Structure/properties relationship for bis(phenoxyamine)Zr(IV)-based olefin polymerization catalysts: A simple DFT model to predict catalytic activity

Article abstract of DOI:10.1021/ma300343c

The productivity of a number of bis(phenoxyamine)Zr(IV)-based catalysts (bis(phenoxyamine) = N,N′-bis(3-R₁-5-R₂-2-O-C ₆H₂CH₂)-N,N′-(R₃) ₂-(NCH₂CH₂N)) in eth

Full text of DOI:10.1021/ma300343c

Article
pubs.acs.org/Macromolecules
Structure/Properties Relationship for Bis(phenoxyamine)Zr(IV)-Based
Olefin Polymerization Catalysts: A Simple DFT Model To Predict
Catalytic Activity
Gianluca Ciancaleoni,†,‡ Natascia Fraldi,‡,§ Roberta Cipullo,§ Vincenzo Busico,§ Alceo Macchioni,†
and Peter H. M. Budzelaar*,⊥
†Dipartimento di Chimica, Universita
̀
di Perugia, via Elce di Sotto, 8, 06123 Perugia, Italy
‡Dutch Polymer Institute (DPI), P.O. Box 902, 5600 AX Eindhoven, The Netherlands
§Dipartimento di Chimica, Universita
̀
di Napoli Federico II, Via Cintia, 80126 Napoli, Italy
⊥Department of Chemistry, University of Manitoba, Winnipeg, MB R3T 2N2, Canada
S
* Supporting Information
ABSTRACT: The productivity of a number of bis-
(phenoxyamine)Zr(IV)-based catalysts (bis(phenoxyamine)
=
N,N′-bis(3-R1 -5-R2 -2-O-C6 H2 CH2 )-N,N′-(R3 )2 -
(NCH2CH2N)) in ethene and propene polymerization was
evaluated for different R1/R2/R3 combinations. In previous
studies on this class we demonstrated that the cations that
form upon precatalyst activation (e.g., by methylalumoxane)
adopt a “dormant” mer-mer geometry, and an endothermic
isomerization to the active fac-fac geometry is the necessary
first step of the catalytic cycle. Herewith we report a clear correlation between catalyst activity and the DFT-calculated energy
difference ΔEi between the active and dormant state. The correlation only holds when the calculations are run on ion pairs,
which is less obvious than it may appear because the anion in these systems is not at the catalyst front. This finding provides a
comparatively simple and fast method to predict the activity of new catalysts of the same class.
Among the published approaches, some comprehensively
examined the 3-dimensional interaction between the monomer
and the catalyst active pocket;10a−c others focused on specific
supposedly key aspects, such as e.g. the energy of ion pair
separation vs ligand cone angle.10d Moderate success has been
achieved in ranking catalysts with ligand frames featuring
limited structural diversity within prototypical classes. However,
it is fair to admit that the picture remains poorly defined as
soon as one looks at novel structures, particularly in the rapidly
growing and widely differentiated area of “post-metallocene”
catalysis.12,13
An interesting case is that of [bis(phenoxyamine)]Zr(IV)-
based catalysts (bis(phenoxyamine) = N,N′-bis(3-R1-5-R2-2-O-
C6H2CH2)-N,N′-(R3)2-(NCH2CH2N), Scheme 1; short nota-
tion, ONNO),14 for which we recently elucidated an unusual
polymerization mechanism (Figure 1 and Scheme 2).13 In spite
of the tetradentate nature of the ligand, which may suggest a
certain stereorigidity, the first step entails the endothermic
rearrangement of the Zr−alkyl cation from the mer-mer
conformation C2x (the lowest energy isomer for an [ONNO]-
Zr(R)(□)+ species; □ = coordination vacancy) to the fac-fac
INTRODUCTION
■
The history of olefin polymerization is tightly liaised with
serendipity since the very beginning, with the polyethylene
breakthroughs in 1935 and 1953.1 Methylaluminoxane (MAO,
the preferred activator for new-generation catalysts2) represents
a more recent but similarly paradigmatic case of fortuitous
discovery.3 Last but not least, the vast majority of olefin
polymerization catalysts have been found by trial-and-error, an
approach now boosted by the introduction of high throughput
experimentation/screening (HTE/HTS) tools and methods,4,5
particularly in industrial discovery programs.6,7
Catalyst design has often been claimed as a feasible option,8
but the obvious prerequisite is a reliable model of structure/
properties relationships.9 The discovery of group 4 metallocene
systems offered the first realistic modeling opportunities for
well-defined active species,10 and since then several important
aspects of catalyst performance (e.g., the enantio- and
regioselectivity in the insertion of prochiral monomers, the
relative rates of the accessible chain transfer pathways, and
evento some extentthe rate balance of the latter with chain
propagation) have been more or less successfully quantified
with computational methods (primarily density functional
theory (DFT) based11). In spite of this remarkable progress,
however, understanding of the quintessential relation, that is
the one between catalyst structure and activity, remains elusive.
Received: February 19, 2012
Revised: April 16, 2012
© XXXX American Chemical Society
A
dx.doi.org/10.1021/ma300343c | Macromolecules XXXX, XXX, XXX−XXX

Macromolecules
Article
to predict the activity of new catalysts within this family, thus
opening the door to in silico catalyst prescreening.
Scheme 1. Numbering of Complexes
RESULTS AND DISCUSSION
■
The precatalyst structures considered for this investigation are
shown in Scheme 1 (P1−P10). They were chosen so as to
ensure a wide steric and electronic diversity for the R1/R2
substituents and some variation on R3 as well. All neutral
[ONNO]ZrX2 precursor complexes had X = benzyl (previous
studies demonstrated, not surprisingly, that catalytic activity is
basically independent of the nature of X when X = alkyl15).
When exploring quantitative relationships between catalyst
structure and performance, one should use reliable and truly
comparable data. In order to minimize effects possibly arising
from different materials and protocols, we decided to test ex
novo all catalysts, even where previous literature results were
available. The precatalysts were activated with MAO/2,6-di-tert-
butylphenol (TBP) mixtures17 in toluene, and tested in ethene
and propene polymerization at 25 °C under similar conditions,
as described in the Experimental Section. The observed
productivities for polyethylene (RE) and polypropylene (RP),
expressed as kg (polymer) mol (Zr)−1 [CnH2n]−1 h−1, are
summarized in Table 1; in general, the agreement with pre-
Table 1. Average Productivities for Polyethylene (RE) and
a
Polypropylene (RP) in Toluene at 25 °C for Catalysts
Obtained from Precursors P1−P10 upon Activation with
MAO/TBP
PE: RE
PP: RP
PE: RE
PP: RP
Figure 1. DFT-calculated propene insertion profile for a model of
[ONNO]Zr(IV) catalyst (R1 = CMe2Ph, R2 = R3 = Me).
P1
P2
P3
P4
P5
2.7 × 102
4.0 × 102
4.6 × 103
2.0
6.2
2.5
76
P6
P7
P8
P9
P10
5.6 × 105
5.8 × 104
4.6 × 102
4.2 × 104
2.9 × 103
1.2 × 104
6.2 × 102
1.0
2.6 × 102
33
Scheme 2. Geometrical Isomers of [(ONNO)ZrX2]
Complexes
0.2
2.4 × 103
7.4
a
In kg (polymer) mol (Zr)−1 [CnH2n]−1 h−1.
existing data is good.14f In most cases, the catalysts featured
moderate activity (RE < 4000; RP < 100), e.g., compared with
typical zirconocenes under similar conditions.18 Notable
exceptions are those derived from precursors P6, P7, and P9;
the former, in particular, is more active than many competent
zirconocenes.
Altogether, the R values in Table 1 span a range of more than
6 orders of magnitude and should therefore represent a valid set
to benchmark models of structure−activity relationship. At first
sight it is not easy to come up with simple qualitative
interpretations. For polyethylene, RE clearly tends to increase
with increasing size of the frontal R1 substituents. A
conceptually similar trend was reported for metallocenes and
traced to a progressive loosening of the ion pairing;10b in the
present case, however, the same interpretation looks ques-
tionable because we demonstrated that the anion is not at the
catalyst front.11,13 For polypropylene, in turn, no clear
relationship of RP with R1 size can be seen.19 For both
monomers, as already noted, the productivity has a dramatic
maximum for R1 = 9-anthracenyl (P6), a large but flat moiety
that in the proper orientation leaves the catalyst “mouth” as
open as with R1 = methyl (P1) (for a visual comparison see
Figure 2). An interesting trend was observed when replacing
methyl groups with practically isosteric Cl atoms as R1 and/or
R2: substitution at R1 (P1 vs P9) resulted in a large increase of
productivity (from 270 to 42 000 for RE; from 6.2 to 260 for
conformation C2, passing through an intermediate fac-mer
conformation C1.15 C2 in turn is the most stable isomer
whenever a Lewis base, and in particular the olefin, saturates the
Zr coordination sphere15 and can undergo monomer insertion
according to the classical Cossee−Arlman scheme.16 Polymer
microstructure analysis14d and DFT studies of the reaction
path13b,14d consistently point to it as the active species;
microstructure analysis also rules out propagation at the C1
conformation.14f
In this paper, we disclose a simple and quantitative
correlation between catalyst activity and the DFT-computed
energy difference ΔEi between ion pairs featuring C2x and C2
cations, whichwe believecan be used as a convenient tool
B
dx.doi.org/10.1021/ma300343c | Macromolecules XXXX, XXX, XXX−XXX

Macromolecules
Article
Figure 2. Optimized geometries of ion pairs M1_C2x, M1_C2, M6_C2x, and M6_C2. Hydrogen atoms are omitted for clarity.
RP); a smaller but still very significant effect was associated with
substitution at R2 (P1 vs P10), ending up with RE = 2900 and
RP = 33; changing methyl for Cl at both R1 and R2 (P7) yielded
the largest productivities in the subset (RE = 58 000, RP = 620).
A truly surprising and puzzling effect was observed upon
replacing R3 = methyl with the slightly larger ethyl group; this
caused a vertical drop of productivity (by roughly 3 orders of
magnitude; compare P3−P4 and P7−P8). Although not
unprecedented in coordination catalysis, it is rather uncommon
that a fragment remote from the active pocket impacts so
strongly on the activity.
catalystEXSY cannot be used because the peaks broaden and
coalesce already at low temperature. On the other hand,
assuming that the mechanism of Figure 1 is general for
[ONNO]Zr(IV)-based catalysts (which based on current
knowledge looks like a plausible hypothesis), estimating the
relative stabilities of C2x and C2 for a given ligand frame by
means of DFT can represent a convenient predictive tool in
alternative to experiment.
With this in mind, we studied computationally model catalyst
ion pairs M1−M10 with the same ancillary ligand frames as
precatalysts P1−P10. In all cases, the cation featured an
isobutyl group (to mimic a growing polypropylene chain) and a
Searching for a key to interpret this picture, it is worth noting
that according to the polymerization mechanism of Figure 1 the
largest contribution to the activation enthalpy of the polymer-
ization comes from the initial isomerization of C2x to C2, even
though the rate-determining step can be monomer insertion (as
seems to be the case for propene).13b Therefore, it seemed a
reasonable working hypothesis that decreasing the enthalpy
difference between C2x and C2 should speed up the overall
reaction, and in this work we test this hypothesis.
15
−
coordination vacancy, and the anion was MeB(C6F5)3 . The
latter is convenient from the computational standpoint; since
the anion is always in the second coordination sphere, it is
representative of other competent counterions (including the
ill-defined anion of MAO) for the catalysts of interest, as found
in previous experimental studies.15 The geometries of both fac-
fac (C2) and mer-mer (C2x) isomeric cations were optimized,
with the anion in the positions identified in previous studies
(Figure 2).13,15 Not surprisingly, in the fac-fac geometry in
general the isobutyl group turned out to interact with the Zr
center through a β-agostic interaction; this is not feasible in the
mer-mer geometry, where the vacant site is trans to the isobutyl
moiety. The structures of all isolated MC1−MC10 cations were
optimized as well, starting from the respective ion pair
structures and removing the anion. The main results of the
calculations are summarized in Tables 2 and 3. Table 2, in
particular, lists the values of the four Cα−Zr−X angles (where
Cα is the α carbon of the isobutyl and X represents a N or O
atom of the ligand) for the C2x isomers of all cations; these
values represent the only appreciable differences between the
1
We recently reported a H-EXSY NMR methodology to
measure this isomerization rate.13a,15 The drawback of this
procedure is that it requires the synthesis and isolation of the
ion pair, which is a delicate job due to the high reactivity of this
moiety. The EXSY study was undoubtedly very useful to get
insight into the C2x−C2 equilibrium, including the activation
parameters, but it cannot be proposed as a practical way to
predict catalyst activity because once a given ion pair is actually
available testing it in catalysis is definitely faster than carrying
out the EXSY measurements. Moreover, the spectroscopic
approach has a limited window of applicability: in particular, if
the isomerization is very fastas it should be for a “good”
C
dx.doi.org/10.1021/ma300343c | Macromolecules XXXX, XXX, XXX−XXX

Macromolecules
Article
The energy differences between the C2 and C2x isomers for
the M1−M10 ion pairs (ΔEi) as well as the isolated MC1−
MC10 cations (ΔEi,c) are collected in Table 3.
Plots of ln RX (X = E or P; Table 1) vs ΔEi (Table 3) for ion
pairs M1−M10 (Figure 3) show a nice linear correlation. This
Table 2. Values of the Cα−Zr−X Angles (in deg) in the
DFT-Optimized Geometries of the Cations in mer-mer
Geometry for Ion Pairs (“M”) and Isolated Cations (“Mc”)
(See Text)
X = O1
X = N1
X = N2
X = O2
M1_C2x
Mc1_C2x
M2_C2x
Mc2_C2x
M3_C2x
Mc3_C2x
M4_C2x
Mc4_C2x
M6_C2x
Mc6_C2x
M7_C2x
Mc7_C2x
M8_C2x
Mc8_C2x
M9_C2x
Mc9_C2x
M10_C2x
Mc10_C2x
98.7
101.8
101.2
103.7
99.8
105.1
109.8
105.2
107.8
105.2
108.6
107.9
109.2
106.2
104.3
104.5
110.2
109.6
110.0
104.3
109.9
104.4
109.8
97.7
100.2
97.6
102.1
107.1
104.6
107.6
104.1
107.9
106.0
107.9
103.6
110.0
100.4
106.7
107.0
107.1
96.7
99.3
96.6
103.2
100.7
101.6
99.7
99.1
97.0
98.2
97.3
104.4
96.7
99.6
96.2
100.2
98.2
99.8
100.4
100.2
100.5
100.2
100.9
100.3
101.5
97.1
■
Figure 3. Plot of catalyst productivity for polyethylene (ln RE, ) and
101.6
97.2
107.1
96.4
●
polypropylene (ln RP, ) for catalysts derived from precursors P1−
P10 (Table 1) vs ΔEi for ion pairs M1−M10 (Table 3). The straight
lines through the data points represent the best linear fits, with
correlation coefficients of 0.952 and 0.938, respectively.
101.7
107.0
Table 3. DFT-Calculated Energy Differences between the C2
and C2x Isomers for the Studied Complexes
supports our hypothesis that for the considered [ONNO]Zr-
(IV)-based catalysts the activity is mainly governed by the
relative stabilities of the C2x (dormant) and C2 (active)
isomers. As already noted before, from this one should not
conclude that the C2x−C2 isomerization is the rate-
determining step, but rather that the energetics of the
subsequent steps in the reaction path (i.e., olefin uptake and
insertion, see Figure 1) are rather similar for all catalysts, i.e.,
relatively insensitive to the specific substitution pattern of the
ancillary ligand, at least as far as the impact on the reaction rate
is considered. The most active catalysts in the set (namely, P6
and P7) feature values of ΔEi < 7 kcal mol−1, whereas at the
other extreme catalysts with very poor activity (namely P1, P2,
P4, P8) all have ΔEi > 9.5 kcal mol−1. While the correlation is
not perfect, it accommodates all screened substituent effects,
both steric and electronic ones, independent of whether or not
a specific substituent is close to or far from the active pocket.
Of course, second-order effects are present which are not taken
into account by this simple scheme, as indicated by the minor
scattering of the data points in Figure 3.
It would be convenient if the relative stabilities of the isolated
cations followed those of the ion pairs; in such a case, the DFT
modeling would be much less demanding in terms of
computational time/resources. Unfortunately, plots of ln RX
vs ΔEi,c for MC1−MC10 show a large scatter and no significant
correlation (Figure 4), which indicates that the anion must be
included explicitly for a meaningful prediction of catalyst
activity. This may look somehow surprising, becauseas noted
repeatedlyfor these systems the anion is not at the catalyst
front; we conclude that ion pairing impacts on the complex
summation of interactions defining cation energetics to
different extents for different geometrical isomers.
ion pairs
ΔEi (kcal mol−1
)
isolated cations
ΔEi,c (kcal mol−1
)
M1
M2
M3
M4
M5
M6
M7
M8
M9
M10
9.73
10.08
9.17
Mc1
Mc2
Mc3
Mc4
Mc5
Mc6
Mc7
Mc8
Mc9
Mc10
8.37
10.28
5.38
12.31
10.25
5.87
12.37
10.34
12.26
5.54
6.90
9.52
5.10
7.05
5.79
9.48
8.00
computed geometries of the cations in the two cases (ion pairs
vs isolated cations). In the absence of the anion, the [ONNO]
ligand in the mer-mer conformation appears to “relax”, and all
four Cα−Zr−X angles (in the range between 98° and 110°)
systematically increase by 2°−10°; this is an indication that the
ion pairing interaction, while being an important element of
stabilization, introduces some strain in the six-membered Zr−
O−C−C−C−N ring. We believe that this is the reason why,
when all the Cα−Zr−X angles are forced to approach 90°
following the coordination of a nucleophile (e.g., THF, an
olefin, but also the methyl group of the MeB(C6F5)3− anion) to
the sixth Zr coordination site, the mer-mer geometry is no
longer the most stable one, as found before;13,15 in fact, it
appears that the strain introduced by the interaction with the
nucleophile is not compensated by its binding energy, and
when the nucleophile does bind, the preferred structure has the
fac-fac geometry. Importantly, the effect of the anion on the
values of the Cα−Zr−X angles turned out to be vanishingly
small for structures M4 and M8, that is, when R3 = ethyl (rather
than methyl); this can be traced to the fact that in such cases
the anion, which is located at the back of the ligand frame, is
practically prevented from interacting with the cation due to
the larger steric demand of R3. We will come back to this later.
Two cases are specially paradigmatic in this respect. For the
M6 ion pairs (R1 = 9-anthracenyl) the value of ΔEi (= 5.87 kcal
mol−1) is the lowest in Table 3, whereas for the MC6 cations
that of ΔEi,c (= 12.26 kcal/mol) is one of the highest. This can
be understood on inspection of the DFT-optimized structures.
D
dx.doi.org/10.1021/ma300343c | Macromolecules XXXX, XXX, XXX−XXX

Macromolecules
Article
one. On the other hand, modulating this gap by means of a
rational design of the ancillary ligand remains difficult, as
discussed above, and even in case of success major drawbacks
can show up; productivity is only one of several properties of
interest, and other important features of catalyst performance
may vary in an undesired manner. The catalyst derived from
precursor P6 is once again a good example; with its low ΔEi it
is exceedingly active but yields polymers of very low average
molecular mass and in the case of polypropylene practically
atactic.
Until now, we have not yet identified R1/R2/R3 combina-
tions ending up with well-balanced properties. In spite of this,
the calculation of ΔEi is highly recommended before the actual
synthesis of a similar catalyst with a new combination of
substituents. If the calculated ΔEi is higher than 8 kcal/mol, the
catalyst is probably not worth trying, and valuable laboratory
time is saved.
■
Figure 4. Plot of catalyst productivity for polyethylene (ln RE, ) and
●
polypropylene (ln RP, ) for catalysts derived from precursors P1−
P10 (Table 1) vs ΔEi,c for isolated cations MC1−MC10 (Table 3). The
straight lines through the data points represent the best linear fits, with
correlation coefficients of 0.26 and 0.13, respectively.
Of course, changing other structural features of the ligand is
also an option, but then the possibility that the structure/
activity correlation changes in nature should be considered. For
the investigated [ONNO]Zr(IV)-R cations, the root of the
problem is that the two six-membered M-N-C-C-C-O rings
tend to adopt the inactive mer-mer geometry. The literature
demonstrates that proper modifications can be very beneficial;
in particular, (“salalen”)Ti(IV)-22 and bis(phenoxy ether)M-
(IV)-based catalysts23 with high activities and selectivities have
been disclosed. We are currently looking at these and other
formally related systems with [OYYO] or [YYYY] ligands (Y =
heteroatom) to find out if the mechanism elucidated for
[ONNO]M(IV) species is more general. We believe that this
can be the case as long as the role of inner-sphere ion pairs
(ISIP)24 in the catalytic cycle is marginal; as a matter of fact,
whenever an ISIP is the most stable product of precatalyst
activation, as is the case with metallocenes25,26 and a number of
“post-metallocene”12 catalysts, the activity is crucially and
directly affected by the cation−anion interaction, which makes
general interpretation schemes less likely.
In M6_C2x (Figure 2) the distance between the Zr atom and
the nearest F atom of the anion is 3.94 Å (to be compared, e.g.,
with 3.23 Å for M1_C2x, with R1 = methyl); this is due to the
orientation of the anthracenyl moiety, which points directly
toward the anion, pushing it away, weakening the anion−cation
interaction, and ultimately destabilizing the ion pair. In M6_C2,
on the other hand, the anion and the anthracenyl do not
interact at all. The net result is a peculiarly low value of ΔEi for
this complex. For the isolated cations the situation is reversed;
in fact, the MC6_C2x model species turned out to be very
stable because the two nearly parallel anthracenyl moieties do
not cause the steric congestion typical of complexes with close-
to-spherical R1’s. This justifies the large value of ΔEi,c.
A much more subtle case is that of M4 and M8, with R3 =
ethyl. Both feature large ΔEi values, which correlates well with
the very low productivities of the catalysts based on them.
Their methyl-substituted homologues M3 and M7, in turn,
have lower ΔEi values and, consistently, yield catalysts with
higher productivities. Why this is so, however, is hard to
understand by visual inspection of the optimized structures;20
the substituent effects here are more diffuse, which makes them
difficult to identify.
EXPERIMENTAL SECTION
■
All experiments were performed under an atmosphere of dry argon or
in a argon-filled glovebox. Solvents were purchased by Aldrich as
HPLC grade. Toluene and diethyl ether were previously purified in a
MBraun SPS unit. n-Hexane was dried on metallic Na by standard
procedures. Tetrabenzylzirconium was synthesized according to the
literature.27 NMR samples of complexes were prepared in oven-dried
J-Young NMR tubes. NMR spectra for ligands and metal complexes
were recorded on Bruker Avance DRX 400 spectrometer and
referenced to residual protons in benzene-d6 (δ 7.15) and chloro-
form-d1 (δ 7.26) and to 13C chemical shift of benzene (δ 128.0). The
following abbreviations were used for describing NMR multiplicities: s,
singlet; d, doublet; t, triplet; q, quartet; m, multiplet; br, broad signal.
MAO (10% w/w solution in toluene) was purchased from Chemtura.
1H NMR analysis revealed it to contain 38% of Al as “free” AlMe3.
TBP and 4-methyl-2-(1-methyl-1-phenylethyl)phenol were purchased
from Aldrich. Propene (polymerization grade) was purchased from
SON and used as received. Complexes P1,14a P2,14a P3,14c P5,14f
P6,14c and P728 were synthesized according to the literature.
CONCLUSIONS
■
On the basis of the results of the previous section, we conclude
that the correlation between ΔEi and activity for the
investigated catalyst class holds and can be used as a simple
and convenient predictive tool, regardless of whether or not a
detailed structural interpretation of the DFT-calculated ΔEi
value can be found. This means that, although the approach was
inspired by a precise element of mechanistic understanding, the
search for improved catalysts through a systematic exploration
of the substituent effects remains largely empirical (albeit
carried out in silico). What differentiates our model from a
typical QSAR one21 is that one single parameter with a well-
defined molecular kinetic meaning (namely, ΔEi) is adequate to
describe the property of interest, whereas less fortunate cases
require a “black-box” treatment with complex “cocktails” of
descriptors whose individual relationships with the targeted
variable(s) can be obscure. In the investigated case, in fact, the
correlation between catalyst structure and activity is dominated
by one key element, that is, the energy gap between a dormant
species representing the catalyst resting state and the active
Synthesis of the (ONNO) Ligands. The tetradentate [ONNO]
ligands L1−L10 were prepared by reacting N,N′-dimethylethylenedi-
amine or N,N′-diethylethylenediamine with formaldehyde and the
appropriate phenol, as described in the literature.14 10 mmol of N,N′-
dimethylethylenediamine, 10 mmol of formaldehyde (37% solution in
water), and 10 mmol of the appropriate phenol were added to 30 mL
of methanol and kept under reflux for 3 h. The precipitated product
was filtered off, washed with cold methanol, and dried in an oven, at 65
E
dx.doi.org/10.1021/ma300343c | Macromolecules XXXX, XXX, XXX−XXX

Macromolecules
Article
°C under vacuum for 3 h. A second crop of product could be obtained
by keeping the methanol solution in a fridge for several days. Total
yields were ca. 60%.
129.3 (C−H), 129.4 (C−H), 129.7 (C−H), 145.5 (Cipso), 154.4
(Cipso). Anal. Calcd for C34H36Cl4N2O2Zr: C, 55.28; H, 5.05; N, 3.79.
Found: C, 55.38; H, 5.25; N, 3.71.
1
L4 (R1 = Cumyl, R2 = Me, R3 = Et). 1H NMR (400 MHz, C6D6, 300
K): δ 0.48 (t, 6H, N−CH2−CH3), 1.84 (s, 12H, C(CH3)2), 1.90 (m,
4H, N−CH2−CH3), 1.93 (s, 4H, NCH2), 2.28 (s, 6H, CH3), 3.05 (s,
4H, NCH2Ar), 6.71 (d, 2H, Ar), 7.08 (t, 2H, Ar), 7.18 (t, 4H, Ar),
7.29 (d, 2H, Ar), 7.37 (d, 4H, Ar), 10.46 (bs, 2H, OH). 13C NMR
(100.62 MHz, C6D6, 300 K): δ 11.1 (N−CH2−CH3), 21.1 (CH3),
29.5 (C(CH3)2), 42.2 (C(CH3)2), 47.5 (N−CH2−CH3), 50.0
(NCH2), 58.3 (NCH2Ar), 122.0 (Cipso), 124.9 (C−H), 126.0 (C−
H), 126.9 (Cipso), 127.4 (C−H), 127.8 (C−H), 128.1 (C−H), 135.6
(Cipso), 151.5 (Cipso), 154.4 (Cipso). Anal. Calcd for C40H52N2O2: C,
81.04; H, 8.85; N, 4.73. Found: C, 81.22; H, 8.95; N, 4.66.
Complex P9. H NMR (400 MHz, C6D6, 300 K): δ 0.79 (d, 2H,
NCH2), 1.82 (s, 6H, CH3), 1.99 (s, 6H, NCH3), 2.02 (d, 2H, ZrCH2),
2.25 (d, 2H, NCH2Ar), 2.36 (d, 2H, NCH2), 2.40 (d, 2H, ZrCH2),
3.83 (d, 2H, NCH2Ar), 6.28 (d, 2H, Ar−H), 6.87 (t, 2H, Ar−H), 7.09
(t, 4H, Ar−H), 7.15 (d, 2H, Ar−H), 7.25 (d, 4H, Ar−H). 13C NMR
(100.62 MHz, C6D6, 300 K): δ 20.2 (CH3,), 44.6 (NCH3), 53.0
(NCH2), 62.4 (NCH2Ar), 63.2 (ZrCH2), 122.9 (Cipso), 123.0 (C−H),
126.6 (Cipso), 128.78 (Cipso), 129.1 (C−H), 129.3 (C−H), 129.6 (C−
H), 130.5 (C−H),145.7 (Cipso), 153.6 (Cipso). Anal. Calcd for
C32H34Cl2N2O2Zr: C, 59.98; H, 5.35; N, 4.37. Found: C, 60.08; H,
5.39; N, 4.34.
1
1
Complex P10. H NMR (400 MHz, C6D6, 300 K): δ 0.71 (d, 2H,
L8 (R1 = R2 = Cl, R3 = Et). H NMR (CDCl3, 400 MHz, 300 K): δ
NCH2), 1.6 (s, 6H, NCH3), 1.92 (d, 2H, NCH2), 2.12 (d, 2H,
NCH2Ar), 2.13 (d, 2H, ZrCH2), 2.20 (s, 6H, CH3), 2.24 (d, 2H,
ZrCH2), 3.70 (d, 2H, NCH2Ar), 6.62 (d, 2H, Ar−H), 6.86 (t, 2H, Ar-
H), 7.00 (d, 4H, Ar−H), 7.07 (t, 4H, Ar−H), 7.11 (d, 2H, Ar−H). 13C
NMR (100.62 MHz, C6D6, 300 K): δ 16.9 (CH3,), 44.2 (NCH3), 52.7
(NCH2), 62.6 (ZrCH2), 62.7 (NCH2Ar), 122.8 (C−H), 123.4 (Cipso),
126.0 (Cipso), 127.6 (C−H), 128.7 (C−H), 128.8 (C−H), 130.6 (C−
H), 146.0 (Cipso), 156.7 (Cipso). Anal. Calcd for C32H34Cl2N2O2Zr: C,
59.98; H, 5.35; N, 4.37. Found: C, 60.03; H, 5.36; N, 4.40.
Polymerization Runs. All ethene and propene homopolymeriza-
tion runs were carried out at 25 °C, in a 250 mL magnetically stirred
jacketed Pyrex reactor with two necks (one of which capped with a
silicone rubber septum, the other connected to a Schlenk manifold),
according to the following procedure, unless otherwise specified. The
reactor was charged under nitrogen with 150 mL of toluene (toluene
HPLC, Aldrich, previously purified in an MBraun SPS unit) containing
5.0 mL of MAO (Chemtura, 10% w/w solution in toluene) and 1.7 g
of 2,6-di-tert-butylphenol (Aldrich, 99%) and thermostated at 25 °C.
After 1 h, 5.0 mL of the liquid phase was syringed out and used to
dissolve the proper amount of precatalyst in a glass vial. The reactor
was then evacuated to remove nitrogen and saturated with the
monomer (SON, polymerization grade) at a partial pressure of 0.8−
1.6 bar for ethene and 2.0 bar for propene. The polymerization was
started by syringing in the catalyst solution, left to proceed at constant
monomer pressure for the appropriate time, and quenched with 5 mL
of methanol/HCl(aq, conc) (95/5 v/v). The polymer was coagulated
with excess methanol/HCl, filtered, washed with more methanol, and
vacuum-dried. Results of polymerization experiments are given in
Table 4 (ethene) and Table 5 (propene).
1.11 (s, 6H, N−CH2−CH3), 2.60 (q, 4H, N−CH2−CH3), 2.73 (s, 4H,
N−CH2), 3.73 (s, 4H, NCH2Ar), 6.85 (d, 2H, Ar), 7.27 (d, 2H, Ar),
10.8 (br, 2H, OH). 13C NMR (100.62 MHz, CDCl3, 300 K): δ 11.1
(N−CH2−CH3), 47.9 (N−CH2−CH3), 50.3 (NCH2), 57.4
(NCH2Ar), 121.7 (Cipso), 123.7 (Cipso), 126.8 (C−H), 128.9 (C−H),
152.9 (Cipso). Anal. Calcd for C20H24Cl4N2O2: C, 51.52; H, 5.19; N,
6.01. Found: C, 51.54; H, 5.26; N, 6.00.
L9 (R1 = Cl, R2 = R3 = Me). 1H NMR (CDCl3, 400 MHz, 300 K): δ
2.23 (s, 6H, CH3), 2.32 (s, 6H, N−CH3), 2.73 (s, 4H, N−CH2), 3.69
(s, 4H, NCH2Ar), 6.68 (d, 2H, Ar), 7.08 (d, 2H, Ar), 10.6 (br, 2H,
OH). 13C NMR (100.62 MHz, CDCl3, 300 K): δ 20.3 (CH3), 41.9
(N−CH3), 54.3 (NCH2), 61.4 (NCH2Ar), 120.5 (Cipso), 122.3 (Cipso),
127.7 (C−H), 129.1 (Cipso), 129.7 (C−H), 154.1 (Cipso). Anal. Calcd
for C18H22Cl2N2O2: C, 58.54; H, 6.00; N, 7.59. Found: C, 58.50; H,
6.05; N, 7.55.
L10 (R1 = Me, R2 = Cl, R3 = Me). 1H NMR (CDCl3, 400 MHz, 300
K): δ 2.20 (s, 6H, CH3), 2.29 (s, 6H, N−CH3), 2.66 (s, 4H, N−CH2),
3.65 (s, 4H, NCH2Ar), 6.80 (d, 2H, Ar), 7.04 (d, 2H, Ar), 10.6 (br,
2H, OH). 13C NMR (100.62 MHz, CDCl3, 300 K): δ 15.7 (CH3),
41.7 (N−CH3), 53.9 (NCH2), 61.4 (NCH2Ar), 122.2 (Cipso), 123.2
(Cipso), 125.8 (C−H), 127.1 (Cipso), 129.8 (C−H), 154.5 (Cipso). Anal.
Calcd for C18H22Cl2N2O2: C, 58.54; H, 6.00; N, 7.59. Found: C,
58.55; H, 6.03; N, 7.60.
Synthesis of Zr Complexes. The synthesis was done according to
the literature procedure.14 5 mmol of ligand was weighted in a Schlenk
flask and dissolved in 10 mL of dry toluene (heating the mixture helps
the dissolution of the compound). The resulting solution was added to
another Schlenk flask containing a solution of 5 mmol of Zr(Benzyl)4
in 10 mL of the same solvent, under an argon atmosphere. The
mixture was kept at 65 °C for 2 h, and then the solvent was evacuated
to give the product as a pale yellow powder. Recrystallization from
diethyl ether afforded the product in 60% yield.
Table 4. Ethene Polymerization Results
Complex P4. 1H NMR (400 MHz, C6D6, 300 K): δ 0.20 (t, 6H, N−
CH2−CH3), 1.35 (d, 2H, N−CH2), 1.84 (s, 6H, CH3), 1.94 (m, 2H,
N−CH2−CH3), 2.02 (s, 6H, CH3), 2.10 (m, 2H, overlapped, N−
CH2−CH3), 2.10 (d, 2H, overlapped, NCH2), 2.19 (d, 2H, Zr−CH2),
2.20 (s, 6H, CH3), 2.71 (d, 2H, Zr−CH2), 2.76 (d, 2H, N−CH2Ar),
3.20 (d, 2H, N−CH2Ar), 6.43 (d, J = 2.3 Hz, 2H, Ar), 6.79 (t, 2H, Ar),
6.95 (d, 4H, Ar), 7.08 (t, 2H, Ar), 7.09 (t, 4H, Ar), 7.23 (t, 4H, Ar),
7.34 (d, 2H, Ar), 7.50 (d, 4H, Ar). 13C NMR (100.62 MHz, C6D6, 300
K): δ 4.3 (CH3), 21.0 (CH3), 29.2 (CH3), 32.6 (CH3), 42.7
(C(CH3)2), 43.4 (N−CH2), 44.6 (NCH2), 56.8 (NCH2Ar), 72.1
(Zr−CH2), 120.9 (C−H), 125.5 (C-H), 125.9 (Cipso), 126.5 (C−H),
126.8 (C-H), 127.1 (Cipso), 127.9 (C−H), 128.1 (C−H), 129.6 (C−
H), 129.9 (C−H), 137.0 (Cipso), 149.8 (Cipso), 151.6 (Cipso), 156.9
(Cipso). Anal. Calcd for C52H60N2O2Zr: C, 74.68; H, 7.23; N, 3.35.
Found: C, 74.60; H, 7.28; N, 3.30.
[C2H4]
(M)
[Zr]
(μmol)
yield
(g)
RE (kgPE/
(molcat h [C2H4]))
Cat.
tp (h)
P1
P2
P3
P4
P5
P6
P7
P8
P9
P10
0.17
0.17
0.17
0.21
0.17
0.11
0.21
0.21
0.21
0.21
38
0.17
0.05
0.25
1.5
0.3
270
400
42
0.14
1.0
5.1
30
4600
2
0.02
0.19
0.56
0.52
0.35
0.89
0.69
3.8
0.135
0.51
9
0.12
0.067
0.083
0.5
2400
560000
58000
460
1.2
4.6
0.083
0.25
42000
2900
Computational Details. Density functional calculation were
performed with the Turbomole program29 (version 5.8) in
combination with the OPTIMIZE routine of Baker and co-workers.30
All geometries were fully optimized at the restricted RI31-BP8632 level,
using the SV(P)33 basis set (small-core pseudopotential on Zr34). The
cations in these systems have well-defined geometries, but for the ion
pairs the movement of the anion relative to the cation is very soft. A
number of plausible starting geometries were explored for the basic
system, and the “best” optimized structures were then used to
Complex P8. 1H NMR (400 MHz, C6D6, 300 K): δ 0.28 (t, 6H, N−
CH2−CH3), 1.27 (d, 2H, N−CH2), 1.91 (d, 2H, N−CH2), 1.98 (d,
2H, Zr−CH2), 2.34 (d, 2H, Zr−CH2), 2.35 (m, 2H, N−CH2−CH3),
2.53 (d, 2H, N−CH2Ar), 2.78 (m, 2H, N−CH2−CH3), 3.33 (d, 2H,
N−CH2Ar), 6.47 (d, 2H, Ar), 6.86 (t, 2H, Ar), 7.07 (t, 4H, Ar), 7.14
(d, 4H, Ar), 7.30 (d, 2H, Ar). 13C NMR (100.62 MHz, C6D6, 300 K):
δ 5.2 (CH3), 44.9 (N−CH2), 45.7 (NCH2), 56.5 (NCH2Ar), 64.1
(Zr−CH2), 123.2 (C−H), 123.5 (Cipso), 124.0 (Cipso), 128.8 (C−H),
F
dx.doi.org/10.1021/ma300343c | Macromolecules XXXX, XXX, XXX−XXX

Macromolecules
Article
(4) High Throughput Screening in Catalysis; Hagemeyer, A., Strasser,
P. A. F., Jr., Eds.; Wiley-VCH: Weinheim, 2004.
(5) Boussie, T. R.; Diamond, G. M.; Goh, C.; Hall, K. A.; LaPointe,
A. M.; Leclerc, M.; Lund, C.; Murphy, V.; Shoemaker, J. A. W.; Tracht,
U.; Turner, H.; Zhang, J.; Uno, T.; Rosen, R. K.; Stevens, J. C. J. Am.
Chem. Soc. 2003, 125, 4306.
(6) (a) Frazier, K. A.; Boone, H. W.; Vosejpka, P. C.; Stevens J. C.
Dow Global Technologies Inc., US 6,953,764, 2005. (b) Boussie, T.
R.; Diamond, G. M.; Goh, C.; Hall, K. A.; LaPointe, A. M.; Leclerc,
M.; 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. Arriola, D. J.; Carnahan, E. M.; Hustad,
P. D.; Kuhlman, R. L.; Wenzel, T. T. Science 2006, 312, 714.
(7) Busico, V.; Pellechia, R.; Cutillo, F.; Cipullo, R. Macromol. Rapid
Commun. 2009, 30, 1697 and references within.
Table 5. Propene Polymerization Results
[C3H6]
(M)
[Zr]
(μmol)
yield
(g)
RP (kgPP/
Cat.
tp (h)
(molcat h [C3H6]))
P1
P2
P3
P4
P5
P6
P7
P8
P9
P10
1.32
37
63
9
3.5
2
1.1
6.2
2.5
0.42
0.45
0.05
0.35
3.3
0.5
4
76
50
18
3.5
4.6
46
37
37
0.2
2
7.4
0.08
0.5
4
8900
600
1
1.9
0.25
2.4
2
260
33
2
3.2
(8) For examples of in silico catalyst design, see: (a) Wondimagegn,
T.; Wang, D. Q.; Razavi, A.; Ziegler, T. Organometallics 2009, 28,
1383. (b) Wondimagegn, T.; Wang, D.; Razavi, A.; Ziegler, T.
Organometallics 2008, 27, 6434.
(9) (a) Gibson, V. C.; Spitzmesser, S. K. Chem. Rev. 2003, 103, 283.
(b) Britovsek, G. J. P.; Gibson, V. C.; Wass, D. F. Angew. Chem., Int.
Ed. 1999, 38, 428. (c) Busico, V. Macromol. Chem. Phys. 2007, 208, 26
and references therein. (d) Matsugi, T.; Fujita, T. Chem. Soc. Rev.
2008, 37, 1264. (e) Corradini, P.; Guerra, G.; Cavallo, L. Acc. Chem.
Res. 2004, 37, 231. (f) Coates, G. W. Chem. Rev. 2000, 100, 1223.
construct starting geometries for the substituted systems. These
systems, in particular the ion pairs, are so large that calculation of the
Hessian is impractical; with computer times of weeks this would also
defeat the purpose of using the calculations as an efficient tool for
prediction of catalyst performance. Finally, the ion pairs are so “floppy”
that thermal corrections arising from a vibrational analyses would not
be very meaningful. Therefore, no vibrational analyses were carried for
any of the substituted systems, and energy values cited in the text are
pure electronic energies.
Single-point solvent corrections were calculated using the
conductor-like screening model (COSMO)35 with ε = 2.37 to
model an nonpolar solvent (e.g., toluene). All energies mentioned in
the text include this solvation correction, unless noted otherwise. To
check for the sensitivity of the results to details of the calculation
method, we also carried out single-point calculations with a larger basis
set (def-TZVP36) and calculated the nonelectrostatic solvation terms
separately.37 These results are discussed in the Supporting
Information, where we conclude that the simple recipe described
here is preferable as a prescreening method.
(g) Mohring, P. C.; Coville, N. J. Coord. Chem. Rev. 2006, 250, 18.
̈
(h) Mohring, P. C.; Coville, N. J. J. Mol. Catal. 1992, 77, 41.
̈
(10) (a) Cruz, V. L.; Ramos, J.; Martinez, S.; Munoz-Escalona, A.;
̃
Lafuente, P.; Pena, B.; Martinez-Salazar, J. Polymer 2004, 45, 2061.
̃
(b) Cruz, V. L.; Ramos, J.; Martinez, S.; Munoz-Escalona, A.;
Martinez-Salazar, J. Organometallics 2005, 24, 5095. (c) Cruz, V. L.;
̃
̀ ̀
Martinez, S.; Martinez-Salazar, J.; Polo-Ceron, D.; Gomez-Ruiz, S.;
Fajardo, M.; Prashar, S. Polymer 2007, 48, 4663. (d) Manz, T. A.;
Phomphrai, K.; Medvedev, G.; Krishnamurthy, B. B.; Sharma, S.; Haq,
J.; Novstrup, K. A.; Thomson, K. T.; Delgass, W. N.; Caruthers, J. M.;
Abu-Omar, M. M. J. Am. Chem. Soc. 2007, 129, 3776.
Initial geometries and reasonable starting Hessians were obtained
from PM3 computations with the Spartan package from Wavefunction
Inc.38
(11) For a recent review, see: Srebro, M.; Michalak, A.
Computational Modeling of Polymerization Catalysts. In Handbook
of Transition Metal Polymerization Catalysis; Hoff, R., Mathers, R. T.,
Eds.; Wiley: Hoboken, NJ, 2010.
(12) (a) Zuccaccia, C.; Busico, V.; Cipullo, R.; Talarico, G.; Froese,
R. D. J.; Vosejpka, P. C.; Hustad, P. D.; Macchioni, A. Organometallics
2009, 28, 5445. (b) Froese, R. D. J.; Hustad, P. D.; Kuhlman, R. L.;
Wenzel, T. T. J. Am. Chem. Soc. 2007, 129, 7831.
(13) (a) Ciancaleoni, G.; Fraldi, N.; Budzelaar, P. H. M.; Busico, V.;
Macchioni, A. Dalton Trans. 2009, 8824. (b) Ciancaleoni, G.; Fraldi,
N.; Budzelaar, P. H. M.; Busico, V.; Cipullo, R.; Macchioni, A. J. Am.
Chem. Soc. 2010, 132, 13651.
(14) (a) Tshuva, E. T.; Goldberg, I.; Kol, M. J. Am. Chem. Soc. 2000,
122, 10706. (b) Busico, V.; Cipullo, R.; Friederichs, N.; Ronca, S.;
Togrou, M. Macromolecules 2003, 36, 3806. (c) Busico, V.; Cipullo, R.;
Friederichs, N.; Ronca, S.; Talarico, G.; Togrou, M.; Wang, B.
Macromolecules 2004, 37, 8201. (d) Segal, S.; Goldberg, I.; Kol, M.
Organometallics 2005, 24 (2), 200. (e) Yeori, A.; Groysman, S.;
Goldberg, I.; Kol, M. Inorg. Chem. 2005, 44 (13), 4466. (f) Busico, V.;
Cipullo, R.; Pellecchia, R.; Ronca, S.; Roviello, G.; Talarico, G. Proc.
Natl. Acad. Sci. U. S. A. 2006, 103, 15321.
ASSOCIATED CONTENT
* Supporting Information
■
S
Summary and discussion of TZVP results and nonelectrostatic
solvent corrections; DFT-optimized structures of all species
studied as a zipped archive of pdb files usable with various
molecule viewers. This material is available free of charge via
the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
*E-mail: budzelaa@cc.umanitoba.ca.
■
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
This work is part of the Research Programme of the Dutch
Polymer Institute, Eindhoven, The Netherlands, Project #633.
■
REFERENCES
(15) Ciancaleoni, G.; Fraldi, N.; Busico, V.; Macchioni, A.; Budzelaar,
P. H. M. Organometallics 2011, 30, 3096.
(16) (a) Cossee, P. J. Catal. 1964, 3, 80. (b) Arlman, E. J.; Cossee, P.
J. Catal. 1964, 3, 99. (c) Cossee, P. Rec. Trav. Chim. Pays-Bas 1966, 85,
1151.
■
(1) (a) McMillan, F. M. The Chain Straighteners; MacMillan
Publishing Co.: London, 1979. (b) Malpass, D. B. Introduction to
Industrial Polyethylene: Properties, Catalysts, Processes; Scrivener
Publishing: Salem, 2010.
(17) (a) Busico, V.; Cipullo, R.; Cutillo, F.; Friederichs, N.; Ronca,
S.; Wang, B. J. Am. Chem. Soc. 2003, 125 (41), 12402.
(b) Rouholahnejad, F.; Mathis, D.; Chen, P. Organometallics 2010,
29, 294.
(2) Chen, E. Y.-X.; Marks, T. J. Chem. Rev. 2000, 100, 1391.
(3) (a) Andresen, A.; Cordes, H. G.; Herwig, J.; Kaminsky, W.; Merk,
A.; Mottweiler, R.; Pein, J.; Sinn, H.; Vollmer, H.-J. Angew. Chem., Int.
Ed. Engl. 1976, 15, 630. (b) Sinn, H.; Kaminsky, W.; Vollmer, H.-J.;
Woldt, R. Angew. Chem., Int. Ed. Engl. 1980, 19, 390.
(18) See e.g.: Alt, H. G.; Koppl, A. Chem. Rev. 2000, 100, 1205.
̈
G
dx.doi.org/10.1021/ma300343c | Macromolecules XXXX, XXX, XXX−XXX

Macromolecules
Article
(19) This can be due to an intrinsic different behavior or to the fact
that measuring the productivity for poorly active catalyst is more
difficult, and the experimental errors can be higher.
(20) Comparing M3_C2 with M4_C2 and M7_C2 with M8_C2,
they have the same anion−cation distance, and even if R1 and R3 are
slightly closer in the case of the ethyl group, there is not enough steric
congestion to justify the increase of ΔEi. On the other hand, M4_C2x
and M8_C2x do not show any evident stabilizing element.
(21) Muir, R. M.; Hansch, C. Nature 1962, 194, 178.
(22) Press, K.; Cohen, A.; Goldberg, I.; Venditto, V.; Mazzeo, M.;
Kol, M. Angew. Chem., Int. Ed. 2011, 50, 3529.
(23) Boussie, T. R.; Brummer, O.; Diamond, G. M.; Goh, C.;
Lapointe, A. M.; Leclerc, M. K.; Shoemaker, J. A. Symyx Technologies,
Inc., WO 03/091262 A1, 2003.
W.; Wong, M. W.; Gonzalez, C.; Pople, J. A. Gaussian 03, C.02;
Gaussian, Inc.: Wallingford, CT, 2004.
(38) Shao, Y.; Molnar, L. F.; Jung, Y.; Kussmann, J.; Ochsenfeld, C.;
Brown, S. T.; Gilbert, A. T. B.; Slipchenko, L. V.; Levchenko, S. V.;
O’Neill, D. P.; DiStasio, R. A., Jr.; Lochan, R. C.; Wang, T.; Beran, G. J.
O.; Besley, N. A.; Herbert, J. M.; Lin, C. Y.; Van Voorhis, T.; Chien, S.
H.; Sodt, A.; Steele, R. P.; Rassolov, V. A.; Maslen, P. E.; Korambath,
P. P.; Adamson, R. D.; Austin, B.; Baker, J.; Byrd, E. F. C.; Dachsel, H.;
Doerksen, R. J.; Dreuw, A.; Dunietz, B. D.; Dutoi, A. D.; Furlani, T. R.;
Gwaltney, S. R.; Heyden, A.; Hirata, S.; Hsu, C.-P.; Kedziora, G.;
Khalliulin, R. Z.; Klunzinger, P.; Lee, A. M.; Lee, M. S.; Liang, W. Z.;
Lotan, I.; Nair, N.; Peters, B.; Proynov, E. I.; Pieniazek, P. A.; Rhee, Y.
M.; Ritchie, J.; Rosta, E.; Sherrill, C. D.; Simmonett, A. C.; Subotnik, J.
E.; Woodcock, H. L., III; Zhang, W.; Bell, A. T.; Chakraborty, A. K.;
Chipman, D. M.; Keil, F. J.; Warshel, A.; Hehre, W. J.; Schaefer, H. F.;
Kong, J.; Krylov, A. I.; Gill, P. M. W.; Head-Gordon, M. Phys. Chem.
Chem. Phys. 2006, 8, 3172.
(24) Macchioni, A. Chem. Rev. 2005, 105, 2039.
(25) (a) Schaper, F.; Geyer, A.; Britzinger, H. H. Organometallics
2002, 21, 473. (b) Zuccaccia, C.; Stahl, N. G.; Macchioni, A.; Chen,
M.-C.; Roberts, J. A.; Marks, T. J. J. Am. Chem. Soc. 2004, 126, 1448.
(c) Rocchigiani, L.; Bellachioma, G.; Ciancaleoni, G.; Macchioni, A.;
Zuccaccia, D.; Zuccaccia, C. Organometallics 2011, 30, 100.
(26) Rocchigiani, L.; Ciancaleoni, G.; Zuccaccia, C.; Macchioni, A.
Angew. Chem., Int. Ed. 2011, 50, 11752.
(27) Zucchini, U.; Albizzati, E.; Giannini, U. J. Organomet. Chem.
1971, 26, 357.
(28) Segal, S.; Goldberg, I.; Kol, M. Organometallics 2005, 24, 200.
(29) (a) Ahlrichs, R.; Bar, M.; Baron, H.-P.; Bauernschmitt, R.;
̈
Bocker, S.; Ehrig, M.; Eichkorn, K.; Elliott, S.; Furche, F.; Haase, F.;
̈
Haser, M.; Hattig, C.; Horn, H.; Huber, C.; Huniar, U.; Kattannek, M.;
Kohn, A.; Kolmel, C.; Kollwitz, M.; May, K.; Ochsenfeld, C.; Ohm, H.;
Schafer, A.; Schneider, U.; Treutler, O.; Tsereteli, K.; Unterreiner, B.;
̈
̈
̈
̈
̈
von Arnim, M.; Weigend, F.; Weis, P.; Weiss, H. Turbomole, Version 5;
Theoretical Chemistry Group, University of Karlsruhe, 2002.
(b) Treutler, O.; Ahlrichs, R. J. Chem. Phys. 1995, 102, 346.
(30) (a) PQS, version 2.4; Parallel Quantum Solutions: Fayetteville,
AK, 2001 (the Baker optimizer is available separately from PQS upon
request). (b) Baker, J. J. Comput. Chem. 1986, 7, 385.
̈
(31) Eichhorn, K.; Treutler, O.; Ohm, H.; Haser, M.; Ahlrichs, R.
Chem. Phys. Lett. 1995, 242, 652.
̈
(32) Becke, A. D. Phys. Rev. A 1988, 38, 3098. Perdew, J. P. Phys. Rev.
B 1986, 33, 8822. Treutler, O.; Ahlrichs, R. J. Chem. Phys. 1995, 102,
346.
(33) Schafer, A.; Horn, H.; Ahlrichs, R. J. Chem. Phys. 1992, 97, 2571.
̈
(34) Andrae, D.; Haussermann, U.; Dolg, M.; Stoll, H.; Preuss, H.
̈
Theor. Chim. Acta 1990, 77, 123.
(35) Schafer, A.; Klamt, A.; Sattel, D.; Lohrenz, J. C. W.; Eckert, F.
̈
Phys. Chem. Chem. Phys. 2000, 2, 2187.
(36) Schafer, A.; Huber, C.; Ahlrichs, R. J. Chem. Phys. 1994, 100,
5829.
(37) Using the PCM method (a−c) of Gaussian03 (d): (a) Cances,
E.; Mennucci, B.; Tomasi, J. J. Chem. Phys. 1997, 107, 3032.
(b) Mennucci, B.; Tomasi, J. J. Chem. Phys. 1997, 106, 5151. (c) Cossi,
M.; Barone, V.; Mennucci, B.; Tomasi, J. Chem. Phys. Lett. 1998, 286,
253. (d) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;
Robb, M. A.; Cheeseman, J. R.; Montgomery Jr., J. A.; Vreven, T.;
Kudin, K. N.; Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.;
Barone, V.; Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.;
Petersson, G. A.; Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.;
Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao,
O.; Nakai, H.; Klene, M.; Li, X.; Knox, J. E.; Hratchian, H. P.; Cross, J.
B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R.
E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.;
Ayala, P. Y.; Morokuma, K.; Voth, G. A.; Salvador, P.; Dannenberg, J.
J.; Zakrzewski, V. G.; Dapprich, S.; Daniels, A. D.; Strain, M. C.;
Farkas, O.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman,
J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.; Clifford, S.; Cioslowski, J.;
Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.;
Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.;
Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.; Johnson, B.; Chen,
H
dx.doi.org/10.1021/ma300343c | Macromolecules XXXX, XXX, XXX−XXX