Bis-amides and Amine Bis-amides as Ligands for Olefin Polymerization Catalysts Based on V(IV), Cr(IV) and Mn(IV). A Density Functional Theory Study

1
Bis-amides and Amine Bis-amides as Ligands for
Olefin Polymerization Catalysts Based on V(IV),
Cr(IV) and Mn(IV). A Density Functional Theory
Study

• Timothy K. Firman and Tom Ziegler
• University of Calgary

2
Introduction
Following the discovery of catalytic olefin
polymerization activity of group IV
metallocene-based systems 19761, a great variety
of homogeneous transition-metal based olefin
polymerization catalysts have been discovered.
Many d0 group III or group IV metal based
systems, with a widening variety of ligands were
found, from constrained geometry catalysts2 to
the diamide system of McConville et al.3
Computational modeling of d0 systems has also
progressed rapidly.4 More recently, a number of
late transition metal catalysts have been
developed. Brookhart et al. discovered d8 Ni and
Pd compounds5 which are active polymerization
catalysts with certain bulky ligand systems.
Several systems with a Cr center and a few d
electrons have recently been found to be active
catalysts.6,7 With these indications of
catalytic potential in compounds with electron
counts between zero and eight, we present a
computational study of d1, d2, and d3 metal
systems with nitrogen-based ligands. A previous
study8 had indicated some potential in bis-amide
systems, we explore here systems with an
additional amine. The amine was added in the
hope of destabilizing the termination transition
state. We use an analysis of the localized
electron density to explore how the nature and
placement of the ligands affect catalytic
properties.

3
Computational Details
All structures and energetics were calculated
with the Density Functional Theory (DFT) program
ADF9. All atoms were modeled using a frozen core
approximation. V, Cr, and Mn were modeled with a
triple-z basis of Slater type orbitals (STO)
representing the 3s, 3p, 3d, and 4s orbitals with
a single 4p polarization function added. Mo, Ru,
and Pd were similarly modeled with a triple-z STO
representation of the 4s, 4p, 4d, 5s, and a
single 5p polarization function. Main group
elements were described by a double-z STO
orbitals with one polarization function (3d for
C, N and 2p for H.)10 In each case, the local
exchange-correlation potential11 was augmented
with electron exchange functionals12 and
correlation corrections13 in the method known as
BP86. First-order scalar relativistic
corrections14 were added to the total energy of
all systems. In most cases, transition states
were located by optimizing all internal
coordinates except for a chosen fixed bond
length, iterating until the local maximum was
found, with a force along the fixed coordinate
less than .001 a.u. For b-hydride transfer,
transition states were found using a standard
stationary point search to a Hessian with a
single negative eigenvalue. All calculations were
spin unrestricted and did not use symmetry. The
Boys and Foster method was used for orbital
localization, and the orbitals were displayed
using the adfplt program written by Jochen
Autschbach.
4
d1V, d2Cr , d3Mn a Comparison

• All have formal oxidation states of 4, but will
  each have a net charge of about 1.
• Each has six valence orbitals (an s and 5 d).
• After filling the SOMOs, V has 5 empty orbitals,
  Cr 4 and Mn 3
• Each available orbital has a bonding interaction
  with the ligands.
• These orbitals must be orthogonal
• Individual orbitals of V will be more ligand
  centered, to balance charge
• Metal bonding orbitals are often shared between
  ligands, e.g. trans- ligands typically share a
  single s-bonding metal orbital.

5
Metal Alkyl Structures

• d1V is nearly tetrahedral
• NH2 are flat, with p bonds not in the same plane
• a b-agostic hydride
• d2Cr is nearly tetrahedral
• NH2 are flat, p aligned(shared)
• no b-agostic hydride
• d3Mn includes a 146angle
• NH2 are bent out of plane due to weakened p
  -interactions
• no b-agostic hydride

6
Olefin Adduct

• d1V is trigonal bipyramidal
• NH2 are flat, p bonds unshared
• b-agostic hydride
• d2Cr is trigonal bipyramidal
• NH2 are flat, p bonds aligned
• No b-agostic hydride
• d3Mn is trigonal bipyramidal
• NH2 are bent out of plane
• No b-agostic hydride
• NH2 is apical instead of ethene

7
Insertion
DE

• Insertion barriers are similar and fairly small
• Geometries are quite different from one another
• Each has a ligand trans to a forming or breaking
  bond

8.3kcal/mol 12.5 kcal/mol 13.6 kcal/mol
8
Termination (b-hydride transfer)

• Each termination barrier is substantially higher
  than the insertion
• No b-hydride elimination TS found lower in energy
• The amine is either trans to the hydride, or to
  one of the reacting M-C bonds

DE
16.4 kcal/mol 19.6 kcal/mol 20.0
kcal/mol
9
Enthalpies Summary

• Insertion and termination numbers are promising,
  particularly for a system lacking steric bulk
• Uptake energy is too low.
• Entropy will be unfavorable by about 12-15
  kcal/mol
• Displacement of counterion will also effect
  uptake energetic

10
LocalizedOrbitals
Olefin insertion of d2Cr
11
More Localized Orbitals
Chain Termination-d2 Cr (via b-hydride transfer)
12
The Second Row Transition Metals

• Good second row olefin polymerization catalysts
  exist, including d0 Zr and d8 Pd
• Olefin uptake energies are expected to increase
  due to generally stronger bonds
• These compounds are found to be low spin
• Compounds with a like number of occupied metal
  orbitals may be analogous
• Model systems with d2 Mo,d4 Ru, and d6 Pd were
  calculated

13
Second Row Results
While the uptake energies are substantially
improved, these combinations of ligand and metal
do not result in good catalysts.
14
Tethered Nitrogen Ligands

• Electronically similar to the previous systems
• Chelation will keep the ligands bound
• All three nitrogen will stay on one side this
  will leave the other side vacant and may help
  uptake
• Limited conformational flexibility
• Sterically unhindered, as in the untethered case

15
Uptake Enthalpy of Linked System
DEreorganization is the energy required to
distort the alkyl minimum to the shape of the
adduct (minus the ethylene)

• The metal-ethylene bond energy would be about 20
  kcal/mol in each case, but large differences in
  reorganization energy result in differences in
  uptake energies.
• The shapes of the untethered ethylene adducts
  predict energetics
• In the untethered Cr adduct, the two NH2 groups
  are near the NH3 group with a hydrogen pointing
  toward it. The tethers hold it in just this
  position.
• V has the N ligands close together, but must
  twist one of the NH2 groups.
• Mn has an NH2 trans to the NH3, which cannot
  occur with a tether, so the uptake energy is
  actually worse with the tether than without.

16
Catalytic Properties with Tether

• The tether again has a large effect on the
  energetics that is very different for different
  metals
• In the V system, The N ligands are held in a
  position close to both transition states the
  energies of both are decreased.
• In Cr, the insertion is close to the tethered
  case, but the untethered termination prefers
  trans NH2 groups, which is impossible for the
  tethered case. The resulting energy is much
  higher.
• In Mn, the tether is different from all three
  shapes, causing them each a similar energy
  penalty. The net result is similar DEs.

17
Conclusions

• The occupation of metal orbitals has a
  substantial effect on molecular properties
• The binding changes in transition states
  substantially in flexible, untethered systems
• Tethering the ligands alters the energetics
  differently for different transition states
  matching tether types with metal is important

18
Acknowledgements
This research was supported by the Natural
Sciences and Engineering Research Council of
Canada (NSERC) and Novacor Research
and Technology Corporation.
References
(1) Andersen, A. A. Cordes, H. G. Herwig, J.
Kaminsky, W. Merck, A. Mottweiler, R. Pein,
J. Sinn, H. Vollmer, H. J. Angew. Chem.,
Int. Ed. Engl. 1976, 15, 630. (2) a) Stevens, J.
C. Timmers, F. J. Wilson, D. R. Schmidt, G.
F. Nicklas, P. N. Rosen, R. K. Knight, G. W.
Lai, S. Y. Eur. Pat. Appl., EP 416815,
1990. b) Stevens, J. C. Neithammer, D. R. Eur.
Pat. Appl., EP 418044, 1990. c) Canich, J. A. M.
Eur. Pat. Appl., EP 420436, 1990. (3)
Scollard, J. D. McConville, D. H. Payne, N. C.
Vittal, J. J. Macromolecules 1996, 29, 5241. (4)
Margl, P. Deng, L. Ziegler, T. Organometallics
1998, 17, 933. (5) Johnson, L. K. Killian, C.
M. Brookhart, M. J. Am. Chem. Soc. 1995, 117,
6414. (6) White, P. A. Calabrese, J. Theopold,
K. H. Organometallics 1996, 15, 5473. (7) Emrich,
R. Heinemann, O. Jolly, P. W. Krüger, C.
Verhovnik, G. P. J. Organometallics 1997, 16,
1511. (8) Schmid, R. Ziegler, T. Submitted for
publication. (9) a) ADF 2.3.3, Theoretical
Chemistry, Vrije Universiteit, Amsterdam
b) Baerends, E. J. Ellis, D. E. Ros, P.
Chem.Phys. 1973, 2, 41. c) te Velde, G
Baerends, E. J. J. Comp. Phys. 1992, 99, 84. (10)
Snijders, J. G. Baerends, E. J. Vernoijs, P.
At. Nuc. Data Tables 1982, 26, 483. (11) Vosko,
S. H. Wilk, L. Nusair, M. Can. J. Phys. 1980,
58, 1200. (12) Becke, A. Phys. Rev. A 1988, 38,
3098. (13) a) Perdew, J. P. Phys. Rev. B 1986,
34, 7406. b) Perdew, J. P. Phys. Rev. B 1986,
34, 8822. (14) a) Snijders, J. G. Baerends, E.
J. Mol. Phys. 1978, 36, 1789. b)
Snijders, J. G. Baerends, E. J. Ros, P. Mol.
Phys. 1979, 38, 1909.