Head

Mechanism, Stereochemistry and Applications of Homogeneous Ziegler-Natta Polymerization

T. WilsonSED Group Meeting

Jan. 27th 2008

Poly-Olefins in Industry and Society

The three major classes of polyethylene:-High density polyethylene (HDPE) is a linear, semi-crystalline homopolymer of ethene-Linear low-density polyethylene (LLDPE) is a random co-polymer of ethylene and alpha-olefins-Low density polyethylene (LDPE) is a branched ethylene homopolymer

Poly-olefins -Fastest-growing segment of the polymer industry-World-wide production is in excess of 160 billion pounds/year

Uses of Poly-olefins-Vary from asphalt blends and sealants (amorphous) to surgical prostheses (Linear, Ultra-HDPE)

Polymer Properties

Isotactic

Syndiotactic (antiotactic)

Isotactic-Atactic Stereo-block

Atactic

x y

Stereochemistry

MeCat.

• Polymer tacticities, connectivities, and lengths

- Goal of modern polymer chemistry is to control all factors of a polymers microscopic structure

Regio-irregular

Regio-regular

Connectivity

n m

Degree of Polymerization

Historical Work of Ziegler and Natta

• Early polymerization w/ aluminum: The “Aufbaureaktion”

Ziegler, K et. al. Angew. Chem. 1955, 67, 425

Al(Et)3 H2C CH2

(100 atm.)

100 °CAl

(C2H4)m-C2H5

(C2H4)n-C2H5

(C2H4)o-C2H5

O2, H2O

m,n,o < 100

(C2H4)m-C2H5HO

m = 6-8- Large scale production of detergents

• Polymerization of ethylene

• Natta!s Contribution

-high molecular weight, Linear PE-low pressures and temperatures

H2C CH2

EtAlCl/TiCl4

n1-5 atm, 23 - 60 °C

EtAlCl/VCl4

n1-5 atm, 23 - 60 °C

Natta, G. et. al J. Am. Chem. Soc. 1955, 77, 1708.

-1st example of a crystallineisotactic polypropylene

Nobel Prize in Chemistry to Ziegler and Natta in 1963

Homogeneous Metallocene Catalysts

Jan. 5 , 1959 SOLUBLE CAWLYSTS FOR ETHYLENE POLYMERIZATION 83

I .1

n-Heptane

aluminum chlorides. Several things are obvious. Foremost is the fact that in many examples these catalysts were fully as active as a typical Ziegler catalyst prepared from titanium tetrachloride.

TABLE I

EFFECT OF ALUMINUM : TITANIUM RATIO ON POLYMERJZATION

OF ETHYLENE WITH A SOLUBLE CATALYSTO Pressure

2 0 . 5 50 1.9

4 <0.5 50 1 . 2

6 2.75 50 1.9

10 1 . 5 50 1.6

2 5 49 4 . 0

drop b AI A1:Ti Time, lb,/sq.

Solvent compound ratio hr. I D . R.S.V.6

limoles of bis-(cyclopentadieny1)-titanium dichloride and ten millimoles of diethylaluminum chloride in one liter of toluene. After the solution had turned blue, ethylene containing 0.003% oxygen was passed in a t atmospheric pressure and 15' solution temperature. A slow polymerization took place and 17.8 g. of polymer was formed in 80 minutes; the rate was not increased by the addition of diethylaluminum chloride. If, however, ethylene containing 0.025y0 oxygen was introduced, the solution turned a greenish-brown ; the addition of diethylaluminum chloride to main- tain this color resulted in the formation of 300 g. of polymer in four hours. Actually, almost any desired rate of polymerization can be realized by a judicious balance of oxygen and alkylaluminum compound. On the assumption that the oxygen was acting to convert trivalent titanium to tetravalent, a polymerization was carried out by adding diethylaluminum chloride to a solution of the tetravalent compound, bis-(cyclopentadieny1)- titanium dichloride, in toluene while passing in ethylene containing 0.003% oxygen. Polymeriza- tion took place smoothly and 125 g. of polymer was formed in 90 minutes.

These results leave little doubt that some tetravalent titanium must be present for these catalysts to show their high activity.13 It remains to be proved, however, whether a tetravalent titanium species is a sole requirement, or whether perhaps a combination of tetravalent and trivalent titanium is necessary for high activity; this question is under active investigation.

These results shed considerable light on the puzzling behavior observed in our early work. Undoubtedly, oxygen reacts with the trivalent titanium species to oxidize it to a tetravalent form. This reacts with the alkylaluminum compound to form an active catalyst, which probably is reduced; ;.e., an oxidation-reduction system is set

up. If the alkyl compound is too highly active, it either reacts with the oxygen preferentially or reduces the Ti(1V) too rapidly and a good catalyst does not result; this serves to explain the variable results with triethylaluminum, which reacts with oxygen very rapidly.

It is our belief that the active catalyst is a species such as (CsH&TiEtCl.EtAlCL or a reaction product of this with ethylene, and that the polymer grows by insertion of a monomer molecule between the alkyl group and the titanium. The

C1 C2Hs CI C,Hi, a+/ \6-/ ,/ ' /

\ \ ~ C ~ H , ) , T I AI-CI + (CsHs)>T1 AI-CI

C4Hr CI

first step in the polymerization would probably in- volve a n-type complex between the titanium and the olefin; if our structure is correct, the aluminum

(13) Natta and co-workersev" have indicated recently that soluble complexes containing trivalent titanium will polymerize ethylene

slowly at Q 5 O and 40 atmospheres pressure.

(14) G. Natta, P. Pino, G. Mazzanti, U. Giannini, E. Mantica and M. Peraldo. Chim. e ird. (Mi lan) , 86, 19 (1957); J . Pdymcr Sr i . . 26.

120 (1967).

• Synthesis of Ti(III) metallocene/Al complex

TiCl

Cl

5 mmol

Cl Al

10 mmol

H2C CH2 (g)

w/ 0.025 mol% O2

1 L Toluene, 15 °C, 1h n

-174 g PP-Similar activity to Ziegler Catalysts-Linear polymer (0.05% Methyl content)

- Authors propose that oxygen oxidizes the Ti(III) to the active Ti(IV) catalyst- Raises question of the role of the aluminum activator- Earliest example of homogeneous metallocene cat.

TiCl

ClCl Al

Heptane Blue Crystaline Solid (Air Sensitive)

-EA, EPA, and Colorimetric Analysis suggest Ti(III)/Aluminum complex(C5H5)2Ti(III)Cl•1/2(C2H5)2AlCl•1/2C2H5AlCl

Cold

Breslow, D et. al. JACS, 1959, 81, 81.

2

Intermittent-Growth Model

• Kinetic studies of Reichert, Fink and Eisch

- Kinetic studies on alkylaluminum activated metallocene catalysts revealed that polymer growth alternated b/w dormant and active states.

PnM

ClCp

CpAl

Cl

Cl

Et

Lewis Acid/BaseAdduct

ClM

ClCp

CpPnM

Cp

Cp

ClAl

Cl

Cl

Et

Dormant Active

Tight ion pair

Pn+1M

ClCp

CpAl

Cl

Cl

Et

H2C CH2Cl2AlEt

H.-H. Brintzinger et al. REVIEWS

)I: = C'-P,-, ---) C'-P, + C'-P,,, - C-P,-, c-P,

Scheme 4. "Intermittent-growth" model involving equilibria between polymer- bearing. but inactive primary complexes (C-P,) and active catalyst species (C*-P.),

generated by excess alkykdlumlnum halide, as proposed by Fink and co-workers

[31]. AI2 = (AIEtCl,),, Al; = unknown. P. = polymer chain withn monomer units;

C-P,, here Cp,TiP,,CI ' . ' AICI,Et.

reactive, olefin-separated ion pairs by displacement of an alumi-

nate anion from the metal center. At any rate, the limitation of

homogeneous catalyst systems to the polymerization of only

ethene was a crucial obstacle for progress in this field for many

years. Fortunately, this impediment was overcome by a series of

serendipitous observations,[37- 391 which led, around 1980, to

the discovery by Kaminsky, Sinn, and co-workers that

metallocenes are activated for the catalytic polymerization of

propene and higher olefins by methyl alumin~xanes.[~". 391

2.2. Polymerization of Propene and Higher Olefins

Water, which had long been considered to be a "poison" for

Ziegler-Natta catalysts, was first reported by Reichert and

Meyer to cause a surprising increase in the rate of ethene poly-

merization by the catalyst system Cp,TiEtCl/AIEtCl, .[371 Sub-

sequent studies by Long and Breslow on the effects of water in

the otherwise inactive system Cp,TiCl,/Me,AICl led to the no-

tion that formation of a dimeric aluminoxane, ClMeAl-0 - Al-

CIMe, by partial hydrolysis of Me,AICI might generate an ex-

ceptionally strong Lewis acid and, hence, a potent activator for

Cp,TiMeCI toward ethene polymerization.[381

While studying halogen-free systems such as Cp,ZrMe,/

AIMe,, Sinn and Kaminsky noticed that addition of water im-

parts to this otherwise inactive reaction system a surprisingly

high activity for ethene polymerization which was, furthermore, unprecedentedly constant over extended reaction b1

Sinn and Kaminsky observed that an interaction between

Cp,ZrMe, and AIMe, occurred only when water had been

added. The suspected formation of methyl aluminoxane (MAO)

by partial hydrolysis of AIMe, was subsequently supported by

its direct synthesis and characterization as a mixture of

oligomers of approximate composition (MeAIO), . Activation

of Cp,ZrMe, and Cp,ZrCI, with preformed M A 0 did indeed

yield exceedingly active catalysts for the polymerization of

ethene.[39b1 Similar activities were obtained with MAO-activat-

ed Cp,TiCI,; however, at temperatures above 0 "C this catalyst

system is rapidly deactivated, most likely by reduction to the

Ti"' stage.[401

Sinn, Kaminsky, and co-workers noticed furthermore that

MAO-activated homogeneous metallocene catalysts were-in

contrast to previously studied metallocene catalysts activated by

aluminum halides--capable of polymerizing propene and

higher ole fin^.^^". 39c-g3 Although the achiral metallocene

catalysts were still lacking the stereoselectivity of heterogeneous

Ziegler-Natta systems, aluminoxane-activated metallocene

catalysts now came to be most promising as model sys-

tems.

While oligomeric alkyl aluminoxanes have been known for

more than 30 years, for example as initiators for the polymeriza-

tion of o~iranes,[~' l their exact composition and structure are

still not entirely clear. When the hydrolysis of AIMe,, which is

is conducted under controlled conditions, it appears to generate

mostly oligomers Me,AI-[0-AIMe],-OAIMe, with n z 5 - 20,[39jI

Investigations in quite a number of research groups by

cryoscopy, UV, vibrational and NMR spectroscopy, chro-

matography, and other means[38. 3 9 j , 4 2 - 5 0 c 1 yield the following

picture for aluminoxane solutions. Residual AIMe, in MA0

seems to participate in equilibria that intercon-

vert different M A 0 oligomers[42b.43-461 and possibly also

cyclic and branched olig~mers.[~~'-j . 461 Cross-linking by

methyl-free oxoaluminum centers has been proposed to gener-

ate a microphase with an A1,0, core.[471 Aluminoxane clusters

[RAI(p,-0]),, with R = tevt-butyl and n = 4, 6, or 9, have been

isolated and structurally characterized by Barron and his

Complexes with four-coordinate A1 centers seem to

predominate in M A 0 s o l ~ t i o n s [ ~ ~ " . ~ ~ ~ and might contain in-

tramolecular A1,O +Al or AI-CH, +Al br idge~. [~ '~1 The pres-

ence of three-coordinate A1 centers in M A 0 solutions has been

deduced by Siedle and co-workers from "A1 NMR data.[sOb.cl

While species of exceptional Lewis acidity are certainly present

in MA0 solutions, their exact composition and structure is still

not adequately understood.["]

When toluene solutions of Cp,ZrCI, are treated with MAO,

a fast, initial ligand exchange reaction generates primarily the

monomethyl complex Cp,ZrMeCI 44b1 excess M A 0 leads

to Cp,ZrMe, .[39i1 These systems become catalytically active

when the concentration of excess M A 0 is raised to A1:Zr ratios

of about 200:l or The ways in which excess M A 0

induces this activity have been investigated largely by spectro-

of the Al centers in MA0 have an exceptionally high propensity

to abstract a CH; ion from Cp,ZrMe, and to sequester it in a

weakly coordinating ion CH,-MAO-. A fast, reversible trans-

fer of 13CH, groups from Cp,ZrMe, to the Al centers of a M A 0

activator was observed by Siedle et al.[sOb,cl Barron and co-

w o r k e r ~ ~ ~ ~ ~ ' obtained NMR spectroscopic evidence that

Cp,ZrMe, and alumoxane clusters like (p,-O),Al,tBu, form

complexes of the type [Cp,ZrMe+ . . . (p,-O),Al,(tBu),Me-] in

[DJtoluene solution, which polymerize ethene. The tendency of

four-coordinate Al centers in these aluminoxane clusters to ab-

stract a methyl anion is ascribed by these authors to the relief of

ring strain upon formation of the methyl complex.

91Zr and 13CNMR spectra of Cp,ZrMe,/MAO solu-

t i o n ~ [ ~ ~ ~ . ~ ~ and solid-state XPS[521 and 13C NMR[53"1 studies

indicate formation of a cation [Cp,ZrR]+, which is most likely

stabilized by coordinative contact with its CH, -MAO- counter-

ion, for example through bonding like that in A1,O + Zr or

AI-CH, + Zr. These contacts appear to give way, in the pres-

ence even of substituted olefins, to olefin-separated ion pairs

[Cp,ZrR(olefin)]+ CH, - MAO-, the presumed prerequisite for

olefin insertion into the Zr-R bond. This hypothesis-that the

unusually low coordinating capability of the anion A- in the ion

pair [Cp,ZrMe]' A- is crucial for catalytic activity[sOel-led to

the discovery of a series of highly active cationic metallocene

highly exothermic (and indeed potentially 42b1 1,

scopic methods,[39i. 5 0 - 5 2 , 53al It is ' generally assumed that some

1146 Angcw. Chem. In l . Ed. EngI. 1995, 34, 11 43 - 11 70

- Lead to the development of the intermittent-growth model

Al2 = (AlEtCl)2Al' = (unknown)Pn = Polymer chain

C = Inactive stateC* = Active state

Riechert, K. H. angew. makromol. chem. 1970, 12, 175

Activator-Free Polymerization

J . Am. Chem. SOC. 1986, 108, 741 1-7413 741 1

'11

1'7

Figure 1. Structure of the Cp,Zr(CH,)(THF)+ cation. The BPh4- structure is normal. Important bond lengths (A) and bond angles (deg) are as follows: Zr-C(35) 2.256 ( I O ) ; Zr -O 2.122 (14); Zr-Cp(av) 2.487 (41); Zr-CNT(av) 2.174; O-C(36) 1.443 (27); 0-C(39) 1.464 (25); CNT-Zr-CNT 129.6; C(35)-Zr-0 97.4 (5); C(39)-0-C(36) 104.5 (16); C N T indicates the centroid of a C p ring.

Figure 2. Structure of the Cp,Zr(CH,)(THF)+ cation viewed down the 0-Zr bond. The P-carbons of the T H F ligand have been removed for clarity.

Cp2Yb(CH3)(THF) (4), in which steric effects should dominate directional bonding effects, the T H F ligand lies nearly in the Me-Yb-0 plane (dihedral angle = I6").l6 This difference suggests that there is a significant T component in the Zr-0 bond of 3." Consistent with this proposal, after correction for metal size difference,I8 the M-CH, and average M-Cp distances in 3 are ca. 0.04 8, greater than in 4, while the M - 0 distance in 3 in ca. 0.04 A shorter than in 4.19-22

(16) Evans, W. J.; Dominguez, R.; Hanusa, T. P. Organometal/ics 1986, 5, 263.

(17) In this orientation the T H F 0 p orbital can overlap with the l a , Zr LUMO. Lauher, J . W.; Hoffmann, R. J . Am. Chem. SOC. 1976, 98, 1729 and references therein.

(18) Yb" is ca. 0.146 A larger than Zr4+. Shannon, R. D. Acta Crys- tallogr., Sect. A 1976, A32, 751.

(19) (a) Other structural comparisons su port the proposed Zr-0 ?r-

bonding in 3. The Zr-O bond of 3 is ca. 0.09 1 shorter than the Zr-O bond of Cp,Zr(?2-CPh,0CH,)C1196 and 0.26 A shorter (after correction for the difference in 0 and N radii) than the Zr-N bond of (C5Me5),Zr(pyri- dine)(C,0-s2-OCCH2).19C (b) Erker, G.; Dorf, U.; Atwood, J. L.; Hunter, W. E. J . Am. Chem. SOC. 1986, 108, 2251. (c) Moore, E. J.; Straus, D. A,; Armantrout, J.; Santarsiero, B. D.; Grubbs, R. H.; Bercaw, J. E. J . Am, Chem. SOC. 1983, /05, 2068.

(20) The chemical significance of Zr-0 ?r-bonding in 3 is under investigation. For example, it provides an explanation2' for the observation that 3 undergoes hydrogenolysis slowly (2 days) under conditions ( 1 atm of H,, room temperature) where the analogous PMe, complex Cp2Zr(CHp)(PMe3)+ undergoes instantaneous reaction.22

(21) Gell, K. 1.; Posin, B.; Schwartz, J.; Williams, G. M. J . Am. Chem. SOC. 1982, 104, 1846.

(22) Jordan, R. F.; Bajgur, C. S.; Dasher, W. E.; Scott, B.; Rheingold, A. L., submitted for publication in Organomefallics.

0002-7863/86/1508-7411$01.50/0

In CH2C12 solvent, cationic complex 3 is an ethylene polymerization catalyst, producing linear polyethylene with a minimum activity of ca. 0.2 g/(mmol of catalyst) min atm at 25 "C and 1-4 atm of ethylene.23 This activity is relatively low" due to the presence of the T H F ligand which, though labile, competes with ethylene for the Zr coordination site. Addition of T H F or other donor ligands slows the polymerization rate dramatically and in T H F or CH3CN solvents no activity is observed.24

The observation of ethylene polymerization by 3 supports the original Long-Breslow-Newburg mechanism for olefin polymerization by the soluble Ziegler-Natta systems. It also satisfies a necessary condition of the proposal3 that cationic complex 2 is the active species; Le., that Cp2M"R+ complexes prepared in the absence of AI cocatalysts should polymerize ethylene. Detailed spectroscopic studies of this and related systems are in progress.

Acknowledgment. Support from the Research Corporation, the donors of the Petroleum Research Fund, administered by the American Chemical Society, and the Washington State University Research and Arts Committee is gratefully acknowledged. The X-ray diffraction system was acquired with the aid of NSF grant CHE8408407 and the Boeing Co.

Supplementary Material Available: Summary of X-ray data collection parameters; ORTEP diagram of BPh4-; stereoview of [Cp2Zr(CH3)(THF)] [BPh,]; and tables of atomic coordinates and isotropic thermal parameters, bond lengths, bond angles, aniso- tropic thermal parameters, H atom coordinates and isotropic thermal parameters (1 1 pages); tables of observed and calculated structure factors (13 pages). Ordering information is given on any current masthead page.

(23) Typical analysis for polymer produced under these conditions: M, = 18400; M, = 33000; M , / M , = 2.58.

(24) The THF-free species Cp2Zr(CH,)+ (5) (or its CH2CI, solvate) is generated as a transient intermediate in CH2CI, solution by reaction of C P , Z ~ ( C H , ) ~ with [(C5H,Me)2Fe](BPh,]. In the presence of THF, 5 is trapped as 3, while in the absence of potential ligands it decomposes principally to Cp,Zr(CH,)Cl. Generation of 5 in the presence of 1 atm of ethylene results in very rapid polymerization. The activity of 5 is considerably greater than that of 3, confirming the inhibition by T H F and indicating that T H F is not required for polymerization activity. Activity measurements for 5 and polymer characterization are in progress. Jordan, R. F.; Bajgur, C. S.; Echols, S. F., unpublished results.

The Trimethylphosphine Adduct of the Zirconocene-Benzyne Complex: Synthesis, Reactions, and X-ray Crystal Structure

Stephen L. Buchwald* and Brett T. Watson

Department of Chemistry Massachusetts Institute of Technology

Cambridge, Massachusetts 021 39

John C. Huffman

Molecular Structure Center Department of Chemistry, Indiana University

Bloomington, Indiana 47405

Received July 8, 1986

Although several q2-benzyne complexes of transition metals have been reported,2 there are only two examples of mononuclear benzyne complexes which have been structurally characterized. These are the Cp*TaMe2(benzyne) (Cp* = q5-CsMe,) complex reported by Schrockl and Bennett's 1,2-bis[ (dicyclohexyl- phosphino)ethane]nickel(benzyne) complex.2 The existence of

(1 ) McLain, S. J.; Schrock, R. R.; Sharp, P. R.; Churchill, M. R.; Youngs, W. J. J . Am. Chem. SOC. 1979, 101, 263. Churchill, M. R.; Youngs, W. J. Inorg. Chem. 1979, 18, 1691.

(2) Bennett, M. A,; Hambley, T. W.; Roberts, N . K.; Robertson, G. B. Organometallics 1985, 4, 1992 and references therein.

0 1986 American Chemical Society

ZrMe

Me Ag[BPh4] ZrNCMe

Me

BPh4

Ag0 C2H6

80% yield

(AgMe)

not obs.

THF

DCM, 23 °C, 1h nC2H4 (4 atm.)

[Cp2Zr(Me)(THF)][AgBPh4]

• Activity towards polymerization of ethene

-Activity of ca. 0.2g/(mmol cat.)

• Reduced activity attributed to ligand exchange b/w ethene and THF

• Study by Jordan and co-workers

Jordan, R. F. JACS 1986, 108, 7410.

The Role of Water: MAO Activators

• Interesting observation by Kaminsky and Sinn

TiMe

Me AlMe3

C2H4 (7.9 atm)H2O

nToluene, 21 °C

from bis(cyclopentadienyl)methyltitanium(lv) monochloride]

complex formation is almost quantitative"]. Surprisingly, how-

ever, not only halogen-deficient _but also halogen-free systems

exhibited P-hydrogen transfer (and a-hydrogen transfer, but

much more slowly) with formation of dimetall~alkanes['~.

NMR-spectroscopic investigation of such a reaction in the presence of ethylene revealed a weak but lasting insertion

of ethylene with formation of p~lyethylene '~~! Addition of traces of chloride in the form of bis(cyclopentadieny1)titani-

um(1v) chloride lowered the yield of polyethylene, initiated

the known reduction reaction to the expected extent and

accelerated it[s1. On the other hand, the yield increased to

600000 g polyethylene/g titanium and the activity to 400000 g

polyethylene/g titaniumlh when one equivalent of trimethyl-

or triethylaluminum previously treated with water was added

to dialkylbis(cyclopentadienyl)titanium(~v) [alkyl, e. g. methyl]

(cf. Table 1 and example 1).

sence of ethylene leads to slow degradation of the starting com-

pound (curve A3).

If, on the other hand, addition is carried out in the presence

of ethylene a long-wave absorbing species with a band maxi-

mum at 390nm (curves A4 and A,) is formed in the course

of a few minutes. The rate of formation of the compound

characterized by La, = 390nm is strongly dependent on the

concentration of the reaction partners. If trimethylaluminum previously treated with water is added

in the presence of propylene (Fig. 1 b) degradation of the

starting compound is observed. Polypropylene was not found.

On removal and replacement of the propylene by ethylene

the maximum arises at 390nm and polymerization of the

ethylene ensues. That is, the band at 390nm is formed only

on simultaneous presence of halogen-free bis(cyc1openta-

dienyl)titanium(iv) with water-treated trialkylaluminum and

ethylene (Fig. 1 c).

Table 1

Rater

Yield of Time Temp Tltdnium Amount of Re1 Mol Wt

concentration wdter ddded (viscometric)

[mol 'I] [mol~l]

tthylene polymerization with the system bis(cyclopentddienyl)dimethyltitdnium(lV), trimethylaluminum. and

.~ ~ ~~ ~~ ~~ -~ ~~ ~ ~~

polyethylene [h] ["CI

[gi l l ~~~ ._____~ ~~~ ~~ ~~ ~ ~ ~ _ _ _ ~

33.6 I12 21 7 x ~ o - ~ i 0 . 2 5 x 10-2 3 800000

43.2 I .5 21 7 x10-4 4.6 x l o - ' 200000

57.6 I .5 50 3.5x 10-J 4.6 x I0 90000

34.2 1.5 I 2 1 . 1 x 10-6 7.5 x 10 ' ~

~~~

Interestingly, compounds formed on the basis of bis(cyclo-

pentadienyl)zirconium(Iv) and exhibiting no polymerization

activity -even in the case of completely dehalogenated second-

ary product~[~~-can also become very active polymerization

catalysts by addition of trialkylaluminum previously treated

with water. These systems excel in that products with relative

masses between a few million and a few hundred can be

generated at almost constant activity by suitably choosing

the reaction temperature (Table 2).

Tahle 2. Control of molecular weight by choice of temperature i i i the polymeri-

7ation of ethylene with tri(cyclopentadienyI)rirconium -aluminum com-

pounds.

Yield of Time Temp. Zirconium Rel. Mol. Wt.

polyethylene [h] ["C] concentration (viscometric)

-~ ~~~

[niol I] ~ ~~~ ~~~ -

[g I1

81 64 50 2.2 x 10-3 1500000

45 48 , 60 2.2 x 1 0 - 3 363 000

45 64 70 2.2x lo - ' 225000

90 X8 80 2.2 x 1 0 - 3 40000

X 0 64 90 2.2x oils and waxes [a]

[a] Predoiiiinantly x-olefins of empirical formula Cz,,H4,, (,I = 2 , 3. 4...).

~ ~~~~

~ ~ ~.. ~-

Preliminary experiments show a maximum activity at an

A1 : H 2 0 ratio of 2: 1 to 5 : 1. Evidently water and trialkylalu-

minum react with each other, for the activity disappears if

the Al: H 2 0 ratio drops below 1 :3, i .e. all alkyl groups are

hydrolyzed. Presumably alumoxanes are formed"0'.

Aging processes were not observed, and in particular no

reduction which would be comparable to that observed under

the conditions in halogen-containing systems. If the solution of catalyst is mixed with polyethylene and

the solvent removed, "gas-phase polymerization" takes place

after admission of ethylene (see example 2).

We found that the spectrum of bis(cyc1opentadienyl)dimeth-

yltitanium(1v) (Fig. 1 a, curve A,) remains unaltered-aside

from dilution effects (curve A2)-on addition of trimethylalu-

minum (which shows no absorption) in the absence of ethylene. Addition of trimethylaluminum treated with water in the ab-

I

Fig. I . U V spectroscopic measurements (Car) IJI.

a ) Curve A , : 1.5 mmol Cp2Ti(CHr)2'liter toluene. Curve A 2 : A , + 13.4

minol AI(Cll.3).3.1iter toluene. CurveA3: A , +AI(CH3)3 treated with H z O

(14.4 inmol;liter). Recorded after 25 min. Curve A,: A> i n the presence

of etlqlene. Recorded after 5 inin. Curve A s : A+ recorded after 25

min.

b) Curve B , : 1.5 mmol C~,Ti (CH3)~/ l i te r toluene (=curve A , ) . Curve

B 2 : R, +saturation with propylene. Curve B 3 : B2 +AI(CH,)3 treated

with H20 (14.4 mniol#liter). Curve B,: B3 after 5 mins.

c ) Curve C , : 1.5 mniol Cp2Ti(C'H3)2,'liter toluene (=curve A , I . Curve

Cz: C , +A1(CH3), treated with H 2 0 (14.4 n~inol:liter) + \nfuration

with propylene. Recorded after 5 min (=curve BL). Cl ine C 3 : ('L. propy-

lene removed and ethylene admitted under pressure. Recorded after

7 min. Curve C1: C 3 , recorded after 16 min.

Experirnerital

Example 1 : Trimethylaluminum (4 x mol) and water

(2.3 x mol) are added at 12°C to 350ml toluene in a glandless 11 agitator autoclave after careful evacuation and

bake out. Bis(cyclopentadienyl)dimethyltitanium(rv) (3.7 x

lo- ' mol) is then introduced and the mixture allowed to react

with ethylene at 8 bar. The reaction is terminated after 90 min, and the polyethylene filtered off and dried. Yield : 11.4g.

Activity: 1.1 1 x lo6 mol ethylene per mol titanium compound

63 1

-Best activity observed

for Al/H2O ratios of 2:1 to

5:1

- For Al/H2O ratio of 1/3

activity dramatically drops

UV Data for rxn:A1: 1.5 µM Cp2Ti(Me)2 in Tol.

A2: A1 + 14.4 mmol Al(Me)3A3: A1 + Al(Me)3 pre-treated w/H2O

A4: A3 in the presence of C2H4A5: A4 after 25 minutes

from bis(cyclopentadienyl)methyltitanium(lv) monochloride]

complex formation is almost quantitative"]. Surprisingly, how-

ever, not only halogen-deficient _but also halogen-free systems

exhibited P-hydrogen transfer (and a-hydrogen transfer, but

much more slowly) with formation of dimetall~alkanes['~.

NMR-spectroscopic investigation of such a reaction in the presence of ethylene revealed a weak but lasting insertion

of ethylene with formation of p~lyethylene '~~! Addition of traces of chloride in the form of bis(cyclopentadieny1)titani-

um(1v) chloride lowered the yield of polyethylene, initiated

the known reduction reaction to the expected extent and

accelerated it[s1. On the other hand, the yield increased to

600000 g polyethylene/g titanium and the activity to 400000 g

polyethylene/g titaniumlh when one equivalent of trimethyl-

or triethylaluminum previously treated with water was added

to dialkylbis(cyclopentadienyl)titanium(~v) [alkyl, e. g. methyl]

(cf. Table 1 and example 1).

sence of ethylene leads to slow degradation of the starting com-

pound (curve A3).

If, on the other hand, addition is carried out in the presence

of ethylene a long-wave absorbing species with a band maxi-

mum at 390nm (curves A4 and A,) is formed in the course

of a few minutes. The rate of formation of the compound

characterized by La, = 390nm is strongly dependent on the

concentration of the reaction partners. If trimethylaluminum previously treated with water is added

in the presence of propylene (Fig. 1 b) degradation of the

starting compound is observed. Polypropylene was not found.

On removal and replacement of the propylene by ethylene

the maximum arises at 390nm and polymerization of the

ethylene ensues. That is, the band at 390nm is formed only

on simultaneous presence of halogen-free bis(cyc1openta-

dienyl)titanium(iv) with water-treated trialkylaluminum and

ethylene (Fig. 1 c).

Table 1

Rater

Yield of Time Temp Tltdnium Amount of Re1 Mol Wt

concentration wdter ddded (viscometric)

[mol 'I] [mol~l]

tthylene polymerization with the system bis(cyclopentddienyl)dimethyltitdnium(lV), trimethylaluminum. and

.~ ~ ~~ ~~ ~~ -~ ~~ ~ ~~

polyethylene [h] ["CI

[gi l l ~~~ ._____~ ~~~ ~~ ~~ ~ ~ ~ _ _ _ ~

33.6 I12 21 7 x ~ o - ~ i 0 . 2 5 x 10-2 3 800000

43.2 I .5 21 7 x10-4 4.6 x l o - ' 200000

57.6 I .5 50 3.5x 10-J 4.6 x I0 90000

34.2 1.5 I 2 1 . 1 x 10-6 7.5 x 10 ' ~

~~~

Interestingly, compounds formed on the basis of bis(cyclo-

pentadienyl)zirconium(Iv) and exhibiting no polymerization

activity -even in the case of completely dehalogenated second-

ary product~[~~-can also become very active polymerization

catalysts by addition of trialkylaluminum previously treated

with water. These systems excel in that products with relative

masses between a few million and a few hundred can be

generated at almost constant activity by suitably choosing

the reaction temperature (Table 2).

Tahle 2. Control of molecular weight by choice of temperature i i i the polymeri-

7ation of ethylene with tri(cyclopentadienyI)rirconium -aluminum com-

pounds.

Yield of Time Temp. Zirconium Rel. Mol. Wt.

polyethylene [h] ["C] concentration (viscometric)

-~ ~~~

[niol I] ~ ~~~ ~~~ -

[g I1

81 64 50 2.2 x 10-3 1500000

45 48 , 60 2.2 x 1 0 - 3 363 000

45 64 70 2.2x lo - ' 225000

90 X8 80 2.2 x 1 0 - 3 40000

X 0 64 90 2.2x oils and waxes [a]

[a] Predoiiiinantly x-olefins of empirical formula Cz,,H4,, (,I = 2 , 3. 4...).

~ ~~~~

~ ~ ~.. ~-

Preliminary experiments show a maximum activity at an

A1 : H 2 0 ratio of 2: 1 to 5 : 1. Evidently water and trialkylalu-

minum react with each other, for the activity disappears if

the Al: H 2 0 ratio drops below 1 :3, i .e. all alkyl groups are

hydrolyzed. Presumably alumoxanes are formed"0'.

Aging processes were not observed, and in particular no

reduction which would be comparable to that observed under

the conditions in halogen-containing systems. If the solution of catalyst is mixed with polyethylene and

the solvent removed, "gas-phase polymerization" takes place

after admission of ethylene (see example 2).

We found that the spectrum of bis(cyc1opentadienyl)dimeth-

yltitanium(1v) (Fig. 1 a, curve A,) remains unaltered-aside

from dilution effects (curve A2)-on addition of trimethylalu-

minum (which shows no absorption) in the absence of ethylene. Addition of trimethylaluminum treated with water in the ab-

I

Fig. I . U V spectroscopic measurements (Car) IJI.

a ) Curve A , : 1.5 mmol Cp2Ti(CHr)2'liter toluene. Curve A 2 : A , + 13.4

minol AI(Cll.3).3.1iter toluene. CurveA3: A , +AI(CH3)3 treated with H z O

(14.4 inmol;liter). Recorded after 25 min. Curve A,: A> i n the presence

of etlqlene. Recorded after 5 inin. Curve A s : A+ recorded after 25

min.

b) Curve B , : 1.5 mmol C~,Ti (CH3)~/ l i te r toluene (=curve A , ) . Curve

B 2 : R, +saturation with propylene. Curve B 3 : B2 +AI(CH,)3 treated

with H20 (14.4 mniol#liter). Curve B,: B3 after 5 mins.

c ) Curve C , : 1.5 mniol Cp2Ti(C'H3)2,'liter toluene (=curve A , I . Curve

Cz: C , +A1(CH3), treated with H 2 0 (14.4 n~inol:liter) + \nfuration

with propylene. Recorded after 5 min (=curve BL). Cl ine C 3 : ('L. propy-

lene removed and ethylene admitted under pressure. Recorded after

7 min. Curve C1: C 3 , recorded after 16 min.

Experirnerital

Example 1 : Trimethylaluminum (4 x mol) and water

(2.3 x mol) are added at 12°C to 350ml toluene in a glandless 11 agitator autoclave after careful evacuation and

bake out. Bis(cyclopentadienyl)dimethyltitanium(rv) (3.7 x

lo- ' mol) is then introduced and the mixture allowed to react

with ethylene at 8 bar. The reaction is terminated after 90 min, and the polyethylene filtered off and dried. Yield : 11.4g.

Activity: 1.1 1 x lo6 mol ethylene per mol titanium compound

63 1

Authors attributed increased activity to the formation of methyl aluminoxanes (MAO)

Kaminsky, W. et. al. Angew. Chem. Int. Ed. Engl. 1976, 15, 630.

MAO Activators (The “Black-Box”)

1D linear or cyclic chains(3-coordinate Al)

2D-cyclic structures(3- and 4-coordinate Al)

3D-clusters(3- and 4-coordinate Al)

- Prepared by the control hydrolysis of trimethylaluminum- Exact structure and composition are not well understood

- Multiple equilibria present in solution further complicate matters

-Generally accepted that the mechanism of activation involves abstraction of a chloride (L2ZrClMe) or a methyl anion (L2ZrMe2). - Large excesses are need (>200/1, Al/Ti)

Marks, T. et. al. Chem. Rev. 2000 100, 1391.

“Cation-Like” Homogeneous Polymerization

3624 J . Am. Chem. SOC., Vol. 113, No. 9, 1991

Q

Communications to the Editor

Figure 1. Molecular structure of [ 1 ,2-(CH,)2CSH3]2ZrCH3+ CH,B- (C6FJ),- (2). Important bond distances (A) and angles (deg) are as follows: Z r C ( 1 5 ) = 2.252 (4), Z r C ( 3 4 ) = 2.549 (3), B C ( 3 4 ) = 1.663 (5), Zr-H(34A) = 2.71 (3), Zr-H(34B) = 2.25 (3), Z-H(34C) = 2.30 (3), Z r C d n g = 2.500 ( I ) (av), C(15)-ZrC(34) = 92.0 ( I ) , Zr-C(34)-B = 161.8 (2), ring centroid-Zr-ring centroid = 131.3 ( I ) , C(16)-B-C(34) = 108.7 (3), C(22)-B-C(34) = 112.7 (2), C(28)-B-C(34) = 102.8 (3). C(16)-B-C(22) 106.5 (3). C(16)-B-C(28) = 114.3 (3), C(22)-B-C- (28) = 112.0 (3). Thermal ellipsoids are drawn at the 35% probability level.

The reaction of tris(pentafluoropheny1)borane with a variety of zirconocene dimethyl complexes proceeds rapidly and quan- titatively (by NMR) to yield, after recrystallization from hy- drocarbon solvents, methyltriarylborate complexes (eq 1). The

LLZT(CH3h + B(C6Fsh - J&CH,+CH3B(C&h-

1: L = q5-C!5HS

2: L = $-1,2-(CH&C5H,

(1) 3: L = $-(CH,)SCs

new compounds have been characterized by standard analytic/ spectroscopic technique^.'^ Particularly telling in the latter are quadrupolar-broadened IH/I3C NMR spectral features assignable to a CH3BR3- groupI4 and low-field Zr"CH3 signals previously associated with "cation-like" specie^.^^^ The diastereotopic ring CH and CH3 sjgnals in 213 indicate loss of the time-averaged ring centroid-Zr-ring centroid mirror plane in eq 1. Regarding the lability of B-CH3 complexation, N M R line broadening indicates AG* = 18.7 (2) and 19.7 (2) kcal/mol (80 OC) for intramolecular Zr-CH3/B-CH3 interchange in 1 and 2, respectively. However,

w6

mpm-

( ~ Y - ~ ~ F - F e 18.7 Hz. 6 F, m-F). Anal. Calcd for C3,H2,BIFI;Zr: C, 49.83; H, 2.95. Found: C, 49.69; H, 2.83. 3: NMR (C6D6, 25 "C), 'H, 6 1.37 (s, 30 H), 0.29 (s, 3 H, ZrCH,), -0.30 (s, br, 3 H, BCH,); I'C, 6 124.01 (s, Cp), 50.36 (q, 'Jc-H 121.5 Hz, ZrCH,), 14.34 (s, br, BCH,), 11.02 (9. IJC-H = 127.4 Hz, Cp-CH,); "B (C6D6 + THF-d8, 25 "C) 6 -14.01. Anal. Calcd for CaH36BlF,SZr: C, 53.16; H, 4.01. Found: C, 53.07; H, 3.87.

(14) (a) Nbth, H.; Wrackmeyer, B. Nuclear Magnetic Resonance Spec- troscopy of Boron Compounds; Springer-Verlag: Berlin, 1978; Chapters 4,7. (b) Onak, T. Organoborane Chemistry: Academic Press: New York. 1975: Chapter 2.

as evidenced by the high-temperature interchange of the diastereotopic ring signals, inversion of the dissymmetric ion pair structure occurs a t a slightly greater rate (AG* = 18.3 (2) kcal/mol (80 "C)) in 2.

The crystal structure of 215 (Figure 1) consists of a 'bent- sandwich" [ 1 ,2-(CH3)2C5H3]zZrCH3+ cation weakly coordinated to a CH3B(C6F5)3- anion via a nonlinear (161.8 (2)O), highly unsymmetrical Zr(p-CH3)B bridge. With the exception of a shortened Zr-C(l5) bond (cf., 2.273 (9, 2.280 (5) A in Cpz- Zr(CH3)2),16J7 key aspects of the Zr coordination sphere such as the angle ring centroid-Zr-ring centroid and Zr-C",,&av) are unexceptional (132.5O and 2.525 (12) A, respectively, in Cp2Zr(CH3)z16J7). The Zr-CH3(bridge) distance is elongated by ca. 0.3 & I 8 while the B-CH3 distance appears to be norma1,I" and the valence angles about B deviate only slightly from tetra- hedral. The C(34) hydrogen atoms are bent away from B and toward Zr, with the closest Zr-H contact (2.25 (3) A) exceeding typical terminal and bridging Zr-H bond distances (1.78 (2) and 1.94 (2), 2.05 (3) .&)I9 as well as a short Zr-.H(C,,,,) *agostic" distance (2.16 A).2o

Complexes 1-3 are active homogeneous catalysts for olefin polymerization. Using procedures described previously,21 ethylene polymerization proceeds rapidly at 25 OC, 1 atm pressure, with NJ1) = 45 s-I (-4.5 X lo6 g of polyethylene (mol of Zr)-I h-' atm-I), roughly comparable in activity to typical zirconocene/ alumoxane catalysts.lfJ The polyethylene produced is highly linear by NMR22 with relatively high molecular weight (A?, = 124000, Xfn = 61 200).23 With propylene2I a t 25 OC, 1 yields atactic (by NMR)24 polypropylene with N,(1) = 4.2 s-I (aw = 15 600, A?,, = 3000).23 NMR experiments in which a toluene-d8 solution of 1 was exposed to 10 equiv of propylene at -25 OC indicated that >70% of 1 undergoes olefin insertion under these conditions. This argues that the observed catalytic activity is not due to a minor component or impurity.

These results demonstrate the direct abstractive role of an organo-Lewis acid in the stoichiometric conversion of a zirconocene dialkyl to the corresponding "cation-like" zirconocene monoalkyl having high activity for homogeneous a-olefin polymerization.

(15) Crystal data: ZrF15C 4H24B; monoclinic space group P2Jn; a = 12.261 (2) A, b = 20.010 (6) A, c = 13.053 ( 5 ) A, ,9 = 90.80 (2)O at -120 "C; V = 3202 (2) A'; 2 = 4; d,, = 1.700 g cmd. The structure was solved by direct methods (SHELXS-86) and refined with weighted and unweighted difference Fourier syntheses and blocked-matrix least-squares (SHELX-76). R(F) = 0.027 and R,(F) = 0.029 for 3261 absorption-corrected reflections with I > 2.58u(I) measured on a CAD4 diffractometer (Mo Ka radiation, A = 0.71069 A, 28 = 45").

(16) (a) Hunter, W. E.; Hrncir, D. C.; Vann Bynum, R.; Penttila, R. A.; Atwood, J. L. Organometallics 1983,2,750-755. (b) In Cp,Zr(CH,)THP, Zr-CH, = 2.26 ( I ) A, Zr-Ci, = 2.49 (4) (av) A, and the angle ring centroid-Zr-ring centroid = 129.6"."

(17) For a compilation of L2ZrR2 structural data, see: Schock, L. E.; Brock, C. P.; Marks, T. J. Organometallics 1987, 6, 232-241.

(18) (a) Zr-(p-CH,) = 2.559 (7), A in [CpzZr(C,0q2-OCCHCH2C- (CH3),)]z[p-AI(CH3)2](p-CH3).18b (b) Waymouth, R. M.; Santarsiero, B. D.; Coats, R. J.; Bronikowski, M. J.; Grubbs, R. H. J. Am. Chem. Soc. 1986,

(19) Jones, S. B.; Petersen, J. L. Inorg. Chem. 1981, 20, 2889-2894. (20) Jordan, R. F.; Bradley, P. K.; Baenziger, N. C.; La Pointe, R. E. J .

Am. Chem. SOC. 1990, 112, 1289-1291. (21) (a) Jeske, G.; Lauke, H.; Mauermann, H.; Swepston, P. N.; Schu-

mann, H.; Marks, T. J. J . Am. Chem. Soc. 1985,107,8091-8103. (b) Jeske, G.; Schock, L. E.; Swepston, P. N.; Schwann, H.; Marks, T. J. J. Am. Chem. Soc. 1985, 107, 8103-81 10. (c) Ethylene polymerization experiments were carried out for 40-150 s in toluene solution with catalyst concentrations of -0.15 mM. (d) Propylene polymerization was camed out in a quartz Warden pressure vessel (1-5 atm pressure) for 0.5 h in toluene solution with catalyst concentrations of - 1.4 mM. In a typical experiment, 6 mL of propylene in 2 mL of toluene was converted into 3.0 g of polypropylene in 0.5 h.

(22) Encyclopedia of Polymer Science and Engineering; Wiley: New York, 1987; Vol. IO, pp 298-299.

(23) We thank Dr. J. C. Stevens of Dow Chemical Co. for GPC measurements.

(24).(a) By "C NMR in 1,2,4-trichlorobenzene (130 0C);24b*c pentad composition (%): mmmm (5.6). mmmr ( I l.8), rmmr (6.9), mmrr (10.4), nnrr + mmrm (25.4). rmrm (14.7), rrrr (4.9), mrrr (1 l.8), mrrm (8.5). (b) Bovey, F. A. Chain Srructure and Conformation of Macromolecules; Academic Press: New York. 1982; pp 78-91. (c) Ewen, J. A. J . Am. Chem. Soc. 1984,

108, 1427-1441.

106, 6355-6364.

Key Distances:Zr-C(15): 2.273 Å (2.280 for Cp2ZrMe2)Zr-C(34): 2.549 Å (2.249 Å for other µ!CH3)

B-C(34): 1.663 Å (normal)Zr-H(B): 2.25 Å (2.16 Å for other agostic H's)

Crystalline solid characterizedby EA, MS, NMR and X-Ray

ZrMe

Me

MeMe

MeMe

B(C6F5)3

Benzene[1,2-(CH3)2C5H3]2ZrCH3

+ CH3B(C6F5)3-

Key Angles:Zr-C(34)-B =161.8C(16)-B-C(34) = 108.7 C(22)-B-C(34) = 112.7, C(28)-B-C(34) = 102.8 C(16)-B-C(22) = 106.5 C(16)-B-C(28) = 114.3C(22)-B-C(28) = 112.0.

• Study by Marks on “Cation-like” homogeneous Zr-Catalysts

Marks, T. et. al. JACS, 1991, 113, 3623

-Highly unsymmetrical bridging methyl-Shortened Zr-C(15) bond-Boron distance/angles are typical

“Cation-Like” Homogeneous Polymerization

H2CCH2

Toluene, 23 °C, 0.5 h

Zr-Cat.

(1 atm)

n

- 4.5 x 106 g PE (mol of Zr)-1 h-1

- Similar to activity observed with MAO/Zirconocene catalysts

• Polymerization of ethylene and propylene

nToluene, 23 °C, 0.5 h

1.4 mM Zr-Cat.

(5 atm)

- 2.1 x 104 g PE (mol of Zr)-1 h-1

- atactic Polypropylene- NMR control experiment showed that >70% of Zr-cat. undergoes olefin insertion

-Provides direct experimental evidence for abstractive role of group 13 co-catalysts/activators in Ziegler-Natta polymerization

Marks, T. et. al. JACS, 1991, 113, 3623

Mechanism of Ziegler-Natta Polymerization

• Cossee mechanism (1960)

Key Features:- Olefin insertion via cis-opening of the double bond- For prochiral olefins, insertion occurs such that non-bonding interactions are minimized (R-groups are trans in the T.S.)

• Modified Cossee mechanism (Brookhart-Green-Rooney)

-Can also have a 2,1-insertion(regio-irregular)

agostic: derived from a greek word, which appears in Homer (trans: to hold or clasp oneself)

Cossee, P. Tetrahedron Lett. 1960, 17, 17.

Evidence For Agostic Interactions

Cp2ZrCl2 (0.1mol%)

n-C4H9 n-C4H9

CH2D

D

syn (erythro)

n-C4H9 n-C4H9

CH2D

D

anti (threo)

H2 (15 atm), MAO

Toluene, -5 °C, 20h

5% yield (anti : syn = 1.3 : 1)

(E)- or (Z)-1-deuteriohexene

• Kinetic isotope effects in hydrodimerizations

• Mechanism is similar to Z-N Polymerization

• Hydrodimerizations of deuteriohexene (Brintzinger)

ZrCp

Cp

H

R(Cp2)Zr H

RH

(Cp2)Zr H

RH

R

(Cp)2Zr

R R

Higher oligomers

R

H2

H3C

R R

R

- [(Cp)2Zr-H]

Cp2ZrCl2/MAO

H2

Brintzinger, H-H. Angew. Chem. Int. Ed. Engl. 1990, 102, 1412

Evidence for Agostic Interactions

HDH R

Cp2Zr H

HCp2Zr

D HH R

H

Cp2ZrCH2R

D

HDH R

D

Cp2ZrH

CH2R

RHD H

Cp2ZrH

DH

HR

CH2RD

Cp2ZrD

HD

RH

HCH2R

kH kD

!-H agostic

!-Dagostic

H2 H2

H2DC

H

HR

CH2RD

H2DC

D

RH

HCH2R

H2DC

D

HR

HCH2R

threo (anti) syn (erythro)

(E)-seriesHH

D R

Cp2Zr H

HCp2Zr

H HD R

H

Cp2ZrD

CH2R

RDH H

D

Cp2ZrCH2R

H

HHD R

Cp2ZrH

DH

RH

DCH2R

Cp2ZrD

HD

HR

CH2RH

kH kD

!-H agostic

!-Dagostic

H2 H2

H2DC

H

RH

DCH2R

H2DC

D

HR

CH2RH

H2DC

D

RH

CH2RH

threo (anti) syn (erythro)

(Z)-series

- Observed selectivity could be explained by agostic interactions

- Agostic H-interaction lowers T.S. energy relative to agostic-D

Group Exercise

HH

[OpSc(H)(PMe3)H2 (4 atm)

23 °CMe

DD


23 °CCH2D

D

CH2D

D

1.2 : 1 (trans : cis)

D


23 °C

1.2 : 1 (cis : trans)

CH2D

D

CH2D

D

D

>98% yield


a agostic assistance in the insertion step is expected to favor the trans product (vide infra). Hydrolysis and 2H N M R analysis of the resultant mixture of deuteriomethylcyclopentanes revealed a 1 .OO f 0.05 ratio of trans:cis products, arguing against an a

agostic assisted insertion in their system, however. The scandium hydride, {($-CSMe,),SiMe2JSc(PMe3)H

("OpSc( PMe,)H"), cleanly catalyzes the hydrocyclization of 1,Shexadiene to methylcyclopentane.* In light of some very recent theoretical results favoring transition state B,9 we have adapted this catafytic hydrocyclization reaction along the lines of the Grubbs experiment to probe for a agostic assistance with the scandium system. trans,trans-l,6-d2-1 ,5-Hexadiene was em- ployed as substrate, and as expected, OpSc(PMe3)H cleanly catalyzes its hydrocyclization to a mixture of cis- and trans-

d2-methylcyclopentane (eq 2).1° 2H{lH] NMR analysis reveals a ( I . I9 f 0.04): 1 ratio of transcis products a t 25 OC (see supplementary material)."

J . Am. Chem. Soc., Vol. 112, No. 25, 1990 9407

perature (1.07 f 0.03 at 120 'C, 1 . I9 f 0.04 at 25 OC, 1.26 f 0.03 at -10 "C); (2) hydrocyclization of cis,cis-1,6-d2-1,5-hexadiene affords a ratio of 1.20 f 0.04, indicating that insertion of the pendant olefin is not influenced by the geometry about its double bond; and (3) similarly, trans-l-dl-l,5-hexadiene gives the same trans:cis ratio of 1.19 f 0.02 with the single deuteron partitioned equally (2H NMR) between methyl and ring positions of the d,-methylcyclopentane product.

For longer chain diolefins, simple hydrogenation to a,w-d2- alkanes competes with hydrocyclization. Reaction of OpSc- (PMeJH with neat 1,Sheptadiene at -4 OC in the presence of 1 atm of H2 leads to an approximately 60:40 mixture of methylcyclohexane and n-heptane. 1,7-Octadiene is converted ex- clusively to n-octane. Nonetheless, the partial hydrocyclization of trans,trans-1,7-d2- 1,6-heptadiene allowed us to ascertain the ratio of trans- to cis-d2-methylcyclohexane products. We find a (1.2 1 f 0.5): 1 ratio of cis-d2-methylcyclohexane:frans-d2- methylcyclohexane, although the analysis is less accurate due to overlap of the cis-D resonance with that for the methyl deuterons in the 2H N M R spectrum (see supplementary material). This reversal of trans:cis ratio is entirely consistent with the expectation that face selectivity for pendant olefin approach would be opposite to that for formation of the five-membered ring. The developing six-membered ring adopts a chair-like conformation, and the pseudo bicyclic transition state is now trans fused. Thus, as illustrated below, preferential H in the a agostic position leads to the cis isomer.

1

Assuming that an cy agostic interaction is, in fact, responsible,12 the excess of trans-d2-methylcyclopentane may be rationalized as shown in Scheme I . Addition of achiral a,w-diene to achiral OpScH yields a precisely 5050 mixture of R and S 1,6-d2-5- hexenyl-scandium complexes. Due to ring strain there should a strong preference for cis fusion of the pseudo 4,5 ring system in the transition state for olefin insertion.I3 As shown in Scheme I for the R isomer only, the face selection for insertion of the pendant olefin then depends on whether H or D occupies the a agostic position. The expected preference for H to occupy the bridging position'* leads to excess of the R,R (trans) product. A similar analysis for the S enantiomer leads to the same conclusion: the trans isomer ( in that case S,S) is produced in excess, if an a agostic interaction assists olefin insertion into the Sc-C bond (see supplementary material for full analysis).

Additional experiments support the supposition that the par- titioning of stereochemistry is due to a kinetic deuterium isotope effect operat ing a t the CY methylene of the [OpScCH DCH2CH2CH,CH=CH D] intermediate: ( I ) the t ramcis ratio varies in a normal enthalpic manner with tem-

(8) Piers. W. E.; Shapiro, P. J.; Bunel, E. E.; Bercaw, J. E. Synlett 1990,

(9) H. H. Brintzinger, private communication. ( I O ) In a typical procedure, a 10-15-mg sample of OpSc(PMe,)H was

loaded into a 1 SO-mL thick-walled reactor equipped with a 4-mm Kontes needle valve. Toluene (4 mL) was added in vacuo, the scandium complex dissolved with warming, and ca. 100 equiv of rigorously dried trans,trans- /,6-d2-l,5-hexadiene and 1 atm of H2 were added at -196 OC. The reactor was warmed to room temperature and stirred for 30 min at 25 OC. Excess dihydrogen was removed by two freeze-pump-thaw cycles, and all volatile organic compounds were vacuum transferred away from the catalyst residue for analysis by 2H NMR and GC.

( I I ) 'H-coupled 2H spectra, while not entirely base-line resolved, gave the same ratio of trans:cis, confirming that NOES for 2H are insignificant.

(12) An alternative, less likely possibility is that the excess trans isomer results from the smaller size of deuterium in a more conventional transition state such as A. 'Steric" kinetic deuterium isotope effects of 1.16 f 0.03 have been observed for extremely crowded transition states; see: Melander, M.; Saunders, W. H., Jr. Reacrion Rares of Isotopic Molecules; Wiley-lntersci- ence: New York, 1980; p 189. On the other hand, the 1.Oo:l ratio observed by Grubbs and co-workers at -100 OC would appear to rule out a substantial contribution from steric effects.

( I 3) Although strained, some trans alicyclic bicyclo[3.2.0]heptane compounds have becn prepared: Meinwald, J.; Anderson, P.; Tufariello, J. J. J . Am. Chem. Soc. 1966,88, 1301 and references cited therein.

(14) Calvert, R. B.; Shapley, J. R. J . Am. Chem. Soc. 1978. 100,

14-84.

1726-7727.

slightly favored slightly disfavored

Our results provide good evidence for the modified Green- Rooney pathway for chain propagation with these Ziegler-Natta systems.15 Moreover, they suggest a rationale for the apparent requirement that active catalysts be 14-electron alkyl derivatives with two vacant orbitals: one to acoommodate the incoming olefin, another for the a agostic interaction.I6

Acknowledgment. We thank Professors Bob Grubbs, Tom Flood, Maurice Brookhart, and Dennis Dougherty for helpful discussions. We also thank Professor Hans Brintzinger for providing us with information prior to publication. This work was supported by the USDOE Office of Basic Energy Sciences (Grant No. DE-FG03-85ER113431), by Exxon Chemicals Americas, and by Shell Companies Foundation. W.E.P. thanks the National Sciences and Engineering Research Council of Canada and the Izaak Walton Killam Foundation for postdoctoral fellowship support.

Supplementary Material Available: Experimental details for the preparation of the substrates used, 2H NMR measurements, gas chromatographic separations, and complete versions of Scheme I for frans,trans-1,6-d2-1 ,5-hexadiene and trans,trans-l,7-d2- 1,6-heptadiene hydrocyclizations (8 pages). Ordering information is given on any current masthead page.

( 1 5 ) Brintzinger and Krauledat have very recently observed a I :( I .28 & 0.02) ratio of threo-erythro 5-merhyl,6-d2-5-methylundecane isomers in the hydrodimerization of rrans-l-d,-I-hexene with the ($-C5Hs)2ZrC12/ methylalumoxane system (Krauledat, H.; Brintzinger, H. H. Angew. Chem., su bmitted).

(16) B agostic interactions have been identified as the preferred ground- stole structures for Cp*,ScCH2CH, and for the cations [Cp*(L)- CoCH2CH2R]+ and [Cp2(PMe3)ZrCH2CH2R]+ (refs 1 and 8; Jordan, R. F.; Bradley, P. K.; Baenziger, N. C.; LaPornte, R. E. J . Am. Chem. SOC. 1990, 112, 1289). Our results implicate the (less stable due to ring strain) a agostic arrangement in the transifion stare for C-C bond formation. Thus, we tentatively conclude that B agostic structures characterize ground states and a agostic structures transition states for chain propagation. On the other hand, the I . & I 1 ratio observed by Grubbs et al. could be taken as an indication that a agostic assistance is not required for C-C bond formation.

• Further support for agostic interactions from Bercaw

GROUP EXERCISE: 1. Provide a mechanism/stereochemical rationale that accounts for the observed KIE in the formation of d2-methylcyclopentane.2. Provide a mechanistic rationale that explains why the selectivity is opposite in the case of d2-methylcyclohexane.

Bercaw Solution9406 J . Am. Chem. SOC. 1990, I 1 2, 9406-9407

solvents are complex and depend both on the solute-solvent hydrogen bonding and dielectric relaxation. The heteroatom lone pair hydrogen bonding in the absence of dielectric effects can often lead to red shifts.

Acknowledgment. This work is supported in part through a grant from the Office of Naval Research. We gratefully ac- knowledge profitable discussions with Mr. Ricardo Longo.

Q "Agostic" Assistance in Ziegler-Natta Polymerization of Olefins. Deuterium Isotopic Perturbation of Stereochemistry Indicating Coordination of an Q C-H Bond in Chain Propagation

Warren E. Piers and John E. Bercaw*

Contribution No. 81 77, Arnold and Mabel Beckman Laboratories of Chemical Synthesis

California Institute of Technology Pasadena, California 91 125

Received August I, 1990

The well-defined, homogeneous Ziegler-Natta olefin polymerization systems that have been reported recently provide an unprecedented opportunity to investigate the mechanism of this important process. New systems' include ( I ) single-component catalysts such as cationic group 4 metallocenes* or the isoelectronic, neutral group 3 or lanthanide metallocene hydrides or alkyls) and (2) highly active, two-component systems consisting of "methylalumoxane" in combination with a group 4 metallocene derivative, which, with suitable modification of the cyclopentadienyl ligands, may exhibit remarkable iso-or syndiospe- ~ i f ic i t ies .~ While a consensus appears to be developing that in all these systems the active catalysts are the 14-electron, do (or doP) metallocene alkyls, Cp2MR (M = lanthanide or group 3 transition metal) or [Cp2MR]+ (M = group 4 transition metal), the mechanism for chain propagation and the geometry of the transition state for olefin insertion into the metal-carbon bond have not yet been unequivocally established.

The Cossee mechanism and a staggered arrangement of alkyl and olefin substituents (A ) generally have been assumed in ra- tionalizing the stereospecificity of propene polymerization by these metallocene catalyst system^.^ The most popular altemative suited to do metal complexes, the "modified Green-Rooney mechanism",

( I ) A class of one-component ethylene-polymeri~tion catalysts based on late-transition-metal complexes of the type [($C,Me,)(L)M(C,H,)(R)]+ (M = CO. Rh; L = phosphine or phosphite) has also been developed. See: Brookhart, M.: Volpe, A. F., Jr.; Lincoln, D. M. J. Am. Chem. Soc. 1990,112, 5634-5636 and references cited therein.

(2) (a) Jordan, R. F.; LaPointe, R. E.; Bradley, P. K.; Baenziger, N. Organometallics 1989, 8, 2892-2906. (b) Hlatky, G. G.; Turner, H. W.; Eckman, R. R. J. Am. Chem. Sor. 1989, 111 , 2728-2729. (c) Taube, R.; Krukowa, L. J. Organomet. Chem. 1988,347, C9. (c) Eshuis, J. J. W.; Tan, Y. Y.; Teuben. J . H. J. Mol. Catal.. submitted.

(3) (a) Watson, P. L. J. Am. Chem. Sor. 1982, 104, 337. (b) Burger, B. J.; Thompson, M. E.; Cotter, W. D.; Bercaw, J. E. J. Am. Chem. Sor. 1990. 112. 1566-1 577. (c) Jeske, G.; Lauke. H.; Mauermann, H.; Sweptson. P. N.; Schumann, H.; Marks, T. J . J. Am. Chem. Sor. 1985. 107, 8091.

(4) (a) Kaminsky, W.; Kulper, K.; Brintzinger, H. H.; Wild, F. R. W. P. Angew. Chem., Int. Ed. Engl. 1985,24, 507. (b) Ewen, J. A. J . Am. Chem. Sor. 1984,106,635-6364. (c) Ewen, J. A,; Jones, R. L.; Razavi, A. J. Am. Chem. Sor. 1988, 110, 6255-6256. (d) Erker, G.; Nolle, R.; Tsay, Y.-H.; Kruger, C. Angew. Chem., tnt. Ed. Engl. 1989, 29, 629. (e) Mallin, D. T.; Rausch, M. D.; Lin, Y.-G.; Dong, S.; Chien. J . C. W. J. Am. Chem. Soc. 1990, 112,2030-2031. (0 Rexoni. L.: Waymouth. R. M. J. Am. Chem. Soc. 1990, 112.4953-4954.

(5) (a) Pino, P.; Galimbcrti, M. J. Organomet. Chem. 1989,370. 1-7. (b) Pino, P.; Cioni. P.: Wei, J.; Rotzinger. B.; Arizzi, S. Recent Developments in Basic Research on the Stereospecific Polymerization of a-Olefins. In Tran- sition Metal Catalyzed Polymerization: Ziegler-Nut fa and Metathesis Po- lymerization; Quirk, R. p., Ed.; Cambridge Press: New York, 1988; pp 1-24 and tcferences cited therein. (c) Jolly, C. A,; Marynick. D. S. J. Am. Chem. SOC. 1989, I I I . 7968-7974. (d) Waymouth, R.; Pino, P. J. Am. Chem. Sor.

1990, 112. 491 1-4914.

Scheme I

D- ' 0

OpScH I

slightly slightly f a v o r e v \sfavored

I

- OpSCH 1 H2;

I

- OpSCH 1 H2;

invokes CY C-H coordination to assist olefin insertion,6 a quite different transition-state geometry (B).

A B

I n a cleverly conceived experiment, Grubbs et al.' probed for an N agostic interaction in the transition state for olefin insertion. Racemic I-d,-S-hexenylchlorotitanocene was prepared and found to undergo AIC12(CH2CH3)-induced cyclization to a mixture of cis- and tmns-2-dl-cyclopentylmethyl stereoisomers (eq 1). Any

CH,

1 AIEtCI, (-lOO°C) CI / ' 2 bipyndine (-100°C to 25 '5

CPZTI \ P.-

Hr 'D 3 hydrolysis cis'trans = 1 00 t 0 05

(6) (a) Brookhart, M.; Green, M. L. H . J. Organomet. Chem. 1983,250, 395. (b) Brookhart, M.; Green, M. L. H.; Wong, L. frog. tnwg. Chem. 1988, 36, I .

(7) Clawson, L.; Soto, J.; Buchwald, S. L.; Steigerwald, M. L.; Grubbs, R. H J . Am. Chem. SOC. 1985, 107, 3377-3378.

Q 1990 American Chemical Society 0002-78631901 I5 I2-9406$02.50/0


a agostic assistance in the insertion step is expected to favor the trans product (vide infra). Hydrolysis and 2H N M R analysis of the resultant mixture of deuteriomethylcyclopentanes revealed a 1 .OO f 0.05 ratio of trans:cis products, arguing against an a

agostic assisted insertion in their system, however. The scandium hydride, {($-CSMe,),SiMe2JSc(PMe3)H

("OpSc( PMe,)H"), cleanly catalyzes the hydrocyclization of 1,Shexadiene to methylcyclopentane.* In light of some very recent theoretical results favoring transition state B,9 we have adapted this catafytic hydrocyclization reaction along the lines of the Grubbs experiment to probe for a agostic assistance with the scandium system. trans,trans-l,6-d2-1 ,5-Hexadiene was em- ployed as substrate, and as expected, OpSc(PMe3)H cleanly catalyzes its hydrocyclization to a mixture of cis- and trans-

d2-methylcyclopentane (eq 2).1° 2H{lH] NMR analysis reveals a ( I . I9 f 0.04): 1 ratio of transcis products a t 25 OC (see supplementary material)."

J . Am. Chem. Soc., Vol. 112, No. 25, 1990 9407

perature (1.07 f 0.03 at 120 'C, 1 . I9 f 0.04 at 25 OC, 1.26 f 0.03 at -10 "C); (2) hydrocyclization of cis,cis-1,6-d2-1,5-hexadiene affords a ratio of 1.20 f 0.04, indicating that insertion of the pendant olefin is not influenced by the geometry about its double bond; and (3) similarly, trans-l-dl-l,5-hexadiene gives the same trans:cis ratio of 1.19 f 0.02 with the single deuteron partitioned equally (2H NMR) between methyl and ring positions of the d,-methylcyclopentane product.

For longer chain diolefins, simple hydrogenation to a,w-d2- alkanes competes with hydrocyclization. Reaction of OpSc- (PMeJH with neat 1,Sheptadiene at -4 OC in the presence of 1 atm of H2 leads to an approximately 60:40 mixture of methylcyclohexane and n-heptane. 1,7-Octadiene is converted ex- clusively to n-octane. Nonetheless, the partial hydrocyclization of trans,trans-1,7-d2- 1,6-heptadiene allowed us to ascertain the ratio of trans- to cis-d2-methylcyclohexane products. We find a (1.2 1 f 0.5): 1 ratio of cis-d2-methylcyclohexane:frans-d2- methylcyclohexane, although the analysis is less accurate due to overlap of the cis-D resonance with that for the methyl deuterons in the 2H N M R spectrum (see supplementary material). This reversal of trans:cis ratio is entirely consistent with the expectation that face selectivity for pendant olefin approach would be opposite to that for formation of the five-membered ring. The developing six-membered ring adopts a chair-like conformation, and the pseudo bicyclic transition state is now trans fused. Thus, as illustrated below, preferential H in the a agostic position leads to the cis isomer.

1

Assuming that an cy agostic interaction is, in fact, responsible,12 the excess of trans-d2-methylcyclopentane may be rationalized as shown in Scheme I . Addition of achiral a,w-diene to achiral OpScH yields a precisely 5050 mixture of R and S 1,6-d2-5- hexenyl-scandium complexes. Due to ring strain there should a strong preference for cis fusion of the pseudo 4,5 ring system in the transition state for olefin insertion.I3 As shown in Scheme I for the R isomer only, the face selection for insertion of the pendant olefin then depends on whether H or D occupies the a agostic position. The expected preference for H to occupy the bridging position'* leads to excess of the R,R (trans) product. A similar analysis for the S enantiomer leads to the same conclusion: the trans isomer ( in that case S,S) is produced in excess, if an a agostic interaction assists olefin insertion into the Sc-C bond (see supplementary material for full analysis).

Additional experiments support the supposition that the par- titioning of stereochemistry is due to a kinetic deuterium isotope effect operat ing a t the CY methylene of the [OpScCH DCH2CH2CH,CH=CH D] intermediate: ( I ) the t ramcis ratio varies in a normal enthalpic manner with tem-

(8) Piers. W. E.; Shapiro, P. J.; Bunel, E. E.; Bercaw, J. E. Synlett 1990,

(9) H. H. Brintzinger, private communication. ( I O ) In a typical procedure, a 10-15-mg sample of OpSc(PMe,)H was

loaded into a 1 SO-mL thick-walled reactor equipped with a 4-mm Kontes needle valve. Toluene (4 mL) was added in vacuo, the scandium complex dissolved with warming, and ca. 100 equiv of rigorously dried trans,trans- /,6-d2-l,5-hexadiene and 1 atm of H2 were added at -196 OC. The reactor was warmed to room temperature and stirred for 30 min at 25 OC. Excess dihydrogen was removed by two freeze-pump-thaw cycles, and all volatile organic compounds were vacuum transferred away from the catalyst residue for analysis by 2H NMR and GC.

( I I ) 'H-coupled 2H spectra, while not entirely base-line resolved, gave the same ratio of trans:cis, confirming that NOES for 2H are insignificant.

(12) An alternative, less likely possibility is that the excess trans isomer results from the smaller size of deuterium in a more conventional transition state such as A. 'Steric" kinetic deuterium isotope effects of 1.16 f 0.03 have been observed for extremely crowded transition states; see: Melander, M.; Saunders, W. H., Jr. Reacrion Rares of Isotopic Molecules; Wiley-lntersci- ence: New York, 1980; p 189. On the other hand, the 1.Oo:l ratio observed by Grubbs and co-workers at -100 OC would appear to rule out a substantial contribution from steric effects.

( I 3) Although strained, some trans alicyclic bicyclo[3.2.0]heptane compounds have becn prepared: Meinwald, J.; Anderson, P.; Tufariello, J. J. J . Am. Chem. Soc. 1966,88, 1301 and references cited therein.

(14) Calvert, R. B.; Shapley, J. R. J . Am. Chem. Soc. 1978. 100,

14-84.

1726-7727.

slightly favored slightly disfavored

Our results provide good evidence for the modified Green- Rooney pathway for chain propagation with these Ziegler-Natta systems.15 Moreover, they suggest a rationale for the apparent requirement that active catalysts be 14-electron alkyl derivatives with two vacant orbitals: one to acoommodate the incoming olefin, another for the a agostic interaction.I6

Acknowledgment. We thank Professors Bob Grubbs, Tom Flood, Maurice Brookhart, and Dennis Dougherty for helpful discussions. We also thank Professor Hans Brintzinger for providing us with information prior to publication. This work was supported by the USDOE Office of Basic Energy Sciences (Grant No. DE-FG03-85ER113431), by Exxon Chemicals Americas, and by Shell Companies Foundation. W.E.P. thanks the National Sciences and Engineering Research Council of Canada and the Izaak Walton Killam Foundation for postdoctoral fellowship support.

Supplementary Material Available: Experimental details for the preparation of the substrates used, 2H NMR measurements, gas chromatographic separations, and complete versions of Scheme I for frans,trans-1,6-d2-1 ,5-hexadiene and trans,trans-l,7-d2- 1,6-heptadiene hydrocyclizations (8 pages). Ordering information is given on any current masthead page.

( 1 5 ) Brintzinger and Krauledat have very recently observed a I :( I .28 & 0.02) ratio of threo-erythro 5-merhyl,6-d2-5-methylundecane isomers in the hydrodimerization of rrans-l-d,-I-hexene with the ($-C5Hs)2ZrC12/ methylalumoxane system (Krauledat, H.; Brintzinger, H. H. Angew. Chem., su bmitted).

(16) B agostic interactions have been identified as the preferred ground- stole structures for Cp*,ScCH2CH, and for the cations [Cp*(L)- CoCH2CH2R]+ and [Cp2(PMe3)ZrCH2CH2R]+ (refs 1 and 8; Jordan, R. F.; Bradley, P. K.; Baenziger, N. C.; LaPornte, R. E. J . Am. Chem. SOC. 1990, 112, 1289). Our results implicate the (less stable due to ring strain) a agostic arrangement in the transifion stare for C-C bond formation. Thus, we tentatively conclude that B agostic structures characterize ground states and a agostic structures transition states for chain propagation. On the other hand, the I . & I 1 ratio observed by Grubbs et al. could be taken as an indication that a agostic assistance is not required for C-C bond formation.

The forming six-membered ring adopts a chair-like conformation in the pseudo bicyclic T.S., which is trans-fused.

2.

1. cis-fused pseudo bicyclic T.S. is favored due to ring strain in the

Bercaw, J. E., JACS, 1990, 112, 9406

Stereochemical Considerations

REVIEWS Chtral Metallocene Catalysts

could be expected to be retained even under catalysis conditions.

When activated with M A 0 in the manner described above, these

onstr-i~~etallocenesl' 04] were indeed found, in independent stud-

ies by Ewen with (en)(thind),TiC12/MAO[1051 and by Kaminsky

and Kiilper with ruc-(en)(thind),ZrClz/MA0,~1061 to polymer-

ize propene and other cr-olefins to give highly isotactic polymers.

These findings led to extensive exploration of the mechanisms

by which these catalysts control the stereochemistry of polymer

growth and the effects of different metallocene structures on the

tacticities and other properties of the polymers produced.

3.1. Catalyst Structures and Stereoselectivities

Polypropene produced by MAO-activated chiral ansa-zir-

conocenes such as rrrc-(en)(ind),ZrCI, or rac(en)(thind),ZrCl,

has similar isotacticity to polymers produced with heteroge-

neous Ziegler-Natta catalysts. For a more quantitative discus-

sion of these relations. we consider briefly the 13C N M R spec-

troscopic methods1t071 currently used to characterize the

stereoregularity of poly(r-olefins).

The 13C NMR signals of a polymer are most conveniently

related to its microstructure by a stereochemical notation devel-

oped by Bovey.'107"1 Relative configurations of neighboring

units (or "dyads") are designated as m (meso) for equally and r

(rac'cmo) for unequally positioned substituents in a Fischer-type

projection of the polymer chain. In polypropene, the I3C N M R

shift of each CH, group is determined by the configurations of

two neighboring repeat units on either side;[108] each CH, sig-

nal is thus assignable to a particular "pentad" pattern, repre-

sented by the four consecutive m or r designators framing the

CH, group under consideration (Scheme 9). All ten possible

m m m m m m r r r r r r

-LLLL... ...bLLr_L... .--JL$+..

... 1 I T 1 ,... ... JufT... ...+.L,A-.. ...%,... m m m r m m r m r m r T r T r m

r m m r r m r m m r r m

mm-centered mr-centered rr -centered

Scheme 9. The ten possible stereochemical pentads of a polyolefin [107].

pentad signals (mmmm, mmmr, rmmr, mmrr, mmrm, rmrr, rm-

rm, r r r r , rrrm, and mrrm)['Ogl are observed for the randomly

configurated repeat units of atactic polypropene[llO1 (Fig. 4).

Isotactic polypropene, on the other hand, is ideally character-

ized by a single 13C N M R signal for mmmm pentads, since its

repeat units have identical configuration over long segments of

the polymer chain.

The degree of isotacticity of such a polymer is usually ex-

pressed as the ratio of the mmmm pentad integral to the integral

sum of all pentad signals observed; it will be designated in the

following as [mmmm] .I' ' ' I Isotactic polypropene produced by

modern heterogeneous Ziegler-Natta catalysts[" is highly

1

I " " I " " I " " I ' " " ' ' " ~ ' ' "

23.0 22.0 21.0 20.0 19.0

a h

Fig. 4. 13C NMR spectra of isotactic polypropene (CH, region) with [mmmnr] >

90% obtained with ror-MetSi(l-benz[e]indenyl),ZrCl,iMAO [117 h. 120dl (top),

and ofatactic polypropene with random pentad distribution obtained with the f n m

isomer of the same catalyst (bottom).

stereoregular with [mmmm] > 0.95. At ambient temperatures,

typical MAO-activated chiral ansa-metallocene catalysts

yield polypropene with stereoregularities of [mmmm] x 0.8 - 0.9.['063 "'3 ' I 3 ] At increased polymerization temperatures,

however, most of these homogeneous catalysts are distinctly less

stereoselective than typical heterogeneous Ziegler- Natta cata-

The 3C N M R signals associated with occasional stereoerrors

in the isotactic polymers produced by metallocene catalysts indi-

cate that stereoregularity is controlled by the chirality of the

metallocene catalyst and not by the asymmetry of the last insert-

ed unit: Such "chain-end'' control would give rise to the single-

inversion pentad mmrm as a main error signal, since an occa-

sionally inverted chain-end configuration would generally be

followed by units of the same configuration. For this chain-end

control one would expect an error pentad ratio close to

mmmr: mmrr :mmrm : mrrm = 1 : 0: 1 :O. In fact, an error pentad

distribution close to mmmr : mmrr : mmrm : mrrm = 2: 2 : 0: 1 is

observed for these polymers (Scheme 10). As previously estab-

lished for heterogeneous this pentad pattern is ex-

pected when olefin insertion is indeed under "catalytic-site"

control.[105* l l S 1 In this case, r dyads due to stereoinversions will

occur in pairs, since a catalyst center of fixed chirality forces the

enantiofacial orientation of subsequent olefin insertions to re-

turn to the previous preference immediately after an occasional

mistake (Scheme 10).

How the stereoselectivities of chiral catalysts depend on the

structure of the complex, in particular on different bridging

units and substituent patterns, has been the subject of compara-

tive studies in several research groups.['I6- In investigations '

of this kind, it is crucial to take into account the polymerization

temperature TP.[l2l1 Even an unbridged metallocene catalyst

with unsubstituted ring ligands, Cp,TiPh,/MAO, has been

lysts." 14=1

A i i ~ r n C'limi / ) i f . Ed. Engl. 1995, 34. 1143-1170 1151

REVIEWS Chtral Metallocene Catalysts

could be expected to be retained even under catalysis conditions.

When activated with M A 0 in the manner described above, these

onstr-i~~etallocenesl' 04] were indeed found, in independent stud-

ies by Ewen with (en)(thind),TiC12/MAO[1051 and by Kaminsky

and Kiilper with ruc-(en)(thind),ZrClz/MA0,~1061 to polymer-

ize propene and other cr-olefins to give highly isotactic polymers.

These findings led to extensive exploration of the mechanisms

by which these catalysts control the stereochemistry of polymer

growth and the effects of different metallocene structures on the

tacticities and other properties of the polymers produced.

3.1. Catalyst Structures and Stereoselectivities

Polypropene produced by MAO-activated chiral ansa-zir-

conocenes such as rrrc-(en)(ind),ZrCI, or rac(en)(thind),ZrCl,

has similar isotacticity to polymers produced with heteroge-

neous Ziegler-Natta catalysts. For a more quantitative discus-

sion of these relations. we consider briefly the 13C N M R spec-

troscopic methods1t071 currently used to characterize the

stereoregularity of poly(r-olefins).

The 13C NMR signals of a polymer are most conveniently

related to its microstructure by a stereochemical notation devel-

oped by Bovey.'107"1 Relative configurations of neighboring

units (or "dyads") are designated as m (meso) for equally and r

(rac'cmo) for unequally positioned substituents in a Fischer-type

projection of the polymer chain. In polypropene, the I3C N M R

shift of each CH, group is determined by the configurations of

two neighboring repeat units on either side;[108] each CH, sig-

nal is thus assignable to a particular "pentad" pattern, repre-

sented by the four consecutive m or r designators framing the

CH, group under consideration (Scheme 9). All ten possible

m m m m m m r r r r r r

-LLLL... ...bLLr_L... .--JL$+..

... 1 I T 1 ,... ... JufT... ...+.L,A-.. ...%,... m m m r m m r m r m r T r T r m

r m m r r m r m m r r m

mm-centered mr-centered rr -centered

Scheme 9. The ten possible stereochemical pentads of a polyolefin [107].

pentad signals (mmmm, mmmr, rmmr, mmrr, mmrm, rmrr, rm-

rm, r r r r , rrrm, and mrrm)['Ogl are observed for the randomly

configurated repeat units of atactic polypropene[llO1 (Fig. 4).

Isotactic polypropene, on the other hand, is ideally character-

ized by a single 13C N M R signal for mmmm pentads, since its

repeat units have identical configuration over long segments of

the polymer chain.

The degree of isotacticity of such a polymer is usually ex-

pressed as the ratio of the mmmm pentad integral to the integral

sum of all pentad signals observed; it will be designated in the

following as [mmmm] .I' ' ' I Isotactic polypropene produced by

modern heterogeneous Ziegler-Natta catalysts[" is highly

1

I " " I " " I " " I ' " " ' ' " ~ ' ' "

23.0 22.0 21.0 20.0 19.0

a h

Fig. 4. 13C NMR spectra of isotactic polypropene (CH, region) with [mmmnr] >

90% obtained with ror-MetSi(l-benz[e]indenyl),ZrCl,iMAO [117 h. 120dl (top),

and ofatactic polypropene with random pentad distribution obtained with the f n m

isomer of the same catalyst (bottom).

stereoregular with [mmmm] > 0.95. At ambient temperatures,

typical MAO-activated chiral ansa-metallocene catalysts

yield polypropene with stereoregularities of [mmmm] x 0.8 - 0.9.['063 "'3 ' I 3 ] At increased polymerization temperatures,

however, most of these homogeneous catalysts are distinctly less

stereoselective than typical heterogeneous Ziegler- Natta cata-

The 3C N M R signals associated with occasional stereoerrors

in the isotactic polymers produced by metallocene catalysts indi-

cate that stereoregularity is controlled by the chirality of the

metallocene catalyst and not by the asymmetry of the last insert-

ed unit: Such "chain-end'' control would give rise to the single-

inversion pentad mmrm as a main error signal, since an occa-

sionally inverted chain-end configuration would generally be

followed by units of the same configuration. For this chain-end

control one would expect an error pentad ratio close to

mmmr: mmrr :mmrm : mrrm = 1 : 0: 1 :O. In fact, an error pentad

distribution close to mmmr : mmrr : mmrm : mrrm = 2: 2 : 0: 1 is

observed for these polymers (Scheme 10). As previously estab-

lished for heterogeneous this pentad pattern is ex-

pected when olefin insertion is indeed under "catalytic-site"

control.[105* l l S 1 In this case, r dyads due to stereoinversions will

occur in pairs, since a catalyst center of fixed chirality forces the

enantiofacial orientation of subsequent olefin insertions to re-

turn to the previous preference immediately after an occasional

mistake (Scheme 10).

How the stereoselectivities of chiral catalysts depend on the

structure of the complex, in particular on different bridging

units and substituent patterns, has been the subject of compara-

tive studies in several research groups.['I6- In investigations '

of this kind, it is crucial to take into account the polymerization

temperature TP.[l2l1 Even an unbridged metallocene catalyst

with unsubstituted ring ligands, Cp,TiPh,/MAO, has been

lysts." 14=1

A i i ~ r n C'limi / ) i f . Ed. Engl. 1995, 34. 1143-1170 1151

• The tacticity and degree of stereo-regularity in PP

Isotactic [mmmm] >90%

atactic

! 13C is dependent on the configuation

of it's four stereogenic neighbors

r rm m

r = racemo (unequal)m = meso (equal)

[mmmm] =int(mmmm)

SUM[int(pentads)]

[mmmm] > 0.95 is Highly isotactic

Similar eqn applies for syndiotactic

Brintzinger, H-H. Angew. Chem. Int. Ed. Engl. 1995, 34, 1143

Stereochemical Considerations

are now available that can control the molecularweight, molecular weight distribution, comonomerincorporation, and both the relative and absolutestereochemistry of a polymer in a way that is oftenimpossible using conventional hetereogeneous cata-lysts. Although their commercial implementation inthe solution phase is often impractical, they can beheterogenized for efficient gas-phase or flow-throughreaction by attaching them to a solid support. Per-haps most importantly, these defined molecular-based systems allow detailed structural and mecha-nistic studies. Thus, through theoretical and empiricalstudies scientists can rapidly evolve new and im-proved generations of catalysts.

B. Scope of ReviewThis review covers the scientific literature from the

mid-1980s to the present concerning stereoselectivepolymerization by single-site transition metal andf-block metal complexes. Strategies for controlling therelative configuration of main-chain stereogenic cen-ters of chain-growth polyolefins are included; sincethe emphasis is on stereochemical control of polym-erization by the homogeneous catalyst, the polymer-ization of optically active monomers will not becovered. The review will concentrate on examiningstate-of-the-art stereoselective polymerization cata-lysts and will focus on proposed mechanisms ofstereocontrol. Although the emphasis will be onstereochemical control by the catalyst, other impor-tant characteristics such as polymerization activityand polymer properties will be included.

C. Mechanisms, Nomenclature, and Quantificationof Stereoregularity

Both the ligand set of a single-site catalyst and thegrowing polymer chain influence the stereochemistryof the polymerization reaction.13 It is interesting tonote that, unlike the catalytic synthesis of smallmolecules, during a chain-growth polymerizationreaction a polymer chain remains bound to the activemetal center during monomer enchainment. Thus,the stereogenic center from the last enchained mono-mer unit will have an influence on the stereochem-istry of monomer addition; if this influence is signifi-

cant, the mode of stereochemical regulation is referredto as “polymer chain-end control”. It should be notedthat in rare instances more than one stereogeniccenter of the polymer can play a significant role instereoregulation. If the ligand set is chiral andoverrides the influence of the polymer chain end, themechanism of stereochemical direction is termed“enantiomorphic-site control” (Scheme 1). In theformer mechanism, a stereochemical error is propa-gated, while in the latter a correction occurs sincethe ligands direct the stereochemical events.

Scheme 1 introduces the parameters that are usedto describe the stereoselectivity of the monomerenchainment process. For chain-end control, theparameters Pm and Pr refer to the probability of mesoand racemic placements, respectively (the Boveyformalism is a convenient way to describe polymertacticity, with a small “m” for meso, and a small “r”for racemic relationships between adjacent stereo-genic centers). A Pm equal to unity indicates isotac-ticity, while a Pr equal to unity signifies syndiotac-ticity. For site-control mechanisms, the parameter Rrepresents the degree of enantiotopic selectivity ofthe enchainment. When R is either 1 or 0 an isotacticpolymer forms, while an R parameter of 0.5 producesan atactic polymer. Polymer architectures relevantto this review are shown in Figure 1.

There are several techniques for determining thetype of tacticity and degree of stereoregularity of apolymer sample. Commonly used methods includesolubility, X-ray diffraction, IR spectroscopy, andthermal properties (melting point and glass-transi-tion temperature). In the case of chiral polymers,optical rotation can be used to determine the absoluteconfiguration as well as degree of enantiomeric puritywhen the optically pure polymer is available. How-ever the most useful method for determining apolymer’s tacticity classification as well as quantify-ing its stereochemical purity is nuclear magneticresonance (NMR).14,15 In many cases the shifts for thevarious polymer nuclei are sensitive to adjacentstereogenic centers, resulting in fine structure thatcan provide quantitative information about the poly-mer microstructure once the shifts identities areassigned. For example, the methyl region of a high-resolution 13C NMR spectrum of atactic polypropyl-

Scheme 1. Chain-End and Enantiomorphic Site Mechanisms of Stereocontrol

1224 Chemical Reviews, 2000, Vol. 100, No. 4 Coates

• Chain-End Control vs. Site Control

errorspropagated

errors corrected

mmmr : mmrr : mmrm : mrrm = 2 : 2 : 0 : 1

m m m r r m m

Enantiomorphic Site Control Pentad Distributions

mmmr : mmrr : mmrm : mrrm = 1 : 0 : 1 : 0

m m m r m m m

Chain-End Control Pentad Distributions

• Predicted 13C NMR Patterns

(r-dyads)( rr-triads)

Coates, G. W. Chem. Rev. 2000, 100, 1223.

Isotactic PP from Homogeneous Catalysts

nToluene, -60 °C, 4 h

Zr-Cat-1:2(86 µmol, 56% rac, 44% meso)MAO 9.4 (mmol)

n

63% isotactic(insoluble in pent.)

37% atatic(soluble in pent.)

(2.4 mol)6.0 g Polymer

Ti-Cat-1 Ti-Cat-2

13C NMR of insolublepolymer

C13 NMR of crude

• Ewen studies with racemic and meso ansa-metallocenes

- 13C-NMR analysis of the triad fragmentssuggested a mixture of isotactic/atatic-PP

- Author proposed that the isotactic fragmentarose from the C2-symmetric cat-2

• Later confirmed by Brintzinger using isomerically ansa-metallocene

nToluene, 23 °C, 0.5 h

Zr-Cat-3 (3.3 µmol)MAO 5.6 (mmol)

2.7 kg (mol Zr)-1 h-1

>95% isotactic-PPZr-Cat-3Brintzinger, H-H. Angew. Chem. Int. Ed. Engl. 1985, 24, 507

Ewen, J. A. JACS, 1984, 106, 6355.

Ligand-Controlled PP-Synthesis

• Observations of Ewen and Brintzinger led to further optimization of C2-Metallocene catalysts

metallocene productivitybTm(°C) Mw [mmmm]

6 188 132 24 000 0.789 5c 149 4 000 0.9710 1.6d 162 134 000 0.97711 190 137 36 000 0.8212 99 145 195 000 0.8813 403 146 330 000 0.8914 245 150 213 000 0.8915 875 161 920 000 0.99116 47 160 400 000 0.992“4th generation”e 20 162 900 000 >0.99

Productivity: kg of PP/(mmol of M‚h). 4th generation heterogeneous Z-N catalyst.

- Highly isotactic-PP can now be achieved with homogeneous C2 catalysts- Catalyst productivities and polymer properties are close to modern heterogeneous Z-Ncatalysts

-A question that arose was could syndiotactic polymers also be prepared


Ligand Symmetry and PP-tacticity

• Initial studies by Ewen with CS metallocenes led to syndiotactic-PP

n25 °C, 1 h

Zr-Cat-4 (1.3 µmol)MAO (10.7 wt%)

26 g PP[rrrr] = 0.89, Mw/Mn=1.9

(1.2 L)

isotactic blocks in high molecular weight, atacticchains.380 Thermoplastic elastomeric polypropene(TPE-PP) can be defined as a propene-block-ho-mopolymer. As described above, this material is nothomogeneous, as it can be fractionated by solventextraction, and in this respect is similar to the TPE-PP produced with DuPont-type catalysts.381,382 TPE-PP has been previously obtained with several differ-ent heterogeneous catalysts383 and is a material ofcommercial interest.383,384 Possible applications areas a component in car bumpers, in materials formedical applications, and in general as EPR/i-PPcompatibilizer and a-PP substitute. TPE-PP can alsobe produced by C1-symmetric metallocene catalysts,as described by Chien and others,385-395 but in thiscase the polymer is expected to be homogeneous withrespect to composition distribution (see section IV.C).

B. Syndiotactic Polypropene: Cs-SymmetricMetallocenes

The first metallocene -and the first catalyst ingeneral- able to produce highly syndiotactic polypro-pene (s-PP), was the Cs-symmetric Me2C(Cp)(9-Flu)-ZrCl2 (Cs-1 in Chart 18).112 The behavior of thiscatalyst46,113 and the characterization of s-PP397-407

have been extensively reviewed. A number of studieson the thermal behavior, crystal structures, andmorphology of s-PP have appeared in the litera-

ture.397-406 A Tm° of 214 °C and a !Hu of 1.4 kJ/molfor fully syndiotactic polypropene have been extrapo-lated.397 Hence, the Tm° of s-PP would be notablyhigher than that of i-PP (186 °C), although in practices-PP has always lower melting points than i-PP ofcomparable stereoregularity and molecular weights.More recently, a more reasonable Tm° of 182 °C hasbeen obtained by De Rosa.405 Applications of s-PP arestill under investigation; improvement of mechanicalproperties has been reported for s-PP/i-PP blends.408

The invention of syndiospecific Cs-symmetric met-allocenes has marked the turning point in the un-derstanding of the mechanism of stereocontrol withmetallocene catalysts. Again, the presence of isolatedinsertion errors of the type rrrrmmrrr is consistentwith site control (Scheme 27). In the case of thesyndiospecific Me2C(Cp)(9-Flu)ZrCl2 catalyst, in whichthe two sites are enantiotopic, occasional “skipped”insertions produce a minor amount of insertion errorsof the type rrrrmrrrr, which are identical to thoseproduced by chain-end control. In the case of isospe-cific C2-symmetric metallocenes, skipped insertionswould not be observable due to the presence of twohomotopic sites.

The statistics of syndiospecific polymerization havebeen treated in detail by Farina.409

As outlined above, the syndiospecific Cs-symmetriccatalysts are those for which the two available

Chart 18. Examples of Syndiospecific Cs-Symmetric Metallocenes

1300 Chemical Reviews, 2000, Vol. 100, No. 4 Resconi et al.

Zr-Cat-4

• Similar ligand optimizations have led to highly syndio-specific catalysts

n0°C, Liq. C3H6

Zr-Cat-4a-d

MAO

Zirconocene 4a produces polypropylene of syndiotacticitycomparable to {(Et2C)(fluorenyl)(cyclopentadienyl)}ZrCl2.8 Zir-conocene 4b3 exhibits the highest syndiospecificity: at 0°C inliquid propylene good yields of polypropylene are obtainedhaving 98.9% rrrr.

C1 zirconocene 5a performs similarly to Cs symmetric andclosely related 4b. On the other hand, chiral 5b producesmoderately syndiotactic polypropylene at 0 °C in liquid pro-pylene (entry 12), but switches to moderately isospecific at lowerpropylene concentrations and higher temperatures (entries 13and 14). A likely explanation for the behavior of these catalystsis that site epimerization by “chain swinging” may compete withinsertion of propylene at lower propylene concentrations andhigher temperatures, thus lowering syndiospecificity (Scheme1). For catalyst 5b, insertions from the same enantioface ofpropylene occur mostly from only one side of the zirconocenewedge under conditions where site epimerization is faster thanchain propagation. Hence, isotactic polypropylene is produced.

The mechanisms for these processes are presently under furtherinvestigation.

Acknowledgment. The work has been supported by the USDOEOffice of Basic Energy Sciences (Grant No. DE-FG03-85ER13431)and by Exxon Chemicals Americas. We thank Professor Terry Nileat the University of North Carolina, Greensboro, for providing adviceon the preparation of 1,3-diisopropylcyclopentadiene and Drs. TerryBurkhardt and Charles Ruff of Exxon Chemicals Americas forassistance with polymerizations and 13C NMR characterization of someof the polypropylene samples.

Supporting Information Available: Experimental details describ-ing the syntheses and characterization (including NMR data) for 2, 4a-d, and 5a-b, as well as experimental details of the polymerizations ofpropylene and characterization of the polypropylenes (28 pages). Seeany current masthead page for ordering and Internet access instructions.

JA962413Z

(8) In a parallel run (cf. entry 3 of Table 1) in liquid propylene at 60 °C,{(Et2C)(fluorenyl)(cyclopentadienyl)}ZrCl2 with 300 equiv of MAO yielded30 300 g of PP/g of catalyst/h with [r] content of 92.6%.

Table 1. Propylene Polymerizations with Catalyst Derived from4a-d and 5a-ba

entry catalyst activityf [rrrr]g [mmmm]g [r]h [m]i

1 4ab 2160 83.7 0.0 94.0 6.02 4ac 300 27.3 0.0 74.1 25.93 4ad 74200 76.0 0.1 92.6 7.44 4bb 1750 98.9 0.0 99.6 0.45 4bc 230 38.8 0.0 75.6 24.36 4cb 1730 95.9 0.0 99.0 1.07 4cc 230 33.9 0.0 75.2 24.88 4db 722 90.5 0.0 96.9 3.19 4dc 200 29.6 0.0 76.2 23.810 5ab 1500 83.1 0.0 94.4 5.611 5ac 200 20.0 4.8 62.0 38.012 5bb 930 41.8 5.6 73.5 26.513 5bc 156 0.0 61.2 14.6 85.414 5be 110 0.0 58.5 17.6 82.4

a See Supporting Information for full experimental details. b Liquidpropylene at 0 °C; 2 mg of catalyst and 2000 equiv of MAO.c Propylene (40 psig) in 35 mL toluene at 25 °C; 10 mg of catalyst and430 equiv of MAO. d Liquid propylene at 60 °C; 300 equiv of MAO.e Propylene (10 psig) in toluene at 25 °C; 10 mg of catalyst and 430equiv of MAO. f Polymer isolated (g)/catalyst (g)/hour. g Percentageof [rrrr] or [mmmm] pentad by 13C NMR analysis. h [r] ) [rr] +0.5[mr]. i [m] ) [mm] + 0.5[mr].

Scheme 1

Communications to the Editor J. Am. Chem. Soc., Vol. 118, No. 47, 1996 11989






JA962413Z






Scheme 1







JA962413Z






Scheme 1







JA962413Z






Scheme 1


A New Class of Zirconocene Catalysts for theSyndiospecific Polymerization of Propylene and ItsModification for Varying Polypropylene fromIsotactic to Syndiotactic

Timothy A. Herzog, Deanna L. Zubris, and John E. Bercaw*

Arnold and Mabel Beckman Laboratoriesof Chemical Synthesis

California Institute of TechnologyPasadena, California 91125

ReceiVed July 15, 1996

Metallocene catalysts for the polymerization of ethylene,propylene and other R-olefins are the focus of intense currentinterest.1 A wide variety of metallocene catalysts from groups3 and 4 have now been prepared. The tacticity of polypropylenevaries predictably with the structure of the metallocene cata-lyst: C2V symmetric metallocenes generally afford atacticpolypropylene; C2 symmetric metallocenes produce highlyisotactic polypropylene; C1 metallocenes also produce isotacticpolypropylene, but generally with less stereospecificity. Whereasthe types of isospecific metallocene catalysts are structurallyhighly variable, syndiotactic polypropylene has been producedusing essentially a single type of Cs symmetric ansa-metallocenecatalyst 1, or minor variants thereof.2 According to the original

proposals from Ewen and Razavi,2a the syndiospecificity arisesfrom propylene insertions occurring from alternating (enan-tiotopic) sides of the metallocene wedge with the propylenemethyl group directed away from the larger fluorenyl ligand.Our interests in preparing single-component, syndiospecificgroup 3 metallocene catalysts led us to the ligand system ofthe Cs symmetric zirconocene 2,3 seemingly closely related tothat of 1. We anticipated that the bulky [SiMe3] substituents

on the lower cyclopentadienyl ligand would function as doesthe fluorenyl ligand of 1, directing the methyl group towardthe less sterically hindered (upper) cyclopentadienyl ligand.Surprisingly, 2 in combination with 103 equiv of methylalu-

minoxane (MAO) in liquid propylene at 0 °C producesessentially atactic polypropylene (55% r diads).Re-evaluation of the relative importance of the various steric

interactions operating for propylene polymerizations with met-allocene catalysts,4 as well as experiments supporting R C-Hagostic assistance in the transition state for chain propagation,5

led us to conclude that the transition structure for propylenepolymerization with the Ewen/Razavi catalyst system is likely3.6 The preference for the polymer chain extending from the

R methylene group (P) and the methyl group of the propylenemonomer to assume a trans relationship dominates, forcing themethyl group down toward the fluorenyl ligand. The relativelyflat fluorenyl ligand with its open region between the benzosubstituents nicely accommodates this orientation. In contrast,the ligand system for 2 lacks this feature.A new ligand has been designed with 1,2-[SiMe2]2 linking

of cyclopentadienyl and 3,5-diisopropylcyclopentadienyl groups.Zirconocene 4 (eq 1)7 thus incorporates all three features of 1:(a) Cs symmetry, (b) cyclopentadienyls of differing size, and(c) steric bulk flanking the metallocene wedge with an openregion in center. Furthermore, this ligand design allows

systematic variation of the cyclopentadienyl substituent (R)contained in the mirror plane, permitting steric changes for thecatalyst while maintaining Cs symmetry (e.g., 4a-d) or loweringthe overall symmetry to C1 (e.g., 5a,b).When activated with MAO, 4a-d and 5a,b are active

catalysts for the polymerization of propylene (Table 1). Indeed,the Cs symmetric zirconocenes 4a-d are highly syndiospecific.

(1) For recent reviews and leading references, see: (a) Brintzinger, H.H.; Fischer, D.; Mulhaupt, R.; Rieger, B.; Waymouth, R. M. Angew. Chem.,Intl. Ed. Engl. 1995, 34, 1143. (b) Bochmann, M. J. Chem. Soc., DaltonTrans. 1996, 255.(2) (a) Ewen, J. A.; Jones, R. L.; Razavi, A.; Ferrara, J. D. J. Am. Chem.

Soc. 1988, 110, 6255. (b) Patsidis, K.; Alt, H. G.; Milius, W.; Palackal, S.J. J. Organomet. Chem. 1996, 509, 63. (c) Shiomura, T.; Kohno, M.; Inoue,N.; Yokote, Y.; Akiyama, M.; Asanuma, T.; Sugimoto, R.; Kimura, S.;Abe, M. Stud. Surf. Sci. Catal. 1994, 89, 326. (d) Razavi, A.; Atwood, J.L. J. Organomet. Chem. 1993, 459, 117. (e) Winter, J.; Rohrmann, M.;Dolle, V.; Spaleck, W. Eur. Pat. Appl. 0,387,690 and 0,387,691, 1991. (f)Spaleck, W.; Aulbach, M.; Bachmann, B.; Kuber, F.; Winter, A.Macromol.Symp. 1995, 89, 237-247. (g) Ewen, J. A. Macromol. Symp. 1995, 89,181-196.(3) Henling L. M.; Herzog, T. A.; Bercaw, J. E. Acta Crystallogr.

Submitted.

(4) (a) Longo, P.; Grassi, A.; Pellechi, C.; Zambelli, A.Macromolecules1987, 20, 1015. (b) Pino, P.; Galimberti, M. J. Organomet. Chem. 1989,370, 1. (c) Gilchrist, J. C.; Bercaw, J. E. J. Am. Chem. Soc. In press.(5) (a) Piers, W. E.; Bercaw, J. E. J. Am. Chem. Soc. 1990, 112, 9406-

9407. (b) Brintzinger, H. H.; Krauledat, H. Angew. Chem., Int. Ed. Engl.1990, 29, 1412. (c) Brintzinger, H. H.; Leclerc, M. J. Am. Chem. Soc.,1995, 117, 1651. (d) Burger, B. J.; Cotter, W. D.; Coughlin, E. B.; Chacon,S. T.; Hajela, S.; Herzog, T. A.; Kohn, R. O.; Mitchell, J. P.; Piers, W. E.;Shapiro, P. J.; Bercaw, J. E. In Ziegler Catalysts; Fink, G., Mulhaupt, R.,Brintzinger, H. H., Eds.; Springer: Berlin, 1995. (e) Grubbs, R. H.; Coates,G. W. Acc. Chem. Res. 1996, 29, 85.(6) Calculations are in agreement with this overall transition structure:

Cavallo, L.; Guerra, G.; Vacatello, M.; Corradini, P.Macromolecules 1991,24, 1784.(7) (a) Clark, T. J.; Killian, C. M.; Luthra, S.; Nile, T. A. J. Organomet.

Chem. 1993, 462, 247. (b) Cano, A.; Cuenca, T.; Gomez-Sal, P.; Royo, B.;Royo, P. Organometallics 1994, 13, 1688-1694. (c) Lang, H.; Blau, S.;Muth, A.; Weiss, K.; Neugenauer, U. J. Organomet. Chem. 1995, 490, C32.(d) Mengele, W.; Diebold, J.; Troll, C.; Roll, W.; Brintzinger, H. H.Organometallics 1993, 12, 1931-1935. (e) Diamond, G. M.; Rodewald,S.; Jordan, R. F. Organometallics 1995, 14, 5. (f) Christopher, J. N.;Diamond, G. M.; Jordan, R. F. Abstracts of Papers 210th National Meetingof the American Chemical Society, Chicago, Illinois, Fall 1995; AmericanChemical Society: Washington, DC, 1995; INOR310.

11988 J. Am. Chem. Soc. 1996, 118, 11988-11989

S0002-7863(96)02413-4 CCC: $12.00 © 1996 American Chemical Society

Zr-Cat-4a-d

• PP-tacticity is highly dependent on the ligand symmetryCS -> Syndiotactic C2 -> isotactic

Ewen, O., JACS, 1988, 110, 6255.

Bercaw, J. E. JACS, 1996, 118, 11988

Unifying Stereochemical Models

CS -> syndiotactic

C2 -> isotactic

- Polymer chain moves after each insertion (due to metallocyclobutane T.S.)- Coordination sites are homotopic- Same prochiral face of the olefin is inserted, regardless of the coordination site

- Coordination sites are enantiotopic- Opposite prochiral faces of the olefin insert at each coordination site


Mechanisms of Racemization

• Site epimerization in CS catalyst systems (“chain swinging”)

nToluene/or liq. C3H6, 0 °C, 0.5 h

Zr-Cat (2 mg)MAO (2000 eq.)






JA962413Z






Scheme 1


Experimental observation:1. High propene pressure: [rrrr] = 41.8(moderately syndiotactic)2. Low propene pressure: [mmmm] = 61.2(moderately isotactic)

- Author!s proposed that at low [propene] chain swinging is competitive with monomer insertion (enchainment)

R = CHMeCMe3

Bercaw, J. E. et. al. JACS, 1996, 118, 11988

Mechanisms for Racemization

Mechanisms of Chain Termination

1. Chain transfer to aluminum:

2. Beta-hydride transfer:

Chain-end epimerization


Resconi, L. et. al. Macromolecules, 1995, 28, 6667.

Exploiting Site Epimerization

• Catalysts for the synthesis of Hemisotactic-PP

n

Zr-Cat/MAO

- Hemisotactic-PP contain isospecific monomer units separated by monomers of random stereochemistry

• Mechanistic rationale

- Chiral ligand provides aspecific (random) and isospecific coordination sites


Ewen’s Symmetry Rules

-Number of well defined ligandare now available for the synthesisof PP-polymers with varying properties

- Lead to the formulation of Ewen!s symmetry rules

- What can synthetic chemists do with stereoselective homogeneous polymerization?

Resconi, L. et. al. Chem. Rev. 2000, 100, 1253.

Kinetic Resolutions of Olefins

Me

Zn Me

Me

2TiCl4,

20 °C, 236 hPolymer

Me

(2/1 ratio Zn/Ti, ~20 mol%) Recovered monomer: [!]D = -0.75°Eantiopure authentic: [!]D = 36.0°Optically purity = 2%

• Basic idea

• First experiments were performed by Pino in 1963

&

Chiral Z-NCatalyst

kS fast

kR slow

n

R R R

R

(enriched monomer)

(optically active polymer)

R

R

Where k = rate of monomer enchainment

- Selectivities were very low- Proof of principle experiment- Little was known about the catalyst structure

Pino, P. et. al. JACS 1963, 85, 3888.

Kinetic Resolutions with Zr-Metallocenes

- 13C-NMR analysis suggest Isostatic polymers were formed - Low but significant values of s

were observed for all olefins- moderate to good s for 3,4-dimethyl-1-pentene

• Homogeneous Zr-catalysts (Bercaw)

Bercaw, J. E. et. al. JACS, 2004, 126, 8216.

Mechanistic and Stereochemical Rationale

- Authors propose a site epimerization to explain the formation of primarily isotactic polymer. - Enchainment of the (S)-enantiomer on the right side of the wedge is favored due to steric interactions

Structure Activity Relationships

- No further increases in selectivity observed

- Chain-end control could be competing with catalyst

-Author!s suggested that increasing steric bulk in the lower right quadrant could lead to higher selectivities

Bercaw, J. E. et. al. JACS, 2004, 126, 8216.

Chain-End Control vs. Site Control

n!"

#

# '

$

Zr-Cat/MAO

• Polymer contains two stereogenic centers

- How can site-control and chain end control be separated- Authors proposed a co-polymerization with an achiral monomer

Scenario 1: site-control and chain-end control work in concert (s-homopoly > s-copoly)Scenario 2: site-control and chain-end control oppose each other (s-homopoly < s-copoly)

-Side-chain and main chain chirality could be influencing

monomer enchainment

Bercaw, J. E. et. al. PNAS, 2006, 103, 15303.

Kinetic Resolution of Co-polymerization

R H2C CH2

(S)-2 (1-2 mg), MAO (500/1)

Toluene/tetradecane, 25 °C(300 torr)

Co-polymer

-13CNMR and mp data were consistent with polymer structures containing few consecutive chiral repeat units- Data suggest that for all but one olefin scenario 1 is operative (s-homopoly > s-copoly)- For 3-methyl-1-pentene site-control and chain-end control work uncooperatively

- Complex interactions of end-chain and site control make it difficult to rationally design new ligands

Conclusions

Karl Ziegler-the last Al-Chemist“...because he turned aluminum into gold.”

(1) Advent of homogeneous catalysts provided a more well-defined system for a mechanistic analysis of Z-N polymerization(1I) Variation of ligand symmetry in the metallocene allowed for the synthesis of a number of different polypropylene structures(III) Kinetic resolution of olefins using homogeneous catalysts is very promising, but still in it’s infancy

Excellent Reviews To Be Aware Of

I. Excellent reviews on stereochemistry and mechanism of Ziegler-Natta Polymerizations:

(1) Brintzinger, H-H.; Fischer, D.; Mulhaupt, R.; Rieger, B.; Waymouth, R. M. Angew. Chem. Int. Ed. Engl. 1995, 34, 1143-1170.(2) Coates, G. W. Chem. Rev. 2000, 100, 1223-1252.(3) Resconi, L.; Cavallo, L.; Fait, A.; Piemontesi, F. Chem. Rev. 2000, 100, 1253-1346.

II. Review on co-catalysts and activators in Ziegler-Natta Polymerization:Chen, E. Y-X.; Marks, T. J. Chem. Rev. 2000, 100, 1391-1434.

III. Reviews on agostic interactions in organometallic complexes:(1) Brookhart, M.; Green, M. L. H.; Wong, L-L. Prog. Inorg. Chem. 1988, 36, 1-124(2) Brookhart, M.; Green, M. L. H.; Parkin, G. Proc. Nat. Acad. Sci. 2007, 104, 6908-6914.

IV. Chemical Reviews volume 100, issue 4 is dedicated to metal-catalyzed polymerization

Backup: Evidence for MAO-Cluster

Head

Documents

Transcript of Head