Iron-catalyzed chain growth of ethylene: In situ regeneration of ZnEt2 by tandem catalysis

Article abstract of DOI:10.1021/acscatal.5b01231

A dual catalyst system to implement in situ regeneration of ZnEt₂, the Chain Transfer Agent (CTA) in Catalyzed Chain Growth of ethylene (CCG), has been demonstrated. As in typical CCG systems, an Fe homogeneous catalyst is used to grow oligomeric chains that transfer rapidly to ZnEt₂. However, rather than liberating alkane products and destroying the expensive chain transfer agent via acid hydrolysis workup, a second Fe-alkyl catalyst, (BiPy)₂FeEt₂, has been introduced to regenerate ZnEt₂ via ethyl/alkyl exchange and liberate the oligomer chains as α-olefins via ss-hydride elimination. This improvement reduces the ZnEt₂ loading and leaves the chain growth catalyst competent, in contrast to Ni(acac)₂ shown to be unsuitable for in situ tandem catalysis. These findings greatly enhance the industrial viability of the chemistry.

Full text of DOI:10.1021/acscatal.5b01231

Letter
pubs.acs.org/acscatalysis
Iron-Catalyzed Chain Growth of Ethylene: In Situ Regeneration of
ZnEt2 by Tandem Catalysis
Renan Cariou*,† and John W. Shabaker‡
†BP Chemicals, Saltend, Hull HU12 8DS, United Kingdom
‡BP Products North America, 150 West Warrenville Road, Naperville, Illinois 60563, United States
S
* Supporting Information
ABSTRACT: A dual catalyst system to implement in situ regeneration of
ZnEt2, the Chain Transfer Agent (CTA) in Catalyzed Chain Growth of
ethylene (CCG), has been demonstrated. As in typical CCG systems, an Fe
homogeneous catalyst is used to grow oligomeric chains that transfer
rapidly to ZnEt2. However, rather than liberating alkane products and
destroying the expensive chain transfer agent via acid hydrolysis workup, a
second Fe-alkyl catalyst, (BiPy)2FeEt2, has been introduced to regenerate
ZnEt2 via ethyl/alkyl exchange and liberate the oligomer chains as α-olefins
via ß-hydride elimination. This improvement reduces the ZnEt2 loading and leaves the chain growth catalyst competent, in
contrast to Ni(acac)2 shown to be unsuitable for in situ tandem catalysis. These findings greatly enhance the industrial viability of
the chemistry.
KEYWORDS: catalyzed chain growth, chain transfer, α-olefins, ethylene polymerization, tandem catalysis
emarkable control over polyolefins’ microstructures,
molecular weights, and polydispersity has been achieved
deactivation of the chain growth catalyst, 1/MAO, prior to
chain displacement and liberation of α-olefins.3,4
R
due to the development of single-site polymerization catalysts.1
Catalyzed chain growth (CCG) of ethylene and/or α-olefins
offers such control, leading to polyolefin materials described by
Sita as precision hydrocarbons (PHCs) with a broad range of
potential commercial applications such as the production of
lubricant base oils.2a Such a reaction proceeds via chain growth
on main group alkyls (MgR2, ZnR2, AlR3) catalyzed by
lanthanide or transition metal complexes.2 The bis(imino)-
pyridine iron-(II) dichloride complex 1, in combination with
methylaluminoxane (MAO) as cocatalyst, was reported by
Gibson to be highly active in the CCG reaction of ethylene
with ZnEt2 as chain transfer agent (CTA) (Scheme 1).3 The
Furthermore, the use of stoichiometric amounts of expensive
ZnEt2 renders a potential process prohibitively expensive. An
alternative to using smaller amounts of the CTA has been
shown by Sita, who reported the synthesis of PHCs using a
ternary living coordinative chain polymerization based on the
combination of ZnEt2 and AliBu3 CTAs catalyzed by a hafnium
catalyst. Chains are transferred rapidly and reversibly between
Zn and Al centers allowing for the use of ZnEt2 in amounts as
low as 2 equiv/catalyst, although a large stoichiometric excess
of aluminum alkyls is required.2a,5 Another significant develop-
ment in controlling the microstructure of polyolefins is the
report of chain shuttling between two group IV olefin
polymerization catalysts by Dow scientists, who leveraged
chain transfer chemistry for the production of ethylene/α-olefin
block copolymers (OBCs).6 This process involves ZnEt2 as the
Chain Shuttling Agent (CSA) and is used for ethylene/1-
octene copolymerization to afford block copolymers.
Scheme 1. Catalyzed Chain Growth of Ethylene on Zinc
In order to circumvent the challenges of the Fe-catalyzed
chain growth of ethylene, we envisioned the use of a dual
catalyst system able to regenerate ZnEt2 in situ without
deactivation of the chain growth catalyst 1, akin to Dow’s chain
shuttling process. However, our approach differs from that of
Dow in the use of a catalyst effective in ß-hydride elimination
and inactive in olefin polymerization. The one pot tandem
catalytic cycle involves catalyzed chain growth (chain growth/
chain transfer cycle) followed by an olefin production cycle
(Scheme 2). As for CCG, the olefin production cycle relies on
reaction afforded a Poisson distribution of Zn(oligomer)2
products, which after hydrolysis, yielded linear alkanes centered
at C50. Alternatively, a Poisson distribution of linear α-olefins
was produced via a nickel-catalyzed chain displacement reaction
from Zn(oligomer)2. Despite great potential, this process
comes with several hurdles to overcome. Importantly, the
tandem chain growth/displacement experiment requires
Received: June 12, 2015
© XXXX American Chemical Society
4363
DOI: 10.1021/acscatal.5b01231
ACS Catal. 2015, 5, 4363−4367

ACS Catalysis
Letter
Scheme 2. One-Pot Tandem Catalytic Cycle
efficient chain transfer between Zn(oligomer)2 and a M-Et
species to give ZnEt2 and M-oligomer, which readily releases
the chain as an α-olefin via ß-H elimination and generates a M-
H intermediate. Ethylene insertion into the M-H bond reforms
the M-Et species, thus closing the cycle.
Previous reports have shown that efficient chain transfer
between Fe and Zn species is governed by steric hindrance at
the Fe center, as well as matching Zn−C and Fe−C bond
strength.4,7 Hence, two Fe-based organometallic complexes,
NacNacFe-Et 3,8 and (Bipy)2FeEt2 4,9 were selected for
investigation as “transfer” catalysts. Before attempting catalysis,
we investigated whether efficient chain transfer would occur
between a ZnR2 species and the selected “transfer” catalysts 3
or 4. Accordingly, the complexes were tested in a chain
displacement experiment with dihexylzinc under an ethylene
atmosphere (Scheme 3) and monitored by 1H NMR
spectroscopy.
Figure 1. 1H NMR spectra (C6D6, 300 MHz) of dihexylzinc (top) and
in the presence of 5 mol % of 4 (bottom).
5.76 ppm consistent with the formation of 1-hexene. In
addition, a new broad peak was observed at −0.61 ppm.10
Equilibrium conversion was reached within 42 h under these
conditions and indicated that chain transfer of hexyl and ethyl
chains occurred between the Zn and Fe centers, and implying
that ZnEt2 was regenerated (Figure 2).
Scheme 3. Dihexylzinc Chain Displacement Probe Reaction
Using 3 and 4 as Catalysts
Figure 2. Conversion vs time plot of chain displacement of dihexylzinc
catalyzed by 5 mol % of 4 under ethylene atmosphere.
Upon mixing dihexylzinc with 5 mol % of the paramagnetic
complex 3 in C6D6, no change in the H NMR spectrum was
1
On the basis of these observations, we decided to test 4 in
catalysis by adding it to the typical Fe-catalyzed chain growth
system. First, a typical chain growth experiment using, 1/
MAO/ZnEt2, was carried out according to the established
procedure reported by Gibson.3 After 30 min, this procedure
afforded a Poisson distribution of alkanes centered at C46 upon
workup with acidified methanol (Table 1, Entry 1). Note that
alkanes are generated upon acidic workup of the Zn-alkyls
produced during the reaction. Upon repeating the reaction in
the presence of 0.5 mol % of 4 relative to ZnEt2, only liquids
were formed, and no solids precipitated upon acidic methanol
workup at the end of the reaction (Table 1, Entry 2). This
result was in stark contrast to the benchmark experiment
described above. The reaction was monitored over the period
of 30 min by sampling at regular time intervals, with aliquots
quenched in aqueous HCl and filtered through alumina prior to
GC analysis while using nonane as internal standard.
observed apart from broadening of the peaks. Upon exposure
to ethylene, no reaction was observed after 2 h at room
temperature. After heating the sample at 60 °C for 13 h,
ethylene was consumed, and new broad peaks in the olefinic
region (4−6 ppm) were observed.10 Note that the starting
catalyst 3 was still present in solution, indicating no apparent
decomposition during the test experiment. Volatiles were
vacuum transferred to another NMR tube and 1H NMR
spectroscopy indicated the presence of 1-hexene, which was
also confirmed by GC analysis.10 This result indicated that
chain transfer occurred between the Fe and Zn species;
however, elevated temperature was necessary to afford the α-
olefin product most likely due to increased stability of the Fe-
alkyl species. Indeed, NacNac ligands have been shown to
stabilize highly reactive species.11 Complex 4 is conveniently
1
diamagnetic and allowed the reaction to be monitored by H
NMR spectroscopy. Upon mixing dihexylzinc with 5 mol % of
4 in C6D6, an upfield shift from 0.28 to −0.28 ppm was
observed for the methylene protons on the carbon α to the Zn
center (Figure 1). Exposing the solution to an ethylene
atmosphere at room temperature resulted in ethylene
consumption and appearance of olefinic protons at 4.99 and
GC analyses showed the presence of both alkanes and α-
olefins and that over time the concentration of α-olefins
increased at the expense of the concentration of alkanes (Figure
3a), indicating that tandem catalysis occurred. After 30 min, the
product distribution consisted of 77 mol % α-olefins with a
distribution centered at C8. The aim of regenerating and
4364
DOI: 10.1021/acscatal.5b01231
ACS Catal. 2015, 5, 4363−4367

ACS Catalysis
Letter
Table 1. Tandem Catalysis Experiments
toluene soluble fraction
yield (center of
toluene insoluble fraction
chain growth
catalyst
transfer
catalyst
ZnEt2
d
activity
(g/mmol/h)
α-olefin
(mol %)
yield (center of
α-olefin
e,f
e
e
e,
f
entry
(eq)
distribution )
distribution
)
(mol %)
a
1
1
1
1
1
1
1
1
1
-
500
500
250
100
500
500
500
0
-
-
4.1 g (C46)
-
nd
-
1640
456
a
2
4
4
4
2
4
-
1.14 g (C8)
77
93
94
16
63
-
a
3
1.09 g (C8 + C18)
0.38 g (C8)
0.30 g (C24)
99+
80
nd
40
-
556
a
4
0.87 g (C30 + C44)
3.3 g (C46)
3.39 g (C50)
-
500
a
5
0.33 g (C8)
1452
1448
152
b
6
0.23 g (C20)
c
g
7
0.38 g
a
8
4
-
-
0.60 g (PE)
-
240
a
Conditions: 1 (5 μmol), 4 or 2 (12.5 μmol), toluene (45 mL), MAO/1 = 100, 30 min, C2H4 = 5 psig, reaction carried out for 30 min without
b
temperature control followed by quenching with acidified methanol. Conditions: 1 (5 μmol), 4 (5 μmol), toluene (45 mL), MAO/1 = 100, 30 min,
C2H4 = 5 psig reaction carried out for 30 min without temperature control followed by quenching with acidified methanol. Conditions: 1 (5 μmol),
c
bipyridine (12.5 μmol), toluene (45 mL), MAO/1 = 100, 30 min, C2H4 = 5 psig reaction carried out for 30 min without temperature control
d
e
f
followed by quenching with acidified methanol. Equivalents relative to chain growth catalyst 1. Determined by GC. Determined by 1H NMR in
d10-xylene. Schulz−Flory distribution.
g
Figure 3. Alkanes and α-olefin distribution obtained from tandem catalysis experiments at the indicated time intervals: (a) Entry 2, 500 equiv of
ZnEt2 (b) Entry 3, 250 equiv of ZnEt2.
recycling ZnEt2 in situ is to decrease the amount of this
expensive reagent required during catalysis, thus the amount of
ZnEt2 was halved to 250 equiv relative to 1 and resulted in a
similar behavior (i.e., increase of α-olefins concentration at the
expense of alkanes; Table 1, Entry 3). Most importantly, a
bimodal distribution of α-olefins appeared over time, indicating
that the regenerated ZnEt2 re-entered the catalytic cycle
(Figure 3b). Toluene insoluble solids recovered (0.30 g)
upon quenching at the end of the reaction were analyzed by
GC and NMR spectroscopy. High-temperature GC analysis of
the solids revealed a Poisson distribution centered at C24, while
1H NMR spectroscopy indicated chains contained olefinic and
methyl end groups confirming chain transfer and release of the
chain via ß-H elimination.10 Further lowering of the ZnEt2
loading to 100 equiv relative to 1 increased the amount of
1
solids to 0.87 g (Table 1, Entry 4) upon quenching. H NMR
analysis indicated a C38 average distribution consisting of 80
mol % α-olefin. Interestingly, high temperature GC analysis of
the solids revealed a bimodal distribution with peaks centered
at C30 and C44 (Figure 4). GC analysis of the toluene soluble
fraction only shows traces of alkanes, and a C8 centered α-
olefins distribution.10
The original reports from Gibson indicate that Ni(acac)2, 2,
can regenerate ZnEt2 only following catalyst deactivation.3,4
4365
DOI: 10.1021/acscatal.5b01231
ACS Catal. 2015, 5, 4363−4367

ACS Catalysis
Letter
In conclusion, the Fe-catalyzed chain growth of ethylene
producing Poisson distributions of alkanes reported by Gibson
has been reinvestigated for the recycling of the chain transfer
agent, ZnEt2. It has been demonstrated that akin to Ni(acac)2,
simple Fe alkyl complexes also catalyze the conversion of
Zn(oligomer)2 to ZnEt2 and α-olefins under an ethylene
atmosphere. Most importantly, results presented here show
that ZnEt2 can be regenerated in situ without deactivation of
the Fe chain growth catalyst in a tandem catalysis fashion using
the iron “transfer” catalyst, (BiPy)2FeEt2. In contrast, Ni(acac)2
is not suitable under the reaction conditions reported here. The
dual catalyst system presented here offers the possibility to
lower the loading of ZnEt2. In addition, the product
distribution is tunable and offers a wide range of target
products in the synthesis of precision hydrocarbons.
Application of this strategy to other Catalyzed Chain Growth
systems as well as optimization of the reaction conditions such
as temperature, pressure, ratio of catalysts are currently the
focus of our efforts. Furthemore kinetic studies will shade light
on the origins of the bimodal distribution and will be reported
in due course.
Figure 4. GC chromatogram of the solids obtained from tandem
catalysis experiment, Table 1, entry 4 (100 equiv of ZnEt2).
Thus, it was decided to benchmark 2 under identical conditions
for the in situ recycling of ZnEt2 (Table 1, Entry 5). In this
instance, 3.3 g of solids were recovered along with a toluene
soluble fraction of hydrocarbons centered at C8 and consisting
of 16 mol % α-olefins. In contrast to 4, the α-olefin
concentration did not increase with time suggesting little to
no recycling of the chain transfer agent.10 In addition, 1H NMR
spectroscopy of the solids did not show the presence of olefinic
peaks and GC analysis revealed a C46 centered distribution
nearly identical to that of the CCG benchmark experiment
(Entry 1). All together, these results suggest that in contrast to
(BiPy)2FeEt2, Ni(acac)2 does not take part in tandem catalysis,
presumably due to incompatibility with MAO as suggested in
initial reports.3,4 It also indicates that Ni(acac)2 or the product
of the reaction between Ni(acac)2 and MAO do not affect the
chain growth catalyst allowing for the typical catalyzed chain
growth to occur.
ASSOCIATED CONTENT
* Supporting Information
■
S
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acscatal.5b01231.
Experimental details, NMR spectra and GC traces (PDF)
AUTHOR INFORMATION
Corresponding Author
*E-mail: renan.cariou@uk.bp.com.
■
Notes
The authors declare no competing financial interest.
Interestingly, the catalytic activity in the tandem experiments
is reduced to about one-third of the original catalytic activity
presumably due to interaction of free BiPy ligand with the chain
growth catalyst. Indeed, one of the BiPy ligands in 4 is known
to be loosely bound to the Fe center.9 Decreasing the loading
of the transfer catalyst 4 from 2.5 to 1 equiv relative to the
chain growth catalyst 1 restored the catalytic activity and
produced a hydrocarbon distribution centered at C50 consisting
of about 40 mol % α-olefins (Table 1, Entry 6). Additionally,
the toluene soluble fraction contained a distribution centered at
C20 with 63 mol % of α-olefins. The detrimental effect of free
BiPy on the catalytic activity was further demonstrated by the
reduced activity of 1/MAO in the presence of 2.5 equiv of BiPy
(Table 1, Entry 7). The reaction produced a Schulz−Flory
distribution of alkanes, indicating that free BiPy inhibits chain
growth. The importance of ZnEt2 in tandem catalysis was also
investigated by carrying out an experiment without ZnEt2 and
resulted in the formation of fully saturated polyethylene (PE) in
low activity (Table 1, Entry 8), indicating that chains do not
transfer between the two iron centers.
From these results, it is apparent that the shape and position
of the molecular weight distribution is highly dependent on the
reaction conditions (i.e., amount of ZnEt2 and ratio of catalysts
1 to 4). Given that the rate of propagation on 1 and β-H
elimination on 4 are independent of one another and the
amount of Zn reagent, we anticipate that the relative rate of
chain transfer reactions of 1 and 4 with Zn species will dictate
the shape and position of the molecular weight distributions.
This hypothesis will require further kinetic studies.
ACKNOWLEDGMENTS
■
We thank Dr. Russell Taylor and Dr. Glenn Sunley for helpful
discussions and Joanne Medcalf for GC analyses.
REFERENCES
■
(1) (a) Gibson, V. C.; Spitzmesser, S. K. Chem. Rev. 2002, 103, 283−
316. (b) Coates, G. W. Chem. Rev. 2000, 100, 1223−1252. (c) Baier,
M. C.; Zuideveld, M. A.; Mecking, S. Angew. Chem., Int. Ed. 2014, 53,
9722−9744.
(2) (a) Wei, J.; Zhang, W.; Sita, L. R. Angew. Chem., Int. Ed. 2010, 49,
1768−1772. (b) Kempe, R. Chem. - Eur. J. 2007, 13, 2764−2773.
(c) Valente, A.; Mortreux, A.; Visseaux, M.; Zinck, P. Chem. Rev. 2013,
113, 3836−3857.
(3) Britovsek, G. J. P.; Cohen, S. A.; Gibson, V. C.; Maddox, P. J.; van
Meurs, M. Angew. Chem., Int. Ed. 2002, 41, 489−491.
(4) Britovsek, G. J. P.; Cohen, S. A.; Gibson, V. C.; van Meurs, M. J.
Am. Chem. Soc. 2004, 126, 10701−10712.
(5) Wei, J.; Zhang, W.; Sita, L. R. Patent No. WO2011082418.
(6) (a) Arriola, D. J.; Carnahan, E. M.; Hustad, P. D.; Kuhlman, R. L.;
Wenzel, T. T. Science 2006, 312, 714−719. (b) Arriola, D. J.;
Carnahan, E. M.; Cheung, Y. W.; Devore, D. D.; Graf, D. D.; Hustad,
P. D.; Kuhlman, R. L.; Li Pi Shan, C.; Poon, B. C.; Roof, G. R.;
Stevens, J. C.; Stirn,P. J.; Wenzel, T. T. (Dow Global Technologies),
Patent No. WO 2005090427 A2, 2005. (c) Hustad, P. D.; Kuhlman, R.
L.; Arriola, D. J.; Carnahan, E. M.; Wenzel, T. T. Macromolecules 2007,
40, 7061−7064. (d) Hustad, P. D.; Kuhlman, R. L.; Carnahan, E. M.;
Wenzel, T. T.; Arriola, D. J. Macromolecules 2008, 41, 4081−4089.
(7) van Meurs, M.; Britovsek, G. J. P.; Gibson, V. C.; Cohen, S. A. J.
Am. Chem. Soc. 2005, 127, 9913−9923.
4366
DOI: 10.1021/acscatal.5b01231
ACS Catal. 2015, 5, 4363−4367

ACS Catalysis
Letter
(8) Vela, J.; Smith, J. M.; Lachicotte, R. J.; Holland, P. L. Chem.
Commun. 2002, 2886−2887.
(9) Yamamoto, A.; Morifuji, K.; Ikeda, S.; Saito, T.; Uchida, Y.;
Misono, A. J. Am. Chem. Soc. 1968, 90 (7), 1878−1883.
(10) See Supporting Information.
(11) (a) Mindiola, D. J. Angew. Chem., Int. Ed. 2009, 48, 6198−6200.
(b) Rodriguez, M. M.; Bill, E.; Brennessel, W. W.; Holland, P. L.
Science 2011, 334, 780−783. (c) Cowley, R. E.; Holland, P. L. Inorg.
Chem. 2012, 51, 8352−8361. (d) Holland, P. L.; Cundari, T. R.; Perez,
L. L.; Eckert, N. A.; Lachicotte, R. J. J. Am. Chem. Soc. 2002, 124,
14416−14424. (e) Vela, J.; Vaddadi, S.; Cundari, T. R.; Smith, J. M.;
Gregory, E. A.; Lachicotte, R. J.; Flaschenriem, C. J.; Holland, P. L.
Organometallics 2004, 23, 5226−5239.
4367
DOI: 10.1021/acscatal.5b01231
ACS Catal. 2015, 5, 4363−4367