Dihalogeno(diphosphane)metal(II) complexes (metal = Co, Ni, Pd) as pre-catalysts for the vinyl/addition polymerization of norbornene - Elucidation of the activation process with B(C6F5)3/AlEt3 or Ag[closo-1-CB11H12] and evidence for the in situ formation of naked Pd2+ as a highly active species

Article abstract of DOI:10.1039/b302937a

Dihalogenometal(II) complexes with bidentate phosphane ligands of the general type [M{Ph₂P(CH₂)_nPPh₂} X₂] with n = 2 to 5, X = Cl or Br and M = Co, Ni or Pd have been utilized as catalysts for the vinyl/addition polymerization of norbornene. These complexes can be activated with the Lewis-acids methylalumoxane (MAO) or tris(pentafluorophenyl)borane, B(C₆F₅)₃ in combination with triethylaluminium (AlEt₃). The nickel(II) and palladium(II) complexes show very high polymerization activities up to 10⁷ g_polymer mol_metal^-1 h^-1. Yet, the complexes Pd(dppe)Cl₂ (5, 1.9 × 10⁷ g_polymer mol_Pd^-1 h^-1) and Pd(dppp)Cl₂ (6, 3.0 × 10³ g_polymer mol_Pd^-1 h^-1) demonstrated that small changes in the ligand structure could have great effects on the polymerization activity [dppe = 1,2-bis(diphenylphosphino)ethane, Ph₂P(CH₂)₂PPh₂; dppp = 1,3-bis(diphenylphosphino)propane, Ph₂P(CH₂)₃PPh₂]. The activation process of the pre-catalysts 5 and 6 in combination with B(C₆ F₅)₃/AlEt₃ was followed by multinuclear (¹H, ¹⁹F, and ³¹P) NMR investigations and by reaction with B(C₆ F₅)₃ and Ag[closo-1-CB₁₁H₁₂]. The reaction of B(C₆F₅)₃ and AlEt₃ leads to an aryl/alkyl group exchange resulting in the formation of AlEt_3-n(C₆F₅)_n and B(C₆F₅)_3-nEt_n with Al(C₆F₅)₃ and BEt₃ as main products for an about equimolar ratio. AlEt_3-n(C₆F₅)_n will then react with the pre-catalysts and abstract the chloride atoms to form [M{Ph₂ P(CH₂)_nPPh₂}]²⁺ as the active species for the polymerization. The higher polymerization activity of 5/B(C₆F₅)₃/AlEt₃ compared to 6/B(C₆F₅)₃/AlEt₃ can be explained by a ligand redistribution reaction of unstable [Pd^II (dppe)]²⁺ to give inactive and isolable [Pd^II(dppe)₂]²⁺ and highly active, naked Pd²⁺ cations together with the lower coordinating ability of the anionic adduct [Cl-Al(C₆F₅)₃]^- in comparison to [Cl-B(C₆F₅)₃]^-. The Lewis-acid Al(C₆F₅)₃ is much more activating than B(C₆F₅)₃. The [Pd(dppe)₂]²⁺ cation from the ligand redistribution was isolated in the (X-ray) structurally elucidated compounds [Pd^II(dppe)₂]-[ClB(C₆ F₅)₃]₂·4CH₂Cl₂ and [Pd^II(dppe)₂][CB₁₁ H₁₁Cl]₂·3CH₂Cl₂. The stable [Pd(dppp)]²⁺ cation from 6 could be crystallized as [Pd^II(dppp)(CB₁₁H₁₂)] [CB₁₁H₁₂] (CB₁₁H₁₂ = mono-anionic carborane [closo-1-CB₁₁H₁₂]^-).

Full text of DOI:10.1039/b302937a

View Article Online / Journal Homepage / Table of Contents for this issue
Dihalogeno(diphosphane)metal(II) complexes (metal ؍ Co, Ni, Pd)
as pre-catalysts for the vinyl/addition polymerization of
norbornene – elucidation of the activation process with
B(C6F5)3/AlEt3 or Ag[closo-1-CB11H12] and evidence for the
in situ formation of “naked” Pd2؉ as a highly active species†
Paul-Gerhard Lassahn,a Vasile Lozan,a Biao Wu,a Andrew S. Wellerb and Christoph Janiak*a
a Institut für Anorganische und Analytische Chemie, Universität Freiburg, Albertstr. 21,
79104 Freiburg, Germany. E-mail: janiak@uni-freiburg.de
b Department of Chemistry, University of Bath, Bath, UK BA2 7AY.
E-mail: a.s.weller@bath.ac.uk
Received 14th March 2003, Accepted 10th September 2003
First published as an Advance Article on the web 25th September 2003
Dihalogenometal() complexes with bidentate phosphane ligands of the general type [M{Ph2P(CH2)nPPh2}X2] with
n = 2 to 5, X = Cl or Br and M = Co, Ni or Pd have been utilized as catalysts for the vinyl/addition polymerization of
norbornene. These complexes can be activated with the Lewis-acids methylalumoxane (MAO) or tris(pentafluoro-
phenyl)borane, B(C6F5)3 in combination with triethylaluminium (AlEt3). The nickel() and palladium() complexes
show very high polymerization activities up to 107 gpolymer molmetalϪ1 hϪ1. Yet, the complexes Pd(dppe)Cl2 (5, 1.9 × 107
gpolymer molPdϪ1 hϪ1) and Pd(dppp)Cl2 (6, 3.0 × 103 gpolymer molPdϪ1 hϪ1) demonstrated that small changes in the
ligand structure could have great effects on the polymerization activity [dppe = 1,2-bis(diphenylphosphino)ethane,
Ph2P(CH2)2PPh2; dppp = 1,3-bis(diphenylphosphino)propane, Ph2P(CH2)3PPh2]. The activation process of the
pre-catalysts 5 and 6 in combination with B(C6F5)3/AlEt3 was followed by multinuclear (1H, 19F, and 31P) NMR
investigations and by reaction with B(C6F5)3 and Ag[closo-1-CB11H12]. The reaction of B(C6F5)3 and AlEt3 leads
to an aryl/alkyl group exchange resulting in the formation of AlEt3 Ϫ n(C6F5)n and B(C6F5)3 Ϫ nEtn with Al(C6F5)3
and BEt3 as main products for an about equimolar ratio. AlEt3 Ϫ n(C6F5)n will then react with the pre-catalysts
and abstract the chloride atoms to form [M{Ph2P(CH2)nPPh2}]2ϩ as the active species for the polymerization.
The higher polymerization activity of 5/B(C6F5)3/AlEt3 compared to 6/B(C6F5)3/AlEt3 can be explained by a
ligand redistribution reaction of unstable [PdII(dppe)]2ϩ to give inactive and isolable [PdII(dppe)2]2ϩ and highly
active, “naked” Pd2ϩ cations together with the lower coordinating ability of the anionic adduct [Cl–Al(C6F5)3]Ϫ
in comparison to [Cl–B(C6F5)3]Ϫ. The Lewis-acid Al(C6F5)3 is much more activating than B(C6F5)3. The [Pd(dppe)2]2ϩ
cation from the ligand redistribution was isolated in the (X-ray) structurally elucidated compounds [PdII(dppe)2]-
[ClB(C6F5)3]2ؒ4CH2Cl2 and [PdII(dppe)2][CB11H11Cl]2ؒ3CH2Cl2. The stable [Pd(dppp)]2ϩ cation from 6 could be
crystallized as [PdII(dppp)(CB11H12)][CB11H12] (CB11H12 = mono-anionic carborane [closo-1-CB11H12]Ϫ).
high oxidation states.1–3 Only a few reports describe the form-
Introduction
ation of low molar mass oligomeric material with 2,7-connect-
The homopolymerization of norbornene can be accomplished
ivity of the monomer from the cationic or radical homo-
by three different routes with each route leading to its own
polymer type with different structure and properties (Scheme 1).
polymerization of norbornene.4 Finally, the vinyl or addition
polymerization of norbornene yields saturated 2,3-inserted
rotationally constrained polymers in which the bicyclic struc-
tural unit remains intact and only the double bond of the
π-component is opened. Catalysts containing the metals
titanium,5,6 zirconium,7–11 chromium,12 and currently the late
transition-metals cobalt,13–15 nickel16–31 and palladium29,32–48 are
described in the literature for the vinyl/addition polymerization
of norbornene.49–56 The late transition-metal complexes are
commonly activated with methylalumoxane (MAO),17–20,34,49
except for the cationic palladium-complexes [Pd(NCR)4]2ϩ2AϪ
(NCR weakly bound nitrile-ligand; A = “non”-coordinating
counter ion).35–45 Another co-catalytic system for the activation
of metal complexes is the organo-Lewis acid tris(pentafluoro-
phenyl)borane, B(C6F5)3 with or without triethylaluminium
(AlEt3). This co-catalytic system is known from the activation
process of early transition-metal Group 4 metallocene catalysts
in olefin polymerization57–65 and was recently also applied to
the activation of late transition-metal complexes for the
(co)polymerization of cyclopentene,66 ethene,67 norbornene and
norbornene derivatives.16,32,31,50–56
Previously, we have communicated initial polymerization
results of nickel() and palladium() dihalogeno complexes
bearing bidentate phosphane ligands as pre-catalysts for the
Scheme 1
Polymers produced via ring-opening-metathesis-polymeriz-
ation (ROMP) still contain double bonds in the polymer back-
bone and can be obtained for a number of transition metals in
† Electronic supplementary information (ESI) available: Full 1H, 19F
and 31P NMR data of compounds 5 and 6 in combination with different
amounts of B(C6F5)3/AlEt3. See http://www.rsc.org/suppdata/dt/b3/
b302937a/
T h i s j o u r n a l i s © T h e R o y a l S o c i e t y o f C h e m i s t r y 2 0 0 3
D a l t o n T r a n s . , 2 0 0 3 , 4 4 3 7 – 4 4 5 0
4437

View Article Online
vinyl polymerization of norbonene.32 These complexes can be
activated with MAO as well as with B(C6F5)3 in combination
SADABS.70 The structures were solved by direct methods
(SHELXS-97)71; refinement was done by full-matrix least
squares on F 2 using the SHELXL-97 program suite.71 All non-
hydrogen positions were found and refined with anisotropic
temperature factors. The hydrogen atoms were placed at calcu-
lated positions, using appropriate riding models (HFIX 43 for
aromatic CH, HFIX 33 for CH3, HFIX 23 for CH2, HFIX 153
for CH and BH in the carborane, HFIX 83 for OH) and iso-
tropic temperature factors of U(H) = 1.2 Ueq(BH, CH and CH2)
or U(H) = 1.5 Ueq(CH3, O). One of the chlorine atoms in each
of the two independent CH2Cl2 solvent molecules of crystalliz-
ation in the structure of [Pd(dppe)2][ClB(C6F5)3]2ؒ4CH2Cl2 (14)
was found to be disordered. One of the CH2Cl2 solvent mole-
cules in [Pd(dppe)2][CB11H11Cl]2ؒ3CH2Cl2 (15) is disordered
around a crystallographic glide plane. In the structure of
[Pd(dppp)(CB11H12)][CB11H12] (16) about 22% of the carborane
counter anions were found to contain a hydroxy group, and
hence to be of formula [CB11H11(OH)]Ϫ. The OH-group may
have been introduced by the solvent diethyl ether which con-
tained traces of water.72 Thus, the crystal structure would have
with AlEt3 and showed remarkable polymerization activities
Ϫ1
up to 2.0 × 107 gpolymer molmetal
h
Ϫ1. In this paper, we wish to
present the polymerization results of the complete series of
cobalt(), nickel() and palladium() complexes of the general
type [M{Ph2P(CH2)nPPh2}Cl2] with n = 2 to 5 and M = Co, Ni,
Pd (Scheme 2). In the case of the palladium() pre-catalysts 5
and 6 we were able to obtain information about the activation
process with B(C6F5)3 and AlEt3 by using multinuclear (1H, 19F,
and 31P) NMR investigations and crystallization of reaction
products with B(C6F5) or the closo-carborane anion [CB11H12]Ϫ.
to be correctly formulated as [Pd(dppp)(CB11H12)][CB11H11.78
-
(OH)0.22]. The carbon atom in CB11H11Cl in 15 and in the bound
and free cage of CB11H12 [or CB11H12(O0.22)] in 16 were located
as their B–C bond lengths are slightly shorter than those for
B–B. For 15 B–C = 1.701(6)–1.728(8) Å, B–B = 1.733(6)–
1.778(6) Å; for 16, bound cage B–C = 1.693(7)–1.709(6), B–B =
1.748(6)–1.800(6) Å, free cage B–C = 1.687(8)–1.718(7), B–B =
1.726(9)–1.774(8) Å. The location may be different in CB11H11-
(OH), since the carbon atom was not found opposite to the
hydroxide-bound boron (B10) atom. In the anion CB11H11Cl in
15 the carbon atom is opposite to the chlorine bound boron(12)
atom. The B12 vertex is known to be most reactive for electro-
philic substitution and thus would be expected to be substituted
first.73 Also, the carbon atom was found opposite to one of
the BH groups bonded to palladium in the metal-bound cage in
16. This is as expected as metal fragments generally bind
through B(H)12 and B(H)7 when the cage acts in a bidentate
fashion.74,75 Details of the X-ray structure determinations and
refinements are provided in Table 1. Graphics were obtained
with ORTEP 3 for Windows.76 Displacement ellipsoids are
drawn at the 50% probability level and H atoms are shown as
spheres of arbitrary radii.
Scheme 2
The homopolymer vinyl-poly(norbornene) is of interest as a
specialty polymer with good mechanical strength, heat resist-
ance, and optical transparency, e.g. for deep ultraviolet photo-
resists, interlevel dielectrics in microelectronics applications or
as a cover layer for liquid-crystal displays.49,68
Experimental
General procedures
All work involving air- and/or moisture-sensitive compounds
was carried out using standard vacuum, Schlenk or drybox
techniques. FT-IR spectra (KBr pellets) were measured on a
Bruker Optik IFS 25. NMR spectra were recorded with a
Bruker Avance DPX 200 at 300 K and 200 MHz (1H NMR), 81
MHz (31P NMR, calibration against an external standard of
85% H3PO4) and 188 MHz (19F NMR, calibration against an
external standard of CDFCl2). 1H NMR spectra were cali-
brated against the solvent signal (CD2Cl2 5.32 ppm). The NMR
spectra for compound [Pd(dppp)(CB11H12)][CB11H12] (16) were
recorded with a Bruker Avance 500 at 300 K and 500 MHz
(1H NMR) and 202 MHz (31P NMR). The NMR experiments
with air-sensitive materials were performed under an inert-
gas atmosphere using NMR tubes with screw caps (Wilmad),
Teflon-covered septa (Wheaton) and CD2Cl2 as solvent.
CCDC reference numbers 206046–206048.
See http://www.rsc.org/suppdata/dt/b3/b302937a/ for crystal-
lographic data in CIF or other electronic format.
Materials
Cobalt() chloride (Aldrich), nickel() chloride (Aldrich),
nickel() chloride hexahydrate (Acros), nickel() bromide
(Aldrich), palladium() chloride (Merck-Schuchardt), 1,2-bis-
(diphenylphosphino)ethane (dppe, Aldrich), 1,3-bis(diphenyl-
phosphino)propane (dppp, Strem Chemicals), potassium
diphenylphosphide ((C6H5)2PK, 0.5 M solution in tetrahydro-
furan), 1,4-dichlorobutane (Aldrich), 1,5-dichloropentane
(Aldrich), methylalumoxane (10 wt% solution in toluene,
Witco), tris(pentafluorophenyl)borane (B(C6F5)3, Aldrich), and
triethylaluminium (AlEt3, 1 mol LϪ1 solution in hexane, Merck-
Schuchardt) were used as received. Toluene, tetrahydrofuran
and n-hexane were dried over sodium metal, distilled and stored
under nitrogen. Methylene chloride was dried over CaH2.
Norbornene (bicyclo[2.2.1]hept-2-ene, Aldrich) was purified by
distillation and used as a solution in toluene. 1,4-Bis(diphenyl-
Elemental analysis were obtained on
a VarioEL from
Elementaranalysensysteme GmbH. Gel permeation chromato-
graphy (GPC) analyses were performed on a PL-GPC 220
(columns PL gel 10 µm MIXED-B) with polymer solutions in
1,2,4-trichlorobenzene (concentration of 2–3 mg mLϪ1). The
GPC was measured at 140 ЊC with an injection volume of
200 µL and with a rate of 1 mL minϪ1
.
phosphino)butane
(dppb),77
1,5-bis(diphenylphosphino)-
X-Ray crystallography
pentane77 (dpppt) and Ag[CB11H12]78 were synthesized by the
published literature procedures.
X-Ray data were collected with a Bruker AXS with CCD area-
detector and Mo-Kα radiation (λ = 0.71073 Å), graphite mono-
chromator, double-pass method φ–ω-scan. Data collection and
cell refinement with SMART,69 data reduction with SAINT.69
An experimental absorption correction was performed with
Preparation of the pre-catalysts
The nickel() complexes Ni(dppe)Cl2 (1) and Ni(dppp)Cl2 (2)
were synthesized by addition of the ligand in CH2Cl2 to a solu-
D a l t o n T r a n s . , 2 0 0 3 , 4 4 3 7 – 4 4 5 0
4438

View Article Online
Table
1
Crystal data for the palladium complexes [PdII(dppe)2][ClB(C6F5)3]2ؒ4CH2Cl2 (14), [PdII(dppe)2][CB11H11Cl]2ؒ3CH2Cl2 (15) and
[PdII(dppp)(CB11H12)][CB11H12] (16)
Compound
14
15
16
Formula
M
C92H56B2Cl10F30P4Pd
2337.77
C57H75B22Cl8P4Pd
1511.89
C29H50B22O0.22P2Pd
808.37
Crystal size/mm
Crystal description
T/K
0.23 × 0.09 × 0.08
Isometric, colorless
213(2)
0.18 × 0.10 × 0.10
Isometric, brown
213(2)
0.16 × 0.12 × 0.08
Isometric, yellow
213(2)
Crystal system
Space group
Triclinic
P1
Monoclinic
C2/m
Triclinic
P1
¯
¯
a/Å
b/Å
c/Å
α/Њ
10.909(3)
14.192(3)
16.104(4)
103.811(4)
104.192(4)
91.833(4)
2336.2(9)
1
19.750(2)
16.029(1)
13.770(1)
90.00
119.159(1)
90.00
3806.8(5)
2
1.319
10.9600(8)
11.4034(9)
18.4232(14)
102.192(1)
98.503(1)
107.025(1)
2097.1(3)
2
β/Њ
γ/Њ
V/Å3
Z
D/g cmϪ3
1.662
1.280
F(000)
1164
0.661
1538
0.645
824
0.543
µ/mmϪ1
Measured reflections
Unique reflections (Rint
18052
8734 (0.0365)
5827
0.954/Ϫ0.847
0.0454/0.0995
0.0769/0.1135
0.979
11893
4726 (0.0182)
3808
0.483/Ϫ0.469
0.0418/0.1181
0.0549/0.1274
1.042
19180
9809 (0.0416)
5954
1.152/Ϫ1.241
0.0522/0.0933
0.0971/0.1090
1.038
)
Obs. reflections [I>2σ(I )]
Max/min ∆ρa/e ÅϪ3
R1/wR2 [I>2σ(I )]b
R1/wR2 (all data)b
Goodness-of-fit on F 2 c
2
a Largest difference peak and hole. b R1 = [Σ(||Fo| Ϫ |Fc||)/Σ|Fo|]; wR2 = [Σ[w(Fo2 Ϫ Fc2)2]/Σ[w(Fo )2]]1/2
.
c Goodness-of-fit = [Σ[w(Fo2 Ϫ Fc2)2]/(n Ϫ p)]1/2
.
tion of NiCl2ؒ6H2O in ethanol according to a procedure given
by Busby et al.79 The palladium() complexes Pd(dppe)Cl2 (5),
Pd(dppp)Cl2 (6), Pd(dppb)Cl2 (7) and Pd(dpppt)Cl2 (8) were
synthesized by combining a solution of PdCl2 in hot concen-
trated HCl and ethanol or 1-butanol with a solution of the
appropiate ligand in CH2Cl2 or 1-butanol.32 The nickel() com-
plexes Ni(dppb)Cl2 (3), Ni(dpppt)Br2 (4), Ni(dppe)Br2 (13) and
the cobalt() complexes Co(dppe)Cl2 (9), Co(dppp)Cl2 (10),
Co(dppb)Cl2 (11) and Co(dpppt)Cl2 (12) were synthesized
according to a modified procedure given by Sacconi and
Gelsomini.77 The yields were in the range of 70 to 97% and the
purities were checked by CHN-analysis.
aluminium foil to protect the silver salt from light. Pd(dppp)Cl2
(0.089 g, 0.15 mmol) was dissolved in freshly dried CH2Cl2
(40 mL) and added dropwise to the silver salt. After stirring for
4 days the resulting suspension was filtered over Celite. The
Celite was washed with 30 mL of CH2Cl2 and the solvent was
removed in vacuo to a volume of 5 mL. Addition of dried
n-hexane (30 mL) resulted in the formation of a yellow precipi-
tate which was filtered off and dried in vacuo (yield 0.080 g,
66%, mp >200 ЊC). C29H50B22P2Pd (804.93): calcd. C 43.27, H
1
6.26; found C 44.09, H 6.49%. H NMR (500 MHz, CD2Cl2):
δ = 1.0–2.5 (br, 22H, CHB11H11), 2.54 (s, br, 2H, CHB11H11),
2.75 (t, br, 4H, P–CH2–CH2–CH2–P), 2.83 (s, br, 2H, P–CH2–
CH2–CH2–P), 7.30–7.70 (m, 20H, aromatic H-atoms); 31P
NMR (202 MHz, CD2Cl2): δ = 17.9. Crystals suitable for X-ray
diffraction were grown within several days after redissolving the
solid product in a minimum of methylene chloride and overlay-
ering this solution with diethyl ether at room temperature. The
crystals were found to have the chemical composition of the
title compound except for about 22% of the carborane counter
anions that were found to contain a hydroxy group, and hence
to be of formula [CB11H11(OH)]Ϫ.
Synthesis of [PdII(dppe)2][ClB(C6F5)3]2ؒ4CH2Cl2 (14). Pd-
(dppe)Cl2 (0.115 g, 0.2 mmol) and B(C6F5)3 (0.307 g, 0.6 mmol)
were dissolved under an inert-gas atmosphere in a mixture of
freshly dried toluene (5 mL) and methylene chloride (10 mL).
Some colorless, air-sensitive crystals which were suitable for
study by X-ray diffraction were obtained after 4 months at Ϫ18
ЊC and analyzed as the title compound. The small amount of
crystalline material available (∼10 mg) precluded any further
analyses.
Synthesis of [PdII(dppe)2][CB11H11Cl]2ؒ3CH2Cl2 (15). Ag-
[CB11H12] (0.075 g, 0.30 mmol) was placed under an inert-gas
atmosphere into a Schlenk-flask which was covered with alu-
minium foil to protect the silver salt from light. Pd(dppe)Cl2
(0.086 g, 0.15 mmol) was suspended in freshly dried CH2Cl2
(40 mL) and added dropwise to the silver salt. After stirring for
4 days the resulting suspension was filtered over Celite. The
Celite was washed with 30 mL of CH2Cl2 and the solvent was
removed in vacuo to a volume of 5 mL. Some crystals suitable
for X-ray diffraction were grown within several days after over-
laying the CH2Cl2 solution with diethyl ether at room temper-
ature. Single-crystal X-ray analysis yielded a compound with
the formula of the title compound. The small amount of
crystalline material available (∼10 mg) precluded any further
analyses.
Polymerization procedures
General. The pre-catalysts were applied as solutions (for pre-
catalysts 1, 2, 6, 8, 9, 10 and 13) or as a fine suspensions via
ultrasonication in methylene chloride (for precatalysts 3 to 5, 7,
11 and 12). Polymerizations were conducted at room temper-
ature in a water bath to ensure a constant temperature during
the reaction. Polymerization runs were carried out at least three
times to ensure reproducibility. The IR spectra of the poly-
(norbornene)s obtained with the catalysts 1 to 12 showed the
absence of a double bond at 1620 to 1680 cmϪ1. This proved
that vinyl/addition polymerization instead of a ring-opening
metathesis polymerization (ROMP) had occurred. The conver-
sion was calculated by gravimetric analysis of the polymer.
General procedure for the homopolymerization of norbornene
with MAO as co-catalyst. A Schlenk-flask was charged with the
norbornene solution and the MAO-solution was added. After
1 min the solution or suspension of the pre-catalyst was added
Synthesis of [PdII(dppp)(CB11H12)][CB11H12] (16). Ag-
[CB11H12] (0.075 g, 0.30 mmol) was placed under an inert-
gas atmosphere into a Schlenk-flask which was covered with
D a l t o n T r a n s . , 2 0 0 3 , 4 4 3 7 – 4 4 5 0
4439

View Article Online
Table 2 Polymerization of norbornene (NB) with complexes 1 to 12 in combination with methylalumoxane (MAO)
Catalysta
Time/min
Conversion (%)
Activity/gpolymer molmetalϪ1 hϪ1
Mn/g molϪ1
Mw/g molϪ1
Q = Mw/Mn
1/MAO
2/MAO
3/MAO
4/MAO
5/MAO
6/MAO
7/MAO
8/MAO
9/MAO
10/MAO
11/MAO
12/MAO
1–12/–b
5
5
5
5
5
5
60
5
60
60
60
60
60
2.0
3.7
18.4
36.0
3.3
0.1
1.8
8.0
0.4
1.7
0.3
0.6
0
2.1 × 104
1.6 × 106
1.3 × 106
7.1 × 105
1.8 × 105
2.7 × 106
2.6 × 106
1.7 × 106
7.9 × 105
Not soluble
Not soluble
Not soluble
Not soluble
1.4 × 106
1.4 × 106
6.9 × 105
4.4 × 105
–
1.7
1.9
2.4
4.4
4.2 × 104
2.1 × 105
4.1 × 105
3.7 × 104
1.1 × 103
1.7 × 103
9.1 × 104
4.0 × 102
7.2 × 105
5.7 × 105
2.8 × 105
1.5 × 105
–
1.9
2.4
2.5
2.9
–
1.6 × 103
3.0 × 102
5.8 × 102
No activity observed
Conditions: room temperature; toluene/methylene chloride solution.a 10.6 mmol NB, [M] : [NB] = 1 : 1000, [M] : [Al] = 1 : 100, total volume 10.0 mL.
b 10.6 mmol NB, [M] : [NB] = 1 : 1000, total volume 10.0 mL.
Ϫ1
via syringe and the mixture was stirred with a magnetic stirrer.
The polymerization was stopped by the addition of 30 mL of a
10 : 1 methanol/concentrated HCl mixture. The precipitated
polymer was filtered off, washed with methanol and dried in
vacuo for 5 h.
of 3.0 × 102 (11/MAO) to 4.1 × 105 gpolymer molmetal hϪ1
(4/MAO). Only the nickel complexes 3 and 4 and perhaps the
palladium complex 8 exhibit activities around or above 105
gpolymer molmetalϪ1 hϪ1. These complexes contain the six- or seven-
membered dppb and dpppt ligands (Fig. 3). For the other com-
plexes the conversion and activity with MAO was unremark-
able. In accordance with the literature data, the cobalt() com-
plexes 9–12 delivered only traces of poly(norbornene) within a
polymerization time of 1 h.15 The polymers obtained with the
nickel- and cobalt-containing catalysts were soluble in 1,2,4-
trichlorobenzene and could be investigated by gel permeation
chromatography (GPC) whereas the palladium()-catalyzed
poly(norbornene)s were insoluble in common solvents as
expected from the literature.16,29,31–37,49,50 The soluble polymer
samples displayed a monomodal and narrow molar mass distri-
bution with Q = Mw/Mn close to a value of 2 (except for the
highly active 4/MAO with Q = 4.4). The number-average molar
General procedure for the homopolymerization of norbornene
with B(C6F5)3/AlEt3 as co-catalysts. A Schlenk-flask was
charged with the norbornene solution. The solution or suspen-
sion of the pre-catalyst followed by the separate co-catalyst
components [B(C6F5)3 and AlEt3] were quickly added via
syringe and the mixture was stirred with a magnetic stirrer. The
polymerization was stopped by the addition of 40 mL of a
10 : 1 methanol/concentrated HCl mixture. The precipitated
polymer was filtered off, washed with methanol and dried in
vacuo for 5 h.
From studies of a number of metal complexes which did not
show any polymerization activity with B(C6F5)3/AlEt3, we con-
clude that the combination of B(C6F5)3/AlEt3 alone is not
polymerization active.
mass (Mn) was found between 1.5 × 105 and 1.6 × 106 g molϪ1
.
Thus, the average chain length for the polymers lies between
1,600 (12) and 17,000 (1) monomer units (Mnorbornene = 94.16 g
molϪ1). A value of Q = 2 is the theoretical dispersity for a
Schulz–Flory type distribution arising from an ideally behaved
Results and discussion
polymerization reaction with a chain-termination reaction.80
A
dispersity of Q ≈ 2 indicates a single-site character, i.e. a highly
homogeneous structure for the active catalyst species.
Norbornene polymerization
Nickel(), palladium() and cobalt() complexes of the general
type [M{Ph2P(CH2)nPPh2}X2] with n = 2 to 5, X = Cl or Br and
M = Ni (1 to 4), Pd (5 to 8) and Co (9 to 12) were tested in
the homopolymerization of norbornene. The results of the
polymerization activities using methylalumoxane (MAO) as
co-catalyst are summarized in Table 2 and Fig. 1.
In order to gain additional information about the activation
process the activator was changed from MAO to the better
defined co-catalyst B(C6F5)3 with or without triethylaluminium
(AlEt3). Mechanistic investigations with MAO as a co-catalyst
are difficult since the exact composition and structure of MAO
is still not entirely clear.61,65,81–84 The results of the polymeriz-
ation of norbornene with the pre-catalysts 1 to 8 and 10 in
combination with B(C6F5)3/AlEt3 or with B(C6F5)3 alone are
summarized in Table 3 and Fig. 2.
The polymerization activities of the metal–phosphane
complexes 1 to 12 in combination with MAO covered a range
Dichloro-nickel() and -palladium() complexes bearing
bidentate phosphane ligands can be activated with the co-cata-
lytic system B(C6F5)3/AlEt3. Activation of the cobalt complex
10 under similar conditions essentially failed. The activity
trends for the nickel complexes 1–4 and the palladium com-
plexes 6–8 are the same as seen above for activation with MAO.
Again the nickel complexes 3 and 4 and the palladium
complexes 7 and 8 with the longest-chain dppb and dpppt lig-
Ϫ1
ands show activities around or above 105 gpolymer molmetal hϪ1
(Fig. 2). Even the overall activities for these complexes with
MAO or with B(C6F5)3/AlEt3 are quite similar (Fig. 3) although
a direct comparison should not be made in view of the different
metal-to-co-catalyst ratios. For similar or even higher monomer
conversions and polymerization activities the required co-
catalyst quantities were much lower with B(C6F5)3/AlEt3 than
with MAO (molar ratios of Pd : borane : AlEt3 = 1 : 9 : 10 and
Pd : MAO = 1 : 100). This leads to fewer co-catalyst residues in
Fig.
1
Activities of 1–12 with MAO as co-catalyst in the
polymerization of norbornene. Detailed conditions are given in Table 2.
D a l t o n T r a n s . , 2 0 0 3 , 4 4 3 7 – 4 4 5 0
4440

View Article Online
Table 3 Polymerization of norbornene (NB) with complexes 1 to 8 and 10 in combination with B(C6F5)3 and triethylaluminium (AlEt3)
Catalyst
Time/min
Conversion (%)
Activity/gpolymer molmetalϪ1 hϪ1
Mn/g molϪ1
Mw/g molϪ1
Q = Mw/Mn
3.8 × 103
4.1 × 105
3.4 × 105
3.9 × 105
1.2 × 105
1.0 × 106
8.1 × 105
8.4 × 105
4.2 × 105
Not soluble
Not soluble
Not soluble
Not soluble
1.2 × 106
–
2.5
2.4
2.2
3.4
a
1/B(C6F5)3/AlEt3
2/B(C6F5)3/AlEt3
3/B(C6F5)3/AlEt3
4/B(C6F5)3/AlEt3
5/B(C6F5)3/AlEt3
6/B(C6F5)3/AlEt3
7/B(C6F5)3/AlEt3
8/B(C6F5)3/AlEt3
10/B(C6F5)3/AlEt3
60
60
5
4.1
9.0
9.0
14.8
54.3
3.1
8.1
17.4
0.6
0
a
8.6 × 103
1.0 × 105
a
a
a
a
a
a
a
5
1.7 × 105
1/6
60
5
1.9 × 107
3.0 × 103
9.2 × 104
1
9.9 × 105
60
60
60
60
60
5.7 × 102
6.5 × 105
1.8
–
1.9
1.6
–
b
1–8/AlEt3
No activity observed
–
3/B(C6F5)3 c
15.0
11.1
0
1.4 × 104
1.1 × 106
9.5 × 105
–
2.1 × 106
1.5 × 106
–
c
4/B(C6F5)3
1.0 × 104
c
1,2,5–8/B(C6F5)3
No activity observed
Conditions: room temperature; toluene/methylene chloride solution.a 30.1 mmol NB, [M] : [NB] = 1 : 1000, [M] : [B] : [Al] = 1 : 9 : 10, total volume
40.0 mL. b 30.1 mmol NB, [M] : [NB] = 1 : 1000, [M] : [Al] = 1 : 10, total volume 40.0 mL. c 30.1 mmol NB, [M] : [NB] = 1 : 1000, [M] : [B] = 1 : 9, total
volume 40.0 mL.
six-membered chelate ring and then increased again. Similar
activity trends for the nickel- or palladium-series with chelate
ring size may be the result of a similar polymerization mechan-
ism or activation process when using MAO or B(C6F5)3/AlEt3
as a co-catalyst. Still, the activity trends are difficult to explain
in the absence of a more profound knowledge of the active
species. Generally, the active species in MAO- or B(C6F5)3/
AlEt3-activated Ni- or Pd-complexes are not well known.49
Some investigations to be discussed below were aimed at
increasing the understanding of the activation process of these
pre-catalysts to supplement our recent mechanistic study on
the activation of palladium() salts containing [PdCl4]2Ϫ and
[Pd2Cl6]2Ϫ ions with evidence for the in situ formation of
PdCl2.32
An exception upon B(C6F5)3/AlEt3-activation is the
palladium–dppe complex 5. The reproducible and extremely
high activity of 5 of the order of 107 gpolymer molPdϪ1 hϪ1 is note-
Fig. 2 Activities of 1–8 with B(C6F5)3/AlEt3 or B(C6F5)3 alone as co-
worthy (Fig. 2 and 3). A conversion of 50% was reached after
catalyst in the polymerization of norbornene. Detailed conditions are
a polymerization time of 10 s. This may already point to a
different activation mechanism and active species compared to
the other complexes (see below). Furthermore, we note that
only the nickel pre-catalysts 3 and 4 which contain a six- or
seven-membered dppb and dpppt ligand, respectively, could be
activated with B(C6F5)3 alone for the vinyl polymerization of
norbornene, albeit with low activity. The GPC data of the
nickel/borane catalyzed poly(norbornene)s correspond closely
to the data of the polymer samples obtained with MAO as co-
catalyst. All polymers displayed a monomodal and narrow
molar mass distribution around Q = 2. The average chain length
lies between 1,300 (4/B(C6F5)3/AlEt3) and 12,000 (3/B(C6F5)3)
monomer units corresponding to a number-average molar mass
given in Table 3.
(Mn) between 1.2 × 105 and 1.1 × 106 g molϪ1
.
The influence of the halogen ligand was investigated and the
nickel–dppe complexes 1 and 13 with chloro and bromo ligands
were tested in combination with the two co-catalytic systems
MAO and B(C6F5)3/AlEt3 (Table 4 and Fig. 4). The polymeriz-
ation activity can be increased considerably when changing the
halogen ligand in the nickel–dppe complexes from chloro to
bromo. These effects may be explained by the weaker Ni–Br
bond strength.
Fig.
3
Polymerization activities of the metal–phopshine chelate
complexes [M{Ph2P(CH2)nPPh2}Cl2] 1–8 as a function of the chelate
size n (n=2 to 5); activation with MAO or B/Al (= B(C6F5)3/AlEt3).
Detailed conditions are given in Tables 2 and 3.
the polymer which should be highly advantageous for the
prospective optical applications of poly(norbornene).
For the palladium() pre-catalysts 5 and 6 we investigated the
influence of the molar ratio of B(C6F5)3 on the polymerization
activity (see Table 5 and Fig. 5). The Pd–dppe complex 5 and the
dppp-analog 6 had shown dramatic differences in the poly-
merization activities with B(C6F5)3/AlEt3 as co-catalysts.
Both palladium pre-catalysts 5 and 6 show the same activity
trends when varying the molar ratio of Pd : B(C6F5)3 in the
polymerization of norbornene. A decrease of the amount of
B(C6F5)3 from Pd : B(C6F5)3 : Al = 1 : 9 : 10 down to 1 : 2 : 10 led
to a steady decrease in polymerization activity. Yet, even at a
Thus, the activity correlates with the length of the alkyl-
group of the bidentate phosphane ligand, that is the chelate
ring size appears to be independent of the co-catalytic system
which is used to activate the nickel- or palladium-pre-catalysts
(Fig. 3). In the case of nickel the increase in chain length results
in a steady increase in the polymerization activity. Whereas for
the palladium complexes the polymerization activities dropped
significantly when going from n = 2 to 3 that is from a five- to a
D a l t o n T r a n s . , 2 0 0 3 , 4 4 3 7 – 4 4 5 0
4441

View Article Online
Table 4 Polymerization of norbornene (NB) with the Ni–dppe complexes 1 (Cl ligand) and 13 (Br ligand)
Catalyst
Time/min
Conversion (%)
Activity/gpolymer molNiϪ1 hϪ1
Mn/g molϪ1
Mw/g molϪ1
Q =Mw/Mn
1/MAOa
5
5
60
5
60
60
2.0
3.8
4.1
19.2
0
2.1 × 104
1.6 × 106
5.1 × 105
4.1 × 105
1.1 × 105
–
2.7 × 106
9.5 × 105
1.0 × 106
3.1 × 105
–
1.7
1.9
2.5
2.8
–
13/MAOa
4.3 × 104
3.8 × 103
b
1/B(C6F5)3/AlEt3
13/B(C6F5)3/AlEt3
1,13/AlEt3
1,13/B(C6F5)3
2.2 × 105
b
c
No activity observed
d
0
–
–
–
Conditions: room temperature; toluene/methylene chloride solution.a 10.6 mmol NB, [Ni] : [NB] = 1 : 1000, [Ni] : [Al] = 1 : 100, total volume 10.0 mL.
b 30.1 mmol NB, [Ni] : [NB] = 1 : 1000, [Ni] : [B] : [Al] = 1 : 9 : 10, total volume 40.0 mL. c 30.1 mmol NB, [Ni] : [NB] = 1 : 1000, [Ni] : [Al] = 1 : 10,
total volume 40.0 mL. d 30.1 mmol NB, [Ni] : [NB] = 1 : 1000, [Ni] : [B] = 1 : 9, total volume 40.0 mL.
Table 5 Polymerization of norbornene (NB) with the PdCl2 complexes 5 (dppe) and 6 (dppp) and varying ratios of borane in B(C6F5)3/AlEt3
Catalyst
Molar ratio Pd : B : Al
Time/min
Conversion (%)
Activity/gpolymer molPdϪ1 hϪ1
5
1 : 9 : 0
1/6
1/6
1/6
1/6
1/6
60
60
60
60
60
54.3
24.0
12.5
4.5
2.1
3.1
1.0
1.0
0
1.9 × 107
8.1 × 106
4.3 × 106
1.5 × 106
7.1 × 105
3.0 × 103
9.0 × 102
9.5 × 102
–
5
1 : 7 : 10
1 : 5 : 10
1 : 3 : 10
1 : 2 : 10
1 : 9 : 10
1 : 5 : 10
1 : 2 : 10
1 : 1 : 10
1 : 0 : 10
5
5
5
6
6
6
5, 6
5, 6
0
–
Conditions: room temperature; toluene/methylene chloride solution, 30 mmol NB, [Pd] : [NB] = 1 : 1000, total volume 40.0 mL.
complete loss in polymerization activity. No activity could be
detected within a 1 h reaction time. Hence, activation of the
metal–phosphane pre-catalysts 5 and 6 necessitates the presence
of at least two equivalents of B(C6F5)3. This is different from
the results of an analogous study on the activation of bis-
(diphenylglyoximato)nickel with B(C6F5)3/AlEt3 where a high
activity down to a molar ratio of Ni : B : Al = 1 : 1 : 10 was
found.29
In addition, it is known that B(C6F5)3 with trimethyl-
aluminium, AlMe3 85,86 or with AlEt3 29 gives an aryl/alkyl group
exchange (eqn. (1))
2 B(C6F5)3 ϩ 2 AlEt3
2 AlEt3 Ϫ n(C6F5)n ϩ2 B(C6F5)3 Ϫ nEtn (1)
Fig. 4 Polymerization activities of the nickel–dppe complexes 1 and 13
with chloro or bromo ligands; activation with MAO or B(C6F5)3/AlEt3.
Detailed conditions are given in Table 4.
which results in the formation of AlEt3 Ϫ n(C6F5)n and B(C6F5)3 Ϫ n
-
Etn with Al(C6F5)3 as the main product for a molar B : Al ratio
of about 1 : 1. Furthermore, aluminium organyls can work as
halide abstracting agents.61,65,87 Hence, AlEt3 Ϫ n(C6F5)n will then
react with the pre-catalysts and can abstract the two chloride
atoms to presumably form the polymerization active species
(eqn. (2)).
M(L)Cl2 ϩ 2 AlEt3 Ϫ n(C6F5)n
[M(L)]2ϩ ϩ 2 [Cl–AlEt3 Ϫ n(C6F5)n]Ϫ (2)
The bidentate phosphane ligand remains bound to palladium.
It was ascertained in separate experiments that neither
B(C6F5)3 nor AlEt3 alone acted as a co-catalyst towards the
palladium complexes 5 and 6. Thus, the presence and inter-
action of both components, B(C6F5)3 and AlEt3, is necessary
for activation of 5 and 6. The high Lewis acidity of B(C6F5)3
with its perfluorated phenyl ligands plays an important role in
the formation of the active species. It was shown for the other-
wise highly active complexes bis(acetylacetonato)nickel() and
bis(2-ethylhexanoato)nickel() that the weaker Lewis acid
B(C6H5)3 led to no polymerization activity.16
Fig. 5 Polymerization activities of the Pd complexes 5 and 6 with dppe
or dppp ligands as a function of the borane ratio; activation with
B(C6F5)3/AlEt3. Detailed conditions are given in Table 5.
NMR studies
ratio of Pd : B(C6F5)3 : Al = 1 : 2 : 10 the activity could be
considered (very) high for 5. However, a further decrease in the
amount of B(C6F5)3, below a ratio of Pd : B = 1 : 2 resulted in a
The activation hypothesis put forward in eqn. (2) was further
tested by multinuclear (1H, 19F, and 31P) NMR investigations.
D a l t o n T r a n s . , 2 0 0 3 , 4 4 3 7 – 4 4 5 0
4442

View Article Online
Scheme 3 31P NMR data of the interaction between dppe, dppp, dppb, 5 and 6 with B(C6F5)3 (solvent CD2Cl2, except for 5 and 6 alone in d6-DMSO
and dppb alone in CDCl3, room temperature, data given in ppm). See ESI† for complete spectroscopic details.
Instead of MAO the better defined co-catalytic system B(C6F5)3/
AlEt3 was used for the NMR studies. The fluorine atoms in
B(C6F5)3. The 31P signals for 5 or 6 and B(C6F5)3 are slightly
shifted compared to the 31P signal of the complexes as a result
of a (necessary) solvent change and the coordination or
abstraction of the chloride atoms by B(C6F5)3. In the case of
5/B(C6F5)3 the signal at ϩ56.9 ppm is assigned to [Pd(dppe)2]2ϩ
from a ligand redistribution of the unstable [Pd(dppe)]2ϩ (see
below). The NMR chemical shift is close to the one observed
for [Pd(dppe)2](OAc) (58.7 ppm in CD3OD) (see also Scheme
7).88,89 For 6/B(C6F5)3, the 31P NMR signal at ϩ16.0 ppm
assigned to [Pd(dppp)]2ϩ can be compared to the structurally
elucidated cation [Pd(dppp)(closo-CB11H12]ϩ in 16 (see below)
which resonates at ϩ17.9 ppm.
–C6F5 and the phosphane ligands on palladium provide for 19
F
and 31P NMR probes, in addition to 1H, to investigate the activ-
ation process. Detailed multinuclear NMR data of the com-
plexes 5 and 6 and the phosphane ligands dppe and dppp in
combination with different amounts of the Lewis acid B(C6F5)3
is available as ESI.†
The spectral changes in the 1H NMR are indicative of a reac-
tion between dppe and dppp and B(C6F5)3 and also between 5
and 6 and B(C6F5)3. However, in the latter reaction mixture they
do not allow one to distinguish between phosphane co-
ordination to Pd or to B. This is best elucidated by using 31P
NMR. The 31P signals and their interpretation are provided in
Scheme 3. The 31P signals of the dppe or dppp ligands in com-
bination with B(C6F5)3 are shifted with respect to the signals of
the pure ligands, hence indicating a possible P–B coordination.
We ascribe the one signal for dppe/B(C6F5)3 to a 1 : 1 adduct
between the reactants together with a fast equilibrium (on the
NMR time scale) between the two phosphorus donor atoms
(see Scheme 3). For dppp/B(C6F5)3 two signals are observed
which we assign to a 1 : 1 and a 1 : 2 adduct. The 1 : 2 adduct
becomes possible in the case of dppp because of the wider sep-
aration of the two phosphorus donor atoms by the longer alkyl
chain. Similar results were observed in the 1 : 9-spectrum of 1,4-
bis(diphenylphosphino)butane (dppb) in combination with
B(C6F5)3. The signal of the dppb ligand at Ϫ16.1 ppm (CDCl3,
RT) were low-field shifted by the addition of B(C6F5)3 and three
new signals appeared at 9.6 ppm, 13.9 ppm and 14.5 ppm, with
the latter two signals significantly broadened (CD2Cl2, RT).
Again, the signal at 9.6 ppm can be assigned to a 1 : 2-adduct
and the two broad signals at 13.9 ppm and 14.5 ppm should be
the result of the formation of the 1 : 1 adduct mentioned above
(similar chemical shifts compared to the signals of dppp/
B(C6F5)3). The appearance of two signals in the case of dppb/
B(C6F5)3 may be explained by a slow equilibrium between the
two phosphorus donor atoms.
The 19F NMR data supports the interaction of dppe and
dppp with B(C6F5)3. Two sets of ortho, meta and para signals
are observed in the case of a 1 : 9 molar ratio with the more
intense one corresponding to the free borane. The second set is
assigned to the P–B interaction (Scheme 4). The smaller inten-
sity ratio between free and coordinated borane of 2.4 : 1 in the
case of dppp/B(C6F5)3 versus 3.9 : 1 for dppe/B(C6F5)3 is in
accord with the interpretation of a 1 : 2 dppp–borane adduct in
addition to the 1 : 1 adduct (see above). Furthermore, chloride
abstraction from 5 and 6 can be deduced from the 19F NMR
data. The addition of various equivalents of B(C6F5)3 to the
palladium complexes 5 and 6 gives the same spectral features
(see Fig. 6 and Scheme 4). Two sets of ortho, meta and para
signals are again observed. It is a coincidence that the set
labelled with an asterisk (*) appears to have the same chemical
shifts as the set assigned to the P–B interaction from dppe and
dppp (Scheme 4). Alternatively, also the chloride-bridged
adduct [(F5C6)3B–Cl–B(C6F5)3]Ϫ from Et4NϩClϪ and B(C6F5)3
at a 1 : 9 ratio shows the same chemical shifts. The 19F NMR
chemical shifts in B(C6F5)3 are not very sensitive to the type of
donor atom (here P or Cl) bound to boron. Together with the
above 31P NMR results, the features in the 19F NMR from the
reaction of 5 or 6 and B(C6F5)3 (Fig. 6) have to be explained by
a fast equilibrium on the NMR time scale between free borane
and a 1 : 1 boron–chloride adduct, [Cl–B(C6F5)3]Ϫ, together
with an additional 2 : 1 boron–chloride adduct, [(F5C6)3B–Cl–
B(C6F5)3]Ϫ, as depicted in Scheme 4. The averaged signals for
Yet, the chemical shifts of the ligand–B(C6F5)3 adducts are
quite different from those seen from the reaction of 5 or 6 with
D a l t o n T r a n s . , 2 0 0 3 , 4 4 3 7 – 4 4 5 0
4443

View Article Online
Fig. 6 19F NMR spectra of Pd(dppe)Cl2 (5) and Pd(dppp)Cl2 (6) in
combination with different amounts of B(C6F5)3. The letters o, m,
p indicate the ortho, meta, and para fluorine atoms in the fast
equilibrium between B(C6F5)3 and [Cl–B(C6F5)3]Ϫ. The letters o*, m*,
p* indicate the ortho, meta, and para fluorine atoms in the [(F5C6)3B–
Cl–B(C6F5)3]Ϫ species (Scheme 4). The spectra were measured in
CD2Cl2 (0.6 mL) at room temperature; 7 µmol palladium complex and
the corresponding amount of B(C6F5)3. See ESI† for signal multiplicity,
FF coupling constants and further details.
the fast equilibrium between free borane and [Cl–B(C6F5)3]Ϫ are
shifted towards the signals for pure B(C6F5)3 with increasing
borane ratio. The signals for [(F5C6)3B–Cl–B(C6F5)3]Ϫ remain
nearly constant. There may also be an equilibrium between
[(F5C6)3B–Cl–B(C6F5)3]Ϫ and [Cl–B(C6F5)3]Ϫ/free borane which
is, however, slow on the NMR time scale. The assignment and
the formation of a boron–chloride adduct was verified by an
independent investigation of tetraethylammonium chloride,
Et4NϩClϪ with B(C6F5)3 where the same signal pattern could be
observed (Scheme 4).32
Still, the pre-catalysts 5 and 6 could not be activated with
B(C6F5)3 alone and triethylaluminium (AlEt3) had to be applied
simultaneously for activation. The 31P signals of the ternary
system and their interpretation are provided in Scheme 5. The
31P signals of dppe or dppp in combination with B(C6F5)3/AlEt3
are only very little shifted with respect to the signals of the pure
ligands. Such small shifts are in agreement, however, with
changes in 31P chemical shifts found in trimethylaluminium–
phosphane complexes.90,91 Thus, the shifts indicate a P–Al
adduct, possibly as part of a fast equilibrium between the free
components. Concerning the system dppp/10AlEt3 just one
single 31P resonance is observed. Possibly also the 1 : 2 complex
(dppp)(AlEt3)2 is part of the fast equilibrium between the 1 : 1
adducts or no 1 : 2 adduct is formed from AlEt3 alone but
only with Al(C6F5)3. The addition of ten equivalents of AlEt3 to
the complexes 5 and 6 in combination with different amounts
of B(C6F5)3 showed remarkable differences in the 31P NMR
spectra. In the case of 5/B(C6F5)3/AlEt3 the 31P NMR spectrum
was the same as observed before for 5/B(C6F5)3 (see Scheme 3),
that is the signal did not change when adding the AlEt3
to a solution of 5 and B(C6F5)3. The species with a signal
at 56.9 ppm in Schemes 3 and 5 is assigned to [Pd(dppe)2]2ϩ
which is derived from a ligand redistribution of the unstable
[Pd(dppe)]2ϩ (see also below).88,89,92 The redistribution also
gives rise to a Pd2ϩ cation. However, the reaction mixture
from 5/B(C6F5)3 and also 6/B(C6F5)3 is inactive towards poly-
merization. We explain this fact by a stronger, deactivating
coordination of [Cl–B(C6F5)3]Ϫ to the cations. The adduct [Cl–
Al(C6F5)3] is suggested to be less coordinating. In other words,
the Lewis-acid Al(C6F5)3 appears to be more activating than
B(C6F5)3.
Scheme 4 19F NMR data of the interaction between dppe, dppp and
ClϪ (from 5, 6 or Et4NCl) and B(C6F5)3 (CD2Cl2, room temperature,
data given in ppm). See ESI† for signal multiplicity, FF coupling
constants and further details.
normal NMR characteristics, the formation of a paramagnetic
palladium center can be excluded and the broadening of the 31
P
NMR signals should be the result of some fast exchange reac-
tion of the dppp ligand. Still, no signals were found at positions
seen in the dppp/B(C6F5)3/AlEt3 mixture, so that a complete
ligand abstraction by AlEt3 Ϫ n(C6F5)n can be excluded. Intra-
molecular phosphane exchange pathways responsible for spin
loss and signal coalescence were observed in the Pd(0) com-
plexes PdL2 with L = iPr2P(CH2)nPiPr2)2 for n = 3 and 4 (but not
n = 2). Besides, [Pd(iPr2P(CH2)nPiPr2)2] for n = 3 and 4 (but not
n = 2) are coordinatively unsaturated trigonal complexes which
are in equilibrium with the binuclear complexes LPd(η2-L)-
PdL.93 In summary, as for B(C6F5)3 the 31P NMR spectra also
excluded the dppe or dppp ligand abstraction from 5 or 6 by the
B(C6F5)3/AlEt3 components.
The 19F NMR spectra of the ternary mixture of the pal-
ladium complexes 5 or 6 in combination with different equiv-
alents of B(C6F5)3 and ten equivalents of AlEt3 are depicted
in Fig. 7. Their analysis requires an understanding of the
interactions between B(C6F5)3 and AlEt3 or between ClϪ,
B(C6F5)3 and AlEt3. As mentioned above, B(C6F5)3 and AlEt3
lead to a facile aryl/alkyl group exchange which results in the
formation of Al(C6F5)3 as the main product when using a
B(C6F5)3 : AlEt3 ratio of about 1 : 1 (see eqn. (1)). Lower
amounts of B(C6F5)3 or vice versa higher amounts of AlEt3
Upon addition of AlEt3 to a solution of 6 and B(C6F5)3 the
31P NMR signal broadened strongly which we can explain
either by formation of a paramagnetic palladium center or by
some fast exchange reaction involving the dppp ligand. Taking
1
into account that the H NMR and 19F NMR spectra showed
D a l t o n T r a n s . , 2 0 0 3 , 4 4 3 7 – 4 4 5 0
4444

View Article Online
spectroscopy (Scheme 6). In the 1 : 1 : 10 spectrum of Et4NϩClϪ
: B(C6F5)3 : AlEt3 three different groups of signals were found
and according to the intensity ratios and in comparison to the
spectra of B(C6F5)3/AlEt3 with different molar ratios, the sig-
nals can be attributed to the formation of Et2Al(Et)2AlEt(C6F5)
(weak signals), AlEt2(C6F5)/[Cl–AlEt2(C6F5)]Ϫ (broad signals)
and [Et2(F5C6)Al–Cl–AlEt2(C6F5)]Ϫ. The broad signals are
due to fast equilibrium on the NMR time scale between free
AlEt2(C6F5) and [Cl–AlEt2(C6F5)]Ϫ.
Scheme 6 19F NMR data of the interaction between ClϪ, B(C6F5)3 and
AlEt3 (CD2Cl2, room temperature, data given in ppm). See ESI† for
signal multiplicity, FF coupling constants and further details.
Scheme 5 31P NMR data of the interaction between dppe, dppp, 5 and
6 with AlEt3 and different amounts of B(C6F5)3. Al(C6F5)3 is derived
Addition of one equivalent of Et4NϩClϪ to a 9 : 10 mixture
of B(C6F5)3/AlEt3 resulted in a 19F NMR spectrum which
can also be explained by a fast equilibrium between free
from the nearly equimolar (9
: 10) reaction of B(C6F5)3 and
AlEt3 (resulting chemical shifts given in italics) (solvent CD2Cl2, except
for 5 and 6 alone where d6-DMSO was used, room temperature, data
given in ppm). See ESI† for complete spectroscopic details.
Al(C6F5)3 and the
1 : 1 aluminium–chloride adduct
together with an additional 2 : 1 aluminium–chloride adduct
(Scheme 6).
The 19F NMR spectra of the palladium pre-catalysts 5 and 6
in combination with different equivalents of B(C6F5)3 and ten
equivalents of AlEt3 showed remarkable differences (Fig. 7).
With regard to the differences in the polymerization activities
Ϫ1
of the palladium pre-catalysts 5 (1.9 × 107 gpolymer molPd hϪ1
)
and 6 (3.0 × 103 gpolymer molPdϪ1 hϪ1) it was important to concen-
trate on the signals which contain information about the
removal of the chloride atoms of the complexes. Beside the
known signals of the anionic boron-species [EtB(C6F5)3]Ϫ at
Ϫ132.98 (d, 23.7 Hz, o-F), Ϫ165.87 (t, 20.3 Hz, p-F) and
Ϫ168.38 (t, 18.7 Hz, m-F)95 at low B(C6F5)3 concentrations, the
19F NMR spectra were dominated by the signals of monomeric
respectively dimeric aluminium or aluminium–chloride adducts
with different numbers of C6F5-groups. Unfortunately, an
assignment of individual signals cannot be done with high
certainty. In case of the m-F-atoms (ca. Ϫ161 to Ϫ165 ppm)
and the p-F-atoms (ca. Ϫ151 to Ϫ160 ppm) it was possible to
distinguish between the signals of aluminium or aluminium–
chloride species and to extract quantitative information about
the chloride-abstraction from the pre-catalysts 5 and 6 by
B(C6F5)3/AlEt3. Using this information it was obvious that in
the case of the highly-active pre-catalyst 5 the signals of the
aluminium–chloride species were more intense than in the case
of the low-active pre-catalyst 6. Taking the absence of 31P
NMR signals into account (Scheme 5), the interactions between
6 and B(C6F5)3/AlEt3 result only in minor chloride abstractions
and enter instead into fast exchange reactions between
B(C6F5)3/AlEt3 and the dppp ligand. A direct correlation
between the intensities of the 19F NMR signals and the
differences in the polymerization activities of the palladium
pre-catalysts 5 and 6 is difficult and may lead to incomplete
assumptions for the activation process with B(C6F5)3/AlEt3 as
will be outlined below.
Fig. 7 19F NMR spectra of Pd(dppe)Cl2 (5) and Pd(dppp)Cl2 (6) in
combination with different amounts of B(C6F5)3 and 10 equivalents of
AlEt3. The following abbreviations were used: B*=signals of the
anionic species [EtB(C6F5)3]Ϫ at Ϫ132.98 (d, 23.7 Hz, o-F), Ϫ165.87
(t, 20.3 Hz, p-F) and Ϫ168.38 (t, 18.7 Hz, m-F),95 free = chloride-free
aluminium species of the type AlEt3 Ϫ n(C6F5)n, n=1–3, Cl*=1 : 1
aluminium–chloride adduct of the type [Cl–AlEt3 Ϫ n(C6F5)n]Ϫ, n=1–3
together with the corresponding 2 : 1 aluminium–chloride adduct.
Signals are assigned based on Et4NϩClϪ/B(C6F5)3/AlEt3; only the para-
fluorine atoms are indicated for clarity: pЉ=[Cl–AlEt2(C6F5)]Ϫ at
Ϫ155.83 (t, 19.2 Hz, p-F), pЉ2 =[Et2(F5C6)Al–Cl–AlEt2(C6F5)]Ϫ at
Ϫ160.24 (t, 19.4 Hz, p-F), p2 =[(F5C6)3Al–Cl–Al(C6F5)3]Ϫ at Ϫ158.96 (t,
19.2 Hz, p-F), pЈ=[Cl–AlEt(C6F5)2]Ϫ or [Et(F5C6)2Al–Cl–AlEt(C6F5)2]Ϫ
at Ϫ157.41 (t, 19.4 Hz, p-F) or Ϫ157.73 (t, 19.2 Hz, p-F), respectively.
The spectra were measured in CD2Cl2 (0.6 mL) at room temperature
(7 µmol palladium complex, 70 µmol AlEt3 and the corresponding
amount of B(C6F5)3).
indicate a stepwise aryl/alkyl group exchange and show the
formation of AlEt(C6F5)2 or AlEt2(C6F5) as main products,
depending on the B(C6F5)3 : AlEt3 ratio (see Fig. 7).85,94 The
interactions between ClϪ, B(C6F5)3 and AlEt3 were investigated
by the addition of one equivalent of Et4NϩClϪ to a 1 : 10 and
a 9 : 10 mixture of B(C6F5)3 and AlEt3 using 19F NMR
During the NMR experiments we became aware of the form-
ation of a fine, black precipitate following the addition of
B(C6F5)3/AlEt3 to the palladium pre-catalyst 5. This precipitate
was identified as elemental palladium and its formation may
D a l t o n T r a n s . , 2 0 0 3 , 4 4 3 7 – 4 4 5 0
4445

View Article Online
play an important role in the activation process (see below). The
following experiment proved the formation of elemental
palladium after the addition of B(C6F5)3/AlEt3 (Scheme 7).
Scheme 7
Under an inert-gas atmosphere 0.3 mmol 5 and 0.9
mmol B(C6F5)3 were dissolved in 25 mL dried CH2Cl2 and
3.0 mmol AlEt3 (1 mol LϪ1 in hexane) were added (molar ratio
5 : B(C6F5)3 : AlEt3 = 1 : 3 : 10). After two days the black
precipitate, which started to form directly after the addition of
AlEt3 was centrifuged, washed with CH2Cl2 and redissolved in
HNO3/HCl (v/v = 1/3). After the removal of HNO3 via evapor-
ation, a solution of dimethylglyoxime in ethanol was added
to the HCl/water solution. The yellow precipitate was clearly
identified as bis(dimethylglyoxime)palladium() by IR. A
mass balance of bis(dimethylglyoxime)palladium() showed a
reduction of about 38% of the palladium() after addition of
B(C6F5)3/AlEt3. After the removal of elemental palladium
(after two days!) the remaining clear, red-brown CH2Cl2 solu-
tion showed one singlet at 58.6 ppm in its 31P NMR which
proved the coordination of the dppe-ligand to palladium()
(cf. Scheme 5) presumably as [Pd(dppe)2]2ϩ (see below). The
NMR chemical shift is close to the one observed for [Pd-
(dppe)2](OAc) (58.7 ppm in CD3OD).88,89 In the case of the low-
active pre-catalyst 6 in combination with B(C6F5)3/AlEt3 no
precipitation with respect to the reduction of palladium() was
observed over a period of several hours.
Fig. 8 Molecular structure of [Pd(dppe)2][ClB(C6F5)3]2ؒ4CH2Cl2 (14);
only one of the symmetry related anions is shown for clarity. Selected
distances (Å) and angles (Њ): Pd–P1 2.3412(10), Pd–P2 2.3582(9), B–Cl1
1.925(4), B–C 1.626(6)–1.644(5); P1–Pd–P2 97.31(3), P1–Pd–P2_2
82.68(3), Cl–B–C 102.2(3)–112.3(3), C–B–C 107.7(3)–114.2(3);
symmetry operation _2=Ϫxϩ1, Ϫy, ϪzϪ1. CH2Cl2 molecules of
crystallization are omitted for clarity.
from [Pd(dppe)]2ϩ is due to the absence of nucleophilic counter
anions as the opposite occurs in the presence of such coordin-
ating counter anions with the bis-chelate as the kinetic product
and the mono-chelate as the thermodynamic product.88,89,92
In order to provide a more stabilizing anion for the potential
[Pd(L)]2ϩ cation we reacted the monocarba-closo-dodeca-
borane(12) anion, [closo-1-CB11H12]Ϫ as its silver salt with the
palladium complexes 5 and 6. In both cases suitable crystals
were obtained after several days by overlayering a filtered
methylene chloride solution of complex and carborane with
diethyl ether.97 Starting from the palladium–dppe complex 5,
the [Pd(dppe)2]2ϩ cation96 is again found in the crystal structure
of [PdII(dppe)2][CB11H11Cl]2ؒ3CH2Cl2 (15, Fig. 9). Thus, the
instability of mono-ligated [Pd(dppe)]2ϩ and the reaction given
in eqn. (3) must be invoked again. The carborane counterion to
[Pd(dppe)2]2ϩ has picked up a chlorine atom (probably from
CH2Cl2) and become [closo-CB11H11Cl]Ϫ.97 The phenyl rings on
the opposing dppe ligands in the [Pd(dppe)2]2ϩ cations in 14 and
15 are too far apart for a meaningful π–π interaction,98 different
from the situation in the [Ni(dppe)2]2ϩ cation.99
X-Ray structure determinations
To further understand the activity differences of the palladium
complexes 5 and 6 we tried to isolate crystalline products from
the interaction between the complexes and the Lewis acid
B(C6F5)3. We are aware, of course, that isolable crystalline
products may have little to do with the polymerization-active
solution species, hence, any structural results have to be inter-
preted with caution towards the catalytic observations.
Complexes 5 or 6 and B(C6F5)3 (molar ratios = 1 : 3) were
dissolved in a mixture of toluene and methylene chloride (v/v =
1 : 2) and it was only possible to obtain extremely air-sensitive
crystals starting from 5/B(C6F5)3 after several months. Under
the same conditions a crystallization starting from complex 6
was not possible. Structure analysis revealed the formation of
a [Pd(dppe)2]2ϩ fragment96 with [ClB(C6F5)3]Ϫ anions in [PdII-
(dppe)2][ClB(C6F5)3]2ؒ4CH2Cl2 (14) from 5/B(C6F5)3 (Fig. 8).
The isolation of [ClB(C6F5)3]Ϫ supports the chloride abstracting
functionality of B(C6F5)3 which was outlined above. The form-
ation of [Pd(dppe)2]2ϩ must be explained by a ligand redistri-
bution reaction between two [Pd(dppe)]2ϩ moieties (eqn. (3)).
2 [Pd(dppe)]2ϩ
[Pd(dppe)2]2ϩ ϩ Pd2ϩ
(3)
The latter cations are apparently unstable in the absence of
stabilizing anions (see below).
Fig. 9 Molecular structure of [Pd(dppe)2][CB11H11Cl]2)ؒ3CH2Cl2 (15).
Selected distances (Å) and angles (Њ): Pd–P 2.3442(6), B12–Cl1 1.836(4);
P–Pd–P_4 81.88(3), P1–Pd–P_2 98.12(3). CH2Cl2 molecules of
crystallization are omitted for clarity. The palladium atom and the
carborane are cut by a mirror plane passing through Pd, C1b, B3, B10,
B12 and Cl1. Furthermore, Pd sits on a center of inversion. Symmetry
transformations _2=Ϫx, y, Ϫz; _3=Ϫx, Ϫy, Ϫz; _4=x, Ϫy, z.
The same ligand redistribution reaction between two
[Pd(dppe)]2ϩ moieties was formulated in the NMR studies
and catalytic activation reactions of 5 with B(C6F5)3 or
B(C6F5)3/AlEt3 (see Schemes 3 and 5). The concomittant
formation of elemental palladium was demonstrated in Scheme
7. It has been suggested that the formation of [Pd(dppe)2]2ϩ
D a l t o n T r a n s . , 2 0 0 3 , 4 4 3 7 – 4 4 5 0
4446

View Article Online
In contrast, the crystals derived from the palladium–dppp
complex 6 and Ag[closo-1-CB11H12] showed the expected form-
ation of the mono-ligated [Pd(dppp)]2ϩ species in the structure
of [PdII(dppp)(CB11H12)][CB11H12] (16). Here the cation was
stabilized by coordination to one of the carborane anions as
an exo-(diphosphane)palladium-closo-monocarborane com-
plex [Pd(dppp)(closo-CB11H12]ϩ (Fig. 10). This complex cation
crystallizes with a second carborane equivalent [CB11H12]Ϫ
as the counter ion. In the structure of 16 about 22% of the
carborane counter anions were found to contain a hydroxy
group (introduced most likely by traces of water in the solvent)
and hence be of formula [CB11H11(OH)]Ϫ. The 31P NMR signal
of the cation is observed at ϩ17.9 ppm, close to the ϩ16.0 ppm
of the proposed species [Pd(dppp)]2ϩ formed in solution from 6
and B(C6F5)3 (see Scheme 3). Similar structures with [CB11H12]
ligated to a metal in a bidentate mode have already been
described with platinum74 or rhodium.75,100
in a ligand redistribution reaction to form a [Pd(dppe)2]2ϩ
cation and “naked” Pd2ϩ (demonstrated by the 31P NMR
experiments and the formation of elemental palladium under
non-polymerization-conditions and indirectly by the crystal
data of 14 and 15). “Naked” Pd2ϩ cations are supposed to be
the highly-active species for the polymerization of norbornene.
The in-situ formation of “naked” Pd2ϩ appears to be the key to
the high activity which is not seen when applying Pd() salts,
such as PdCl2 in bulk.32 The chloride abstraction and the ligand
redistribution should be a fast two-step-reaction for pre-
catalyst 5 – as shown by the short polymerization time (54%
conversion in 10 s) and the immediate precipitation of
elemental palladium under non-polymerization-conditions. (3)
No ligand redistribution but fast exchange reactions between
B(C6F5)3/AlEt3 and the dppp ligand were observed in the case
of 6/B(C6F5)3/AlEt3. This was indentified by the absence of 31P
NMR signals in the spectra of 6/B(C6F5)3/AlEt3, by the crystal
data of compound 16 and the missing precipitation of
elemental palladium after the addition of B(C6F5)3/AlEt3 to
pre-catalyst 6.
The possibility to undergo a ligand redistribution reaction
and thereby to form “naked” Pd2ϩ as a highly-active species for
the polymerization of norbornene could be the result of the
steric features in the [Pd(L)]2ϩ fragments with L = dppe and
dppp and in the resulting square-planar [Pd(L)2]2ϩ complexes.
Comparison of the crystal data of the pre-catalysts 5 and
6101,102 or compounds 14–16 showed the expected smaller bite
angle on Pd for the five-membered ring formed with dppe. The
bite angle with dppp in a six-membered chelate ring is larger by
about 5Њ (Scheme 9, argument 1).
Fig. 10 Molecular structure of [Pd(dppp)(CB11H12)][CB11H12] (16).
Selected distances (Å) and angles (Њ): Pd–P1 2.257(1), Pd–P2 2.253(1),
Pd–B7a 2.401(4), Pd–H7a 1.945, Pd–B12a 2.406(4), Pd–H12a 1.929;
P2–Pd–P1 89.48(3), P1–Pd–B7a 113.3(1), P2–Pd–B7a 155.6(1), P1–Pd–
B12a 154.8(1), P2–Pd–B12a 114.0(1).
Combining the results of the multinuclear NMR investi-
gations with the pre-catalysts 5 and 6 in combination with
B(C6F5)3 and B(C6F5)3/AlEt3 with the structural features of the
compounds 14, 15 and 16, the activation process of the pre-
catalysts may be formulated in the following way (Scheme 8). (1)
Addition of B(C6F5)3 or B(C6F5)3/AlEt3 to the pre-catalysts 5
and 6 resulted at least in partial chloride abstraction and in the
formation of [Pd(L)]2ϩ fragments (L = dppe, dppp) as shown by
the multinuclear NMR experiments. These [Pd(L)]2ϩ species
exhibit however only a low to medium polymerization activity.
(2) In case of 5, the [Pd(dppe)]2ϩ cations are unstable and react
Scheme 9
It is obvious that in the case of the dppp-ligand the
larger P–Pd–P-angle will increase the steric interaction between
two coordinated dppp-ligands which may explain the not-
observed formation of the [PdII(dppp)2]2ϩ complex. No ligand
redistribution means no formation of “naked” Pd2ϩ. A
palladium(0) species of formula [Pd0(dppp)2] is described in the
literature. The formation of the zero-valent species is the result
of the distorted tetrahedral coordination sphere around the
palladium(0) center which was also found in the correspond-
ing [Pd0(dppe)2] species.103 There are reports on the failed syn-
thesis of [Pd(dppp)2]2ϩ 93,103 together with a recent paper on the
Scheme 8
D a l t o n T r a n s . , 2 0 0 3 , 4 4 3 7 – 4 4 5 0
4447

View Article Online
preparation of [Pd(dppp)2]2ϩ (without any structural proof on
the formation of a bis-chelate complex, however).104 We note
again that the Pd(0) complexes PdL2 with L = iPr2P(CH2)nPiPr2)2
for n = 3 and 4 (but not n = 2) were described as coordinatively
unsaturated trigonal complexes (one P-atom is dangling free)
which are in equilibrium with the binuclear complexes LPd-
(η2-L)PdL.93 There appears to be a principle difference in the
ide atoms. The anions detected were the species [Cl–E(C6F5)3]Ϫ
and [(F5C6)3E–Cl–E(C6F5)3]Ϫ with E = B when using B(C6F5)3
alone, E = Al when employing B(C6F5)3/AlEt3. B(C6F5)3
and AlEt3 lead to an aryl/alkyl group exchange which results
in the formation of AlEt3 Ϫ n(C6F5)n and B(C6F5)3 Ϫ nEtn with
Al(C6F5)3 and BEt3 being the main products for an about equi-
molar ratio of B(C6F5)3 and AlEt3. The extremely high poly-
merization activity of the complex Pd(dppe)Cl2 (5, 1.9 × 107
gpolymer molPdϪ1 hϪ1), for example in comparison to Pd(dppp)Cl2
(6, 3.0 × 103 gpolymer molPdϪ1 hϪ1), upon activation with B(C6F5)3/
AlEt3 is explained by the formation of highly active “naked”
stability (and structure?) of [Pd(dppe)2]2ϩ and [Pd(dppp)2]2ϩ
.
The former has been found to be inactive in CO/ethene co-
polymerization while the latter exhibits an appreciable activity
and can react with CO/H2O.89,92 The possible formation of
[Pd(dppp)]2ϩ as an intermediate is also seen in the formation
Pd2ϩ through
a ligand redistribution of the unstable
of a dimeric, unsaturated palladium() species of the type
[Pd(dppe)]2ϩ cation to yield [Pd(dppe)2]2ϩ and Pd2ϩ together
with the lower coordinating ability of the anionic adduct [Cl–
Al(C6F5)3]Ϫ in comparison to [Cl–B(C6F5)3]Ϫ. [Pd(dppe)2]2ϩ
cations were found in crystalline solids derived from 5
and B(C6F5)3 or Ag[closo-1-CB11H12]. In the absence of the
norbornene monomer the formation of Pd metal is observed
for 5/B(C6F5)3/AlEt3 but not for 6/B(C6F5)3/AlEt3. The find-
ing of “simple” unligated species upon activation in cyclo-
olefin polymerization agrees with our recent observation on the
in situ formation of PdCl2 from palladium() salts containing
[PdCl4]2Ϫ and [Pd2Cl6]2Ϫ ions.32
2ϩ 105
[(dppp)PdI]2
.
In the case of the pre-catalyst 5 the ligand
redistribution reaction and the formation of “naked” Pd2ϩ as a
highly-active species should be the result of a smaller P–Pd–P
angle which reduces the steric interaction between the two
dppe-ligands (Scheme 9, argument 1). For the smaller nickel ion
the formation of [NiII(dppe)2]2ϩ and [Ni0(dppe)2] is also
possible.99,106
In addition to the formation of “naked” Pd2ϩ the difference
in bite angle in the [Pd(L)]2ϩ-fragments with L = dppe and dppp
follows the trend for the polymerization activity (Scheme 3, see
Scheme 9, argument 2). Ligand dppe is described as a “narrow-
bite” ligand towards Pd because of its P–Pd–P-angle.107 After
chloride-abstraction a smaller bite angle offers more space on
the metal center for incoming norbornene units and the grow-
ing polymer chain. In contrast to this a “wide-bite” ligand like
dppp will provide less space and larger steric hindrance for an
insertion-polymerization of norbornene and therefore decrease
the polymerization activity. This argument also holds for the
smaller nickel atom, where the bite angle increases from dppe108
to dppp.109 As an opposing trend towards the increase in bite
angle, the observed rise in activity towards the larger dppb
and dpppt ligands (Scheme 3) can be explained by the lessened
chelate effect which will lead to an easier and more frequent
opening of the chelate ring. This ring opening may be helped by
a P–Al interaction as noted above for 6/B(C6F5)3/AlEt3 (to
explain the absence of 31P NMR signals).
Acknowledgements
The research was supported by the Fonds der Chemischen
Industrie. We appreciate a gift of PdCl2 by Degussa-Hüls AG
and a gift of MAO by Witco GmbH. We thank the referees who
provided helpful comments.
References and notes
1 K. J. Ivin and J. C. Mol, in Olefin Metathesis And Metathesis
Polymerization, Academic Press, San Diego, 1997, p. 407.
2 R. H. Grubbs, in Comprehensive organometallic chemistry,
Pergamon Press, Oxford, 1982, vol. 8, p. 499.
3 C. Janiak, in Moderne Anorganische Chemie, ed. E. Riedel, de
Gruyter, Berlin, 1998, p. 708.
4 J. P. Kennedy and H. S. Makowski, J. Macromol. Sci.-Chem., 1967,
A1, 345–370; N. G. Gaylord, B. M. Mandal and M. Martan,
J. Polym. Sci., Polym. Lett. Ed., 1976, 14, 555–559; N. G. Gaylord,
A. B. Deshpande, B. M. Mandal and M. Martan, J. Macromol.
Sci.-Chem., 1977, A11(5), 1053–1070; N. G. Gaylord and A. B.
Deshpande, J. Polym. Sci., Polym. Lett. Ed., 1976, 14, 613–617.
5 G. Sartori, F. C. Ciampelli and N. Cameli, Chim. Ind. (Milano),
1963, 45, 1478.
The probably necessary formation of a metal–organyl start-
ing group for the insertion polymerization of norbornene is
suggested to occur via C6F5 transfer from B(C6F5)3ؒH2O110–112
113
and Al(C6F5)3 respectively to the metal center or by alkyl-
ation of the metal center from the known alkyl transfer agents
BEt3 or AlEt3.87
6 J. P. Koinzer, U. Langbein and E. Taeger, Ger. Pat., 2421838, 1975;
VEB Leuna-Werke (GDR), Chem. Abstr., 1976, 84, 60227y.
7 W. Kaminsky, A. Bark and M. Arndt, Makromol. Chem., Macromol.
Symp., 1991, 47, 83–93.
8 W. Kaminsky and A. Noll, Polym. Bull., 1993, 31, 175–182.
9 H. Herfert, P. Montag and G. Fink, Makromol. Chem., 1993, 194,
3167–3182.
10 W. Kaminsky, Angew. Makromol. Chem., 1994, 223, 101–120.
11 M. Arndt and I. Beulich, Macromol. Chem. Phys., 1998, 199, 1221–
1232.
Conclusions
Norbornene undergoes a vinyl/addition polymerization using
activated phosphane-chelated metal() halide complexes of the
general type [M{Ph2P(CH2)nPPh2}X2] with n = 2 (dppe), 3
(dppp), 4 (dppb) and 5 (dpppt), M = Co, Ni and Pd, X = Cl, Br.
The nickel() and palladium() complexes gave polymerization
activities of about 103–107 gpolymer molmetalϪ1 hϪ1 when activated
with methylalumoxane (MAO) or with a combination of
12 U. Peuckert and W. Heitz, Macromol. Rapid. Commun., 1998, 19,
159–162.
tris(pentafluorophenyl)borane/triethylaluminium
[B(C6F5)3/
13 B. L. Goodall, L. H. McIntosh III and L. F. Rhodes, Makromol.
Chem., Macromol. Symp., 1995, 89, 421–432.
14 F. P. Alt and W. Heitz, Acta Polym., 1998, 49, 477–481.
15 F. P. Alt and W. Heitz, Macromol. Chem. Phys., 1998, 199, 1951–
1956.
16 P. G. Lassahn, C. Janiak and J. S. Oh, Z. Naturforsch., Teil B, 2001,
56, 1289–1292.
17 F. Peruch, H. Cramail and A. Deffieux, Macromol. Chem. Phys.,
1998, 199, 2221–2227.
18 M. Arndt and M. Gosmann, Polym. Bull., 1998, 41, 433–440.
19 C. Mast, M. Krieger, K. Dehnicke and A. Greiner, Macromol. Rapid
Commun., 1999, 20, 232–235.
20 S. Borkar and P. K. Saxena, Polym. Bull., 2000, 44, 167–172.
21 T. J. Deming and B. M. Novak, Macromolecules, 1993, 26, 7089–
7091.
AlEt3]. It was also possible to activate the nickel complexes
Ni(dppb)Cl2 and Ni(dpppt)Cl2 with the Lewis acid B(C6F5)3
alone. Bromine complexes gave a somewhat higher activity than
the analogous chlorine compounds, apparently due to the
weaker Ni–Br bond. The increase in the alkyl-chain length
resulted for the nickel complexes in a steady increase in the
polymerization activity whereas for the palladium complexes
the polymerization activities dropped first from dppe to the
dppp complex and then increased subsequently. Multinuclear
(31P and 19F) NMR investigations on the complexes Pd(dppe)-
Cl2 and Pd(dppp)Cl2 in combination with B(C6F5)3 and
B(C6F5)3/AlEt3 suggest the formation of [Pd(L)]2ϩ cations in the
activation process. E(C6F5)3 (E = B, Al) reacts with the pre-cata-
lysts [M{Ph2P(CH2)nPPh2}X2] under abstraction of the two hal-
22 W. Massa, N. Faza, H.-C. Kang, C. Focke and W. Heitz,
Acta Polym., 1997, 48, 385–391.
D a l t o n T r a n s . , 2 0 0 3 , 4 4 3 7 – 4 4 5 0
4448

View Article Online
23 H. Maezawa, J. Matsumoto, H. Aiura and S. Asahi, Eur. Pat.,
0445755, 1991; Idemitsu Kosan CO (JP), Chem. Abstr., 1991, 115,
256943g.
24 K.-H. Ostoja Starzewski, Eur. Pat., 0829493, 1998; Bayer AG (DE),
Chem. Abstr., 1998, 128, 2177.
60 G. H. Hlatky, R. R. Eckman and H. W. Turner, Organometallics,
1992, 11, 1413–1416.
61 C. Janiak, Metallocene Catalysts for Olefin Polymerization, in
Metallocenes, ed. A. Togni and R.L. Halterman, Wiley-VCH,
Weinheim, 1998, ch. 9, p. 547.
25 U. Denninger and K.-H. Ostoja Starzewski, Eur. Pat., 0837079,
1998; Bayer AG (DE), Chem. Abstr., 1998, 128, 308903j.
26 B. L. Goodall, L. H. McIntosh III and D. A. Barnes, WO Pat.,
9637529, 1997; B.F. Goodrich Company (USA), Chem. Abstr., 1997,
126, 104554v.
27 B. L. Goodall, G. M. Benedikt, L. H. McIntosh III, D. A. Barnes
and L. F. Rhodes, WO Pat., 9514048, 1995; B.F. Goodrich Company
(USA), Chem. Abstr., 1995, 123, 341322p.
28 Equivalents: US Pat., 5468819, 1995; B.F. Goodrich Company
(USA), Chem. Abstr., 1995, 123, 341322p.
29 B. Berchtold, V. Lozan, P.-G. Lassahn and C. Janiak, J. Polym. Sci.
A: Polym. Chem., 2002, 40, 3604–3614.
30 G. M. Benedikt, E. Elce, B. L. Goodall, H. A. Kalamarides, L. H.
McIntosh III, L. F. Rhodes and K. T. Selvy, Macromolecules, 2002,
35, 8978–8988.
31 D. A. Barnes, G. M. Benedikt, B. L. Goodall, S. S. Huang, H. A.
Kalamarides, S. Lenhard, L. H. McIntosh III, K. T. Selvy, R. A.
Shick and L. F. Rhodes, Macromolecules, 2003, 36, 2623–2632.
32 P.-G. Lassahn, V. Lozan and C. Janiak, Dalton Trans., 2003, 927–
935.
62 M. Bochmann, J. Chem. Soc., Dalton Trans., 1996, 255–270.
63 S. Collins, L. F. Rhodes and B. L. Goodall, US Pat., 5668234, 1997;
B.F. Goodrich Company (USA), Chem. Abstr., 1997, 127, 263591s.
64 I Tritto, R. Donetti, M. C. Sacchi, P. Locatelli and G. Zannoni,
Macromolecules, 1997, 30, 1247–1252.
65 C. Janiak, Metallocene Catalysts, in Kirk-Othmer Encyclopedia
of Chemical Technology Online Edition 2002, DOI: 10.1002/
0471238961.1305200110011409.a01, http://www.mrw.interscience.
wiley.com/kirk/articles/metajani.a01/frame.html.
66 S. J. McLain, J. Feldman, E. F. McCord, K. H. Gardner, M. F.
Teasley, E. B. Coughlin, K. J. Sweetman, L. K. Johnson and
M. Brookhart, Macromolecules, 1998, 31, 6705–6707.
67 B. Y. Lee, G. C. Bazan, J. Vela, Z. J. A. Komon and X. Bu, J. Am.
Chem. Soc., 2001, 123, 5352–5353.
68 N. R. Grove, P. A. Kohl, S. A. B. Allen, S. Jayaraman and R. Shick,
J. Polym. Sci., Part B: Polym. Phys., 1999, 37, 3003–3010.
69 SMART, Data Collection Program for the CCD Area-Detector
System and SAINT, Data Reduction and Frame Integration
Program for the CCD Area-Detector System, Bruker Analytical
X-ray Systems, Madison, WI, USA, 1997.
33 P. G. Lassahn, C. Janiak and J. S. Oh, Macromol. Rapid Commun.,
2002, 23, 16–20.
34 C. Janiak and P. G. Lassahn, Polym. Bull., 2002, 47, 539–546.
35 A. Sen, T.-W. Lai and R. R. Thomas, J. Organomet. Chem., 1988,
358, 567–588.
36 T.-W. Lai and A. Sen, Organometallics, 1982, 1, 415–417.
37 A. Sen and T.-W. Lai, J. Am. Chem. Soc., 1981, 103, 4627–4629.
38 C. Mehler and W. Risse, Makromol. Chem., Rapid Commun., 1991,
12, 255–259.
39 N. Seehof, C. Mehler, S. Breunig and W. Risse, J. Mol. Catal., 1992,
76, 219–228.
40 C. Mehler and W. Risse, Makromol. Chem., Rapid Commun., 1992,
13, 455–459.
41 S. Breunig and W. Risse, Makromol. Chem., 1992, 193, 2915–
2927.
42 J. Melia, E. Connor, S. Rush, S. Breunig, C. Mehler and W. Risse,
Makromol. Chem., Macromol. Symp., 1995, 89, 433–442.
43 T. F. A. Haselwander, W. Heitz, S. A. Krügel and J. H. Wendorff,
Macromol. Chem. Phys., 1996, 197, 3435–3453.
44 T. F. A. Haselwander, W. Heitz and M. Maskos, Macromol. Rapid.
Commun., 1997, 18, 689–697.
70 G. Sheldrick, Program SADABS: Area-detector absorption
correction, University of Göttingen, Germany, 1996.
71 G. M. Sheldrick, SHELXS-97, SHELXL-97, Programs for Crystal
Structure Analysis, University of Göttingen, Germany, 1997.
72 A. J. Welch and A. S. Weller, J. Chem. Soc., Dalton Trans., 1997,
1205–1212.
73 T. Jelinek, P. Baldwin, W. R. Scheidt and C. A. Reed, Inorg. Chem.,
1993, 32, 1982–1990.
74 G. S. Mhinzi, S. A. Litster, A. D. Redhouse and J. L. Spencer,
J. Chem. Soc., Dalton Trans., 1991, 2769–2776.
75 A. S. Weller, M. F. Mahon and J. W. Steed, J. Organomet. Chem.,
2000, 614-615, 113–119.
76 M. N. Burnett, C. K. Johnson, ORTEP-III: Oak Ridge Thermal
Ellipsoid Plot Program for Crystal Structure Illustrations, Report
ORNL-6895, Oak Ridge National Laboratory, Oak Ridge, TN,
1996; L. J. Farrugia, ORTEP 3 for Windows, Version 1.0.1β,
University of Glasgow, 1997; L. J. Farrugia, J. Appl. Crystallogr.,
1997, 30, 565.
77 I. Sacconi and J. Gelsomini, Inorg. Chem., 1968, 7, 291–294.
78 K. Shelly, D. C. Finster, Y. J. Lee, W. R. Scheidt and C. A. Reed,
J. Am. Chem. Soc., 1985, 107, 5955–5959.
45 A. S. Abu-Surrah, U. Thewalt and B. Rieger, J. Organomet. Chem.,
1999, 587, 58–66.
46 J. Lipian, R. A. Mimna, J. C. Fondran, D. Yandulov, R. A. Shick,
B. L. Goodall, L. F. Rhodes and J. C. Huffman, Macromolecules,
2002, 35, 8969–8977.
79 R. Busby, M. B. Hursthouse, P. S. Jarrett, C. W. Lehmann, K. M. A.
Malik and C. Phillips, J. Chem. Soc., Dalton Trans., 1993, 3767–
3770.
80 P. J. Flory, in Principles of Polymer Chemistry, Cornell University
Press, Ithaca, NY, 1986, p. 317.
47 A. D. Hennis, J. D. Polley, G. S. Long, A. Sen, D. Yandulov,
J. Lipian, G. M. Benedikt, L. F. Rhodes and J. Huffman,
Organometallics, 2001, 20, 2802–2812.
48 P. G. Lassan, C. Tzschuke, W. Bannwarth and C. Janiak,
Z. Naturforsch., B, 2003, 58, in press.
81 C. Janiak, B. Rieger, R. Voelkel and H.-G. Braun, J. Polym. Sci.,
Polym. Chem., 1993, 31, 2959–2968.
82 Macromol. Symp., 1995, 97, 1–246, containing the papers of lectures
at the Hamburg macromolecular colloquium on “Alumoxanes” in
1994.
49 C. Janiak and P. G. Lassahn, J. Mol. Catal. A, Chem., 2001, 166,
193–209.
50 L. H. McIntosh III, B. L. Goodall, R. A. Shick and J. Saikumar, US
Pat., 5912313, 1999; B.F. Goodrich Company (USA), Chem. Abstr.,
1997, 127, 110414m.
83 I. Tritto, M. C. Sacchi, P. Locatelli and S. X. Li, Macromol. Chem.
Phys., 1996, 197, 1537–1544.
84 W. Kaminsky, Angew. Makromol. Chem., 1994, 233, 101–120.
85 M. Bochmann and M. J. Sarsfield, Organometallics, 1998, 17, 5908–
5912.
51 B. L. Goodall, W. Risse and J. P. Mathew, US Pat., 5705503, 1998;
B.F. Goodrich Company (USA), Chem. Abstr., 1997, 126, 104553u.
52 B. L. Goodall, L. H. McIntosh III and D. A. Barnes, US Pat.,
5677405, 1997; B.F. Goodrich Company (USA), Chem. Abstr., 1997,
126, 104554v.
53 B. L. Goodall and L. H. McIntosh III, WO Pat., 9856839, 1998;
B.F. Goodrich Company (USA), Chem. Abstr., 1999, 130, 66911g.
54 J.-H. Lipian, A. Bell, B. L. Goodall, L. F. Rhodes, R. A. Shick and
J. Saikumar, US Pat., 6232417, 2001; B.F. Goodrich Company
(USA), Chem. Abstr., 1999, 130, 252843m.
86 J. Klosin, G. R. Roof, E. Y. X. Chen and K. A. Abboud,
Organometallics, 2000, 19, 4684–4686.
87 J. R. Zietz, in Ullmann, Encyclopedia of Industrial Chemistry, VCH,
Weinheim, 1985, vol. A1 pp. 284 and 543.
88 A. Marson, A. B. van Oort and W. P. Mul, Eur. J. Inorg. Chem.,
2002, 3028–3031.
89 C. Bianchini, H. M. Lee, A. Meli, W. Oberhauser, M. Peruzzini and
F. Vizza, Organometallics, 2002, 21, 16–33.
90 C. Janiak, K. C. H. Lange, U. Versteeg, D. Lentz and P. H. M.
Budzelaar, Chem. Ber., 1996, 129, 1517–1529.
55 B. L. Goodall and L. H. McIntosh III, US Pat., 6265506, 2001;
B.F. Goodrich Company (USA), Chem. Abstr., 2001, 135, 122919h.
56 B. L. Goodall, B. F. Goodrich Company (USA) in Book of
Abstracts, 215th ACS National Meeting, Dallas, March 29–April 2,
1998, INOR-263, American Chemical Society, Washington, DC,
1998.
91 A. R. Barron, J. Chem. Soc., Dalton Trans., 1988, 3047–3050.
92 C. Bianchini, A. Meli and W. Oberhauser, Dalton Trans., 2003, 2627–
2635.
93 M. Portnoy and D. Milstein, Organometallics, 1993, 12, 1655–
1664.
94 P.-G. Lassahn and C. Janiak, unpublished results.
95 D. A. Walker, T. J. Woodman, D. L. Hughes and M. Bochmann,
Organometallics, 2001, 20, 3772–3776.
57 X. Yang, C. L. Stern and T. J. Marks, J. Am. Chem. Soc., 1994, 116,
10015–10031.
58 E. Y.-X. Chen and T. J. Marks, Chem. Rev., 2000, 100, 1391–1434.
59 G. H. Hlatky, H. W. Turner and R. R. Eckman, J. Am. Chem. Soc.,
1989, 111, 2728–2729.
96 For comparison with [Pd(dppe)2]Cl2 see: L. M. Engelhardt, J. M.
Patrick, C. L. Raston, P. Twiss and A. H. White, Aust. J. Chem.,
1984, 37, 2193–2200.
D a l t o n T r a n s . , 2 0 0 3 , 4 4 3 7 – 4 4 5 0
4449

View Article Online
97 N. J. Patmore, M. F. Mahon, J. W. Steed and A. S. Weller, J. Chem.
Soc., Dalton Trans., 2001, 277–283.
98 C. Janiak, J. Chem. Soc., Dalton Trans., 2000, 3885–3896.
99 [Ni(dppe)2]-(NO3)2 and -(ClO4)2: A. F. Williams, Acta Crystallogr.,
Sect. C, 1989, 45, 1002–1005.
100 A. Rifat, N. J. Patmore, M. F. Mahon and A. S. Weller,
Organometallics, 2002, 21, 2856–2865.
101 W. L. Steffen and G. J. Palenik, Inorg. Chem., 1976, 15, 2432–
2438.
108 [Ni(dppe)Cl2], P–Ni–P = 87Њ: A. L. Spek, B. P. van Eijck, R. J. F.
Jans and G. van Koten, Acta Crystallogr., Sect. C, 1987, 43, 1878–
1880; S. C. Davies, R. A. Henderson, D. L. Hughes and K. E.
Oglieve, J. Chem. Soc., Dalton Trans., 1998, 425–431.
109 [Ni(dppp-2-Me,OMe)Br2], P–Ni–P = 96Њ: A. Muth, G. Reinhard,
G. Huttner, T. Seitz, T. Klein and L. Zsolnai, Z. Naturforsch.,
Teil B, 1994, 49, 889–897; [Cl2Ni(dppp-2-dppp)NiCl2], P–Ni–P =
94.5 and 96.8Њ: M. R. Mason, C. M. Duff, L. L. Miller, R. A.
Jacobsen and J. G. Verkade, Inorg. Chem., 1992, 31, 2746–2755;
see however [Ni(dppp-2-Me,OMe)Cl2], P–Ni–P = 85.5Њ: S.-T. Liu,
G.-J. Liu, C.-H. Yieh, M.-C. Cheng and S.-M. Peng, J. Organomet.
Chem., 1990, 387, 83–94.
110 C. Janiak, L. Braun, T. G. Scharmann and F. Girgsdies,
Acta Crystallogr., Sect. C, 1998, 54, 1722–1724.
111 A. R. Siedle, W. M. Lamanna, R. A. Newmark, J. Stevens, D. E.
Richardson and M. Ryan, Makromol. Chem., Macromol. Symp.,
1993, 66, 215.
102 S. Singh and N. K. Jha, Acta Crystallogr., Sect. C, 1995, 51, 593–
595.
103 M. R. Mason and J. G. Verkade, Organometallics, 1992, 11, 2212–
2220.
104 C. Bianchini, A. Meli, G. Müller, W. Oberhauser and E. Passaglia,
Organometallics, 2002, 21, 4965–4977.
105 P. H. M. Budzelaar, P. W. N. M. van Leeuwen and C. F. Roobeek,
Organometallics, 1992, 11, 23–25.
106 [Ni(dppe)2]: H. Hartung, U. Baumeister, B. Walther and
M. Maschmeier, Z. Anorg. Allg. Chem., 1989, 578, 177–184.
107 M. Tromp, J. A. van Bokhoven, R. J. van Haaren, G. P. F.
Strijdonck, A. M. J. van der Eerden, P. W. N. M. van Leeuwen and
D. C. Koningsberger, J. Am. Chem. Soc., 2002, 124, 14814–14815.
112 A. A. Danopoulos, J. R. Galsworthy, M. L. H. Green, S. Cafferky,
L. H. Doerrer and M. B. Hursthouse, Chem. Commun., 1998,
2529–2560.
113 H. A. Kalamarides, S. Iyer, J. Lipian, L. F. Rhodes and C. Day,
Organometallics, 2000, 19, 3983–3990.
D a l t o n T r a n s . , 2 0 0 3 , 4 4 3 7 – 4 4 5 0
4450