Mixed-ligand iminopyrrolato-salicylaldiminato group 4 metal complexes: Optimising catalyst structure for ethylene/propylene copolymerisations

Article abstract of DOI:10.1016/j.jorganchem.2007.04.047

Treatment of MCl₃(OC₆H₃-2-^tBu-6-CH{double bond, long}NC₆F₅)(THF) (M = Ti, Zr) with a variety of different potassium iminopyrrolate salts (K⁺[RN{double bond, long}CHC₄H₃N]^-), (R = phenyl, cyclo-hexyl, ethyl) afforded the corresponding titanium and zirconium mixed-ligand complexes MCl₂(N-O)(N-N). The molecular structures of TiCl₂(OC₆H₃-2-^tBu-6-CH{double bond, long}NC₆F₅)(C₂H₅N{double bond, long}CHC₄H₃N) (1c), TiCl₂(OC₆H₃-2-^tBu-6-CH{double bond, long}NC₆F₅)(C₆H₁₁N{double bond, long}CHC₄H₃N) (1b) and ZrCl₂(OC₆H₃-2-^tBu-6-CH{double bond, long}NC₆F₅)(C₆H₁₁N{double bond, long}CHC₄H₃N) (2b) show distorted octahedral geometries with trans-O^-,N^-/cis-Cl₂ arrangements. On activation with MAO the titanium (iminopyrrolato)(salicylaldiminato) complexes show excellent activities in ethylene polymerisation and are significantly more effective ethylene/propylene copolymerisation catalysts, both in terms of activity and propene incorporation, than either of the parent complexes. The ethylene-propylene copolymers show ca. 80% 1,2 regioselectivity and at high propylene incorporation tend towards an alternating structure.

Full text of DOI:10.1016/j.jorganchem.2007.04.047

Journal of Organometallic Chemistry 692 (2007) 4603–4611
www.elsevier.com/locate/jorganchem
Mixed-ligand iminopyrrolato-salicylaldiminato group 4 metal
complexes: Optimising catalyst structure for ethylene/propylene
copolymerisations
a
a
a
a
Lewis M. Broomfield , Yann Sarazin , Joseph A. Wright , David L. Hughes ,
b
b
a,
*
William Clegg , Ross W. Harrington , Manfred Bochmann
a
Wolfson Materials and Catalysis Centre, School of Chemical Sciences and Pharmacy, University of East Anglia, Norwich, NR4 7TJ, UK
b
School of Natural Sciences (Chemistry), University of Newcastle, Newcastle upon Tyne, NE1 7RU, UK
Received 22 March 2007; received in revised form 24 April 2007; accepted 26 April 2007
Available online 10 May 2007
Abstract
Treatment of MCl3(OC6H3-2-tBu-6-CH@NC6F5)(THF) (M = Ti, Zr) with a variety of different potassium iminopyrrolate salts
(K+[RN@CHC4H3N]À), (R = phenyl, cyclo-hexyl, ethyl) afforded the corresponding titanium and zirconium mixed-ligand complexes
MCl2(N–O)(N–N). The molecular structures of TiCl2(OC6H3-2-tBu-6-CH@NC6F5)(C2H5N@CHC4H3N) (1c), TiCl2(OC6H3-2-tBu-6-
CH@NC6F5)(C6H11N@CHC4H3N) (1b) and ZrCl2(OC6H3-2-tBu-6-CH@NC6F5)(C6H11N@CHC4H3N) (2b) show distorted octahedral
geometries with trans-OÀ,NÀ/cis-Cl2 arrangements. On activation with MAO the titanium (iminopyrrolato)(salicylaldiminato) com-
plexes show excellent activities in ethylene polymerisation and are significantly more effective ethylene/propylene copolymerisation cat-
alysts, both in terms of activity and propene incorporation, than either of the parent complexes. The ethylene–propylene copolymers
show ca. 80% 1,2 regioselectivity and at high propylene incorporation tend towards an alternating structure.
ꢀ 2007 Elsevier B.V. All rights reserved.
Keywords: Titanium; Zirconium; Ethene; Copolymerization; Salicylaldiminate; Iminopyrrolate
1. Introduction
fraction of bulky comonomer incorporated, while the activ-
ity is significantly decreased [7]. On the other hand, bis(imi-
nopyrrolato) titanium complexes (structure B) give lower
ethylene polymerisation productivities but achieve higher
incorporation of both cyclic and terminal alkenes in the
polyethylene backbone [8,9].
These limitations can be overcome with hetero-ligated
group 4 metal complexes [10–14] which have been shown
to combine these characteristics and are able to produce
highly active catalysts capable of high levels of comonomer
incorporation. The most successful systems for a-olefin
co-polymerisations are the (iminopyrrolato)(salicylaldimi-
nato)titanium complexes, which havepreviously been shown
to combine high activity with high comonomer incorpora-
tion in copolymerisations of ethylene with 1-hexene,
cyclopentene and norbornene [10]. We report here the syn-
theses and structures of two new titanium (N–N)(N–O)
The development of new, more selective and active
homogeneous polymerisation catalysts remains a driving
force within organometallic chemistry. Of the non-metallo-
cene group 4 metal complexes tested in olefin polymerisa-
tions, octahedral metal dichloride systems containing
bidentate mono-anionic N,NÀ or N,OÀ ligands have
proved to give particularly effective catalysts [1,2]. Bis
(salicylaldiminato) complexes of titanium and zirconium
of type A (Chart 1) are potent catalysts for ethylene homo-
polymerisations, but are inefficient copolymerisation
catalyst [3–6]. They are sensitive to steric factors; for exam-
ple, reducing the size of the ortho-substituent increases the
*
Corresponding author. Tel./fax: +44 1603 592044.
E-mail address: m.bochmann@uea.ac.uk (M. Bochmann).
0022-328X/$ - see front matter ꢀ 2007 Elsevier B.V. All rights reserved.
doi:10.1016/j.jorganchem.2007.04.047

4604
L.M. Broomfield et al. / Journal of Organometallic Chemistry 692 (2007) 4603–4611
tBu
Cl
Cl
O
Ar
N
CH2Cl2
- KCl
Ar
R
N
tBu
M
N
THF
Cl
K
O
+
N
M
N
N
Cl
R
Cl
Ar = C6F5
1a R = C6H5 (M = Ti)
1b R = cyclohexyl (M = Ti)
1c R = Et (M = Ti)
2b R = cyclohexyl (M = Zr)
Chart 1.
mixed-ligand complexes and that of a zirconium analogue.
When activated with methylaluminoxane (MAO) the
titanium complexes are shown to be highly effective ethyl-
ene–propylene copolymerisation catalysts.
19F spectra of all the complexes indicate hindered rotation
of C6F5 groups, rendering the ortho- and meta-fluorines
inequivalent.
The structures of 1b, 1c and 2b were determined by
X-ray crystallography and are shown in Figs. 1–3, respec-
tively. Complexes 1c and 2b have essentially octahedral
geometry, with the two anionic functions trans to one
another while the two chloride ligands are cis. These
structures are consistent with the most stable calculated
geometry for such octahedral complexes bearing two
mono-anionic ligands, and is also the geometry observed
for most bis(salicylaldiminato), bis(iminopyrrolato) as well
as for related mixed-ligand group 4 metal complexes
[4,9–11].
2. Syntheses
The complexes TiCl2(OC6H3-2-tBu-6-CH@NC6F5)-
(R–N@CHC4H3N) (1a, R = Ph; 1b, R = cyclohexyl; 1c,
R = Et) as well as ZrCl2(OC6H3-2-tBu-6-CH@NC6F5)(Cy–
N@CHC4H3N) (2b) were synthesised following previously
published procedures [8,9] by reacting the mono(salicylal-
diminato) complexes MCl3(OC6H3-2-tBu-6-CH@NC6F5)
(THF) (M = Ti, Zr) in dichloromethane with potassium
salts of the corresponding iminopyrroles (Scheme 1). The
products were obtained as dark red to orange crystals.
1
The H NMR and 19F NMR spectra in CDCl3 of the
complexes were consistent with the presence of a single ste-
1
reoisomer at room temperature, with the H NMR spec-
trum showing only two sharp imine peaks, one each for
the iminopyrrolato and the salicylaldiminato ligands. The
tBu
N
O
Cl
Cl
N
Cl
Cl
N
N
R
R
Ti
N
Ti
R
N
R
O
tBu
Fig. 1. ORTEP representation of the structure of 1b showing 50%
probability ellipsoids. Hydrogen atoms have been omitted for clarity.
˚
Ba R = C6H5
Selected bond lengths (A) and angles (ꢁ) with estimated standard
Aa R = C6F5 (M = Ti),
Ab R = C6H5 (M = Ti)
deviations: Ti(1)–Cl(1) 2.282(3), Ti(1)–Cl(2) 2.282(2), Ti(1)–N(1)
2.072(6), Ti(1)–N(2) 2.119(6), Ti(1)–N(3) 2.242(6), Ti(1)–O(1) 1.820(5);
Cl(1)–Ti(1)–Cl(2) 96.84(10), N(1)–Ti(1)–N(2) 76.4(2), N(2)–Ti(1)–N(3)
84.9(2), O(1)–Ti(1)–N(1) 162.7(2).
Bb R = Cyclohexyl
Bc R = Et
Scheme 1.

L.M. Broomfield et al. / Journal of Organometallic Chemistry 692 (2007) 4603–4611
4605
N-C6F5 derivatives 1a–c in the presence of MAO give
highly active catalysts for the homopolymerisation of
ethylene and for ethylene–propene copolymerisations over
a wide compositional range. On the other hand, under
comparable conditions the homopolymerisation of propyl-
ene was unsuccessful; only small traces of polymer were
recovered, even with long reaction times and high catalyst
loadings.
3.1. Ethylene polymerisations
The effect of changing the substituent on the pyrrole-
imino nitrogen atom in 1a–c on the polymerisation of
ethylene has been investigated. Such modification has pre-
viously been found to drastically change the catalytic per-
formance of bis(iminopyrrolato) titanium systems (B),
where it was shown that bulky groups such as SiMe3
almost totally suppress ethylene polymerisation activities
[9]. In marked contrast, all three mixed-ligand pre-catalysts
1a–c, are very highly active, regardless of the nature of the
substituent, whereas the productivities of the zirconium
complex 2b were an order of magnitude lower (Table 1).
Complex 1a was found to be the most effective catalyst,
Fig. 2. ORTEP representation of the structure of 1c showing 50%
probability ellipsoids. Hydrogen atoms have been omitted for clarity.
˚
Selected bond lengths (A) and angles (ꢁ): Ti(1)–Cl(1) 2.3031(8), Ti(1)–Cl(2)
2.2774(7), Ti(1)–N(1) 2.2837(16), Ti(1)–N(2) 2.0736(19), Ti(1)–N(3)
2.1366(18), Ti(1)–O(1) 1.8229(14); Cl(1)–Ti(1)–Cl(2) 95.52(3), N(1)–
Ti(1)–N(2) 83.78(6), N(2)–Ti(1)–N(3) 76.37(7), N(1)–Ti(1)–O(1) 79.59(6).
with an activity of 62,000 kg PE (mol metal)À1 hÀ1 barÀ1
,
followed by 1c and 1b. This order of activity is rather dif-
ferent from that found for the bis(iminopyrrolato) titanium
complexes B where R = cyclohexyl gave the most produc-
tive catalyst, 14,100 kg PE (mol metal)À1 hÀ1 barÀ1 [9].
Increasing the polymerisation temperature had very lit-
tle effect on the activity of catalyst 1a. If one considers
the monomer solubility (see Supporting Information), the
activity of 1b increases twofold as the polymerisation tem-
perature is raised from 20 to 50 ꢁC, whilst the zirconium
analogue 2b shows a fourfold increase. Whereas catalysts
1a, 1b and 2b are stable under catalytic conditions up to
temperatures of 50 ꢁC, this is not the case for the N-ethyl
derivative 1c, the activity of which is approximately halved
on warming from 20 to 50 ꢁC. The lower thermal stability
of 1c can possibly be attributed to the lower steric protec-
tion of the metal centre offered by the ethyl group.
Table 1
Ethylene homopolymerisation results for catalysts 1a–c and 2b/MAOa
Fig. 3. ORTEP representation of the structure of 2b showing 50%
Entry Catalyst/ Temperature/ Time/ Polymer/ Productivityb Mwc
Mw/
probability ellipsoids. Hydrogen atoms have been omitted for clarity.
lmol
ꢁC
min
g
Mn
˚
Selected bond lengths (A) and angles (ꢁ): Zr(1)–Cl(1) 2.4132(11), Zr(1)–
1
2
1a/0.5
1a/0.5
20
50
1.5
1.5
0.74
0.34
62,000
28,000
158,000 2.3
Cl(2) 2.3844(10), Zr(1)–N(1) 2.424(3), Zr(1)–N(2) 2.183(3), Zr(1)–N(3)
2.315(3), Zr(1)–O(1) 1.953(2); Cl(1)–Zr(1)–Cl(2) 96.79(4), N(1)–Zr(1)–
N(2) 82.77(11), N(2)–Zr(1)–N(3) 73.17(12), N(1)–Zr(1)–O(1) 75.12(10).
157,000 2.3
3
4
1b/0.5
1b/0.5
20
50
1.5
1.5
0.26
0.28
22,000
23,000
566,000 2.5
264,000 2.4
5
6
1c/0.5
1c/0.5
20
50
1.5
1.5
0.31
0.05
26,000
5000
275,000 2.5
269,000 2.4
3. Polymerisation studies
7
8
2b/5
2b/5
20
50
1.5
1.5
0.34
0.5
2800
4200
310,000 2.5
212,000 2.2
It has been shown previously that mixed-ligand (imi-
nopyrrolato)(salicylaldiminato) titanium catalysts are more
active ethylene polymerisation catalysts than their zirco-
nium analogues [10,11]. The present studies show that the
Conditions: total volume 250 cm3, 1 bar ethylene, 3 mmol MAO.
In kg PE (mol metal)À1 hÀ1 barÀ1
Measured by gel permeation chromatography calibrated with polystyrene
a
b
c
.
standards.

4606
L.M. Broomfield et al. / Journal of Organometallic Chemistry 692 (2007) 4603–4611
Gel permeation chromatography (GPC) measurements
Propylene incorporation vs. feed ratio
on polyethylene samples produced with 1a–c showed mono-
modal composition, with polydispersities typical of single-
site catalysts. Unlike the bis(salicylaldiminato)Ti catalyst
Aa, catalysts 1a–c are not living. Weight-average molecular
weights of between 158,000 and 566,000 g molÀ1 were
found, with 1a (R = Ph) giving the lowest and 1b (R =Cy)
the highest values.
50
45
40
35
30
25
20
15
10
5
3.2. Ethylene–propylene copolymerisations
Complexes 1a, 1b and 1c/MAO catalyse the copolymer-
isation of ethylene and propylene. The results are summa-
rised in Table 2. The dependence of P incorporation on the
P/E feed ratio is illustrated in Fig. 4. Complexes 1a, 1b and
1c are all highly effective ethylene/propylene copolymerisa-
tion catalysts. Changing the pyrrole-imine nitrogen substitu-
ent R had very little effect on the propylene incorporation.
With a gas feed of 5 L minÀ1 propylene and 0.5 L minÀ1 of
ethylene at 20 ꢁC, all three titanium catalysts incorporate
up to ca. 40 mol% propylene, with activities of ca. 1000 kg
0
0
20
40
60
80
100
Propylene in solution [mol-%]
Fig. 4. Incorporation of propylene (mol%) versus propylene in toluene
solution (mol%) for catalysts 1a (¤), 1b (h) and 1c (m).
systems 1a–c. On the other hand, while Cp2TiCl2 is able
to incorporate high amounts of propylene, it produces
copolymers with far lower molecular weights than those
obtained with 1a–c. Furthermore, although the titanium
bis(iminopyrrolato) catalysts Ba and Bb have previously
been shown to give polymers with good P incorporation,
they suffer from low catalyst activities [6]. Complexes
1a–1c therefore represent a good compromise and combine
high catalytic activities with excellent levels of comonomer
incorporation.
polymer (mol metal)À1 hÀ1 barÀ1
.
The behaviour of 1a–c in ethylene–propylene copoly-
merisations was compared with Cp2TiCl2/MAO (Table 2,
entries 16–18) and with the well-known bis(salicylaldimi-
nato) catalyst Aa (entries 19–21) [3,6,9,14–17]. Under com-
parable conditions catalyst Aa was only able to incorporate
12.5 mol% P, with an activity of 660 kg polymer (mol
metal)À1 hÀ1 barÀ1 (Table 2, entry 20), i.e. the P incorpora-
tion is substantially lower than with the mixed-ligand
Table 2
Ethylene/propylene copolymerisation results for pre-catalysts 1a, 1b and 1ca
Entry
Catalyst/lmol
P:Eb
Time/min
Polymer/g
Productivityc
% P contentd
Mw
Mw/Mn
1
2
3
4
5
6
7
8
1a/1
1a/1
1a/1
1a/1
1a/1
1b/1
1b/1
1b/1
1b/1
1b/1
1c/1
1c/1
1c/1
1c/1
1c/1
1:1
5:1
10:1
1:5
1:10
1:1
5:1
10:1
1:5
1:10
1:1
5:1
10:1
1:5
3
10
10
1
1
3
10
10
1
1
3
10
10
1
1
3
10
1
3
10
1
0.1
2000
720
720
29,000
37,000
7000
2000
1500
16,000
14,000
5000
1800
1400
22,000
24,000
9000
900
34,000
4800
660
15,000
12
36
45
6
ꢀ0
10
27
39
2
ꢀ0
10
30
42
3
ꢀ0
32
60
ꢀ0
3.0
12.5
ꢀ0
163,000
74,000
75,000
2.5
2.6
2.6
3.4
3.2
2.0
2.2
2.0
1.9
1.9
2.2
2.1
1.9
2.0
1.9
2.1
2.6
2.0
1.8
2.0
1.7
0.12
0.12
0.46
0.63
0.33
0.34
0.25
0.27
0.24
0.25
0.31
0.23
0.37
0.40
0.47
0.16
0.59
0.24
0.12
0.25
421,000
452,000
200,000
172,000
127,000
204,000
238,000
152,000
123,000
103,000
158,000
167,000
55,000
9
10
11
12
13
14
15
16
17
18
19
20
21
a
1:10
1:1
10:1
1:10
1:1
Cp2TiCl2/1
Cp2TiCl2/1
Cp2TiCl2/1
Aa/1
Aa/1
Aa/1
7700
188,000
311,000
258,000
368,000
10:1
1:10
Total volume 250 cm3, 20 ꢁC, 1 bar ethylene pressure, 3 mmol MAO.
b
c
P:E = propylene:ethylene gas feed ratio.
Productivity in kg polymer (mol metal)À1 hÀ1 barÀ1
Determined by 13C NMR spectroscopy.
.
d

L.M. Broomfield et al. / Journal of Organometallic Chemistry 692 (2007) 4603–4611
4607
4. Copolymer microstructure
lyst 1b show that P–P diads (aa methylene-carbons,
Fig. 5) are highly unfavourable and only become apparent
at very high propylene incorporation (Fig. 6 spectrum a).
Similar results were observed for catalysts 1a and 1c. At
high P content there is a tendency towards comonomer
alternation, reminiscent of copolymers described by Zam-
belli et al. using vanadium catalysts [20].
Catalysts 1a–1c and Aa all have a much higher affinity
for ethylene than for propylene. 13C NMR spectroscopy
using peak assignments were carried out in accordance
with Carman et al. [18] and Randall [19]. The labelling sys-
tem for the assignments is shown in Fig. 5. Representative
13C NMR spectra for copolymers with differing P content
are shown in Fig. 6. E–P copolymers produced with cata-
The effect of changing R in pre-catalyst 1a–c on the
possible monomer insertions has been investigated by
αα
H
αβ
Me
*
*
H
H
H
H
*
Me
H
H
Me
H
Me
αγ
*
αδ+
*
H
Me
H
*
*
H
H
H
Me
*
*
*
H
H
βγ
Me
αδ+
Me ββ
H
*
γγ
*
αδ+
*
γδ+
H
H
*
*
*
Me
H
H
H
Me
*
*
*
*
*
H
H
βδ+
Me
H
*
Me
βδ+
γδ+
*
αδ+
γδ+
*
αδ+
H
*
*
*
H
H
*
*
Me
H
H
Me
H
*
*
*
*
*
*
*
H
H
βδ+
δ+δ+
Me
Me
βδ+
δ+δ+
Fig. 5. Labels for methylene carbons in 13C NMR spectra [16].
δ+δ+
a) Gas feed 0.5 L min-1 C2 + 5 L min-1 C3
Propene incorporation 39 mol-%
βδ+
γδ+
αδ+
CH3
CHc
CHb
βγ
CHa
αγ
αα
αβ
ββ
b) Gas feed 1 L min-1 C2 + 5 L min-1 C3
Propene incorporation 27 mol-%
H
CHb =
CHa =
CHc =
H
H
H
H
c) Gas feed 1 L min-1 C2 + 1 L min-1 C3
Propene incorporation 10 mol-%
40
30
48
20
Fig. 6. 13C NMR spectra of E-P copolymers prepared with catalyst 1b MAO.

4608
L.M. Broomfield et al. / Journal of Organometallic Chemistry 692 (2007) 4603–4611
Table 3
6. Experimental
E/P reactivity ratios for catalysts 1a–1c for copolymers with a composi-
tional range of 0–45 mol% P
6.1. General
Precatalyst
Fineman-Ross
Kelen-Tudo¨s
¨
re
rp
re rp
re
rp
rerp
Syntheses were carried out under nitrogen using stan-
dard Schlenk line techniques. Solvents were distilled from
sodium–benzophenone (diethyl ether, THF), sodium (tolu-
ene), sodium–potassium alloy (light petroleum, b.p
40–60 ꢁC), or CaH2(dichloromethane). NMR solvents
1a
1b
1c
14.9
50.6
32.4
ꢀ0
ꢀ0
ꢀ0
ꢀ0
17.1
46.4
35.2
ꢀ0
ꢀ0
ꢀ0
ꢀ0
0.01
0.01
ꢀ0
ꢀ0
˚
(CDCl3, CD2Cl2, C6D6) were dried over activated 4 A
molecular sieves and degassed by several freeze-thaw
cycles. NMR spectra were recorded using Bruker DPX-
300 spectrometer and Bruker Avance DRX-400 spectrom-
eters with a 5 mm BBO probe. Chemical shifts are reported
in ppm and referenced to residual solvent resonances (1H,
13C), 19F is relative to CFCl3. Nitrogen, ethylene and pro-
pylene (BOC) were purified by passing through columns of
Fineman-Ross and Kelen-Tudo¨s statistical methods. The
¨
results are summarised in Table 3. Excellent agreement
between the two methods was observed. All three catalysts
show re ꢁ rp, indicative of very high concentrations of iso-
lated propylene units [21].
According to 13C NMR statistical analysis (Carman
method) of ethylene/propylene copolymers made using
precatalyst 1b with MAO, at 20 ꢁC and gas feed ratios
of 5 L minÀ1 of P and 0.5 L minÀ1 of E, propylene inser-
tion with 1,2-regiochemistry was most favourable. 82
mol% of the propylene was inserted in 1,2-fashion mech-
anism, leaving 18% of the propylene units inserted via a
regio-inversion (2,1 insertion) mechanism. Such behav-
iour is again similar to vanadium catalysts [20]. Sterically
unfavourable head-to-head contacts were not detected by
13C NMR spectroscopy (detection limit <0.3 mol%). The
C2 symmetric titanium bis(salicylaldiminato) catalyst Aa
activated with MAO homo-polymerises propylene in syn-
diotactic fashion via a chain-end controlled mechanism
with 2,1-regio-insertions [16]. However, when ethylene is
added to the process the propylene regio-insertions become
mainly 1,2 which is very similar to what is observed here
for 1b.
˚
supported P2O5, with moisture indicator, and activated 4 A
molecular sieves. Gel permeation chromatography GPC
analyses were performed by RAPRA Technology Ltd on
a Polymer Laboratories PL-GPC 220 instrument equipped
˚
with two PLgel 7.5 A MIXED B columns and calibrated
with polystyrene standards. Elemental analyses were per-
formed by London Metropolitan University. The ligands
HO(2-tBu)C6H3-6-CH@NC6F5, C6H5N@CH-2-C4H3NH
and C6H11N@CH-2-C4H3NH were synthesised using stan-
dard literature procedures [22,23].
6.2. Synthesis of C2H5N@CH-2-C4H3NH
To pyrrole-2-carboxaldehyde (4.8 g, 50 mmol) was
added a 70% solution of ethylamine in water (1.75 cm3,
50 mmol). After addition of a few drops of glacial acetic
acid the solution was vigorously stirred until the product
precipitated. To the reaction mixture diethyl ether (40 cm3)
and a 6 N HCl (20 cm3) solution was added at 0 ꢁC. The
organic phase was isolated, washed with saturated aqueous
NaHCO3 (50 cm3), dried over MgSO4 and then concentrated
under vacuum to afford an orange solid. Recrystallisation
from light petroleum at À25 ꢁC gave large orange crystals
(3.78 g, 35 mmol, 70%). 1H NMR (300 MHz, 300 K,
CDCl3): d 9.00 (s, 1H, C4H3NH), 8.07 (s, 1H, NCHAr),
6.88 (br s, 1H, C4H3N), 6.47 (br s, 1H, C4H3N), 6.25–6.21
(br s, 1H, C4H3N), 3.59–3.51 (m, 2H, CH2), 1.27–1.21 (m,
3H, CH3).
5. Conclusions
Titanium (N,NÀ)(N,OÀ) mixed-ligand complexes of
type 1 activated with MAO provide highly active catalysts
for the homopolymerisation of ethylene and for ethylene/
propylene copolymerisations. The ethylene homopolymeri-
sation activity was found to be higher or comparable to
either of the Ti(N,OÀ)2 and Ti(N,NÀ)2 parent complexes
A and B. The N-phenyl iminopyrrolato complex 1a showed
the highest productivity for ethylene polymerisation, while
changing this substituent for cyclohexyl or ethyl reduced
productivity. In ethylene–propylene copolymerisations the
sterically more open mixed-ligand systems show signifi-
cantly enhanced comonomer incorporation accompanied
by high activities and molecular weights, compared to A,
B or Cp2TiCl2/MAO. 13C NMR analysis showed that pro-
pylene is incorporated predominantly with 1,2-regiochem-
istry, although ca. 20% of P units show regio-inversion
(2,1). The reactivity ratios indicate a predominance of iso-
lated P units, and a tendency towards an alternating struc-
ture at high P content.
6.3. Synthesis of TiCl2(OC6H3-2-tBu-6-CH@NC6F5)
(PhN@CHC4H3N) (1a)
In a modification of a literature procedure [10,11], to a
stirred solution of TiCl3(OC6H3-2-tBu-6-CH@NC6F5)-
(THF) (2.31 g, 5 mmol) in dichloromethane (15 cm3) at
À78 ꢁC was slowly added by syringe a suspension of
K+[PhN@CH(C4H3N]À (1.04 g, 5 mmol) in dichlorometh-
ane (15 cm3). The resulting solution was allowed to warm
to room temperature and stirred for 6 h. The volatiles were
removed under vacuum and the resulting red-black solid

L.M. Broomfield et al. / Journal of Organometallic Chemistry 692 (2007) 4603–4611
4609
was extracted with diethyl ether (4 · 60 cm3) and washed
with light petroleum (6 · 25 cm3). The resulting dark red
solid was recrystallised from dichloromethane/light petro-
leum at À30 ꢁC to give very dark red-black needle-shaped
m-F), 163.1 (m, 1F, m-F). Anal. Calc.: C, 49.51; H, 3.81;
N, 7.22. Found: C, 49.49; H, 3.90; N, 7.33%.
6.6. Synthesis of ZrCl2(OC6H3-2-tBu-6-CH@NC6F5)-
(C6H11N@CHC4H3N) (2b)
1
crystals (1.2 g, 40%). H NMR (300 MHz, 300 K, CDCl3):
d 8.41 (s, 1H, CH@N, FI), 7.75 (d, 1H, J = 7.2 Hz, Ar),
7.74 (s, 1H, CH@N, PI), 7.65 (d, 1H, J = 7.7 Hz, Ar),
7.34 (d, 1H, J = 7.6 Hz, Ar), 7.15–7.05 (m, 6H, Ar), 6.68
(d, 1H, J = 7.6 Hz, Ar), 6.15 (m, 1H, Ar), 1.34 (s, 9H,
tBu) (FI = phenoxy-imine, PI = pyrrole-imine). 13C NMR
(75 MHz, 300 K, CDCl3): d 174.4 (C@N), 160.9 (C@N,
156.8, 150.2, 144.4, 139.1, 137.8, 136.5, 134.6, 129.3,
127.1, 125.2, 124.0, 123.5, 122.9, 121.0, 34.1, 29.7. 19F
NMR (282.4 MHz, 300 K, CDCl3): d 141.0 (d, 1F, o-F),
149.3 (d, 1F, o-F), 157.1 (t, 1F, p-F), 161.5, (m, 1F, m-
F), 162.6 (m, 1F, m-F). Anal. Calc.: C, 53.36; H, 3.52; N,
6.67; Cl, 11.25. Found: C, 53.14; H, 3.55; N, 6.65; Cl,
11.58%.
The complex was synthesised in an analogous way to
1a, using the appropriate potassium iminopyrrolate salt
and the zirconium mono(salicylaldiminato) complex as
opposed to the titanium one. The product was isolated as
large block-like orange crystals suitable for X-ray crystal-
1
lography. H NMR (300 MHz, 300 K, CDCl3): d 8.31 (s,
1 H, CH@N, FI), 7.95 (s, 1H, CH@N, PI), 7.79 (d, 1H,
J = 6.0 Hz, Ar), 7.29 (m, 1H, Ar), 7.26 (d, 1H,
J = 8.1 Hz, Ar), 7.11 (t, 1H, J = 7.8 Hz, Ar), 6.54 (m,
1H, Ar), 6.04 (m, 1H, Ar), 3.44 (m, cyclohexyl p-H), 1.61
t
(s, 1H, Bu), 1.22–0.96 (m, 11H, cyclohexyl). (FI = phen-
oxy-imine, PI = pyrrole imine). 19F NMR (282.4 MHz,
300 K, CDCl3): d 143.0 (d, 1F, o-F), d 149.6 (d, 1F, o-F),
d 157.5 (t, 1F, p-F), d 162.0, (1F, m-F), d 163.1 (1F, m-
F). Anal. Calc.: C, 49.48; H, 4.15; N, 6.18. Found: C,
49.38; H, 4.10; N, 6.35%.
6.4. Synthesis of TiCl2(OC6H3-2-tBu-6-CH@NC6F5)-
(C6H11N@CHC4H3N) (1b)
The complex was synthesised in an analogous way to 1a,
using the appropriate potassium iminopyrrolate salt.
1H NMR (300 MHz, 300 K, CDCl3): d 8.30 (s, 1H, CH@N,
FI), 7.77 (s, 1H, CH@N, PI), 7.76 (d, 1H, J = 7.9 Hz, Ar),
7.58 (m, 1H, Ar), 7.36 (d, 1H, J = 8.7 Hz, Ar), 7.18 (t, 1H,
J = 8.6 Hz, Ar), 6.46 (m, 1H, Ar), 6.01 (m, 1H, Ar), 3.74
6.7. Polymerisation of ethylene
A solution of MAO in toluene (250 cm3) was saturated
with ethylene (1 bar), with vigorous stirring (1000 rpm).
Polymerisation was initiated by charging a toluene (1 cm3)
solution of the appropriate pre-catalyst. After polymerisa-
tion had occurred methanol (3 cm3) was added to terminate
the run. The polymer was precipitated with methanol
(300 cm3) and aluminium residues were dissolved by further
addition of 2 M HCl (5 cm3). The polymer was separated by
filtration and washed with methanol (100 cm3), 2 M HCl
(5 cm3), water (20 cm3) and then again methanol (10 cm3).
The resulting polymer was then dried in a vacuum oven at
60 ꢁC to constant mass.
t
(m, cyclo-hexyl p-H), 1.85 (s, 9H, Bu), 1.76–1.34 (m, 11
H cyclo-hexyl). (FI = phenoxy-imine, PI = pyrrole imine).
13C NMR (75 MHz, 300 K, CDCl3): d 174.4 (C@N),
170.0 (C@N), 155.5, 154.0, 142.8, 139.8, 139.5, 137.8,
136.4, 124.7, 125.1, 123.8, 118.1, 112.6, 65.3, 38.5, 35.7,
34.6, 33.9, 32.9, 31.6, 26.2, 25.8, 24.9. 19F NMR
(282.4 MHz, 300 K, CDCl3): d 142.0 (d, 1F, o-F), d 149.6
(d, 1F, o-F), d 157.6 (t, 1F, p-F), d 161.5, (m, 1F, m-F), d
163.3 (m, 1F, m-F). Anal. Calc.: C, 52.85; H, 4.44; N,
6.60. Found: C, 52.96; H, 4.42; N, 6.63%.
6.8. Ethylene–propylene copolymerisations
6.5. Synthesis of TiCl2(OC6H3-2-tBu-6-CH@NC6F5)-
(EtN@CHC4H3N) (1c)
Runs were carried out in the same way as ethylene
homopolymerisations, with the exception that an ethyl-
ene/propylene gas mix was varied using a dual gas flow
meter.
The complex was synthesised in an analogous way to 1a,
using the appropriate potassium iminopyrrolate salt. The
product was isolated as very dark orange plate-like crystals
6.9. X-ray crystallography
1
suitable for X-ray diffraction. H NMR (300 MHz, 300 K,
CDCl3): d 8.37 (s, 1H, CH@N, FI), 7.82 (dd, 1H,
J = 8.5 Hz, Ar), 7.80 (s, 1H, CH@N, PI), 7.72 (m, 1H,
Ar), 7.41 (m, 1H, Ar), 7.19 (t, 1H, J = 7.0 Hz, Ar), 6.47
(m, 1H, Ar), 6.07 (m, 1H, Ar), 3.70 (m, 2H, CH2), 1.63
Intensity data for 1c and 2b were recorded on an Oxford
Diffraction Xcalibur Sapphire 3 diffractometer equipped
with a Spellman DF3 molybdenum sealed-tube source
operating at À50 kV, and fitted with Enhance X-ray optics.
Data for 1b were collected on a Bruker SMART APEX2
CCD diffractometer at Daresbury SRS station 16.2smx.
In all cases, crystals were mounted on glass fibres in per-
fluorinated polyether oil, and held in place by the nitrogen
cold-stream of the instrument. Data collection and reduc-
tion were carried out using either CRYSALIS CCD and RED
t
(s, 9H, Bu), 1.25 (m, 3H, CH3). (FI = phenoxy-imine,
PI = pyrrole imine). 13C NMR (75 MHz, 300 K, CDCl3):
d 174.7, 161.0, 156.7, 150.2, 142.7, 139.6, 137.8, 137.5,
136.5, 134.9, 125.3, 124.0, 118.2, 112.3, 35.5, 30.1, 15.1.
19F NMR (282.4 MHz, 300 K, CDCl3): 139.9 (d, 1F, o-
F), 152.1 (d, 1F, o-F), 157.2 (t, 1F, p-F), 161.4 (m, 1F,

4610
L.M. Broomfield et al. / Journal of Organometallic Chemistry 692 (2007) 4603–4611
Table 4
Crystal data for compounds 1b, 1c and 2b
Compound
1b
1c
2b
Formula
C28H28Cl2F5N3OTi
636.33
Colourless block
0.01 · 0.01 · 0.01
Monoclinic
P21/c
17.992(10)
12.198(6)
14.103(7)
90
108.222(5)
90
C24H22Cl2F5N3OTi
582.25
Orange plate
0.45 · 0.12 · 0.04
Triclinic
C28H28Cl2F5N3OZr
679.65
Yellow block
0.40 · 0.20 · 0.15
Monoclinic
P21/n
18.507(2)
8.2301(8)
20.480(2)
90
112.495(10)
90
Formula weight
Crystal description
Crystal dimensions/mm
Crystal system
Space group
ꢀ
P1
˚
a/A
7.6548(7)
10.5818(18)
15.8130(19)
94.628(12)
95.744(9)
97.746(10)
1257.0(3)
2
˚
b/A
˚
c/A
a/ꢁ
b/ꢁ
c/ꢁ
3
˚
V/A
2940(3)
4
2882.1(5)
4
Z
T/K
l/mmÀ1
120(2)
0.531
140(2)
0.613
180(1)
0.627
h range/ꢁ
3.8–29.7
11,075
4131
2060
0.1401
3.5–34.1
22,750
9016
3681
0.0739
3.4–25.0
35,618
5050
4077
0.0690
Number of data collected
Number of unique data
Number of ‘observed’ data
Rint
R1 [I> 2r(I)], R1 (all data)
wR2 (all data)
0.0728, 0.1681
0.1843
0.0448, 0.1534
0.0779
0.0556, 0.0695
0.0954
[2] T. Fujita, H. Makio, in: R.H. Crabtree, D.M.P. Mingos (Eds.),
Comprehensive Organometallic Chemistry III, vol. 11, Elsevier,
Amsterdam, 2007, p. 691.
[3] R. Furuyama, J. Saito, S. Ishii, H. Makio, M. Mitani, H. Tanaka, T.
Fujita, J. Organomet. Chem. 690 (2005) 4398.
[4] S. Matsui, M. Mitani, J. Saito, Y. Tohi, H. Makio, N. Matsukawa,
Y. Takagi, K. Tsuru, M. Nitabaru, T. Nakano, H. Tanaka, N.
Kashiwa, T. Fujita, J. Am. Chem. Soc. 123 (2001) 6847.
[5] R. Furyama, J. Saito, S. Ishii, M. Mitani, S. Matsui, Y. Tohi, H.
Makio, N. Matsukawa, H. Tanaka, T. Fujita, J. Mol. Catal. A. 200
(2003) 31.
programs (1c, 2b) [24] or APEX2 and SAINT (1b) [25]. Struc-
ture determination was by direct methods in SHELXS-97
[26], and refinement was by full-matrix least-squares meth-
ods using SHELXL-97 [26] within the WINGX program suite
[27]. All non-hydrogen atoms were refined using aniso-
tropic thermal parameters, and hydrogen atoms were
added using a riding model. The crystal and refinement
data are collected in Table 4.
Acknowledgements
[6] A. Sakuma, M.-S. Weiser, T. Fujita, Polym. J. 39 (2007) 193.
[7] R. Furuyama, M. Mitani, J. Mohri, R. Mori, T. Nakano, H. Tanaka,
T. Fujita, Macromolecules 38 (2005) 1546.
[8] Y. Yoshida, T. Nakano, H. Tanaka, T. Fujita, Isr. J. Chem. 42
(2002) 353.
[9] Y. Yoshida, S. Matsui, T. Fujita, J. Organomet. Chem. 690 (2005)
4382.
[10] D.A. Pennington, S.J. Coles, M.B. Hursthouse, M. Bochmann, S.J.
Lancaster, Macromol. Rapid Commun. 27 (2006) 599.
[11] D.A. Pennington, S.J. Coles, M.B. Hursthouse, M. Bochmann, S.J.
Lancaster, Chem. Commun. (2005) 3150.
This work was supported by the Engineering and Phys-
ical Sciences Research Council. We are grateful to Prof. V.
Busico and Dr. R. Cipullo (University of Naples) for
recording the 13C NMR spectra of the copolymers and
helpful discussion.
Appendix A. Supplementary material
[12] D.A. Pennington, R.W. Harrington, W. Clegg, M. Bochmann, S.J.
Lancaster, J. Organomet. Chem. 691 (2006) 3183.
[13] Y. Suzuki, T. Oshiki, H. Tanaka, K. Takai, T. Fujita, Chem. Lett. 34
(2005) 1458.
[14] A.F. Mason, G.W. Coates, J. Am. Chem. Soc. 126 (2004)
10798.
[15] L.J. Fetters, J.H. Lee, R.T. Mathers, P.D. Hustad, G.W. Coates,
L.A. Archer, S.P. Rucker, D.J. Lohse, Macromolecules 38 (2005)
10061.
[16] A.E. Cherian, E.B. Lobkovsky, G.W. Coates, Macromolecules 38
(2005) 6259.
CCDC 641426, 641427 and 641428 contain the supple-
mentary crystallographic data for 1b, 1c and 2b. These data
can be obtained free of charge via http://www.ccdc.cam.a-
c.uk/conts/retrieving.html, or from the Cambridge Crystal-
lographic Data Centre, 12 Union Road, Cambridge CB2
1EZ, UK; fax: (+44) 1223-336-033; or e-mail: depos-
it@ccdc.cam.ac.uk. Supplementary data associated with
this article can be found, in the online version, at
doi:10.1016/j.jorganchem.2007.04.047.
[17] J. Saito, Y. Suzuki, H. Makio, H. Tanaka, M. Onda, T. Fujita,
Macromolecules 39 (2006) 4023.
References
[18] C.J. Carman, R.A. Harrington, C.E. Wilkes, Macromolecules 10
(1977) 536.
[1] L. Resconi, J.C. Chadwick, L. Cavallo, in: R.H. Crabtree, D.M.P.
Mingos (Eds.), Comprehensive Organometallic Chemistry III, vol. 4,
Elsevier, Amsterdam, 2007, p. 1005.
[19] J.C. Randall, Macromolecules 11 (1978) 33.
[20] F.A. Bovey, M.C. Sacchi, A. Zambelli, Macromolecules 7 (1974)
752.

L.M. Broomfield et al. / Journal of Organometallic Chemistry 692 (2007) 4603–4611
4611
[21] Y. Sarazin, G. Fink, K. Hauschild, M. Bochmann, Macromol. Rapid
Commun. 26 (2005) 1208.
[24] Oxford Diffraction, CRYSALIS CCD, RED and ABSPACK, Abdingdon,
Oxford, UK, 2005.
[22] R.K.J. Bott, D.L. Hughes, M. Schormann, M. Bochmann, S.J.
Lancaster, J. Organomet. Chem. 665 (2003) 135.
[23] V.C. Gibson, C. Newton, C. Redshaw, G.A. Solan, A.J.P. White,
D.J. Williams, J. Chem. Soc. Dalton Trans. (2002) 4017.
[25] APEX2 and SAINT, Bruker AXS Inc., Madison, WI, USA, 2004.
[26] G.M. Sheldrick, Programs for Crystal Structure Analysis, University
of Go¨ttingen, Germany, 1997.
[27] L.J. Farrugia, J. Appl. Crystallogr. 32 (1999) 837.