Bis(imino)cyclodiphosph(V)azane complexes of late transition metals: Efficient catalyst precursors for ethene and propene oligomerization and dimerization

Article abstract of DOI:10.1016/j.jcat.2005.12.008

New bis(imino)cyclodiphosph(V)azanes and their Fe, Co, and Ni complexes were synthesized and characterized. The MAO actived Ni and Co complexes exhibited high catalytic activity in ethene dimerization and oligomerization [up to 4000 kg of olefins/(mol_cath)], as well as in propene dimerization, whereas Fe derivatives yielded negligible activity. The dependence of catalytic behavior on oligomerization conditions was studied. The Ni catalysts revealed high selectivity in ethene (butene content up to 94%) and especially in propene dimerization (hexene content up to 100%). The Ni complexes catalyzed also codimerization of ethene and propene, with the yield of the resulting pentenes depending on the initial ethene/propene ratio. The main propagation route in the propene dimerization with Co catalysts was a 1,2 insertion of propene into the MH bond followed by a 2,1 insertion and β-H elimination, whereas with Ni catalysts, the reaction proceeded with equal probabilities via 1,2-1,2 and 1,2-2,1 propene insertions.

Full text of DOI:10.1016/j.jcat.2005.12.008

Journal of Catalysis 238 (2006) 196–205
www.elsevier.com/locate/jcat
Bis(imino)cyclodiphosph(V)azane complexes of late transition metals:
Efficient catalyst precursors for ethene and propene oligomerization
and dimerization
Kirill V. Axenov, Markku Leskelä, Timo Repo ∗
Department of Chemistry, Laboratory of Inorganic Chemistry, PO Box 55, University of Helsinki, FIN-00014 Finland
Received 7 September 2005; revised 8 November 2005; accepted 2 December 2005
Available online 9 January 2006
Abstract
New bis(imino)cyclodiphosph(V)azanes and their Fe, Co, and Ni complexes were synthesized and characterized. The MAO actived Ni and Co
complexes exhibited high catalytic activity in ethene dimerization and oligomerization [up to 4000 kg of olefins/(mol h)], as well as in propene
cat
dimerization, whereas Fe derivatives yielded negligible activity. The dependence of catalytic behavior on oligomerization conditions was studied.
The Ni catalysts revealed high selectivity in ethene (butene content up to 94%) and especially in propene dimerization (hexene content up to
100%). The Ni complexes catalyzed also codimerization of ethene and propene, with the yield of the resulting pentenes depending on the initial
ethene/propene ratio. The main propagation route in the propene dimerization with Co catalysts was a 1,2 insertion of propene into the M–H
bond followed by a 2,1 insertion and β-H elimination, whereas with Ni catalysts, the reaction proceeded with equal probabilities via 1,2–1,2 and
1,2–2,1 propene insertions.
© 2005 Elsevier Inc. All rights reserved.
Keywords: Homogeneous catalysis; Oligomerization; Dimerization; Late transition metal complexes; Cyclodiphosph(V)azanes
1. Introduction
Over the last decade, ligands based on the cyclodiphosph(III
or V)azane framework have been extensively studied because
of their ability to coordinate with main group [17] and tran-
sition metals [18]. In particular, bis(amido)cyclodiphosph(III)-
azane complexes of group 4 metals have been a focus of
research because they exhibit substantial catalytic activity in
homogeneous ethene polymerization after methylalumoxane
(MAO) activation [19–21]. In these complexes, the central
metal atom is coordinated to a ligand by two amido nitro-
gen atoms, and the coordination sphere of complex is often
accomplished by an additional interaction with one of the
endocyclic nitrogen atoms (Fig. 1a). Soft oxidation of the
bis(amino)cyclodiphosph(III)azans with oxygen, chalcogens,
or organic azides leads to corresponding cyclodiphosph(V)-
azane compounds, which are attractive tetradentate ligands
for the synthesis of late transition metal complexes (Fig. 1b)
[22–24].
Over the last three decades, catalytic oligomerization of
ethene has attracted considerable attention as a large-scale
source of linear higher olefins [1]. Oligomers of various sizes
are important intermediates; uses include the production of
detergents (C6–C16) and as comonomers in the production
of linear low-density polyethylene (LLDPE, C6, and C8) [2].
Since the development of the Shell higher olefins process
(SHOP), which is catalyzed by Ni P∩O-chelate complexes
[3–5], late transition metal complexes have played an essen-
tial role in catalytic oligomerization reactions [6]. To date, an
impressive number of late transition metal complexes based on
salen-type [7], cyclopentadienyl [8,9], bidentate diimine [10],
bis(imino)pyridyl [11–13], phosphinopyridine [14,15], and
chelate P∩O ligands [16] have been introduced as catalysts for
olefin oligomerization.
Phosphor(V)-imino complexes of late transition metals
based on bis(aryliminophosphoranyl)methane or bis(arylpho-
sph(V)imino)pyridine ligands also exhibit considerable cata-
*
Corresponding author. Fax: +358 9 191 50198.
E-mail address: timo.repo@helsinki.fi (T. Repo).
0021-9517/$ – see front matter © 2005 Elsevier Inc. All rights reserved.
doi:10.1016/j.jcat.2005.12.008

K.V. Axenov et al. / Journal of Catalysis 238 (2006) 196–205
197
fore use. Trimethylsilylazide (Fluka) and 1-azidoadamantane,
AdN3 (Aldrich), were used as received. A 1.6 M solution of
MeLi in Et2O was received from Aldrich. Anhydrous FeCl2,
NiBr2, and CoCl2 were received from Aldrich and stored in
Schlenk flasks under argon. Methylalumoxane (MAO, 30 wt%
solution in toluene) was received from Borealis Polymers Oy.
2,6-di-i-Propylphenylazide [25] and cis-(t BuNPCl)2 [27] (1)
were prepared according to modified literature procedures.
2.2. Synthesis of ligands
2.2.1. cis-(t BuNPMe)2 (2)
The solution of MeLi (1.6 M, 89 ml, 142.4 mmol) was
added slowly over 40 min to the cooled (−50 ◦C) solution of
(t BuNPCl)2 (1) (18.2 g, 66.2 mmol) in Et2O (30 ml). The
reaction mixture was allowed to warm and stirred overnight.
After separation of LiCl precipitation by filtration through
a glass filter, the volume of the filtrate was diminished to
ca. 30 ml by careful evaporation under low argon pressure
(100 mm Hg). The orange viscous residue was transferred via
syringe to a vacuum distillation system with a short deflegma-
tor. Because the product is thermally sensitive, it must be dis-
tilled quickly. Distillation yielded a yellowish viscosous liquid
(6.2 g, 37.5%). Bp. = 57–64 ◦C/2 mmHg (31–32 ◦C/0.5 mmHg
[28]). This compound was dissolved in toluene and used as a
toluene solution. 1H NMR (400 MHz, C6D6, 29 ◦C): δH = 1.22
(t, 18H, t Bu, P2N2 cycle), 1.62 (m, 6H, MeP). 31P{1H} NMR
(162 MHz, C6D6, 31 ◦C): δP = 173.00 (s) (171.4) (see also
supplementary material). The data resemble those obtained for
cis-(t BuNPMe)2 in previous work [28], in which the exact val-
ues of coupling constants are given. MS(EI): m/z (%) 236 (80,
M+), 220 (80, M+–Me).
Fig. 1. Cyclodiphosph(III and V)azane complexes previously described in lit-
erature.
lytic activity in ethene polymerization [25,26]. Augmenting
our previous studies with bis(amino)cyclodiphosph(III)azane
complexes, and inspired by earlier observations, herein we re-
port the preparation of cyclodihosph(V)azanes bearing bulky
phosph(V)imino groups and the synthesis of the corresponding
Fe, Co, and Ni complexes. Reactivity of the complexes after
MAO activation toward ethene and propene oligomerization
was investigated.
2. Experimental
2.1. General remarks
All manipulations were performed under an inert argon at-
mosphere using standard Schlenk techniques. The hydrocar-
bon and ether solvents were refluxed over sodium and ben-
zophenone, distilled, and stored under an inert atmosphere with
sodium flakes. CH2Cl2 was refluxed with LiAlH4, distilled,
and stored over dry molecular sieves. CDCl3 was dried over
dry molecular sieves, and CD2Cl2 was refluxed with P4O10,
distilled, and stored over dry molecular sieves in a glovebox.
C6D6 was refluxed with sodium, distilled, and stored over
dry sieves in a glovebox. Mass spectra were measured using
2.2.2. cis-[(Me3SiN)(t BuNPMe)]2 (3)
Me3SiN3 (3.64 g, 4.2 ml, 31.6 mmol) was added slowly
via syringe to the stirred solution of cis-(t BuNPMe)2 (3.15 g,
13.45 mmol) in toluene (20 ml). The reaction mixture was
warmed to 90 ◦C and maintained at this temperature overnight.
All volatiles were removed in vacuo, and the residue was treated
with hexane. After filtration, hexane was removed from the fil-
trate to give a colorless solid product (4.89 g, 95.7%). 1H NMR
(200 MHz, C6D6, 29 ◦C): δH = 0.37 (s, 18H, Me3Si), 1.37
(s, 18H, t Bu, P2N2 cycle), 1.72 (m, 6H, MeP). 13C{1H} NMR
(50.3 MHz, C6D6, 29 ◦C): δC = 3.71 (t, JPC = 1.5 Hz, CH3
of Me3Si), 24.10 (d, J1 = 5.7 Hz, CH3 of MeP), 26.24 (d,
J1 = 5.7 Hz, CH3 of MeP), 30.91 (t, JPC = 4.2 Hz, CH3,
t Bu), 53.56 (C, t Bu). 31P{1H} NMR (162 MHz, C6D6, 31 ◦C):
δP = −12.7 (s). MS(EI): m/z (%) 408 (60, M+), 337 (80, M+–
Me3Si), 235 (70, M+–2Me3Si). These characteristics (see also
supplementary material) are in accordance with those reported
for compound 3 earlier [28].
1
a JEOL SX102 spectrometer, and H and 13C NMR spectra
were recorded using a Varian Gemini 200-MHz spectrome-
1
ter. The H and 13C NMR spectra were referenced relative to
CHDCl2 (5.28 and 53.73 ppm, respectively), CHCl3 (7.24 and
77.0 ppm, respectively), and C6D5H (7.24 and 128.0 ppm, re-
1
spectively). 31P NMR and some of the H NMR spectra were
collected with a Bruker AMX 400 spectrometer. Phosphorus
signals were referenced relative to an external 85% H3PO4 so-
lution. Elemental analyses were performed at the Laboratory
of Pharmaceutical Chemistry, Department of Pharmacy, Uni-
versity of Helsinki. Gas chromatography–mass spectroscopy
(GC–MS) of polymerization products was performed with an
Agilent Technologies 6890N Network GC System.
tert-Butylamine and phosphorus trichloride were purchased
from Merck and purified by distillation under argon (t BuNH2
over sodium hydroxide). 2,6-di-i-Propylaniline was received
from Aldrich and distilled in vacuo over sodium hydroxide be-
2.2.3. cis-[(2,6-iPr2C6H3N)(t BuNPMe)]2 (4)
A solution of raw 2,6-iPr2C6H3N3 (2.25 g, 11.1 mmol) in
toluene (10 ml) was added slowly via syringe to the stirred so-
lution of cis-(t BuNPMe)2 (1.00 g, 4.3 mmol) in toluene (10 ml).
The reaction mixture was warmed to 90 ◦C and maintained

198
K.V. Axenov et al. / Journal of Catalysis 238 (2006) 196–205
at this temperature overnight. All volatiles were removed in
vacuo, and the red oil residue was dissolved in hexane. The
volume of solution was diminished to ca. 5 ml, and this mix-
ture was kept at −50 ◦C. After 2 days, the solid product was
precipitated and isolated as a reddish solid (1.12 g, 44.6%).
C34H58N4P2 calculated: C, 69.83; H, 10.00; N, 9.58; found: C,
was carried out as reported for [(Me3SiN)(t BuNPMe)]2FeCl2
(6). Co complex was isolated as a blue solid (1.0 g, 74%).
(C16H42Cl2N4P2Si2Co calculated: C, 35.86; H, 7.64; N, 10.43;
found: C, 35.69; H, 7.86; N, 10.40). MS(EI): m/z (%) 536 (3,
M+), 408 (20, ligand).
1
70.07; H, 9.88; N, 9.31). H NMR (200 MHz, C6D6, 29 ◦C):
2.3.3. [(Me3SiN)(t BuNPMe)]2NiBr2 (8)
t
δH = 1.19 (s, 18H, Bu, P2N2 cycle), 1.37 (d, 24H, CH3 of
Solid NiBr2 (1.47 g, 6.7 mmol) and cis-[(t BuNPMe)2
(NSiMe3)2] (3) (1.37 g, 3.4 mmol) were treated in THF (40 ml)
as described above, and separation of the desired product was
carried out as reported for 6. Ni complex was isolated as a green
solid (1.0 g, 47%). (C16H42Br2N4P2Si2Ni calculated: C, 30.87;
H, 6.86; N, 8.95; found: C, 30.64; H, 6.75; N, 8.93). MS(EI):
m/z (%) 627 (1.2, M+), 468 (5, M+–2Br), 408 (30, ligand).
i
iPr), 1.99 (m, 6H, MeP), 3.66 (m, 4H, CH of Pr), 7.11 (m,
4H, Ar), 7.26 (d, 2H, Ar). 13C{1H} NMR (50.3 MHz, C6D6,
i
29 ◦C): δC = 22.60 (d, CH3 of MeP), 24.44 (CH3, Pr), 25.30
i
(d, CH3 of MeP), 28.78 (CH, Pr), 30.68 (t, JPC = 4.2 Hz,
t
t
CH3, Bu), 53.37 (C, Bu), 120.84 (Ar), 123.24 (Ar), 140.43
(t, J = 1.9 Hz, Ar), 141.46 (t, J = 3.4 Hz, Ar). 31P{1H} NMR
(162 MHz, C6D6, 31 ◦C): δP = −33.07 (s). MS(EI): m/z (%)
584 (60, M+), 409 (80, M+–ArN), 235 (80, M+–2ArN).
2.3.4. [(2,6-iPr2C6H3N)(t BuNPMe)]2FeCl2 (9)
2.2.4. cis-[(AdN)(t BuNPMe)]2 (5)
Solid FeCl2 (0.40 g, 3.2 mmol) was added to the stirred
solution of cis-[(t BuNPMe)2(2,6-iPr2C6H3N)2] (4) (1.64 g,
2.8 mmol) in 30 ml of THF, and the reaction mixture was stirred
at an ambient temperature for 2 days. After drying under vac-
uum, the residue was washed several times with hexane and dis-
solved in CH2Cl2. Hexane was added until the unreacted gray
FeCl2 began to precipitate. After filtration, all volatile compo-
nents were removed in vacuo to yield a raw product. This pu-
rification procedure was repeated several times to finally yield
a dark-violet solid product (1.20 g, 60%). (C34H58Cl2N4P2Fe
calculated: C, 57.39; H, 8.22; N, 7.87; found: C, 57.87; H, 8.27;
N, 7.66). 1H NMR (200 MHz, CD2Cl2, 29 ◦C): δH = 1.17 (br.
A solution of AdN3 (3.33 g, 18.8 mmol) in toluene (10 ml)
was added slowly via syringe to the stirred solution of cis-
(t BuNPMe)2 (2.00 g, 8.55 mmol) in toluene (20 ml). The yel-
low reaction mixture was warmed to 110 ◦C and maintained at
this temperature overnight. The resulting mixture was filtrated,
and all volatiles were removed from the filtrate in vacuo to yield
a yellow sticky solid product (3.67 g, 80.7%). (C30H54N4P2
calculated: C, 67.64; H, 10.22; N, 10.52; found: C, 67.19; H,
10.37; N, 10.79). 1H NMR (200 MHz, C6D6, 29 ◦C): δH = 1.38
(s, 18H, t Bu, P2N2 cycle), 1.75 (m, 6H, MeP), 1.77 (br. m, 12H,
CH2 of Ad), 2.16 (br. m, 12H, CH2 of Ad), 2.25 (br. m, 6H, CH
of Ad). 13C{1H} NMR (50.3 MHz, C6D6, 29 ◦C): δC = 21.50
(d, CH3 of MeP), 23.50 (d, CH3 of MeP), 30.18 (CH, Ad),
31.68 (dd, J1 = 3.8 Hz, J2 = 1.5 Hz, CH3, t Bu), 37.32 (CH2,
i
d, 24H, CH3 of Pr), 1.31 (s, 18H, t Bu, P2N2 cycle), 2.17 (br.
m, 6H, MeP), 3.50 (br. s, 4H, CH of iPr), 6.77 (br. s, 2H, Ar),
6.98 (br. d, 4H, J = 6.6 Hz, Ar). 13C{1H} NMR (50.3 MHz,
CD2Cl2, 29 ◦C): δC = 21.76 (s, CH3 of MeP), 22.92 (CH3,
t
Ad), 42.88 (CH2, Ad), 51.54 (d, JPC = 15.26 Hz, C, Bu),
52.71 (dd, J1 = 8.4 Hz, J2 = 1.1 Hz, CN, Ad). 31P{1H} NMR
(162 MHz, C6D6, 31 ◦C): δP = 46.2 (s). MS(EI): m/z (%) 532
(50, M+), 383 (75, M+–AdN), 235 (80, M+–2AdN).
i
t
iPr), 27.39 (CH, Pr), 29.59 (t, JPC = 4.2 Hz, CH3, Bu),
118.74 (Ar), 121.66 (Ar), 139.28 (m, Ar), 140.45 (m, Ar).
31P{1H} NMR (162 MHz, CD2Cl2, 31 ◦C): δP = −31.88 (s).
MS(EI): m/z (%) 675 (5, M+–Cl), 641 (36, M+–2Cl), 584 (80,
ligand).
2.3. Synthesis of complexes
2.3.1. [(Me3SiN)(t BuNPMe)]2FeCl2 (6)
2.3.5. [(2,6-iPr2C6H3N)(t BuNPMe)]2CoCl2 (10)
Solid FeCl2 (0.65 g, 5.1 mmol) was added to the stirred so-
lution of cis-[(t BuNPMe)2(NSiMe3)2] (3) (1.00 g, 2.45 mmol)
in 30 ml of THF, and the reaction mixture was refluxed for
2 days. After drying under vacuum, the residue was washed sev-
eral times with hexane and dissolved in CH2Cl2. Hexane was
added until the unreacted gray FeCl2 began to precipitate. Af-
ter filtration, all volatile components were removed in vacuo to
yield a raw product. This purification procedure was repeated
several times to finally yield a green solid product (0.58 g,
44.3%). (C16H42Cl2N4P2Si2Fe calculated: C, 47.03; H, 10.36;
N, 13.71; found: C, 47.30; H, 10.62; N, 13.78). MS(EI): m/z
(%) 500 (2, M+–Cl), 408 (20, ligand).
Solid CoCl2 (0.42 g, 3.2 mmol) and cis-[(t BuNPMe)2(2,6-
iPr2C6H3N)2] (4) (1.64 g, 2.8 mmol) were treated in THF
(30 ml) as described for [(2,6-iPr2C6H3N)(t BuNPMe)]2FeCl2
(9), and separation of the desired product was carried out as re-
ported for 6. Co complex was isolated as a green solid (1.57 g,
78.5%). (C34H58Cl2N4P2Co calculated: C, 57.14; H, 8.18; N,
7.84; found: C, 56.82; H, 8.16; N, 7.86.) 1H NMR (200 MHz,
CD2Cl2, 29 ◦C): δH = 1.20 (d, 24H, CH3 of iPr), 1.34 (s, 18H,
t Bu, P2N2 cycle), 2.19 (m, 6H, MeP), 3.53 (m, 4H, CH of
iPr), 6.79 (m, 2H, Ar), 7.00 (d, 4H, J = 7.3 Hz, Ar). 13C{1H}
NMR (50.3 MHz, CD2Cl2, 29 ◦C): δC = 21.86 (s, CH3 of
i
i
MeP), 23.03 (CH3, Pr), 27.53 (CH, Pr), 29.71 (t, JPC
=
4.2 Hz, CH3, t Bu), 118.62 (m, Ar), 121.80 (d, J = 3.4 Hz, Ar),
139.258 (m, Ar), 140.64 (m, Ar). 31P{1H} NMR (162 MHz,
CD2Cl2, 31 ◦C): δP = −31.78 (s). MS(EI): m/z (%) 679 (2,
M+–Cl), 640 (8, M+–2Cl), 584 (80, ligand).
2.3.2. [(Me3SiN)(t BuNPMe)]2CoCl2 (7)
Solid CoCl2 (0.35 g, 2.7 mmol) and cis-[(t BuNPMe)2
(NSiMe3)2] (3) (1.0 g, 2.5 mmol) were treated in THF (30 ml)
as described above, and separation of the desired product

K.V. Axenov et al. / Journal of Catalysis 238 (2006) 196–205
199
2.3.6. [(2,6-iPr2C6H3N)(t BuNPMe)]2NiBr2 (11)
ods for their preparation is the Staudinger reaction [29]. This
reaction is the key to the many interesting structures, espe-
cially those with inorganic chelating P=N groups [30]. To pro-
vide effective metal coordination, the chelating phosphor(V)-
imino groups must be in cis-orientation. However, all cy-
clodiphosph(III or V)azanes cannot adopt the thermodynami-
cally stable cis-configuration, but undergo cis–trans isomeriza-
tion and appear as inseparable equilibrium mixtures of cis and
trans isomers [17]. Only the cis-cyclodiphosph(III or V)azanes
Solid NiBr2 (0.92 g, 4.2 mmol) was added to the stirred
solution of cis-[(t BuNPMe)2(2,6-iPr2C6H3N)2] (4) (1.64 g,
2.8 mmol) in 30 ml of THF, and the reaction mixture was re-
fluxed for 2 days. Separation of the desired product was carried
out as reported for 6. Ni complex was isolated as a green-
ish solid (1.65 g, 73.4%). (C34H58Br2N4P2Ni calculated: C,
50.84; H, 7.28; found: C, 51.21; H, 7.58.) 1H NMR (200 MHz,
CD2Cl2, 29 ◦C): δH = 1.16 (d, 24H, CH3 of iPr), 1.30 (s, 18H,
t Bu, P2N2 cycle), 2.16 (m, 6H, MeP), 3.48 (m, 4H, CH of iPr),
6.75 (m, 2H, Ar), 6.96 (d, 4H, J = 7.7 Hz, Ar). 31P{1H} NMR
(162 MHz, CD2Cl2, 31 ◦C): δP = −31.65 (s). MS(EI): m/z (%)
802 (1, M+), 722 (2, M+–Br), 641 (16, M+–2Cl), 584 (80, lig-
and).
t
with small substituents at the phosphorus (e.g., Me) and Bu
groups at the nitrogen atoms do not undergo such an iso-
merization process [31]. One of these stable compounds, cis-
(ClPNt Bu)2 (1), was chosen as a starting material for our
study. After methylation of 1 by MeLi, the corresponding cis-
(MePNt Bu)2 (2) was isolated with moderate yield (Scheme 1);
the substance was sensitive to the harsh temperatures and partly
decomposed during the distillation. In 1H NMR, methyl groups
at phosphorus(III) show specific spin–spin coupling with phos-
phorus (A3XXꢀAꢀ3 spin system) [28].
After purification, 2 was introduced into the Staudinger re-
action with three different organic azides (Scheme 1). Two of
these azides, Me3SiN3 and AdN3 (Ad – 1-adamantyl), were
chosen because they were commercially available and relatively
stable substances. The preparation of the third azide, 2,6-di-i-
propylphenylazide (2,6-iPr2C6H3N3), has been described pre-
viously; this substance can be simply obtained from the corre-
sponding aniline [25]. In addition, it was assumed that sterically
bulky substituents, such as Me3Si, 1-adamantyl, and 2,6-di-i-
propylphenyl, would be beneficial to the catalytic activity of
the late transition metal complexes.
2.3.7. [(AdN)(t BuNPMe)]2FeCl2 (12)
Solid FeCl2 (0.4 g, 3.2 mmol) and cis-[(t BuNPMe)2(AdN)2]
(5) (1.63 g, 3.1 mmol) were treated in THF (30 ml). Reac-
tion and separation of the desired product were carried out
as reported for 6. Fe complex was isolated as a brown solid
(1.20 g, 57.8%). (C30H54Cl2N4P2Fe calculated: C, 54.64; H,
8.25; found: C, 54.61; H, 8.67). MS(EI): m/z (%) 657 (4, M+),
622 (10, M+–Cl), 532 (54, ligand).
2.3.8. [(AdN)(t BuNPMe)]2CoCl2 (13)
Solid CoCl2 (0.4 g, 3.1 mmol) and cis-[(t BuNPMe)2(AdN)2]
(5) (1.63 g, 3.1 mmol) were treated in THF (30 ml). Reac-
tion and separation of the desired product were carried out
as reported for 6. Co complex was isolated as a blue solid
(1.23 g, 60.6%). (C30H54Cl2N4P2Co calculated: C, 54.38; H,
8.21; found: C, 53.81; H, 8.17). MS(EI): m/z (%) 662 (6, M+),
626 (14, M+–Cl), 532 (75, ligand).
After treatment of (t BuNPMe)2 (2) with Me3SiN3 at 90 ◦C
for 1 day, the 31P NMR spectrum of the raw product re-
vealed a marked upfield shift of the phosphorus signal [from
δP = 173.00 ppm for (t BuNPMe)2 to δP = −12.7 ppm], cor-
roborating the formation of phosphorus(V) species. In addi-
tion, a small amount of the trans isomer was recognizable
from the 31P NMR data (δP = −2.86 ppm). Purification of
the raw product with hexane extraction gave the desired com-
pound, [(t BuNPMe)2(NSiMe3)2] (3), with high yield. The bulk
2.4. Polymerization experiments
A 1-L steel autoclave was charged with 200 ml of toluene
and cocatalyst (MAO), thermostated at the temperature required
for the experiment, saturated with ethene or propene, and sup-
plemented with the desired amount of precatalyst solution.
The monomer pressure ( 50 mbar) and temperature ( 0.5 ◦C)
were kept constant during each polymerization run. Monomer
consumption, polymerization temperature, and pressure were
controlled by real-time monitoring. The polymerizations were
quenched by opening the system to air, and a set amount (2 ml)
of cyclooctane was added as a GC standard to the solution.
Probes from the mixture were filtrated through a layer of Al2O3
to remove Al-containing species and catalyst decomposition
products, then analyzed by GC methods.
1
composition and purity of the product were confirmed by H
NMR and 13C NMR data and by elemental analysis [28].
The Staudinger reaction between (t BuNPMe)2 and AdN3 or
2,6-iPr2C6H3N3 gave cis-[(t BuNPMe)2(AdN)2] (4) and cis-
[(t BuNPMe)2(2,6-iPr2C6H3N)2] (5), respectively (Scheme 1).
The specific spin–spin coupling of MeP protons in 1H NMR
spectra was observed for 3–5 (see supplemental material). Ev-
1
idently, such coupling in H NMR indicates the presence of a
MeP(V)2N2 cycle and can be used as a characteristic finger-
print for these phosphazene systems. The solid-state structure
of 3 was revealed by single-crystal X-ray diffraction studies
(see supplemental material). However, the crystals obtained for
X-ray measurements consisted of 3 in trans configuration. This
can be explained by contamination of the product by a trans
isomer (see above), which seems to crystallize better than the
cis isomer.
3. Results
3.1. Ligand synthesis
Compounds containing a phosphorus–nitrogen double bond
are well known, and one of the most effective synthetic meth-

200
K.V. Axenov et al. / Journal of Catalysis 238 (2006) 196–205
Scheme 1.
3.2. Complex synthesis
13C NMR data revealed C2v symmetry for Fe, Co, and Ni com-
plexes (9–11) in solution. The same specific spin–spin coupling
of phosphorus and PMe protons as in the parent ligand 4 was
also observable in 1H NMR spectra of the complexes (see sup-
plemental material).
Treatment of the bis(imino)cyclodiphosph(V)azanes 3–5
with anhydrous Fe(II), Co(II) dichlorides, or NiBr2 yielded the
metal complexes 6–13 (Scheme 1). Ligands 3–5 displayed only
moderate ability to chelate with the metal precursors, indicated
by moderate yields; high temperatures, long reaction times, and
an excess of metal halides were needed for the complex prepa-
ration. To increase the yields of the complexes with ligands 3
and 5, the reactions were performed in refluxing THF. Unfortu-
nately, under these conditions, ligand 4 gave only unidentified
products; thus, the complexations with 4 were carried out at
ambient temperature. Of the metal precursors, NiBr2 revealed
the poorest propensity to form complexes, and all attempts to
prepare a Ni complex with 1-adamantyl substituted ligand 5
failed.
3.3. Ethene oligomerization
On activation by MAO in toluene, the Ni and Co complexes
revealed high catalytic activity in ethene oligomerization, typ-
ically in the range of 700–1600 kg/(molcat h) (Table 1). When
the reactions were terminated by addition of acidified methanol
to the resulting toluene phase, only a faint precipitation of inor-
ganic material originating from MAO appeared. GC–MS data
from samples, filtered through Al2O3, revealed the presence of
a mixture of olefins (from butenes to C18–C20 fractions) (Ta-
ble 1). The quantities of C12–C20 oligomers were calculated
directly from the GS data, which were calibrated with cyclooc-
tane. The α value (characteristic for Schulz–Flory oligomer
distribution) [32,33] was determined from the molar ratio of
C16/C14, and the mol% of lower (C4–C10) alkenes in the prod-
uct mixture was calculated by backward extrapolation from C12
on the basis of the α value [34].
The central metal atom in MAO-activated bis(imino)cyclo-
diphosph(V)azanes complexes has a significant influence on the
product distribution in ethene oligomerization. After MAO ac-
tivation, the nickel complexes gave mixtures with a high (up
to 94%) proportion of butenes together with hexenes and a
recognizable C10 fraction (<1%), whereas the cobalt catalysts
were prone to produce C4–C20 oligomers with Schulz–Flory
In general, purification of the complexes was straightfor-
ward. THF was evaporated from the reaction mixture, and
unreacted ligand was removed by hexane extraction, after
which an excess of metal halides was precipitated from the
dichlorometane solution of the complex by adding hexane.
After filtration, these solvents were removed in vacuum, and
target complexes 6–13 were isolated as colored powders. The
composition was confirmed by elemental analysis and mass
spectroscopy data. As expected, the complexes bearing ligands
3 and 5 appeared to be paramagnetic, and their NMR investiga-
tions failed. Surprisingly, for the Fe, Co, and Ni derivatives of
1
cis-[(2,6-Pri2C6H3N)(ButNPMe)]2, reliable H, 13C, and 31P
NMR data were recorded. In the 31P NMR spectra, only sig-
1
nals related to the desired complexes were found, and H and

K.V. Axenov et al. / Journal of Catalysis 238 (2006) 196–205
201
Table 1
a
Ethylene olygomerization and dimerization results
b
c
Entry
no.
Catalyst
Product
yield (g)
α
Activity
C
(wt%)
C
(wt%)
C
C
(wt%)
C
(wt%)
C
(wt%)
C
(wt%)
C
18
(wt%)
4
6
8
10
12
14
16
(wt%)
1
2
3
4
5
7
8
10
11
13
7.75
7.95
7.54
5.60
3.56
0.75
–
0.66
–
1550
1590
1510
1120
720
14
89
19
94
71
15
10
19
6
15
–
17
–
14
1
14
–
13
–
11
–
12
–
8
10
–
7
–
5
7
–
5
–
–
–
–
10
–
5
4
5
a
◦
Conditions: MAO (Al/M = 1000), 200 ml of toluene, 8 bar of ethylene (under these conditions [C H ] is ∼1 mol/l), reaction temperature 30 C.
2
4
b
c
In kg
/(mol h [C H ]).
cat
2 4
olefins
Calculated based on α value.
distribution (see the GC data in supplemental material). Quite
unexpectedly, the iron complexes displayed very low catalytic
activity.
Table 2
Influence of the oligomerization temperature on the catalytic behaviour of the
complexes
a,b
c
c
Entry Catalyst Yield of
no.
α
T
Activity
C
C
6
Stability of the cyclodiphosph(V)azane catalysts in the
oligomerization reactions also depended on the nature of the
central atom. At the start of oligomerization, the Co catalysts
were highly active, but they deactivated rapidly, manifesting as
a decline in ethene consumption during the reaction. The initial
activity of the Ni catalysts was lower than that of the Co cata-
lysts, but activity remained nearly constant during the oligomer-
ization experiments. The Co and Ni catalysts thus seem to
provide similar average productivity in ethene oligomerization
during the first 30 min (Table 1).
The substitution pattern at the ligand moiety also had an im-
portant role in defining the catalytic activity of the complexes.
Co catalysts 7 and 10 bearing Me3Si and bulky 2,6-i-Pr2C6H3,
respectively, exhibited similar catalytic activities, whereas 13
bearing the 1-adamantyl substituted ligand gave 50% lower ac-
tivity values under similar reaction conditions (Table 1).
The influence of reaction temperature and MAO/M ratio on
activity and catalyst selectivity was also investigated for Ni and
Co complexes (Tables 2 and 3). The catalysts revealed high
sensitivity to the oligomerization temperature (Table 2); ex-
periments conducted at above 50 ◦C gave only low activities,
and oligomerization was terminated above 70 ◦C. Despite the
decrease in catalytic activity with increasing oligomerization
temperature, the selectivity of the catalysts remained similar.
In ethene dimerization with the Ni catalysts, the proportion of
the C4 fraction increased slightly with increasing temperature,
whereas with the Co catalysts, the α value for the Shulz–Flory
distribution rose marginally (Table 2), indicating an increase in
the amount of C12–C16 fractions.
4
◦
oligomers (g)
( C)
(wt%) (wt%)
∗
1
2
3
4
5
6
7
8
9
10
11
12
13
14
7
7
7
8
8
8.14
7.75
4.67
2.86
7.95
7.06
3.97
7.54
3.11
1.62
5.60
2.18
3.56
2.09
0.72 20
0.75 30
0.75 40
–
–
–
0.63 20
0.66 30
0.68 40
–
–
–
–
–
4180
–
–
–
87
89
91
–
–
–
92
94
27
71
51
–
–
–
13
10
9
–
–
1550
1050
1470
∗
20
30
40
1590
1420
2035
8
∗
10
10
10
11
11
13
13
13
1510
700
830
–
∗
20
30
20
30
40
8
6
11
10
16
1120
1120
∗
720
470
a
b
In kg
/(mol h [C H ]).
cat
olefins
2 4
Conditions: MAO (Al/M = 1000), 200 ml of toluene, 8 bar of ethylene
∗
( , ethene pressure was 3 bar).
c
Only for Ni catalysts (for Co complexes, α value is present).
Table 3
Influence of MAO concentration on catalytic behaviour of complexes
a,b
c
c
Entry Catalyst Yield of
no.
α
MAO/M Activity
C
C
6
4
oligomers (g)
(wt%) (wt%)
1
2
3
4
5
6
7
8
7
7
8
9.13
7.75
5.71
7.96
7.16
7.54
5.44
5.60
0.75 500
0.75 1000
1830
1550
1140
1590
1430
1510
1090
1120
–
–
–
–
10
10
–
–
–
500
1000
90
89
–
8
10
10
11
11
0.70 500
0.66 1000
–
–
–
–
500
1000
93
94
7
6
Changes in MAO concentrations did not strongly affect the
catalytic activity of the Ni and Co complexes or the selectiv-
ity of ethene dimerization by Ni catalysts. However, higher
MAO/Co ratios increased the amount of the C4 fraction, in-
dicating an increased probability of a termination reaction (Ta-
ble 3).
a
In kg
/(mol h [C H ]).
cat
olefins
2 4
b
◦
Conditions: oligomerization temperature 30 C, 200 ml of toluene, 8 bar of
ethylene.
c
Only for Ni catalysts (for Co complexes, α value is present).
exhibited signals of C4 olefins, toluene (from MAO solution),
and a small amount of C6 fraction (Fig. 2). Integration of the
spectra revealed a butenes-2/butene-1 ratio of 1.6–1.7:1. trans-
Butene-2 was the dominating butene-2 isomer in the product
mixture, with a ratio of 1.4–1.7:1. These observations are in
agreement with the values reported for enthalpies of formation
for various butenes (Table 6); such thermodynamic control in
The Ni catalysts produced a mixture of olefins with a very
high quantity of C4 alkenes (>90%). Detailed NMR stud-
ies were performed to determine the ratio between different
butene isomers. Dimerization experiments were carried out in
deuterated benzene at 30 ◦C and 8 bar ethene pressure, and fil-
tered samples from the resulting solutions were analyzed by 1H
NMR. These results confirmed the GC data, because the spectra

202
K.V. Axenov et al. / Journal of Catalysis 238 (2006) 196–205
1
t
Fig. 2. H NMR spectrum of product mixture obtained in ethene dimerization experiments in C D where [(Me SiN)( BuNPMe)] NiBr /MAO (8) was used as a
6
6
3
2
2
catalyst.
Table 4
Propene dimerization results
a
Entry Catalyst Mass of the
T
MAO/M Propene
pressure
Activity Hexene-1 Hexene-2
4-methyl- 4-methyl-2-
1-pentene pentene and
C
total
C
C
12
6
9
◦
no.
products (g)
( C)
(%)
and
(wt%) (wt%)
(bar)
hexene-3
(%)
(%)
2,3-dimethyl-
1-butene (%)
(wt%)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
7
7
8
4.2
0.4
11.3
3.3
2.1
1.5
18.0
5.8
3.2
25.4
1.4
19.6
1.6
20
40
20
40
20
40
20
40
20
20
20
20
20
20
20
1000
1000
1000
1000
1000
1000
1000
1000
500
500
500
500
1000
1000
1000
7
7
7
7
7
7
7
7
7
7
7
7
3.5
3.5
3.5
54.2
15.8
143.9
137.5
27.4
13
16
5
71
65
42
39
65
62
53
41
74
43
94
32
63
35
42
7
5
10
13
5
5
9
12
4
9
1
12
5
10
8
9
14
43
40
10
11
33
39
7
43
1
48
10
49
44
83
71
∼100
∼100
65
13
21
4
8
<1 <1
<1 <1
24 11
14 11
<1 <1
<1 <1
11
<1 <1
11
<1 <1
27 21
<1 <1
<1 <1
8
8
10
10
11
11
7
20
22
5
8
15
5
4
8
22
6
6
62.5
75
229.7
240.3
41.5
323.5
17.9
250.8
113.1
328.6
657.2
∼100
∼100
88
1
8
∼100
10
11
7
8
11
80
9
∼100
52
4.6
9.3
∼100
∼100
a
◦
◦
◦
kg
/(mol h [C H ]); at 20 C and 7 bar, [C H ] = 15.67 mol/l; 40 C and 7 bar, [C H ] = 4.80 mol/l; 20 C and 7 bar, [C H ] = 2.83 mol/l.
cat
olefins
3
6
3
6
3
6
3 6
butene distribution can be the result of a fast and reversible iso-
merization reaction, catalyzed by Ni–H species (see below).
fines, whereas the Ni catalysts produced mainly C6 alkenes with
traces of higher olefins (see the GC data in supplemental mate-
rial).
3.4. Propene dimerization
NMR and GC methods were used to analyze the proportions
of the hexene isomers in the C6 fraction in detail. Consider-
ing all possible combinations for propene dimerization, nine
different hexenes (12 with cis/trans isomers) can be expected
(Scheme 3). The propene dimerization reaction with Ni catalyst
10 was carried out in deuterobenzene, after which the reaction
solution was refluxed for a long time to remove propene, which
is very soluble in toluene and benzene. 1H and 13C NMR spec-
tra of Al2O3-filtrated samples were measured and compared
Propene oligomerization properties of the Ni and Co cat-
alysts were evaluated by applying standard reaction condi-
tions: 20 ◦C and 7 bar propene pressure. Under these con-
ditions, the Co catalysts showed low catalytic activity [70–
100 kg/(molcat h)], whereas the Ni catalysts were highly active
[up to 3600 kg/(molcat h)] (Table 4). The resulting toluene so-
lutions were analyzed by GC–MS methods as described above.
The samples produced by Co complexes included dissolved
propene and C6 fraction with admixtures of C9 and C12 ole-
1
with H and 13C NMR spectra of pure alkenes available from
the Aldrich NMR Library. Based on this comparison, six differ-

K.V. Axenov et al. / Journal of Catalysis 238 (2006) 196–205
203
ent hexenes were identified: hexene-1, hexenes-2, hexenes-3,
Table 5
Propene ethene codimerization results
4-methyl-1-pentene, 4-methyl-2-pentene, and 2,3-dimethyl-1-
1
butene. To verify these results, H and 13C NMR spectra of
a
Entry Catalyst Ethene
no.
Ethene
Propene Activity C /C C /C
6
5
6
5
consump- pressure pressure
ratio
ratio
pure hexenes (purchased from Aldrich) and their mixtures with
appropriate compositions were recorded and used for analysis
of propene dimerization products.
tion (l)
(bar)
(bar)
(mol) (wt)
1
2
3
4
5
6
8
8
8
11
11
11
2.43
1.98
1.46
2.87
1.99
1.23
3
2
1
3
2
1
1
2
3
1
2
3
640
520
390
760
525
325
1.2
1.0
0.75
0.35
0.4
1.4
1.2
0.9
0.4
0.5
0.8
In the original GC chromatograms, five signals correspond-
ing to C6 olefins were detected. To ensure reliable data on
product composition in the reaction mixture, these results were
compared with the GC data obtained for pure alkenes and their
combinations. With the applied GC method, the cis/trans iso-
mers were noted to have almost the same retention times, ex-
cluding cis- and trans-2-hexenes. This information was then
used to calculate the quantities of different hexenes in the prod-
ucts (Table 4). Summarizing the results, 1-hexene content in the
mixtures remained <20% for the Ni and Co catalysts. The Co
catalysts produced hexene-2 and hexene-3 as the main prod-
ucts, whereas the Ni catalysts produced mixtures of olefins
with two main fractions: hexene-2/hexene-3 (the result of 1,2–
2,1 propene insertions) and 4-methyl-2-pentene/2,3-dimethyl-
1-butene (the result of 1,2–1,2 propene insertions, Scheme 3)
(Table 4).
0.7
a
Calculated from ethene consumption.
The influence of oligomerization conditions on catalysis
(dimerization activity and content of the resulting olefin mix-
tures) was studied with the Ni complexes. The catalysts were
found to be sensitive to reaction temperature and excessive
MAO, and the best activities were achieved at low poly-
merization temperatures and MAO concentrations and high
propene pressure (Table 4). In all experiments, the content
of hexene-1 and 4-methyl-1-pentene in products remained
constant, whereas the hexene-2/hexene-3 fraction decreased
slightly with increasing temperature and propene pressure (Ta-
ble 4). Only C6 olefins with very small traces of higher fractions
were recognized in GC–MS experiments.
Scheme 2.
acterized. In the olefin oligomerization experiments, these com-
plexes exhibited behavior similar to that of known late transi-
tion metal catalysts (e.g., those used in SHOP). The complex
species containing M–H bonds are widely considered to be the
catalytic sites for the oligomerization reaction. They are thought
to be formed via β-H elimination from the intermediate metal
alkyl complexes (Scheme 2) [35,36].
To draw a full analogy with the accepted mechanism
(Scheme 2), Co and Ni hydride species are proposed to be active
intermediates in Co and Ni bis(imino)cyclodiphosph(V)azane-
catalyzed olefin oligomerization processes. Complex activation
by MAO can lead to methyl-substituted cationic species, which
can be transformed into cationic Co and Ni hydride species af-
ter one or two ethene insertions followed by elimination.
The nature of the central metal in catalyst species plays a sig-
nificant role in defining the catalytic behavior of the complex,
particularly the composition of the resulting olefin mixture. In
ethene oligomerization, the bis(imino)cyclodiphosph(V)azane
Co complexes produced olefins with the Schulz–Flory distrib-
ution from C4 to C20, but mixtures of alkenes with high (up
to 95%) ethene dimerization products were obtained with Ni
catalysts. Dimerization can be considered a particular case of
oligomerization in which the termination reaction (β-hydrogen
3.5. Ethene and propene codimerization
As described above, bis(imino)cyclodiphosph(V)azane Ni
complexes catalyze selective ethene or propene dimerization re-
actions. The mixtures of olefins obtained with these catalysts
had high proportions (up to 90–100%) of butenes or hexenes.
Based on these results, possible codimerization of ethene and
propene was expected. The codimerization experiments were
carried out in toluene at 30 ◦C. The set amounts of propene
were introduced into the autoclave, which was then filled with
ethene up to 4 bar total pressure. The resulting olefin mixtures
were investigated by GC methods.
Appearance of a new fraction between C4 and C6 olefins was
detected (see the GC data in supplemental material). The con-
tent of the C5 olefins in the resulting mixtures depended on the
initial ratio of the monomers (Table 5). Propene was observed
in all mixtures.
4. Discussion
In this study, the bis(imino)cyclodiphosph(V)azane com-
plexes of the late transition metals were synthesized and char-

204
K.V. Axenov et al. / Journal of Catalysis 238 (2006) 196–205
transfer to the metal) is much faster than olefin insertion into a
metal–alkyl bond (Scheme 2) [37]. Ni alkyl species apparently
revealed much higher instability toward β-H elimination than
the analogous Co intermediates, which caused the selectivity of
Ni catalysts in ethene dimerization.
In propene oligomerization, both the Ni and Co complexes
produced mixtures of alkenes with high proportions (up to
100%) of dimerization products (C6 olefins). The selective
propene dimerization by Co catalysts compared with their be-
havior in ethene oligomerization indicates that the β-H transfer
to cobalt and ethene insertion into the Co–alkyl bond occur
more rapidly than the propene insertion. This means that in the
case of the Co catalysts, ethene insertion is a clearly more fa-
vored process than propene insertion.
Table 6
Thermodynamic data for butenes and hexenes
◦
Entry no.
Compounds
ꢀH (g) at 298 K (kJ/mol)
f
1
2
3
4
5
6
7
8
9
1-Butene
cis-2-Butene
trans-2-Butene
1-Hexene
cis-2-Hexene
−0.63
−7.70
−10.80
−42.09
−52.77
−54.32
−48.04
−54.87
−51.64
−57.89
−61.91
−66.77
trans-2-Hexene
cis-3-Hexene
trans-3-Hexene
4-Methyl-1-pentene
cis-4-Methyl-2-pentene
trans-4-Methyl-2-pentene
2,3-Dimethyl-1-butene
10
11
12
The species containing the metal–hydrogen bonds, which
are responsible for alkene oligomerization, are generally con-
sidered to also catalyze the isomerization of the produced olefin
oligomers through a reversible addition/elimination to the C–C
bond [35,36]. In ethene and propene oligomerizations with
bis(imino)cyclodiphosph(V)azane Co and Ni catalysts, this re-
sults in the formation of 1- and 2-butenes and hexenes with
different double-bond positions.
The higher content of linear 2- and 3-hexenes compared
with 1-hexene in propene dimerization products (Table 4) can
be explained by a fast and reversible isomerization of hexene-
1, leading to the thermodynamically more preferable hexene-2
and hexene-3 (Table 6).
According to Scheme 3, hexene-1 can be formed as a result
of 1,2 propene insertion into the M–H bond, followed by a 2,1
insertion. With Co catalysts, this dimerization route is favored,
as revealed by the ratio of linear/branched hexenes in the prod-
uct mixtures (Table 4). In contrast, the probabilities for a 1,2
propene insertion followed by a 1,2 or a 2,1 insertion are al-
most equal in propene dimerization catalyzed by Ni complexes
(Table 4). They produced mixtures of olefins with two main
fractions: hexene-2/hexene-3 (result of 1,2–2,1 propene inser-
tions) and 4-methyl-2-pentene/2,3-dimethyl-1-butene (result of
1,2–1,2 propene insertions, Scheme 3). In this case, the hex-
ene isomerization process also plays a major role in forming
the thermodynamically preferred hexene-2, hexene-3, and 4-
methyl-2-pentene. 2,3-Dimethyl-1-butene can be the result of
a 1,2 propene insertion into the iso-propyl Ni species formed
by a Markovnikoff addition of Ni–H particles to the propene
(2,1 propene insertion into Ni–H bond) (Scheme 3).
The ethene–propene codimerization experiments with Ni
catalysts revealed no significant difference in the activation en-
ergies of the ethene and propene insertions into the Ni–alkyl
bond (propagation step, Scheme 2), because the content of the
resulting olefin mixtures is regulated by the initial ratios of the
comonomers. Such catalytic behavior was observed for both Ni
complexes investigated (Table 6). In addition, Ni compound 11,
bearing di-i-propylphenyl groups, produced the olefin mixtures
with higher content of pentenes relative to the C6 fraction than
Scheme 3.

K.V. Axenov et al. / Journal of Catalysis 238 (2006) 196–205
205
Me3Si-substituted Ni catalyst 8. In the presence of the bulky
aryl substituents near the metal center of catalyst 11, coordina-
tion of ethene versus a larger propene molecule can be sterically
preferable for codimerization. As a result of this steric con-
trol, the probability of formation of ethene–ethene and ethene–
propene dimers is enhanced.
[2] D. Vogt, in: B. Cornils, W.A. Herrmann (Eds.), Applied Homogeneous
Catalysis with Organometallic Compounds, vol. 1, VCH, Weinheim, New
York, Basel, Cambridge, Tokyo, 1996, pp. 246–248.
[3] W. Keim, in: M. Graziani (Ed.), Fundamental Research in Homogeneous
Catalysis, vol. 4, Plenum, New York, 1984, p. 131.
[4] E.F. Lutz, J. Chem. Educ. 63 (1986) 202.
[5] W. Keim, New J. Chem. 11 (1987) 531.
[6] D. Vogt, in: B. Cornils, W.A. Herrmann (Eds.), Aqueous-Phase Organo-
metallic Catalysis, VCH, Weinheim, 1998, p. 547.
[7] M. Wang, H. Zhu, K. Jin, D. Dai, L. Sun, J. Catal. 220 (2003) 392.
[8] S. Bogärt, T. Chenal, A. Mortreux, G. Nowogorcki, W.C. Lehmann, J.-F.
Carpentier, Organometallics 20 (2001) 199.
[9] K.-D. Hungenrg, J. Kerth, F. Langhauser, H.-J. Müller, P. Müller, Angew.
Makromol. Chem. 227 (1995) 159.
[10] C.M. Killian, L.K. Johnson, M. Brookhart, Organometallics 16 (1997)
2005.
[11] B.L. Small, M.J. Carney, D.M. Holman, C.E. O’Rourke, J.A. Halfen,
Macromolecules 37 (2004) 4375.
[12] B.L. Small, M. Brookhart, J. Am. Chem. Soc. 120 (1998) 7143.
[13] K.P. Tellmann, V.G. Cibson, A.J.P. White, D.J. Williams, Organo-
metallics 24 (2005) 280.
[14] F. Speiser, P. Braunstein, Organometallics 23 (2004) 2633.
[15] F. Speiser, P. Braunstein, Organometallics 23 (2004) 2625.
[16] J. Heinicke, M. Köhler, N. Peulecke, W. Keim, J. Catal. 225 (2004) 16.
[17] L. Stahl, Coord. Chem. Rev. 210 (2000) 203.
[18] G.G. Briand, T. Chivers, M. Krahn, Coord. Chem. Rev. 233–234 (2002)
237.
[19] L.P. Grocholl, L. Stahl, R.J. Staples, Chem. Commun. (1997) 1465.
[20] K.V. Axenov, V.V. Kotov, M. Klinga, M. Leskelä, T. Repo, Eur. J. Inorg.
Chem. (2004) 4702.
[21] K.V. Axenov, M. Klinga, M. Leskelä, T. Repo, Organometallics 24 (2005)
1336.
[22] T. Chivers, M. Krahn, G. Schatt, Inorg. Chem. 41 (2002) 4348.
[23] G.R. Lief, C.J. Carrow, L. Stahl, R.J. Staples, Chem. Commun. (2001)
1562.
[24] T. Chivers, C. Fedorchuk, M. Krahn, M. Parvez, G. Schatt, Inorg.
Chem. 40 (2001) 1936.
5. Conclusion
In this study, highly active ethene and propene oligomer-
ization and dimerization catalysts based on bis(imino)cyclodi-
phosph(V)azane Co and Ni complexes are described. The cat-
alytic activity of these complexes is sensitive to high oligomer-
ization temperature and monomer concentration. In contrast to
Co catalysts, which produce olefins with a Schulz–Flory distri-
bution, Ni complexes can selectively dimerize ethene. This may
be due to the increased instability of Ni–alkyl intermediates to-
ward β-H elimination. Both Co and Ni complexes catalyzed
propene dimerization. A 1,2 insertion of propene followed by
a 2,1 insertion into a M–H bond is the main reaction pathway
in propene dimerization by Co catalysts. In Ni complexes, the
probabilities of 1,2–1,2 and 1,2–2,1 propene insertions are al-
most equal, reflecting the composition of the resulting hexene
mixtures.
Acknowledgments
This work was supported by the Academy of Finland
(projects 209739 and 204408) and the Finnish National Tech-
nology Agency (TEKES). The authors thank Professor Markku
Ahlgrén (University of Joensuu) and Docent Martti Klinga
(University of Helsinki) for their help with the single-crystal
X-ray structural studies of trans-[(t BuNPMe)2(NSiMe3)2].
[25] S. Al-Benna, M.J. Sarsfield, M. Thornton-Pett, D.L. Ormsby, P.J. Maddox,
P. Bres, M. Bochmann, J. Chem. Soc., Dalton Trans. (2000) 4247.
[26] M. Bochmann, M.J. Sarsfield, US Patent WO 2000047592.
[27] T.G. Hill, R.C. Haltiwanger, M.L. Thompson, S.A. Katz, A.D. Norman,
Supplementary material
Inorg. Chem. 33 (1994) 1770.
[28] The synthesis of [( BuNPMe) (NSiMe ) ] was reported by O.J. Scherer,
t
Supplementary material includes the NMR data of obtained
ligands and complexes, structural data for trans-[(t BuNPMe)2-
(NSiMe3)2], trans-(3) (CCDC-283071) in cif format, and GC
chromatograms of the ethene dimerization and oligomerization
and ethene–propene codimerization.
2
3 2
G. Schnabl, Chem. Ber. 109 (1976) 2996.
[29] R.A. Shaw, Phosphorus and Sulfur 4 (1978) 101.
[30] W.Q. Tian, Y.A. Wang, J. Org. Chem. 69 (2004) 4299.
[31] I. Silaghi-Dumitrescu, I. Haiduc, Phosphorus and Sulfur 91 (1994) 21.
[32] P.J. Flory, J. Chem. Am. Soc. 62 (1940) 1561.
Please visit doi: 10.1016/j.jcat.2005.12.008.
[33] G.V. Schulz, Z. Phys. Chem. B 30 (1935) 379.
[34] This method was used in Ref. [13] to calculate the olefin mixture compo-
sition
[35] J. Scupinˇska, Chem. Rev. 91 (1991) 613.
References
[36] S.M. Pillai, M. Ravindranathan, S. Sivaram, Chem. Rev. 86 (1986) 353.
[37] D. Vogt, in: B. Cornils, W.A. Herrmann (Eds.), Applied Homogeneous
Catalysis with Organometallic Compounds, vol. 1, VCH, Weinheim, New
York, Basel, Cambridge, Tokyo, 1996, p. 260.
[1] D. Vogt, in: B. Cornils, W.A. Herrmann (Eds.), Applied Homogeneous
Catalysis with Organometallic Compounds, vol. 1, VCH, Weinheim, New
York, Basel, Cambridge, Tokyo, 1996, p. 245.