Control of Chain Walking by Weak Neighboring Group Interactions in Unsymmetrical Catalysts

Article abstract of DOI:10.1021/jacs.7b08975

A combined theoretical and experimental study shows how weak attractive interactions of a neighboring group can strongly promote chain walking and chain transfer. This accounts for the previously observed very different microstructures obtained in ethylene polymerization by [κ²-N,O-{2,6-(3′,5′-R₂C₆H₃)₂C₆H₃-Rfnet=C(H)-(3,5-X,Y-2-O-C₆H₂)}NiCH₃(pyridine)], namely hyperbranched oligomers for remote substituents R = CH₃ versus high-molecular-weight polyethylene for R = CF₃. From a full mechanistic consideration, the alkyl olefin complex with the growing chain cis to the salicylaldiminato oxygen donor is identified as the key species. Alternative to ethylene chain growth by insertion in this species, decoordination of the monomer to form a cis β-agostic complex provides an entry into branching and chain-transfer pathways. This release of monomer is promoted and made competitive by a weak η²-coordination of the distal aryl rings to the metal center, operative only for the case of sufficiently electron-rich aryls. This concept for controlling chain walking is underlined by catalysts with other weakly coordinating furan and thiophene motifs, which afford highly branched oligomers with >120 branches per 1000 carbon atoms.

Full text of DOI:10.1021/jacs.7b08975

Subscriber access provided by University of Florida | Smathers Libraries
Article
Control of Chain Walking by Weak Neighbouring
Group Interac-tions in Unsymmetric Catalysts
Laura Falivene, Thomas Wiedemann, Inigo Göttker-
Schnetmann, Lucia Caporaso, Luigi Cavallo, and Stefan Mecking
J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.7b08975 • Publication Date (Web): 20 Dec 2017
Downloaded from http://pubs.acs.org on December 20, 2017
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of the American Chemical Society is published by the American Chemical
Society. 1155 Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 9
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Control of Chain Walking by Weak Neighbouring Group Interac-
tions in Unsymmetric Catalysts
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Laura Falivene,c,† Thomas Wiedemann,a,† Inigo Göttker-Schnetmann,a Lucia Caporaso b,*, Luigi Caval-
loc and Stefan Mecking a,*
a) Chair of Chemical Materials Science, Department of Chemistry, University of Konstanz, 78464 Konstanz, Germany.
b) Department of Chemistry, University of Salerno, Via Giovanni Paolo II, 84084-Fisciano (SA), Italy.
c) King Abdullah University of Science and Technology, Chemical and Life Sciences and Engineering, Kaust Catalysis
Center, Thuwal 23955-6900, Saudi Arabia.
†) Both authors contributed equally
KEYWORDS:
Catalysis;
Polymerization;
Isomerization;
Mechanism;
Weak
Interactions;
Selectivity.
ABSTRACT: A combined theoretical and experimental study shows how weak attractive interactions of a neighbouring group
can strongly promote chain walking and chain transfer. This accounts for the previously observed very different microstructures
obtained in ethylene polymerization by [κ2-N,O-{(2,6-(3',5'-R2C6H3)2C6H3-N=C(H)-(3,5-X,Y2-2-O-C6H2)}]NiCH3(pyridine)],
namely hyperbranched oligomers for remote substituents R = CH3 versus. high molecular weight polyethylene for R = CF3. From a
full mechanistic consideration the alkyl olefin complex with the growing chain cis to the salicylaldiminato oxygen donor is identi-
fied as the key species. Alternative to ethylene chain growth by insertion in this species, decoordination of the monomer to form a
cis ß-agostic complex provides an entry into branching and chain transfer pathways. This release of monomer is promoted and
made competitive by a weak 2-coordination of the distal aryl rings to the metal center, operative only for the case of sufficiently
electron rich aryls. This concept for controlling chain walking is underlined by catalysts with other weakly coordinating furane and
thiophene motifs, which afford highly branched oligomers with
>
120 branches per 1000 carbon atoms.
Scheme 1. Synthetic relevance of chain walking processes:
oligomerization of ethylene to hyperbranched oligomers.
INTRODUCTION
Catalytic insertion reactions promoted by late transition
metal complexes are unique in their synthetic and practical
scope. Other than established polyolefin processes,1,2 they are
often tolerant towards heteroatom-containing functional
groups and they provide unique branching structures. A fun-
damental ubiquitous issue is the occurrence of ß-hydride elim-
ination in these reactions. In a classical olefin chain growth
polymerization, the ratio of insertion chain growth rate vs. rate
of chain transfer via ß-hydride elimination will determine the
products’ molecular mass. Moreover, ß-hydride elimination is
also a key step of ‘chain walking’, that is, migration of a metal
center along a hydrocarbon chain via a series of ß-H elimina-
Despite the relevance and utility of chain walking processes
and the underlying step of ß-H elimination, a concept is miss-
ing to rationally promote or suppress their occurrence and
competition with other pathways.
tions and reinsertions.3,4, This can result in the formation of
5
highly branched polymers from ethylene as the sole monomer
(Scheme 1). Such ‘chain walking’ processes are not restricted
to olefin polymerizations, but also occur in various other in-
Given this background, neutral nickel(II) salicylaldiminato
complexes provide a unique case example. These catalysts,
introduced by Grubbs et al.19 and Johnson et al.,20 are out-
standing in being highly active for ethylene polymerization,
stable towards water, and compatible with the presence of
,
,
, ,
,
,
,
-
stances.6 7 8 9 10 11 12 13 14 For example, they are essential to
terminal functionalizations from internal double bonds, which
are particularly attractive for the valorization of plant oils.15
The latter example illustrates a situation frequently encoun-
tered: an insertion reaction is strongly favored for the linear
metal-alkyl even though branched alkyls are also accessible,
resulting in a strong preference for the non-branched, linear
product. In fact, even for cases of thylene polymerization to a
strictly linear polymer, extensive underlying chain walking
-
certain vinyl monomers.21 25 We have advanced these catalysts
H2C=CH2
R = Me
R = CF3
occurs.16
-18
ACS Paragon Plus Environment
liquid, amorphous
solid, semicrystalline
R1-pyr

Journal of the American Chemical Society
Page 2 of 9
by introduction of N-terphenyl motifs (R1-pyr) which allow
Overall the underlying mechanisms are not understood.
1
2
3
4
5
6
7
8
for an essential tuning of their catalytic properties. This is
based on the serendipitous observation that R-substituents in
an actually rather remote position to the metal site alter the
outcome of ethylene polymerization entirely.26 For example,
for R = CF3 a linear semicrystalline polymer is obtained vs.
hyperbranched liquid low-molecular weight oligomers for R =
CH3 (Figure 1). At the same time, the catalysts stability and
activity a very similar.
We now develop a conclusive picture from a comprehensive
combined theoretical and experimental study, which shows
how weak interactions of a neighbouring group can direct the
competition between chain growth and ß-hydride elimination.
RESULTS AND DISCUSSION
9
Figure 1. Products of ethylene polymerization depending on the
nature of the catalysts’ remote substituents R (reaction conditions:
60 °C, 20 atm ethylene pressure).
Phenomenologically, electron rich substituents in the 3,5-
R
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
positions of the N-terphenyl moiety of 1-pyr make the nickel
center more prone for β-hydride elimination. This increases
chain walking and chain transfer reactions explaining the
lower molecular mass oligomers with high degrees of branch-
ing produced by these complexes. Regarding the open ques-
tion how the electronic character of the substituents translates
to the nickel center, one possibility is a transfer of electron
density via increasing the overall electron density of the con-
jugated system of the salicylaldiminato ligand. A second sce-
nario comprises discrete interactions of the salicylaldiminato
ligand with the nickel center additional to its chelating N,O-
coordination, e.g. by direct π-interaction between the ligand
aryl moieties and the metal center.
-
29
Due to its practical utility27 and possibly conceptually in-
structive nature, the origin of this peculiar effect of the remote
substituents is being sought.30 -34 Marks and coworkers sug-
gest an attractive F···H interaction of a remote –CF3 group
with the growing chain to disfavor ß-H elimination,30 but a
fluorine-free electron withdrawing group yields virtually iden-
tical product microstructures.35 In a related study, the two
complexes 2a and 2b were both found to form highly
branched ethylene oligomers, with a higher observed activity,
three-fold increased molecular mass and slightly higher degree
of branching for X = SO2. This is suggested to result from a
-
hemilabile interaction36 of the sulfonyl group.31 However,
the question for a conclusive mechanistic scenario that ac-
counts for increased molecular masses (favored by suppressed
ß-H elimination) and similar or slightly increased branching
(favored by increased ß-H elimination) remains open.
Catalyst precursors and pressure reactor experiments.
With regard to the aforementioned issues, molecular structures
of the catalyst precursors can be instructive. These neutral
Ni(II) salicylaldiminato complexes are stabilized by a P-, N-
or other donor like trimethylphosphine or pyridine. These
often bind relatively strongly, and compared with ethylene
complexes as an example of species that occur during catalysis
they are more stable (note that concerning activation of the
precursors for catalysis, this is overcome by the large excess
of olefin and the high dilution under pressure reactor condi-
tions). This will alter the role of any additional weak interac-
tions in the catalyst precursor, and likely reduce their rele-
vance. For related anilinotropone complexes Jenkins and
Brookhart were able to observe ß-agostic complexes at low
temperatures.4 2 These are devoid of such additional strongly
coordinating ligands. Applying analogous synthetic methodol-
ogy to salicylaldiminato complexes to date did not yield clean
well-characterized products.4 3
39
Our interpretation from studies of a range of substitution
patterns35- is that the electron withdrawing or donating na-
4 1
R
ture of the remote substituent (R) is the decisive feature in 1-
pyr. However, this is a purely phenomenological account to
date.
2
ACS Paragon Plus Environment

Page 3 of 9
Journal of the American Chemical Society
Table 1. Polymerization results with complexes bearing potential coordinating motifs in the ligand backbone.a
1
2
3
entry
catalyst
T [°C]
yield
[g]
TOFb
Mn (NMR)
[g mol-1]c
Mn (GPC)
[g mol-1]d
Mw/Mn
(GPC)d
branches
/1,000 Ce
precursor
1f
2
CF31-pyr
Me1-pyr
3-pyr
50
40
40
60
80
40
60
40
40
23.4
10.9
9.9
6.2
1.8
41,800
19,500
17,700
11,100
3,200
3,900
550
-
19,000f
3,500
5,900
5,100
5.1
1.7
1.9
2.0
2.5
1.7
1.7
1.9
1.9
10
77
4
5
6
7
8
9
1,300
4,800
2,600
1,200
4,100
2,200
1,500
1,600
3
109
115
4
5
3-pyr
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
3-pyr
1,800
8,600
3,500
2,200
2,800
112
6
7
4-pyr
2.2
0.3
2.1
127
123
124
145h
4-pyr
8
9g
2a-pyr
2a-PMe3
3,800
10,000
2.8
areaction conditions: 20 µmol of catalyst in 100 mL of toluene for 1 h at 20 bar of ethylene. b TOF x mol [C2H4] x mol-1 [Ni] h-1. cmolecu-
1
d
lar weights calculated from H NMR intensity ratio of unsaturated end groups vs. overall integral. in THF vs. polystyrene standards.
1
edegree of branching calculated from H NMR intensity ratio of methyl groups (corrected for saturated end groups) vs. overall integral.
fdata from ref. 26: 40 µmol of catalyst for 30 min at 40 bar, GPC data from high temperature GPC in trichlorobenzene. gdata from ref. 31:
10 µmol of catalyst, [Ni(cod)2] as cocatalyst at 8 bar for 40 min. hnot corrected for endgroups
In pressure reactor experiments, they yield ethylene oligo-
mers with very high degrees of branching (Table 1). Like for
Me1-pyr, the oligomers possess a hyperbranched microstruc-
ture as concluded from observable sec-Bu branches (cf. Sup-
porting Information). In detail, degrees of branching exceed
those of the terphenyl system (Me1-pyr) while molecular
weights are slightly higher. Concerning the branching patterns,
3-pyr and 4-pyr produce a higher proportion of methyl
branches and less higher branches compared to Me1-pyr (Table
S1 in the Supporting Information). In conclusion, the observed
catalytic behavior of complexes 3-pyr and 4-pyr underlines
that electron rich potentially coordinating distal substituents
promote pathways for branch formation and chain transfer.
This is shown here for heteroaromatic rings, going beyond
previous findings on substituted phenyl groups.
Figure 2. Comparison of X-ray crystal structures (50 % probabil-
ity ellipsoids, H atoms omitted for clarity) of CF31-pyr and Me1-
pyr.
Notwithstanding, X-ray diffraction analyses of N-terphenyl
salicylaldiminato nickel pyridine complexes show that in the
Cyclic voltammetry on the catalyst precursors showed dis-
tinct effects of the nature of the distal aryl rings on the elec-
solid state structures, one of the distal aryl rings of the ter-
tronic properties of the nickel center. All complexes showed
phenyl moiety is located close to an axial position of the nick-
oxidation and reduction transitions during cyclic voltammetry
el center with typical nickel-aryl distances in the range of 3.05
measurements in the potential range expected for a four coor-
to 3.47 Å (Figure 2), the shorter distances being found for Me1-
dinate square planar Ni(II)/Ni(III) pair. Measurements were
pyr.26,40,41 These values suggest the possibility of a π-
carried out using different scan rates from 25 to 2,000 mV s-1
interaction of an aromatic ring with the metal center in this
and the half-wave potential was found to be independent of
the scan rate. The oxidation of all complexes is only partially
class of complexes.4 4
To further probe for the role of such possible -interactions,
we prepared complexes 3-pyr and 4-pyr with electron-rich
furanyl- and thiophenyl distal aryl motifs.
reversible which is attributed to a relatively fast decomposi-
tion of the Ni(III)-species. Cyclic voltammograms (cf. Figure
S1 in the Supporting Information) of the corresponding salic-
ylaldimines did not show any redox transitions at the poten-
3
ACS Paragon Plus Environment

Journal of the American Chemical Society
tials observed for the complex. Therefore, the oxidation and
reduction processes during cyclic voltammetry measurements
are assumed to be metal centered. The considerably higher
half wave potential of CF31-pyr compared to Me1-pyr shows
that the nature of the remote substituents clearly affects the
electron density at the Ni center, as reflected by the easier
oxidation of the more electron rich Me1-pyr (Table 2). The E1/2
values of 3 and 4 are in line with a relatively electron rich
nature.
Page 4 of 9
1
2
3
4
5
6
7
8
Table 2: Half-wave potentials of catalyst precursors ac-
cording to cyclic voltammetry measurements.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
entry
complex
CF31-pyr
Me1-pyr
3-pyr
E1/2 [mV]a
1
306
-33
84
Figure 3. NCI analysis of the electron density of Me1-pyr. The
isosurface corresponds to a reduced density gradient isosurface of
s = 0.5 a.u. The surface is colored on a blue-to-red scale according
to the sign of (2) , ranging from -0.3 to 0.3 a.u. (see the SI for
details). Within this scale blue indicates weak attractive interac-
tions, and red indicates weak repulsive interaction. For an addi-
tional perspective of the map see Figure S4.
2
3
4
4-pyr
76
a determined from referenced cyclic voltammograms
An increased electron density at the metal center is expected
to result in an increased propensity for ß-hydride elimination
and resulting chain transfer. Indeed, molecular masses of the
oligomers obtained with the different catalyst precursors and
their half wave potentials decrease from CF31-pyr > 3 ≈ 4 >
Me1-pyr. Regarding degrees of branching of the polymers and
oligomers, CF31-pyr stands out with its low degree of branch-
ing and low electron density as reflected by its high E1/2 value.
For the more electron rich catalysts Me1-pyr, 3-pyr and 4-pyr,
which all yield highly branched oligomers, there is no clear
correlation. Degrees of branching are not only governed by the
propensity of the catalyst for ß-hydride elimination, but also
its relative ability for insertion into a linear alkyl vs. different
branched alkyls (vide infra). Thus, it appears reasonable that
sterics might have a different (higher) impact on chain trans-
fer, and add to the effect of electron density at the metal.
To explore whether this interaction can be more relevant
during catalysis, when a vacant coordination site can be pre-
sent on the nickel, we analyzed the coordinatively unsaturated
Ni-methyl fragment Me1 and CF31 (that is the product of pyri-
dine dissociation from Me1-pyr and CF31-pyr). For both sys-
tems two different isomers were located (Figure 4), one of
them showing a clear η2-interaction of one of the distal aro-
matic rings of the terphenyl moiety of the ligand to the metal
center (Figure 4 left).
These findings show that the nature or substitution pattern,
respectively, of the distal aryl rings clearly affects the elec-
tronic situation at the metal site as reflected by its redox prop-
erties, without implications on how this effect originates.
Active species and intermediates during catalysis. The role
of the distal aryl substituents in catalysis was accessed by a
comprehensive theoretical DFT study. The reported free
energies were built through solvent single point energy calcu-
lations on the BP86/6-31G geometries using the M06 func-
tional and the triple-ζ TZVP with the Gaussian09 package. For
more details see the SI.
Figure 4. Two possible isomers of the dissociated three coordi-
nate nickel-methyl species Me1 and their relative GTol in kcal mol-
1(GTol for CF31 isomers are also given).
Notably, for Me1 the isomer featuring the η2-interaction of
the aryl ring of the terphenyl moiety (Figure 4 left) is preferred
by 3.1 kcal mol-1, whereas for CF31 the isomer having no nick-
el-aryl η2-interaction, and having the methyl group trans to
oxygen atom is favored by 1.1 kcal mol-1 (Figure 4 right). This
difference is of course related to the different coordinating
ability of the aryl group in Me1-pyr and CF31-pyr and provides
a first indication of a non covalent interaction possibly rele-
vant during catalysis,47 which could account for the different
catalytic behavior observed for Me1-pyr and CF31-pyr.
To have a better understanding of the impact of fluorines on
the electronic structure of the catalysts, first we analyzed the
potential π-interaction between one of the aryl rings of the
terphenyl moiety and the nickel center, suggested on the base
of the short nickel-aryl distances in the X-ray structure of Me1-
pyr. To this end we performed a non-covalent interaction
(NCI) analysis using the approach developed by W. Yang and
coworkers.45,46 The gradient isosurface of the reduced density
gradient (s = 0.5 a.u.) for Me1-pyr (Figure 3) indicates weak
attractive interactions (blue, a.u.) along the axis con-
necting the Ni atom and the vicinal C atoms of the aryl ring,
confirming an attractive interaction between the Ni center and
the aryl ring of the terphenyl ligand.
4
ACS Paragon Plus Environment

Page 5 of 9
Journal of the American Chemical Society
complex were calculated to involve transition states with
1
2
extremely high free energies and therefore can be ruled out
(See SI).
3
4
5
6
7
8
An alternative pathway to reach 1--C is through coordina-
tion of ethylene to 1--T followed by isomerization of the
ethylene complex 2-Coor-T to 2-Coor-C – steps already
considered in the linear chain growth reaction of Figure 5 –
followed by ethylene dissociation from 2-Coor-C, leading to
the β-agostic intermediate 1--C.
9
Figure 5. Energy profile (GTol in kcal mol-1) for the ethylene
insertion with complexes Me1-pyr (blue) and CF31-pyr (orange).
Energies are relative to the β-agostic trans-polymeryl nickel
complex 1--T.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Linear chain growth. Starting from 1--T (referenced as
the zero energy point, Figure 5), ethylene coordination occur-
ring by opening of the β-agostic interaction requires overcom-
ing a free energy barrier of 11.2 (CF31) and 9.2 (Me1) kcal mol-
1. For both complexes the resulting 2-Coor-T intermediate is
slightly higher in energy relative to 1--T
+ C2H4
(ΔG = 1.9 kcal mol-1). Due to the higher energy for the direct
ethylene insertion from the 2-Coor-T intermediate (almost +4
kcal mol-1 respect to the insertion from 2-Coor-C), we investi-
gated the systems’ isomerization to the slightly more energetic
2-Coor-C intermediate via the tetrahedral four-coordinated
transition state TS-2isom. Migratory insertion via TS-2ins yields
the stable β-agostic complex, 2--T. The overall free energy
barrier from 1--T to TS-2ins amounts to 12.7 and
13.7 kcal mol-1 for Me1 and CF31, respectively.
The two systems behave similarly in the linear chain growth
step, with CF31 showing only a slightly higher energy profile,
probably due to a slightly higher steric hindrance in CF31, as
evidenced by topographic steric maps (Figure 6). The main
steric hindrance is located in the eastern hemisphere,48 hosting
the monomer in the insertion transition state TS-2ins. Slightly
higher steric hindrance in CF31 can also be observed near the
north and south poles, with the CF3 groups generating steric
hindrance in the western hemisphere.
Figure 7. Two potential pathways for the decoordination of eth-
ylene to form 1--C. Free energies (GTol in kcal mol-1) of all
intermediates for Me1 (blue) and CF31 (orange).
We found that release of ethylene from 2-Coor-C can occur
via two pathways, see Figure 7. The most straightforward
pathway is direct dissociation of ethylene via TS-2decoor, and
an energy barrier of 12.3 and 10.8 kcal mol-1 for Me1 and CF31,
respectively. The other ethylene dissociation pathway pro-
ceeds in two steps.5 0 At first ethylene is displaced by a η2-
interaction of the distal aryl ring of the terphenyl moiety with
the Ni center trough TS-2decoorAr, see Figure 8. In this transi-
tion state the complex assumes a square pyramid like geome-
try with the η2-interaction of the distal ring lying on the metal
plane and the ethylene in the apical position at almost 3 Å
from the metal. After displacement of ethylene (i.e. formation
of 1-Pr-C), the ligand-Ni interaction is replaced by a β-agostic
hydrogen bond.
Figure 6. Topographic steric maps of the fragments Me1-T and
CF31-T. The complexes are oriented as shown for CF31-T on the
left.
Branch Formation and Chain Transfer Reactions. Both
branch formation via chain walking and chain transfer proceed
via β-hydride elimination (BHE) as the initial step.4 9 Starting
from 1--T, BHE leads to the nickel hydride species 1-BHE-
T with a propene molecule coordinated trans to the oxygen
atom. Due to the unfavorable kinetics and thermodynamics
(see Figure 10 in the last section) this possibility was ruled out
for both complexes. Hence, BHE has to occur from the cis
isomer 1--C, leading to the sterically less crowded and ener-
getically favored isomer 1-BHE-C. However, all potential
pathways for a direct cis/trans isomerization of the β-agostic
Figure 8. Geometry of TS-2decoorAr for Me1-pyr system.
5
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 6 of 9
As shown in Figure 7, for CF31 the direct pathway is favored
and 12.7 kcal mol-1 in Figures 9 and 5), which explains the
by 2.5 kcal mol-1 over the two steps pathway (compare barriers
of 10.8 and 13.3 kcal mol-1 in Figure 7). In contrast, for Me1,
the two steps pathway, is 2.0 kcal mol-1 lower in energy then
the direct pathway (compare barriers of 12.3 and 10.3
kcal mol-1 in Figure 7). This decisive difference is due to
aforementioned stronger interaction between Ni and the elec-
tron rich aryl rings of Me1 in the transition state (TS-2decoorAr).
Hereby, an entry into branching and chain transfer pathways is
enabled (vide infra).
high amount of branching obtained by Me1-pyr.
1
2
3
4
5
6
7
8
As for CF31, both the chain termination and branching path-
ways are of higher energy compared to the Me1 pathways, see
Figure 8. Combined with the more difficult access to 1--C
calculated for CF31 (Figure 7), this explains the much higher
degree of branching of products obtained by Me1-pyr com-
pared to those by CF31-pyr, and the low molecular weight
oligomers obtained with Me1-pyr in pressure reactor experi-
ments in contrast to the high molecular weight polyethylene
produced by CF31-pyr.
9
As result, for Me1 formation of 1--C from 2-Coor-C, with
a barrier of 10.3 kcal mol-1, becomes competitive with chain
growth from 2-Coor-C, with a barrier of 10.0 kcal mol-1, by
only 0.3 kcal mol-1 (Figure 7). Conversely, for CF31, ethylene
insertion from 2-Coor-C, with a barrier of 8.7 kcal mol-1, is
favored by 2.1 kcal mol-1 over ethylene dissociation, with a
barrier of 10.8 kcal mol-1. The small energy difference be-
tween ethylene dissociation and ethylene insertion allows Me1
to enter easily into the chain branching and termination path-
ways (Figure 9).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The overall picture emerging (see Figure 10) from DFT cal-
culations indicates that the ethylene complex (2-Coor-C), is
the key intermediate of the catalytic cycle. It gives access to
branch formation and chain transfer reactions that compete
with linear chain growth. This competition affects the product
molecular weight and microstructure, i.e. branched or linear
products. In line with its decisive role in the catalytic cycle,
the largest differences between the two catalytic systems stud-
ied are found here. Starting from 2-Coor-C, for both catalysts
linear chain growth has the lowest energy barrier, in agree-
ment with the formation of long chain products including
linear sequences (rather than short chain oligomers like butene
or hexene). However, for Me1-pyr, branch formation and chain
transfer pathways are accessible with a ΔΔG≠ of only
0.3 kcal mol-1 relative to the chain growth. This is the result of
an energetically favorable pathway involving a weak interac-
tion of the distal aryl ring of the ligand with the metal, which
facilitates the releasing of ethylene to form the cis -agostic
complex key to chain transfer and branch formation. This
pathway is only viable for distal rings with a sufficient coordi-
nation ability. Electron donating substituents promote and
enable this pathway, whereas for the case of electron with-
drawing trifluoromethyl groups this pathway is blocked. This
results in a distinctively different outcome of the overall cata-
lytic reaction, namely linear polymer vs. highly branched
oligomers.
Figure 9. Free energies (GTol in kcal mol-1) of all important
intermediates for chain transfer and branch formation with Me1
(blue) and CF31 (orange).
Focusing on Me1 energetics, 1--C can undergo BHE with
the relatively small barrier of 8.8 kcal mol-1 to form the stable
1-BHE-C intermediate with propene coordinated in the less
sterically crowed position. Intermediate 1-BHE-C can evolve
either to chain transfer or chain branching. Chain transfer
occurs via coordination and direct insertion of an incoming
ethylene molecule via the five coordinated intermediate 1-
Transf-1 and an overall barrier of 19.8 kcal mol-1 (Figure 9,
top). The following facile release of propene leads to the β-
agostic complex 1-Transf-2, which can start the growth of a
new polymer chain.
As further evidence that energy differences between chain
growth vs. release of ethylene from 2-Coor-C to 1--C
(ΔΔG≠
in kcal mol-1) define the polyethylene micro-
prop -elim
structure, calculations performed on catalysts analogous to 1
with other remote substituents on the terphenyl distal rings
qualitatively agree with their previously found propensity to
form highly branched, low molecular weight polymers (MeO1-
pyr) or low-branched linear polymers (NO21-pyr) under pres-
sure reactor conditions (See Table 3).
Table 3. ΔΔG≠ as a measure for the relative propensity for
chain growth vs. access to chain transfer and branching
pathways and experimental branching.
Moving to the branch formation pathway (Figure 9, bot-
tom), the coordinated propene in 1-BHE-C can reinsert in a
2,1-fashion via TS-12,1-ins, leading to the β-agostic resting state
3--C. Coordination of a new ethylene molecule displaces the
agostic interaction, leading to intermediate 3-Coor-C. Finally,
monomer insertion into the Ni-isopropyl bond via TS-3ins
leads to the methyl branched product 4--T. The rate deter-
mining step along the branch formation pathway is the open-
ing of the agostic interaction by ethylene coordination. Ac-
cording to the numbers reported in Figure 9, for Me1-pyr chain
branching is favored over chain termination by 5.9 kcal mol-1
(compare barriers of 13.9 and 19.8 kcal mol-1 in Figure 9). The
overall barrier for branching is only 1.2 kcal mol-1 higher than
the overall barrier for chain growth (compare barriers of 13.9
entry
catalyst
ΔΔG≠
branches
/1,000 Cd
10
prop -elim
1a
2b
3c
4a
5a
CF31-pyr
NO21-pyr
3-pyr
Me1-pyr
MeO1-pyr
-2.1 kcal mol-1
-1.6 kcal mol-1
-0.4 kcal mol-1
-0.3 kcal mol-1
-0.2 kcal mol-1
11
109c
77
79
a) Taken from ref. 26, polymer obtained at 40 bar ethylene, 50 °C.
b) Taken from ref. 26, polymer obtained at 40 bar ethylene, 50 °C.
6
ACS Paragon Plus Environment

Page 7 of 9
Journal of the American Chemical Society
c)
d)
This work, polymer obtained at 20 bar ethylene, 40 °C. De-
molecular weight highly branched oligomers compared to
termined by 13C NMR spectroscopy.
CF31-pyr (for complete DFT data cf. SI). Interestingly, the key
interaction of the ligand with the metal occurs also in this case
through a η2-coordination of the ring rather than via the lone
pair of the furanyl-oxygen, as reported in Figure 11 for 1-T.
1
2
3
4
Finally, for a structurally different catalyst 3-pyr with ter-
phenyl distal rings replaced by 2-furanyl substituents, the
calculated ΔΔG≠
= -0.4 kcal mol-1 for system 3-pyr
prop -elim
5
agrees with its experimentally observed ability to form low
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 10. Gibbs free energies (GTol in kcal mol-1) of all important intermediates for linear chain growth, chain transfer
and branch formation with Me1 (blue) and CF31 (orange).
CONCLUSIVE SUMMARY
From the combined theoretical and experimental study of
reaction pathways during catalysis and complementing exper-
imental properties of the catalyst precursors a conclusive
mechanistic account evolves. This reveals an additional weak
interaction of a neighbouring group as the source of the pecu-
liar effect of the rather remote substituents.
A key underlying element is the unsymmetric nature of the
catalysts, with an N,O-coordinating chelating ligand. Chain
growth occurs from the alkyl olefin species with the growing
chain cis to the oxygen donor. Alternative to this chain
growth, release of the olefin to the corresponding cis ß-agostic
alkyl can occur. This can undergo facile ß-H elimination,
which by further reaction steps results in chain transfer or
branch formation. It is important to note that the trans ß-
agostic alkyl – which is the product of the aforementioned
migratory chain growth event – does not undergo ß-H elimina-
Figure 11. Two possible isomers 1-T of the dissociated three
coordinate nickel-methyl starting species of catalyst 3-pyr and
tion which is associated with a high barrier for this isomer.
their relative GTol in kcal mol-1.
The olefin release step, which is decisive for the microstruc-
ture obtained, can be facilitated by additional weak binding
interactions of the chelating N,O-ligand: coordination of the
distal aryl rings stabilizes an ethylene-free intermediate, from
7
ACS Paragon Plus Environment

Journal of the American Chemical Society
which the cis ß-agostic alkyl forms. This pathway is only
viable for sufficiently strongly binding motifs, and in the
system studied it is not accessible for aromatics that are elec-
tron deficient as a result of electron withdrawing substituents.
Page 8 of 9
1
2
3
4
5
6
7
8
9
These findings explain the formation of very different, use-
ful products in preparative polymerization experiments with
these robust and very active catalysts as controlled by the
choice of substituents. More generally, they provide a concept
for addressing the competition between ß-H elimination and
chain growth via weak attractive interactions of a specific site
of the (chelating) ligand. As ß-H elimination is the decisive
step for chain walking as well as release of substrate from the
catalyst, this applies to polymerization as well as isomeriza-
tion catalytic reactions.15,6,7,8,9,10,11,12,13,14-15 Isomerizing func-
tionalization schemes are increasingly developed for the selec-
tive generation of complex molecules with biochemical activi-
ty. The concept revealed here can contribute to also advancing
such synthetic schemes, in addition to controlling polymer
materials’ properties.
(3)
Johnson, L. K.; Killian, C. M.; Brookhart, M. J. Am. Chem.
Soc. 1995, 117, 6414–6415.
(4)
1001–1003.
(5)
428–441.
(6)
Chem. 2000, 65, 3966–3970.
(7)
107, 213–219.
(8)
2006, 25, 3040–3044.
(9)
(10) Kochi, T.; Hamasaki, T.; Aoyama, Y.; Kawasaki, J.;
Kakiuchi, F. J. Am. Chem. Soc. 2012, 134, 16544–16547.
(11) Werner, E. W.; Mei, T.-S.; Burckle, A. J.; Sigman, M. S.
Science 2012, 338, 1455–1458.
(12) Zhang, C.; Santiago, C. B. Kou, L.; Sigman, M. S. J. Am.
Chem. Soc. 2015, 137, 7290–7293.
(13) Juliá-Hernandez, F.; Moragas, T.; Cornella, J.; Martin, R.
Nature 2017, 545, 84–89.
(14) Lin, L.; Romano, C.; Mazet, C. J. Am. Chem. Soc. 2016,
138, 10344–10350.
(15) Goldbach, V.; Roesle, P.; Mecking, S. ACS Catal. 2015, 5,
5951–5972.
(16) Kanazawa, M.; Ito, S.; Nozaki, K. Organometallics 2011,
30, 6049–6052.
(17 ) Cowley, M. P.; Jordan, R. F. Angew. Chem. Int. Ed. 2011,
50, 3744–3746.
(18 ) Noda, S.; Nakamura, A.; Kochi, T.; Chung, L. W.;
Morokuma, K.; Nozaki, K. J. Am. Chem. Soc. 2009, 131, 14088–
14100.
(19) Younkin, T. R.;Connor, E. F.; Henderson, J. I.; Friedrich,
S. K.; Grubbs, R. H.; Bansleben, D. A. Science 2000, 287, 460–462.
(20 ) Johnson, L. K.; Bennett, A. M. A.; Ittel, S. D.; Wang, L.;
Parthasarathy, A.; Hauptman, E.; Simpson, R. D.; Feldman, J.;
Coughlin E. B. (DuPont) Eur. Pat. 952997 B1, 2004.
(21) Mu, H.; Pan, L.; Song, D.; Li, Y. Chem. Rev. 2015, 115,
12091–12137.
(22) Leblanc, A.; Grau, E.; Broyer, J.-P.; Boisson, C.; Spitz, R.;
Monteil, V. Macromolecules 2011, 44, 3293–3301.
(23) Radlauer, M. R.; Buckley, A. K.; Henling, L. M.; Agapie,
T. J. Am. Chem. Soc. 2013, 135, 3784–3787.
(24 ) Xi, Z.; Bazzi, H. S.; Gladysz, J. J. Am. Chem. Soc. 2015,
137, 10930–10933.
(25 ) Ölscher, F.; Göttker-Schnetmann, I.; Monteil, V.; Mecking,
S. J. Am. Chem. Soc. 2015, 137, 14819–14828.
(26 ) Zuideveld, M. A.; Wehrmann, P.; Röhr, C.; Mecking, S.
Angew. Chem Int. Ed. 2004, 43, 869–873.
(27 ) Osichow, A.; Rabe, C.; Vogtt, K.; Narayanan, T.; Harnau,
L. ; Drechsler, M.; Ballauff, M.; Mecking, S. J. Am. Chem. Soc. 2013,
135, 11645–11650..
(28 ) Wiedemann, T.; Voit, G.; Tchernook, A.; Roesle, P.;
Göttker-Schnetmann, I.; Mecking, S. J. Am. Chem. Soc. 2014, 136,
2078–2085.
(29) Wiedemann, T.; Tchernook, A.; Göttker-Schnetmann, I.;
Mecking, S.; Bessel, M.; Omeis, J.; Frank, A. EP 2891511 A1.
Priority: November 22, 2013. Applicant: Byk-Chemie GmbH.
(30 ) Weberski, M. P.;Chen, C.; Delferro, M.; Zuccaccia, C.;
Macchioni, A.; Marks, T. J. Organometallics 2012, 31, 3773–3789.
(31) Stephenson, C. J.;McInnis, J. P.; Chen, C.; Weberski, M.
P.; Motta, A.; Delferro, M.; Marks, T. J. ACS Catalysis 2014, 4, 999–
1003.
Moehring, V. M.; Fink, G. Angew. Chem. Int. Ed. 1985, 24,
Guo, L.; Dai, S.; Sui, X.; Chen, C. ACS Catal. 2016, 6,
Wakamatsu,, H.; Nishida, M.; Adachi, N.; Mori, M. J. Org.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Behr, A.; Obst, D.; Westfechtel, A. Eur. J. Lipid Sci. 2005,
Ghebreyessus, K. Y.; Angelici, R. J. Organometallics
Grotjahn, D. B. Pure Appl. Chem. 2010, 82, 635–647.
ASSOCIATED CONTENT
Supporting Information.
Synthetic procedures for preparation of new ligands and com-
1
plexes. H, 13C and 19F NMR spectra for new compounds. Proce-
dures for ethylene polymerisation. Polymer analytical data. Com-
putational details, absolute energies (in Hartrees) of all molecules
whose geometries were optimized. This material is available free
of charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
Email: Stefan.Mecking@uni-konstanz.de
Email: lcaporaso@unisa.it
ACKNOWLEDGEMENT
Financial support by the DFG (Me1388/14-1) and by Byk is
gratefully acknowledged. Stefano Santacroce, Florian Wimmer
and Giusi Boffa contributed as a part of their undergraduate stud-
ies. We thank Rainer Winter for discussions of cyclic voltamme-
try data.
TOC entry
REFERENCES
(1)
Stürzel, M.; Mihan, S.; Mülhaupt, R. Chem. Rev. 2016,
116, 1398–1433.
(32) Wang, J.; Yao, E.; Chen, Z.; Ma, Y. Macromolecules 2015,
48, 5504–5510.
(2)
Baier, M. C.; Zuideveld, M. A.; Mecking, S. Angew. Chem.
Int. Ed. 2014, 53, 9722–9744.
8
ACS Paragon Plus Environment

Page 9 of 9
Journal of the American Chemical Society
1
2
3
4
5
6
7
(33) Hu, Y.; Dai, S.; Chen, C. Dalton Trans. 2016, 45, 1496–
1503.
(34 ) Chen, Z.; Mesgar, M.; White, P. S.; Daugulis, O.;
Brookhart, M. ACS Catalysis 2015, 5, 631–636.
(35 ) Osichow, A.; Göttker-Schnetmann, I.; Mecking, S.
Organometallics 2013, 32, 5239–5242.
8
9
(36 ) Jeffrey, J. C.; Rauchfuss, T. B. Inorg. Chem. 1979, 18,
2658-2666.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(37 ) Mecking, S.; Keim, W. Organometallics 1996, 15, 2650-
2656.
(38 ) Bader, A.; Lindner, E. Coord. Chem. Rev. 1991, 108, 27-
110.
(39) Slone, C. S.; Weinberger, D. A.; Mirkin, C. A.: The
Transition Metal Coordination Chemistry of Hemilabile Ligands. In
Progress in Inorganic Chemistry; John Wiley & Sons, Inc.: 2007, p
233-350.
(4 0 ) Göttker-Schnetmann, I.; Wehrmann, P.; Röhr, C.;
Mecking, S. Organometallics 2007, 26, 2348–2362.
(4 1) Bastero, A.; Göttker-Schnetmann, I.; Röhr, C.; Mecking, S.
Adv. Synth. Catal. 2007, 349, 2307–2316.
(4 2) Jenkins, J. C.; Brookhart, M. J. Am. Chem. Soc. 2004, 126,
5827–5842.
(4 3) Voit, G.; Mecking, S. unpublished results.
(4 4 ) In fact, also for 2a in the solid state the distal aryl ring is
located over the Ni center. Cf. ref. 31.
(45) Johnson, E. R.; Keinan, S.; Mori-Sánchez, P.; Contreras-
García, J.; Cohen, A. J.; Yang, W., J. Am. Chem. Soc. 2010, 132,
6498–6506.
(46) Contreras-García, J.; Johnson, E. R.; Keinan, S.; Chaudret,
R.; Piquemal, J.-P.; Beratan, D. N.; Yang, W., J. Chem. Theory
Comput. 2011, 7, 625–632.
(47) Neel, A. J.; Hilton, M. J.; Sigman, M. S.; Toste, F. D.,
Nature 2017, 543, 637–646.
(48) Falivene, L., Credendino, R., Poater, A., Petta, A., Serra,
L., Oliva, R., Scarano, V.; Cavallo, L. Organometallics 2016, 35,
2286–2293.
(49) The chain transfer reaction involving a monomer assisted
β-hydride elimination (BHT) is not energetically competitive.
(50) Jian, Z.; Falivene, L.; Boffa, G.; Ortega Sánchez, S.;
Caporaso, L.; Grassi, A.; Mecking, S. Angew. Chem. Int. Ed. 2016,
128, 14590–14595.
9
ACS Paragon Plus Environment