin)(L)] (M = Pt).[14] A recently
calculated structure of 1b is in fairly good agreement with the
crystal structure data of 1b (and 3b).[15] However, allyl bonds
À
À
À
À
C1 C2 and C2 C3 and metal–ligand bonds Pd1 C3, Pd1 C7,
À
and Pd1 P1 are significantly shorter in 1b and 3b than in the
calculated structure. The crystal structure of 1b is isomor-
phous to the nickel analogue (H) reported earlier by Taube
et al.,[16] although the larger size of palladium is clearly
reflected in the metal–ligand bond lengths.
The solution structures of complexes 1b–5b were studied
by multinuclear (variable-temperature) NMR spectroscopy
and by high-resolution ESI-MS (see the Supporting Informa-
Scheme 2. Preparation of 1b–5b. Conditions: CH2Cl2, 1 h, room tem-
perature. [a] For 5b: CH2Cl2/MeOH (1:1, v/v), 3 equiv HBF4 and
1 equiv PR3 were added separately.
1
tion). Room temperature H NMR spectra of all complexes
show that none of the protons of the h5-2,7-octadien-1-yl
ligand are magnetically equivalent, indicating structural
rigidity. Variable-temperature (VT) NMR measurements of
complexes bearing phosphines with relatively small cone
angles (i.e. 1b, 2b, and 5b) support this conclusion, as the
spectra are largely temperature-independent. Importantly,
the results of 1H{31P} and 1H-1H COSY measurements
confirm that the structure of the cyclic metal fragment in
solution is comparable to the structure observed for 1b and
3b in the solid state. Notably, the NMR spectra of 3b and 4b
are temperature-dependent. This dependency is most pro-
1
nounced for complex 4b. Both H and 31P NMR data show
that 4b is in dynamic exchange between three distinct
rotamers. The dynamic behavior observed for 4b is fairly
comparable to that observed for other P(o-tolyl)3 complexes
as described by Buchwald and co-workers.[11] The NMR data
of TOMPP complex 3b shows a similar dynamic behavior;
however, as a result of the lower steric demand of the o-
methoxy substituents, rotational barriers are lowered consid-
erably. These results show that the h5-2,7-octadien-1-yl ligand
remains rigid in noncoordinating solvents and chelates even
under the influence of increased steric hindrance.
Figure 1. ORTEP diagram of the X-ray structure of 3b. Hydrogen
atoms and BF4À anion are omitted for clarity. Ellipsoids are drawn at
50% probability.
Table 1: Selected bond lengths [ꢀ] of 1b and 3b (crystal), 1b (calcu-
lated), and of nickel analogue H.
3b
M=Pd
R=2-MeOC6H4
1b (1)
M=Pd
R=Ph
1b[15]
M=Pd
R=Ph
H[16]
M=Ni
R=Ph
In telomerization, it has been shown that the addition of
bases such as NR3 and ROÀ is beneficial[5] and particularly in
case of Pd/TPPTS-catalyzed telomerization of 1,3-butadiene
even required for a successful reaction, suggesting that the
nucleophilic addition of NuÀ to C is the rate-determining step
in catalysis. To investigate the role of C in catalysis, the
stoichiometric reaction of complexes 1b–5b with NaOMe was
studied by UV/Vis spectroscopy (Figure 2 and Scheme 3).
GC–MS analysis identified 1-methoxy-2,7-octadiene (n-
telomer) and 3-methoxy-1,7-octadiene (iso-telomer) as main
reaction products in all cases. From the obtained data, initial
rates (D(De)/Dt) were extrapolated and are reported relative
to 1b (Table 2). Surprisingly, p-methoxy complex 2b exhibits
M1–C1
M1–C2
M1–C3
M1–C7
M1–C8
M1–P1
2.163(3)
2.185(3)
2.242(4)
2.300(3)
2.236(3)
2.3390(8)
2.142(7)
2.128(7)
2.256(7)
2.322(6)
2.257(5)
2.358(2)
2.167
2.219
2.335
2.415
2.307
2.428
2.08(1)
2.02(1)
2.16(2)
2.18(2)
2.04(1)
2.225(4)
the best of our knowledge, these are the first crystal structures
of intermediate C of the Pd-catalyzed telomerization of 1,3-
butadiene. The two structures show that the 2,7-octadien-1-yl
ligand is bonded in h5 fashion to Pd through an h3-allylic
fragment (C1, C2, C3) and a chelating h2-olefin tail (C7, C8).
Angew. Chem. Int. Ed. 2010, 49, 7972 –7975
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
7973

Communications
this fact, as a much higher turnover number (TON) can be
achieved with glycerol in a shorter time period (3b: 1520,
0.5 h, 2b: 18, 5 h) even though the reaction rate of 3b with
methoxide is six times slower than that of 2b. For TPPTS
complex 5b, however, the sluggish reaction rate is well in line
with previous observations[5] and demonstrates that because
of the proximity of the anionic sulfonate groups, the approach
of the methoxide anion to 5b is unfavorable. In addition, the
attractive electrostatic interaction of the sulfonated ligand
with the cationic metal causes complex 5b to be additionally
stabilized despite of its relatively low electron donation
(based on the value of nCO of the corresponding [Ni(CO)3-
(PR3)] complexes, see Table 2). The stability of 5b is
exemplified by the observation that it degrades much slower
to palladium black in solution than 1b, 2b, or 4b. The striking
difference in reactivity and stability between regioisomers 2b
and 3b can essentially be attributed to a similar influence now
exerted by the o-methoxy substituents in 3b. The proximity of
these electronegative substituents to the cationic metal center
Figure 2. Time-resolved UV/Vis spectroscopic changes De
(LmolÀ1 cmÀ1) (l=310–315 nm) for the nucleophilic addition of
~
~
*
NaOMe to 1b ( , l=309 nm), 2b ( , l=316 nm), 3b ( ,
À
(Pd1 O2: 3.211(6) ꢀ) considerably increases the stability of
*
^
l=314 nm), 4b ( , l=311 nm), and 5b ( , l=310 nm). Conditions:
3b by additional electrostatic attraction (i.e. 3b degrades
much more slowly to palladium black in solution than 1b, 2b,
and 4b). Conversely, the increased reaction rate of 4b with
respect to 1b also clearly shows that the strong (repulsive)
interaction of the o-methyl groups (q = 1948) with the
Pd(1,2,3,7,8-h5-2,7-octadien-1-yl) fragment destabilizes the
complex and enhances its reactivity. Nevertheless, the cata-
lytic results of Pd/TTP (TTP = tris(o-tolyl)phosphine) show
that increased reactivity of 4b is not necessarily productive, as
it does not lead to increased telomerization activity. For 3b a
similar steric effect (q = 1768) could be expected, but the
NMR data clearly show that the steric influence of the o-
methoxy groups is much less pronounced. In addition to this
steric, destabilizing effect the TOMPP ligand also stabilizes
complex C by both electron donation through phosphorus
(nCO = 2058.3 cmÀ1) and electrostatic interaction with the o-
methoxy substituents. The influence of these substituents in
the other steps of the cycle may therefore also be considered.
The same properties that destabilize saturated palladium(0)
species by promoting ligand exchange (steps D to A, and E to
A) at the same time stabilize unsaturated species by hemi-
labile coordination of the o-methoxy groups. Furthermore,
the oxidative coupling of A to B is promoted by strong
electron donation, whereas complex B is also stabilized by the
electrostatic interactions considered for complex C. It thus
appears that o-methoxy groups possess the most favorable
combination of steric and electronic properties for each
intermediate in the catalytic cycle.
MeOH, room temperature, 10 min.
Scheme 3. Stoichiometric nucleophilic addition of NaOMe to 1b–5b.
Conditions: MeOH, room temperature, overnight.
Table 2: Relative initial rates of the nucleophilic addition of NaOMe to
1b–5b, TON of Pd/PR3-catalyzed telomerization of glycerol with 1,3-
butadiene,[8a] cone angles q [8] of coordinated PR3 ligands, and nCO [cmÀ1
]
of [Ni(CO)3(PR3)] complexes.
Rel. rate TON Cone angle q [8] nCO
[cmÀ1
]
1b (R=Ph)
1.00
–
145[17]
2068.9[17]
2066.1[17]
2058.3[17]
2066.6[17]
2070[18]
2b (R=4-MeOC6H4) 1.07
3b (R=2-MeOC6H4) 0.16
18 145[17]
1520 176[12]
51 194[17]
4b (R=2-MeC6H4)
5b (R=3-SO3C6H4)
2.37
0.02
–
152–166[18]
a similar rate as 1b, whereas o-methoxy complex 3b is
approximately six times slower than 1b. In contrast, o-methyl
complex 4b is more than twice as fast as 1b, and TPPTS
complex 5b is the slowest of the series and reacts 50 times
slower than 1b. These significant rate differences clearly
illustrate that properties of the supporting phosphine strongly
influence the reactivity of complex C.
When the rates are compared to previously reported
catalytic data,[8a] it is clear that the addition of the nucleophile
to C is not always the rate-limiting step in catalysis. For
example, the results of Pd/TPMPP (2b; TPMPP = tris(4-
methoxyphenyl)phosphine) and Pd/TOMPP (3b) illustrate
In summary, a simple and efficient one-pot procedure for
the preparation of stable cationic complexes C (1b–5b),
which represent the key reactive intermediate in the Pd/PR3-
catalyzed telomerization of 1,3-butadiene, is presented. The
employed method not only avoids the use of 1,3-butadiene as
reactant, but more importantly enables detailed mechanistic
studies for a large variety of Pd/PR3-based catalyst systems.
Furthermore, the solid-state structures of 1b and 3b con-
stitute the first crystallographic evidence for intermediate C.
Solution-state NMR spectroscopic and reactivity studies show
that both the steric and electronic properties of the phosphine
7974 www.angewandte.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2010, 49, 7972 –7975

Angewandte
Chemie
Magerlein, S. Klein, J. Krause, M. Beller, Adv. Synth. Catal. 2001,
343, 29 – 33.
ligand impart a strong influence on the complex. The
combination of electron donation, electrostatic interaction,
and steric repulsion is particularly important for controlling
the reactivity of complex C. These insights have important
implications for rational catalyst design for telomerization
and other Pd-mediated reactions.
[6] a) A. Behr, G. Ilsemann, W. Keim, C. Krꢁger, Y. H. Tsay,
Organometallics 1986, 5, 514 – 518; b) U. Storzer, O. Walter, T.
Zevaco, E. Dinjus, Organometallics 2005, 24, 514 – 520.
[7] E. Kuntz, J.-M. Basset, D. Bouchu, G. Godard, F. Lefebvre, N.
Legagneux, C. Lucas, D. Michelet, Organometallics 2010, 29,
523 – 526.
Received: May 21, 2010
Published online: September 10, 2010
[8] a) R. Palkovits, I. Nieddu, C. A. Kruithof, R. J. M. Klein
Gebbink, B. M. Weckhuysen, Chem. Eur. J. 2008, 14, 8995 –
9005; b) R. Palkovits, I. Nieddu, R. J. M. Klein Gebbink, B. M.
Weckhuysen, ChemSusChem 2008, 1, 193 – 196.
[9] R. Palkovits, A. N. Parvulescu, P. J. C. Hausoul, C. A. Kruithof,
R. J. M. Klein Gebbink, B. M. Weckhuysen, Green Chem. 2009,
11, 1155 – 1160.
[10] P. J. C. Hausoul, P. C. A. Bruijnincx, R. J. M. Klein Gebbink,
B. M. Weckhuysen, ChemSusChem 2009, 2, 855 – 858.
[11] R. A. Widenhoefer, H. A. Zhong, S. L. Buchwald, Organo-
metallics 1996, 15, 2745 – 2754, and references therein.
[12] L. Hirsivaara, L. Guerricabeitia, M. Haukka, P. Suomalainen,
R. H. Laitinen, T. A. Pakkanen, J. Pursiainen, Inorg. Chim. Acta
2000, 307, 48 – 56.
[13] T. A. Albright, R. Hoffmann, J. C. Thibeault, D. L. Thorn, J. Am.
Chem. Soc. 1979, 101, 3801 – 3812.
[14] K. Miki, Y. Kai, N. Kasai, H. Kurosawa, J. Am. Chem. Soc. 1983,
105, 2482 – 2483.
[15] C. F. Huo, R. Jackstell, M. Beller, H. Jiao, J. Mol. Model. 2010,
16, 431 – 436.
Keywords: diene ligands · palladium · phosphanes ·
.
reactive intermediates · telomerization
[1] a) A. Behr, M. Becker, T. Beckmann, L. Johnen, J. Leschinski, S.
Reyer, Angew. Chem. 2009, 121, 3652 – 3669; Angew. Chem. Int.
Ed. 2009, 48, 3598 – 3614; b) N. D. Clement, L. Routaboul, A.
Grotevendt, R. Jackstell, M. Beller, Chem. Eur. J. 2008, 14,
7408 – 7420; c) M. J.-L. Tschan, E. J. Garcia-Suarez, Z. Freixa, H.
Launay, H. Hagen, J. Benet-Buchholz, P. W. N. M. van Leeuwen,
J. Am. Chem. Soc. 2010, 132, 6463 – 6473.
[2] Y. Tokitoh, T. Higashi, K. Hino, M. Murasawa, N. Yoshimura
(Kuraray), US 5118885, 1992.
[3] J. Briggs, J. Patton, S. Vermaire-Louw, P. Margl, H. Hagen, D.
Beigzadeh (Dow Global Technologies Inc.), WO 2010019360,
2010.
[4] a) R. Benn, P. W. Jolly, R. Mynott, B. Raspel, G. Schenker, K.-P.
Schick, G. Schroth, Organometallics 1985, 4, 1945 – 1953; b) P. W.
Jolly, R. Mynott, B. Raspel, K. P. Schick, Organometallics 1986,
5, 473 – 481.
[5] a) F. Vollmꢁller, J. Krause, S. Klein, W. Magerlein, M. Beller,
Eur. J. Inorg. Chem. 2000, 1825 – 1832; b) F. Vollmꢁller, W.
[16] R. Taube, J.P Gehrke, P. Bꢂhme, J. Kꢂttnitz, J. Organomet.
Chem. 1990, 395, 341 – 358; J. Sieler, R. Kempe, S. Wache, R.
Taube, J. Organomet. Chem. 1993, 455, 241 – 246.
[17] C. A. Tolman, Chem. Rev. 1977, 77, 313 – 3348.
[18] T. Bartik, B. Bartik, B. E. Hanson, I. Guo, I. Toth, Organo-
metallics 1993, 12, 164 – 170.
Angew. Chem. Int. Ed. 2010, 49, 7972 –7975
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
7975