A survey of pendant donor-functionalised (N,O) phosphine ligands for Cr-catalysed ethylene tri- and tetramerisation

Article abstract of DOI:10.1039/c4cy00457d

In this study three classes of ligands are explored for ethylene tri-/tetramerisation in conjunction with chromium and both triethylaluminium (AlEt₃) and methylaluminoxane (MAO) co-catalysts. Hydrazine based ligands containing an N-H functionality [PN(NH)P], analogous to Rosenthal's previously reported PNPNH system, show selectivity towards 1-hexene and 1-octene formation in conjunction with AlEt₃, and act as PNP tetramerisation analogues when MAO is employed. PNP ligands containing non-protic pendant donor moieties generally show poor activity and selectivity when AlEt₃ is employed as an activator, however when MAO is used good activities and selectivities are achieved. n-Propylcyclopentane and 2-propenylcyclopentane, and higher homologues, are produced during catalysis when oxygen is the donor atom. The formation of such products is discussed with respect to the generation of methylenecyclopentane and methylcyclopentane by PNP based tetramerisation catalysts. Simple phosphine ligands containing O-H functionalisation are also explored and it was shown that the catalyst selectivity is highly dependent on both the activator and structural features of the ligand employed.

Full text of DOI:10.1039/c4cy00457d

Catalysis
Science &
Technology
View Article Online
View Journal | View Issue
PAPER
A survey of pendant donor-functionalised (N,O)
phosphine ligands for Cr-catalysed ethylene
tri- and tetramerisation†
Cite this: Catal. Sci. Technol., 2014,
4, 2574
a
b
James A. Suttil, Peter Wasserscheid, David S. McGuinness, a Michael G. Gardinera
*
*
and Stephen J. Evansc
In this study three classes of ligands are explored for ethylene tri-/tetramerisation in conjunction with
chromium and both triethylaluminium (AlEt3) and methylaluminoxane (MAO) co-catalysts. Hydrazine
based ligands containing an N–H functionality [PN(NH)P], analogous to Rosenthal's previously reported
PNPNH system, show selectivity towards 1-hexene and 1-octene formation in conjunction with AlEt3,
and act as PNP tetramerisation analogues when MAO is employed. PNP ligands containing non-protic
pendant donor moieties generally show poor activity and selectivity when AlEt3 is employed as an activa-
tor, however when MAO is used good activities and selectivities are achieved. n-Propylcyclopentane and
2-propenylcyclopentane, and higher homologues, are produced during catalysis when oxygen is the
donor atom. The formation of such products is discussed with respect to the generation of methyl-
enecyclopentane and methylcyclopentane by PNP based tetramerisation catalysts. Simple phosphine
ligands containing O–H functionalisation are also explored and it was shown that the catalyst selectivity
is highly dependent on both the activator and structural features of the ligand employed.
Received 11th April 2014,
Accepted 8th May 2014
DOI: 10.1039/c4cy00457d
www.rsc.org/catalysis
In 2004, researchers from Sasol reported the first catalysts
for the selective conversion of ethylene to 1-octene which
Introduction
While the production of linear alpha olefins (LAOs) is tradi-
tionally by full-range processes, these processes increasingly
do not meet the market demand for short chain LAOs
(1-butene, 1-hexene and 1-octene) which are important feed-
stocks for the production of linear low density polyethylene.
Attempts to meet the high demand for these feedstocks
may result in the over production of higher molecular weight
fractions and as such there has been significant focus on
developing alternate selective oligomerisation technologies
for ethylene conversion which will circumvent this issue. A
number of such technologies have been reported in both the
patent and academic literature and have been the focus of
a number of reviews in recent years.1–7 Investigation of
the mechanistic aspects of this reaction also continues to
receive interest.8–11 While a number of systems are known for
the selective trimerisation of ethylene to 1-hexene, there are
fewer known systems with high activity and selectivity for
ethylene tetramerisation.
consisted of a chromium source, methylaluminoxane and a PNP
ligand (I).12,13 In the initial reports activities of 591 000 g per g
of Cr h−1 and selectivities in excess of 70% 1-octene were
achieved however in subsequent reports activities exceeding
3 000 000 g per g of Cr h−1 have been reported.12–14 Since their
discovery, there has been numerous studies into the
electronic and steric factors that govern the PNP ligands novel
selectivity and excellent activity, as well the effects of the
ligand backbone length and composition.1,4,5 At the same
time, new ligand structures which support tetramerisation
catalysis continue to be developed.15–18 A particularly interest-
ing ligand, bearing marked similarity to Sasol's PNP system,
was described in the patent literature in 2009.19,20 The ligand
(II) contains the PNP backbone and, when unactivated, binds
to chromium in a similar fashion with coordination to the
metal centre through both phosphorous atoms as a bidentate
ligand.13,21–23 In the original report it was stated that, in con-
junction with a chromium source and MAO, the oligomer dis-
tribution consists of 1-hexene and 1-octene; a result that was
later confirmed by Wass and co-workers.20,22 Perhaps more
interesting is that combinations of the ligand with triethyl-
aluminium yield a catalyst capable of trimerising ethylene
with high selectivity (87% 1-hexene) and with an activity of
72 500 g per g of Cr h−1. Reduction of the ligand steric bulk
(III) was shown to convert the catalyst from predominantly
a School of Physical Sciences – Chemistry, University of Tasmania, Private Bag 75,
Hobart 7001, Australia. E-mail: david.mcguinness@utas.edu.au
b Lehrstuhl für Chemische Reaktionstechnik, Universität Erlangen-Nürnberg,
Egerlandstraße 3, 91058 Erlangen, Germany
c Sasol Technology Ltd, PO Box 1183, Sasolburg 1947, South Africa
† CCDC 995241–995245. For crystallographic data in CIF or other electronic
format see DOI: 10.1039/c4cy00457d
2574 | Catal. Sci. Technol., 2014, 4, 2574–2588
This journal is © The Royal Society of Chemistry 2014

View Article Online
Catalysis Science & Technology
Paper
producing 1-hexene to mainly 1-octene (80%), however the
activity was somewhat reduced (10 500 g per g of Cr h−1).19
The activation of the system with triethylaluminium is
of industrial relevance as it replaces costly MAO as the
co-catalysts.24–30
third weak donor. Such an arrangement has been demon-
strated by Gambarotta and co-workers in the case of Cr/pyrrole/
triethylaluminium (Phillips type) trimerisation catalysts,42,43 as
well as in other cases.45
The highly selective nature of the PNPNH systems in
conjunction with their long catalyst lifetime and the low
cost of the co-catalyst make them of interest to industry,
while their well-defined nature make them ideal for struc-
tural and mechanistic studies. Given these previous studies,
we were interested in investigating
a series of PNP
derivatives bearing either a protic nitrogen connected to
the PNP nitrogen backbone, which could react with the
alkylaluminium activator, or bearing donor functionalities,
capable of forming a Lewis pair with alkylaluminium com-
pounds. Additionally, it was off interest to see if this con-
cept could be extended to the use of hydroxy moieties,
both with bidentate phosphine ligands and simple mono-
phosphine ligands. Herein we report the preparation of
such ligands and their catalytic activity in conjunction with
a chromium source and triethylaluminium or MAO.
Since the development of this system there has been sig-
nificant work to determine the extent of the interaction
between the alkylaluminium co-catalyst, the chromium
source and the ligand. Early studies showed that the ligand
could readily coordinate to AlR3, however an increase in tem-
perature was required to deprotonate the PNPNH ligand
backbone.21,31–34 It was also demonstrated that further
heating resulted in rearrangement of the ligand backbone to
give an NPPN species which was shown to be inactive for eth-
ylene oligomerisation and was speculated to be the mode by
which the catalyst deactivated.21,32 The catalyst's temperature
dependence was also demonstrated in kinetic studies where
it was shown that at lower temperatures an induction period
of ~30 minutes is required before trimerisation commences.
However, at temperatures above that which ligand deproton-
ation can occur, ethylene trimerisation commences with no
induction period. Such kinetic studies have also demon-
strated the catalyst's long lifetime with ethylene uptake and
1-hexene production shown to be continuous over a period of
2 hours.35,36 In addition to this, supporting the ligand on a
series of functionalised polystyrene resins was shown to be a
viable method for preparing a recyclable catalyst material.37,38
These studies demonstrated that a supported system could
generate 1-hexene and that the catalyst material remained
active for a period greater than 40 hours and through eight
batch runs. In more recent reports Rosenthal and co-workers
have demonstrated the beneficial effect of addition of a chlo-
ride source to the catalyst mixture.39,40 A three-fold improve-
ment in activity for some chloride additives in combination
with Cr(acac)3 was demonstrated compared to standard runs
with CrCl3(thf)3. Interestingly, in this report combinations of
Cr(acac)3/PNPNH/AlEt3 were shown to be completely inactive
for catalysis, and as such the authors conclude that the active
species most likely contains chloride; a result which is further
supported by a subsequent structural study and is analogous
to previously reported ethylene trimerisation systems.33,34,41–44
However, while there is no doubt that addition of a chloride
source to the reaction mixture yields a positive result, previ-
ous reports by Rosenthal and co-workers have also demon-
strated that catalytic mixtures containing no chloride can
readily give high trimerisation activity and selectivity and as
such the exact role of chloro groups in the catalytic mecha-
nism remains unclear.19,23 At this stage most of the evidence
points towards a coordination arrangement such as shown in
IV, in which the chloro group attached to aluminium acts as a
Results and discussion
The ligand motifs explored in this study are shown in Scheme 1.
Ligand 1 is the PNPNH system reported by Rosenthal and co-
workers, and is included in the study to provide a benchmark
against which new catalysts could be compared.21 The ligands
upon which the present study is based fall into three general
classes. Ligands 2–8 are potentially bidentate and contain an
amine functionalisation either directly (2, 3), or pendant to
(4–8), the ligand backbone. With the exception of 2, these
ligands all contain the PNP ligand structure. They, therefore,
probably bear the closest similarities to ligand 1 of the systems
studied herein. The second class of ligands (9–13) contain a
pendant donor group attached, via an aromatic (9 and 10) or
an ethyl (11–13) bridge, to the central nitrogen of the PNP
ligand. While this donor cannot be deprotonated, it is expected
to be capable of coordinating to the aluminium cocatalyst,
Scheme
1 Ligand motifs surveyed for Cr-catalysed ethylene
oligomerisation.
This journal is © The Royal Society of Chemistry 2014
Catal. Sci. Technol., 2014, 4, 2574–2588 | 2575

View Article Online
Paper
Catalysis Science & Technology
which could potentially lead to analogous behaviour to that
observed with 1. Finally, we were also interested in surveying
some simple phosphines containing hydroxy functionalisation
(14–16). In these cases the cocatalyst could deprotonate and
interact with the hydroxyl group. Attempts to prepare ligands
with hydroxy functionality pendant to the PNP structure have
not been successful thus far, as discussed below.
Ligand synthesis
Compounds 1–3, 8, 14 and 15 were either prepared via the lit-
erature procedures or purchased and used as received (see
Experimental section).21,46–48 Ligands 7 and 9 have been previ-
ously reported in the literature and prepared by other methods
but were readily synthesised according to reaction (1).49,50 An
assortment of novel PNP ligands with protic nitrogen moieties
(4–6) or bearing donor functionalities (10–13) were prepared
according to reaction (1) and characterised by 1H, 13C and
31P NMR spectroscopy and either elemental analysis or mass
spectrometry. Each of the ligands shows a single phosphorous
resonance in the range of δ 60–70 which is consistent with
previously reported PNP analogue.14 Compound 10 shows
peak broadening for the phosphine bound phenyl rings in
Fig. 1 Molecular structure of ligand 4. Thermal ellipsoids are shown at
the 50% probability level. All methyl and aromatic-ring hydrogen atoms
are omitted for clarity. Selected bond lengths (Å) and angles (°): N1–P1/2
1.7204(10), 1.7150(10), N1–N2 1.4443(13), P–C 1.8239(12)–1.8331(12),
N2–H2N 0.903(18), C–P–C/N 101.23(5)–103.33(5), 102.02(5)–104.10(5),
P
(angles about NP2N/CHN) 359.40, 326.7.
1
the H and 13C NMR spectra, which is indicative of hindered
rotation around the C–P bond on the NMR time scale, inter-
estingly the analogous compound Ph2PN((o-CH3)C6H4)PPh2
wherein the trimethylsilyl ether is replaced with a methyl
ether shows no such restriction.51 Demethylation of (R,R)-1,2-
bis[(2-hydroxyphenyl)(phenylphosphino)]ethane, following a pre-
viously published analogous method, with boron tribromide
furnished ligand 16 in 54% yield (reaction (2)) with a single
phosphorous resonance at δ −40.37.52
Fig. 2 Molecular structure of ligand 13. Thermal ellipsoids are shown
at the 50% probability level. All hydrogen atoms are omitted for clarity.
Selected bond lengths (Å) and angles (°): N1–P1/2 1.7088(10)–1.7181(11),
P–C 1.8306(12)–1.8428(13), N–C 1.4743(14)–1.4862(14), C25–C26 1.5272(16),
P
C–P–C/N 100.69(5)–107.66(6), (angles about N1/2) 359.89, 339.67.
(1)
is consistent with the previously reported analogous compound
N,N-bis(diphenylphosphino)-2-(diphenylphosphino)ethanamine.
Compound 13 displays sp2 hybridisation for the P2NC core with
the sum of the bond angles in 13 equal to 359.89° compared to
360° for the analogous compound. Similarly the P–N1
(1.7088(10)–1.7181(11) Å/1.716(3)–1.718(3) Å), N1–C25 (1.4862(14) Å/
1.472(4) Å) and C25–C26 (1.5272(16) Å/1.519(4) Å) bond lengths
are in good agreement between the two compounds.55 In both
4 and 13, the pendant nitrogen donor and phosphines display
trigonal pyramidal geometry indicating that the electron lone
pairs are available for donation.
In addition to the preparation of a series of secondary
amine substituted PNP ligands, attempts have been made to
prepare PNP ligands bearing a hydroxyl moiety on the nitro-
gen backbone. While the target ligands have ultimately not
been synthesised, we have investigated some interesting
chemistry of the PNP ligands. Given that diphenylphosphine
chloride reacts preferentially with alcohols rather than
amines, standard synthetic methods could not be employed
to prepare the desired ligands and as such other strategies
(2)
Crystals of 4 and 13 suitable for X-ray diffraction were
grown from saturated solutions in anhydrous ethanol. The
solid state structure of 4 (Fig. 1) is broadly comparable to a
series of substituted 2,2-bis(diphenylphosphino)hydrazines
all of which contain a trigonal planar P2NN core; the core
bond angles of compound 4 sum to 359.40° which is in
agreement with the range of 356.7° to 359.9° for the analo-
gous compounds.50,53,54 The reported compound 1-tert-butyl-
2-bis(diphenylphosphino)hydrazine shows the greatest struc-
tural similarity to 4 with good agreement between the N1–N2
(1.4443(13) Å/1.445(1) Å) and N2–H2N (0.903(18) Å/0.89(2) Å)
bond lengths.50 The solid state structure of ligand 13 (Fig. 2)
2576 | Catal. Sci. Technol., 2014, 4, 2574–2588
This journal is © The Royal Society of Chemistry 2014

View Article Online
Catalysis Science & Technology
Paper
have been explored. Epoxide ring opening reactions were ini-
tially investigated as one means of introducing a hydroxyl
functionality to the PNP ligands. Surprisingly, the reaction
between bis(diphenylphosphino)amine and cyclohexene oxide
did not lead to opening of the epoxide ring, even in refluxing
tetrahydrofuran. Deprotonation of the PNP precursor with
n-butyllithium, to generate the PNP derived lithium amide,
followed by introduction of cyclohexene oxide (Scheme 2)
did lead to ring opening, however rather than the desired
N-functionalised compound, the reaction yielded 17 in mod-
erate yields after hydrolytic work up. 31P NMR spectroscopy of
geometry) indicating a significant reduction in its basicity and
hence a decrease in its capacity to react as a nucleophile in the
epoxide ring opening reaction.58 Examples of triarylphosphines
undergoing epoxide ring opening reactions have previously
been demonstrated in the literature and are typically facilitated
by the presence of an acid to act as the alcohol proton source
and the phosphorane counter anion.59–61 In the case of PNP
ligands it appears that no acid source is required as the
phosphorane is stabilised by the amides anionic character
while the lithium cation most likely generates a lithium alkox-
ide, upon ring opening, which is readily converted to the alco-
hol during hydrolytic work up.
2
17 shows two sets of doublets with a Jpp splitting of 94 Hz
which is comparable with similarly reported compounds.53,56
Crystals of 17 were grown from a saturated solution in hot
methanol (Fig. 3) and structural analysis shows PN–P bond
lengths broadly comparable with those reported for the anal-
ogous compounds 7-(Ph2–NPPh2)-8-NH2-quinoline and P,P-
diphenyl-N-(1,1,2,2-tetraphenyl-1-diphosphanylidene)phosphinous
amide.53,57 Upon initial inspection, it is somewhat surprising that
the ring opening reaction has occurred on a tertiary phosphorous
atom rather than the secondary amine core. However, the PNP
crystal structures reported herein (Fig. 1 and 2), as well as the
crystal structure of lithium[bis(diphenylphosphino)amide],
show an sp2 hybridised nitrogen core (trigonal planar
The use of protecting group chemistry has also been
employed in an attempt to prepare the desired derivatives.
Attempts to cleave the benzyl functionality of 12 by hydrogena-
tion with 5% palladium on carbon according to an analogous
literature procedure yielded only starting material and a sec-
ondary product wherein one phosphorous/nitrogen bond had
been cleaved; cleavage of the benzyl ether bond was not
observed.62 Similarly, no reaction of 12 was observed with other
hydrogenation catalysts.
Silyl ethers have often been employed as alcohol
protecting groups as they are easily and selectively prepared
in the presence of amines and can be readily cleaved either
with tetrabutylammonium fluoride (TBAF) and a proton
source or by hydrolysis. Deprotection of 11 with TBAF
gave only an intractable residue from which no species
could be definitively identified while both 10 and 11 proved
resistant to hydrolysis, even at elevated temperatures. It
has previously been demonstrated that phenolic trimethyl-
silyl ethers can be readily cleaved by titanium(IV) chloride
or its tetrahydrofuran adduct.63 However, introduction of
10 and 11 to TiCl4 in dichloromethane showed no cleavage
of the silyl ether but rather a bidentate coordination of
the ligand through the phosphorous donors (reaction (3)).
The stability of the silyl ether is quite remarkable and
somewhat unexpected given the oxophilic nature of tita-
nium. Complex 18 has been characterised by 1H, 13C and
31P NMR spectroscopy and elemental analysis while crystals
Scheme
2
Reaction
of
cyclohexene
oxide
with
lithium
bis(diphenylphosphino)amide to afford compound 17.
suitable for X-ray diffraction were grown from
a hot
dichloromethane–petroleum spirits solution (Fig. 4). Struc-
tural analysis shows a distorted octahedral arrangement
of the ligands around the titanium atom. In an attempt
to prepare crystals of 10 from a hot methanol solution it was
found that, under these conditions, the ligand undergoes
facile deprotection of the trimethylsilyl ether followed by a
rapid intramolecular reaction to generate
a monocyclic
hydrophosphorane (19, reaction (4)). Compound 19 crystal-
lises readily from saturated diethyl ether solutions to give
colourless needles suitable for X-ray diffraction studies. The
solid state structure of 19 (Fig. 5) shows a trigonal bipyramidinal
arrangement of the ligands around the phosphorane core with
the phenoxide moiety and proton occupying the axial posi-
tions. Several similar hydrophosphorane structures have been
reported but, to the best of the authors' knowledge, this is the
first example based on a PNP backbone.64–70
Fig. 3 Molecular structure of 17. Thermal ellipsoids are shown at the
50% probability level. Disorder in a phenyl ring (C25) is omitted for
clarity. All methylene, methine and aromatic-ring hydrogen atoms are
omitted for clarity. Selected bond lengths (Å) and angles (°): N1–P1,P2
1.5846(13), 1.6853(13), P–C 1.8092(15)–1.8413(18), O1–H1 0.91(4), N1⋯H1
1.90, C–P1–C 106.42(6)–108.37(7), N1–P1–CPh/Cy 112.87(7)–114.66(7), 107.00(7),
C–P2–C/N1 96.77(7)–102.16(8).
This journal is © The Royal Society of Chemistry 2014
Catal. Sci. Technol., 2014, 4, 2574–2588 | 2577

View Article Online
Paper
Catalysis Science & Technology
Ethylene oligomerisation and polymerisation
While there have been numerous ethylene trimerisation stud-
ies with 1 previously reported in the literature,21–23,32–36,38,40
it was of interest to optimise the ligand under our conditions
in order to determine a series of conditions with which to
benchmark the novel ligand systems. The results of catalytic
testing of 1/Cr in conjunction with triethylaluminium are
shown in Table 1. In accordance with the literature the cata-
lyst shows stable selectivity across a range of temperatures
and ethylene pressures yielding, in most cases, in excess of
95% C6 isomers which consist almost exclusively of 1-hexene.
We do note some differences in the relative amounts of the
minor products when compared to other studies. In particu-
lar, less C4 and C10 products, and slightly more polymer.
This might be related to different reaction solvents (cyclohex-
ane versus toluene), but has not been explored further. As
expected, increasing the ethylene pressure yields a corre-
sponding increase in the activity of the system (cf. Table 1,
entries 1–3) due to the first order dependence on ethylene.35
More profound however is the effect of temperature on activ-
ity; an increase of 10 °C shows an almost doubling in activity
(cf. Table 1, entries 3 and 4), which has previously been attrib-
uted to higher temperatures facilitating deprotonation of the
nitrogen functionality by triethylaluminium.21,32 It has also been
reported that only moderate amounts of cocatalyst are required
to generate an active trimerisation system and that lower cata-
lyst loadings are beneficial to catalyst activity as they reduce the
rate of ligand isomerisation to an inactive species.35 While we
were able to reproduce this trend (cf. Table 1, entries 4 and 5)
for subsequent testing of new catalysts it was reasoned that a
higher cocatalyst loading may be beneficial as an impurity scav-
enger. While the inclusion of chloro-containing modifiers can
have a beneficial effect when used in conjunction with 1, it has
also been shown that employing chromium(III) chloride tetrahy-
drofuran adduct as the metal source is sufficient.40
Fig. 4 Molecular structure of 18. Thermal ellipsoids are shown at
the 50% probability level. A second molecule of similar geometry
and a DCM lattice solvent molecule is not shown (geometries given
below are for the range found in both molecules). All hydrogen
atoms are omitted for clarity. Selected bond lengths (Å) and angles (°):
Ti–P 2.6314(11)–2.6393(12), Ti–Cltrans
2.2796(8)–2.3022(8),
P/trans Cl
2.2360(14)–2.2694(12), N–P 1.711(4)–1.720(4), P–Ti–P 62.99(3), 63.38(3),
Cl–Ti–Pcis/trans 81.00(4)–91.47(5), 153.13(5)–154.65(4), Cl–Ti–Cl
114.77(4)–115.33(4),92.25(4)–94.95(4),164.40(3)–166.82(3), (angles about
N) 356.6–357.8.
cis equat/cis axial/trans
P
Screening of the ligands 1–4 and 9–16 was undertaken
(Table 2) employing the optimised conditions determined for
ligand 1 and the behaviour of the different classes of ligand
are discussed below.
Fig. 5 Molecular structure of 19. Thermal ellipsoids are shown at the
50% probability level. All aromatic-ring hydrogen atoms are omitted
for clarity. Selected bond lengths (Å) and angles (°): P1–N1,O1,C7,C13
1.7384(11), 1.7666(9), 1.8281(14), 1.8259(13), 1.337(18), H1–P1–O1,N1,C7,
C13 175.1(8), 88.3(8), 90.4(8), 92.4(8), N1–P1–C7,C13 124.01(6)–124.18(6),
C7–P1–C13 111.81(6).
Ligands containing N–H functionality (2–8). Ligand 2,
which is envisaged to be bidentate through the phosphine
and pyridine nitrogen donors with a protic nitrogen bridge,
shows poor oligomerisation selectivity in conjunction with
triethylaluminium yielding a series of LAOs and a high
degree of polymer (Table 2, entry 2). Ligands 3 and 4 are PNP
analogues that contain amine moieties in their ligand
structure which we envision could be deprotonated in a similar
fashion to that reported by Rosenthal and co-workers.21 Ligand
3 has previously been reported as an ineffective ethylene tri-/
tetramerisation system in conjunction with MAO yielding a
broad range of oligomeric products.14 Similarly, when
triethylaluminium is employed as the activator, 3 yields a
broad range of oligomeric products (Table 2, entry 3); attempts
to improve the selectivity and activity by varying both the
temperature and cocatalyst loading (Table 2, entries 4 and 5)
(3)
(4)
2578 | Catal. Sci. Technol., 2014, 4, 2574–2588
This journal is © The Royal Society of Chemistry 2014

View Article Online
Catalysis Science & Technology
Paper
Table 1 Optimisation of 1/Cr/AlEt3 for ethylene trimerisationa
Entry
Pressure (bar)/T (°C)
Activityb
% C4
% C6 (1-C6)
% C8 (1-C8)
% Other
% PE
1
2
3
30/50
45/50
50/50
50/60
50/60
19 970
22 270
24 150
44 380
50 260
0.1
0.1
0.1
0.2
0.0
95.1 (99.9)
95.4 (99.9)
94.6 (99.9)
95.6 (99.9)
95.4 (99.9)
3.4 (3.4)
3.3 (5.5)
3.3 (5.6)
2.6 (5.0)
2.7 (4.6)
0.4
0.4
0.4
0.3
0.3
1
0.9
1.6
1.3
1.5
4
5c
a
b
c
17.5 μmol of ligand 1, 10 μmol CrCl3(thf)3, 70 equiv. AlEt3, 200 mL cyclohexane total volume, 30 min. g product per g of Cr h−1
.
Run
employed only 15 equivalents of AlEt3.
a
Table 2 Ethylene oligomerisation and polymerisation with ligands 1–4 and 9–16 in conjunction with Cr/AlEt3
Entry
Catalyst
Activityb
% C4
% C6 (1-C6)
% C8 (1-C8)
% Other
% PE
1
2
1
2
3
3
3
4
9
10
11
12
13
14
15
16
44 380
7630
2760
4550
3710
4590
3600
—
—
—
4280
3200
3280
910
0.2
0.3
1.3
0.9
17.2
1.6
0.6
Trace
Trace
Trace
2.8
95.6 (99.9)
1.5 (94.7)
9.3 (88.2)
1.6 (69.5)
12.9 (92.8)
11.1 (90.4)
3.1 (94.4)
Trace
Trace
Trace
28.8 (98.6)
5.1 (96.4)
38.3 (98.8)
17.7 (99.9)
2.6 (5.0)
2.0 (34.8)
8.2 (53.3)
1.7 (75.9)
6.9 (41.5)
17.6 (85.7)
4.1 (44.3)
—
0.3
10.5
16.3
2.6
6.1
7.0
22.8
—
—
—
3.4
16.5
0
1.3
85.7
64.9
93.1
56.9
62.7
69.5
—
—
—
63.3
72.1
58.7
76.2
3
4c
5d
6
7
8e
9e
10e
11
12
13
14e
—
—
1.7 (75.4)
5.1 (30.3)
1.6 (75.3)
1.0 (50.0)
1.2
1.4
3.3
1.8
a
b
17.5 μmol of ligand, 10 μmol CrCl3(thf)3, 70 equiv. AlEt3, 200 mL cyclohexane total volume, 30 min, 50 bar ethylene pressure, 60 °C.
g
product per g of Cr h−1
.
Run performed at 40 °C. Run employed 300 equivalents of AlEt3. 17.5 μmol of ligand, 10 μmol CrCl3(thf)3, 70
c
d
e
equiv. AlEt3, 50 mL cyclohexane total volume, 30 min, 50 bar ethylene pressure, 60 °C.
gave little improvement towards selective oligomerisation.
Interestingly, the hydrazine derived ligand 4, when activated
with triethylaluminium, shows selectivity in the liquid phase
towards both 1-hexene (overall 10%) and 1-octene (overall 15%,
Table 2, entry 6, combined 1-C6 and 1-C8 represents 67% of liq-
uid products). Such a liquid phase distribution is comparable
with that observed for the Sasol PNP ligands, however in our
case there is still a significant contribution from polymer for-
mation. Work with 1 has shown that temperature and AlEt3
concentration can have a profound effect on the activity and
selectivity of the catalyst and as such it was of interest to
attempt to improve our system by varying these factors
(Table 3).32,35 While reducing the temperature does indeed
increase the activity it also significantly increases polymer for-
mation (cf. Table 3, entries 1 and 2). However, increasing the
temperature was also found to have a detrimental effect yield-
ing a significant increase in the formation of higher oligomers
(cf. Table 3, entries 2 and 4). These results are somewhat con-
sistent with those reported for 1 wherein a narrow temperature
range is required for the deprotonation of the amine by tri-
ethylaluminium to occur, generating an active trimerisation
system, but excessive heating results in catalyst deactivation.32
Increasing the loading of the cocatalyst (Table 3, entry 3) shows
a slight increase in the selectivity towards 1-hexene however a
significant amount of 1-butene is now formed, indicating that
the ethylene oligomerisation has shifted from a somewhat
selective distribution towards a Schulz–Flory type distribution.
Structural factors of the hydrazine derived PNP ligands and
their effect on catalytic selectivity have also been investigated
(Table 4). In all cases where a bulkier nitrogen substituent was
Table 3 Optimisation of 4/Cr/AlEt3 for ethylene tri-/tetramerisationa
Entry
Temperature (°C)
Activityb
% C4
% C6 (1-C6)
% C8 (1-C8)
% Other
% PE
1
40
60
60
75
5970
4590
4057
4380
1.3
1.6
13.7
1.0
3.5 (91.8)
11.1 (90.4)
14.6 (90.7)
8.3 (93.0)
8.7 (92.0)
17.6 (85.7)
15.9 (76.4)
10.9 (79.4)
3.0
7.0
8.9
83.5
62.7
46.9
52.8
2
3c
4
27.0
a
b
17.5 μmol of ligand, 10 μmol CrCl3(thf)3, 70 equiv. AlEt3, 200 mL cyclohexane total volume, 30 min, 50 bar ethylene pressure, 60 °C.
g
product per g of Cr h−1
. Run employed 300 equivalents of AlEt3.
c
This journal is © The Royal Society of Chemistry 2014
Catal. Sci. Technol., 2014, 4, 2574–2588 | 2579

View Article Online
Paper
Catalysis Science & Technology
a
Table 4 Ethylene oligomerisation and polymerisation with hydrazine derived ligands 4–8 in conjunction with Cr/AlEt3
Entry
Catalyst
Activityb
% C4
% C6 (1-C6)
% C8 (1-C8)
% Other
% PE
1
2
3
4
5
4
5
6
7
8
4590
2455
2310
2180
1130
1.6
2.9
2.5
4.5
9.1
11.1 (90.4)
18.7 (93.2)
18.6 (98.4)
16.1 (96.5)
32.8 (97.2)
17.6 (85.7)
12.9 (67.1)
12.2 (60.3)
11.2 (73.9)
13.4 (66.9)
7
62.7
47.3
50
61.6
33.9
18.2
16.7
6.6
10.8
a
b
17.5 μmol of ligand, 10 μmol CrCl3(thf)3, 70 equiv. AlEt3, 200 mL cyclohexane total volume, 30 min, 50 bar ethylene pressure, 60 °C.
g
product per g of Cr h−1
.
used the liquid phase selectivity shifts from predominantly
1-octene towards 1-hexene. Ligands 5, 6 and 7 show overall
1-hexene selectivities of 17%, 18% and 16% respectively
(Table 4, entries 2–4), however a significant amount of higher
oligomers was also recorded for ligands 5 and 6. Screening of
ligand 8 resulted in the most selective trimerisation system of
the series with an overall 1-hexene selectivity of 32%, albeit the
activity of this ligand is also the lowest which might perhaps
be attributed to its rearrangement, that has been known to
occur upon deprotonation.46 It is notable that, for ligands 4, 7
and 8, the combined 1-hexene and 1-octene selectivity within
the liquid product fraction is reasonably high (62–67%), how-
ever the formation of polyethylene lowers the overall selectivity.
PNP ligands containing pendant donor functionalities (9–13).
Previous structural studies of PNP ligands bearing pendant
ether and thioether donor functionalities from the nitrogen
backbone have shown that, for ethyl and propyl bridges
between the nitrogen and donor atom, rather than the
expected tridentate binding mode to chromium the ligands
adopt a bidentate P,P coordination mode with the pendant
donor remaining uncoordinated.51,71 As such we envisaged
that analogous pendant donor functionalised ligands may be
able to form Lewis acid/base adducts with triethylaluminium
through their donor moiety, bringing the aluminium cocatalyst
into close proximity with the chromium metal centre and
potentially generating an active tri-/tetramerisation system. This
concept has been demonstrated by Rosenthal and co-workers
for their ethylene trimerisation system 1 (structure IV).21 Unfor-
tunately the majority of pendant donor functionalised systems
screened either led to poor liquid phase selectivity (Table 2,
entry 7) or, in the case of the ether functionalised ligands, no
catalytic activity was noted (Table 2, entries 8–10). However,
ligand 13 shows good selectivity towards ethylene trimerisation
in the liquid phase (Table 2, entry 11, overall 28% 1-hexene,
76% of liquid products), but again suffers from high polymer
formation. The reason behind the lack of activity demonstrated
by the ether functionalised systems in comparison to their
amine functionalised counterparts is currently unclear.
Phosphine ligands containing hydroxyl functionality (14–16).
Screening of the simple phosphine ligand 14 yielded approx-
imately 28% liquid phase oligomers however the selectivity
towards short chained linear alpha olefins was only moderate
(Table 2, entry 12). Conversely, ligands 15 and 16, where the
hydroxyl/phosphine bridge is converted from a saturated
alkane to an aromatic bridge, shows an improvement in selec-
tivity towards 1-hexene yielding overall 38% and 17% 1-hexene
respectively (Table 2, entries 13 and 14). Previous studies have
shown the benefits of bidentate phosphine ligands for selective
oligomerisation, and as such it was of interest to increase the
ratio of 15 : chromium in an attempt to improve selectivity.1,4–7
Increasing the ligand : Cr ratio from 1.75 :1 to 2.2:1 yields a
two-fold increase in overall 1-hexene selectivity from 37.8% to
75.3%, however a decrease in activity is also noted (cf. Table 5,
entries 1 and 2). This indicates that the shift in selectivity
results not from an increase in the production of 1-hexene but
rather from a suppression of polymer formation. Interestingly,
activation of 15, regardless of the ligand ratio, with MAO
yielded only a polymerisation catalyst (Table 5, entries 3 and 4).
Ligands of the type 15 and 16, in conjunction with triethyl-
aluminium, are reminiscent of aryloxide systems first disclosed
by IFP Energies Nouvelles with both systems (15 and IFP's
bis(2,6-diphenylphenoxy)isobutylaluminium) yielding compa-
rable activities and selectivities.72,73 Moreover, workers from
Sasol Technology have demonstrated that combinations of
a chromium source, triethylaluminium and certain phenols
when combined in situ can yield effective ethylene
Table 5 Ethylene oligomerisation with 15/Cr and AlEt3 or MAO
Entry
Ligand (μmol)
Activitya
% C4
% C6 (1-C6)
% C8 (1-C8)
% Other
% PE
1b
2b
3c
4c
17.5
22.0
10.0
20.0
3280
2040
22 310
10 190
1.4
2.4
—
38.3 (98.8)
76.2 (98.8)
—
—
1.6 (75.3)
2.5 (74.0)
—
0
0
—
—
58.7
18.9
100
—
—
100
a
b
g product per g of Cr h−1
.
10 μmol CrCl3(thf)3, 70 equiv. AlEt3, 200 mL cyclohexane total volume, 30 min, 50 bar ethylene pressure, 60 °C.
c
10 μmol Cr (acac)3, 500 equiv. MAO, 200 mL cyclohexane total volume, 30 min, 45 bar ethylene pressure, 50 °C.
2580 | Catal. Sci. Technol., 2014, 4, 2574–2588
This journal is © The Royal Society of Chemistry 2014

View Article Online
Catalysis Science & Technology
Paper
trimerisation catalysts, however high ligand : chromium ratios
are required.74,75
high activities known for other PNP derivatives. Similarly, this
system displays good selectivity towards 1-octene; overall 55%
of the total products.
Screening of 2 and 4–16 in conjunction with MAO activation.
While it is of industrial importance to attempt to replace MAO
with significantly cheaper cocatalysts such as triethylaluminium,
there is still significant interest in generating new catalytic
systems for ethylene oligomerisation, and as such we have also
tested our novel systems in conjunction with MAO (Table 6),
which is known to facilitate selective ethylene oligomerisation.
Ligands 1 and 3 in conjunction with MAO have been tested
previously,14,22 while ligand 15 has been discussed above (see
Table 5).
Ligand 2 shows only low selectivity towards selective tri-/
tetramerisation due to the significant degree to which higher
LAOs were also formed (Table 6, entry 1). The hydrazine
derived ligands (4–8) act as direct PNP analogues in conjunc-
tion with MAO, albeit with high polymer formation, displaying
selectivity towards 1-hexene (overall between 7–12%) and
1-octene (overall 30–60%). Ligand 4, which has the lowest steric
bulk on the nitrogen backbone, again yields both the highest
1-octene selectivity and overall activity of hydrazine derived
ligands. Interestingly, ligand 8 once again shows poor activity
compared to the alkyl substituted species which is most likely
due to its poor stability.46
Activation of the pendant donor functionalised PNP
ligands 9–13 with MAO yielded a series of selective oligomerisation
systems. Ligand 9 displayed both the lowest activity and selec-
tivity of the pendant donor functionalised systems screened
(Table 6, entry 7), similar results have previously been demon-
strated by Hor and co-workers for other pyridine functionalised
PNP ligands.71 Conversely, the amine donor functionalised
ligand 13 yielded the most active system screened in conjunc-
tion with MAO, giving activities in excess of 420 000 g product
per g of Cr h−1 (Table 6, entry 11). While no direction compari-
son can be made, this system is beginning to approach the
The ether functionalised ligands 10–12 also show good
activity and a high degree of selectivity towards 1-octene for-
mation (Table 6, entries 8–10). Analogous ligands have previ-
ously been explored by Hor and co-workers and Bercaw and
co-workers. Our results are generally in agreement with these
studies.51,71 It is noteworthy that in previous reports the lon-
ger chain oligomeric products are reported to be either C11+
linear alpha olefins71 or C10 isomers (through the co-
trimerisation of 1-hexene and ethylene) and C12 isomers
(through the co-trimerisation of 1-octene and ethylene),51
however the co-trimeric isomers were not present in our cata-
lytic mixtures. Detailed analysis of the liquid phase for
ligands 10–12 in conjunction with MAO reveals some inter-
esting features. Each oligomeric fraction consists of a distri-
bution of four isomers; for example the C6 oligomeric
fraction from the catalytic mixture of ligand 10 comprises of
38.8% 1-hexene, 5.4% hexane, 23.9% methylcyclopentane
and 31.9% methylenecyclopentane. Although the C8 fraction
is enriched in 1-octene (~95%), octane, n-propylcyclopentane
and 2-propenylcyclopentane are also detected; while the
higher oligomers (C10+) show approximately equal portions of
each of the four isomers. A previous investigation into the
formation of methylcyclopentane and methylenecyclopentane
by PNP based ethylene tetramerisation systems has proposed
two possible mechanisms by which the species can form.76
In that report the authors suggest that the readily formed
chromacycloheptane can rearrange either via a concerted
mechanism (Scheme 3, pathway a) or a formal β-hydride
transfer to chromium and subsequent cyclisation of the
5-hexenyl moiety by reinsertion (Scheme 3, pathway b) to
generate the cyclopentylmethylchromium intermediate 20.
This intermediate is proposed to decompose either via a
Table 6 Ethylene oligomerisation and polymerisation with ligands 2 and 4–16 in conjunction with Cr/MAOa
Entry
Catalyst
Activityb
% C4
% C6 (1-C6)
% C8 (1-C8)
% Other
% PE
1
2
3
4
5
6
2
4
5
6
7
8
9
10
11
12
13
14
15
15
16
14 320
38 260
17 370
20 050
14 820
8880
2.4
1.6
2.1
1.9
1.8
3.3
5.2
3.1
2.4
2.2
1.2
1.3
—
7.6 (60.6)
12.1 (62.4)
11.5 (64.8)
13.0 (67.6)
13.5 (86.3)
13.6 (81.0)
11.3 (55.2)
23.5 (38.8)
16.7 (35.4)
15.0 (41.1)
10.5 (57.1)
12.8 (73.2)
—
13.2 (87.0)
62.6 (96.4)
46.2 (96.4)
52.4 (96.6)
39.5 (98.0)
30.8 (96.0)
10.0 (87.5)
55.1 (94.5)
52.8 (95.7)
41.7 (96.0)
58.0 (96.1)
55.2 (97.7)
—
24.4
17.7
11.4
11.6
5
52.4
6.0
28.8
21.1
40.2
43.2
62.2
4.2
11.8
22.5
15.7
22.1
100
9.1
7c
8d
9d
10d
11
12e
13
14e
15d
5870
11.2
14.1
16.7
18.6
14.6
8.7
—
—
2.9
37 810
40 530
11 350
425 020
46 140
22 310
10 190
1460
—
1.2
—
—
100
93.8
1.3 (80.0)
0.8 (86.1)
a
b
10 μmol of ligand, 10 μmol Cr(acac)3, 500 equiv. MAO, 200 mL cyclohexane total volume, 30 min, 45 bar ethylene pressure, 50 °C. g product
per g of Cr h−1
.
Run performed at 60 °C, 50 bar ethylene pressure. 10 μmol of ligand, 10 μmol Cr(acac)3, 500 equiv. MAO, 50 mL
c
d
cyclohexane total volume, 30 min, 45 bar ethylene pressure, 50 °C. e 20 μmol of ligand employed.
This journal is © The Royal Society of Chemistry 2014
Catal. Sci. Technol., 2014, 4, 2574–2588 | 2581

View Article Online
Paper
Catalysis Science & Technology
Summary and conclusions
Herein we have prepared a series of ligands that fall
into three general classes, each of which has been tested in
conjunction with triethylaluminium and MAO. The first
series contain an N–H functionality and are structurally similar
to Rosenthal and co-worker's active PNPNH trimerisation
system.21 In conjunction with triethylaluminium, the hydra-
zine derived PNP ligands 4–8, show selectivity towards
1-hexene and 1-octene formation. Unfortunately, this liquid
phase selectivity is accompanied by a high degree of poly-
mer formation, which ultimately could not be overcome.
When MAO is employed as the activator these ligands act
as PNP analogues yielding a high degree of selectivity
towards 1-octene.
Scheme
3 Postulated mechanisms for the formation of
cyclopentylmethylchromium intermediate 20; (a) concerted, and (b)
stepwise β-hydride transfer and reinsertion.
While our results with the hydrazine-based ligands do not
match those achieved with the remarkable system of Rosenthal
and co-workers, this new system does offer promise for MAO
free tri- and tetramerisation, and provides directions for future
development. The formation of two distinct product categories,
short chain α-olefins and polyethylene, perhaps suggests that a
number of different active species (and mechanisms) are
formed upon activation. This is a behaviour we have observed
previously with Ti-based catalysts.78,79 The key to improving
overall selectivity might therefore lie in controlling activation
and such multi-mechanism behaviour, which is, unfortunately,
unlikely to be a trivial matter.
The second class of catalysts consist of pendant donor
functionalised PNP ligands. In the majority of cases activa-
tion with triethylaluminium leads to either low activity and
poor selectivity or no activity at all. However, under the same
conditions the amine functionalised ligand 13 shows selectivity
towards 1-hexene. Broadly speaking, the pendant functionalised
ligands, in conjunction with MAO, act as PNP derived tri-/
tetramerisation systems and yield results consistent with
previous studies. Ligand 13 shows the most remarkable activity
(in excess of 425 000 g per g of Cr h−1) of the systems screened
in this study. The presence of n-propylcyclopentane and
2-propenylcyclopentane, and higher homologues, formed by the
ether functionalised ligands 10–12 during catalysis, strongly
supports the formation of such species via a cyclisation-chain
growth-disproportionation mechanism rather than a radical
based mechanism.
Screening of simple phosphine ligands containing an O–H
functionality employing triethylaluminium as the cocatalyst
has shown that with an aromatic bridge between the phos-
phine and hydroxyl moiety, the catalysts are effective for eth-
ylene trimerisation and that polymerisation is suppressed at
high ligand : chromium ratios. Such findings are consistent
with previous work on aryloxy-chromium catalysts.75 Con-
versely, in conjunction with MAO ligands 15 and 16 yield
polymerisation systems, however 14 (aliphatic phosphine/
hydroxyl bridge) shows selectivity for tetramerisation. As
such, it is clear that both the choice of co-catalyst and ligand
structure are important in determining the selectivity for this
ligand motif.
Scheme 4 Ethylene insertion into intermediate 20 to yield higher
n-alkylcyclopentanes and alkenylcyclopentanes.
disproportionation process (either mono- or bimetallic) or
via cyclopentylmethyl radicals (although 5-hexen-1-yl radicals
cannot be discounted); however experimental evidence
disfavoured the latter mechanism.76 The formation of the
longer chained homologues of methylcyclopentane and
methylenecyclopentane by ligands 10–12 is readily ratified
by ethylene insertion (Cossee–Arlmann type chain growth)
into the alkyl–Cr bond of intermediate 20 (Scheme 4) and
subsequent decomposition via a disproportionation process;
however the formation of such oligomers by radical interme-
diates seems unlikely. As such, the findings in this study are
complimentary to the results previously published by Overett
and co-workers.76
Somewhat surprisingly, ligand 14 which gave poor selectiv-
ity upon activation with triethylaluminium yields an active
tri-/tetramerisation system when two ligand equivalents are
employed with MAO (Table 6, entry 12), with good selectivity
to 1-octene observed (overall 1-octene selectivity of 54%). To
the best of the authors' knowledge this is the first report of
this simple ligand motif for selective ethylene tri-/
tetramerisation. Ligand 16, which is structurally similar to SK
Energy's bridged diphosphine ligands for ethylene tri-/
tetramerisation, surprisingly yielded only poor short chain
LAO selectivity with low activity (Table 6, entry 15).77
2582 | Catal. Sci. Technol., 2014, 4, 2574–2588
This journal is © The Royal Society of Chemistry 2014

View Article Online
Catalysis Science & Technology
Paper
3.98 mmol) was added dropwise and the resulting pale yellow
solution was stirred overnight. The solution was washed with
3 × 30 mL of degassed, deionised water, dried with sodium
sulphate, filtered and the volatiles removed under reduced
pressure to give a cream coloured residue. The residue was
extracted with 8 mL of anhydrous ethanol from which the
title compound crystallized overnight as a white solid in 21%
yield (0.20 g, 0.42 mmol). 1H NMR (CDCl3, 400 MHz): δ 7.42–7.48
(m, 8H, aryl-H), 7.21–7.31 (m, 12H, aryl-H), 3.78 (s, 1H, NH), 2.46
(t, J = 10.5 Hz, 1H, NCH), 1.42 (m, 5H, cyclohexyl-H), 0.85 (m, 3H,
cyclohexyl-H), 0.59 (m, 2H, cyclohexyl-H). 31P NMR (CDCl3,
161.8 MHz): δ 69.60 (s). 13C NMR (CDCl3, 100.5 MHz): δ 139.3
(t, J = 6.2 Hz, aryl-Cipso), 133.2 (t, J = 11.6 Hz, aryl-C), 128.9
(s, aryl-C), 128.0 (t, J = 2.9 Hz, aryl-C), 58.5 (t, J = 3.8 Hz, NCH),
31.3 (s, cyclohexyl-C), 26.0 (s, cyclohexyl-C), 24.8 (s, cyclohexyl-
C). Anal. calcd for C30H32P2N2: C 74.67, H 6.68, N 5.81. Found:
C 74.95, H 6.93, N 6.10.
Preparation of (Ph2P)2N–N(i-Pr)H (6). Compound 6 was
prepared according to the procedure outlined for compound 5,
employing isopropylhydrazine hydrochloride (0.22 g, 1.99 mmol).
Compound 6 was isolated as an off-white solid in 21% yield
(0.19 g, 0.42 mmol). 1H NMR (CDCl3, 400 MHz): δ 7.40–7.51
(m, 8H, aryl-H), 7.20–7.34 (m, 12H, aryl-H), 3.81 (d, J = 2.0 Hz,
1H, NH), 2.94 (m, 1H, CH(CH3)2), 0.59 (d, J = 6.3 Hz, 6H,
CH(CH3)2). 31P NMR (CDCl3, 161.8 MHz): δ 68.79 (s). 13C NMR
(CDCl3, 100.5 MHz): δ 139.4 (t, J = 5.9 Hz, aryl-Cipso), 133.3 (t,
J = 11.4 Hz, aryl-C), 128.9 (s, aryl-C), 128.0 (t, J = 2.8 Hz, aryl-C),
50.1 (t, J = 4.9 Hz, CH(CH3)2), 20.7 (s, CH(CH3)2). Anal. calcd
for C27H28P2N2: C 73.29, H 6.38, N 6.33. Found: C 73.33, H
6.09, N 6.31.
Experimental
General considerations
(2-Hydroxyphenyl)diphenylphosphine (Sigma-Aldrich, 97%)
and N,N-bis(diphenylphosphino)amine (Strem, 98%) were
purchased and used as received. Ph2PN(i-Pr)P(Ph)N(i-Pr)H
(1),21 Ph2PN(2-pyridinel)H (2),48 (Ph2P)2N–N(Ph)H (8) (ref. 46)
and (2-hydroxyethyl)diphenylphosphine (14) (ref. 47) were pre-
pared according to literature procedures. Syntheses involving
air/moisture sensitive reactants or products were carried out
under argon or nitrogen using standard Schlenk techniques, or
in a glove box. Solvents were purified by passage through an
Innovative Technologies purification system and, where appro-
priate, stored over a sodium mirror. MAO was supplied by
Albemarle as a 10% solution in toluene and MMAO-3A was
supplied by AkzoNobel as a 7% solution in heptane.
Nuclear magnetic resonance studies were performed using
an Oxford AS400 or Varian Mercury 300Plus NMR spectrome-
ter at room temperature. 1H NMR spectra were acquired at
400 MHz or 300 MHz, 13C NMR spectra at 100.5 MHz or
75 MHz and 31P NMR spectra at 161.8 MHz. H NMR spectra
1
were internally referenced to the residual solvent peaks, and
13C NMR spectra to deuterated solvent resonances. 31P NMR
spectra were referenced to triphenylphosphine standard.
The liquid products resulting from ethylene oligomerisation
were analysed by GC on a Varian 3900 or a Shimadzu GC-2014
utilising helium carrier gas and FID detector. Quantification
was achieved via addition of a known volume of cyclohexane or
nonane internal standard upon completion of the catalytic run.
Preparation of (Ph2P)2N–N(t-Bu)H (7). Compound 7 was
prepared according to the procedure outlined for compound 5,
employing tert-butylhydrazine hydrochloride (0.25 g, 2.01 mmol).
Compound 7 was isolated as colourless plates in 25% yield
(0.23 g, 0.51 mmol). 1H NMR (CDCl3, 400 MHz): δ 7.45–7.51
(m, 8H, aryl-H), 7.19–7.25 (m, 12H, aryl-H), 3.76 (bs, 1H, NH),
0.73 (s, 9H, NC(CH3)3). 31P NMR (CDCl3, 161.8 MHz): δ 71.21 (s).
13C NMR (CDCl3, 100.5 MHz): δ 139.7 (d, J = 14.0 Hz, aryl-Cipso),
133.7 (d, J = 23.5 Hz, aryl-C), 128.8 (s, aryl-C), 127.7 (d, J = 6.4 Hz,
aryl-C), 55.4 (s, NC(CH3)3), 28.9 (s, NC(CH3)3). Anal. calcd
for C28H30P2N2: C 73.67, H 6.62, N 6.14. Found: C 73.42, H
6.51, N 5.98.
Synthesis
Preparation of (Ph2P)2N–N(Me)H (4). To 50 mL of stirred
anhydrous tetrahydrofuran at
0 °C was added excess
triethylamine (15 mL), methylhydrazine (0.53 g, 5.70 mmol)
and diphenylphosphine chloride (1.06 mL, 5.70 mmol). The
resulting suspension was stirred for 30 minutes and a second
equivalent of diphenylphosphine chloride (1.06 mL, 5.70 mmol)
was added dropwise. The mixture was stirred for 24 hours,
solvent removed under vacuum, and the residue washed with
60 mL and 25 mL portions of degassed deionised water. The
resulting oil was extracted with 2 × 5 mL portions of anhydrous
ethanol which upon standing crystallised the title compound as
small colourless cubes in 29% yield (0.14 g, 0.33 mmol). 1H
NMR (C6D6, 400 MHz): δ 7.52–7.57 (m, 8H, aryl-H), 6.99–7.11
(m, 12H, aryl-H), 3.40 (bd, J = 5.3 Hz, 1H, NH), 2.15 (d, J = 4.5
Hz, 3H, NCH3) 31P NMR (C6D6, 161.8 MHz): δ 65.85 (s). 13C
NMR (C6D6, 100.5 MHz): δ 140.3 (t, J = 5.7 Hz, aryl-Cipso), 133.2
(t, J = 11.1 Hz, aryl-C), 128.8 (s, aryl-C), 128.0 (s, aryl-C), 39.1 (t,
J = 5.6 Hz, NCH3). Anal. calcd for C25H24P2N2: C 72.44, H 5.84,
N 6.76. Found: C 73.02, H 5.87, N 6.76.
Preparation
of
Ph2PN(2-pyridine)PPh2
(9).
2-
Aminopyridine (0.25 g, 2.66 mmol) was dissolved into 30 mL
of anhydrous dichloromethane, to which excess triethylamine
(5 mL) was added. One equivalent of diphenylphosphine
chloride (0.50 mL, 2.66 mmol) was added dropwise with
stirring. The mixture was allowed to stir for 30 minutes and a
second equivalent of diphenylphosphine chloride (0.50 mL,
2.66 mmol) was added, upon which the mixture was stirred
for 3 days. The resulting precipitate was removed via cannula
filtration and the filtrate passed through a 2 cm plug of
neutral alumina. The volatiles were removed in vacuo to give a
yellow residue which was washed with 3 × 5 mL portions of
anhydrous ethanol to give the title compound as a white solid
Preparation
of
(Ph2P)2N–N(cyclohexyl)H
(5).
Cyclohexylhydrazine hydrochloride (0.30 g, 1.99 mmol) was
suspended in 30 mL of anhydrous dichloromethane with
stirring. Excess triethylamine (1.4 mL) was added, generating
a colourless solution. Diphenylphosphine chloride (0.74 mL,
1
in 87% yield (1.11 g, 2.31 mmol). H NMR (CDCl3, 400 MHz):
This journal is © The Royal Society of Chemistry 2014
Catal. Sci. Technol., 2014, 4, 2574–2588 | 2583

View Article Online
Paper
Catalysis Science & Technology
δ 8.18 (ddd, J = 0.8, 2.1, 5.0 Hz, 1H, pyridine-H), 7.45–7.51
(m, 8H, aryl-H), 7.17–7.28 (m, 13H, aryl-H), 6.72–6.77 (m, 2H,
pyridine-H). 31P NMR (CDCl3, 161.8 MHz): δ 59.31 (s). 13C
NMR (CDCl3, 100.5 MHz): δ 159.5 (t, J = 5.0 Hz, pyrindyl-Cipso),
147.9 (s, pyrindyl-C), 138.1 (m, aryl-Cipso), 136.2 (s, pyrindyl-C),
133.1 (d, J = 22.6 Hz, aryl-C), 128.9 (s, aryl-C), 127.9 (t, J = 2.4 Hz,
aryl-C), 118.3 (s, pyrindyl-C), 117.9 (t, J = 3.3 Hz, pyrindyl-C).
Anal. calcd for C29H24P2N2: C 75.30, H 5.23, N 6.06. Found: C
75.02, H 5.07, N 6.09.
Preparation of (Ph2P)2NCH2CH2OBz (12). Compound 12
was prepared according to the procedure outlined for
compound 10, employing O-benzyl-2-aminoethanol (4.97 g,
32.9 mmol). Compound 12 was isolated as a colourless liquid
in 20% yield (3.33 g, 6.42 mmol) after washing with anhydrous
methanol. 1H NMR (CD2Cl2, 299.89 MHz): δ 7.11–7.43 (m, 25H,
aryl-H), 4.14 (s, 2H, OCH2Ph), 3.50 (m, 2H, (Ph2P)2NCH2), 3.04
(t, J = 6.90 Hz, 2H, CH2OBz). 31P NMR (CD2Cl2, 161.8 MHz): δ
63.60 (s). 13C NMR (CD2Cl2, 75.41 MHz): δ 139.8 (m, aryl-Cipso),
138.8 (m, aryl-Cipso), 133.1 (m, aryl-C), 129.2 (s, aryl-C), 128.5
(m, aryl-C), 127.9 (s, aryl-C), 127.7 (s, aryl-C), 73.11 (s, OCH2Ph),
70.1 (m, (Ph2P)2NCH2), 51.84 (t, J = 9.7 Hz, CH2OBz). MS
(electrospray): m/z 520.1 [M + H]+.
Preparation of O-trimethylsilyl-2-aminophenol. 2-Aminophenol
(3.18 g, 29.1 mmol) was degassed with five vacuum/argon
cycles and suspended in 60 mL of dichloromethane.
Triethylamine (5 mL) was added and stirring was commenced.
Trimethylsilyl chloride (4.85 mL, 38.4 mmol) was added as a
stream and the resulting suspension was stirred for 1 hour.
The resulting triethylamine hydrochloride was removed via
cannula filtration and the yellow supernatant concentrated in
vacuo to give a yellow liquid. The liquid was flash distilled at
95 °C under full pump vacuum to give the title compound as a
Preparation of (Ph2P)2NCH2CH2N(i-Pr)2 (13). Compound 13
was prepared according to the procedure outlined for com-
pound 4, employing N,N-diisopropylethylenediamine (0.38 g,
2.64 mmol). Compound 13 was isolated as colourless plates
1
in 24% yield (0.32 g, 0.63 mmol). H NMR (CDCl3, 400 MHz):
δ 7.35–7.44 (m, 8H, aryl-H), 7.26–7.31 (m, 12H, aryl-H), 3.26
(m, 2H, (Ph2P)2NCH2), 2.67 (septet, J = 6.6 Hz, 2H, NCH(CH3)2),
2.15 (m, 2H, (i-Pr)2NCH2), 0.73 (d, J = 6.2 Hz, 12H, NCH(CH3)2).
31P NMR (CDCl3, 161.8 MHz): δ 62.06 (s). 13C NMR (CDCl3,
100.5 MHz): δ 139.7 (d, J = 13.4 Hz, aryl-Cipso), 132.8 (t, J =
11.2 Hz, aryl-C), 128.7 (s, aryl-C), 128.1 (t, J = 3.0 Hz, aryl-C), 54.3
(t, J = 10.5 Hz, (Ph2P)2NCH2), 48.9 (s, NCH(CH3)2), 46.4 (t, J =
3.9 Hz, (i-Pr)2NCH2), 20.9 (s, NCH(CH3)2). Anal. calcd
for C32H38P2N2: C 74.98, H 7.47, N 5.46. Found: C 74.78, H
7.77, N 5.55.
1
pale yellow liquid in 43% yield (2.29 g, 12.6 mmol). H NMR
(CD2Cl2, 299.89 MHz): δ 6.60–6.83 (m, 4H, aryl-H), 3.75 (bs, 2H,
NH2), 0.32 (s, 9H, Si(CH3)3). 13C NMR (CD2Cl2, 75.41 MHz): δ
142.9 (aryl-Cipso), 138.9 (aryl-Cipso), 115.7, 118.4, 119.0, 122.3
(aryl-C), 0.43 (Si(CH3)3). MS (electrospray): m/z 182.0 [M + H]+.
Preparation of (Ph2P)2N(2-(Me3SiO)-phenyl) (10). O-
Trimethylsilyl-2-aminophenol (4.79 g, 32.9 mmol) was dissolved
in 50 mL of anhydrous dichloromethane with stirring. To this
was added diphenylphosphine chloride (4.64 mL, 25.2 mmol)
dropwise. The resulting suspension was stirred overnight. The
organic phase was then washed with 3 × 50 mL of degassed,
deionised water, dried on magnesium sulphate and filtered
before the volatiles were removed under reduced pressure. The
resulting residue was washed with 2 × 10 mL portions of diethyl
ether to give 7 as a white solid in 76% yield (5.29 g, 9.73 mmol).
1H NMR (CD2Cl2, 299.89 MHz): δ 7.20–7.40 (m, 20H, aryl-H),
7.00 (m, 1H, aryl-H), 6.78 (d, J = 7.80 Hz, aryl-H), 6.54 (m, 2H,
aryl-H), 0.09 (s, 9H, Si(CH3)3). 31P NMR (CD2Cl2, 161.8 MHz): δ
62.95 (s). 13C NMR (CD2Cl2, 75.41 MHz): δ 152.8 (s, aniline aryl-
Cipso), 140.6 (m, aryl-Cipso), 139.1 (m, aniline aryl-Cipso), 133.7
(broad m, aryl-C), 131.9 (m, aniline aryl-C), 127.8–129.8 (broad
m, aryl-C), 126.8 (s, aniline aryl-C), 120.3 (s, aniline aryl-C),
118.4 (s, aniline aryl-C), 0.64 (s, Si(CH3)3). Anal. calcd for
C33H33P2NSiO: C 72.11, H 6.05, N 2.55. Found: C 72.02, H 5.98,
N 2.37.
Preparation of (Ph2P)2NCH2CH2OSiMe3 (11). Compound
11 was prepared according to the procedure outlined for
compound 10, employing O-trimethylsilyl-2-aminoethanol
(1.68 g, 12.6 mmol). Compound 11 was isolated as a colourless
liquid in 46% yield (2.92 g, 5.83 mmol) after washing with
anhydrous methanol. 1H NMR (CD2Cl2, 299.89 MHz): δ 7.25–7.43
(m, 20H, aryl-H), 3.39 (m, 2H, (Ph2P)2NCH2), 3.12 (t, J = 2.81 Hz,
2H, CH2OSiMe3), 0.11 (s, 9H, Si(CH3)3). 31P NMR (CD2Cl2,
161.8 MHz): δ 62.83 (s). 13C NMR (CD2Cl2, 75.41 MHz): δ 139.9
(m, aryl-Cipso), 133.1 (m, aryl-C), 128.2 (s, aryl-C), 128.5 (t, J = 3.5 Hz,
aryl-C), 61.9 (m, (Ph2P)2NCH2), 46.6 (s, CH2OSiMe3), −0.66 (s,
9H, Si(CH3)3). MS (electrospray): m/z 502 [M + H]+.
Preparation
of
(R,R)-1,2-bis[(2-
hydroxyphenyl)(phenylphosphino)]ethane (16). R,R-DIPAMP
(0.11 g, 0.24 mmol) was degassed with five vacuum/argon
cycles before being dissolved in 5 mL of dichloromethane.
The solution was cooled to −95 °C and boron tribromide
(0.11 mL, 1.09 mmol) was added. The solution was allowed
to return to room temperature and was stirred overnight. The
volatiles were then removed under reduced pressure yielding
a white solid which was suspended in 8 mL of degassed,
deionised water and heated to 90 °C for four hours. Upon
cooling, solid sodium hydrogen carbonate was added until a
neutral pH was achieved. The aqueous solution was then
extracted with 3 × 20 mL of diethyl ether and 1 × 20 mL of
dichloromethane. The organic phases were combined, dried
on magnesium sulphate, filtered and the solvent removed in
vacuo to give the title product as a colourless solid in 54% yield
(54.6 mg, 0.127 mmol). 1H NMR (CD2Cl2, 299.89 MHz): δ 7.25–7.32
(m, 12H, aryl-H), 7.07–7.11 (m, 2H, aryl-H), 6.86–6.91 (m, 4H,
aryl-H), 6.60 (bs, 2H, OH), 2.17 (m, 4H, CH2). 31P NMR (CD2Cl2,
161.8 MHz): δ −40.37 (s). 13C NMR (CD2Cl2, 75.41 MHz): δ
160.2 (t, J = 9.7 Hz, aryl-Cipso), 137.3 (m, aryl-Cipso), 133.2, 131.8,
129.0, 121.4, 115.8 (s, aryl-C), 132.3 (t, J = 9.1 Hz, aryl-C), 128.9
(t, J = 3.4 Hz, aryl-C), 121.2 (m, aryl-Cipso), 22.8 (m, CH2). MS
(electrospray): m/z 431 [M + H]+. Anal. calcd for C26H24P2O2: C
72.55, H 5.62. Found: C 71.64, H 5.39.
Preparation of Ph2PNP(2-hydroxycyclohexyl)Ph2 (17). N,
N-Bis(diphenylphosphino)amine (0.50 g, 1.30 mmol) was
dissolved into 20 mL of tetrahydrofuran and cooled to −95 °C
2584 | Catal. Sci. Technol., 2014, 4, 2574–2588
This journal is © The Royal Society of Chemistry 2014

View Article Online
Catalysis Science & Technology
Paper
with stirring. n-Butyllithium (0.66 mL, 1.05 mmol, 1.6 M in
hexanes) was added dropwise to the solution. The resulting
yellow solution was stirred for 30 minutes at −95 °C then
allowed to return to room temperature where it was stirred
for 5 hours. Cyclohexene oxide, freshly distilled from phos-
phorous pentoxide, was added and the mixture was heated to
67 °C for two days. The orange solution was then cooled to
room temperature and quenched with 2 mL of degassed,
deionised water. The volatiles were removed from the pale
yellow solution to give a cream solid which was extracted with
2 × 10 mL of toluene. The organic extracts were combined,
dried on sodium sulphate, filtered and then concentrated to
give a colourless residue. The residue was washed with 3 × 5 mL
portions of diethyl ether to give the title compound as a
colourless powder in 44% yield (0.28 g, 0.58 mmol). Crystals
suitable for X-ray diffraction were prepared by recrystallisation
from hot methanol. 1H NMR (CD2Cl2, 299.89 MHz): δ 7.12–7.72
(m, 20H, aryl-H), 3.73 (m, 1H, CH-N), 2.72 (m, 1H, CH–OH),
2.03 (m, 1H, cyclohexyl-H), 1.16 (s, 1H, OH), 0.78–1.76 (m, 7H,
cyclohexyl-H). 31P NMR (CD2Cl2, 161.8 MHz): δ 38.70 (d, J = 94
Hz, PN), 31.87 (d, J = 94 Hz, P-N). 13C NMR (CD2Cl2, 75.41
MHz): δ 132.3 (dod, J = 2.9, 9.1 Hz, aryl-C), 132.0 (dod, J = 2.9,
9.1 Hz, aryl-C), 128.5 (d, J = 4.0 Hz, aryl-C), 128.3 (d, J = 4.0 Hz,
aryl-C), 127.5, 128.0, 128.1, 128.7, 129.0, 129.1, 129.2, 129.8,
130.0, 130.9, 131.2, 134.0, 134.1 128.5 (aryl-C), 24.5, 25.9, 26.1,
26.3, 36.1, 36.3, 42.5, 43.6, 69.9, 70.0 (cyclohexyl-C). MS (electro-
spray): m/z 484.1 [M + H]+. Anal. calcd for C30H31P2NO: C 74.51,
H 6.47, N 2.90. Found: C 74.74, H 6.55, N 2.76.
(CD2Cl2, 75.41 MHz): δ 110.1, 115.1, 118.0, 118.1, 119.2,
121.9, 123.8, 128.6, 128.8, 128.9, 129.0, 129.1, 129.4, 130.4,
130.7, 130.9, 131.2, 131.3, 131.4, 131.5, 131.6, 132.0, 134.6,
134.7 (aryl-C). MS (electrospray): m/z 478 [M + H]+. Anal.
calcd for C30H25P2NO: C 75.46, H 5.28, N 2.93. Found: C
75.25, H 5.23, N 2.80.
Representative procedure for ethylene oligomerisation/
polymerisation
A 0.45 L stainless steel Parr Reactor, fully temperature and
pressure controlled and equipped with solvent/catalyst injec-
tion port and stirrer, was preheated to 120 °C and evacuated
for a minimum of two hours before being purged with argon.
The reactor was cooled to the appropriate temperature and
charged with a total of 200 mL cyclohexane and the required
amount of activator. Stirring was then commenced and the
reactor charged with 4/5 the total required pressure of ethyl-
ene. The metal source and ligand solutions were then com-
bined in the injection port and were injected into the reactor
via a positive pressure gradient yielding the desired ethylene
pressure. During the reaction, the pressure was kept constant
with a replenishing flow of ethylene. Samples of the liquid
phase were taken via an outlet every five minutes and imme-
diately filtered and analysed by gas chromatography. After
30 minutes run time the replenishment of ethylene was
ceased and the reactor cooled before purging of excess ethyl-
ene to atmospheric pressure. The solid and liquid phases
were collected in a preweighed beaker and dried over night at
60 °C with reduced pressure (~50 mbar) to quantify non-
volatile species.
Preparation
of
(Ph2P)2N(2-(Me3SiO)-
phenyl)titanium(IV)tetrachloride (18). Compound 10 (0.25 g,
0.45 mmol) was taken up in 5 mL of dichloromethane and
added slowly to TiCl4 (0.06 mL, 0.50 mmol) in 10 mL of
dichloromethane at −95 °C. The resulting red solution was
allowed to return to room temperature and stirred overnight.
The volatiles were removed in vacuo and the remaining red
solid was washed with 10 mL of toluene to yield the title
compound in 97% yield (0.33 g, 0.44 mmol). Crystals suitable
for X-ray diffraction were grown from a hot solution of
X-ray crystallography
Data for 4, 13 and 17 were collected at −173 °C on crystals
mounted on a Hampton Scientific cryoloop at the MX1
beamline of the Australian Synchrotron, while data for 18
and 19 were similarly collected on the MX2 beamline.80 The
structures were solved by direct methods with SHELXS-97,
refined using full-matrix least-squares routines against F2
with SHELXL-97,81 and visualised using X-SEED.82 All non-
hydrogen atoms were refined anisotropically. Compound 17
featured disorder of a PPh2 unit that was modelled as a two
site complementary occupancy of the 5 remote carbon atoms
of a phenyl ring and associated hydrogen atoms. Details of
the disorder modelling are provided in the cif file. PH, NH
and OH protons were located and positionally refined.
Refinement for 18 is presented in P21, as P21/m leads to dis-
order in the DCM lattice solvent across the mirror plane (not
apparent in the P21 refinement) and significantly higher R
(>10%). All other hydrogen atoms were placed in calculated
positions and refined using a riding model with fixed C–H
distances of 0.95 Å (sp2CH), 0.99 Å (CH2), 0.98 Å (CH3). The
thermal parameters of all hydrogen atoms were estimated as
Uiso(H) = 1.2Ueq(C) except for CH3 where Uiso(H) = 1.5Ueq(C).
A summary of crystallographic data is given below. CCDC
995241–995245.
1
dichloromethane/petroleum spirits. H NMR (CD2Cl2, 299.89
MHz): δ 7.18–7.80 (m, 22H, aryl-H), 6.88 (dt, J = 1.50, 8.55 Hz,
1H, aryl-H), 6.60 (dd, J = 1.50, 8.10 Hz, 1H, aryl-H), −0.30
(s, 9H, Si(CH3)3). 31P NMR (dichloromethane-d2, 161.8 MHz):
δ 61.38. 13C NMR (dichloromethane-d2, 75.41 MHz): δ 119.4,
120.7, 128.4, 128.5, 128.6, 129.8, 130.6, 131.8, 131.9, 134.0,
154.2 (aryl-C), −0.00 (s, 9H, Si(CH3)3). Anal. calcd for
C33H33P2NSiOTiCl4(dichloromethane)2/3: C 50.83, H 4.35, N 1.76.
Found: C 50.95, H 4.42, N 1.75.
Preparation of (Ph2P)N(2-(O)-phenyl)PHPh2 (19). A portion
of 10 (0.11 g, 0.20 mmol) was dissolved in hot methanol and
stirred overnight. The volatiles were removed in vacuo and
the resulting off-white solid was recrystallised from diethyl
ether to give 19 in 50% yield (0.05 g, 0.10 mmol) as colourless
needles suitable for X-ray diffraction. 1H NMR (CD2Cl2,
299.89 MHz): δ 6.31–8.14 (m, 24H, aryl-H), 3.65 (d, JP,H = 14.1
Hz, 1H, P-H). 31P NMR (CD2Cl2, 161.8 MHz): δ 37.52 (d, JP,P
=
202 Hz, P(III)), −34.71 (d, JP,P = 202 Hz, P(V)). 13C NMR
This journal is © The Royal Society of Chemistry 2014
Catal. Sci. Technol., 2014, 4, 2574–2588 | 2585

View Article Online
Paper
Catalysis Science & Technology
Crystal data for 4: C25H24N2P2, M = 414.40, monoclinic,
9 L. H. Do, J. A. Labinger and J. E. Bercaw, Organometallics,
2012, 31, 5143.
10 J. Rabeah, M. Bauer, W. Baumann, A. E. McConnell,
W. F. Gabrielli, P. B. Webb, D. Selent and A. Bruckner, ACS
Catal., 2013, 3, 95.
11 I. Y. Skobelev, V. N. Panchenko, O. Y. Lyakin, K. P. Bryliakov,
V. A. Zakharov and E. P. Talsi, Organometallics, 2010, 29, 2943.
12 K. Blann, A. Bollmann, J. T. Dixon, A. Neveling,
D. H. Morgan, H. Maumela, E. Killian, F. Hess, S. Otto,
L. Pepler, H. Mahomed and M. Overett, WO 2004/056479,
Sasol Technology, 2004.
a = 14.4930(3), b = 10.3290(8), c = 15.3920(3) Å, β = 109.523(2)°,
U = 2171.68(18) Å3, T = 100 K, space group P21/n (no. 14), Z = 4,
35591 reflections measured, 5447 unique (Rint = 0.0421),
5163 > 4σ(F), R = 0.0366 (observed), Rw = 0.0960 (all data).
Crystal data for 13: C32H38N2P2, M = 512.58, triclinic, a =
10.2160(7), b = 10.3180(3), c = 14.3790(6) Å, α = 79.664(2), β =
76.635(2), γ = 71.842(5)°, U = 1391.70(12) Å3, T = 100 K, space
¯
group P1 (no. 2), Z = 2, 23 874 reflections measured, 6344
unique (Rint = 0.0407), 5917 > 4σ(F), R = 0.0390 (observed),
Rw = 0.1072 (all data). Crystal data for 17: C30H31NOP2, M =
483.50, monoclinic, a = 8.6080(7), b = 22.5050(8), c = 13.7650(8) Å,
13 A. Bollmann, K. Blann, J. T. Dixon, F. M. Hess, E. Killian,
H. Maumela, D. S. McGuinness, D. H. Morgan, A. Neveling,
S. Otto, M. Overett, A. M. Z. Slawin, P. Wasserscheid and
S. Kuhlmann, J. Am. Chem. Soc., 2004, 126, 14712.
14 K. Blann, A. Bollmann, H. de Bod, J. T. Dixon, E. Killian,
P. Nongodlwana, M. C. Maumela, H. Maumela, A. E. McConnell,
D. H. Morgan, M. Overett, M. Prétorius, S. Kuhlmann and
P. Wasserscheid, J. Catal., 2007, 249, 244.
3
¯
β = 106.913(2)°, U = 2551.3(3) Å , T = 100 K, space group P21/n
(no. 14), Z = 4, 41 516 reflections measured, 6283 unique (Rint
=
0.0328), 6184 > 4σ(F), R = 0.0475 (observed), Rw = 0.1066 (all
data). Crystal data for 18: C33H33Cl4NOP2SiTi·0.5(CH2Cl2) M =
781.80, monoclinic, a = 11.5300(8), b = 15.3370(7), c =
20.8500(17) Å, β = 103.737(3)°, U = 3581.6(4) Å3, T = 100 K,
¯
space group P21 (no. 4), Z = 4, 77 400 reflections measured,
20509 unique (Rint = 0.0647), 18 238 > 4σ(F), R = 0.0450
(observed), Rw = 0.1448 (all data). Crystal data for 19:
C30H25NOP2, M = 477.45, triclinic, a = 8.701(2), b = 10.6660(10),
15 O. L. Sydora, T. C. Jones, B. L. Small, A. J. Nett, A. A. Fischer
and M. J. Carney, ACS Catal., 2012, 2, 2452.
16 Y. Shaikh, K. Albahily, M. Sutcliffe, V. Fomitcheva,
S. Gambarotta, I. Korobkov and R. Duchateau, Angew. Chem.,
Int. Ed., 2012, 51, 1366.
c = 13.9380(13) Å, α = 70.2480(10), β = 84.203(3), γ = 81.578(3)°,
3
¯
U = 1202.4(3) Å , T = 100 K, space group P1 (no. 2), Z = 2,
25361 reflections measured, 6695 unique (Rint = 0.0370),
6170 > 4σ(F), R = 0.0435 (observed), Rw = 0.1145 (all data).
17 Y. Shaikh, J. Gurnham, K. Albahily, S. Gambarotta and
I. Korobkov, Organometallics, 2012, 31, 7427.
18 T. E. Stennett, T. W. Hey, L. T. Ball, S. R. Flynn, J. Radcliffe,
C. L. McMullin, R. L. Wingad and D. F. Wass,
ChemCatChem, 2013, 5, 2946.
Acknowledgements
The Australian Research Council is thanked for financial sup-
port through grant FT100100609 to DSM and an Australian
Postgraduate Award to JAS. We thank Sasol Technology Ltd
for funding and permission to publish this work. Albemarle
Asia Pacific is thanked for donating the MAO solution. Noel
Davies and Thomas Rodemann of the UTas Central Science
Laboratory are thanked for MS and elemental analysis mea-
surements respectively. Christina Wronna at the Lehrstühle
für Anorganische Chemie, Friedrich-Alexander Universität
Erlangen-Nürnberg is thanked for elemental analysis mea-
surements. Data for the structures of complexes 4, 13 and
17–19 were obtained on the MX1 and MX2 beamlines at the
Australian Synchrotron, Victoria, Australia.
19 V. Aliyev, M. H. Al-Hazmi, F. Mosa, P. M. Fritz, H. Bölt,
A. Wöhl, W. Müller, F. Winkler, A. Wellenhofer,
U. Rosenthal, B. H. Müller, M. Hapke and N. Peulecke, WO
09/068157, Linde and Sabic, 2009.
20 P. M. Fritz, H. Bölt, A. Wöhl, W. Müller, F. Winkler,
A. Wellenhofer, U. Rosenthal, M. Hapke, N. Peulecke,
M. H. Al-Hazmi, V. O. Aliyev and F. M. Mosa WO 09/006979,
Linde and Sabic, 2009.
21 S. Peitz, N. Peulecke, B. R. Aluri, S. Hansen, B. H. Müller,
A. Spannenberg, U. Rosenthal, M. H. Al-Hazmi, F. M. Mosa,
A. Wöhl and W. Müller, Eur. J. Inorg. Chem., 2010, 1167.
22 A. Dulai, C. L. McMullin, K. Tenza and D. F. Wass,
Organometallics, 2011, 30, 935.
23 B. R. Aluri, N. Peulecke, S. Peitz, A. Spannenberg,
B. H. Müller, S. Schulz, H.-J. Drexler, D. Heller, M. H. Al-
Hazmi, F. Mosa, A. Wöhl, W. Müller and U. Rosenthal,
Dalton Trans., 2010, 39, 7911.
24 D. S. McGuinness, M. Overett, R. P. Tooze, K. Blann,
J. T. Dixon and A. M. Z. Slawin, Organometallics, 2007, 26,
1108.
Notes and references
1 T. Agapie, Coord. Chem. Rev., 2011, 255, 861.
2 A. W. Al-Sa'doun, Appl. Catal., A, 1993, 105, 1.
3 J. T. Dixon, M. J. Green, F. M. Hess and D. H. Morgan,
J. Organomet. Chem., 2004, 689, 3641.
4 D. S. McGuinness, Chem. Rev., 2011, 111, 2321.
5 P. W. N. M. van Leeuwen, N. D. Clément and
M. J.-L. Tschan, Coord. Chem. Rev., 2011, 255, 1499.
6 G. P. Belov, Pet. Chem., 2012, 52, 139.
25 D. S. McGuinness, A. J. Rucklidge, R. P. Tooze and
A. M. Z. Slawin, Organometallics, 2007, 26, 2561.
26 L. E. Bowen, M. F. Haddow, A. G. Orpen and D. F. Wass,
Dalton Trans., 2007, 1160.
7 D. F. Wass, Dalton Trans., 2007, 816.
8 W. H. Monillas, J. F. Young, G. P. A. Yap and
K. H. Theopold, Dalton Trans., 2013, 42, 9198.
27 A. J. Rucklidge, D. S. McGuinness, R. P. Tooze,
A. M. Z. Slawin, J. D. A. Pelletier, M. J. Hanton and
P. B. Webb, Organometallics, 2007, 26, 2782.
2586 | Catal. Sci. Technol., 2014, 4, 2574–2588
This journal is © The Royal Society of Chemistry 2014

View Article Online
Catalysis Science & Technology
Paper
28 T. E. Stennett, M. F. Haddow and D. F. Wass,
Organometallics, 2012, 31, 6960.
29 A. Sattler, J. A. Labinger and J. E. Bercaw, Organometallics,
2013, 32, 6899.
48 P. A. Aguirre, C. A. Lagos, S. A. Moya, C. Zúñiga, C. Vera-Oyarce,
E. Sola, G. Peris and J. C. Bayón, Dalton Trans., 2007, 5419.
49 S. M. Aucott, M. L. Clarke, A. M. Z. Slawin and
J. D. Woollins, J. Chem. Soc., Dalton Trans., 2001, 972.
50 A. N. Kornev, V. V. Sushev, Y. S. Panova, N. V. Belina,
O. V. Lukoyanova, G. K. Fukin, S. Y. Ketlov, G. A. Abakumov,
P. Lönnecke and E. Hey-Hawkins, Inorg. Chem., 2012, 51, 874.
51 P. R. Elowe, C. McCann, P. G. Pringle, S. K. Spitzmesser and
J. E. Bercaw, Organometallics, 2006, 25, 5255.
30 Y. Chen, E. Callens, E. Abou-Hamad, N. Merle,
A. J. P. White, M. Taoufik, C. Copéret, E. Le Roux and
J.-M. Basset, Angew. Chem., Int. Ed., 2012, 51, 11886.
31 S. Peitz, N. Peulecke, B. R. Aluri, B. H. Müller,
A. Spannenberg, U. Rosenthal, M. H. Al-Hazmi, F. M. Mosa,
A. Wöhl and W. Müller, Organometallics, 2010, 29, 5263.
32 S. Peitz, N. Peulecke, B. R. Aluri, B. H. Müller,
A. Spannenberg, U. Rosenthal, M. H. Al-Hazmi, F. M. Mosa,
A. Wöhl and W. Müller, Chem. – Eur. J., 2010, 16, 12127.
33 S. Peitz, N. Peulecke, B. Müller, A. Spannenberg,
H.-J. Drexler, U. Rosenthal, M. H. Al-Hazmi, K. Al-Eidan,
A. Wöhl and W. Müller, Organometallics, 2011, 30, 2364.
34 B. H. Müller, N. Peulecke, A. Spannenberg, U. Rosenthal,
M. H. Al-Hazmi, R. Schmidt, A. Wöhl and W. Müller,
Organometallics, 2012, 31, 3695.
35 A. Wöhl, W. Müller, S. Peitz, N. Peulecke, B. R. Aluri,
B. H. Müller, D. Heller, U. Rosenthal, M. H. Al-Hazmi and
F. M. Mosa, Chem. – Eur. J., 2010, 16, 7833.
36 W. Müller, A. Wöhl, S. Peitz, N. Peulecke, B. R. Aluri,
B. H. Müller, D. Heller, U. Rosenthal, M. H. Al-Hazmi and
F. M. Mosa, ChemCatChem, 2010, 2, 1130.
37 V. O. Aliyev, M. H. Al-Hazmi, F. Mosa, U. Rosenthal,
B. H. Müller, M. Hapke, N. Peulecke, A. Wöhl, P. M. Fritz,
H. Bölt, W. Müller, F. Winkler and A. Wellenhofer, WO 09/
121456, Linde and Sabic, 2009.
52 S. Torker, A. Müller, R. Sigrist and P. Chen, Organometallics,
2010, 29, 2735.
53 A. N. Kornev, N. V. Belina, V. V. Sushev, J. S. Panova,
O. V. Lukoyanova, S. Y. Ketkov, G. K. Fukin, M. A. Lopatin
and G. A. Abakumov, Inorg. Chem., 2010, 49, 9677.
54 V. V. Sushev, N. V. Belina, G. K. Fukin, Y. A. Kurskiy,
A. N. Kornev and G. A. Abakumov, Inorg. Chem., 2008,
47, 2608.
55 A. D. Chowdhury, S. M. Mobin, S. Mukherjee, S. Bhaduri
and G. K. Lahiri, Eur. J. Inorg. Chem., 2011, 3232.
56 N. J. Rahier, J.-N. Volle, M. A. Lacour and M. Taillefer,
Tetrahedron, 2008, 64, 6645.
57 D. A. Dickie and R. A. Kemp, Acta Crystallogr., Sect. E: Struct.
Rep. Online, 2008, 64, o1449.
58 M. Nieger, N. Poetschke, E. Niecke and K. Airola, Z.
Kristallogr. - New Cryst. Struct., 1997, 212, 247.
59 M. Wada and A. Tsuboi, J. Chem. Soc., Perkin Trans. 1,
1987, 151.
60 S.-I. Yamamoto, K. Okuma and H. Ohta, Bull. Chem. Soc.
Jpn., 1988, 61, 4476.
38 N. Peulecke, B. H. Müller, S. Peitz, B. R. Aluri, U. Rosenthal,
A. Wöhl, W. Müller, M. H. Al-Hazmi and F. Mosa,
ChemCatChem, 2010, 2, 1079.
39 A. Wöhl, U. Rosenthal, B. H. Müller, N. Peulecke, S. Peitz,
W. Müller, H. Bölt, A. Meiswinkel, B. R. Aluri, M. H. Al-Hazmi,
M. Al-Masned, K. Al-Eidan and F. M. Mosa WO 2010/115520,
Linde and Sabic, 2010.
40 B. H. Müller, N. Peulecke, S. Peitz, B. R. Aluri, U. Rosenthal,
M. H. Al-Hazmi, F. Mosa, A. Wöhl and W. Müller, Chem. –
Eur. J., 2011, 17, 6935.
61 H. Christol, D. Greley, M. R. Darvich, F. Fallouh, F. Plenat
and H.-J. Cristau, Bull. Soc. Chim. Fr., 1989, 4, 477.
62 H.-J. Cristau, A. Hervé and L. D. Virieux, Synthesis, 2003,
14, 2216.
63 J. A. Suttil, D. S. McGuinness, M. Pichler, M. G. Gardiner,
D. H. Morgan and S. J. Evans, Dalton Trans., 2012, 41, 6625.
64 M. A. H. Laramay and J. G. Verkade, J. Am. Chem. Soc., 1990,
112, 9421.
65 S. Kojima, K. Kajiyama, M. Nakamoto and K.-Y. Akiba, J. Am.
Chem. Soc., 1996, 118, 12866.
41 A. Jabri, P. Crewdson, S. Gambarotta, I. Korobkov and
R. Duchateau, Organometallics, 2006, 25, 715.
66 K. Ishihara, Y. Karumi, S. Kondo and H. Yamamoto, J. Org.
Chem., 1998, 63, 5692.
42 A. Jabri, C. B. Mason, Y. Sim, S. Gambarotta, T. J. Burchell
and R. Duchateau, Angew. Chem., Int. Ed., 2008, 47, 9717.
43 I. Vidyaratne, G. B. Nikiforov, S. I. Gorelsky, S. Gambarotta,
R. Duchateau and I. Korobkov, Angew. Chem., Int. Ed., 2009,
48, 6552.
44 J. Zhang, A. Li and T. S. A. Hor, Organometallics, 2009, 28, 2935.
45 S. V. Kulangara, C. Mason, M. Juba, Y. Yang, I. Thapa,
S. Gambarotta, I. Korobkov and R. Duchateau, Organometallics,
20121, 31, 6438.
46 V. V. Sushev, A. N. Kornev, Y. A. Min'ko, N. V. Belina,
Y. A. Kurskiy, O. V. Kuznetsova, G. K. Fukin, E. V. Baranov,
V. K. Cherkasov and G. A. Abakumov, J. Organomet. Chem.,
2006, 691, 879.
67 S. Kojima, M. Sugino, S. Matsukawa, M. Nakamoto and
K.-Y. Akiba, J. Am. Chem. Soc., 2002, 124, 7674.
68 S.-Y. Nakafuji, J. Kobayashi, T. Kawashima and
M. W. Schmidt, Inorg. Chem., 2005, 44, 6500.
69 S. Kojima, K. Kajiyama, M. Nakamoto, S. Matsukawa and
K.-Y. Akiba, Eur. J. Org. Chem., 2006, 218.
70 V. R. Chintareddy, A. Ellern and J. G. Verkade, J. Org. Chem.,
2010, 75, 7166.
71 Z. Weng, S. Teo and T. S. A. Hor, Dalton Trans., 2007, 3493.
72 D. Commereuc, S. Drochon and L. Saussine US 6031145,
IFP, 2000.
73 D. Commereuc, S. Drochon and L. Saussine, US 2001/
0023281, IFP, 2001.
47 S. Vargas, M. Rubio, A. Suárez, D. del Río, E. Álvarez and
A. Pizzano, Organometallics, 2006, 25, 961.
74 C. Grove, D. H. Morgan, S. L. Schwikkard and
B. C. B. Bezuidenhoudt, WO 02/083306, Sasol Technology, 2002.
This journal is © The Royal Society of Chemistry 2014
Catal. Sci. Technol., 2014, 4, 2574–2588 | 2587

View Article Online
Paper
Catalysis Science & Technology
75 D. H. Morgan, S. L. Schwikkard, J. T. Dixon, J. J. Nair and
R. Hunter, Adv. Synth. Catal., 2003, 345, 939.
79 J. A. Suttil, M. F. Shaw, D. S. McGuinness, M. G. Gardiner
and S. J. Evans, Dalton Trans., 2013, 42, 9129.
76 M. J. Overett, K. Blann, A. Bollmann, J. T. Dixon,
D. Hassbroek, E. Killian, H. Maumela, D. S. McGuinness and
D. H. Morgan, J. Am. Chem. Soc., 2005, 127, 10723.
77 T. K. Han, S. S. Chae, S. O. Kang, K. R. Wee and S. K. Kim,
WO 2009/022770, SK Energy, 2009.
80 T. M. McPhillips, S. E. McPhillips, H. J. Chiu, A. E. Cohen,
A. M. Deacon, P. J. Ellis, E. Garman, A. Gonzalez,
N. K. Sauter, R. P. Phizackerly, S. M. Soltis and P. Kuhn,
Synchrotron Radiat., 2009, 9, 401.
81 G. M. Sheldrick, SHELX97 Programs for Crystal Structure
Analysis, Universität Göttingen, Germany, 1998.
78 J. A. Suttil, D. S. McGuinness, M. G. Gardiner and S. J. Evans,
Dalton Trans., 2013, 42, 4185.
82 L. J. Barbour, J. Supramol. Chem., 2001, 1, 189.
2588 | Catal. Sci. Technol., 2014, 4, 2574–2588
This journal is © The Royal Society of Chemistry 2014