agostic bani seminar2

7/30/2019 Agostic Bani Seminar2

1/47

Agostic Interaction in d0 TransitionMetal Alkyl Complexes

Bani Kanta Sarma


2/47

Introduction:

Definition,Historical perspective and Importance.

Characterization of Agostic Interactions.

Theoretical Treatment of the Agostic Interaction.

Revised Bonding Model for d0 System and Evidences.

Driving force and Bond Ellipticity

Manipulation of Agostic Interactions:

Metal Polarization and the Extent of Delocalization.

A Note on dn (n 0) Agostic Complexes.

Conclusion

Plan of Talk

1


3/47

Original definition: to refer specifically to situations in which a hydrogen

atom is covalent lybonded to both a carbon and to a transition metal atom.

Brookhart, M.; Green, M. L. H. J. Organometall. Chem. 1983, 250, 395.

Generalized definition : distortion of an organometallic moietywhich brings

an appended C-H bondinto close proximitywith the metal center.

Scherer, W.; McGrady, G. S.Angew. Chem., Int. Ed. Engl. 2004, 43, 1782.

Such a definition accommodates most of the reported examples, and separates

the nature of the phenomenon and the driving force behind it from its observable

chemical consequences.

Definition

LnM

H2CCH2

H

2


4/47

Development of the Concept

La Placa S. J.; Ibers, J. A. Inorg. Chem. 1965, 4, 778.

- is about what is expected from van der Waals radii-

Considered it to be a true five coordinated species.

Bailey, N. A.; Jenkins, J. M.; Mason, R.; Shaw, B. L. J.

Chem. Soc. Chem. Commun. 1965, 237.

-a distorted octahedron, the sixth coordination

being occupied by a hydrogen

Sum of rvdw = 3.1

3


5/47

MountingEvidence

Trofimenko, S. J. Am. Chem. Soc. 1968, 90, 4754.

Inorg. Chem.1970, 9, 2493.

-Observation Hydridic(-2.6 ppm) character in the 1H NMR

hydrogens are intruding into a suitable empty metal orbital.

Cotton, F. A.; LaCour, T.; Stanislowski, A. G.J. Am. Chem. Soc. 1974, 96, 754.

- studied by single-crystal X-ray diffraction

a three center, two-electron bond encompassing the CHMo atoms

- analogy to bonding concept in borane chemistry.

Mo

OC

N N

CO

N NB

H2CCH3

Ph

CH3

H

4


6/47

Mounting Evidence

PdH distance isLess than the sum of the van der Waals radii

Roe, D. M.; Bailey, P. M.; Moseley, K.; Maitlis, P.M. J.Chem.Soc.Chem. Commun. 1972, 1237.

This type of interaction has not been observed before, though Trofimenko noted that

the methylene hydrogens of the ethyl groups in [Pd{Et2B(pz)2}(3-C3H4Ph)(CO)2] and

[Ni{Et2B(pz)2}2] are shifted upfield.

Sum of rvdw = 3.1

5


7/47

Conclusive Evidence: X-ray and Neutron diffraction Studies

Brown, R. K.; Williams, J. M.; Schultz, A. J.; Stucky, G.

D.; Ittel, S. D.; Harlow, R. L. J. Am. Chem. Soc. 1980,

102, 981.1.874

a very short FeH distance of 1.874 at 30 K.

stretchedaliphatic C-H bond (1.164 ), the longest ever observed in a

crystal.

6


8/47

In 1982, Green et al. reported what are now the textbook examples of MHC -

and -agostic interactions in the transition metal alkyl complexes [RTiCl3(dmpe)]

(dmpe = Me2PCH2CH2PMe2; R = Me or Et).

Green et al coined the term agostic to describe the phenomenon. They also

introduced a halfarrow convention to represent the interaction.

The term agostic comes from the Greek word which occurs in

Homer and translates as to clasp or hold to oneself.

Growing Evidences

a) Green et al J.Chem. Soc. Chem. Commun.1982, 1410.

b) Green et al, J. Chem. Soc. Dalton Trans.1986, 1629.

C

TiP

P Cl

H

H

H

Cl

ClH2C

Ti

HH

H

P

PClCl

Cl

C-H M

7


9/47

Reflection of Growing Importance

Fig 1. Number of publications N with the keyword

agostic which appeared in the chemical literature

between 1980 and 2000.

8


10/47

Why do we care?1. Olefin elimination from transition metal alkyls and the reverse reaction-olefin

insertion into M-H bond - are both believed to proceed through -agostic

interaction.

2. ZieglerNatta Polymerization:

LnM

H2C

CH3

LnM

H2CCH2

H

Ln

M

H2C CH2

H

Conversion of -agostic interaction to -agostic interaction in a step wise manner.

C

Ti

H CH2

CH2

P

C

Ti

H CH2

CH2

P

C

Ti

H CH2

CH2

P

C

Ti

H

P

(i)

(ii)

(iii)

(iv) C C+

-agostic-agosticGreen et al. J. Organometall. Chem. 1983, 250, 395 9


11/47

3. Hydroformylation:

Why do we care?

C Co

CO

CO

CO

O

H2C

H2

C

H

H2C Co

CO

CO

CO

H2C H

HCo(CO)4

a

HCo(CO)3

Co(CO)3

H

RCo(CO)4

RCO-Co(CO)3

H2

RCHO

CO

b

cd

e

+ CO

II

IIIIIIII

HCo(CO)4 + CO + H2 CH3CH2CHOH2C CH2

Ziegler, T. Can. J. Chem.1995, 73, 743.

10


12/47

4. Activation of C-H bonds: Functionalization of inert C-H bond.

Why do we care?

O

B

O

B

O

O

B

O

O

+ H2

BR2

RhR2B

C H

66

Rh

R2B H

R2B C

RhR2B H

R2B C

RhR2B C H+

H

RhR2B+CR2B

+

BR2

Catalyst

= Cp*Rh(4-C6Me6)

11


13/47

X-ray diffraction study of a single crystal represents by far the most commonly

used structural technique to reveal agostic interactions.

Characterization of Agostic Interaction

X-ray Diffraction

Low scattering factor of the hydrogen atom

The relative X-ray scattering ratio of Z(M)/1 is unfavorable.

Discrepancies of more than 0.1 in the MH separations.

Problem arises with the accurate location of H atoms.

12


14/47

More accurate location of the hydrogen atoms

The neutron scattering ratio M versus H is more favorable than in the X-

ray case.

Neutron Diffraction

Only a handful of agostic alkyl systems has been characterized to date by

neutron diffraction studies.

Long collection times and large and high quality single crystals required.

13


15/47

Study of isolated molecule free from intermolecular forces

No perturbation of the equilibrium structure

Gas Electron Diffraction

To date, no compound has been shown unambiguously to

possess an agostic structure by GED.

Problems of Thermal stability, involatility.

The complicated diffraction pattern

Does not discriminate well between comparable interatomic distances

The relatively weak scattering of H atoms as in X-ray diffraction.

14


16/47

1H chemical shifts range ( typically 0 to -16 ppm for dn ; n 0 systems)

0 to 3 ppm for d0

systems

Coupling Constants for the J(13C-1H); (typically 75 -100 Hz)

NMR Spectroscopy

Long timescale of the NMR experiment (ms or longer).

Weak nature of the interaction (1-10 kcal/mol)

Fluxionality averages the changes between two or more C-H bonds.

Static spectra are in the best case observable at low temperatures

(e.g. 80 to -100 C). 15


17/47

A striking -agostic complex [EtNi(dbpe)]+ (dbpe=

tBu2PCH2CH2PtBu2) reported by Spencer et al.

Ambient temperature

1H (NMR); = - 1.24 ppm

Characterization by Low Temperature NMR

At -100 C this splits into normal methylene and agostic resonances

at = +1.1 and - 5.8 ppm, respectively.Coupling constant 1J(C-H) = 68 Hz (agostic H)

= 151 Hz (terminal H)

a) Spencer et al J.Chem. Soc. Chem. Commun.1991, 1601.

b) Spencer et al Organometallics1991, 10, 49.16


18/47

J(C- H)ave

= 120-130 Hz

J(C H)ave

= 80 - 90 Hz

J(C- H)ave

= 120-130 Hz

Fluxional System

C

H

H

M

H

C

C

H

H

M

B

C

H

H

M

A

HH

J(C-Hbridging) 80-100 Hz.J(C-Hterminal) 140 Hz.

This arises from an increase in s-characterin the terminal C-H bonds.

value in 1H NMR - a poor criterion for distinguishing between A, B, and C.

J(C- H)

ave= 120 Hz

17


19/47

Isotopic Perturbation of Resonance (IPR) Studies

J( 13C-1H) 110 Hz rules out C

Isotopic Perturbation of Resonance (IPR) found rule out A

1. 1H chemical shift (CH3) > (CH2D > (CHD2).

2. 13C-H coupling J(13C-H) values

J(C-H) for CH3 > J(C-H) for CH2D > J(C-H) for CHD2

C-H-M and C-D-M small zero point energy difference

C-H and C-D large zero point energy difference

This arises because there is a thermodynamic preference forhydrogen rather than deuterium to occupy the bridging position.

18


20/47

The C-H stretching vibration, (C-H) of the agostic C-H bond should

move to lower frequency.

The lower energy vibrations involving C-H deformation and rocking

motions should likewise be significantly perturbed.

Vibrational Spectroscopy

Advantage is the short timescale of vibrational transitions.

(C-H)28002350 cm-1

has in many instances been interpretedas diagnostic evidence of a MHC interaction.

19


21/47

Complications with Vibrational Spectroscopy

The individual C-H oscillators on a carbon atom couple to give symmetric and

antisymmetric combinations.

Overtones of deformation modes around 1400 cm-1 fall close to the region of

interest.

Fermi resonance of overtones with the (C-H) fundamentals, leads to hybrid

features which are difficult to deconvolute.

All but one C-H bond in an alkyl group is substituted with deuterium.

The (C-H) fundamentals are thus decoupled from all other vibrations.

Partial Deuteriation

20


22/47

Theoretical Studies

Hypothetical model [MeTaH4]n- (n=2 or 4) using the Extended Huckel (EHMO)

Methyl group could in principle distort like a carbene group.

According to this study secondary interaction weakens theC-H bond and attracts the -hydrogen to the metal.

R. J. Goddard, R. Hoffmann, E. D. Jemmis, J. Am. Chem. Soc.1980, 102, 7667.

Ti

H

H

H

C

H

H

HH

[MeTiH5]2-MeTiCl3(dmpe)

O. Eisenstein, Y. Jean, J. Am. Chem. Soc.1985, 107,1177.

Ta

H

CH3

H

H

Hn-

n = 2 or 4

(CH3) acceptor orbitals of the metal

multiple bonding between Ti (d0) centerand the Me group.

(C-H) acceptor orbitals of Ti

C

TiP

PCl

H

H

H

Cl

Cl

21


23/47

The Valence-Bond Approach

A valence-bond approach

1) The valence electron (VE) count ( 16 VE) of the transition metal M.

2) The Lewis acidity and charge of M.

3) The degree of steric congestion at the M, measured primarily by itscoordination number (CN); and

4) The presence of an available acceptor orbital of suitable symmetry at M.

Valence-bond approach suffers from poor predictive power.

22


24/47

Experimental and theoretical studieshave unambiguously shown topossess no agostic interactions of any

significance.

Fig 1. Molecular structures of MeTiCl3(gas electron diffraction)

Fig 2. molecular structure of MeTiCl3(dmpe)(neutron diffraction study)

Agostic methyl configuration wasconfirmed by a neutron diffraction studyon a single crystal.

1. VE count 8.

2. CN 4.

1. VE count 12.

2. CN 6.

Limitations of Valence Bond Approach

23


25/47

A Revised Bonding Model for d0 System

1. The agostic interaction is a phenomenon driven by electron delocalization.

2. Hyperconjugative delocalization of the M-C bonding electrons.

3. This delocalization causes a global bonding redistribution within the metal

alkyl fragment.

geometrical distortions and significant changes in force constants.

The development of an agostic interaction is assisted by:

A softening of the M-C-X (X = C or H) bending potential and a flexible MLn

coordination geometry, each of which facilitates the canting of the alkyl group.

The presence of sites of locally enhanced Lewis acidity.

Scherer, W.; McGrady, G. S.Angew. Chem. Int. Ed.2004, 43, 1782. 24


26/47

1. Agostic interactions in d0 transition-metal complexes do not rely on the

presence of 3c-2e interactions between the metal atom and C-H bonds.

Contrasts between New and Previous models

3. It is not the total but the local Lewis acidity at the metal center which controlsthese agostic interactions.

2. The bonding between the metal and an agostic alkyl group is accomplished

mainly by one bonding orbital and its associated pair of electrons.

Scherer, W.; McGrady, G. S.Angew. Chem. Int. Ed.2004, 43, 1782.

No C-H activation

VE count 16 is not important

Ligand environment is important

25


27/47

Examples as Evidence of Revised Bonding Model

1. -agostic interactions are generally stronger than their- or-counterparts; and

2. An acute M-C-C angle -agostic interaction

The gas electron diffraction structure of

CH3CH2TiCl3 and both X-ray and neutron diffraction

data for CH3CH2TiCl3(dmpe) are now well

established.

How important is VE count and CN around the metal centre?

V. E. = 8 V. E. = 12

no dependence on the precise location of H atoms.

H2C

TiCl

Cl

Cl

H

H

HH2C

Ti

HH

H

P

PCl

Cl

Cl

anagostic agostic

26


28/47

Fig. A) Gas electron diffraction (at 293 K) [EtTiCl3]

B) The molecular structures of [EtTiCl3(dmpe)] based on a high

resolution charge density study (sin/l=1.097 -1; T=105 K)

Calculated values at the B3LYP/6-311G(d,p) level with

additional f-polarization function for Ti.

A B

Structural Studies of CH3CH2TiCl3 and CH3CH2TiCl3(dmpe)

TiH

= 2.23

27


29/47

Vibrational Spectroscopic Studies

The IR spectrum of the gaseous [CH3CH2TiCl3] [Fig. (a)]

(CH) 28003100 cm-1

.

The spectrum of the CHD2CD2 isotopomer, [Fig. (b)]

is (CH) 2953 cm-1 (A) and

is (CH) 2927 cm-1 (B)

D2C

TiCl

Cl

Cl

H'

D''

D''

D2C

TiCl

Cl

Cl

D'

H''

D''

Fig. IR spectra of (a) [CH3CH2TiCl3]

and (b) [CH3CH2TiCl3]

Normal or anagostic structure for [CH3CH2TiCl3]

G. S. McGrady et al. Chem. Commun.1997, 1547. 28


30/47

The IR spectrum of solid [CH3CH2TiCl3(dmpe)] [Fig.

(CHterminal) 28003100 cm-1

(CHagostic) 2615 cm-1 (C).

The spectrum of the CHD2CD2 isotopomer, [Fig. (d)] -

is (CHterminal) 2933 cm-1 (D) and

is (CHagostic) 2585 cm-1 (E)

The IR spectral pattern clearly shows an agostic structure for [CH3CH2TiCl3(dmpe)]

Vibrational Spectroscopic Studies

Fig. IR spectra of (c) [CH3CH2TiCl3(dmpe)]and (d) [CHD2CD2TiCl3(dmpe)]

2615 cm-1

2585 cm-1

G. S. McGrady et al. Chem. Commun.1997, 1547.

D2C

Ti

D''D''

H'

P

PCl

D2C

Ti

H''D''

D'

P

PClCl

Cl

Cl

Cl

29


31/47

NMR Spectroscopic Studies

/C (CH3)

ppm

(CH2)

ppm

J(C-H)

Hz

J(C-H)

Hz

CH3CH2TiCl3 0 2.0 3.3 134.3 130.0CH3CH2TiCl3(dmpe] 0 2.65 2.61 147.3 125.7

[CH3CH2TiCl3(dmpe)]

/C (CH3)

ppm

(CH2)

ppm

-90 2.70 2.53

-70 2.70 2.55

-50 2.70 2.58

-30 2.67 2.61

0 2.65 2.65

20 2.61 2.69

40 2.57 2.76

[CH3CH2TiCl3(dmpe)]CH3CH2TiCl3

Swang et al. J. Am. Chem. Soc.1998, 120, 3762.

D2C

Ti

H''H''

H

P

PClCl

ClH2C

TiCl

Cl

Cl

H'

H''

H''

30


32/47

CH3CH2TiCl3 and CDH2CH2TiCl3

= (CH3) - (CH2D)0.019

And varies slightly with temperature

CH3CH2TiCl3(dmpe) and CDH2CH2TiCl3(dmpe)

= (CH3) - (CH2D)

varies from0.05 to0.01 in the temperature

range 193-313 K.

No significant differences are observed in eitherH or1JC-H for the -methyl group

of CH3CH2TiCl3(dmpe) compared with CH3CH2TiCl3.

Isotopic Perturbation of Resonance (IPR) Studies

H2C

TiCl

Cl

Cl

H'

H''

D''H2C

TiCl

Cl

Cl

H

H''

H''

H2C

Ti

H''D''

H

P

PClCl

ClH2C

Ti

H''H''

H

P

PClCl

Cl

31


33/47

DFT Calculations and Frontier MO Analysis

The LUMO of [TiCl3]+ corresponds to a nearly pure dz

2 metal atom orbital.

The HOMO of [C2H5]- resembles an sp3 lone pair orbital on C, considerablydelocalized.

The main bonding interaction is between Ti and C, with a slight antibonding

interaction between the metal atom and the C-H moiety.

W. Scherer et al. Chem. Eur. J.2002, 8, 2324.

Fig. Frontier orbital interaction diagram and

shape of the HOMO and LUMO for [EtTiCl3]

(KohnSham orbitals based on B3LYP/6-311g(d,p) calculations).

32

DFT C l l ti d F ti MO A l i


34/47

Fig. Frontier orbital interaction diagram and shape of the HOMO and LUMO for

[EtTiCl3(dmpe)], espectively (KohnSham orbitals based on B3LYP/6-311G(d,p) level.

The binding of dmpe in CH3CH2TiCl3(dmpe), lengthens the corresponding

Ti-C and Ti-Cl bonds (by ca. 0.06 and 0.2 , respectively),

This canting of the ethyl group is then driven by a positive bonding

overlap between the torus of the Ti dz2 orbital and the C atom.

DFT Calculations and Frontier MO Analysis

HOMO LUMO

33


35/47

1. The geometry of the metalethyl group is such that bonding to both

Cand C is effected by the same orbital on M.

2. The -agostic interaction can be understood only in terms of a Ti-C

bonding orbital that is delocalized over the entire alkyl group.

3. Thus, electron delocalization, rather than MHC donation, appears

to be the primary driving force and the VE count is not a controlling

factor.

Summary of DFT Calculations and Frontier MO Analysis


34


36/47

Fig a) Theoretical bond paths and 2(r) contour map in the Ti-C-C-H planefor [CH3CH2TiCl3(dmpe)] (B3LYP/6-311G(d,p) ) b) Experimental bondpaths and 2(r) contour map for [CH3CH2TiCl3(dmpe)].

a b

Atoms In Molecules (AIM) Approach

No bond CP is evident between Ti and Hatom.

TiH interacrtion plays a minor role in the overall agostic interaction.

The existence of a MHbond CP is not a reliable criteria.


H2C

Ti

HH

H

P

PCl

35


37/47

M-C bonding region:

CH3CH2TiCl3 CH3CH2TiCl3(dmpe)

CP (r) (e-3) 0.84 0.66

Bond length (Exp.) () 2.090 2.154

C-C bonding region:

Analysis of the Laplacian 2 (r)and Local Energy Density H(r)

CH3CH2TiCl3 CH3CH2TiCl3(dmpe)

CP (r) (e-3) 1.576 1.607

Bond length (Exp.) () 1.526 1.513


H2C

TiCl

Cl

Cl

H

H

H

H2C

Ti

HH

H

P

PClCl

Cl

36


38/47

Driving Force Delocalization of M-C bonding electrons

Delocalization is evident if bond ellipticity is traced along the full C-C bond path.

Bond ellipticity 0partial character in a bond, or electronic distortion away from symmetry along

the bond path.

1. The ellipticity profile of [CH3CH2TiCl3(dmpe)]along C-C bond path reveals significant

character.

2. On evidence of its ellipticity profile system

is very distorted.

3. Two maximums with

max = 0.17 and max = 0.03Figure: Calculated ellipticity profile

W. Scherer et al. Chem. Eur. J.2002, 8, 2324. 37

Ellipticity Profile


39/47

Ellipticity Profile

The maximum max= 0.17 is located

closer to the -carbon atom

At the bond CP = 0.10.

Second maximum close to C (max = 0.03).

This second maximum

hypervalent character induced by an additional TiC interaction.

The ellipticity profiles of C-C bonds may be used in a general manner to reveal

delocalization of M-C bonding electron density into the C-C bonding region.

Fig: Calculated ellipticity profile

38


40/47

1. The existence of a very weak TiH interaction,

2. An increase in the C-C bond order, and

3. Weakening of the M-C and C-H bonds by hyperconjugative

electron delocalization.

The analysis of agostic complex [CH3CH2TiCl3(dmpe)] relative to anagosticcomplex CH3CH2TiCl3, reveals

Summary of the AIM analysis

39

M i l ti f A ti I t ti


41/47

Manipulation of Agostic Interactions:

Metal Polarization and the Extent of Delocalization

Analysis of the electronic structures of a series of complexes

[EtTiCl2(L)]+ where L is

H2C

Ti

C

L

Cl

H

HH

Cl

+

1. a strong -acceptor (CO or PF3).

2. a weak -acceptor (PMe3).

3. a -donor (H, CH3 or NMe3) or

4. a -donor ligand (Cl, F or OMe2).


[EtTiCl2]+ have the electronic features of metal-ethyl bonding in EtTiCl3(dmpe)

avoids complicaions from steric factors and

shows more pronounced -agostic interaction than does EtTiCl3(dmpe)

40

M i l ti f ti i t ti


42/47

Manipulation of agostic interaction

1. All -acceptor ligands display acute C-Ti-L.

2. Thus, the -acceptor ligands face directly cis-LICC(2), which corresponds to a

site of locally reduced Lewis acidity.

3. This site is avoided by all -donor or-donor ligands, which prefer instead to

coordinate between cis-LICC(2) and trans-LICC - a site of locally enhanced

Lewis acidity at Ti.

41

Differences for accepting and donor ligands


43/47

Differences for-accepting, - and -donor ligandsThe magnitude of the charge

concentration LICC(1) decreases with

increasing -acceptor character of L.

cis-LICC(1) = 309 e-5(L=PMe3)

cis-LICC(1) = 282 e-5(L=PF3)

cis-LICC(1) = 299 e-5(L=CO)

shift of LICC(1) towards BCC.

L = NMe3

L = OMe2

Agostic interaction is no longer favored in - and -donor ligands

BCC

trans-LICC

C C

H

Shift direction

cis-LICC(1)L = -acceptor

L = and-donor

cis-LICC(2) TiAtomic centre

X

~ 0.4

Fig. Ligand-Induced polarization of Ti atom

in [EtTiCl2 L]+

82.0

80.1

The pronounced local Lewis acidity of the Ti site facing C is reduced.

42


44/47

1. -acceptor ligands trans to the -C-H favors agostic interaction.

- and -donor ligands trans to the -C-H disfavor the interaction.

2. Not global but the locally induced sites of increased Lewis acidity

3. Controlled by local ligand effects the polarization of the metal center.

Manipulation of agostic interaction

Favors agostic interaction

43


45/47

A Note on dn (n 0) Agostic Complexes

1. Ellipticity profiles resemble closely those of alkene complexes.

2. These systems are close to -hydride elimination, possessing a high degree of

C=C and M-H bond character.

3. The occupied C-H bonding orbital acts as a donor and the empty *(C-H) orbital

acts as an acceptor.

4. Significant C-H bond activation is found (C-H elongation > 0.1) than in d0

transition-metal alkyl complexes.

LnM

H2C

CH3

LnM

H2C

CH2

H

LnM

H2

C CH2

H

I II III

44

C l i


46/47

Conclusion

1. Agostic stabilization in early d0 transition-metal alkyl complexes has little or no

dependence on MHC electron donation, but it arises rather from thenegative hyperconjugative delocalization of the M-C bonding electrons.

2. The bonding between the metal atom and the ethyl group is effectively

stabilized by one electron pair/molecular orbital.

3. Analysis of CCs and of bond ellipticities provides a novel and general method

for quantifying the extent of electron delocalization.

4. An understanding of the way in which the ancillary ligands induce polarizationat the metal center, affords the possibility ofpredictingand hence controlling

and manipulating the development of an agostic interaction.

45


47/47

Thank you

agostic bani seminar2

Documents