Structure-activity correlation in titanium single-site olefin polymerization catalysts containing mixed cyclopentadienyl/aryloxide ligation

Article abstract of DOI:10.1021/ja0640849

In this work, we report the synthesis of eighteen titanium cyclopentadienyl aryloxide complexes and their propagation rate constants for 1-hexene polymerization. A correlation between k_p values and the underlying catalyst structures is developed using DFT-computed ligand cone angles and ion pair separation energies as structural and electronic descriptors. The correlation takes the form of an Arrhenius-like relationship where the pre-exponential factor is correlated to the ligand cone angles and the activation energy term is correlated to the ion pair separation energies. This finding is consistent with the idea that the ability of the monomer to access the metal center is a key factor affecting the reaction rate. Copyright

Full text of DOI:10.1021/ja0640849

Published on Web 03/10/2007
Structure-Activity Correlation in Titanium Single-Site Olefin Polymerization
Catalysts Containing Mixed Cyclopentadienyl/Aryloxide Ligation
Thomas A. Manz,† Khamphee Phomphrai,‡ Grigori Medvedev,† Balachandra B. Krishnamurthy,†
Shalini Sharma,‡ Jesmin Haq,† Krista A. Novstrup,† Kendall T. Thomson,† W. Nicholas Delgass,†
James M. Caruthers,*,† and Mahdi M. Abu-Omar*,‡
School of Chemical Engineering and Department of Chemistry, Purdue UniVersity, West Lafayette, Indiana 47907
Received June 9, 2006; E-mail: caruther@ecn.purdue.edu; mabuomar@purdue.edu
Scheme 1. Catalyst Precursor Synthesis
Single-site polymerization catalysts are commercially important
because the polymer’s molecular architecture can be varied by chang-
ing the catalyst structure.1 Ti and Zr mixed cyclopentadienyl arylox-
ide complexes have recently received considerable interest as olefin
polymerization catalysts.2 A catalyst structure library can be created
by varying the substituents on the cyclopentadienyl and aryloxide
ligands, where the catalyst’s performance depends on the nature
of the ligands.
Our structure-activity correlation is based on the idea that partial
displacement of the counterion from the metal center to allow space
for monomer coordination and insertion is a key factor affecting
the polymerization rate. On the basis of this idea, the reaction rate
should increase when separation of the counterion is easier and
when the metal center is less sterically hindered, because these two
factors increase monomer access to the metal. Ion pair separation
energies (EIPS) and ligand cone angles were selected as quantitative
descriptors of these two factors. EIPS was calculated by subtracting
the DFT-computed SCF energy of the contact ion pair from that
of the bare cation [Cp′Ti(OAr)Me]+ and counterion [MeB(C6F5)3]-
in a toluene-like solvent. The ligand cone angle was defined as the
largest X-M-Y angle where M is the center of the metal atom
and X and Y are two points on the van der Waals surface of the
ligand outside the metal’s van der Waals sphere.4 Using the DFT
optimized contact ion pair geometries, this definition provided the
We report the synthesis of eighteen titanium cyclopentadienyl
aryloxide complexes and their propagation rate constants for
1-hexene polymerization. A correlation between kp and the catalyst
structures is developed using DFT-computed ligand cone angles
and ion pair separation energies (EIPS). This correlation takes the
form of an Arrhenius-like relationship, where the pre-exponential
factor (k0) is correlated to the ligand cone angles and the activation
energy term (Ea) is correlated to EIPS
.
A series of compounds Cp′Ti(OAr)Me2, where Cp′ ) C5H5 (Cp)
or C5Me5 (Cp*) were synthesized from three general routes (Scheme
1). For very bulky aryloxides such as OC6H-2,3,5,6-Ph4, a simple
deprotonation by Cp′TiMe3 failed. The easiest way to incorporate
bulky aryloxides was from the reaction of Cp′TiCl3 and LiOAr
followed by complete methylation (method A). For moderately
i
bulky aryloxides such as OC6H3-2,6-R2 (R ) Me, Et, Pr), the
reaction of Cp′TiCl3 and LiOAr in method A yielded a mixture of
mono- and bis(aryloxide) titanium complexes. The most effective
preparation of this type of catalyst was from deprotonation of the
parent phenol with Cp′TiMe3 at low temperature (method B). For
less bulky aryloxides having no ortho substituents, the formation
of the bis(aryloxide) byproduct became problematic in methods A
and B. The solution was to react Cp′TiMe2Cl with LiOAr to exclude
the possibility of bis(aryloxide) complex formation (method C).3
The addition of 1 equiv B(C6F5)3 to Cp′Ti(OAr)Me2 in toluene
immediately gave thermally unstable contact ion pairs [Cp′Ti(OAr)-
Me]+[MeB(C6F5)3]- that were active for the polymerization of 1-hex-
ene. Polymerization reactions were followed by 1H NMR at 0 °C,
and polymer molecular weights were determined by GPC analysis.
A kinetic model containing initiation, propagation, and deactivation
steps was fit to the ln([1-hexene]/[1-hexene]0) versus time concen-
tration profiles and Mn data to obtain the kp values (Table 1). The
model equations and fitted kinetic profiles are presented in the
Supporting Information.
ligand steric descriptors θCp′ and θOAr
.
In Gaussian 03,5 the OLYP/6-311++G** method was chosen
because of its high accuracy and much lower computational cost
than hybrid methods like B3LYP.6 Vacuum and PCM-optimized
geometries for the ion pair of catalyst 1 were computed and found
to be nearly identical. As PCM geometry optimization was very
costly, all geometries were optimized in vacuum. The PCM model
was then used to compute solvation energies on these geometries.3
The solid angle subtended by cone angle θ equals 4π sin2(θ/4).
Thus, the solid angle 4πγ available for monomer approach to the Ti
site when the counterion is partially displaced is approximated by
γ ) 1 - sin2(θCp′/4) - sin2(θOAr/4) - f
(1)
where f is a factor that accounts for ligand orientation and space
blocked by the growing chain and partially displaced counterion.
The experimental data is fit to a correlation in the form of an
Arrhenius-like relationship:
In chain propagation, the monomer first coordinates to the metal
to form a π-complex and then inserts to extend the polymer chain.1e
The chemical structure of the counterion is known to have a large
effect on the catalyst’s activity, and it has generally been established
that weakly coordinating counterions are needed to achieve high
polymerization rates.1a,b Detailed experiments can be used to extract
the rate constants for the different reaction steps.1d,f
IPS/RT
kpred ) k0e-E /RT ) γa0e-E /RTe-RE
(2)
a
0
where kpred is the predicted value of kp, R is the gas constant, T is
absolute temperature, and a0, E0, and R are model parameters.
Figure 1 shows the experimental kp plotted as a function of EIPS and
divided into different catalyst families where the smaller EIPS was used
for catalysts 13, 15, and 17. Family A contains the Cp catalysts,
while the Cp* catalysts are divided as (B) no methoxy, coordinating,
† School of Chemical Engineering.
‡ Department of Chemistry.
9
3776
J. AM. CHEM. SOC. 2007, 129, 3776-3777
10.1021/ja0640849 CCC: $37.00 © 2007 American Chemical Society

C O M M U N I C A T I O N S
Table 1. Summary of Experimental and Computed Parameters for [Cp′Ti(OAr)Me]+[MeB(C6F5)3]- Catalysts
1
catalyst
Cp
′
ligand
OAr ligand
family
k
p (M-1 s-
)
M
n (g/mol)
θCp′ (deg)
θ
OAr (deg)
E
IPS (kcal/mol)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
Cp
Cp
Cp
OC6H2-2,6-Me2-4-Br
OC6H3-2,6-Et2
OC6H3-2,6-iPr2
OC6H5
OC6H4-4-F
OC6H4-4-Cl
A
A
A
B
B
B
B
B
B
C
C
C
B
C
D
D
E
0.22
0.34
0.42
0.42
0.27
0.24
0.26
0.38
0.48
0.51
0.78
0.92
0.74
0.28
0.27
1.36
0.25
0.20
3200
5200
6200
4500
4700
5200
5200
4800
5100
28500
44700
62300
17200
27900
14400
9500
126.3
127.3
127.5
153.8
154.7
154.9
154.6
154.5
154.8
153.1
153.5
153.3
154.4
153.7
153.6
152.9
154.7
154.6
127.5
131.7
137.2
94.9
94.8
94.6
94.6
94.6
94.5
126.3
131.3
135.7
125.8
126.2
111.9
162.7
94.6
35.7
34.6
34.1
31.4
32.1
32.1
31.8
31.2
31.0
28.9
28.4
26.6
Cp*
Cp*
Cp*
Cp*
Cp*
Cp*
Cp*
Cp*
Cp*
Cp*
Cp*
Cp*
Cp*
Cp*
Cp*
OC6H4-4-Br
OC6H4-4-Ph
OC6H4-4-tBu
OC6H3-2,6-Me2
OC6H3-2,6-Et2
OC6H3-2,6-iPr2
OC6H4-2-(cyclohexyl)
OC6H2-2,6-Me2-4-Br
OC6H4-2-(CH2Ph)
OC6H-2,3,5,6-Ph4
OC6H4-3-(OMe)
OC6H4-4-(OMe)
30.2 (d), 30.9 (p)a
29.6
23.9 (d), 31.1 (p)a
14.1
3100
5400
30.7 (d), 31.1 (p)a
30.2
E
95.0
a For aryloxide ligands with only one ortho or meta substituent group, the substituent group can be either distal (d) or proximal (p) to the growing chain.
collected at the same temperature, a single parameter A ) a0e-E /RT
0
was used for each catalyst family. The correlation provides a good
fit to the experimental data (Figure 2), suggesting its underlying
physical basis is correct. This means a portion of Ea for chain
propagation is due to separation of the ion pair. Because the ion
pair only separates to a finite distance, only part of EIPS contributes
to Ea and thus 0 < R < 1. The correlation also shows that steric
factors are important and can be approximately described in terms
of ligand cone angles. Other physical factors which are not explicitly
accounted for are implicitly grouped into different A’s for different
catalyst families. We are now studying how these additional factors
relate to catalyst structure with the goal of developing a more
comprehensive predictive model.
Figure 1. Experimental 1-hexene polymerization rate constants (kp) versus
computed ion pair separation energies in toluene (EIPS).
Acknowledgment. Financial support was provided by the U.S.
Department of Energy Catalysis Science Grant No. DE-FG-
0203ER15466. Supercomputing resources were provided by In-
formation Technology at Purdue and the National Center for
Supercomputing Applications, Grant No. MCA04N010.
Supporting Information Available: Complete ref 5; experimental
details, kinetic and error analysis, and model optimization; DFT
geometries and solvation calculations; X-ray crystal structure for Cp*Ti-
(OC6H-2,3,5,6-Ph4)Me2. This material is available free of charge via
the Internet at http://pubs.acs.org.
References
Figure 2. Prediction of the propagation rate using eqs 1 and 2 with R )
0.300, f ) 0.187, and the following values of A (M-1 s-1) according to the
indicated catalyst family: A, 3.01 × 108 (green square); B, 5.22 × 107
(red circle); C, 2.65 × 107 (blue diamond); D, 6.88 × 105 (crossed lines);
and E, 1.77 × 107 (green triangle).
(1) (a) Chen, M. C.; Roberts, J. A. S.; Marks, T. J. J. Am. Chem. Soc. 2004,
126, 4605-4625. (b) Chen, E. Y.-X.; Marks, T. J. Chem. ReV. 2000, 100,
1391-1434. (c) Chem. ReV. 2000, 100 (4). (d) Veghini, D.; Henling, L.
M.; Burkhardt, T. J.; Bercaw, J. E. J. Am. Chem. Soc. 1999, 121, 564-
573. (e) Lanza, G.; Fragala, I. L.; Marks, T. J. Organometallics 2002, 21,
5594-5612. (f) Liu, Z.; Somsook, E.; White, C. B.; Rosaaen, K. A.;
Landis, C. R. J. Am. Chem. Soc. 2001, 123, 11193-11207.
or 6 substituents; (C) alkyl substituents in both 2, 6 positions; (D)
substituent phenyl rings that partially coordinate to the Ti center in
the bare cation; and (E) methoxy substituents. Catalyst 13 is includ-
ed in family B because in the state with lower EIPS (distal conforma-
tion) the cyclohexyl group points away from the growing chain
providing negligible steric hindrance. A linear correlation between
log(kp) and EIPS emerges for each catalyst family. The methoxy cata-
lysts (family E) are structurally similar to family B but have lower
kp values for comparable EIPS. This could potentially be caused by a
coordination between methoxy groups and either a second Ti center or
the B(C6F5)3 activator, since both of these are prone to bind ethers.
The experimental data were fit to eqs 1 and 2 by minimizing
the sum of (log(kp) - log(kpred))2. Since the experimental data were
(2) (a) Byun, D. J.; Fudo, A.; Tanaka, A.; Fujiki, M.; Nomura, K.
Macromolecules 2004, 37, 5520-5530. (b) Nomura, K.; Tsubota, M.;
Fujiki, M. Macromolecules 2003, 36, 3797-3799. (c) Nomura, K.; Fudo,
A. J. Mol. Catal. A: Chem. 2004, 209, 9-17. (d) Nomura, K.; Itagaki,
K.; Fujiki, M. Macromolecules 2005, 38, 2053-2055. (e) Wang, W.;
Fujiki, M.; Nomura, K. J. Am. Chem. Soc. 2005, 127, 4582-4583. (f)
Fenwick, A. E.; Phomphrai, K.; Thorn, M. G.; Vilardo, J. S.; Trefun, C.
A.; Hanna, B.; Fanwick, P. E.; Rothwell, I. P. Organometallics 2004, 23,
2146-2156. (g) Phomphrai, K.; Fenwick, A. E.; Sharma, S.; Fanwick, P.
E.; Caruthers, J. M.; Delgass, W. N.; Abu-Omar, M. M.; Rothwell, I. P.
Organometallics 2006, 25, 214-220.
(3) See Supporting Information for details.
(4) Tolman, C. A. J. Am. Chem. Soc. 1970, 92, 2956-2965.
(5) Frisch, M. J.; et al. Gaussian 03; Gaussian Inc.: Pittsburgh, PA, 2003.
(6) Baker, J.; Pulay, P. J. Chem. Phys. 2002, 117, 1441-1449.
JA0640849
9
J. AM. CHEM. SOC. VOL. 129, NO. 13, 2007 3777