Synthesis and scale-up of imino-enamido hafnium and zirconium olefin polymerization catalysts

Article abstract of DOI:10.1021/om400189w

New imino-enamido hafnium and zirconium trimethyl complexes were prepared in high yield by reacting a mixture of an imino-enamine ligand and hafnium or zirconium tetrachloride with four equivalents of methylmagnesium bromide in toluene. A significantly im

Full text of DOI:10.1021/om400189w

Article
pubs.acs.org/Organometallics
Synthesis and Scale-up of Imino−Enamido Hafnium and Zirconium
Olefin Polymerization Catalysts
Philip P. Fontaine, Ruth Figueroa, Scott D. McCann, Darrek Mort, and Jerzy Klosin*
Corporate R&D, The Dow Chemical Company, 1776 Building, Midland, Michigan 48674, United States
S
* Supporting Information
ABSTRACT: New imino−enamido hafnium and zirconium trimethyl complexes were prepared in high yield by reacting a
mixture of an imino−enamine ligand and hafnium or zirconium tetrachloride with four equivalents of methylmagnesium bromide
in toluene. A significantly improved imine formation reaction in the final step of the ligand synthesis was developed that involves
the reaction between the keto−enamine and a reagent formed in situ by mixing 0.55 equivalents of titanium tetrachloride and 5.5
equivalents of n-butylamine. Ethylene/1-octene polymerization evaluations at 120 °C revealed that polymerization characteristics
of the imino−enamido hafnium and zirconium trimethyl derivatives were very similar to those of the tribenzyl analogues. A scale-
up of the hafnium trimethyl derivative, performed on a 255 g scale, was accomplished in an overall yield of 57% in four steps
starting with commercially available cyclohexa-1,2-dione. Hafnium and zirconium complexes derived from two isomeric imino−
enamine ligands each resulted in very high molecular weight ethylene/1-octene copolymers, which differ significantly, however,
in the level of incorporated octene. These features make these new catalysts good candidates for the preparation of olefin block
copolymers via chain-shuttling polymerization.
produce ethylene/1-octene copolymers with higher molecular
weight and 1-octene content than the imino−amido analogues.
INTRODUCTION
■
Molecular olefin polymerization catalysts1 have attracted a lot
of attention over the last 20 years due to their ability to produce
tailored polyolefins. Research in this area has been motivated by
the desire to discover new catalysts for the production of new
polyolefin products2 and to introduce improvements in the
current polymerization processes.3 Recently we4 investigated
catalysts based on an imino−amido ligand framework5,6 (e.g.,
1) and discovered that such catalysts exhibit several desirable
features including good catalytic activities at reactor temper-
atures above 100 °C, the ability to form very high molecular
weight polymers, and the capability to undergo reversible chain
transfer with diethylzinc to produce olefin block copolymers.4b
One major drawback inherent to these precatalysts, however, is
their thermal instability, which results in the formation of
isomeric complexes (by 1,2-Me shift) that exhibit inferior
polymerization characteristics.4c To eliminate decomposition
pathways found in imino−amido complexes, we designed and
prepared imino−enamido complexes (e.g., 2, 3) that are
resistant to isomerization and decomposition at elevated
temperatures.7 Imino−enamido complexes not only are
thermally robust but exhibit higher catalytic activities and
The reported synthesis of imino−enamido complexes 2 and
3 is reproducible and effective,7 but it has some disadvantages
for the larger scale preparation of such complexes. Herein, we
report the preparation and olefin polymerization evaluation of
trimethyl hafnium and zirconium analogues of 2 and 3 using a
much improved synthetic protocol, the effectiveness of which
Received: March 7, 2013
Published: May 10, 2013
© 2013 American Chemical Society
2963
dx.doi.org/10.1021/om400189w | Organometallics 2013, 32, 2963−2972

Organometallics
Article
was subsequently demonstrated by a large-scale preparation of
the hafnium analogue. Additionally, the synthesis and
evaluation of complexes derived from the isomeric imino−
enamine ligand are also described.
from the titanium byproducts, and the need to use an expensive
metal precursor (HfBn4) in the final step of the synthesis.
We desired to develop a more cost-effective and practical
synthesis of these precatalysts to accelerate our discovery and
scale-up efforts. The two steps in the synthesis that required the
most significant improvements were the last step in the ligand
synthesis (imine formation) and the preparation of the metal
complex. To improve the ligand synthesis, we focused our
attention on an in situ preparation of the titanium reagent and
its reaction with keto−enamine 6. The synthesis and utility of
TiCl4/amine reagents have been previously described by
Carlson and Nilsson in the preparation of enamines8 and
imines9 from carbonyl compounds. After an optimization of key
reaction parameters, the best conditions identified involved the
preparation of the titanium reagent from the reaction of 0.55
equivalents of TiCl4 and 5.5 equivalents of n-BuNH2 in toluene
at ambient temperature (exothermic reaction) and its
subsequent reaction with one equivalent of keto−enamine 6,
which provided the desired ligand 8 in 96% yield in 3 h
(Scheme 2). During the workup, a solution of 8 was stirred
RESULTS AND DISCUSSION
■
Development of New Synthesis for Imino−Enamido
Precatalysts. The original synthetic route we developed for
the synthesis of imino−enamido complexes with different
substituents at the imino and amido nitrogen atoms is shown in
Scheme 1.7 Cyclohexa-1,2-dione (4) was condensed with
Scheme 1. Original Route to Imino−Enamido Precatalyst 27
Scheme 2. Improved Synthesis of Ligand 8
over K2CO3 to ensure removal of any acidic byproducts. This
procedure provided 8 together with about 1 mol % of 9, the
double-bond isomer of 8. This procedure is far superior to the
one outlined in Scheme 1, as it is significantly less expensive, as
well as easier and faster to execute.10
morpholine to afford the morpholine-based keto−enamine 5,
which was subsequently transaminated with 2,6-diisopropylani-
line to produce keto−enamine 6. The imine functionality was
then installed by treating 6 with the titanium reagent 7, which
in turn was prepared and isolated from the reaction of
tetrakis(dimethylamino)titanium (Ti(NMe2)4) and an excess of
n-butyl amine. Titanium-mediated imine formation was found
to be necessary for synthesizing 8; when organic acids such as
formic acid were used instead, a complete rearrangement of 8
to its isomeric imino−enamine 9 (with double-bond trans-
position) occurred. The desired precatalyst 2 was then obtained
from the reaction of the ligand 8 with a metal precursor,
tetrabenzylhafnium (HfBn4).
Although this method is reliable and can be used for the
preparation of imino−enamido analogues, it is not very
practical for the larger scale synthesis of such precatalysts.
There are several issues with the above synthesis including the
need to prepare and isolate the discrete titanium−imido
reagent 7 from the expensive precursor Ti(NMe2)4, very long
(2 days) reaction times between keto−enamine 6 and 7 to form
the desired ligand 8, difficult separation of the ligand 8 away
With the new straightforward synthesis of ligand 8
developed, we moved our attention to improving the
metalation step to provide the final precatalyst. Since complexes
containing metal−methyl bonds are often used as precatalysts
for olefin polymerization, and methylating reagents are less
expensive than benzylating reagents, we decided to explore the
preparation and evaluation of the trimethyl analogue of
complex 2. It was hoped that the switch from tribenzyl to
trimethyl derivatives would result in catalysts exhibiting
identical or very similar polymerization behavior.11 To test
this hypothesis, the hafnium trimethyl derivative 11 was
prepared initially by the reaction sequence shown in Scheme
3. Reaction of iodine (three equivalents) with 2 in methylene
chloride gave the desired hafnium triiodo derivative 10 in 83%
yield. Complex 10 exhibits Cs symmetry in solution, as shown
by NMR spectroscopy.12 In addition to 1D and 2D NMR and
elemental analysis, complex 10 was also characterized by single-
crystal X-ray analysis (vide infra). Reaction of 10 with three
equivalents of MeMgBr in toluene led to clean formation of the
1
desired trimethyl complex 11 in 99% yield. The H NMR
2964
dx.doi.org/10.1021/om400189w | Organometallics 2013, 32, 2963−2972

Organometallics
Article
1
yellow crystalline solid in very high purity, as indicated by H
Scheme 3. Synthesis of 11 by Two Different Routes
NMR spectroscopy. This one-step procedure is analogous to
the one introduced by Resconi for the preparation of bis-
metallocene14 and CGC15 dimethyl complexes, whereby the
respective ligands and ZrCl4/TiCl4 were reacted with four
equivalents of MeLi in a toluene solution.
Synthesis of Complexes 12−14. The zirconium analogue
12 was also prepared using an analogous one-step procedure as
shown in Scheme 4. The reaction of a mixture of ligand 8 and
Scheme 4. Synthesis of Complex 12
ZrCl4 in toluene with four equivalents of MeMgBr gave the
1
desired complex 12 in 96% yield after workup. The H NMR
spectrum of 11 shows the characteristic singlet at 0.50 ppm,
which is assigned to the protons of the three methyl groups
bound to the hafnium metal center. Additionally, the 1H NMR
spectrum of 11 shows one i-Pr methine signal (at 3.46 ppm)
and two doublets (at 1.21 and 1.42 ppm) corresponding to the
i-Pr methyl groups, indicating that both i-Pr groups have the
same chemical environment. This chemical shift pattern is the
same as that observed in 2 except that all the resonances of 11
are shifted downfield by about 0.1−0.2 ppm. The appearance of
two separate resonances for the i-Pr methyl groups is due to the
hindered rotation of the 2,6-diisopropylphenyl fragment along
the N−C(ipso) bond, which leads to a different chemical
environment of the methyl groups pointed toward and away
from the cyclohex-2-enylidene bridge. 1D NOESY experiments
indicated that the upfield doublet (at 1.21 ppm) corresponds to
the methyl groups pointing toward the bridge of the ligand. 1D
TOCSY spectroscopy is very useful in characterizing these
imino−enamido complexes, as it allows for unequivocal
identification of the protons associated with the two separate
spin systems of the cyclohex-2-enylidene and N-n-alkyl
fragments. For example, irradiation of the triplet at 0.77 ppm
corresponding to the methyl group of the butyl chain in 11
clearly identified the remaining protons of the butyl group
resonances at 1.08, 1.34, and 3.34 ppm. In addition to 1D and
2D NMR spectroscopy, 11 was also characterized by elemental
analysis and single-crystal X-ray analysis at 200 K (vide infra).13
A polymerization evaluation with 11 revealed that its catalytic
characteristics are very similar to those of the tribenzyl analogue
2 (vide infra). The synthesis via the hafnium triiodo derivative
10 was adequate for providing a large enough sample of 11 for
polymerization evaluations, but it was not sufficient for a larger
scale preparation of this precatalyst due to its length and the
expense of HfBn4. This difficulty was overcome by the
successful synthesis of 11 by a much simpler route involving
the addition of four equivalents of MeMgBr to a suspension of
8 and HfCl4 in toluene at ambient temperature (Scheme 3).
The initially formed yellow suspension changed color within
minutes to black, which is typical for this type of reaction. After
a short reaction time (∼1 h) the solvent was removed under
reduced pressure, the product was extracted with hexane, and
the resulting suspension was filtered to remove the insoluble
salts. Precatalyst 11 was obtained in 79% yield as a bright
spectrum of 12 is very similar to that of 11, except for the
chemical shift of the Zr(CH3)3 groups, which resonate at δ 0.74
ppm, 0.24 ppm downfield relative to the Hf(CH3)3 peak.
Interestingly, the chemical shift trend is reversed in the
13C{1H} NMR spectra, with Zr(CH3)3 resonating at 48.60 ppm
and Hf(CH3)3 appearing downfield at 60.40 ppm. The
remaining resonances in the 13C{1H} spectra of 11 and 12
are very similar to each other.
Although hafnium tribenzyl analog containing the isomerized
ligand 9 resulted in lower polymerization activities,7 it was of
interest to us to determine if the trimethyl complexes could also
be prepared from ligand 9 using the one-step procedure and
what their detailed polymerization characteristics would be. A
mixture of ligand 9 and HfCl4 or ZrCl4 was reacted with four
equivalents of MeMgBr, giving the desired complexes 13 and
14 in 45% and 39% yield, respectively (Scheme 5). The lower
Scheme 5. Synthesis of Complexes 13 and 14
yields for these isomers arise from the formation of a small
amount (∼20%) of a second, unidentified product,16 which
needs to be separated from the desired complexes by
crystallization.
1
The H NMR spectrum of 13 shows the same number of
resonances with the same splitting patterns as its isomeric
hafnium complex 11, but surprisingly there are substantial
chemical shift differences observed for the respective protons in
these two complexes. All resonances associated with the six-
membered ring and the n-butyl fragments in 13 are shifted
downfield relative to those found in 11. For example, the
chemical shifts of the double bond and methylene (α to the
imine nitrogen atom) protons in 13 are shifted downfield by
0.53 and 0.33 ppm, respectively, compared to the same
2965
dx.doi.org/10.1021/om400189w | Organometallics 2013, 32, 2963−2972

Organometallics
Article
resonances in 11, which appear at 4.66 and 3.34 ppm. On the
other hand, all resonances (both aromatic and aliphatic)
associated with the 2,6-diisopropylphenyl group are shifted
upfield in 13 compared to those in 11. This upfield chemical
shift is most likely due to magnetic anisotropic shielding
induced by the imine bond, which is much closer to the 2,6-
diisopropylphenyl group in 13 than the distance between the
carbon−carbon double bond and 2,6-diisopropylphenyl group
in 11. Very similar chemical shift differences are also observed
for the zirconium analogues 12 and 14. The only resonances
that are very similar between isomers 11/13 and 12/14 belong
to the Hf(CH3)3 and Zr(CH3)3 fragments, respectively.
Single-Crystal X-ray Structures of 10−14. Molecular
structures of 10, 11, and 13 are shown in Figures 1−3,
Figure 3. Molecular structure of 13. Hydrogen atoms were removed
for clarity. Thermal ellipsoids are shown at 40% probability. Selected
bonds (Å) and angles (deg): Hf−N1 = 2.0998(13), Hf−N2 =
2.3550(13), Hf−C23 = 2.2153(17), Hf−C24 = 2.2150(17), Hf−C25
= 2.2774(16), C1−N1 = 1.379(2), C6−N2 = 1.297(2), C1−C2 =
1.355(2), C5−C6 = 1.500(2), N1−Hf−N2 = 69.34(5), N1−Hf1−C25
= 151.43(5).
example, the C5−C6 bond lengths of 1.343(3) and 1.345(3) Å
for 11 and 12, respectively, clearly indicate the presence of the
double bond in the cyclohex-2-enylidene fragment bonded to
the N−Ar fragment. On the other hand, the C5−C6 bond
lengths in 13 and 14 equal 1.500(2) and 1.498(2) Å,
respectively, indicating the presence of a single bond. As
expected, bond lengths and angles in tribenzyl complexes 2 and
3 are very similar to those found in trimethyl analogues 11 and
12. The zirconium analogues consistently have slightly longer
bond lengths around the metal center than the hafnium
analogues. The difference between the M−N(imino) and M−
N(enamido) bond lengths in complexes 2, 3, 10, 11, and 12
falls within a narrow range of 0.198−0.213 Å, which is
noticeably smaller than those found in complexes 13 and 14
(0.255−0.273 Å) containing isomerized ligand 9. Although the
difference between the Hf−N(imino) and Hf−N(enamido)
bond lengths in the triiodo derivative 10 is very similar (0.198
Å) to that found in the other complexes containing ligand 8,
both Hf−N(imino) and Hf−N(enamido) bond lengths are
shorter by about 0.05 Å than those in 2 and 11. Also, the N1−
Hf(Zr)−C25 bond angles are 161.19(8)° and 156.68(7)° for
11 and 12, respectively, noticeably smaller than the analogous
angle (N1−Hf−I2) in 10 (169.75(5)°).
Figure 1. Molecular structure of 10. Hydrogen atoms were removed
for clarity. Thermal ellipsoids are shown at 40% probability. Selected
bonds (Å) and angles (deg): Hf−N1 = 2.244(2), Hf−N2 = 2.046(2),
Hf−I1 = 2.7388(2), Hf−I2 = 2.7620(2), Hf−I3 = 2.7350(2), C1−N1
= 1.300(3), C6−N2 = 1.412(3), C1−C2 = 1.498(3), C5−C6 =
1.343(3), N2−Hf−N1 = 73.08(7), N1−Hf1−I2 = 169.75(5).
Polymerization Results. Ethylene/1-octene copolymeriza-
tion reactions were conducted in a 2 L batch reactor under
constant ethylene pressure. Precatalysts were activated with 1.2
equivalents (relative to precatalyst) of [HNMe(C18H37)2][B-
(C6F5)4]. All polymerization reactions were carried out for 10
min and stopped by venting the ethylene pressure. Polymer-
ization data are shown in Table 2. The catalytic efficiency and
molecular weight (Mw) of polymers produced by 11 and 12 are
very comparable to those obtained with tribenzyl analogues 2
and 3 (runs 1−4). The efficiency for the hafnium-based
catalysts (2, 11) is about 4 times higher than the zirconium
analogues (3, 12). Octene incorporation, as determined by IR
spectroscopy, decreases from about 7−8 mol % for the hafnium
analogues to 5 mol % for the zirconium analogues. As
anticipated from the previous study,7 the efficiency of the
isomeric complexes 13 and 14 is lower than that observed for
11 and 12. Similarly, octene incorporation for 13 (2.5 mol %)
and 14 (1.3 mol %) is significantly lower than that observed for
Figure 2. Molecular structure of 11. Hydrogen atoms were removed
for clarity. Thermal ellipsoids are shown at 40% probability. Selected
bonds (Å) and angles (deg): Hf−N1 = 2.3194(16), Hf−N2 =
2.1196(15), Hf−C23 = 2.214(2), Hf−C24 = 2.232(2), Hf−C25 =
2.222(2), C1−N1 = 1.288(2), C6−N2 = 1.397(2), C1−C2 =
1.504(3), C5−C6 = 1.343(3), N1−Hf−N2 = 70.81(5), N1−Hf1−
C25 = 161.19(8).
respectively, and selected bond lengths and angles for 10−14
are presented in Table 1.17 Metric parameters obtained for 10−
14 complexes clearly confirm the identity of the individual
isomers originally established by 1D NOESY experiments. For
2966
dx.doi.org/10.1021/om400189w | Organometallics 2013, 32, 2963−2972

Organometallics
Article
Table 1. Bond Lengths (Å) and Angles (deg) for Complexes 2, 3, and 10−14
bond/angle
10
27
11
37
12
13
14
Hf(Zr)−N1
2.244(2)
2.046(2)
2.313(2)
2.100(2)
2.243(2)
2.288(2)
2.266(2)
1.498(3)
1.523(4)
1.518(4)
1.491(3)
1.347(3)
1.459(3)
1.290(3)
1.486(3)
1.403(2)
1.446(2)
72.03(6)
101.0(1)
120.24(8)
98.03(9)
2.3194(16)
2.1196(15)
2.214(2)
2.232(2)
2.222(2)
1.504(3)
1.472(6)
1.457(6)
1.502(3)
1.343(3)
1.469(2)
1.288(2)
1.475(2)
1.397(2)
1.444(2)
70.81(5)
109.55(9)
96.30(11)
102.37(10)
2.332(2)
2.132(2)
2.266(2)
2.303(3)
2.286(2)
1.508(3)
1.517(4)
1.517(4)
1.487(3)
1.349(3)
1.460(3)
1.290(3)
1.478(3)
1.399(3)
1.447(2)
71.21(6)
102.6(1)
122.38(8)
98.22(9)
2.3357(13)
2.1442(12)
2.2339(18)
2.2577(18)
2.237(2)
2.0998(13)
2.3550(13)
2.2153(17)
2.2150(17)
2.2774(16)
1.355(2)
1.499(2)
1.524(2)
1.530(2)
1.500(2)
1.466(2)
1.379(2)
1.463(2)
1.297(2)
1.4470(19)
69.34(5)
107.32(7)
97.13(6)
97.21(6)
2.1077(11)
2.3810(10)
2.2361(14)
2.2372(14)
2.2602(13)
1.3559(18)
1.4937(19)
1.5193(18)
1.5291(18)
1.4984(17)
1.4700(17)
1.3803(16)
1.4606(15)
1.2970(16)
1.4444(15)
68.61(4)
Hf(Zr)−N2
Hf(Zr)−C2318
Hf(Zr)−C2418
Hf(Zr)−C2518
C1−C2
C2−C3
C3−C4
C4−C5
C5−C6
C1−C6
N1−C1
N1−C7
N2−C6
1.498(3)
1.522(4)
1.521(4)
1.502(3)
1.343(3)
1.461(3)
1.300(3)
1.485(3)
1.412(3)
1.455(3)
73.08(7)
1.504(2)
1.497(3)
1.505(3)
1.502(2)
1.345(2)
1.4713(19)
1.2874(19)
1.4720(18)
1.3962(18)
1.4396(17)
70.35(4)
N2−C11
N1−Hf(Zr)−N2
C23−Hf(Zr)−C24
C24−Hf(Zr)−C2518
C23−Hf(Zr)−C2518
109.36(7)
94.87(9)
108.35(6)
97.34(5)
105.11(8)
97.65(5)
a
Table 2. Polymerization Data for Complexes 2, 3, 11−14, and CGC
b
run no.
catalyst (μmol)
polymer yield (g)
efficiency (g of polymer/ mmol of metal)
Mw × 10−3/PDI
Tm (°C)
octene content (mol %)
1
2
3
4
5
6
7
2 (0.2)
48.8
11.1
49.6
11.8
24.8
16.5
82.2
244 000
55 500
248 000
59 000
31 000
20 625
411 000
682/3.5
581/2.7
743/3.1
584/2.5
641/3.4
347/2.7
100/2.2
80.3
95.2
8.3
5.1
3 (0.2)
11 (0.2)
12 (0.2)
13 (0.8)
14 (0.8)
CGC (0.3)
83.1
7.2
93.0
5.2
114.4
118.3
52.8
2.5
1.3
15.9
a
Polymerization conditions: 2 L batch reactor, 533 g of Isopar-E; temp = 120 °C; 250 g of 1-octene; ethylene pressure = 460 psi;
b
precatalyst:activator = 1:1.2; activator: [HNMe(C18H37)2][B(C6F5)4]; 10 μmol of MMAO; reaction time 10 min. Determined by IR; CGC = {(5η-
C5Me4)(SiMe2-N-t-Bu)}TiMe2.
11 (7.2 mol %) and 12 (5.2 mol %). It is interesting that the
complexes derived from ligands 8 and 9 comprise almost
identical steric environments around the metal center and yet
lead to copolymers with substantial differences in octene
incorporation. Since imino−enamido complexes were pre-
viously found to undergo very effective chain transfer reactions
with diethyl zinc,7 complexes 11−14 have potential utility for
the preparation of olefin block copolymers via chain shuttling
polymerization with variable density splits between hard and
soft segments.2e,4b Important to such chemistry is the ability of
the catalysts to produce very high molecular weight
copolymers. Molecular weights of copolymers produced by
11−14 are very high, with Mw values ranging between 350 000
and 750 000.19 Polydispersity indexes for the polymer samples
obtained from all precatalysts are somewhat larger than 2,
which is indicative of multisite catalyst behavior. The identity of
the active species formed under catalytic conditions is unknown
at the present time.
Scale-up of 11. When assessing the industrial viability of
any catalyst, critical factors to consider include the ease and
cost associated with obtaining large quantities of the desired
species, along with the mitigation of any potential safety
hazards. Additionally, many of the laboratory techniques that
are typically utilized for smaller scale reactions (e.g., column
chromatography, multiple solvent washes) are often undesir-
able for larger scale reactions, for which the simplicity of the
reaction workup is also an important criterion. For this reason,
high-yielding and robust synthetic methodologies are usually
sought prior to scale-up. In the present case, we wished to
evaluate our optimized synthetic methodology by applying it to
the preparation of a comparatively large amount (255 g) of 11
needed for pilot plant evaluations. Generally speaking, it is
desired to reduce solvent volumes in order to concentrate the
reagents and thereby speed up the product-forming reactions.
However, it is also important to have an appropriate amount of
solvent to accommodate the heat rise associated with an
exothermic reaction. For the initial transformation (4 → 5),
however, the calculated exotherm is negligible, and in fact the
reaction requires external heat and water removal to proceed.
Ultimately, a reaction concentration of 0.84 M in 4 was chosen,
which led to a convenient reaction time of 3 h (Scheme 6).
Subsequent to the reaction, the solution was cooled to room
temperature, and immiscible dark oil settled at the bottom of
the flask. This oil, presumably a small amount of decomposed
cyclohexa-1,2-dione, is easily removed from the product-
containing solution by decantation. Removal of the solvent
and trituration with hexane furnished 5 as an orange-brown
solid in high purity and 95% yield. The material was carried
along to the next step without further purification.
The second step in the scale-up, conversion of 5 into the
keto−enamine derivative 6 (the transamination reaction), was
carried out on a Schlenk line in a similar fashion to the previous
2967
dx.doi.org/10.1021/om400189w | Organometallics 2013, 32, 2963−2972

Organometallics
Article
workup times for the scale-up of this step since even a very
small amount of acid can result in the rapid and irreversible
isomerization of the product. It is not trivial to quantify the
thermochemistry of this complicated reaction via calculation, so
we opted to determine the reaction heat by measuring it
directly during a small-scale reaction in a well-insulated reactor.
A heat release of 8.9 cal/g of titanium was measured for the
addition of TiCl4 to n-BuNH2, and 2.8 cal/g of 6 was measured
for the reaction of 6 with the in situ generated titanium reagent.
Considering this energy release, a 0.2 M solution of TiCl4 in
toluene was implemented for the scale-up step. Care must be
taken to avoid the high temperatures that would be expected to
accelerate the isomerization reaction (8 → 9). Upon the
addition of 0.31 mol of TiCl4 to a solution of 3.07 mol of n-
BuNH2 in 1.5 L of toluene, a temperature rise up to 60 °C was
observed and the solution was allowed to cool back down to 30
°C before 6 was added. The product was isolated as an oil and
was stored at −20 °C until used in the final step of the scale-up.
Metalation of 8 was also carried out analogously to what has
been described in a small-scale preparation of 11. One practical
difference that arose during this step was that the addition of
the Grignard reagent took about 1 h to complete (to maintain a
reaction temperature of less than 40 °C). Analysis of the
product 11 by 1H NMR showed excellent purity with less than
2% of isomer 13, for a net yield of 255 g (92%). Very good
yields were obtained in each synthetic step, with an overall yield
for the four steps of 57%.
Scheme 6. Summary of the Large-Scale Synthesis of 11
CONCLUSIONS
■
The main objectives of this work were to identify a new version
of the original imino−enamido precatalyst, develop its practical
and economical synthesis, and then apply it to its macroscale
preparation. Polymerization results indicate that trimethyl
complexes 11 and 12 exhibit very similar behavior to their
tribenzyl analogues 2 and 3. The synthesis of 11 and 12 was
accomplished via a much simpler and more economical route
than that used in the preparation of the original imino−
enamido complexes 2 and 3. The most significant improvement
in the synthesis of the trimethyl complexes came during the
ligand preparation step. The new method for ligand synthesis
involves the use of an in situ generated titanium reagent (from
TiCl4 and n-BuNH2) rather than the application of an
expensive reagent to promote the imine formation, which
resulted in shortening the reaction time from 2 days to 3 h
while improving the isolated yield to almost quantitative (96%).
Additionally, this new procedure requires much less solvent,
which further improves the efficiency of the synthesis. In the
last step of the synthesis, the metalation reactions, the desired
complexes were obtained by the reaction of the ligand 8 and
the respective metal chlorides (HfCl4 and ZrCl4) with four
equivalents of MeMgBr in toluene in 80−90% yields. This
approach is significantly more cost-effective than the previous
synthesis, which required the use of HfBn4 and ZrBn4 as
synthetic reagents. The practicality of this new method was
demonstrated by applying it successfully to the macroscale
synthesis of 11, resulting in quantities suitable for pilot plant
evaluations. Specifically, the scale-up resulted in the synthesis of
255 g of 11 in a good overall yield of 57%. The new synthetic
method also allowed for the preparation of hafnium and
zirconium complexes derived from the isomerized imino−
enamine ligand 9. When activated with trialkylammonium
borate, these complexes (13, 14) gave polymers with very high
molecular weights and low 1-octene incorporation. On the
step with the notable exception that a mechanical stirrer was
used to ensure adequate mixing of the reaction mixture. In this
case, lower than reflux temperatures were used, and the
formation of a large quantity of precipitate (morpholino
sulfonate) resulted in a thick slurry, which proved difficult to
stir with magnetic stirring. This in turn resulted in a high degree
of variability in the reaction times for the various test reactions.
With a mechanical stirrer, and a concentration of 0.55 M in 5,
the scale-up could be readily executed with reaction times of
less than 2 h. Thermodynamics of this step are not a concern,
as the transamination is expected to be close to thermoneutral,
with the driving force being the precipitation of morpholino
sulfonate byproduct. After workup and recrystallization, the
desired species was isolated in good yield (71%) and excellent
purity. A net yield for 6 of 185.3 g was obtained after
recrystallization. While the product of this reaction forms very
cleanly, the crude residue typically contains small amounts of
2,6-diisopropylaniline, which is a solubilizing agent and reduces
yields. It is worth noting that it is important to isolate 6 in high
purity, as the subsequent step yields an oil that is not amenable
to easy purification and so must be carried through to the final
metalation in essentially crude form.
The final step of the ligand synthesis (step 3), conversion of
6 into ligand 8, requires the greatest caution from a
thermochemistry standpoint. It has been described previously7
that undesirable acid-catalyzed isomerization of 8 can occur and
is thermodynamically favored. Reaction conditions for this step
were proven to work very well on a small scale, but there was a
concern related to heat transfer differences and reaction and
2968
dx.doi.org/10.1021/om400189w | Organometallics 2013, 32, 2963−2972

Organometallics
Article
1-yl)-2,6-bis(1-methylethyl)benzenaminato-κN)-tris(phenylmethyl)
hafnium (2)7 dissolved in 20 mL of CH2Cl2 at room temperature was
added 1.54 g (6.07 mmol) of iodine dissolved in 60 mL of CH2Cl2
over a period of 20 min. After stirring for 10 min at ambient
temperature, the solvent was removed under reduced pressure. To the
residue was added 30 mL of hexane, and the suspension was stirred for
15 min. Yellow product was collected on the frit, washed with 15 mL
of hexane, and dried under reduced pressure to give 1.47 g of product
(82.7% yield). 1H NMR (C6D6, 500 MHz, 30 °C): δ 7.24 (m, 1H, iPr2-
other hand, complexes derived from ligand 8 (11, 12)
produced copolymers with noticeably higher octene incorpo-
ration. Each of the four complexes investigated (11−14)
resulted in ethylene/1-octene copolymers with very high
molecular weights but with different octene contents. These
features coupled with the demonstrated ability of imino−
enamido complexes to undergo very effective chain transfer
reactions should make complexes 11−14 good candidates for
the preparation of a variety of olefin block copolymers via
chain-shuttling polymerization.
i
3
para-Ph), 7.17 (m, 2H, Pr2-meta-Ph), 4.72 (t, 1H, JH−H = 4.9 Hz,
3
H5), 3.59 (m, 2H, H7), 3.56 (sept. 2H, JH−H = 6.8 Hz, CH(CH3)2),
3
3
1.68 (q, 2H, JH−H = 5.7 Hz, H4), 1.66 (t, 2H, JH−H = 6.4 Hz, H2),
3
1.61 (d, 6H, JH−H = 6.9 Hz, CH(CH3)2), 1.58 (m, 2H, H8), 1.11 (d,
EXPERIMENTAL SECTION
■
6H, 3JH−H = 6.8 Hz, CH(CH3)2), 1.10 (sex., 2H, 3JH−H = 7.1 Hz, H9),
General Considerations. All solvents and reagents were obtained
from commercial sources and used as received unless otherwise noted.
Toluene, hexanes, CH2Cl2, and C6D6 were dried and degassed
according to published procedures. NMR spectra were recorded on
Varian VNMRS-400 and VNMRS-500 and Bruker Avance-400
3
3
1.07 (p, 2H, JH−H = 6.0 Hz, H3), 0.79 (t, 3H, JH−H = 7.4 Hz, H10).
13C{1H} NMR (C6D6, 125 MHz, 30 °C): δ 176.25 (NC), 149.75
(quat.), 144.81 (quat.), 143.98 (quat.), 128.20 (iPr2-para-Ph), 125.05
(iPr2-meta-Ph), 123.80 (C5), 52.23 (C7), 30.53 (C8), 28.95
(CH(CH3)2), 28.70 (C2), 26.21 (CH(CH3)2), 24.42 (C4), 24.35
(CH(CH3)2), 21.92 (C3), 20.98 (C9), 13.84 (C10). Anal. Calcd for
C25H42N2HfI3: C, 29.87; H, 3.76; N, 3.17. Found: C, 30.12; H, 3.62;
N, 3.33.
1
spectrometers. H NMR data are reported as follows: chemical shift
(multiplicity (br = broad, s = singlet, d = doublet, t = triplet, q =
quartet, p = pentet, quint. = quintet, sex. = sextet, sept. = septet and m
= multiplet), integration, and assignment). Chemical shifts for 1H
NMR data are reported in ppm downfield from internal
tetramethylsilane (TMS, δ scale) using residual protons in the
deuterated solvent (C6D6, 7.15 ppm) as references. 13C NMR data
were determined with 1H decoupling, and the chemical shifts are
reported in ppm vs tetramethylsilane (C6D6, 128 ppm). Elemental
analyses were performed at Midwest Microlab, LLC.
Synthetic Details. Preparation of N-((6E)-6-(Butylimino)-1-
cyclohexen-1-yl)-2,6-bis(1-methylethyl)benzenamine (8). To 6.065
g (82.9 mmol) of n-BuNH2 dissolved in 40 mL of toluene was added
(within two minutes) 1.573 g (8.3 mmol) of TiCl4 dissolved in 5 mL
of toluene. The reaction mixture turned red during the addition and
then orange a few minutes after the addition was complete. The
temperature increased during the addition from 27 to 49 °C. After
stirring for 0.5 h at ambient temperature 4.055 g (14.9 mmol) of 2-
((2,6-bis(1-methylethyl)phenyl)amino)-2-cyclohexen-1-one (6) was
added as a solid. The reaction mixture was stirred at ambient
temperature for 3.25 h and then filtered (light yellow filtrate) into a
vessel containing 2.1 g of anhydrous K2CO3. The suspension was
stirred for 2 h with K2CO3 and then filtered to give a colorless solution
(K2CO3 absorbed yellow color from the solution). The solvent was
removed under reduced pressure, leaving 4.66 g of colorless oil. Yield:
95.6%. 1H NMR (C6D6, 500 MHz, 30 °C): δ 7.23−7.17 (m, 3H, i-Pr2-
Ph), 6.89 (s, 1H, NH), 4.78 (t, 1H, J = 4.6 Hz, H5), 3.39 (sept., 2H, 3J
Preparation of (N-((6E)-6-(butylimino-κN)-1-cyclohexen-1-yl)-2,6-
bis(1-methylethyl)benzenaminato-κN)trimethylhafnium (11). To a
30 mL toluene solution containing 0.7486 g (0.85 mmol) of (N-((6E)-
6-(butylimino-κN)-1-cyclohexen-1-yl)-2,6-bis(1-methylethyl)-
benzenaminato-κN)triiodohafnium (10) was added 0.88 mL (2.65
mmol) of 3 M MeMgBr solution in ether at ambient temperature. The
reaction mixture was stirred for 20 min followed by solvent removal
under reduced pressure. The residue was extracted with 30 mL of
hexane and filtered. The solvent was removed from the filtrate, leaving
0.462 g of product as a yellow crystalline solid. Yield: 99.4%. 1H NMR
i
(C6D6, 500 MHz, 30 °C): δ 7.19−7.24 (m, 3H, Pr2-Ph), 4.66 (t, 1H,
3JH−H = 4.9 Hz, H5), 3.46 (sept. 2H, 3JH−H = 7.0 Hz, CH(CH3)2), 3.34
3
3
(m, 2H, H7), 1.89 (t, 2H, JH−H = 6.6 Hz, H2), 1.88 (q, 2H, JH−H
=
3
3
= 6.9 Hz, CH(CH3)2), 3.26 (t, 2H, J = 6.8 Hz, H7), 2.09 (pseudo t,
5.5 Hz, H4), 1.42 (d, 6H, JH−H = 7.0 Hz, CH(CH3)2), 1.34 (m, 2H,
2H, 3J = 6.6 Hz, H2), 1.96 (q, 2H, 3J = 5.5 Hz, H4), 1.69 (p of m, 2H,
3
3
H8), 1.28 (p, 2H, JH−H = 6.2 Hz, H3), 1.21 (d, 6H, JH−H = 6.9 Hz,
3
3J = 7.8 Hz, H8), 1.53 (quint., 2H, J = 6.5 Hz, H3), 1.44 (sex. of m,
CH(CH3)2), 1.08 (sex., 2H, 3JH−H = 7.5 Hz, H9), 0.77 (t, 3H, 3JH−H
=
2H, 3J = 7.0 Hz, H9), 1.22 (d, 12H, 3J = 6.8 Hz, CH(CH3)2), 0.94 (t,
7.4 Hz, H10), 0.50 (s, 9H, Hf−CH3). 13C{1H} NMR (C6D6, 125
MHz, 30 °C): δ 175.30 (NC), 152.07 (quat.), 144.98 (quat.),
144.31 (quat.), 126.07 (iPr2-para-Ph), 124.16 (iPr2-meta-Ph), 114.71
(C5), 60.40 (Hf-CH3), 49.57 (C7), 31.00 (C8), 28.48 (CH(CH3)2),
28.00 (C2), 26.00 (CH(CH3)2), 24.69 (C4), 24.53 (CH(CH3)2),
23.12 (C3), 21.00 (C9), 13.82 (C10). Anal. Calcd for C25H42HfN2: C,
54.68; H, 7.71; N, 5,10. Found: C, 54.94; H, 7.56; N, 5,10.
3
3H, J = 7.3 Hz, H10) ppm. 13C NMR (C6D6, 125 MHz, 30 °C): δ
160.61, 147.38, 140.90, 137.89, 126.83, 123.80, 103.26, 49.84, 33.71,
28.69, 26.80, 24.23, 23.92, 23.39, 21.03, 14.21. HSQC (C6D6, 500
MHz): δ (7.23−7.17, 126.83/123.80), (4.78, 103.26), (3.39, 28.69),
(3.26, 49.84), (2.09, 26.80), (1.96, 23.92), (1.69, 33.71), (1.52, 23.39),
(1.44, 21.03), (1.22, 24.23), (0.94, 14.21) ppm. HRMS (ESI, M +
H)+): (m/z) calcd for C22H35N2 327.277, found 327.280.
Preparation of (N-((6E)-6-(butylimino-κN)-1-cyclohexen-1-yl)-2,6-
bis(1-methylethyl)benzenaminato-κN)trimethylhafnium (11). To a
toluene solution (60 mL) of N-((6E)-6-(butylimino)-1-cyclohexen-1-
yl)-2,6-bis(1-methylethyl)benzenamine (8) (3.03 g, 9.28 mmol) and
HfCl4 (2.97 g, 9.28 mmol) was added 12.68 mL (38.05 mmol) of a 3.0
M MeMgBr ether solution. During the addition of MeMgBr gas
evolution was observed. Within minutes the reaction mixture changed
color from yellow to black. The solution was stirred for 1 h, after
which time the solvent was removed under reduced pressure. To the
residue was added 50 mL of hexane, and the resulting yellow solution
was filtered. The salts were washed with an additional 30 mL of
hexane. The solvent was removed under reduced pressure, leaving 4.03
g of yellow crystalline solid. Yield: 79.1%.
Preparation of (N-((6E)-6-(Butylimino-κN)-1-cyclohexen-1-yl)-2,6-
bis(1-methylethyl)benzenaminato-κN)triiodohafnium (10). To
1.5624 g (2.01 mmol) of (N-((6E)-6-(butylimino-κN)-1-cyclohexen-
Preparation of (N-((6E)-6-(Butylimino-κN)-1-cyclohexen-1-yl)-2,6-
bis(1-methylethyl)benzenaminato-κN)trimethylzirconium (12). To
2969
dx.doi.org/10.1021/om400189w | Organometallics 2013, 32, 2963−2972

Organometallics
Article
(C7), 31.60 (C5), 30.51 (C8), 28.97 (CH(CH3)2), 25.33 (C3), 24.38
(CH(CH3)2), 23.64 (CH(CH3)2), 23.61 (C4), 21.43 (C9), 14.25
(C10). Anal. Calcd for C25H42HfN2: C, 54.68; H, 7.71; N, 5,10.
Found: C, 54.66; H, 7.69; N, 5.07.
a toluene suspension (12 mL) containing 0.3901 g (1.19 mmol) of N-
((6E)-6-(butylimino)-1-cyclohexen-1-yl)-2,6-bis(1-methylethyl)-
benzenamine (8) and 0.266 g (1.19 mmol) of ZrCl4 was added 1.67
mL (5.02 mmol) of 3 M MeMgBr ether solution at ambient
temperature. During the addition the reaction mixture changed color
from yellow to black. After stirring for 2 h, 45 mL of hexane was
added, and the reaction mixture was filtered. The solvent was removed
from the yellow filtrate under reduced pressure to give 0.5305 g of
Preparation of [N-[2-(Butylamino-κN)-2-cyclohexen-1-ylidene]-
2,6-bis(1-methylethyl)benzenaminato-κN]trimethylzirconium (14).
To a toluene solution (40 mL) of 1.2833 (3.93 mmol) of N-[2-
(butylamino)-2-cyclohexen-1-ylidene]-2,6-bis(1-methylethyl)-
benzenamine (9)7 and 0.916 g (3.93 mmol) of ZrCl4 was added 5.4
mL (16.11 mmol) of 3 M MeMgBr ether solution at ambient
temperature. During addition of the MeMgBr gas evolution was
observed. Within minutes the reaction mixture changed color from
yellow to black. The solution was stirred for 1 h. The solvent was
removed under reduced pressure. To the residue was added 10 mL of
toluene followed by 50 mL of hexane, and the yellow solution was
filtered. The solvent was removed under reduced pressure to give a
dark orange crystalline solid. The solid was dissolved in 2 mL of
toluene and 10 mL of hexane. The solution was filtered and put into a
freezer overnight. The solvent was decanted, and the large orange
crystals that had formed were washed with 5 mL of cold hexane and
dried under reduced pressure to give 0.693 g of product. NMR
1
yellow, highly crystalline product. Yield: 96.1%. H NMR (C6D6, 500
i
MHz, 30 °C): δ 7.21 (s, 3H, Pr2-Ph), 4.70 (t, J = 5.0 Hz, 1H, H5),
3.48 (hept, 2H, J = 6.9 Hz,CH(CH3)2), 3.36−3.29 (m, 2H, H7), 1.96
(dd, 2H, J = 7.1, 5.8 Hz, H2), 1.85 (q, 2H, J = 5.7 Hz, H4), 1.40 (d,
6H, J = 6.9 Hz, CH(CH3)2), 1.38−1.27 (m, 4H, H3/H8), 1.22 (d, 6H,
J = 6.9 Hz, 6H, CH(CH3)2), 1.08 (sex, 2H, J = 7.4 Hz, H9), 0.77 (t,
3H, J = 7.4 Hz, H10), 0.74 (s, 9H, Zr-CH3). 13C NMR (C6D6, 126
MHz, 30 °C): δ 174.54 (N=C), 151.62 (quat.), 144.89 (quat.), 144.18
(quat.), 126.13 (iPr2-para-Ph), 124.23 (iPr2-meta-Ph), 113.42 (C5),
49.91 (C7), 48.60 (Zr-CH3), 31.21 (C8), 28.57 (CH(CH3)2), 27.79
(C2), 26.02 (CH(CH3)2), 24.77 (C4), 24.56 (CH(CH3)2), 23.18
(C3), 20.99 (C9), 13.84 (C10). Anal. Calcd for C25H42ZrN2: C, 65.02;
H, 9.17; N, 6.07. Found: C, 64.92; H, 9.08; N, 6.07.
1
spectroscopy showed formation of clean product. Yield: 38.6%. H
NMR (500 MHz, C6D6): δ 7.10−6.99 (m, 3H, iPr2Ph), 5.21 (t, 1H, J =
5.0 Hz, H2), 3.76 (m, 2H, H7), 2.57 (hept, 2H, J = 6.8 Hz,
CH(CH3)2), 2.12 (q, 2H, J = 5.7 Hz, H3), 2.06 (m, 2H, H5), 1.78 (m,
2H, H8), 1.42 − 1.29 (m, 4H, H4 and H9), 1.17 (d, 6H, J = 6.8 Hz,
CH(CH3)2), 0.92 (d, 6H, J = 6.8 Hz, CH(CH3)2), 0.91 (t, 2H, J = 7.4
Hz, H10), 0.71 (s, 9H, Zr(CH3)3). 13C NMR (75 MHz, C6D6): δ
181.71 (CN), 148.31 (quat.), 145.59 (quat.), 138.98 (quat.), 126.61
(iPr2Ph), 124.04 (iPr2Ph), 109.95 (C2), 47.00 (Zr(CH3)3), 45.77
(C7), 31.51 (C5), 30.71 (C8), 29.21 (CH(CH3)2), 25.48 (C3), 24.48
(CH(CH3)2), 23.75 (CH(CH3)2), 23.68 (C4), 21.58 (C9), 14.42
(C10). Anal. Calcd for C25H42ZrN2: C, 65.02; H, 9.17; N, 6.07.
Found: C, 64.98; H, 9.06; N, 6.01.
Preparation of [N-[2-(Butylamino-κN)-2-cyclohexen-1-ylidene]-
2,6-bis(1-methylethyl)benzenaminato-κN]trimethylhafnium (13).
To a toluene suspension (40 mL) of 1.0168 g (3.11 mmol) of N-
[2-(butylamino)-2-cyclohexen-1-ylidene]-2,6-bis(1-methylethyl)-
benzenamine (9)7 and 0.997 g (3.11 mmol) of HfCl4 was added 4.25
mL (12.77 mmol) of 3 M MeMgBr ether solution. During the addition
of the MeMgBr gas evolution was observed. After stirring for 2 h at
ambient temperature, the solvent was removed under reduced
pressure. To the residue was added 20 mL of toluene followed by
20 mL of hexane, and the yellow solution was filtered. The solvent was
removed under reduced pressure, leaving a yellow crystalline solid.
NMR spectroscopy showed formation of the desired product along
with a small amount of impurities. The solid was dissolved in 2 mL of
warm toluene followed by addition of 18 mL of hexane. The solution
was filtered and put into a freezer (at −26 °C) overnight. Precipitated
yellow solid was collected on a frit, washed with 3 mL of cold hexane,
and dried under reduced pressure to give 433 mg of product. The
solvent was removed from the filtrate, and the residue was dissolved in
15 mL of hexane and put into a freezer. After 2 days the solvent was
decanted, and the large crystals that had formed were washed with 2
mL of cold hexane and dried under reduced pressure to give 333 mg of
product. Combined yield: 766 mg (44.8%). 1H NMR (500 MHz,
C6D6): δ 7.08−7.00 (m, 3H, iPr2Ph), 5.19 (t, 1H, J = 5.0 Hz, H2), 3.67
(m, 2H, H7), 2.59 (hept, 2H, J = 6.8 Hz, CH(CH3)2), 2.18 (q, 2H, J =
5.7 Hz, H3), 2.01 (dd, 2H, J = 7.1, 5.7 Hz, H5), 1.76 (m, 2H, H8),
1.39−1.28 (m, 4H, H4 and H9), 1.19 (d, 6H, J = 6.8 Hz, CH(CH3)2),
0.92 (t, 3H, J = 7.4 Hz, H10), 0.91 (d, 6H, J = 6.8 Hz, CH(CH3)2),
0.48 (d, 9H, J = 0.8 Hz, Hf(CH3)3). 13C NMR (126 MHz, C6D6): δ
183.04 (CN), 149.37 (quat.), 145.30 (quat.), 139.37 (quat.), 126.83
(iPr2Ph), 123.99 (iPr2Ph), 111.32 (C2), 60.12 (Hf(CH3)3), 44.72
Scale-up of 2-(4-Morpholinyl)-2-cyclohexen-1-one (5). A 3 L, 3-
necked round-bottomed flask equipped with a Dean−Stark trap and a
condenser was placed under nitrogen, then charged with toluene (1.60
L) and sparged with nitrogen for 30 min. Morpholine (145.8 mL,
1.672 mol) and 1,2-cyclohexanedione (150.0 g, 1.338 mol) were then
added under a stream of nitrogen (caution: care must be taken to avoid
prolonged exposure of 1,2-cyclohexanedione to air, as the white crystalline
material quickly becomes yellow and partially liquefies in air). The
mixture was then heated under reflux for 3 h, with water collecting in
the Dean−Stark trap during this time. The reaction mixture was
cooled to room temperature, and a dark oily residue (unwanted
byproduct) separated from the solution and settled out on the bottom
of the flask. The solution was isolated from the immiscible oil by
decantation, and the solvent was removed under vacuum. The
resulting solid residue was ground up with a spatula and then triturated
with pentane, furnishing orange-brown crystalline material, which was
1
1
shown by H NMR to be of high purity (230.0 g, 95%). H NMR
3
(C6D6, 300 MHz, 30 °C): δ 5.35 (t, 1H, J = 4.6 Hz), 3.67−3.64 (m,
4H), 2.70−2.66 (m, 4H), 2.18−2.13 (m, 2H, H2), 1.87−1.81 (m, 2H,
H4), 1.42 (quint., 2H, J = 6.4 Hz, H3) ppm. 13C NMR (C6D6, 75
3
MHz, 30 °C): δ 194.30, 147.03, 124.72, 67.32, 50.81, 40.49, 25.89,
23.63 ppm. GC/MS (CI) mass spectrum: m/z 182 (M + H).
Scale-up of 2-((2,6-Bis(1-methylethyl)phenyl)amino)-2-cyclohex-
en-1-one (6). A 3 L, three-necked round-bottomed flask equipped
with a mechanical stirrer was placed under a nitrogen atmosphere,
then charged with toluene (1.75 L), which was sparged with nitrogen
2970
dx.doi.org/10.1021/om400189w | Organometallics 2013, 32, 2963−2972

Organometallics
Article
for 30 min. Molecular sieves (100 g, 4 Å) and 2-(4-morpholinyl)-2-
cyclohexen-1-one (5) (175.0 g, 0.9656 mol), followed by 2,6-
diisopropylaniline (196.1 mL, 0.9656 mol), were added under a
stream of nitrogen, and the resulting mixture was heated to 70 °C. At
this point, p-toluenesulfonic acid (183.7 g, 0.966 mol) was added,
producing a light brown cloudy solution. The temperature was
increased to 80 °C, and the solution was mechanically stirred for 1 h,
at which point the solution became clear (an olive green color), and
GC/MS analysis indicated complete conversion. The mixture was
allowed to cool, then concentrated to dryness under vacuum. The
thick residue was broken up with a large spatula, triturated with
pentane to remove excess toluene, extracted in hot pentane, and
filtered through a medium-porosity glass frit. The filtrate was reduced
under vacuum and recrystallized at −20 °C. The white crystals (130.0
g) were harvested, the mother liquor was reduced in volume, and a
second recrystallization was performed at −20 °C, furnishing an
mixed and transferred to a catalyst addition tank and injected into the
reactor. The polymerization conditions were maintained for 10 min
with ethylene added on demand. Heat was continuously removed from
the reaction vessel through an internal cooling coil. The resulting
solution was removed from the reactor, quenched with isopropyl
alcohol, and stabilized by addition of 10 mL of a toluene solution
containing approximately 67 mg of a hindered phenol antioxidant
(Irganox 1010 from Ciba Geigy Corporation) and 133 mg of a
phosphorus stabilizer (Irgafos 168 from Ciba Geigy Corporation).
Between polymerization runs, a wash cycle was conducted in which
850 g of mixed alkanes was added to the reactor and the reactor was
heated to 150 °C. The reactor was then emptied of the heated solvent
immediately before beginning a new polymerization run. Polymers
were recovered by drying for about 12 h in a temperature-ramped
vacuum oven with a final set point of 140 °C.
Polymer Characterization. Melting (Tm) and glass transition (Tg)
temperatures of polymers were measured by differential scanning
calorimetry (Q2000 DSC, TA Instruments, Inc.). Samples were first
heated from room temperature to 200 °C using the ‘Jump To’ feature.
After being held at this temperature for 4 min, the samples were
cooled to −90 °C at 10 °C/min, held for 4 min, and then heated again
to 200 °C. Molecular weight distribution (Mw, Mn) information was
determined by analysis on a custom Dow-built robotic-assisted
dilution high-temperature gel permeation chromatographer (RAD-
GPC). Polymer samples were dissolved for 90 min at 160 °C at a
concentration of 5−7 mg/mL in 1,2,4-trichlorobenzene (TCB)
stabilized by 300 ppm of BHT in capped vials while stirring. They
were then diluted to 1 mg/mL immediately before a 400 μL aliquot of
the sample was injected. The GPC utilized two Polymer Laboratories
PLgel 10 μm MIXED-B columns (300 mm × 10 mm) at a flow rate of
2.0 mL/min at 150 °C. Sample detection was performed using a
PolyChar IR4 detector in concentration mode. A conventional
calibration of narrow polystyrene (PS) standards was utilized, with
apparent units adjusted to homopolyethylene (PE) using known
Mark−Houwink coefficients for PS and PE in TCB at this
temperature. To determine octene incorporation, 30 mg/mL solutions
of each sample were dissolved at 160 °C for 120 min while covered
and shaking with glass stir-bars in vials. A 90 μL amount of each
polymer solution was deposited onto a silicon wafer, heated under a
nitrogen purge at 145 °C until the trichlorobenzene had evaporated,
and analyzed using a Nicolet Nexus 670 FTIR running Omnic (v. 7.1)
software equipped with an AutoPro automated sampler.
Structure Determinations of 10−14. X-ray intensity data were
collected on a Bruker SMART diffractometer using Mo Kα radiation
(λ = 0.71073 Å) and an APEXII CCD area detector. Raw data frames
were read by the program SAINT20 and integrated using 3D profiling
algorithms. The resulting data were reduced to produce hkl reflections
and their intensities and estimated standard deviations. The data were
corrected for Lorentz and polarization effects, and multiscan
absorption corrections were applied. The structure was solved and
refined in SHELXTL6.1, using full-matrix least-squares refinement.
The non-H atoms were refined with anisotropic thermal parameters,
and H atoms were calculated in idealized positions and refined riding
on their parent atoms. The refinement was carried out using F2 rather
than F values. R1 is calculated to provide a reference to the
conventional R value, but its function is not minimized.
Structure 10. C22H33HfI3N2, Mw = 884.69, orthorhombic, Pbca
(0.38 × 0.15 × 0.10 mm3), a = 14.8620(4) Å, b = 16.5798(5) Å, c =
21.8610(6) Å, β = 94.2550(10)°, temp = 100(2) K, Z = 4, V =
5386.7(3) Å3, R1 = 0.0163, 0.0202, wR2 = 0.0326, 0.0336 (I > 2σ(I),
all data), GOF = 1.068.
Structure 11. C25H42HfN2, Mw = 549.10, monoclinic, P2(1)/c (0.22
× 0.22 × 0.21 mm3), a = 10.8416(2) Å, b = 17.4171(3) Å, c =
13.5117(3) Å, β = 94.2550(10)°, temp = 200(2) K, Z = 4, V =
2544.37(9) Å3, R1 = 0.0155, 0.0183, wR2 = 0.0381, 0.0399 (I > 2σ(I),
all data), GOF = 1.035.
1
additional 55.3 g of product. The net yield was 185.3 g (71%). H
NMR (C6D6, 300 MHz): δ 7.17−7.06 (m, 3H), 6.00 (br s, 1H), 4.95
(t, 1H, J = 4.7 Hz), 3.15 (septet, 2H, J = 6.9 Hz), 2.21−2.16 (m, 2H,
H2), 1.75 (q, 2H, J = 5.5 Hz, H4), 1.41 (quint., 2H, J = 6.3 Hz, H3),
1.10 (d, 12H, J = 6.9 Hz) ppm. 13C NMR (C6D6, 75 MHz): δ 194.09,
147.17, 140.82, 135.90, 127.47, 124.04, 112.08, 38.15, 28.91, 24.53,
24.36, 24.04 ppm. GC/MS (CI) mass spectrum: m/z 272 (M + H).
HRMS (ESI, (M + Na)+): (m/z) calcd for C18H25NONa 294.180,
found 294.183.
Scale-up of N-((6E)-6-(Butylimino)-1-cyclohexen-1-yl)-2,6-bis(1-
methylethyl)benzenamine (8). To a stirred solution of n-butylamine
(303.7 mL, 3.074 mol) in toluene (1.50 L) in a 3 L round-bottomed
flask in a nitrogen drybox was added TiCl4 (33.7 mL, 0.307 mol)
dropwise via addition funnel over 15 min (temperature rise from 28
°C up to 60 °C was observed due to the exothermic reaction). The
resulting solution was allowed to stir for an additional 1 h, at which
point the temperature decreased to 30 °C, and 2-((2,6-bis(1-
methylethyl)phenyl)amino)-2-cyclohexen-1-one (6) (150.1 g, 0.5532
mol) was added over ∼2 min (causing a temperature rise up to 38 °C).
The resulting brown mixture was stirred for 3 h and then filtered
through a pad of Celite in a medium-porosity glass frit into a flask
containing anhydrous K2CO3 (75 g) and stirred for 1 h. The mixture
was refiltered through a fine-porosity glass fiber filter and dried under
1
vacuum, furnishing the product as a yellow oil (166.4 g, 92%). H
NMR analysis revealed <2% of isomer 9.
Scale-up of (N-((6E)-6-(Butylimino-κN)-1-cyclohexen-1-yl)-2,6-
bis(1-methylethyl)benzenaminato-κN)trimethylhafnium (11). To a
stirred solution of N-((6E)-6-(butylimino)-1-cyclohexen-1-yl)-2,6-
bis(1-methylethyl)benzenamine (8) (165.0 g, 0.5053 mmol) in
toluene (1.60 L), in a 3 L round-bottomed flask in a nitrogen drybox,
was added HfCl4 (161.9 g, 0.5053 mol) in one portion. To the
resulting tan-colored turbid solution was added MeMgBr (690.6 mL, 3
M in ether, 2.072 mol) dropwise via addition funnel over a period of
∼1 h, to maintain the temperature below 40 °C. The reaction mixture
was stirred for an additional 1 h, at which point a dark black color had
developed. The solvent was removed under vacuum. The dried residue
was extracted in several portions with Isopar-E and filtered through a
pad of Celite in a medium-porosity glass frit. The total extraction
volume was 4.4 L. A portion of the solution was analyzed by 1H NMR
to determine the purity (with <2% reverse isomer) and gravimetrically
analyzed as 8.1 wt % 11 (0.106 M, net yield is calculated from this to
be 92%).
Ethylene-1-octene Polymerization Procedures and Polymer
Characterization. Ethylene-1-octene Copolymerization. A 2 L Parr
reactor was used in the polymerizations. All feeds were passed through
columns of alumina and Q-5 catalyst (available from Engelhard
Chemicals Inc.) prior to introduction into the reactor. Precatalyst and
cocatalyst (activator) solutions were handled in the glovebox. A stirred
2 L reactor was charged with about 533 g of mixed alkanes solvent and
250 g of 1-octene comonomer. The reactor contents were heated to
the polymerization temperature of 120 °C, while 10 μmol of MMAO
was added to the reactor as a scavenger for trace O2 and water. Once at
temperature, the reactor was saturated with ethylene at 460 psig (3.4
MPa). Catalysts and cocatalysts, as dilute solutions in toluene, were
Structure 12. C25H42N2Zr, Mw = 461.83, monoclinic, P2(1)/c (0.38
× 0.38 × 0.36 mm3), a = 10.8529(4) Å, b = 17.0807(7) Å, c =
13.7356(5) Å, β = 95.355(2)°, temp = 200(2) K, Z = 4, V =
2971
dx.doi.org/10.1021/om400189w | Organometallics 2013, 32, 2963−2972

Organometallics
Article
2535.13(17) Å3, R1 = 0.0259, 0.0288, wR2 = 0.0712, 0.0746 (I >
2σ(I), all data), GOF = 1.054.
Structure 13. C25H42HfN2, Mw = 549.10, monoclinic, P2(1)/c (0.48
× 0.42 × 0.35 mm3), a = 13.5113(4) Å, b = 10.1690(3) Å, c =
19.3016(6) Å, β = 110.2520(10)°, temp =100(2) K, Z = 4, V =
2488.02(13) Å3, R1 = 0.0133, 0.0144, wR2 = 0.0309, 0.0312 (I >
2σ(I), all data), GOF = 1.167.
Structure 14. C25H42N2Zr, Mw = 461.83, monoclinic, P2(1)/c (0.45
× 0.35 × 0.30 mm3), a = 13.5205(6) Å, b = 10.1652(5) Å, c =
19.3393(11) Å, β = 110.388(2)°, temp = 100(2) K, Z = 4, V =
2491.5(2) Å3, R1 = 0.0219, 0.0234, wR2 = 0.0598, 0.0612 (I > 2σ(I),
all data), GOF = 1.087.
(7) Figueroa, R.; Froese, R. D.; He, Y.; Klosin, J.; Theriault, C. N.;
Abboud, K. A. Organometallics 2011, 30, 1695−1709.
(8) Carlson, R.; Nilsson, A. Acta Chem. Scand. 1984, B38, 49−53.
(9) Carlson, R.; Larsson, U.; Hansson, L. Acta Chem. Scand. 1992, 46,
1211−1214.
(10) Primary amines other than n-BuNH2 can be used to prepare
different imino−enamines using this method. Figueroa, R.; Klosin, J.
Int. Pat. Appl. 2010022228, 2010.
(11) It is not necessarily the case that tribenzyl and trimethyl
derivatives would have the same polymerization characteristics. This
depends on details of the polymerization mechanism for these
complexes, which is currently unknown. One important question is
whether there are two polymer chains growing at the same cationic
metal center or if one of the alkyl groups (Bn or Me) is a spectator
ligand. If the former scenario is operating, then one could expect very
similar polymerization behavior for both tribenzyl and trimethyl
precatalysts.
ASSOCIATED CONTENT
■
S
* Supporting Information
Molecular structures of 13 and 14. 1D and 2D NMR spectra for
all new compounds. This material is available free of charge via
the Internet at http://pubs.acs.org.
(12) See Supporting Information for 1D and 2D spectra for all new
compounds.
(13) Crystals crack below 200 K.
(14) Balboni, D.; Camurati, I.; Prini, G.; Resconi, L.; Galli, S.;
Mercandelli, P.; Sironi, A. Inorg. Chem. 2001, 40, 6588−6597.
(15) Resconi, L.; Camurati, I.; Grandini, C.; Rinaldi, M.; Mascellani,
N.; Traverso, O. J. Organomet. Chem. 2002, 664, 5−26.
(16) We were unable to isolate this side product in pure form and
characterize it fully. NMR spectroscopy of this byproduct clearly
indicates it has C1 symmetry (e.g., four doublets and two heptets are
observed for the i-Pr groups together with two distinct resonances for
the methylene protons α to nitrogen atom).
(17) Molecular structures of 12 and 14 are isostructural with 11 and
13, respectively. See Supporting Information for details.
(18) For structures 2 and 3, these values refer to bond lengths and
angles between Hf and the methylene carbons of the benzyl groups.
(19) GPC data were obtained on a conventional GPC calibrated
using narrow polystyrene standards. This GPC measures backbone
molecular weight rather than absolute molecular weight of polymers.
Previously published GPC data for 2 and 37 were obtained on a GPC
utilizing a light-scattering detector that measures the absolute
molecular weights. This is the reason that Mw data for 2 and 3
shown in Table 2 are somewhat lower than what was reported
previously.
AUTHOR INFORMATION
■
Corresponding Author
*E-mail: jklosin@dow.com.
Notes
The authors declare no competing financial interest.
REFERENCES
■
(1) (a) Britovsek, G. J. P.; Gibson, V. C.; Wass, D. F. Angew. Chem.,
Int. Ed. 1999, 38, 428−447. (b) Gibson, V. C.; Spitzmesser, S. K.
Chem. Rev. 2003, 103, 283−316. (c) Makio, H.; Kashiwa, N.; Fujita, T.
Adv. Synth. Catal. 2002, 344, 477−493. (d) Park, S.; Han, Y.; Kim, S.
K.; Lee, J.; Kim, H. K.; Do, Y. J. Organomet. Chem. 2004, 689, 4263−
4276. (e) Braunschweig, H.; Breitling, F. M. Coord. Chem. Rev. 2006,
250, 2691−2720. (f) Makio, H.; Terao, H.; Iwashita, A.; Fujita, T.
Chem. Rev. 2011, 111, 2363−2449.
(2) (a) Ittel, S. D.; Johnson, L. K.; Brookhart, M. Chem. Rev. 2000,
100, 1169−1203. (b) Tian, J.; Hustad, P. D.; Coates, G. W. J. Am.
Chem. Soc. 2001, 123, 5134−5135. (c) Bollmann, A.; Blann, K.; Dixon,
J. T.; Hess, F. M.; Killian, E.; Maumela, H.; McGuinness, D. S.;
Morgan, D. H.; Neveling, A.; Otto, S.; Overett, M.; Slawin, A. M. Z.;
Wasserscheid, P.; Kuhlmann, S. J. Am. Chem. Soc. 2004, 126, 14712−
14713. (d) Boussie, T. R.; Diamond, G. M.; Goh, C.; Hall, K. A.;
LaPointe, A. M.; Leclerc, M. K.; Murphy, V.; Shoemaker, J. A. W.;
Turner, H.; Rosen, R. K.; Stevens, J. C.; Alfano, F.; Busico, V.; Cipullo,
R.; Talarico, G. Angew. Chem., Int. Ed. 2006, 45, 3278−3283.
(e) Arriola, D. J.; Carnahan, E. M.; Hustad, P. D.; Kuhlman, R. L.;
Wenzel, T. T. Science 2006, 312, 714−719. (f) Valente, A.; Mortreux,
A.; Visseaux, M.; Zinck, P. Chem. Rev. 2013, ASAP.
(20) Sheldrick, G. M. SHELXTL6.1, Crystallographic Software
Package; Bruker AXS, Inc.: Madison, WI, USA, 2008.
(3) (a) Klosin, J.; Kruper, W. J.; Nickias, P. N.; Roof, G. R.; De
Waele, P.; Abboud, K. A. Organometallics 2001, 20, 2663−2665.
(b) Konze, W. V.; Vanderlende, D. D. Int. Pat. Appl. 2007136497,
2007. (c) Nabika, M.; Katayama, H.; Watanabe, T.; Kawamura-
Kuribayashi, H.; Yanagi, K.; Imai, A. Organometallics 2009, 28, 3785−
3792.
(4) (a) De Waele, P.; Jazdzewski, B. A.; Klosin, J.; Murray, R. E.;
Petersen, J. L.; Theriault, C. N.; Vosejpka, P. C. Organometallics 2007,
26, 3896−3899. (b) Kuhlman, R. L.; Klosin, J. Macromolecules 2010,
43, 7903−7904. (c) Froese, R. D. J.; Jazdzewski, B. A.; Klosin, J.;
Kuhlman, R. L.; Theriault, C. N.; Welsh, D. M.; Abboud, K. A.
Organometallics 2011, 30, 251−262.
(5) (a) Murray, R. E. U.S. Pat. 6,096,676, 2000. (b) Murray, R. E.;
George, V. M.; Nowlin, D. L.; Schultz, C. C.; Petersen, J. L. Polym.
Prepr. (Am. Chem. Soc., Div. Polym. Chem.) 2002, 43, 294−295.
(c) Murray, R. E. Int. Pat. Appl. 2003051935, 2003.
(6) (a) Mashima, K.; Ohnishi, R.; Yamagata, T.; Tsurugi, H. Chem.
Lett. 2007, 36, 1420−1421. (b) Tsurugi, H.; Ohnishi, R.; Kaneko, H.;
Panda, T. K.; Mashima, K. Organometallics 2009, 28, 680−687.
2972
dx.doi.org/10.1021/om400189w | Organometallics 2013, 32, 2963−2972