Olefin polymerization and copolymerization by complexes bearing [ONNO]-Type salan ligands: Effect of ligand structure and metal type (titanium, zirconium, and vanadium)

Article abstract of DOI:10.1002/pola.27218

A series of novel titanium(IV) complexes bearing tetradentate [ONNO] salan type ligands: [Ti{2,2′-(OC₆H₃-5-t-Bu) ₂-NHRNH}Cl₂] (Lig¹TiCl₂: R = C ₂H₄; Lig²TiCl₂: R = C ₄H₈; Lig³TiCl₂: R = C ₆H₁₂) and [Ti{2,2′-(OC₆H ₂-3,5-di-t-Bu)₂-NHC₆H₁₂NH}Cl ₂] (Lig⁴TiCl₂) were synthesized and used in the (co)polymerization of olefins. Vanadium and zirconium complexes: [M{2,2′-(OC₆H₃-3,5-di-t-Bu)₂-NHC ₆H₁₂NH}Cl₂] (Lig⁴VCl₂: M = V; Lig⁴ZrCl₂: M = Zr) were also synthesized for comparative investigations. All the complexes turned out active in 1-octene polymerization after activation by MAO and/or Al(i-Bu)₃/[Ph ₃C][B(C₆F₅)₄]. The catalytic performance of titanium complexes was strictly dependent on their structures and it improves for the increasing length of the aliphatic linkage between nitrogen atoms (Lig¹TiCl₂ << Lig²TiCl₂ < Lig³TiCl₂) and declines after adding additional tert-Bu group on the aromatic rings (Lig³TiCl₂ < Lig⁴TiCl₂). The activity of all titanium complexes in ethylene polymerization was moderate and the properties of polyethylene was dependent on the ligand structure, cocatalyst type, and reaction conditions. The Et₂AlCl-activated complexes gave polymers with lover molecular weights and bimodal distribution, whereas ultra-high molecular weight PE (up to 3588 kg mol^-1) and narrow MWD was formed for MAO as a cocatalyst. Vanadium complex yielded PE with the highest productivity (1925.3 kg mol _v^-1), with high molecular weight (1986 kg mol ^-1) and with very narrow molecular weight distribution (1.5). Copolymerization tests showed that titanium complexes yielded ethylene/1-octene copolymers, whereas vanadium catalysts produced product mixtures.

Full text of DOI:10.1002/pola.27218

JOURNAL OF
POLYMER SCIENCE
WWW.POLYMERCHEMISTRY.ORG
ARTICLE
Olefin Polymerization and Copolymerization by Complexes Bearing
[ONNO]-Type Salan Ligands: Effect of Ligand Structure and Metal Type
(Titanium, Zirconium, and Vanadium)
Marzena Białek, Monika Pochwała, Grzegorz Spaleniak
Faculty of Chemistry, Opole University, Oleska 48, 45-052 Opole, Poland
Correspondence to: M. Białek (E-mail: marzena.bialek@uni.opole.pl)
Received 28 January 2014; accepted 17 April 2014; published online 6 May 2014
DOI: 10.1002/pola.27218
ABSTRACT: A series of novel titanium(IV) complexes bearing
tetradentate [ONNO] salan type ligands: [Ti{2,20-(OC6H3-5-t-
Bu)2-NHRNH}Cl2] (Lig1TiCl2: R 5 C2H4; Lig2TiCl2: R 5 C4H8; Lig3-
was moderate and the properties of polyethylene was depend-
ent on the ligand structure, cocatalyst type, and reaction condi-
tions. The Et2AlCl-activated complexes gave polymers with lover
molecular weights and bimodal distribution, whereas ultra-high
molecular weight PE (up to 3588 kg mol21) and narrow MWD
TiCl2: R 5 C6H12
)
and [Ti{2,20-(OC6H2-3,5-di-t-Bu)2-NHC6H12NH}
Cl2] (Lig4TiCl2) were synthesized and used in the (co)polymer-
ization of olefins. Vanadium and zirconium complexes: [M{2,20-
(OC6H3-3,5-di-t-Bu)2-NHC6H12NH}Cl2] (Lig4VCl2: M 5 V; Lig4ZrCl2:
M 5 Zr) were also synthesized for comparative investigations.
All the complexes turned out active in 1-octene polymerization
after activation by MAO and/or Al(i-Bu)3/[Ph3C][B(C6F5)4]. The
catalytic performance of titanium complexes was strictly
dependent on their structures and it improves for the increasing
length of the aliphatic linkage between nitrogen atoms (Lig1TiCl2
<< Lig2TiCl2 < Lig3TiCl2) and declines after adding additional
tert-Bu group on the aromatic rings (Lig3TiCl2 < Lig4TiCl2). The
activity of all titanium complexes in ethylene polymerization
was formed for MAO as
a cocatalyst. Vanadium complex
yielded PE with the highest productivity (1925.3 kg molv21), with
high molecular weight (1986 kg mol21) and with very narrow
molecular weight distribution (1.5). Copolymerization tests
showed that titanium complexes yielded ethylene/1-octene
copolymers, whereas vanadium catalysts produced product mix-
C
V
tures.
2014 Wiley Periodicals, Inc. J. Polym. Sci., Part A:
Polym. Chem. 2014, 52, 2111–2123
KEYWORDS: copolymerization; polyolefins; salan complexes;
Ziegler-Natta polymerization
The salan ligands that possess an aliphatic bridge between the
two amine donors produce C2-symmetric octahedral complexes
having the two labile groups in the cis position, which makes
them appropriate precatalysts for polymerization of olefins.8
Such complexes, having appropriate substituents on the pheno-
late rings and having appropriate metallic centers, can produce
highly isospecific PP and iPP-block-PE,19 or low-molecular-
weight atactic and ultrahigh-molecular-weight poly(1-hexene)
of different isotacticities.17 They are also active in isospecific
and living 1-hexene polymerization.8 The chiral salan com-
plexes are based on trans-1,2-diaminocyclohexane12 and the
chiral bipyrrolidine20 backbone. The former produce poly
(a-olefin)s whose isotacticity depends on the ligand’s bulk:
non-bulky groups lead to atactic polymers, whereas complexes
with tert-Bu and 1-adamantyl substituents lead to highly iso-
tactic products.12 The complexes with salan ligands arranged
around chiral 2,20-bipyrrolidine are relatively high isospecific
([mmmm] 5 87%) in 1-hexene polymerization.20 The salan zir-
conium complex with the binaphthyl-amine bis(phenolate)
INTRODUCTION The so-called postmetallocene complexes,
after activation, have been used for about two decades as
catalysts in ethylene and a-olefin homopolymerization and
copolymerization processes.1–7 In particular, attention was
paid to the development of complexes bearing dianionic
ancillary tetradentate ligands, amongst which [ONNO]-type
salan ligands for Group 4 metal centers can be distin-
guished:8–20 mainly for Zr8–20 and occasionally Ti9,10,17,20
and Hf9,10 ones. The salan ligands are attractive because
their structures can be modified easily by introduction vari-
ous substituents on aromatic rings and/or by changing
bridge between donor nitrogen atoms. That way of tailoring
the coordination environment around the metal centre may
affect the catalytic properties of the produced complexes.
The diamine-bis(phenolate) complexes of this type were
used in homopolymerization of ethylene14 and propyl-
ene,9,14,16,18,19 in propylene/ethylene copolymerization,19 and
in polymerization of higher a-olefins8–12,14,15,17,20 or other
monomers.10,12,13,21
Additional Supporting Information may be found in the online version of this article.
C
V
2014 Wiley Periodicals, Inc.
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2014, 52, 2111–2123
2111

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
SCHEME 1 A route for the synthesis of salans.
EXPERIMENTAL
ligand turned out to produce atactic PP and isotactic polymers
of higher a-olefins.14 The salophen ligands containing the aro-
matic linkage between nitrogen atoms, which are closely anal-
ogous to salan ligands, formed complexes which are highly
active in 1-hexene polymerization.22 The activity of some of
them was considerably higher than that of salan complexes
but the produced polymers were atactic.22 The C1-symmetric
achiral and chiral zirconium complexes with salan ligands
were also derived from nonsymmetric salan ligands which
bear a halo-substituted phenolate ring and an alkyl-substituted
phenolate ring.11,15,16 These [ONNO0]ZrBn2 complexes yielded
low-molecular-weight polypropylene with different tacticity,
from atactic to isotactic ([mmmm]5 78%),16 and from atactic
to highly isotactic poly(1-hexene).15
All air-sensitive and/or moisture-sensitive compounds were
handled under an inert atmosphere, using standard Schlenk
line techniques and a glove box.
Materials
Argon (grade 5.0, Linde Gas), methylaluminoxane (MAO) (10
wt %, Sigma-Aldrich), Et2AlCl (1.0 M, Sigma-Aldrich), EtAlCl2
(25%, Sigma-Aldrich), and (i-Bu)3Al (1.1 M, Agros Organic),
[Ph3C][B(C6F5)4] (Strem Chemicals), B(C6F5)3 (98%, Strem
Chemical), NaH (60%, Sigma-Aldrich), TiCl4, ZrCl4, VCl4
(Sigma-Aldrich), CDCl3 (99.8%, Sigma-Aldrich), 1,2-dichloro-
benzene-d4 (Deutero GmbH), DMSO-d6 (99.8%, Euriso-top),
ethylenediamine (99.5%, POCH), MgSO4 (POCH), NaBH4
(98%, Sigma-Aldrich), butanediamine (99%, Sigma-Aldrich),
4-t-butylophenol (99%, Sigma-Aldrich), 2,4-di-t-butylophenol
(99%, Sigma-Aldrich), hexanediamine (98%, Sigma-Aldrich),
silica gel (0.063–0.2 mm, Fluka Chemika) were used as
received. Tetrahydrofurane, hexane, and toluene were
refluxed over sodium/benzophenone or sodium (toluene). 1-
Octene (98%, Sigma-Aldrich) was dried under argon over 4A
molecular sieves. Ethylene (Grade 3.0, Linde Gas) and nitro-
gen (Messer) were used after having been passed through a
column with sodium metal supported on Al2O3. Chloroform,
methylene chloride, ethanol, and methanol (POCH) were
used after previous distillation.
The nonchiral salan ligands and appropriate salan com-
plexes, as introduced until today, were based on the
N,N0-dimethyl-ethylenediamine or ethylenediamine backbone
only.8–11,15,17–19,21 We therefore decided to expand that fam-
ily of ligands and complexes. The new salan precursors were
synthesized (Scheme 1), featuring the hexanediamine,
butanediamine, and ethylenediamine backbones, and either
tert-butyl phenolates or di-tert-butyl phenolates. They were
then used for the synthesis of titanium, zirconium, and vana-
dium complexes; their catalytic properties were investigated
in 1-octene and ethylene homopolymerization and in ethyl-
ene/1-octene copolymerization processes. In particular, we
were interested in the research of vanadium systems as only
salan complexes of Group 4 transition metal had been
reported in literature until now.
General Methods
The infrared spectra were recorded with the Nicole Nexus
2002 Fourier transform infrared (FTIR) spectrometer.
2112
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2014, 52, 2111–2123

JOURNAL OF
POLYMER SCIENCE
WWW.POLYMERCHEMISTRY.ORG
ARTICLE
Samples for FTIR analysis were prepared as follows: PEs,
copolymers, and salans were made as powder pills with KBr,
poly(1-olefin)s were placed between two KBr plates and the
complexes were analyzed in nujol. Compositions of copoly-
ans with NaBH4. In case of Lig2–4H2 different procedure
for isolation was applied. First, they were extracted
with methylene chloride and then purified by flesh
chromatography. All ligands were recrystallized from
hexane.
1
mers were found by the modified IR method.23 The H NMR
and 13C NMR spectra of the ligands as well as of the homo-
and copolymers were recorded with the Ultrashield Bruker
spectrometer (400 MHz). The copolymers were analyzed in
1,2-dichlorobenzene-d4 at 120 ꢀC. The peak of the residual
undeuterated solvent at 7.249 was chosen as the internal
N,N-bis(5-t-butyl-2-hydroxybenzyl)-1,2-diaminoethane
(Lig1H2)
Tm 5 161.5–164.0 C, 1H NMR: d 1.28 (s, 18H, ArAC(CH3)3);
ꢀ
2.86 (s, 4H, CH2-NH); 3.99 (s, 4H, NH-CH2-Ar); 6.77 (d,
J 5 8.47 Hz, 2H, ArAH); 6.98 (d, J 5 2.38, 2H, ArAH); 7.19
(dd, J 5 8.5; 2.4 Hz, 2H, ArAH). 13C NMR: d 31.04 (6C,
C(CH3)3); 33.41 (2C, C(CH3)3); 47.47 (2C, (CH2ANH); 52.52
(2C, NH-CH2-Ar); 115.26 (2C, Ar); 120.85 (2C, Ar); 124.72
(2C, Ar); 125.07 (2C, Ar); 141.34 (2C, Ar); 154.91 (2C, Ar).
FTIR (cm21) 3259 (NH); 2958 (CH3); 2901 (CH2); 1612,
1516, 1454 (C5C Ar); 1282, 1267 (PhAO); 1392 (OH);
1377, 1362 (C(CH3)3).
1
reference for the H NMR analysis, and the chemical shifts in
the 13C NMR analysis were referenced internally to the
major backbone methylene carbon resonance, which was
taken to be 30.00 ppm. The samples of ligands (10 mg) and
poly(1-octene)s were dissolved in CDCl3 and the analysis
was performed at room temperature. The melting tempera-
tures (Tm) of PEs and copolymers, their crystallinity (v), and
glass transition temperatures (Tg) of poly(1-olefin)s were
determined by differential scanning calorimetry (DSC) with a
2010 DSC calorimeter from TA Instruments at the heating
rate 10 ꢀC min21 (unless otherwise stated). Molecular
weights and molecular weight distributions of (co)polymers
were determined by gel permeation chromatography using
an Alliance 135 GPCV 2000 apparatus equipped with two
columns: HT 3 and HT 6E, at 135 ꢀC, using 1,2,4-trichloro-
benzene (1.0 mL min21) as a solvent. The average molar
mass values were calculated from the polystyrene calibration
curve obtained with narrow PDI standards. The molecular
N,N-bis(5-t-butyl-2-hydroxybenzyl)21,2-diaminobutane
(Lig2H2)
Tm 5 119.0–124.0 C, 1H NMR: d 1.28 (s, 18H, ArAC(CH3)3);
ꢀ
1.60 (m, 4H, (CH2)2); 2.70 (m, 4H, CH2-NH); 3.98 9(s, 4H,
NHACH2AAr); 6.76 (d, J 5 8.53, 2H, ArAH); 6.98 (d, J 5 2.51,
2H, ArAH); 7.18 (dd, J 5 8.53, 2.51 Hz, 2H, ArAH). 13C NMR:
d 26.70 (2C, (CH2)2); 31.05 (6C, C(CH3)3); 33.41 (2C,
C(CH3)3); 47.94 (2C, CH2ANH); 52.61 (2C, NHACH2AAr);
115.20 (2C, Ar); 121.15 (2C, Ar); 124.57 (2C, Ar); 124.87
(2C, Ar); 141.13 (2C, Ar); 155.17 (2C, Ar). FTIR (cm21) 3291
(NH); 2957 (CH3); 2924 (CH2); 1613, 1514, 1452 (C5C Ar);
1277 (Ph-O); 1390 (OH); 1386, 1362 (C(CH3)3; 747 ((CH2)n;
n ꢁ 4).
ꢀ
weights of poly(1-olefin)s were determined at 100 C.
Synthesis of 5-t-butyl-2-hydroxybenzaldehyde
5-t-butyl-2-hydroxybenzaldehyde was prepared in accord-
ance with the procedure presented in ref. 24.
N,N-bis(5-t-butyl-2-hydroxybenzyl)-1,2-diaminohexane
(Lig3H2)
1H NMR: d 1.32 (s, 9H, ArAC(CH3)3); 6.92 (d, J 5 8.72, 1H,
ArAH); 7.50 (d, J 5 7.50–7.51, 1H, ArAH); 7.57 (dd, J 5 8.72;
2.51 Hz, ArAH); 9.87 (s, 1H, ArACHO); 10.87 (s, 1H,
Tm 5 84.0–85.5 ꢀC, 1H NMR: d 1.28 (s, 18H, ArAC(CH3)3);
1.37 (m, 4H, (CH2)2); 1.54 (m, 4H, ACH2A); 2.68 (t, J 5 7 Hz,
4H, ACH2ANHA); 3.98 (s, 4H, NH-CH2-Ar); 6.76 (d, J 5 8,47
Hz, 2H, ArAH); 6.98 (d, J 5 2.38 Hz, 4H, ArAH); 7.17 (dd,
J 5 8.41; 2.5 Hz, ArAH). 13C NMR: d 26.37 (2C, (CH2)2);
28.96 (2C, ACH2A); 31.05 (6C, C(CH3)); 33.40 (2C, C(CH3)3);
48.14 (2C, CH2ANH); 52.63 (2C, NHACH2AAr); 115.18 (2C,
Ar); 121.26 (2C, Ar); 124.51 (2C, Ar); 124.81 (2C, Ar);
141.03 (2C, Ar); 155.26 (2C, Ar). FTIR (cm21) 3283 (NH);
2960 (CH3); 2930 (CH2); 1615, 1597, 1503 (C5C Ar); 1260
(Ph-O); 1399 (OH); 1376, 1361 (C(CH3)3; 728 ((CH2)n;
n ꢁ 4).
ArAOH). 13C NMR:
d 30.68 (3C, C(CH3)3); 33.52 (1C,
C(CH3)3); 116.65 (1C, Ar); 119.47 (1C, Ar); 129.21 (1C, Ar);
134.14 (1C, Ar); 142.20 (1C, Ar); 158.92 (1C, Ar); 196.29
(1C, CHO).
Synthesis of 3,5-di-t-butyl-2-hydroxybenzaldehyde
3,5-Di-t-butyl-2-hydroxybenzaldehyde
was
synthesized
exactly in the same way as 5-t-butyl-2-hydroxybenzaldehyde.
1H NMR:
d 1.33 (s, 9H, ArAC(CH3)3); 1.44 (s, 9H,
ArAC(CH3)3); 7.35 (d, J 5 2.51 Hz, 1H, ArAH); 7.60 (d,
J 5 2.51 Hz, 1H, ArAH); 9.87 (s, 1H, CHO); 11.65 (s, 1H,
ArAOH). 13C NMR:
d 28.72 (3C, C(CH3)3); 30.77 (3C,
N,N-bis(3,5-di-t-butyl-2-hydroxybenzyl)-1,2-
diaminohexane (Lig4H2)
C(CH3)3); 33.70 (1C, C(CH3)3); 34.48 (1C, C(CH3)3); 119.43
(1C, Ar); 127.32 (1C, Ar); 131.37 (1C, Ar); 137.03 (1C, Ar);
141.08 (1C, Ar); 158.55 (1C, Ar); 196.83 (1C, CHO).
Tm 5 110.5–112.5 ꢀC, 1H NMR:
d
1.30 (s, 18H,
ArAC(CH3)3);1.37 (m, 4H, (CH2)2); 1.44 (s, 18H,
ArAC(CH3)3); 1.56 (m, 4H, ACH2A); 2.70 (t, J 5 7,03 Hz, 4H,
CH2-NH); 3.96 (s, 4H, NHACH2-Ar); 6.88 (d, J 5 2.26, 2H,
ArAH), 7.24 (d, J 5 2.26, ArAH). 13C NMR: d 26.39 (2C,
(CH2)2); 28.96 (2C, ACH2A); 29.10 (6C, C(CH3)3); 31.15 (6C,
C(CH3)3); 33.59 (2C, C(CH3)3); 34.35 (2C, C(CH3)3); 48.12
(2C, CH2ANH); 53.04 (2C, NH-CH2-Ar); 121.64 (2C, Ar);
Synthesis of Ligands
The ligands were prepared according to literature pro-
cedure25 by reaction of 5-t-butyl-2-hydroxybenzaldehyde
or 3,5-t-butyl-2-hydroxybenzaldehyde with appropriate
diamine followed by reduction of resulting salens to sal-
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2014, 52, 2111–2123
2113

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
122.32 (2C, Ar); 122.53 (2C, Ar); 135.31 (2C, Ar); 139.80
(2C, Ar); 154.19 (2C, Ar). FTIR (cm21) 3312 (NH); 2957
(CH3); 2902 (CH2); 1605, 1460 (C5C Ar); 1254, 1238 (Ph-
O); 1390 (OH); 1377, 1360 (C(CH3)3; 725 ((CH2)n; n ꢁ 4).
1H NMR (400 MHz, DMSO-d6): d 5 9.97 (s, 2H, N-H), 7.46 (d,
4
J 5 2.3 Hz, 2H, ArAH), 7.23 (dd, J 5 8.6 Hz, J 5 2.3 Hz, 2H,
3
4
ArAH), 6.87 (d, 3 J5 8.6 Hz, 2H, ArAH), 4.03 (br, 4H, ArACH2A
N), 2.85 (br, 4H, NACH2ACH2ACH2ACH2ACH2ACH2AN), 1.65
(br, 4H, NACH2ACH2ACH2ACH2ACH2ACH2AN), 1.30 (br, 4H,
Synthesis of [Ti{2,20-(OC6H3-5-t-Bu)2-NHC2H4NH}Cl2]
NACH2ACH2ACH2ACH2ACH2ACH2AN),
1.24
(s,
18H,
(Lig1TiCl2)
C(CH3)3). There are also present signals at shift values 3.60 (m)
and 1.76 (m), derived from THF. The ratio 8:1 (complex:THF).
FTIR (cm21) 1251 (Ph-O); 1593, 1563, 1508 (C5C Ar); 1376,
1364 (C(CH3)3); 728 ((CH2)n, nꢁ 4).
Synthesis of [Ti{2,20-(OC6H323,5-di-t-Bu)2-
NHC6H12NH}Cl2] (Lig4TiCl2)
NaH (0.10 g; 2.60 mmol) was added portionwise to a solu-
tion of Lig1H2 (0.50 g; 1.30 mmol) in 50 mL THF, with stir-
ring, at room temperature. When the addition operation
was completed, the mixture was stirred for 1 h. After that
time, the solution of titanium tetrachloride (0.14 cm3; 1.30
mmol) in 20 mL toluene was added dropwise which
changed the coloration to dark-red. The reaction mixture
was stirred for another 3 h at room temperature and then
it was filtered to remove insoluble impurities. The solvent
was expelled under reduced pressure to obtain 0.69 g of a
dark-red powder.
The complex Lig4TiCl2 was synthesized in the same way as
Lig2TiCl2 by reacting 0.90 mmol (0.5 g) of Lig4H2 with 1.81
mmol NaH (0.07 g) and a solution of TiCl4 (0.10 mL; 0.90
mmol) in toluene (15 mL). The procedure yielded 0.60 g of
a dark-red solid.
1H NMR (400 MHz, DMSO-d6): d 5 8.63 (s, 2H, N-H), 7.24 (d,
4
1H NMR (400 MHz, DMSO-d6): d 5 10.00 (s, 2H, N-H), 7.46
(d, J 5 2.3 Hz, 2H, ArAH), 7.25 (dd, J 5 8.4 Hz, J 5 2.3
J 5 7.15 Hz, 2H, ArAH), 7.17 (d, J 5 7.15 Hz, 2H, ArAH), 4.17
4
4
3
4
(br, 4H, ArACH2-N), 2.94 (br, 4H, NACH2ACH2ACH2A
CH2ACH2ACH2AN), 1.63 (br, 4H, NACH2ACH2ACH2ACH2A
Hz, 2H, ArAH), 6.87 (d, J 5 8.4 Hz, 2H, ArAH), 4.12 (br, 4H,
3
ArACH2-N), 1.24 (s, 18H, C(CH3)3), the additional signals at
the shift values 3.64 (t, J 5 6.7 Hz, 2H) 3.40 (t, J 5 6.3 Hz,
2H), 1.72 (m, 2H), 1.51 (m, 2H), N-CH2-CH2-N] suggests the
possibility of the existence of the complex in two conforma-
tions in the ratio 1:1. FTIR (cm21) 1265, 1246 (Ph-O); 1601,
1523, 1456 (C5C Ar); 1376, 1365 (C(CH3)3).
CH2ACH2AN),
1.36,
1.26
(br,
40H,
2xC(CH3)3,
NACH2ACH2ACH2ACH2ACH2ACH2AN). FTIR (cm21) 1231,
1218 (Ph-O); 1597, 1457 (C5C Ar); 1364, 1376 (C(CH3)3);
720 ((CH2)n, n ꢁ 4).
Synthesis of [V{2,20-(OC6H323,5-di-t-Bu)2-
NHC6H12NH}Cl2] (Lig4VCl2)
Synthesis of [Ti{2,20-(OC6H3-5-t-Bu)2-NHC4H8NH}Cl2]
(Lig2TiCl2)
The complex Lig4VCl2 was synthesized in the same way as
Lig1TiCl2 by reacting 1.27 mmol (0.7 g) of Lig4H2 with 2.53
mmol NaH (0.10 g) and a solution of VCl4 (0.13 mL; 1.27
mmol) in toluene (15 mL). The second stage of the synthesis
was conducted for 4 hours. The procedure yielded 0.55 g of
a dark-green solid. FTIR (cm21) 1236, 1231 (Ph-O); 1593,
1452 (C5C Ar); 1375, 1365 (C(CH3)3); 721 ((CH2)n, n ꢁ 4).
The complex Lig2TiCl2 was synthesized in the same way as
Lig1TiCl2 by reacting 0.73 mmol (0.3 g) of Lig2H2 with 1.45
mmol NaH (0.06 g) and a solution of TiCl4 (0.08 mL; 0.73
mmol) in toluene (15 mL). The second stage of the synthesis
was conducted for 2 h. The procedure yielded 0.83 g of a
dark-red solid.
Synthesis of [Zr{2,20-(OC6H323,5-di-t-Bu)2-
1H NMR (400 MHz, DMSO-d6): d 5 9.99 (s, 2H, NAH), 7.49
(d, J 5 2.1 Hz, 2H, ArAH), 7.22 (dd, J 5 8.6 Hz, J 5 2.1
NHC6H12NH}Cl2] (Lig4ZrCl2)
4
3
4
The complex Lig4ZrCl2 was synthesized in the similar way to
that for Lig1TiCl2 but the salan deprotonation step was omit-
ted. The reaction involved 0.80 mmol (0.44 g) of Lig4H2,
0.80 mmol (0.18 g) of ZrCl4 (0.08 mL; 0.73 mmol) and 50
mL of toluene. The reaction was performed for 46 h. The
procedure yielded 0.32 g of a white solid.
Hz, 2H, ArAH), 6.88 (d, J 5 8.6 Hz, 2H, ArAH), 4.02 (br, 4H,
3
ArACH2-N), 2.89 (br, 4H, NACH2ACH2ACH2ACH2AN), 1.74
(br, 4H, NACH2ACH2ACH2ACH2AN), 1.24 (s, 18H, C(CH3)3);
additional signals at the shift values 3.64m (t, J 5 6.6 Hz,
3
NACH2ACH2ACH2ACH2AN), 1.50 (m, NACH2ACH2ACH2A
CH2-N) suggests the possibility of the existence of the com-
plex in two conformations in the ratio 8:1. There are also
present signals at shift values 3.59 (m) and 1.76 (m),
1H NMR (400 MHz, DMSO-d6): d 5 8.59 (s, 2H, N-H), 7.29 (d,
4J 5 2.5 Hz, 2H, ArAH), 7.26 (d, J 5 2.5 Hz, 2H, ArAH), 4.10
4
derived from THF. The ratio 8:1 (complex:THF). FTIR (cm21
)
(br, 4H, ArACH2-N), 2.94 (br, 4H, NACH2ACH2ACH2A
CH2ACH2ACH2AN), 1.65 (br, 4H, NACH2ACH2ACH2A
CH2ACH2ACH2-N), 1.37(s, 18H, C(CH3)3), 1.34 (br, 4H,
NACH2ACH2ACH2ACH2ACH2ACH2AN), 1.26 (s, 18H,
C(CH3)3). FTIR (cm21) 1225 (Ph-O); 1603, 1452 (C5C Ar);
1375, 1364 (C(CH3)3); 722 ((CH2)n, n ꢁ 4).
1267 (Ph-O); 1598, 1504, 1459 (C5C Ar); 1377, 1366
(C(CH3)3); 721 ((CH2)n, n ꢁ 4).
Synthesis of [Ti{2,20-(OC6H3-5-t-Bu)2-NHC6H12NH}Cl2]
(Lig3TiCl2)
The complex Lig3TiCl2 was synthesized in the same way as
Lig1TiCl2 by reacting 1.13 mmol (0.5 g) of Lig3H2 with 2.27
mmol NaH (0.09 g) and a solution of TiCl4 (0.12 mL; 1.13
mmol) in toluene (15 mL). The procedure yielded 0.87 g of
a dark-red solid.
General Procedure for Ethylene Polymerization and
Copolymerization
The ethylene polymerization reactions were performed in a
€
Buchi glass autoclave equipped with the magnetic stirrer and
2114
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2014, 52, 2111–2123

JOURNAL OF
POLYMER SCIENCE
WWW.POLYMERCHEMISTRY.ORG
ARTICLE
SCHEME 2 A route for the synthesis of complexes.
heating-cooling jacket. Initially, 150 mL of toluene and the
required amounts of the activator was charged to the auto-
clave, and when the required temperature was reached, the
complex was introduced to the reactor. Then, the ethylene gas
was fed. The ethylene feed pressure (5 bar) and the reactor
temperature were kept constant throughout the runs. After
the prescribed time, that is 30 min, the ethylene gas feed was
stopped and the obtained mixture was transferred to a dilute
solution of hydrochloric acid in methanol. The polymer was
filtered, washed a few times with methanol and dried under
vacuum. The ethylene/1-octene copolymerization process fol-
lowed the same procedure as ethylene homopolymerization
and the only difference was that 1-olefin was charged to the
reactor after the solvent had been added there.
zolidine32,33 as parent substances. Our research was based on
a two-step synthesis process to which the aldehyde compound
was fed as obtained from the Reimer-Tiemann reaction
(Scheme 1).24,34 The reaction of aldehyde and diamine (with a
suitable chain length) produced salens (Scheme 1) which pre-
cipitated from the reaction mixtures as yellow sediments.
Salens were reduced with sodium borohydride at another
stage (Scheme 1) to yield corresponding salans. Lig1H2 and
Lig3H2 were obtained as white precipitates, while Lig2H2 and
Lig4H2 were recovered from the reaction mixtures by adding
water and extracting the resulting aqueous solutions with
methylene chloride. The ligands Lig2–4H2 were additionally
purified by flash chromatography. Four salan structures were
synthesized within this work. Two of them, containing the
butylene and hexylene linkages between the nitrogen atoms
and one t-Bu group in each phenolate ring (Lig2H2 and
Lig3H2) have not so far been reported in literature.
General Procedure for 1-Octene Polymerization
The monomer (6 mL; 37 mmol) and the organoaluminium
activator were added to a 100 mL flask. After the required
temperature level had been reached (30, 60, or 80 ꢀC), the
complex was added to the reaction mixture. When
[Ph3C][B(C6F5)4] was used as an activator, it was added as
the final component. The mixture was stirred for the pre-
scribed period of time (120 min), and the reaction was ter-
minated by adding the HCl/MeOH solution. The obtained
product was washed with methanol and dried.
The complexes with salan ligands may be obtained directly
from salans28,35,36 or from their salts14 under various reac-
tion conditions, that is at various temperatures and with the
use of various solvents. The salan complexes of titanium and
vanadium were prepared in a two-stage process in our
research. The first stage, deprotonation of salans with the
use of sodium hydride, was performed in tetrahydrofuran
(Lig1–3H2) or in toluene (Lig4H2) at room temperature. At
another stage, the in situ ligand salts were reacted with the
toluene solution of metal tetrachloride (Scheme 2). The reac-
tions were conducted at room temperature over 2 to 4 h,
and then, after removing impurities by filtration, the solvent
was removed under reduced pressure. The obtained titanium
complexes were red-colored and the vanadium complex was
RESULTS AND DISCUSSION
Synthesis of Ligands and Complexes
Single-step and multi-step methods are known and utilized for
the production of salans; those methods involve phenols,26–29
aldehydes,25,30,31 or 1,3-bis(2’-hydroxy-5’-methylbenzyl)imida-
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2014, 52, 2111–2123
2115

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
TABLE 1 Data for 1-Octene Polymerization by Lig1–4TiCl2 Complexes
T
(ꢀC)
Yield
(g)
Productivity
21
C
Mw (g
mol21
)
Mw/
Entry
Complex
Activator
(kg molTi
)
(%)
Mn
MP5
Lig1TiCl2
Lig1TiCl2
Lig1TiCl2
Lig1TiCl2
Lig2TiCl2
Lig2TiCl2
Lig2TiCl2
Lig3TiCl2
Lig3TiCl2
Lig3TiCl2
Lig3TiCl2
Lig4TiCl2
Lig4TiCl2
Lig4TiCl2
MAO
30
30
30
60
30
30
60
RT
30
30
60
30
30
60
0.012
0.005
0.240
traces
0.930
0.012
0.789
0.764
0.969
0.022
0.546
0.189
0.026
0.238
0.40
0.17
8.00
–
0.28
0.12
5.71
–
–
–
–
MP6
Et2AlCl
Al(iBu)3/B
MAO
–
MP11
MP18
MP12
MP14
MP16
MP9
236,000
428
–
–
MAO
31.00
0.40
26.30
25.47
32.30
0.73
18.20
6.30
0.87
7.93
22.12
0.28
18.77
18.17
23.05
0.52
12.99
4.49
0.62
5.66
–
–
Al(iBu)3/B
MAO
–
–
–
–
MAO
–
–
MP3
MAO
–
–
MP4
Al(iBu)3/B
MAO
–
–
MP1
–
–
MP13
MP15
MP17
MAO
–
–
Al(iBu)3/B
MAO
–
–
37,207
100
Polymerization conditions: 0.03 mmol of complex; 3 mmol of organoaluminium activator; 1-octene (6.0 mL, 37.47 mmol); 2 h; Al(iBu)3/B 5 Al(iBu)3/
[Ph3C][B(C6F5)4], 0.45 mmol/0.045 mmol; C—conversion; RT—room temperature.
green-colored. The zirconium complex, because of poor solu-
bility of zirconium tetrachloride and the complex itself in tol-
uene, was prepared with no salan deprotonation step. The
reaction of zirconium tetrachloride and toluene solution of
salan (Scheme 2) was conducted at room temperature over
46 h.
with Al(i-Bu)3/[Ph3C][B(C6F5)4]. A reverse relation was
observed only for the complex Lig1TiCl2. The activity per-
formance for the catalytic systems Lig1–3TiCl2/MAO with one
tert-butyl group on the aromatic rings was improving up to
the following pattern: Lig1TiCl2 << Lig2TiCl2 < Lig3TiCl2
(Fig. 1). Hence, a higher length of the aliphatic bridge [from
(CH2)2 to (CH2)4] turned out to improve considerably the
catalytic activity of the complex studied, probably due to
improved accessibility of the active site for coordination of a
higher 1-olefin. Any further extension of the connecting link
between donor nitrogen atoms does not matter much. The
relation between the catalytic activity and the type of the
bridge in a complex was the same also at higher tempera-
tures, that is at 60 ꢀC (Fig. 1). At the same time, however,
higher reaction temperatures resulted in lower reaction
yields in each case which was evidence of no thermal stabil-
ity of the tested catalysts. The activity of the system
Lig3TiCl2/MAO at ambient temperature (entry MP9) was
lower than at 30 ꢀC and higher than at 60 ꢀC. The complex
Lig4TiCl2, which had the additional tert-butyl group on its
The FTIR analysis revealed that the bands which represented
stretching vibrations of PhAO in the spectra of complexes
were shifted towards lower frequencies, when compared
with their locations in the spectra of corresponding salans.
Some changes were also apparent within the range which is
specific for stretching vibrations of NAH. There are intensive
bands in the spectra of salans: at 3259 cm21 for Lig1H2, at
3291 cm21 for Lig2H2, at 3283 cm21 for Lig3H2 and at 3312
cm21 for Lig4H2. In the spectra of complexes, that band is
either absent or its height is very low and it is shifted
toward lower frequencies (Lig1TiCl2, Lig4ZrCl2). Those
changes speak for the formation of complex compounds.
Additionally, the bands were identified in the FTIR spectra of
complexes Lig2TiCl2 and Lig3TiCl2 which represent stretching
vibrations in the CAO-C group which suggests the presence
of solvent (THF) molecules, either fixed by coordination or
encapsulated in the crystal lattice.
Polymerization of 1-Octene
Zirconium and titanium complexes with salan ligands are
usually tested as precatalysts for a-olefin polymerization in
combination with MAO and Al(i-Bu)3/[Ph3C][B(C6F5)4] or
9,10,12,14,17,20
B(C6F5)3.
The titanium complexes synthesized by
us were activated with the use of the first two compounds
and they were tested in 1-octene polymerization (neat
monomer) at 30 ꢀC (Table 1). The complexes Lig2–4TiCl2
were remarkably more active in conjunction with MAO than
FIGURE 1 Effect of ligand structure and reaction temperature on
the yield of 1-octene polymerization catalyzed by Lig1–4TiCl2/MAO.
2116
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2014, 52, 2111–2123

JOURNAL OF
POLYMER SCIENCE
WWW.POLYMERCHEMISTRY.ORG
ARTICLE
TABLE 2 Data for 1-Octene Polymerization by Lig4MCl2 (M 5 Zr, V) Complexes
T
(ꢀC)
Yield
(g)
Productivity
21
C
Mn (g
mol21
)
Mw/
Entry
Complex
Activator
(kg molM
)
(%)
Mn
MP35
MP34
MP39
MP56
MP33
MP32
Lig4VCl2
Lig4VCl2
Lig4VCl2
Lig4ZrCl2
Lig4ZrCl2
Lig4ZrCl2
Et2AlCl
MAO
RT
RT
30
30
80
80
0.039
0.105
0.092
0
1.30
3.50
3.07
–
0.93
2.50
2.19
–
–
–
234,000
125
–
MAO
–
Al(iBu)3/B
Al(iBu)3/B
MAO
–
–
0.127
0
4.23
–
3.02
–
316
–
1.3
–
Polymerization conditions: 0.03 mmol of complex; 3 mmol of organoaluminium activator; 1-octene (6.0 mL, 37.47 mmol); 2 h; Al(iBu)3/B 5 Al(iBu)3/
[Ph3C][B(C6F5)4], 0.45 mmol/0.045 mmol; C—conversion; RT—room temperature.
aromatic rings, turned out less active than its equivalent
with only one tert-butyl group (Lig3TiCl2) (Table 1, Entries
MP3 and MP13). Yet, it was thermally more stable than com-
plexes with one tert-Bu substituent (Fig. 1).
products contained two fractions of macromolecules with
clearly diversified molecular weights: very high and very
low, which can be observed in GPC chromatograms (Fig. 3).
As regards other catalytic systems, the products were oils
[Fig. 2(b)], irrespective of the catalyst composition and reac-
tion conditions. Among them the only product with higher
degree of polymerization and a higher average molecular
weight, about 37,200 g mol21 (Fig. 3), was produced by the
catalytic system Lig4TiCl2/MAO. The other products were
low-viscosity liquids and could not be analyzed with GPC.
No attempt has been made so far to investigate the catalytic
performance of salan-type titanium complexes together with
simple organoaluminium compounds. Thus, the complex
Lig1TiCl2 was also used in 1-octene polymerization, after its
activation with Et2AlCl. This catalytic system showed how-
ever a very low activity (entry MP6).
The poly(1-octene)s with higher molecular weight, samples
MP11 and MP 34, were analyzed by 13C-NMR. The obtained
spectra indicate that catalysts Lig1TiCl2/Al(iBu)3/[Ph3C]
[B(C6F5)4] and Lig4VCl2/MAO produced about 67% and 46%
isotactic polymers, respectively.
The vanadium complex Lig4VCl2, just like Lig1TiCl2, turned
out poor catalyst in conjunction with Et2AlCl (Table 2, entry
MP35). The efficiency of MAO activated complex was not
much higher, 2.5% only. That catalyst was used at ambient
temperature since increased reaction temperatures up to 30
ꢀC had a low but negative impact on the process yield (Table
2, Entries MP34 and MP39). Lig4ZrCl2 was active in conjunc-
tion with Al(iBu)3/[Ph3C][B(C6F5)4], its activity was rather
low and polymer was obtained only at high temperature
(MP33). The monomer conversion after 2 h exceeded 3%.
Obtained products were analyzed by FTIR method to see if the
cocatalyst type and complex structure had any effect on the
macromolecules structure and thus on the reaction mecha-
nism. The examples of FTIR spectra were provided in Figure 4.
The spectra for the products obtained with the use of titanium
complexes, irrespective of their structures, and MAO as the
activator contain high-intensity bands which are specific for
unsaturated terminal groups: trans-vinylene group (ca., 967
cm21) and vinyl group (910 cm21, 993 cm21). There is also a
The products obtained from polymerization of 1-octene,
when catalyzed by the systems Lig1TiCl2/Al(i-Bu)3/[Ph3C]
[B(C6F5)4] and Lig4VCl2/MAO, had the form of sticky, gluey
solids [Fig. 2(a)]. The GPC analysis revealed that those prod-
ucts had high molecular weights (higher than 200,000 g
mol21). The MWD for those polymers was very wide as the
FIGURE 3 GPC traces of the products of 1-octene polymeriza-
tion obtained with: 1—Lig4VCl2/MAO (MP34); 2—Lig1TiCl2/
Al(iBu)3/[Ph3C][B(C6F5)4] (MP11); 3—Lig4TiCl2/MAO (MP17).
FIGURE 2 Poly(1-octene)s synthesized with Lig4VCl2/MAO (a)
and Lig2TiCl2/MAO (b).
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2014, 52, 2111–2123
2117

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
the same differences in the types of terminal groups of
poly(1-octene)s produced by the vanadium and titanium cat-
alytic systems as observed in FTIR spectra. Additionally, in
case of products given by Lig1–3TiCl2/MAO, very low signals
were present at about 5.1 ppm; they are indicative of a small
share of 1,1,2-trisubstituted groups. Such tri-substituted end
groups can be formed in re-arrangement of the polymer
chain which is then terminated by b-H transfer.37 That signal
was invisible in the spectrum of poly(1-octene) produced by
Lig4TiCl2/MAO, while the fraction of vinylidene groups was
clearly higher in that polymer (the shares of vinylidene
groups in terminal groups of the product obtained from
Lig3TiCl2 and Lig4TiCl2 systems amounted to 1.3% and
13.4%, respectively). Hence, the chain termination reactions
were controlled by the type of transition metal in the com-
plex, and by the presence of the additional t-Bu substituent.
FIGURE 4 FTIR spectra (section 500–2250 cm21) of products
produced by: 1—Lig4TiCl2/MAO (MP13); 2—Lig4VCl2/MAO
(MP39).
The thermal properties of some samples were also investi-
gated by means of DSC. When the products were heated up
at 5 ꢀC min21 within 2100 to 50 ꢀC no exotherms and/or
endotherms caused by crystallization and/or melting were
observed. Only the glass transition temperatures were visible
in thermograms, with the values close to 290 ꢀC for oils
(MP13: 289.0 ꢀC). Some higher Tg was noted for the sticky
solid product MP34 (275.4 ꢀC) which could resulted from
the molecular weight of that polymer which was higher.
band at about 890 cm21 which represents the vinylidene
group but that peak is very low. Vinylidene groups are formed
as a result of the b-elimination reaction of the hydrogen atom
after 1,2-insertion of 1-octene (Scheme 3, eq 1) or as a result
of the b-H transfer reaction to the monomer (Scheme 3, eq 2).
The trans-vinylene and vinyl groups as present within poly-
mer chains may be formed by the b-elimination reaction after
2,1-insertion of 1-octene (eq 3) and b-alkyl elimination reac-
tion, respectively (eq 4). As regards the spectra for the prod-
ucts obtained with the use of the catalytic systems Lig4VCl2/
MAO and Lig4ZrCl2/Al (i-Bu)3/[Ph3C][B(C6F5)4], only high-
intensity peaks for vinylidene and trans-vinyl groups at 967
and 890 cm23 were present.
Ethylene Polymerization and Copolymerization
With Et2AlCl as a cocatalyst, ethylene polymerization tests
promoted by Lig1–4TiCl2 under 5 bar of ethylene, at 30 and
50 ꢀC were performed. The typical results were summarized
in Table 3. The performance of the complexes Lig1–3TiCl2 was
ꢀ
ꢀ
found a bit better at 50 C than at 30 C, while the activity of
Lig4TiCl2 was independent on temperature in practice. At
50 ꢀC, the catalytic activity of the complexes changed in the
line: Lig2TiCl2 < Lig4TiCl2 < Lig1TiCl2 < Lig3TiCl2 (Fig. 5). A
Products which had been synthesized with the use of
Lig4VCl2 and Lig1–4TiCl2, activated by MAO, were also investi-
gated by the 1H NMR method. The 1H NMR spectra showed
SCHEME 3 Termination reaction.
2118
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2014, 52, 2111–2123

JOURNAL OF
POLYMER SCIENCE
WWW.POLYMERCHEMISTRY.ORG
ARTICLE
TABLE 3 Results of Ethylene Polymerization Catalyzed by Salan Complexes
T
(ꢀC)
Yield
(g)
Productivity
21
v
Mwꢂ1023 (g
Mw/
Entry
Complex
Activator
(kg molM
)
Tm (ꢀC)
(%)
mol21
)
Mn
MP45
MP53
MP47
MP20
MP22
MP23
MP8
Lig1TiCl2
Lig1TiCl2
Lig1TiCl2
Lig2TiCl2
Lig2TiCl2
Lig2TiCl2
Lig3TiCl2
Lig3TiCl2
Lig3TiCl2
Lig4TiCl2
Lig4TiCl2
Lig4TiCl2
Et2AlCl
Et2AlCl
MAO
30
50
50
30
50
50
30
50
50
30
50
50
30
30
65
80
1.407
1.527
1.116
1.691
1.978
0.218
1.284
1.332
0.074
1.894
1.847
1.196
0.157
1.348
0.451
0.311
20.1
21.8
15.9
24.2
28.3
3.1
129.9a
–
140
72
92.5b
241.6
3.4
146.5b
78.9
–
106.4; 124.9
–
135.5
128.5a
66
84
–
775
563
89
Et2AlCl
Et2AlCl
MAO
92.0; 105.4; 122.7
134.5
126.8a
28
83
–
–
Et2AlCl
Et2AlCl
MAO
18.3
19.0
1.1
253
56
147.5b
MP19
MP7
96.8; 106.7; 121.9
–
133.9
129.0a
10
83
–
–
–
MP24
MP38
MP42
MP43
MP26
MP31
MP37
Et2AlCl
Et2AlCl
MAO
27.1
26.4
17.1
224.3
1,925.3
6.4
2,250
1.6
3,588
–
996.9b
93.0; 105.3; 122.6
135.7
–
61
31
57
50
50
3.4
–
Lig4VCl2
MAO
133.5
c
Lig4VCl2
EtAlCl2
MAO
138.4
1,968
987
–
1.5
7.7
–
c
Lig4ZrCl2
Lig4ZrCl2
136.9
MAO
4.4
138.0
a
b
c
Polymerization conditions: 0.07 mmol of complex; 10.5 mmol of MAO;
7 mmol of Et2AlCl; 4 mmol of EtAlCl2; 5 bar of ethylene, 30 min; v—
crystallinity.
Broad peak with shoulder.
Bimodal.
0.0007 mmol of complex.
ꢀ
similar trend was observed at 30 ꢀC. The activity of all tita-
nium complexes activated by MAO was tested at 50 C. It was
ucts with lower melting points (128.5–129.9 C) when poly-
merization was performed at 30 ꢀC. On DSC curves of these
polymers the melting endotherms with shoulders are
observed. For polyethylenes produced at 50 ꢀC more than
one melting points were obtained and they were placed at
lower temperatures region. The examples of PE thermo-
grams were presented in Figure 6.
ꢀ
lower than the activity of the same complexes activated by
Et2AlCl, and the catalysts containing Lig4TiCl2 and Lig1TiCl2
were most active. The fact which is noteworthy is that these
two complexes showed the lowest activity in 1-octene poly-
merization. The reason was a larger size of the monomer to
be inserted into the polymer chain; it was more difficult for
the reaction catalyzed by a more sterically hindered complex
to convert a bigger molecule (1-octene) than a smaller one
(ethylene).
The molecular weight values for polyethylenes synthesized
with Lig1–4TiCl2/Et2AlCl were relatively low and they were
strongly dependent on the reaction temperature (Table 3).
The increase in temperature from 30 to 50 ꢀC resulted in a
The activator type was essential for the product properties.
In general, MAO-activated complexes produced linear poly-
ethylenes with high melting points (133.9–135.7 ꢀC),
whereas titanium complexes activated by Et2AlCl gave prod-
FIGURE 6 Thermograms of polyethylene synthesized with: 1—
Lig2TiCl2/MAO (MP23) at 50 ꢀC; 2—Lig2TiCl2/EtAlCl2 (MP20A) at
30 ꢀC; 3—Lig2TiCl2/EtAlCl2 at 50 ꢀC (MP22).
FIGURE 5 Effect of ligand structure and type of activator on
productivity of titanium catalysts at 50 ꢀC.
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2014, 52, 2111–2123
2119

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
The molecular weight value of PE was also significantly
affected by the ligand structure. The increasing number of
ACH2A groups in the bridge decreased the average molecu-
lar weight, while the additional t-Bu group on the ligand
rings produced substantially higher values. That might result
from steric changes around the active site. The steric hin-
drance arising from a shorter bridge and/or the presence of
an additional t-Bu group restrict the polymer chain termina-
tion reactions. The use of MAO instead of Et2AlCl also
increased the Mw value of polymers considerably. The MAO-
activated complexes produced polyethylenes with Mw over 3
million g mol21 and with not very narrow distributions of
molecular weights, Mw/Mn 5 3.4 (the GPC curves had signifi-
cant shoulders, Fig. 7, curve 3).
FIGURE 7 GPC traces of PEs produced by: 1—Lig4TiCl2/EtAlCl2
(MP38) at 50 ꢀC; 2—Lig4TiCl2/EtAlCl2 at 30 ꢀC (MP24);
3—Lig4TiCl2/MAO (MP42) at 50 ꢀC.
Highly interesting, the FTIR spectra for polyethylenes
obtained in the presence of Lig1–4TiCl2/Et2AlCl showed not
only the vinyl terminal groups, which are typical for ethylene
polymerization and which are formed by the b-hydrogen
elimination and/or transfer to the monomer, but also high
absorption bands at 888 cm21 which were indicative for the
vinylidene groups in polyethylene macromolecules. Those
terminal groups in ethylene homopolymerization could be
produced by reinsertion of vinyl terminated chains (macro-
mers) into the growing polyethylene chain. That incorpora-
tion could yield branched PE structures. To learn more
about the PE macromolecules, the 13C NMR spectra were
taken for two polymers: MP8a and MP19a. Those spectra,
however, did not confirm the presence of any branching in
PE chains; there were only peaks at 14.02, 22.85, 29.56,
30.00, and 32.17 ppm which come from low molecular
weights in the test PE sample. There was no signal for terti-
ary carbon atoms. Thus, addition of the ethylene macromer
molecule was probably followed by the chain termination
reaction.
few times lower values of Mw. The molecular weight distri-
bution patterns for PE were bimodal, irrespective of the
reaction temperature, with the higher and lower molecular
weight fractions being clearly separated (Fig. 7). Polymer
with bimodal molecular weight distribution is usually cre-
ated by the presence of two type of active sites. One site
produces high molecular weight fraction and the other gives
low molecular weight fraction. The share of the lower molec-
ular weight fraction was growing for higher temperatures
(Fig. 7, Curves 1 and 2).
The average molecular weight values and molecular weight
distributions for each fraction were calculated for selected
polymers (Table 4). The average values of Mw for higher
molecular weight fractions fell within 1310–3063 kg mol21
,
and higher Mw values within that fraction were noted for the
products obtained at lower temperatures and with the use
of the catalyst with the additional t-Bu groups (Lig4TiCl2).
Similar relations could be observed for the fraction of very
low molecular weight products, of the order of several thou-
sand (1500–8000 g mol21). The molecular weight distribu-
tions were very narrow for both fractions. To verify whether
liquid oligomers were produced in addition to low molecular
weight PE in some polymerization processes (MP20 and
MP24), the liquid post-reaction mixtures were analyzed by
the GC-MS method. No oligomers were found present. Hence,
activation of titanium complexes with diethylchloroalumi-
nium produced two types of active sites, with clearly differ-
ent catalytic performances.
The ethylene polymerization experiments were also con-
ducted in the presence of the salan complex of vanadium,
ꢀ
activated by MAO or EtAlCl2, at 30 C (Table 3, Entries MP43
and MP26). The catalytic system Lig4VCl2/EtAlCl2 turned out
definitely superior in terms of activity. It also yielded a poly-
mer product with a higher melting point. Moreover, its struc-
ture was linear (1.4 group CH3/1000 CH2), its molecular
weight was high (19,60,000 g mol21) and its MWD was very
narrow (1.54). This can indicate that the used catalyst is a
single-site one. MAO-activated Lig4ZrCl2 was also used in
TABLE 4 Molecular Weight and Molecular Weight Distribution of Selected Bimodal Polyethylenes Produced by Titanium Com-
plexes Activated by Et2AlCl
Entry
Complex
T (ꢀC)
Total
Peak 1
Peak 2
Mwꢂ1023 (g mol21
)
Mw/Mn
92.5
Mwꢂ1023 (g mol21
1,310
)
Mw/Mn
1.7
Mwꢂ1023 (g mol21
)
Mw/Mn
3.1
MP45
MP53
MP24
MP38
Lig1TiCl2
Lig1TiCl2
Lig4TiCl2
Lig4TiCl2
30
50
30
50
140
4.4
2.4
8.0
1.5
72
241.6
996.9
–
1,681
1.6
2.4
2,250
1.6
3,063
1.6
2.9
2,264
2.4
1.9
2120
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2014, 52, 2111–2123

JOURNAL OF
POLYMER SCIENCE
WWW.POLYMERCHEMISTRY.ORG
ARTICLE
TABLE 5 Results of Ethylene/1-Octene Copolymerization by the Lig4VCl2/EtAlCl2 and Lig4TiCl2/MAO Catalytic Systems
1-Oct in
1-Octfeed
(mol dm23
Yield
(g)
Productivity
21
copolymer
(mol%)
Mw 1023
(g mol21
)
Entry
Complex
Lig4TiCl2
Activator
MAO
)
(kg molM
)
T
m(ꢀC)
v (%)
Mw/Mn
MP48
MP49
MP50
MP51
MP52
MP26
MP29
MP27
MP30
MP28
0
0.705
2.662
3.550
3.228
2.737
1.411
0.836
0.349
0.335
0.138
10.1
135.6
–
126.9b
59
–
–
1,232
930
–
4.1a
4.1
–
14.4c
17.6c
1.5
2.6
2.7
–
0.20
0.39
0.56
0.73
0
38.0
1.6
2.6
–
50.7
33
–
46.1
–
836
229
1,968
291
238
–
39.1
122.3b
138.4
124.6b
108.7; 128.6
131.3
127.9
27
57
38
–
3.7d
Lig4VCl2
EtAlCl2
2,015.9
1,193.5
498.2
478.4
197.6
–
0.20
0.39
0.56
0.73
–
5.7
–
37
19
8.0
263
2.4
b
c
Polymerization conditions: 0.0007 mmol of Lig4VCl2; 0.07 mmol of Lig4-
TiCl2; 4 mmol of EtAlCl2; 10.5 mmol of MAO; 5 bar of ethylene, 30 min,
150 cm3 of toluene (solvent); 30 ꢀC; v—crystallinity.
Broad peak.
Bimodal—in chromatogram an additional, low intensity, and low
molecular weight peak can be seen.
d
a
Bimodal.
Polymer analyzed in the form of foil.
polymerization of ethylene but its activity was clearly low in
that process.
data in Table 5 indicate that the titanium catalyst offers a
moderate ability only to incorporate a comonomer. The
incorporation level reaches 1.6 mol % at the concentration
of 0.2 mol dm23 and it grows up to 3.7 mol % for 0.73 mol
dm23 of 1-octene in the medium. An increase in the comono-
mer incorporation leads to changes in the polymer proper-
ties. The polymer crystallinity and the melting point become
decreased and the melting peaks for copolymers are defi-
nitely wider than those for PEs. The increasing initial
1-octene concentration gradually decreases the molecular
weight of the resultant copolymers while the molecular
weight distribution broadens and even, at higher comonomer
incorporation levels, it becomes bimodal with a very low
share of the low molecular weight fraction. Thus, the
obtained copolymers are heterogeneous. The 13C NMR spec-
trum of the poly(ethylene-co-1-octene) copolymer, containing
about 3.7 mol % of 1-octene units, produced at the highest
comonomer concentration in the feed (0.73 mol dm23) is
similar to those presented in other papers for copolymers
with not very high comonomer contents, and it exhibits the
signals at 14.06 ppm (CH3), 22.88 ppm (CH2(2)), 27.37 ppm
(CH2(5) and bACH2), 30.04 ppm (eACH2, dACH2 and
CH2(4)), 32.24 ppm (CH2(3)), 34.63 ppm (aACH2, 6ACH2),
and 38.29 ppm (CH). There are no resonances within 40 to
42 ppm for carbon in the block type sequence or at about
25 ppm for carbon in the alternative sequence.
The complexes Lig4MCl2 (where M 5 Ti or V) after activation
by MAO and EtAlCl2, respectively, were used in ethylene
copolymerization with 1-octene. It was for the first time that
the copolymerization process of ethylene and a higher olefin
in the presence of salan complexes of titanium and vanadium
was researched. The reaction conditions were exactly the
same as adopted for ethylene homopolymerization when cat-
alyzed by those systems. Both the catalysts were homogene-
ous but the observed performance was different in the
process studied. As regards the titanium catalyst, addition of
a comonomer and increasing its concentration to 0.39 mol
dm23 improved the activity of the catalyst versus its per-
formance in the homopolymerization process. A further
increase of the comonomer concentration in the reaction
medium, up to 0.73 mol dm23, lowered the process yield
but its value was still nearly four times higher than in homo-
polymerization. In contrast, the catalyst Lig4VCl2/EtAlCl2
exhibited a lower activity when a comonomer was added
and its activity was declining for the growing comonomer
concentration in the reaction medium. The change in the
reaction rate for copolymerization of olefins versus homopo-
lymerization is known as the “comonomer effect” ant it was
described in numerous reports.38–41 The extent of that effect
can be different for different catalytic systems and different
comonomers, and the reasons for that effect may have chem-
ical or physical grounds,38 like for example modification of
active sites and increase in the number of catalytic centers,
changes in monomer diffusion, changes in propagation con-
stant (kp).38,39,42
As regards copolymerization in the presence of a vanadium
complex, the changes in the properties of the polymer prod-
uct for changing comonomer concentrations are not unequiv-
ocal. The melting points for copolymers obtained at higher
comonomer concentrations in the reaction medium are
higher than Tm for copolymers obtained at lower concentra-
tions, and the share of built-in comonomer segments, as
found by the FTIR method, is growing. This can suggests the
Incorporation of 1-octene into the polyethylene chain makes
another important marker for the catalytic performance. The
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2014, 52, 2111–2123
2121

JOURNAL OF
POLYMER SCIENCE
ARTICLE
WWW.POLYMERCHEMISTRY.ORG
3 V.C. Gibson, S. K. Spitzmesser, Chem. Rev. 2003, 103, 283–
315.
presence of mixed products. The copolymer product was
split into two fractions. One of them, which was insoluble in
boiling hexane, was an ethylene homopolymer. The other
fraction, which was soluble in boiling hexane, was a copoly-
merization product. As the amount of the product was very
low it was impossible to investigate for further specifications
of the fractions.
4 H. Makio, H. Terao, A. Iwashita, T. Fujita, Chem. Rev. 2011,
111, 2363–2449.
5 Y. Pan, T. Xu, S. Fu, G.-W. Yang, X.-B. Lu, Macromolecules
2013, 46, 8790–8796.
6 T. Xu, J. Liu, G.-P. Wu, X.-B. Lu, Inorg. Chem. 2011, 50,
10884–10892.
7 S. R. Golisz, J. E. Bercaw, Macromolecules 2009, 42, 8751–
8762.
CONCLUSIONS
8 E. Y. Tshuva, I. Goldberg, M. Kol, J. Am. Chem. Soc. 2000,
122, 10706–10707.
In summary, we synthesized a series of new titanium(IV)
complexes bearing tetradentate [ONNO]-type salan ligands.
Those complexes differ in the type of the linkage between
the donor nitrogen atoms and in the number of t-Bu groups
on the aromatic rings (one or two). The vanadium and zirco-
nium complexes with hexene bridges were also prepared as
reference comparative materials. These have not been so far
reported in literature, either.
9 A. Cohen, J. Kopilov, M. Lamberti, V. Venditto, M. Kol, Mac-
romolecules 2010, 43, 1689–1691.
10 G.-J. M. Meppelder, H.-T. Fan, T. Spaniol, J. Okuda, Organo-
metallics 2009, 28, 5159–5165.
11 A. Cohen, A. Yeori, J. Kopilov, I. Goldberg, M. Kol, Chem.
Commun. 2008, 2149–2151.
12 A. Yeori, I. Goldberg, M. Shuster, M. Kol, J. Am. Chem. Soc.
2006, 128, 13062–13063.
All the complexes exhibited catalytic activities towards 1-
octene polymerization in the presence of MAO and/or
borates as a cocatalyst, producing oily products. Only two
catalytic systems: Lig1TiCl2/Al(i-Bu)3/[Ph3C][B(C6F5)4] and
Lig4VCl2/MAO, produced solid materials with high average
molecular weights and bimodal distribution.
13 A. Yeori, I. Goldberg, M. Kol, Macromolecules 2007, 40,
8521–8523.
14 M. Strianese, M. Lamberti, M. Mazzeo, C. Pellecchia, Macro-
mol. Chem. Phys. 2008, 209, 585–592.
15 A. Cohen, J. Kopilov, I. Goldberg, M. Kol, Organometallics
2009, 28, 1391–1405.
16 A. Cohen, G. W. Coates, M. Kol, J. Polym. Sci. Part A:
Polym. Chem. 2013, 51, 593–600.
Polymerization of ethylene in the presence of salen complexes
has not so far been investigated in practice,14 while there is
completely no information available on their use in copoly-
merization of ethylene and higher olefins. The studied cata-
lysts show moderate activity in both processes and – just
alike in homopolymerization of 1-octene—the activator type
and the complex structure, for example, the type of linkage
between nitrogen atoms, the presence of additional t-Bu
group on each aromatic ring, and the transition metal type,
greatly influenced the catalytic activities and molecular weight
of the polymeric products. Investigations revealed also that
the titanium catalyst gave the copolymer product with moder-
ate incorporation and typical microstructure, while the vana-
dium catalyst produced the mixture of homopolymer and
copolymer.
17 S. Segal, I. Goldberg, M. Kol, Organometallics 2005, 24,
200–202.
18 V. Busico, R. Cipullo, S. Ronca, P. H. M. Budzelaar, Macro-
mol. Rapid Commun. 2001, 22, 1405–1410.
19 V. Busico, R. Cipullo, N. Friederichs, S. Ronca, G.
Talarico, M. Togrou, B. Wang, Macromolecules 2004, 37,
8201–8203.
20 E. Sergeeva, J. Kopilov, I. Goldberg, M. Kol, Chem. Com-
mun. 2009, 3053–3055.
21 S. Segal, A. Yeori, M. Shuster, Y. Rosenberg, M. Kol, Macro-
molecules 2008, 41, 1612–1617.
22 S. Gendler, A. L. Zelikoff, J. Kopilov, I. Goldbegr, M. Kol,
J. Am. Chem. Soc. 2008, 130, 2144–2145.
23 M. Sudoł, K. Czaja, M. Białek, Polimery(W) 2000, 45, 405–
410.
24 J. M. Kerr, C. J. Suckling, P. Bamfield, J. Chem. Soc., Prekin
Trans. 1990, 1, 887–895.
25 H. Schmitt, R. Lomoth, A. Magnuson, J. Park, J. Fryxelius,
M. Kritikos, J. Martensson, L. Hammarstrom, L. Sun, B.
Akermark, Chem. Eur. J. 2002, 8, 3757–3768.
ACKNOWLEDGMENTS
This work was supported by the research grant (grant No. N
N209 140840 within 2011–2014) from the Polish National Sci-
ence Centre. M. Pochwała is a recipient of a Ph.D. fellowship
from a project funded by the European Social Fund “Stypendia
26 E. Y. Tshuva, N. Gendeziuk, M. Kol, Tetrahedron Lett. 2001,
42, 6405–6407.
27 R. Olszewski, F. Stefaniak, A. Mrozek-Niecko, A. Nawrocki
(Przedsiebiorstwo Produkcyjno-Consultingowe ADOB sp. z o.o.
sp. k.), EP2039679A1.
ꢀ
doktoranckie—inwestycja w kadreR naukowa˛ wojewodztwa
opolskiego.”
28 D. Peri, S. Meker, C. M. Manna, E. Y. Tshuva, Inorg. Chem.
2011, 50, 1030–1038.
29 S. Loukiala, J. Ratilainen, J. Volkonen, K. Rissanen, Acta
Chem. Scand. 1997, 51, 1161–1168.
REFERENCES AND NOTES
30 I. Correia, J. C. Pessoa, M. T. Duarte, M. F. Minas da
Piedade, T. Jackush, T. Kiss, M. M. C. A. Castro, C. F. G.
C. Geraldes, F. Avecilla, Eur. J. Inorg. Chem. 2005, 732–
744.
1 S. D. Ittel, L. K. Johnson, M. Brookhart, Chem. Rev. 2000,
100, 1169–1203.
2 G. J. P. Britovsek, V. C. Gibson, D. F. Wass, Angew. Chem.
Int. Ed. Engl. 1999, 38, 428–447.
2122
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2014, 52, 2111–2123

JOURNAL OF
POLYMER SCIENCE
WWW.POLYMERCHEMISTRY.ORG
ARTICLE
31 C. M. Manna, E. Y. Tshuva, Dalton Trans. 2010, 39, 1182–1184.
32 A. Rivera, J. J. Rojas, J. Salazar-Barrios, M. Maldonado, J.
37 I. Kim, J.-M. Zhou, H. Chung, J. Polym. Sci. Part A: Polym.
Chem. 2000, 38, 1687–1697.
ꢀ
Rıos-Motta, Molecules 2010, 15, 4102–4110.
38 J. A. M. Awudza, P. J. T. Tait, J. Polym. Sci. Part A: Polym.
Chem. 2008, 46, 267–277.
33 A. Rivera, J. Camacho, J. Rios-Motta, M. Pojarova, M.
Dusek, Acta Cryst. 2011, E67, o2028.
39 R. van Grieken, A. Carrero, I. Suarez, B. Paredes, Macromol.
Symp. 2007, 259, 243–252.
34 D. S. Kemp, J. Org. Chem. 1971, 36, 202–204.
40 G. Xu, E. Ruckenstein, Macromolecules 1998, 31, 4724–4729.
35 S. H. Kim, J. Lee, D. J. Kim, J. H. Moon, S. Yoon, H. J. Oh,
Y. Do, Y. S. Ko, J.-H. Yim, Y. Kim, J. Org. Chem. 2009, 694,
3409–3417.
41 R. Kravchenko, R. M. Waymouth, Macromolecules 1998, 31,
1–6.
36 T. A. Immel, U. Groth, T. Huhn, Chem. Eur. J. 2010, 16,
2775–2789.
42 M. L. Ferreira, G. B. Galland, D. E. Damiani, M. A. Villar,
J. Polym. Sci. Part A: Polym. Chem. 2001, 39, 2005–2018.
WWW.MATERIALSVIEWS.COM
JOURNAL OF POLYMER SCIENCE, PART A: POLYMER CHEMISTRY 2014, 52, 2111–2123
2123