Synthesis of group 4 complexes that contain the tridentate diamido/donor ligands [(ArylNCH2CH2)2O]2- and zirconium complexes that contain [(ArylNCH2CH2)2S]2- and an evaluation of their activity for the polymerization of 1-hexene

Organometallics 1998, 17, 4795-4812

4795

Syn th esis of Gr ou p 4 Com p lexes Th a t Con ta in th e

Tr id en ta te Dia m id o/Don or Liga n d s [(Ar ylNCH₂CH₂)₂O]^2-

a n d Zir con iu m Com p lexes Th a t Con ta in

[(Ar ylNCH₂CH₂)₂S]^2-a n d a n Eva lu a tion of Th eir Activity

for th e P olym er iza tion of 1-Hexen e

Michael Aizenberg, Laura Turculet, William M. Davis,

Florian Schattenmann, and Richard R. Schrock*

Department of Chemistry 6-331, Massachusetts Institute of Technology,

77 Massachusetts Avenue, Cambridge, Massachusetts 02139

Received J une 22, 1998

Compounds of the type (ArylNHCH₂CH₂)₂O (Aryl ) 2,6-Me₂C₆H₃(H₂[1a ]), 2,6-Et₂C₆H₃

(H₂[1b]), 2,6-i-Pr₂C₆H₃(H₂[1c])) can be prepared by treating (TsOCH₂CH₂)₂O (TsO ) tosylate)

with the lithium anilides in THF. [1a ,b]TiCl₂, [1a ,b]TiMe₂, [1a ]Ti(CH₂Ph)₂, [1a -c]M(NMe₂)₂

(M ) Zr or Hf), [1a -c]MCl₂, and [1a -c]MR₂(R ) Me, Et, i-Bu) were prepared. An X-ray

study of [1a ]Ti(CH₂Ph)₂revealed the structure to be a distorted trigonal bipyramid (type B)

in which the two amido nitrogens and one benzyl ligand occupy equatorial positions. An

X-ray study of [1a ]ZrMe₂showed it to be a distorted trigonal bipyramid that contains “axial”

amido groups (type A), while an X-ray study of [1c]HfEt₂revealed it to have a structure

halfway between type A and type B, i.e., a distorted square pyramid with one alkyl in the

apical position. Analogous compounds were also prepared that contain a sulfur donor instead

of oxygen, i.e., [(2,6-Me₂C₆H₃NHCH₂CH₂)₂S] (H₂[2a ]), [(2,6-i-Pr₂C₆H₃NHCH₂CH₂)₂S] (H₂[2c]),

[2a ,c]Zr(NMe₂)₂, [2a ,c]ZrCl₂, [2a ,c]ZrMe₂, and [2c]Zr(CH₂CHMe₂)₂. An X-ray study of [2a ]-

ZrMe₂revealed it to be closest to a type B structure. Addition of 1 equiv of [PhNMe₂H]-

[B(C₆F₅)₄] in C₆D₅X (X ) Br, Cl) to [1a ,c]MMe₂(M ) Zr, Hf) gave cationic complexes that

contain coordinated dimethylaniline, with which free aniline does not exchange readily on

the NMR time scale at 60 °C. Addition of excess ether to {[1a ]MMe(NMe₂Ph)}[B(C₆F₅)₄]

(M ) Zr, Hf) led to {[1a ]MMe(ether)}[B(C₆F₅)₄] (M ) Zr, Hf) complexes in high yield.

Analogous cations can be prepared in the sulfur ligand system, but they do not appear to be

as stable as in the oxygen ligand system. Zr and Hf dimethyl complexes that contain an

oxygen donor or a sulfur donor ligand can be activated with [Ph₃C][B(C₆F₅)₄] to yield efficient

catalysts for polymerization of 1-hexene, although the molecular weight of the poly(1-hexene)

chains is limited to ∼20 000-∼25 000 under the conditions employed. Neither {[1c]ZrMe-

(ether)}[B(C₆F₅)₄] nor {[1c]HfMe(ether)}[B(C₆F₅)₄] will polymerize 1-hexene in C₆D₅Br at

room temperature, and neither will polymerize ethylene readily at 1 atm and 25 °C. It is

proposed that a solvated five-coordinate cation must lose the solvent in order to react with

an olefin and that â-hydride elimination in the four-coordinate cation limits chain length.

In tr od u ction

bound to the metal.⁵Most recently group 4 complexes

that contain chelating diamido ligand systems have

received increased attention as potential olefin polym-

erization catalysts.^6-32Two systems of greatest inter-

est, largely because of their ability to behave as living

Significant advances have been made in the synthesis

of “well-defined” olefin polymerization catalysts of the

early metals (largely group 4) in the last 15 years, the

vast majority of which contain two cyclopentadienyl-like

rings or, more recently, one cyclopentadienyl-like ring.^1-4

The active species in such complexes is now widely

regarded to be a group 4 metal cation formed in the

presence of a “weakly coordinating” anion such as

[B(C₆F₅)₄]^-. However, in the past decade in particular

researchers have been looking for group 4 metal cations

that do not have one or more cyclopentadienyl rings

(5) For leading references see: (a) Martin, A.; Uhrhammer, R.;

Gardner, T. G.; J ordan, R. F.; Rogers, R. D. Organometallics 1998, 17,

382. (b) Bei, X. H.; Swenson, D. C.; J ordan, R. F. Organometallics 1997,

16, 3282. (c) Tjaden, E. B.; Swenson, D. C.; J ordan, R. F.; Petersen, J .

L. Organometallics 1995, 14, 371.

(6) Scollard, J . D.; McConville, D. H. J . Am. Chem. Soc. 1996, 118,

10008.

(7) Scollard, J . D.; McConville, D. H.; Payne, N. C.; Vittal, J . J .

Macromolecules 1996, 29, 5241.

(8) Scollard, J . D.; McConville, D. H.; Vittal, J . J . Organometallics

1995, 14, 5478.

(9) Scollard, J . D.; McConville, D. H.; Rettig, S. J . Organometallics

1997, 16, 1810.

(10) Aoyagi, K.; Gantzel, P. K.; Kalai, K.; Tilley, T. D. Organome-

tallics 1996, 15, 923.

(1) Brintzinger, H. H.; Fischer, D.; Mulhaupt, R.; Rieger, B.;

Waymouth, R. M. Angew. Chem., Int. Ed. Engl. 1995, 34, 1143.

(2) Kaminsky, W.; Ardnt, M. Adv. Polym. Sci. 1997, 127, 144.

(3) Bochmann, M. J . Chem. Soc., Dalton Trans. 1996, 255.

(4) Chen, Y.-X.; Marks, T. J . Organometallics 1997, 16, 3649.

Publication on Web 09/30/1998

4796 Organometallics, Vol. 17, No. 22, 1998

Aizenberg et al.

polymerization catalysts, are dialkyl complexes of tita-

nium that contain diamidoligands such as [ArylNCH₂CH₂-

CH₂NAr]^2-(Aryl ) 2,6-i-Pr₂C₆H₃),^6-9and dialkyl com-

plexes of zirconium that contain the “diamido/donor”

ligand [((t-Bu-d₆)N-o-C₆H₄)₂O]^2-([NON]^2-).^11,12The

living intermediates in the latter system are stable in

the absence of 1-hexene and have been observed in ¹³C

labeling experiments to be consistent with 1,2-insertion

of the olefin into the metal-carbon bond.¹²We recently

reported in a preliminary fashion complexes that con-

tain “diamido/donor” variations of the [NON]^2-ligand,

[(2,6-R₂C₆H₃NCH₂CH₂)₂O]^2-(R ) Me, i-Pr), and pre-

liminary experiments concerning their use as catalysts

for the polymerization of 1-hexene.²⁵In this paper we

report the full details of studies concerning complexes

of Ti, Zr, and Hf that contain a tridentate ligand of the

general formula [(ArylNCH₂CH₂)₂O]^2-, along with more

abbreviated related studies of zirconium complexes that

contain a tridentate ligand of the general formula

separated readily from the (ArylNHCH₂CH₂)₂O prod-

ucts. The N-substituted morpholine is the major prod-

uct when Aryl ) 2-t-BuC₆H₄because of the facile

intramolecular displacement of the second tosyl group

after the first substitution. Therefore, the anilide must

be protected first with a trimethylsilyl group, which

then allows the protected product to be synthesized in

good yield (eq 2). Subsequent hydrolysis with aqueous

(TsOCH₂CH₂)₂O ^{2LiN(SiMe )Aryl}8

3

[(ArylNCH₂CH₂)₂S]^2-

.

THF

[ArylN(SiMe₃)CH₂CH₂]₂O (2)

Resu lts

Aryl ) 2-t-BuC₆H₄

Syn t h esis of Com p lexes Th a t Con t a in [(Ar yl-

NCH₂CH₂)₂O]^2-Liga n d s. Compounds of the type

(ArylNHCH₂CH₂)₂O (Aryl ) 2,6-Me₂C₆H₃, 2,6-Et₂C₆H₃,

2,6-i-Pr₂C₆H₃) can be prepared by treating (TsOCH₂-

CH₂)O (TsO ) tosylate) with the lithium anilides in

THF (eq 1). The crude products, which are sufficiently

pure for further use, are produced in 50-80% yield as

oils (H₂[1b] or H₂[1c]) or as a crystalline solid (H₂[1a ]).

The side products, N-substituted morpholines, can be

HCl yields (2-t-BuC₆H₄NHCH₂CH₂)₂O (H₂[1d ]) in an

overall yield of 50%.

The reaction between TiCl₄(THF)₂and 2 equiv of Li₂-

[1a ] or Li₂[1b] (prepared in situ) gave the titanium

dichloride complexes [1a ]TiCl₂and [1b]TiCl₂, respec-

tively (eq 3). The dichloride complexes react smoothly

Li₂[1a ,b] + TiCl₄(THF)₂f [1a ,b]TiCl₂^2MeMgCl8

[1a ,b]TiMe₂(3)

(11) Baumann, R.; Davis, W. M.; Schrock, R. R. J . Am. Chem. Soc.

1997, 119, 3830.

(12) Baumann, R.; Schrock, R. R. J . Organomet. Chem. 1998, 557,

69.

(13) Horton, A. D.; de With, J .; van der Linden, A. J .; van de Weg,

H. Organometallics 1996, 15, 2672.

(14) Horton, A. D.; de With, J . Chem. Commun. 1996, 1375.

(15) Cloke, F. G. N.; Geldbach, T. J .; Hitchcock, P. B.; Love, J . B. J .

Organomet. Chem. 1996, 506, 343.

(16) Cloke, F. G. N.; Hitchcock, P. B.; Love, J . B. J . Chem. Soc.,

Dalton Trans. 1995, 25.

(17) Clark, H. C. S.; Cloke, F. G. N.; Hitchcock, P. B.; Love, J . B.;

Wainwright, A. P. J . Organomet. Chem. 1995, 501, 333.

(18) Warren, T. H.; Schrock, R. R.; Davis, W. M. Organometallics

1996, 15, 562.

(19) Warren, T. H.; Schrock, R. R.; Davis, W. M. Organometallics

1998, 17, 308.

(20) Gue´rin, F.; McConville, D. H.; Vittal, J . J . Organometallics

1997, 16, 1491.

(21) Gue´rin, F.; McConville, D. H.; Payne, N. C. Organometallics

1996, 15, 5085.

with methyl Grignard reagent to yield the titanium

dimethyl complexes [1a ,b]TiMe₂as yellow to yellow-

brown powders. Proton and carbon NMR spectra of [1a ]-

TiMe₂and [1b]TiMe₂are consistent with species that

contain two mirror planes on the NMR time scale.

Sharp singlet resonances belonging to the methyl groups

bound to titanium are found in the proton NMR

spectrum at 0.90 and 0.89 ppm and in the carbon NMR

spectrum at 60.75 and 61.14 ppm, for [1a ]TiMe₂and

[1b]TiMe₂, respectively. In [1a ]TiMe₂the two methyl

groups are still equivalent on the NMR time scale at

-70 °C (toluene-d₈, 500 MHz). On the basis of these

data we cannot determine if the ground-state structure

is of type A or type B for [1a ]TiMe₂and [1b]TiMe₂(eq

4). If the lowest energy structure is of type B, which

(22) Gue´rin, F.; McConville, D. H.; Vittal, J . J . Organometallics

1995, 14, 3154.

(23) Gue´rin, F.; McConville, D. H.; Vittal, J . J . Organometallics

1996, 15, 5586.

(24) Gibson, B. C.; Kimberley, B. S.; White, A. J . P.; Williams, D.

J .; Howard, P. Chem. Commun. 1998, 313.

(25) Schrock, R. R.; Schattenmann, F.; Aizenberg, M.; Davis, W. M.

Chem. Commun. 1998, 199.

(26) Male, N. A. H.; Thornton-Pett, M.; Bochmann, M. J . Chem. Soc.,

Dalton Trans. 1997, 2487.

(27) Herrmann, W. A.; Denk, M.; Albach, R. W.; Behm, J.; Herdtweck,

E. Chem. Ber. 1991, 124, 683.

(28) Kempe, R.; Brenner, S.; Arndt, P. Organometallics 1996, 15,

1071.

(29) Friedrich, S.; Schubart, M.; Gade, L. H.; Scowen, I. J .; Edwards,

A. J .; McPartlin, M. Chem. Ber.-Recl. 1997, 130, 1751.

(30) Tsuie, B.; Swenson, D. C.; J ordan, R. F.; Petersen, J . L.

Organometallics 1997, 16, 1392.

(31) Tinkler, S.; Deeth, R. J .; Duncalf, D. J .; McCamley, A. Chem.

Commun. 1996, 2623.

would be analogous to the structure of [1a ]Ti(CH₂Ph)₂

(see below), then a structure of type A must be readily

accessible by a Berry-type process in which the two

equatorial amido nitrogens swing into axial positions

while R_axand the oxygen donor end up in equatorial

positions. As we shall see, structures A and B are close

(32) Schattenmann, F.; Schrock, R. R.; Davis, W. M. Organometallics

1998, 17, 989.

Group 4 Complexes with Diamido/ Donor Ligands

Organometallics, Vol. 17, No. 22, 1998 4797

Å), perhaps in part as a consequence of the greater

donor ability of the dialkyl ether oxygen, although

differences associated with different degrees of flexibility

of the diamido/donor ligand backbone in [(t-BuN-o-

C₆H₄)₂O]TiMe₂and [1a ]Ti(CH₂Ph)₂cannot be dis-

counted. Somewhat surprisingly, both benzyl ligands

are bound in an η¹fashion, as evidenced by the values

of the angles Ti-C(1)-C (122.9(6)°) and Ti-C(2)-C

(121.0(6)°). These structural features suggest that the

relatively complex set of resonances observed in solution

at low temperature in the ¹H NMR spectrum of

[1a ]Ti(CH₂Ph)₂does not involve formation of an η^x-CH₂-

Ph ligand where x > 1 and, furthermore, that the

relatively slow equilibration of the benzyl groups in [1a ]-

Ti(CH₂Ph)₂relative to the methyl groups in [1a ]TiMe₂

might be ascribed to a greater degree of steric crowding

in an intermediate structure of type A for [1a ]Ti(CH₂-

Ph)₂than in a structure of type A for [1a ]TiMe₂. The

angle between the N(1)-Ti-O and N(2)-Ti-O planes

is 137°, 17° larger than in an ideal trigonal bipyramid,

as a consequence of repulsion between the two substi-

tuted amido nitrogen atoms. This angle can be em-

ployed as a measure of the structural type (A or B).

The reaction between M(NMe₄)₄(M ) Zr, Hf) and H₂-

[1a -c] produced the white crystalline [1a -c]M(NMe₂)₂

complexes in high yield (eq 5). The [1a -c]M(NMe₂)₂

F igu r e 1. ORTEP drawing of the structure of [1a ]Ti-

(CH₂C₆H₅)₂.

in energy; therefore, the titanium dimethyl and dibenzyl

complexes may not have similar structures.

Red, crystalline [1a ]Ti(CH₂Ph)₂can be prepared in a

slow reaction between Ti(CH₂Ph)₄and H₂[1a ], although

the isolated yield is low (∼15%). It was also prepared

in ∼35% isolated yield (on a small scale and not

optimized) by treating [1a ]TiCl₂with 2 equiv of benzyl

Grignard reagent. Its 300 MHz room-temperature

proton NMR spectrum exhibits broad resonances for the

benzyl groups (δ(H_R) ∼2.32 ppm) and somewhat broad-

ened resonances for the protons of the backbone and the

ligand methyl groups. In the room-temperature ¹³C-

{¹H} spectrum, measured at either 75 or 125 MHz, all

resonances can be observed except that for the benzylic

carbon of the titanium-bound benzyl groups. At 65 °C

a benzylic carbon resonance can be observed as a broad

singlet at δ 91.7 ppm (125 MHz, C₆D₆). At -70 °C (500

MHz in toluene-d₈) the proton NMR spectrum is con-

sistent with a static structure on the NMR time scale

in which two different benzyl groups are present,

although its complexity prevented ready assignment of

all resonances.

M(NMe₂)₄+ H₂[1a -c] f

2HNMe₂+ [1a -c]M(NMe₂)₂(5)

M ) Zr, Hf

[1a -c]M(NMe₂)₂+ 2Me₃SiCl f

[1a -c]MCl₂+ 2Me₃SiNMe₂(6)

M ) Zr, Hf

[1a -c]MCl₂+ 2RMgX f [1a -c]MR₂

M ) Zr, Hf; R ) Me, Et, i-Bu

(7)

complexes react with an excess of Me₃SiCl in ether to

form relatively insoluble [1a -c]MCl₂complexes (eq 6).

The relatively low solubility of the [1a -c]MCl₂species

suggests that they may be dimers in the solid state, as

is true of {[NON]ZrCl₂}₂.³³Finally, the [1a -c]MCl₂

species can be alkylated by Grignard reagents to afford

the corresponding dialkyl complexes [1a -c]MR₂(R )

Me, Et, i-Bu) in excellent yield (eq 7). The isobutyl

species are stable at room temperature both in the solid

state and in solution, which suggests that they are not

prone to facile decomposition by â-hydride processes,

e.g., to give isobutane and “[1a ]M(CH₂CMe₂),” or by

γ-hydride abstraction or elimination processes. Several

exceptions to these general reactivity patterns should

be mentioned. Attempts to prepare complexes that

contain the unsymmetrical ligand H₂[1d ] invariably led

to mixtures of isomers for bis(dimethylamido), dichlo-

ride, and dimethyl complexes; NMR spectra suggest that

they are mixtures of meso and racemic species formed

as a consequence of restricted rotation (on the NMR

time scale) around the N-C_ipsobond. Full assignment

of NMR spectra therefore was impractical. We do not

An X-ray study of [1a ]Ti(CH₂Ph)₂(Figure 1, Tables 1

and 2) confirmed the structure to be one of type B; the

two nitrogens and one benzyl ligand (C(1)) occupy

“equatorial” positions, while the oxygen and another

benzyl group (C(2)) occupy approximately axial sites. In

accord with this description, the C(2)-Ti-O angle is

169.3(3)° while the C(1)-Ti-O angle is only 92.4(3)°.

The latter should be compared with the O-Ti-N(1)

angle (76.1(3)°) and the O-Ti-N(2) angle (75.2(3)°). The

sum of angles around the Ti center in the equatorial

plane equals 351.6(3)°. The Ti-C(1) and Ti-C(2) bond

lengths (2.129(10) and 2.118(8) Å) are statistically

indistinguishable and close to the values observed in

[(t-BuN-o-C₆H₄)₂O]TiMe₂,¹¹while both equatorial Ti-N

bond lengths (1.940(7) and 1.962(6) Å) are in the range

of other Ti-amido bond lengths reported recently (1.84-

1.98 Å).^{10,11,21,29,31}The Ti-N-C bond angles (127.0(5)

and 129.4(6)°) are larger than corresponding angles in

other structures to be discussed below, which is perhaps

evidence of a greater degree of steric crowding in the

“coordination pocket” of [1a ]Ti(CH₂Ph)₂. The Ti-O

bond length is substantially shorter in [1a ]Ti(CH₂Ph)₂

(2.224(5) Å) than in [(t-BuN-o-C₆H₄)₂O]TiMe₂(2.402(4)

(33) Baumann, R.; Schrock, R. R.; Davis, W. M. Unpublished results.

4798 Organometallics, Vol. 17, No. 22, 1998

Aizenberg et al.

Ta ble 1. Cr ysta llogr a p h ic Da ta , Collection P a r a m eter s, a n d Refin em en t P a r a m eter s for [1a ]Ti(CH₂P h )₂,

a

[1a ]Zr Me₂, [1c]HfEt₂, a n d [2a ]Zr Me₂

[1a ]Ti(CH₂Ph)₂

[1a ]ZrMe₂

[1c]HfEt₂

[2a ]ZrMe₂

empirical formula

fw

C₃₄H₄₀N₂OTi

540.58

C₂₂H₃₂N₂OZr

431.73

C₃₂H₅₂N₂OHf

659.27

C₂₂H₃₂N₂SZr

447.79

cryst color

cryst dimens (mm)

cryst syst

a (Å)

b (Å)

c (Å)

red

colorless

0.20 × 0.20 × 0.12

orthorhombic

12.741(3)

22.663(5)

7.518(2)

90

colorless

0.40 × 0.25 × 0.25

monoclinic

12.396(2)

16.662(4)

16.306(4)

90

109.70(2)

90

3170.8(11), 4

P2₁/c

colorless

0.20 × 0.20 × 0.20

monoclinic

11.143(3)

14.812(3)

13.976(3)

90

105.84(2)

90

2219.2(8), 4

P2₁/c

0.18 × 0.10 × 0.08

orthorhombic

32.803(8)

42.342(13)

8.547(4)

90

R (deg)

â (deg)

90

γ (deg)

V (Å³), Z

space group

D_calcd(Mg/m³)

11872(7), 16

Fdd2

1.210

0.316

4608

2170.9(8), 4

Pnma

1.321

0.519

904

1.381

3.314

1352

1.340

0.597

936

µ (mm^-1

F₀₀₀

)

no. of rflns collected

no. of indep rflns

R_int

6849

2636

0.0888

8145

1607

0.0703

12802

4534

0.0282

8920

3189

0.0419

data/restraints/param

R1 [I > 2σ(I)]

wR2

2507/1/344

0.0609

0.1269

1606/0/125

0.0368

0.1009

4530/0/326

0.0202

0.0526

3187/0/236

0.0321

0.0864

GOF on F²

1.095

1.179

1.086

1.137

a

All data were collected on a Siemens SMART/CCD diffractometer with λ(Mo KR) ) 0.710 73 Å using ω scans at 183(2) K and solved

using a full-matrix least-squares refinement on F². No absorption correction was applied in any case.

Ta ble 2. Selected Bon d Len gth s (Å) a n d An gles (d eg) in [1a ]Ti(CH₂P h )₂, [1a ]Zr Me₂, [1c]HfEt₂, a n d

a

[2a ]Zr Me₂

[1a ]Ti(CH₂Ph)₂

[1a ]ZrMe₂

[1c]HfEt₂

[2a ]ZrMe₂

M-N1

M-N2

M-C1

M-C2

M-D

1.962(6)

1.940(7)

2.129(10)

2.118(8)

2.224(5)

2.084(3)

2.255(6)

2.253(6)

2.336(3)

2.087(3)

2.081(3)

2.232(4)

2.224(4)

2.343(2)

2.073(3)

2.067(3)

2.261(3)

2.279(3)

2.8046(9)

N1-M-N2

N1-M-C1

N1-M-C2

N1-M-D

N2-M-C1

N2-M-C2

N2-M-D

C1-M-C2

C1-M-D

C2-M-D

C3-D-C4

M-D-C4

M-D-C3

M-N1-C

M-N2-C

M-C1-C

M-C2-C

N1-M-D/

N2-M-D^b

128.6(3)

109.2(3)

96.6(3)

76.1(3)

113.8(3)

104.3(3)

75.2(3)

97.4(4)

92.4(3)

169.3(3)

113.3(7)

111.4(5)

107.9(5)

127.0(5)

129.4(6)

122.9(6)

121.0(6)

137

139.1(2)

102.9(1)

103.1(1)

69.54(9)

102.9(1)

103.1(1)

69.54(9)

100.0(2)

129.7(2)

130.2(2)

113.7(4)

114.9(2)

120.2(2)

136.37(11)

104.09(12)

101.68(13)

70.66(10)

103.33(13)

103.75(14)

70.52(10)

104.0(2)

103.05(11)

152.89(13)

113.2(3)

116.2(2)

115.7(2)

125.27(10)

112.58(12)

102.06(11)

71.49(7)

110.46(12)

100.91(11)

70.55(7)

101.37(13)

97.96(10)

160.60(9)

105.0(2)

88.42(11)

99.44(11)

110.1(2)

120.1(2)

112.7(2)

117.1(3)

116.4(2)

180

160

140

a

b

D ) donor (O or S); M ) Ti, Zr, Hf. The external angles between the planes, ideally 120° in type B, 180° in type A; obtained from

a Chem 3D model.

know whether meso and rac isomers are configuration-

ally stable on the chemical time scale at room temper-

ature. The reaction between [1a ]ZrCl₂and 2 equiv of

Me₃CCH₂MgCl gave only [1a ]Zr(CH₂CMe₃)Cl, while

[1a ]Zr(CH₂CMe₃)₂was formed using neopentyllithium

as an alkylating reagent. This behavior is a fairly

typical consequence of some resistance to formation of

a relatively crowded dialkyl species using the Grignard

reagent, a generally “weaker” alkylating agent than a

lithium reagent.

backbone are found as two separate triplet resonances

(sometimes distorted toward an AB pattern) for the

dimethylamido and dichloride complexes, while for the

dialkyl complexes these resonances are observed as two

second-order multiplets; (ii) only one set of resonances

for two metal-bound dimethylamido or alkyl groups is

observed in both proton and carbon NMR spectra,

indicative of either a structure in which two mirror

planes are present (type A) or a process that rapidly

interconverts two type B structures via a type A

intermediate on the NMR time scale; (iii) all complexes

that contain the [1c]^2-ligand reveal two sets of reso-

nances for the isopropyl methyl protons but only one

resonance belonging to the methine protons, consistent

1

The H and ³¹C{¹H} NMR spectra of the Zr and Hf

diamido, dichloro, and dialkyl complexes are entirely

analogous. The most noteworthy NMR spectral features

are the following: (i) the protons of the OCH₂CH₂N

Group 4 Complexes with Diamido/ Donor Ligands

Organometallics, Vol. 17, No. 22, 1998 4799

of this structure is approximately halfway between the

core structure of [1a ]ZrMe₂and that of [1a ]Ti(CH₂Ph)₂,

as shown by the angle between the N(1)-M-O and

N(2)-M-O planes (160°), which is virtually halfway

between the corresponding angles in [1a ]ZrMe₂(180°)

and [1a ]Ti(CH₂Ph)₂(137°). The angles around the

oxygen donor are similar in all three structures, as are

the sums of the angles at oxygen (332.6, 343.5, and

345.1°). The C(1)-Hf-C(2) angle (104.0(2)°) is slightly

larger than the C-M-C angles in [1a ]ZrMe₂and

[1a ]Ti(CH₂Ph)₂. Since the preferred orientation of a 2,6-

disubstituted phenyl ring appears to be roughly per-

pendicular to the MNC_ipsoplane, a structure of type A

appears to be more capable of accommodating large

substituents in the ortho positions of the aryl ring than

a structure of type B. However, steric interaction

between large ortho substituents and large alkyl groups

on the metal in a structure of type A appears to lead to

a distortion of that structure toward type B, even

though a type B structure may not be possible with the

largest ortho substituents on the aryl rings. Therefore,

the observed structure of [1c]HfEt₂is the best compro-

mise between type A and type B, perhaps on the basis

of steric interactions alone.

F igu r e 2. ORTEP drawing of the structure of [1a ]Zr-

(CH₃)₂.

The structural data for [1a ]Ti(CH₂Ph)₂, [1a ]ZrMe₂,

and [1c]HfEt₂, combined with the solution NMR behav-

ior, suggest that the energy differences between struc-

tures of type A and structures of type B, or any variation

between A and B, are not large. Within the oxygen

donor system we propose that the structure of type A

is favored until steric interactions distort it toward a

type B structure and, if sterically possible, all the way

to a type B structure.

Syn th esis of Zir con iu m Com p lexes Th a t Con -

ta in [(Ar ylNCH₂CH₂)₂S]^2-Liga n d s. [(2,6-Me₂C₆H₃-

NHCH₂CH₂)₂S] (H₂[2a ]) and [(2,6-i-Pr₂C₆H₃NHCH₂-

CH₂)₂S] (H₂[2c]) can be prepared in two steps, as shown

in eq 8. The ditosylate intermediate was formed es-

F igu r e 3. ORTEP drawing of the structure of [1c]Hf(CH₂-

CH₃)₂.

with the presence of a mirror plane that interconverts

aryl rings and with hindered rotation of aryl rings about

N-C_ipsobonds on the NMR time scale.

An X-ray study of [1a ]ZrMe₂(Figure 2, Tables 1 and

2) shows it to be a type A structure in which a mirror

plane is defined by Zr, O, C(1), and C(2). None of the

Zr-ligand bond lengths is unusual, being ∼0.12(1) Å

longer than in [1a ]Ti(CH₂Ph)₂, as one would expect. The

two zirconium-bound methyl groups are technically

inequivalent, as a consequence of the fact that the

oxygen donor is not planar (the sum of the angles

around O is 343.5°), but the Zr-C(1) and Zr-C(2) bond

lengths are statistically identical, as are the O-Zr-C(1)

and O-Zr-C(2) angles, and there is no evidence that

any structure of lower than C_2vsymmetry on the NMR

time scale is accessible down to -80 °C. The angle

between the N(1)-M-O and N(2)-M-O planes in [1a ]-

ZrMe₂is 180°, which is characteristic of a structure of

type A. Interestingly, the C(1)-M-C(2) angles in [1a ]-

ZrMe₂and [1a ]Ti(CH₂Ph)₂are within 3° of one another.

Therefore, the process of converting a structure of type

B to a structure of type A could be viewed as a “rotation”

of the C(1)-M-C(2) unit in the C(1)-M-O-C(2) plane

as the angle between the N(1)-M-O and N(2)-M-O

planes increases from 137 to 180°.

excess

(HOCH₂CH₂)₂S ^2TsCl8 (TsOCH₂CH₂)₂S

8

KOH

ArylNH₂

(ArylNHCH₂CH₂)₂S

(8)

H₂[2a ,c]

H₂[2a ], Aryl ) 2,6-Me₂C₆H₃; H₂[2c], Aryl )

2,6-i-Pr₂C₆H₃

sentially quantitatively and was used without purifica-

tion in order to minimize the consequences of its

apparent slow decomposition. Reaction of the ditosylate

with the neat, excess 2,6-disubstituted aniline followed

by removal of the unreacted aniline in vacuo gave high

yields of H₂[2a ] and H₂[2c]. H₂[2a ] could be crystallized

in 70% yield from cold pentane, while H₂[2c] was

obtained as a viscous orange oil that could be used

without further purification. Dimethylamido, dichloro,

and dimethyl complexes ([2a ,c]Zr(NMe₂)₂, [2a ,c]ZrCl₂,

and [2a ,c]ZrMe₂) could all be prepared by methods

analogous to those shown in eqs 5-7. [2a ]ZrMe₂

appears to be thermally unstable at room temperature

(a solution in benzene changed color from light yellow

to dark brown in less than 1 h), while [2c]ZrMe₂is much

more stable in solution. [2c]Zr(CH₂CHMe₂)₂is the

product (in 46% yield) of the reaction between 2 equiv

An X-ray study of [1c]HfEt₂(Figure 3, Tables 1 and

2) suggests that this molecule is best described as a

distorted square pyramid with C(1) in the apical position

and C(2), N(1), N(2), and O in the basal plane. The core

4800 Organometallics, Vol. 17, No. 22, 1998

Aizenberg et al.

F igu r e 4. Proton NMR spectra (in toluene-d₈) of [2a ]Zr-

(NMe₂)₂(*, resolved and average resonances for the NMe₂

ligands; #, resolved and average resonances for the methyl

groups in the 2,6-dimethylphenylimido ligand).

F igu r e 5. Variable-temperature proton NMR spectra of

[2a ]ZrMe₂in toluene-d₈(*, ZrMe₂resonance(s); #, aryl

methyl resonances in [2a ]^2-).

of (Me₂CHCH₂)MgCl and [2c]ZrCl₂; NMR samples of

[2c]Zr(CH₂CHMe₂)₂in toluene-d₈show no signs of

decomposition after being heated to 80 °C in the process

of recording NMR spectra. We believe that the instabil-

ity of [2a ]ZrMe₂is a consequence of some intermolecular

decomposition reaction that is inherently faster for

methyl complexes that contain the [2a ]^2-ligand than

for methyl complexes that contain the [1a ]^2-ligand,

while intermolecular decomposition is inherently slower

for complexes that contain the bulkier [2c]^2-ligand than

for complexes that contain the [2a ]^2-ligand.

methyl resonances in the 2,6-diisopropylphenylimido

group. Complexes [2a ,c]ZrCl₂appear to contain only a

C₂axis on the basis of four different isopropyl methyl

groups being observed in NMR spectra at room tem-

perature. Therefore, we believe that complexes [2a ,c]-

ZrCl₂are dimers containing two bridging chlorides but

having no plane of symmetry.

In [2a ]ZrMe₂the two zirconium-methyl resonances

that are observed at 0.6 and -0.2 ppm at -20 °C or

below coalesce at ∼7 °C (Figure 5). The average ZrMe

resonance is again not found strictly at the average

position, as a consequence of temperature-dependent

chemical shifts over the 120 °C range. In [2c]ZrMe₂the

two zirconium methyl resonances coalesce at ∼14 °C.

Since ν_oat the coalescence temperature is not known

accurately from these data, we cannot determine the

barrier for this rearrangement process accurately. If

we do assume the value of ν_ofound at low temperatures,

then we find that the rate constants at the coalescence

temperature in [2a ]ZrMe₂and [2c]ZrMe₂are 530 and

600 s^-1, respectively, or ∆G^q≈ 13 kcal mol^-1. In the

¹³C NMR spectrum of [2a ]Zr(¹³CH₃)₂at -40 °C (in

CD₂Cl₂) the two zirconium-methyl carbon resonances

are observed at 41.9 and 38.0 ppm, while in [2c]Zr-

(¹³CH₃)₂at 0 °C (in toluene-d₈) they are observed at 45.5

and 39.3 ppm.

A major difference between the sulfur-donor com-

plexes and the oxygen-donor complexes is that type B

complexes are readily observable in sulfur donor com-

plexes. For example, proton spectra of dimethylamido

complexes at -60 °C or below reveal two types of

dimethylamido ligands (each containing equivalent

methyl groups) and two types of ortho substituents in

the 2,6-disubstituted phenylamido ligand, consistent

with a “trigonal” structure of type B in which rotation

about the N-C_ipsobond is slow on the NMR time scale.

In [2a ]Zr(NMe₂)₂, for example, the dimethylamido pro-

ton resonances are found at 2.26 and 3.15 ppm at -60

°C (Figure 4). Above 60 °C the resonances for the two

dimethylamido ligands coalesce to give an average

resonance at 2.58 ppm. The ortho methyl group reso-

nances and backbone resonances also coalesce to give

resonances consistent with C_2vsymmetry on the NMR

time scale. In the ¹³C{¹H} NMR spectrum of [2a ]Zr-

(NMe₂)₂at room temperature a broad amido methyl

carbon resonance is observed at 43.3 ppm. It should

be noted that there is a dramatic temperature depen-

dence of most resonances over the 160 °C range in the

proton NMR spectra shown in Figure 4, with the

“average” resonances not appearing at the average

chemical shift. Therefore, we believe that over the 160

°C temperature range the temperature-dependent pro-

cess cannot be described as simply as interconversion

of two structures of type B via an intermediate of type

A but as a change in the nature of the average lowest

energy conformation itself. As we alluded to in the

previous section, the energy differences between various

nonideal intermediate structures are small, and over a

temperature range of 160 °C the “average” of the

distribution could change considerably. Interestingly,

rotation about the C_ipso-N bond is slow on the NMR

time scale even at 100 °C in [2c]Zr(NMe₂)₂, where the

molecules have a plane of symmetry on the NMR time

scale, as evidenced by the observation of two isopropyl

An X-ray structural study of [2a ]ZrMe₂revealed the

structure to be closest to type B (Tables 1 and 2; Figure

6), on the basis of the angle between the N(1)-Zr-S

and N(2)-Zr-S planes being 140° and the S-Zr-C(2)

angle being 160.60(9)°. The C(1)-Zr-C(2) angle (101.37-

(13)°) is again approximately what it is in the other

three structures reported here. The Zr-S bond length

(2.8046(9) Å) is slightly longer than that reported for

zirconium-sulfur dative bonds in [(t-BuN-o-C₆H₄)₂S]-

ZrCl(NMe₂) (2.7273(10) Å)³⁴or [Cp^*ZrMe(tht)]⁺[BPh₄]^-

2

(2.730(4) Å).³⁵The C(3)-S-C(4) angle (105.0(2)°) is

distinctively smaller than the C-O-C angle in the three

compounds that contain the oxygen donor (∼113°),

comparable to the corresponding C-S-C angle of [(t-

BuN-o-C₆H₄)₂S]ZrCl(NMe₂) (106.1(2)°).³⁴The angle at

sulfur should be compared to the corresponding angle

in titanium complexes that contain the substituted [(O-

o-C₆H₄)₂S]^2-ligand^36-40(e.g., 100.8(2)° in [(O-2,4-Me₂o-

C₆H₄)₂S]Ti(O-i-Pr)₂³⁹). It appears that the structure is

(34) Graf, D. D.; Schrock, R. R.; Davis, W. M. Unpublished results.

(35) Eshuis, J . J .; Tan, Y. Y.; Meetsma, A.; Teuben, J . H.; Renkema,

J .; Evens, G. G. Organometallics 1992, 11, 362.

Group 4 Complexes with Diamido/ Donor Ligands

Organometallics, Vol. 17, No. 22, 1998 4801

relatively clean according to ¹H NMR spectra. For

example, the room-temperature proton NMR spectra of

{[1a ]ZrMe(NMe₂Ph)}[B(C₆F₅)₄] and [1c]ZrMe(NMe₂-

Ph)}[B(C₆F₅)₄] are presented in parts a and b, respec-

tively, of Figure 7. Their most important features are

the following: (i) four (rather than two) multiplets are

observed for the OCH₂CH₂N backbone protons; (ii)

broad resonances near 6 ppm and a broadened singlet

near 2.7 ppm can be assigned to aromatic protons and

the equivalent methyl groups of coordinated N,N-

dimethylaniline, respectively, in a molecule having a

single plane of symmetry; (iii) a singlet at 0.1 to 0.3 ppm

integrated as three protons can be ascribed to the

zirconium-bound methyl group; (iv) four methyl reso-

nances belonging to the isopropyl groups in the ortho

positions of the aryl rings are consistent with no rotation

about the N-C_ipsobonds. (The two resonances for ortho

methyl groups in the zirconium complex whose spec-

trum is shown in Figure 7a are accidentally coincident

at room temperature, but at 40-60 °C two separate

singlets of area 6 each are found.) One equivalent of free

NMe₂Ph was added to C₆D₅Br solutions of the com-

plexes {[1a ,c]ZrMe(NMe₂Ph)}[B(C₆F₅)₄]. In both cases

resonances due to the aromatic protons of free NMe₂Ph

(300 and 500 MHz) were distinctly separated from those

of bound NMe₂Ph at room temperature. (See, for

example, Figure 7a.) For {[1a]ZrMe(NMe₂Ph)}[B(C₆F₅)₄]

resonances due to the NMe groups in free and bound

dimethylaniline also were well-separated, while in {[1c]-

ZrMe(NMe₂Ph)}[B(C₆F₅)₄] they accidentally overlapped.

In the ¹³C{¹H} NMR spectrum (125 MHz) of {[1c]ZrMe-

(NMe₂Ph)}[B(C₆F₅)₄] in the presence of 1 equiv of free

NMe₂Ph all the resonances belonging to free NMe₂Ph

are sharp and located at the expected chemical shifts,

while those for bound NMe₂Ph are broad. These results

suggest that exchange of free and coordinated NMe₂Ph

in {[1c]ZrMe(NMe₂Ph)}[B(C₆F₅)₄] is slow on the NMR

time scale. They also suggest that the broad aromatic

resonances for coordinated dimethylaniline in Figure 7a

must be attributed to a phenomenon other than dimeth-

ylaniline exchange. Most resonances in the spectrum

of {[1c]HfMe(NMe₂Ph)}[B(C₆F₅)₄] at room temperature

are not sharp, but at -20 °C the spectrum is analogous

to the room-temperature spectrum of {[1c]ZrMe(NMe₂-

Ph)}[B(C₆F₅)₄] described above and shown in Figure 7b,

with the exception that now two separate methine

septet resonances are observed and the resonances

belonging to aromatic protons of PhNMe₂ligand are

better resolved. The phenomenon that gives rise to

broadened resonances in the Zr and Hf cationic com-

plexes is not known. We propose that it involves some

restricted rotation about the N-C_ipsobond in bound

dimethylaniline and/or some restricted movement within

the diamido/donor ligand itself.

F igu r e 6. (a, top) ORTEP drawing of the structure of [2a ]-

ZrMe₂. (b, bottom) Top view (Chem 3D) of the structure of

[2a ]ZrMe₂.

being “held” in approximately a type B conformation by

the restriction placed on the C-S-C angle. The rela-

tively small alkyl groups on Zr and in the ortho positions

of the aryl amido substituents are allowing the molecule

to distort toward a structure of type A, but the combina-

tion of the longer Zr-S bond and the smaller C-S-C

angle does not allow the C-S-C unit to “rotate” readily

(without changing the C-S-C angle significantly)

toward an intermediate in which the angle between the

N(1)-Zr-S and N(2)-Zr-S planes is approximately

180° and the N-Zr-N angle is ∼140°. It should be

noted that the Zr-N-C angles (110.1(2) and 116.4(2)°)

are ∼4° smaller in [2a ]ZrMe₂than in [1a ]ZrMe₂or [1c]-

HfEt₂, as if the amido substituents are pushed toward

the coordination pocket. At this stage we believe this

to be a consequence of the large size of S relative to O,

making the MSC₂N ring effectively larger than an

MOC₂N ring.

Ca tion ic Alk yl Com p lexes of Zr a n d Hf. Addition

of 1 equiv of [PhNMe₂H][B(C₆F₅)₄] in C₆D₅X (X ) Br,

Cl) to [1a ,c]MMe₂(M ) Zr, Hf) at -30 °C followed by

warming to room temperature leads to protonolysis of

one of the methyl groups and formation of cationic

complexes having a single plane of symmetry that

contain coordinated dimethylaniline. The reactions are

All attempts to isolate dimethylaniline adducts in the

solid state have not yet been successful. However,

addition of excess ether and then pentane to freshly

prepared chlorobenzene solutions of {[1a ]MMe(NMe₂-

Ph)}[B(C₆F₅)₄] (M ) Zr, Hf) led to precipitation of {[1a ]-

MMe(ether)}[B(C₆F₅)₄] (M ) Zr, Hf) complexes in high

yield in the form of colorless powders. The complexes

are poorly soluble in benzene, forming biphasic mix-

tures. Their NMR spectra in C₆D₅Br are completely

analogous to those for the dimethylaniline complexes.

(36) Miyatake, T.; Mizunuma, K.; Kakugo, M. Makromol. Chem.-

Macromol. Symp. 1993, 66, 203.

(37) Miyatake, T.; Mizunuma, K.; Seki, Y.; Kakugo, M. Makromol.

Chem.-Rapid Commun. 1989, 10, 349.

(38) Porri, L.; Ripa, A.; Colombo, P.; Miano, E.; Capelli, S.; Meille,

S. V. J . Organomet. Chem. 1996, 514, 213.

(39) Fokken, S.; Spaniol, T. P.; Kang, H.-C.; Massa, W.; Okuda, J .

Organometallics 1996, 15, 5069.

(40) van der Linden, A.; Schaverien, C. J .; Meijboom, N.; Ganter,

C.; Orpen, A. G. J . Am. Chem. Soc. 1995, 117, 3008.

4802 Organometallics, Vol. 17, No. 22, 1998

Aizenberg et al.

F igu r e 7. (a, top) Proton NMR spectrum of {[1a ]Zr(NMe₂Ph)(Me)}[B(C₆F₅)₄] in chlorobenzene-d₅at 22 °C (*, free PhNMe₂)

and at 60 °C in the presence of ∼1 equiv of added PhNMe₂. (b, bottom) Proton NMR spectrum of {[1c]Zr(NMe₂Ph)(Me)]-

[B(C₆F₅)₄] in chlorobenzene-d₅.

The proton spectra contain sharp, well resolved triplet

and quartet resonances due to one coordinated ether

molecule, and the two ethyl groups are equivalent on

the NMR time scale. The resonances for ether that has

been added to a sample of {[1a ]HfMe(NMe₂Ph)}-

[B(C₆F₅)₄] are found in the expected positions in both

the Zr and Hf examples, and free and coordinated ether

do not exchange on the NMR time scale at room

temperature. The coordinated ether methyl resonances

are found at unusually high field (0.21 and 0.17 ppm

for the Zr and Hf complexes, respectively) in each case,

presumably as a consequence of the ether donor being

“sandwiched” between aryl groups at a crowded coor-

dination site. It should be noted that the two types of

ortho methyl groups on the aryl ring in {[1a ]HfMe-

(ether)}⁺complexes are inequivalent and do not inter-

convert readily, consistent with slow rotation about the

N-C_ipsobond on the NMR time scale. The Zr complex

is thermally unstable in solution and in the solid state.

However, at -30 °C in C₆D₅Br it remains virtually

unchanged after 1 week. Decomposition of {[1a ]ZrMe-

(ether)}[B(C₆F₅)₄] in C₆D₅Br at room temperature pro-

duces red needles over the course of several hours; this

complex is insoluble in common solvents and was not

characterized further. (No free ether is detected in the

resulting solution.) In contrast, {[1a ]HfMe(ether)}-

[B(C₆F₅)₄] is thermally stable and can be stored for days

at room temperature both as a solid and in solution. All

data are consistent with a solvent being bound relatively

strongly to hafnium compared to zirconium and with

decomposition of a four-coordinate cationic species

formed upon loss of the solvent donor from the five-

coordinate solvent adduct.

Addition of [PhNMe₂H]]B(C₆F₅)₄] to [2a ]ZrMe₂in

either chlorobenzene-d₅or bromobenzene-d₅in the

presence of THF results in the clean formation of the

THF-stabilized species {[2a ]ZrMe(THF)_x}[B(C₆F₅)₄] and

free dimethylaniline. In this species the zirconium

Group 4 Complexes with Diamido/ Donor Ligands

Organometallics, Vol. 17, No. 22, 1998 4803

Ta ble 3. GP C a n d Yield Da ta for P oly(1-h exen e)

Usin g Zr or Hf P r ecu r sor s Activa ted w ith

methyl resonance is observed at 0.2 ppm and the two

aryl methyl resonances are seen at 2.20 and 1.89 ppm.

On the basis of the formation of only monosolvated

cations containing ether or dimethylaniline in the

oxygen donor system, we believe that most likely x ) 1.

When the same reaction is carried out in chlorobenzene-

d₅in the absence of THF, a relatively clean proton NMR

spectrum is obtained at room temperature, consistent

with formation of {[2a ]ZrMe(PhNMe₂)}[B(C₆F₅)₄], in

which the methyl resonance is found at 0.43 ppm.

Similar results were found for the complexes obtained

upon addition of [PhNMe₂H][B(C₆F₅)₄] to [2c]ZrMe₂.

However, the cations in the sulfur donor system do not

appear to be formed as cleanly and do not appear to be

as stable as those generated in the oxygen donor system.

[P h ₃C][B(C₆F ₅)₄] (A) or [P h NMe₂H][B(C₆F ₅)₄] (B)^a

M_n

M_w/

M_nyield

%

g

no.

complex/activator

(theory) M_n

M_w

1

2

3

4

5

[1a ]ZrMe₂/A

[1b]ZrMe₂/A

[1c]ZrMe₂/A

[1a ]HfMe₂/A

[1b]HfMe₂/A

16 800 11 700 18 000 1.6 100

16 800 9 200 11 100 1.2

16 800 2 500 4 00 1.7

94

96

94

95

16 800 18 100 24 700 1.4

16 800 17 500 20 500 1.2

6

7

8

16 800 17 600 24 000 1.3 100

25 200 25 400 31 700 1.3 100

25 200^b21 200 28 700 1.4

98

9

33 600 21 200 26 800 1.3 100

10

11

12

33 600^c24 600 30 600 1.3 100

33 600^d24 600 31 700 1.3

33 600^e20 400 28 000 1.4

16 800 7 200 9 700 1.3

97

96

99

13 [1c]HfMe₂/A

14

15

At this stage we have not been able to observe

monomethyl Zr and Hf cations generated with [Ph₃C]-

[B(C₆F₅)₄] in chlorobenzene or bromobenzene, perhaps

as a consequence of the poor donating ability of bro-

mobenzene or chlorobenzene or the unexpected com-

plexity of the NMR spectra. (Cationic methyl species

generated in chlorobenzene with [Ph₃C][B(C₆F₅)₄] are

effective initiators for 1-hexene polymerization, as

described below; thus, we assume that they are formed

and do not decompose readily to a significant extent.)

We also have not been able to observe monomethyl

cations, even dimethylaniline adducts, that contain Ti.

25 200 7 500 10 200 1.4 100

33 600 6 000 8 300 1.4 100

16 [1d ]HfMe₂/A

17 [1a ]ZrMe₂/B

18 [1b]HfMe₂/B

19 [2a ]ZrMe₂/A^a

20 (3 h)^a

16 800 3 800 6 300 1.6

16 800 9 000 13 000 1.4

16 800 23 000 32 600 1.4

16 800 16 000 18 400 1.2

16 800 15 900 18 400 1.2 100

25 200 20 900 25 000 1.2

77

70

21

86

21

a

22 (3 h)^a

25 200 19 100 23 000 1.2 100

23 [2a ]ZrMe₂/B

24

25 [2a ]ZrMe₂/B (5 h)

26

27

28

29

16 800^f16 900 23 200 1.4

16 800^f16 900 22 600 1.3

21 880 20 800 24 500 1.2

25 200^f16 900 22 500 1.3

33 600^f19 400 28 700 1.5

33 600^f20 700 30 900 1.5

50 400^f28 400 36 100 1.3

16 800^f4 300 5 500 1.3

16 800^f6 100 8 200 1.4

94

96

94

88

84

87

77

75

85

82

80

P olym er iza tion of 1-Hexen e. Addition of 25 equiv

of 1-hexene to an NMR mixture of [1a ]TiR₂(R ) Me,

CH₂Ph) and [PhNMe₂H][B(C₆F₅)₄] in C₆D₅Br results in

only short-term polymerization activity. Some monomer

is not consumed. New olefinic resonances appear that

are centered around δ 5.3 ppm, which suggests that

â-hydride elimination accompanies decomposition of

“{[1a ]TiR}⁺”. Use of [Ph₃C][B(C₆F₅)₄] as an initiator led

to immediate decomposition and formation of dark oily

precipitates.

30 [2c]ZrMe₂/B

31

32 [2c]Zr(CH₂CHMe₂)₂/B 16 800^f4 200 6 800 1.6

33 [2c]Zr(CH₂CHMe₂)₂/B 16 800^f4 300 6 400 1.5

a

Conditions: 0.040 mmol of the metal complex and activator,

0 °C, 10 mL of chlorobenzene, 1 h, unless otherwise noted.

b

Conditions: 0.025 mmol of the metal complex, 0.022 mmol of the

activator, 6.60 mmol of the monomer. ^cConditions: 0.017 mmol

of the metal complex, 0.015 mmol of the activator, 6.00 mmol of

d

the monomer. Conditions 0.012 mmol of the metal complex, 0.008

mmol of the activator, 3.20 mmol of the monomer. ^eThe monomer

was added in two 200 equiv portions with 25 min interval, total

experiment time 1 h 15 min. ^fConditions: 50 µmol of the catalyst,

50 µmol of the activator, 13 mL of chlorobenzene, 3 h at 0 °C.

^gDetermined by GPC on-line light scattering (Wyatt Technology).

In contrast, complexes [1a -d ]MMe₂(M ) Zr, Hf),

when activated with [Ph₃C][B(C₆F₅)₄] in chlorobenzene,

were found to be efficient initiators for polymerization

of 1-hexene. The temperature was kept at 0 °C in order

to minimize decomposition that would result from a

significant exotherm. Data can be found in Table 3. The

low value of dn/dc for poly(1-hexene) in dichloromethane

limited the accuracy of the molecular weight data (see

Experimental Section), so only broad trends will be

discussed. We will first discuss systems that contain

oxygen donor ligands and will assume complete activa-

tion and a quantitative yield of initiator before 1-hexene

is added. To provide a basis for comparison most of the

comments will be based on a theory concerning base

lability. (In reactions initiated with [Ph₃C][B(C₆F₅)₄] it

will be assumed that the C₆D₅Cl solvent is functioning

as the base, as that it is likely to be more labile than

dimethylaniline.) We will not discuss polydispersity, as

the polydispersity values are not accurate beyond two

significant figures. Several poly(1-hexene) samples

were examined by ¹³C NMR and found to be atactic.

to Hf versus Zr. The molecular weight is also higher

than theory based on equivalents of monomers added,

which could be attributed to a significantly stronger

binding of base to the initiating species versus the

propagating species.

(ii) Ar yl Gr ou p . For Zr catalysts increasing the

steric bulk of the aryl groups (runs 1-3) leads to

formation of shorter polymers, although polymer yields

are uniformly high. Therefore, after termination of a

given chain, a new chain can be initiated without

significant catalyst decomposition; i.e., the process is a

chain transfer. These results are consistent with a

greater lability of the base as the size of the substituents

on the aryl ring increase, with consequently a more

competitive chain transfer process. These results sug-

gest that at least for 1-hexene under the conditions we

have employed, the olefin does not add rapidly enough

to the four-coordinate cation to compete with chain

transfer.

(i) In itia tor . Run 17 (compared with run 1) provides

evidence in the form of a relatively low yield of poly(1-

hexene) (in 1 h reaction time) that {[1a ]ZrMe(Ph-

NMe₂)}⁺is a slower catalyst than {[1a ]ZrMe(PhCl)}⁺.

Runs 5 and 18 confirm that trend and are consistent

with the expected stronger binding of dimethylaniline

A similar trend is observed for hafnium (runs 4, 5,

and 13). Polymer yields are again high, but polymers

shorter than expected are found only in run 13. Mo-

4804 Organometallics, Vol. 17, No. 22, 1998

Aizenberg et al.

lecular weights are again somewhat higher than ex-

pected in runs 4 and 5. We propose that hafnium yields

higher molecular weight poly(1-hexene) as a conse-

quence of the base being bound more strongly, thereby

preventing â-hydride elimination in a four-coordinate

base-free cation. A large k_p/k_iratio would also contrib-

ute to a larger than expected molecular weight that

would counteract the consequences of â-hydride elimi-

nation. Although the rate of polymerization would be

expected to be slower for hafnium than for zirconium,

run 13 (unlike run 18) evidently is complete in 1 h at 0

°C and the yield of poly(1-hexene) is high, presumably

since the “base” in this case is a more labile chloroben-

zene.

metal, while dimethylaniline, and especially chloroben-

zene, are relatively labile.

Similar trends were observed for catalytic reactions

that employ the sulfur donor ligands. A comparison of

runs 1, 19, and 20 suggest that the [2a ]ZrMe₂/A initiator

is slower than the [1a ]ZrMe₂/A initiator but that â-hy-

dride elimination does not seem to be as rapid in the

sulfur donor ligand system. In fact, run 20 is reminis-

cent of run 5 or 6. Use of the [2a ]ZrMe₂/B as the

initiator again leads to shorter poly(1-hexene) chains,

as â-hydride elimination again becomes more competi-

tive with chain growth, and approximately the same

limitation to polymer chain length (20 000 to 25 000).

The results using [2c]Zr(CH₂CHMe₂)₂/B as an initiator

(runs 32 and 33) are the same as the results of runs 30

and 31, which suggests that higher alkyl complexes can

be activated successfully. So far the data are not

indicative of any dramatic instability of the cations

formed in the sulfur system. However, qualitative

evidence suggests that the rates of polymerization are

slightly slower in the sulfur donor system than in the

oxygen donor system.

Use of [1d ]HfMe₂leads to the formation of a relatively

low molecular weight polymer in 77% yield (entry 16).

In an analogous polymerization using 50 equiv of

1-hexene in C₆D₅Br at 0 °C and followed by proton

NMR, olefinic proton resonances were observed as a

broad singlet centered at δ 5.4 ppm. If we assume that

these olefinic resonances arise from vinylidene end

groups in the poly(1-hexene), then we can conclude that

â-hydride elimination is minimized when 2,6-disubsti-

tuted aryl groups are employed. One ortho substituent

can block only one side of a structure of type A, leaving

the other side relatively accessible for â-hydride elimi-

nation.

Discu ssion

Although group 4 amido complexes that contain a bis-

(trimethylsilyl)amido ligand have been known for some

time,^41-43group 4 amido ligands that do not contain a

silyl substituent and that are multidentate are rela-

tively recent additions to the area of early-transition-

metal amido chemistry. Although trimethylsilyl sub-

stituents add a significant amount of steric bulk to an

amido ligand and also create an amido ligand that has

suitable electronic characteristics for binding to an early

transition metal, silyl substituents are prone to hydroly-

sis or to CH activation.⁴³Monodentate amido ligands

also can be displaced from the coordination sphere,

either as an anion (rarely) or upon protonation or

alkylation. The combination of one aryl and one alkyl

(tert-butyl) substituent in an amido ligand has recently

been employed to considerable advantage in early-

transition-metal chemistry, especially the chemistry of

molybdenum.^44-49Therefore, the tridentate diamido/

donor ligands of the type reported here are attractive

candidates for a variety of chemistry involving relatively

reactive metal centers. They are related to diamido/

donor ligands of the type [(t-BuN-o-C₆H₄)₂O]^2-reported

recently,^11,12in that the amido nitrogens are bound to

both an alkyl and to an aryl substituent. However, the

tert-butyl groups in the [(t-BuN-o-C₆H₄)₂O]^2-ligands

probably create a significantly greater degree of steric

(iii) Am ou n t of Mon om er . We chose to examine

the result of varying the amount of monomer with [1b]-

HfMe₂as the catalyst precursor and [Ph₃C][B(C₆F₅)₄]

as the activator, since this combination proved to be

relatively well-behaved in the experiments at the scale

of 200 equiv of 1-hexene (entries 5 and 6). The results

obtained are summarized in entries 5-12. Under

standard conditions (see Experimental Section), increas-

ing the amount of 1-hexene from 200 (entries 5 and 6)

to 300 equiv (entry 7) produced a higher molecular

weight in proportion to the amount of additonal mono-

mer employed. Lowering the catalyst concentration

from 0.040 to 0.022 mmol/10 mL produced a similar

result (entry 8). However, increasing the amount of the

monomer to 400 equiv in experiments where catalyst

concentrations of 0.040, 0.015, and 0.008 mmol/10 mL

were employed (entries 9-11), or adding the monomer

in two equal portions (entry 12), did not lead to a

proportional increase in the molecular weight. An

analogous trend is observed for complexes that contain

[1c]^2-, although at much lower molecular weight levels

(entries 13-15), as noted earlier. We conclude that

there is a limit to the polymer chain length as a

consequence of â-hydride elimination under the condi-

tions employed.

(41) Andersen, R. A. Inorg. Chem. 1979, 18, 2928.

(42) Andersen, R. A. J . Organomet. Chem. 1980, 192, 189.

(43) Planalp, R. P.; Andersen, R. A.; Zalkin, A. Organometallics

1983, 2, 16.

(44) Stokes, S. L.; Davis, W. M.; Odom, A. L.; Cummins, C. C.

Organometallics 1996, 15, 4521.

(45) Peters, J . C.; Odom, A. L.; Cummins, C. C. Chem. Commun.

1997, 1995.

(46) Laplaza, C. E.; J ohnson, M. J . A.; Peters, J . C.; Odom, A. L.;

Kim, E.; Cummins, C. C.; George, G. N.; Pickering, I. J . J . Am. Chem.

Soc. 1996, 118, 8623.

(47) J ohnson, M. J . A.; Odom, A. L.; Cummins, C. C. Chem.

Commun. 1997, 1523.

(48) J ohnson, M. J . A.; Lee, P. M.; Odom, A. L.; Davis, W. M.;

Cummins, C. C. Angew. Chem., Int. Ed. Engl. 1997, 36, 87.

(49) Fickes, M. G.; Odom, A. L.; Cummins, C. C. Chem. Commun.

1997, 1993.

The proposal that a five-coordinate cation must lose

a base in order to react with 1-hexene is supported by

the finding that neither {[1c]ZrMe(ether)}[B(C₆F₅)₄] nor

{[1c]HfMe(ether)}[B(C₆F₅)₄] will initiate the polymeri-

zation of 1-hexene in C₆D₅Br at room temperature; no

change is observed after 1 h. Moreover, neither com-

pound initiates the polymerization of ethylene at 1 atm

and 25 °C. Therefore, even ethylene cannot attack the

five-coordinate base adduct to produce a pseudooctahe-

dral intermediate. These results also make it clear that

diethyl ether essentially does not dissociate from the

Group 4 Complexes with Diamido/ Donor Ligands

Organometallics, Vol. 17, No. 22, 1998 4805

Ch a r t 1

prediction is consistent with the fact that the structures

of dialkyl complexes found here can be rationalized on

the basis of steric arguments alone.

The low electron count for dialkyl complexes would

seem to contribute to their instability with respect to

â-hydride elimination or abstraction reactions. Some

hafnocene monoalkyl⁵⁰and dialkyl complexes⁵¹have

been isolated where the alkyl contains a â-proton, and

stable Zr or Hf diamido dialkyl complexes have been

known for some time.^41-43Therefore, the stability of

some of the neutral Zr and Hf species reported here is

not a complete surprise. It is beginning to appear that

diamido group 4 dialkyl complexes can be relatively

stable toward â-elimination, with the stability increas-

ing significantly in the order Ti < Zr < Hf. Neverthe-

less, the relatively high stability of zirconium and

hafnium cationic alkyl intermediates in polymerization

reactions reported here was surprising to us, especially

in view of the relatively low electron count (assuming

one M-N π bond). The instability of the analogous

cationic titanium alkyl complexes was closer to the

expected behavior. It should be pointed out, however,

that the mode of decomposition of no neutral or cationic

alkyl complex reported here has been elucidated, and

therefore we should not presume that the mechanism

of decomposition is uniform or that the diamido/donor

ligand is never involved.

congestion in complexes that contain that ligand than

in the complexes reported here, and the phenylene

backbone is not likely to be as flexible as in the ligands

reported here. As a consequence of the relatively

flexible backbone in the ligands reported here, the donor

atom could dissociate from the metal and, after rotation

about the C-donor bonds, actually turn away from the

metal. However, we believe that dissociation of an O

or S donor, especially in cationic complexes, is not taking

place to any significant degree, at least for Zr and Hf,

where metal-ligand bond strengths are relatively high

compared to Ti-donor bond strengths. However, we

cannot be certain that a donor never dissociates from

the metal, especially in some neutral complexes, and

especially in complexes in which the bond strength is

relatively low (Ti).

All evidence at this time suggests that 12e monosol-

vated monoalkyl complexes (for convenience of type A

shown in eq 9) must lose a base to yield a 10e four-

coordinate cation. This four-coordinate cation most

An important issue in amido complexes in general is

the degree of π bonding between the amido nitrogen and

the metal. In the diamido/donor ligands reported here

the aryl substituent is generally turned 90° from where

it should be for maximum conjugation of the aryl π

system with the nitrogen lone pair. Therefore, the lone

pair on nitrogen is likely to be involved primarily in π

bonding to the metal. However, the amount of π

bonding in the two structures (type A and type B)

should be roughly the same. The major π-bonding

interaction in the structure of type A is that between

the unsymmetric combination of p orbitals on the two

amido nitrogens and the d_xyorbital (in the coordinate

system shown in Chart 1), while in the structure of type

B it is that between the symmetric combination of p

likely has a core geometry that is as close as possible to

a pseudotetrahedron; i.e., the axial amido nitrogens in

the type A structure most likely would move toward the

position occupied by the base as the base leaves the

coordination sphere (eq 9). This movement is the same

as that which leads to a conversion of a type A structure

to a type B structure. In a five-coordinate structure of

type B, in which the base is in an axial position,

formation of the four-coordinate pseudo-tetrahedral

species would require (after loss of S) only some

relatively small movement of the R group. A pseudo-

tetrahedral structure is a sensible and accessible ge-

ometry for a d⁰complex and contrasts markedly with

the distorted-trigonal geometry of a typical cationic

metallocene monoalkyl complex that is believed to be

formed upon activating group 4 metallocene dialkyl

complexes. Formation of inactive monosolvated [Cp₂-

MR(S)]⁺species (M ) Zr, Hf) and required loss of

solvent in order to generate catalytically active species

2

orbitals on the two amido nitrogens and the d_{x -y}

orbital. (The N-M-N system is viewed from a position

trans to the donor in both type A and type B structures.)

In the structure of type B the donor in the axial position

cannot π bond to a significant degree with the metal.

In contrast, an sp²-hybridized donor in a structure of

type A could be involved in π bonding to the metal to

some degree through the d_xzorbital. However, planarity

or near-planarity of the donor in a structure of type A

alone is not evidence for a significant degree of π

bonding, since the electron density could be centered

largely on the donor in a p orbital, especially in the case

of a relatively electronegative oxygen atom donor. If

only one significant π bond is invoked in an electron-

counting scheme, all neutral dialkyl complexes or cat-

ionic monoalkyl complexes (that contain an additional

external σ donor) of either type A or type B are

12-electron species. Therefore, there does not seem to

be any significant electronic (π bonding) advantage to

a structure of type A compared to one of type B or to

any intermediate version between A and B. This

(50) Guo, Z.; Swenson, D. C.; J ordan, R. F. Organometallics 1994,

13, 1424.

(51) Erker, G.; Schlund, R.; Kru¨ger, C. Organometallics 1989, 8,

2349.

4806 Organometallics, Vol. 17, No. 22, 1998

Aizenberg et al.

have been amply demonstrated.^50,52,53If we assume for

the moment that decomposition of intermediate four-

coordinate cations is of the same type for all three

metals and that a base protects against that decomposi-

tion by forming a five-coordinate adduct, then all of the

behavior we have seen so far in these systems can be

explained. We are ready to modify this viewpoint as

more data become available.

were found to behave best in the sense that the

molecular weight of poly(1-hexene) is approximately

equal to the amount of monomer employed, and the

polydispersity is relatively low. â-Hydride elimination

becomes an important factor in polymers that contain

more than 300 equiv of monomer and limits the molec-

ular weight to ∼25 000. Dialkylzirconium complexes

that contain a [(ArylNHCH₂CH₂)₂S]^2-ligand, as well

as cationic complexes, appear to be less stable than

analogous [(ArylNHCH₂CH₂)₂O]^2-complexes. Never-

theless, cationic complexes that contain the sulfur donor

ligand behave like complexes that contain the oxygen

donor ligand in the studies conducted so far.

The results reported here should be contrasted with

those of McConville and co-workers, where Ti catalysts

supported by (ArylNCH₂CH₂CH₂NAryl)^2-ligands (e.g.,

Ar ) 2,6-i-Pr₂C₆H₃) are found to yield poly(1-hexene)

in a living manner in the presence of excess 1-hexene.^6-9

As the 1-hexene is depleted, a decomposition step

involving the anion competes with polymerization and

destroys the catalyst.⁹The reason the titanium com-

plexes prepared and studied here are not relatively well-

behaved polymerization catalysts for 1-hexene, while

Ti[ArylNCH₂CH₂CH₂NAryl]R₂compounds yield sys-

tems that are living for the polymerization of 1-hexene

in the presence of excess 1-hexene, is not known.

Interestingly, there has been no report of [ArylNCH₂CH₂-

CH₂NAryl]^2-zirconium or hafnium catalysts behaving

in a living manner as 1-hexene polymerization catalysts,

and although relatively rigid diamido/donor systems

based on 2,6-disubstituted pyridine have been prepared

for both Ti and Zr,^21,23no polymerization activity was

reported. It should be noted that the number of group

4 diamido or diamido/donor systems in which no N-Si

bond is present is still small and they have not yet been

explored in detail as polymerization catalysts, so it is

perhaps too early to attempt to rationalize results

obtained in different systems at this stage.

Exp er im en ta l Section

Gen er a l P r oced u r es. All experiments were performed

under a nitrogen atmosphere in a Vacuum Atmospheres

drybox or by standard Schlenk techniques unless otherwise

specified. Tetrahydrofuran and diethyl ether were sparged

with nitrogen and passed through two columns of activated

alumina. Toluene was distilled from sodium/benzophenone.

Pentane was sparged with nitrogen, then passed through one

column of activated alumina, and then passed through another

of activated Q5. Chlorobenzene was distilled from CaH₂under

nitrogen. All solvents were stored in the drybox over 4 Å

molecular sieves. 1-Hexene was refluxed over sodium for 4

days, distilled, and stored over 4 Å sieves.

NMR data were recorded at 500 or 300 MHz (¹H), 75 or 125

MHz (¹³C), and 470 MHz (¹⁹F). Chemical shifts are listed in

parts per million downfield from tetramethylsilane or CFCl₃

(for ¹⁹F), unless specified otherwise. A standard variable-

temperature unit was used to control the probe temperature

in variablem temperature runs, and temperatures are consid-

ered accurate to (1 °C. NMR solvents were sparged with

nitrogen and stored over 4 Å molecular sieves. Elemental

analyses were performed by Microlytics of Deerfield, MA.

Chloride analysis was performed by Schwarzkopf Microana-

lytical Laboratory, Inc., of Woodside, NY. X-ray data were

collected on a Siemens SMART/CCD diffractometer.

All gel permeation chromatography (GPC) experiments were

carried out using two J ordi-Gel DVB Mixed Bed columns in

series, a Wyatt Mini Dawn light scattering detector operating

at 690 nm, and a Knauer refractometer. Solutions of samples

dissolved in CH₂Cl₂were filtered through a Millex-SR 0.5 µm

filter. All GPC data were analyzed using Astrette 1.2 (Wyatt

Technology). Values of dn/dc were obtained under the as-

sumption that all sample eluted from the column and were

averaged for various runs. The low value of dn/dc (the change

in refractive index versus concentration) for poly(1-hexene) in

dichloromethane limited the accuracy of the molecular weight

data obtained via light scattering.

Zr(NMe₂)₄, Hf(NMe₂)₄, and Ti(CH₂Ph)₄were prepared ac-

cording to the literature procedure. [Ph₃C][B(C₆F₅)₄] and

[PhNMe₂H][B(C₆F₅)₄] were provided by Exxon Chemicals. 2,2′-

Thiodiethanol, 2,6-Dimethylaniline, 2,6-diethylaniline, 2,6-

diisopropylaniline, and 2-tert-butylaniline were purchased

from Aldrich.

The studies reported here should be regarded as only

an outline. The nature of the base, the size of the alkyl

in both the five-coordinate adducts and four-coordinate

cations that contain the growing polymer chain, the

reactivity of the monomer, the degree of base labilization

in five-coordinate cations, and the protection against

â-hydride elimination afforded by the ortho substituents

or structures of type B are all important features of

polymerization reactions that remain to be fully ex-

plored.

Con clu sion s

Titanium, zirconium, and hafnium dialkyl complexes

that contain diamido/donor ligands of the general

formula (ArylNHCH₂CH₂)₂O can be prepared readily.

Dimethyl complexes can be activated by protonation or

oxidative cleavage of the alkyl to yield cationic species,

many of which can be observed for Zr and Hf, especially

when an oxygen donor such as diethyl ether is bound

to the metal. If the base is sufficiently labile (dimethyl-

aniline or chlorobenzene), then 1-hexene can be polym-

erized relatively efficiently. The vast majority of the

results can be explained in terms of a required loss of a

coordinated base (solvent) from a five-coordinate base

adduct and subsequent olefin attack on a four-coordi-

nate cationic intermediate. â-Hydride elimination is

proposed to occur in a four-coordinate cationic interme-

diate. Hafnium systems (activated with [Ph₃C][B(C₆F₅)₄])

P r ep a r a tion of Com p ou n d s. H₂[1a ]. Solid (TsOCH₂-

CH₂)₂O (6.47 g, 15.6 mmol) was added to a chilled (-30 °C)

solution of LiNH(2,6-Me₂C₆H₃) (4.0 g, 31.5 mmol) in THF (40

mL). The reaction mixture was stirred at room temperature

for 72 h, and all volatile components were then removed in

vacuo. The residue was extracted with pentane, and the

extract was concentrated in vacuo until white crystals ap-

peared. Recrystallization at -30 °C from pentane afforded

colorless crystals (3.0 g, 9.6 mmol, 62%): ¹H NMR δ 2.24 (s,

12, ArMe), 2.95 (dt, 4, CH₂N), 3.15 (t, 4, OCH₂), 3.50 (t, 2, NH),

[6.88 (dd, 2), 6.99 (d, 2] (H_aryl); ¹³C{¹H} NMR δ 18.9 (ArMe),

48.6 (CH₂N), 70.8 (OCH₂), [122.7, 129.6, 130.2, 146.8] (C_aryl).

(52) J ordan, R. F. Adv. Organomet. Chem. 1991, 32, 325.

(53) Wu, Z.; J ordan, R. F. J . Am. Chem. Soc. 1995, 117, 5867.

Group 4 Complexes with Diamido/ Donor Ligands

Organometallics, Vol. 17, No. 22, 1998 4807

Anal. Calcd for C₂₀H₂₈N₂O: C, 76.88; H, 9.03; N, 8.97.

Found: C, 76.93; H, 8.96; N, 8.83.

(s), 0.08 (s)] (18, SiMe₃), [1.41 (s), 1.42 (s)] (18, CMe₃), [3.07

(m, 2), 3.31 (overlapped m, 6)] (OCH₂CH₂N), [6.90 (m, 2), 7.02

(m, 2), 7.10 (m, 2), 7.45 (dt, 2)] (H_aryl); ¹³C{¹H} NMR δ 0.78

(SiMe₃), 33.06 (CMe₃), 36.43 (CMe₃), 51.91 (CH₂N), 69.95

(OCH₂), [125.29, 125.91, 129.96, 133.90, 146.65, 148.16] (C_aryl).

A solution of 10.6 g (20.7 mmol) of [(SiMe₃)(2-t-BuC₆H₄)-

NCH₂CH₂]₂O in 100 mL of ether was stirred at 40 °C with

200 mL of 3 N HCl for 13 h. The reaction mixture was treated

with 3 N KOH until basic. The organic phase was separated,

and the aqueous phase was extracted with an additional 100

mL of ether. The extracts were combined, washed with water

until neutral, and dried over anhydrous Na₂SO₄. Removal of

the ether in vacuo yielded 7.33 g (96%) of H₂[1d ] as a rose-

colored oil which solidified after 2 days. An analytically pure

sample in the form of almost colorless crystals was obtained

by recrystallization from pentane at -30 °C: ¹H NMR (CDCl₃)

δ 1.45 (s, 18, CMe₃), 3.41 (t, 4, CH₂N), 3.85 (t, 4, OCH₂), 4.4

(vbr s, 2, NH), [6.75 (m, 4), 7.18 (td, 2), 7.29 (dd, 2)] (H_aryl);

¹³C{¹H} NMR (CDCl₃) δ 30.08 (CMe₃), 34.36 (CMe₃), 44.31

(CH₂N), 69.61 (OCH₂), [112.25, 117.53, 126.46, 127.30, 134.02,

146.57] (C_aryl). Anal. Calcd for C₂₄H₃₆N₂O: C, 78.21; H, 9.85;

N, 7.60. Found: C, 78.24; H, 10.08; N, 7.54.

[1a ]TiCl₂. A hexane solution of butyllithium (0.8 mL of 2.5

M, 2.0 mmol) was added to a solution of 312 mg (1.0 mmol) of

H₂[1a ] in 10 mL of ether at -30 °C. The mixture was warmed

to room temperature. After 40 min the reaction mixture was

again cooled to -30 °C and a suspension of TiCl₄(THF)₂(334

mg, 1.0 mmol) was added to the stirred mixture. The color of

the mixture changed immediately to deep red, and a red-brown

precipitate was formed. The mixture was stirred for 1 h at

room temperature, and solvents were then removed in vacuo.

The dry residue was extracted with methylene chloride, the

extract was filtered, and the solvent was removed in vacuo to

yield brick red [1a ]TiCl₂; yield 244 mg (56.9%): ¹H NMR δ

2.42 (s, 12, ArMe), 3.27 (t, 4, CH₂N), 3.65 (t, 4, OCH₂), 6.94-

7.20 (m, 6, H_aryl); ¹³C{¹H} NMR δ 19.18 (ArMe), 59.35 (CH₂N),

72.33 (OCH₂), [126.79, 129.06, 131.33, 155.85] (C_aryl). Anal.

Calcd for C₂₀H₂₆Cl₂N₂OTi: C, 55.97; H, 6.11; N, 6.53. Found:

C, 55.51; H, 6.25; N, 6.23.

[1a ]TiMe₂. A 3.0 M solution of MeMgCl in THF (0.33 mL)

was added to a cold (-30 °C) stirred suspension of 200 mg

(0.47 mmol) of [1a ]TiCl₂in 6 mL of ether. The color of the

mixture changed immediately to brown. The mixture was

stirred for 35 min at room temperature, and the solvent was

removed in vacuo. The residue was extracted with 15 mL of

pentane. The extract was filtered and taken to dryness in

vacuo to yield 95 mg (52.5%) of pure [1a ]TiMe₂in a form of a

yellow powder: ¹H NMR δ 0.90 (s, 6, TiMe), [2.508 (s, 6), 2.509

(s, 6H)] (ArMe), 3.36 (m, 8, OCH₂CH₂N), [7.01 (m, 2), 7.14 (m,

4)] (H_aryl); ¹³C{¹H} NMR δ 19.17 (ArMe), 57.03 (CH₂N), 60.75

(TiMe), 70.67 (OCH₂), [125.39, 128.99, 131.76, 152.92] (C_aryl).

[1a ]Ti(CH₂P h )₂. (a ) F r om Ti(CH₂P h )₄. To a red solution

of 824 mg (2.0 mmol) of TiBz₄in 10 mL of pentane was added

a solution of 624 mg (2.0 mmol) in 5 mL of pentane. The

reaction mixture was stirred for 40 h, and all volatile compo-

nents of the mixture were then removed in vacuo, leaving an

oily dark red residue. This residue was washed with 7 mL of

pentane and then 10 mL of ether. The remaining red-orange

precipitate was isolated and dried in vacuo; yield 165 mg

(15%). Red needles suitable for an X-ray structure determi-

nation were grown from a concentrated benzene-d₆solution

at room temperature: ¹H NMR δ 2.43 (br s, 12, ArMe), 2.66

(br s, 4, TiCH₂Ph), 2.99 (t, 4, CH₂N), 3.31 (t, 4, OCH₂), [6.56

(br s, 4), 6.81 (t, 2), 7.02 (m), 7.12 (d), 10] (H_aryl); ¹³C{¹H} NMR

δ 19.71 (ArMe), 58.07 (CH₂N), 70.65 (OCH₂), [121.75, 125.72,

126.11, 127.81, 129.35, 133.49, 150.01, 154.96] (C_aryl). The

resonances due to the benzylic carbons that are not observed

at room temperature appear in the spectrum measured at 65

°C (125 MHz) at 91.7 ppm (v br). Anal. Calcd for C₃₄H₄₀N₂-

OTi: C, 75.54; H, 7.46; N, 5.18. Found: C, 73.90; H, 7.05; N,

4.80.

H₂[1b]. LiNH(2,6-Et₂C₆H₃) (3.44 g, 22.2 mmol) was added

in two portions to a chilled (-30 °C) solution of 4.59 g (11.1

mmol) of (TsOCH₂CH₂)₂O in 80 mL of THF. The resulting

yellow solution was warmed slowly to room temperature and

was stirred for 30 h. The solvent was removed in vacuo, and

the resulting viscous solid was extracted with two 30 mL

portions of pentane. Removal of the pentane from the filtered

extract in vacuo yielded 3.4 g of a slightly yellow oil. Most of

the 1-(2,6-diethylphenyl)morpholine side product of the reac-

tion was distilled away from the product at 40-50 °C/200

mTorr to yield 2.5 g (52%) of an orange oil which was shown

by NMR to be the 90% (by mass) pure H₂[1b]: ¹H NMR δ 1.21

(d, 12, ArCH₂CH₃), 2.69 (q, 8, ArCH₂CH₃), 3.00 (t, 4, OCH₂),

3.27 (t, 4, CH₂N), 3.60 (br t, 2, NH), 6.98-7.08 (m, 6, H_aryl);

¹³C{¹H} NMR δ 15.14 (CH₂CH₃), 24.76 (CH₂CH₃), 49.93

(CH₂N), 70.67 (OCH₂), [123.25, 127.13, 136.80, 145.29] (C_aryl).

It was used without further purification.

H₂[1c]. Solid (TsOCH₂CH₂)₂O (5 g, 12.0 mmol) was added

to a chilled (-30 °C) solution of LiNH(2,6-i-Pr₂C₆H₃) (4.53 g,

24.8 mmol) in THF (30 mL). The reaction mixture was stirred

at room temperature for 24 h, and all volatile components were

removed in vacuo. The residue was extracted with pentane,

and the extract was filtered. Removal of all volatile compo-

nents in vacuo gave an orange oil (4.2 g, 82%), which was used

without further purification. Upon standing, the oil would

crystallize in some cases. An analytically pure sample was

obtained by recrystallization from a concentrated pentane

solution at -30 °C: ¹H NMR δ 1.06 (d, 24, CHMe₂), 3.07 (q, 4,

CH₂N), 3.35 (t, 4, OCH₂), 3.48 (heptet, 4, CHMe₂), 3.60 (t, 2,

1

NH), 7.18-7.14 (m, 6, H_aryl); H NMR (CDCl₃) δ 1.27 (d, 24,

CHMe₂), 3.14 (br t, 4, CH₂N), 3.36 (heptet, 4, CHMe₂), 3.57

(br, 2, NH), 3.70 (t, 4, OCH₂), 7.05-7.18 (m, 6, H_aryl); ¹³C{¹H}

NMR (CDCl₃) δ 24.3 (CHMe₂), 27.6 (CHMe₂), 51.3 (CH₂N), 70.7

(OCH₂), [123.6, 123.7, 142.5, 142.9] (C_aryl). Anal. Calcd for

C

₂₈H₄₄N₂O: C, 79.19; H, 10.44; N, 6.60. Found: C, 79.14; H,

10.23; N, 6.30.

H₂[1d ]. To a solution of 8.96 g (60 mmol) of 2-t-BuC₆H₄-

NH₂in 150 mL of THF precooled to -30 °C was added in two

portions over 15 min 24 mL of 2.5 M solution of BuLi in

hexanes (60 mmol). The reaction mixture was stirred for 40

min more after which to the resulting yellowish solution of

LiNH(2-t-BuC₆H₄) was quickly added at room temperature

6.65 g (60 mmol) of Me₃SiCl. The color was immediately

discharged, and a white precipitate formed. The mixture was

stirred for an additional 40 min, after which the liquid phase

was filtered from the precipitate. The solvent was removed

in vacuo, and the resulting oil was reextracted into pentane

to completely remove the remaining amount of LiCl. The

extract was filtered and evacuated to yield 13.25 g (99%) of

NH(SiMe₃)(2-t-BuC₆H₄) as a yellowish oil: ¹H NMR (CDCl₃)

δ 0.32 (s, 9, SiMe₃), 1.44 (s, 9, CMe₃), 3.70 (br s, 1, NH), [6.73

(td, 1), 6.78 (dd, 1), 7.09, (td, 1), 7.27 (dd, 1)] (H_aryl).

To a precooled (-30 °C) solution of 13.25 g (60 mmol) of NH-

(SiMe₃)(2-t-BuC₆H₄) in 100 mL of THF was added in three

portions over a period of 1 h 24 mL of a 2.5 M solution of

butyllithium in hexane (60 mmol). After 2 h 100 mL of THF

was added, the mixture was cooled again to -30 °C, and 12.43

g (30 mmol) of solid (TsOCH₂CH₂)₂O was added over a period

of 10 min. The reaction mixture was stirred for 12 h at room

temperature, and the greenish solution was separated from a

white precipitate by filtration. The filtrate was concentrated

in vacuo until it became viscous. The oil was extracted with

two portions of pentane (70 mL each), and the extracts were

taken to dryness in vacuo to yield 11.0 g of an almost colorless

oil that was shown by NMR to be a mixture of NHSiMe₃(2-t-

BuC₆H₄) and [(SiMe₃)(2-t-BuC₆H₄)NCH₂CH₂]₂O in about a 1:2

molar ratio. NHSiMe₃(2-t-BuC₆H₄) (2.8 g) was removed by

distillation in vacuo (46 °C/170 mTorr) to leave 7.8 g (50.7%)

of [(SiMe₃)(2-t-BuC₆H₄)NCH₂CH₂]₂O: ¹H NMR (CDCl₃) δ [0.07

4808 Organometallics, Vol. 17, No. 22, 1998

Aizenberg et al.

(b) F r om [1a ]TiCl₂a n d Ben zyl Gr ign a r d Rea gen t. A

1.0 M solution of PhCH₂MgCl in ether (0.47 mL) was added

to a stirred suspension of 100 mg (0.23 mmol) of [1a ]TiCl₂in

3 mL of ether. The suspension quickly turned dark red. The

mixture was stirred for 25 min at room temperature, and the

solvent was removed in vacuo. The residue was washed with

two portions of pentane (5 mL each) and extracted with

methylene chloride. The extract was filtered, and the solvent

was removed in vacuo. The dark red solid was washed with

pentane; yield 72 mg of crude product. The crude product was

washed with minimal cold ether to give the pure product as a

red-orange powder; yield 44 mg (35%).

[1a ]Zr (NMe₂)₂. A solution of Zr(NMe₂)₄(1.07 g, 4.0 mmol)

in pentane (4 mL) was added to a solution of H₂[1a ] (1.26 g,

4.05 mmol) in pentane (14 mL) at room temperature. Crystals

formed upon mixing. After 12 h the crystals were isolated by

filtration and the mother liquor was cooled to -30 °C to yield

a second crop; total yield 1.90 g (97%): ¹H NMR δ 2.43 (s, 12,

ArMe), 2.65 (s, 12, ZrNMe₂), 3.21 (t, 4, CH₂N), 3.55 (t, 4, OCH₂),

[6.92 (t, 2), 7.11 (d, 4)] (H_aryl); ¹³C{¹H} NMR δ 19.6 (ArMe),

42.1 (ZrNMe₂), 54.1 (CH₂N), 72.8 (OCH₂), [123.7, 128.9, 134.9,

152.7] (C_aryl). Anal. Calcd for C₃₂H₅₄N₄OZr: C, 58.85; H, 7.82;

N, 11.44. Found: C, 59.16; H, 7.95; N, 11.46.

[1a ]Zr (CH₂CMe₃)₂. A solution of LiCH₂CMe₃(68 mg, 0.87

mmol) in ether (2 mL) was added to a suspension of [1a ]ZrCl₂

(200 mg, 0.42 mmol) in ether (6 mL) at -30 °C. [1a ]ZrCl₂

dissolved immediately, and a fine precipitate of LiCl formed

rapidly. After 15 min the product was isolated by crystalliza-

tion from ether at -30 °C to yield 146 mg (63%) of colorless

crystals in two crops: ¹H NMR δ 0.91 (s, 4, ZrCH₂CMe₃), 0.99

(s, 18, ZrCH₂CMe₃), 2.53 (s, 12, ArMe), 3.26 (t, 4, CH₂N), 3.43

(t, 4, OCH₂C), [6.96 (t, 2), 7.11 (d, 4)] (H_aryl); ¹³C{¹H} NMR δ

20.3 (ArMe), 35.0 (CH₂CMe₃), 36.5 (CH₂CMe₃), 55.7 (CH₂N),

73.4 (OCH₂), 89.1 (ZrCH₂), [125.4, 129.5, 135.1, 152.3] (C_aryl).

Anal. Calcd for C₃₀H₄₈N₂OZr: C, 66.24; H, 8.89; N, 5.15.

Found: C, 66.17; H, 8.55; N, 4.98.

[1a ]Hf(NMe₂)₂. A solution of 2.50 g (7.06 mmol) of Hf-

(NMe₂)₄in 5 mL of pentane was added at room temperature

to a solution of 2.19 g (7.02 mmol) of H₂[1a ] in 35 mL of

pentane. After 40 min all volatile components of the reaction

mixture were removed in vacuo to leave analytically pure [1a ]-

Hf(NMe₂)₂as a white powder; yield 3.91 g (96%): ¹H NMR δ

2.45 (s, 12, ArMe), 2.69 (s, 12, HfNMe₂), 3.25 (t, 4, CH₂N), 3.52

(t, 4, OCH₂), [6.90 (t, 2), 7.12 (d, 4)] (H_aryl); ¹³C{¹H} NMR δ

19.14 (ArMe), 41.54 (HfNMe₂), 53.88 (CH₂N), 72.61 (OCH₂),

[123.39, 128.46, 134.90, 152.52] (C_aryl). Anal. Calcd for C₂₄H₃₈

-

ON₄Hf: C, 49.95; H, 6.64; N, 9.71. Found: C, 50.01; H, 6.27;

N, 9.50.

[1a ]Zr Cl₂. Neat Me₃SiCl (5.2 g, 48 mmol) was added to a

solution of [1a ]Zr(NMe₂)₂(1.17 g, 2.39 mmol) in 80 mL of

diethyl ether at room temperature. The solutions were mixed

thoroughly. After 46 h colorless needles (0.99 g) were isolated

by filtration (88% yield). Using less solvent led to cocrystal-

lization of intermediate [1a ]Zr(NMe₂)Cl with the desired

product: ¹H NMR δ 2.43 (s, 12, ArMe), 3.09 (t, 4, CH₂N), 3.49

[1a ]HfCl₂. To a solution of 2.02 g (3.5 mmol) of [1a ]Hf-

(NMe₂)₂in 80 mL of ether at room temperature was added

3.80 g (35 mmol) of Me₃SiCl. After 4 days the reaction mixture

was taken to dryness in vacuo and the white residue was

washed twice with 10 mL portions of ether and dried in vacuo

to afford 1.85 g (94%) of analytically pure [1a ]HfCl₂as a

colorless powder: ¹H NMR δ 2.46 (s, 12, ArMe), 3.19 (t, 4,

CH₂N), 3.47 (t, 4, OCH₂), [6.94 (m, 2), 7.04 (d, 4)] (H_aryl); ¹³C-

{¹H} NMR δ 19.06 (ArMe), 54.82 (CH₂N), 73.87 (OCH₂),

1

(t, 4, OCH₂), [6.94 (dd, 2), 7.01 (d, 4)] (H_aryl); H NMR (CD₂-

Cl₂) δ 2.43 (s, 12, ArMe), 3.77 (t, 4, CH₂N), 4.62 (t, 4, OCH₂),

[7.01 (dd, 2), 7.10 (d, 4)] (H_aryl); ¹³C{¹H} NMR δ 19.4 (ArMe),

55.8 (CH₂N), 73.8 (OCH₂), [126.7, 129.7, 134.6, 148.5] (C_aryl).

Anal. Calcd for C₂₀H₂₆Cl₂N₂OZr: C, 50.83; H, 5.55; N, 5.93.

Found: C, 50.83; H, 5.57; N, 5.84.

[125.86, 129.21, 134.65, 149.06] (C_aryl). Anal. Calcd for C₂₀H₂₆

-

Cl₂N₂OHf: C, 42.91; H, 4.68; N, 5.00. Found: C, 42.82; H,

4.33; N, 4.90.

[1a ]Zr Me₂. A chilled (-30 °C) solution of MeMgBr (3.0 M

in ether, 0.580 mL, 1.74 mmol) was added to a suspension of

[1a ]ZrCl₂(400 mg, 0.85 mmol) in CH₂Cl₂(10 mL) at -30 °C.

After the mixture was stirred for 20 min at room temperature,

dioxane (159 mg, 1.80 mmol) was added. After 20 min of

additional stirring, the reaction mixture was filtered. Upon

addition of pentane (5 mL) the filtrate turned slightly cloudy

and was filtered again. All volatile components were then

removed from the filtrate. The residue was dissolved in ether,

and the solution was cooled to -30 °C to yield 252 mg (69%)

of colorless cubes in two crops. Crystals suitable for X-ray

crystallography were obtained by crystallization from ether

at -30 °C: ¹H NMR δ 0.27 (s, 6, ZrMe), 2.44 (s, 12, ArMe),

3.27 (m, 8, OCH₂CH₂N), [7.01 (dd, 2), 7.14 (d, 4)] (H_aryl); ¹³C-

{¹H} NMR δ 19.3 (ArMe), 43.4 (ZrMe), 55.5 (CH₂N), 74.7

(OCH₂), [125.7, 129.4, 136.5, 148.7] (C_aryl). Anal. Calcd for

[1a ]HfMe₂. To a stirred suspension of 1.73 g (3.09 mmol)

of [1a ]HfCl₂in 50 mL of ether was added 2.15 mL of a 3.0 M

solution of MeMgCl in THF. After 50 min at room tempera-

ture the reaction mixture was filtered and the white residue

was washed with an additional 30 mL of ether. The ether

fractions were combined, filtered through Celite, and evapo-

rated to dryness in vacuo to yield crude [1a ]HfMe₂. Analyti-

cally pure product was obtained in 78% yield as a colorless

powder by recrystallization from CH₂Cl₂/Et₂O (1:1) mixtures

at -30 °C: ¹H NMR δ 0.06 (s, 6, HfMe), 2.46 (s, 12, ArMe),

3.31 (m, 8, OCH₂CH₂N), [6.98 (t, 2), 7.12 (d, 4)] (H_aryl); ¹³C-

{¹H} NMR δ 18.95 (ArMe), 53.74 (HfMe), 55.07 (CH₂N), 73.88

(OCH₂), [125.02, 128.99, 135.96, 148.86] (C_aryl). Anal. Calcd

for C₂₂H₃₂N₂OHf: C, 50.91; H, 6.21; N, 5.40. Found: C, 50.48;

H, 6.32; N, 5.40.

C

₂₂H₃₂N₂OZr: C, 61.21; H, 7.47; N, 6.49. Found: C, 61.31; H,

[1a ]Hf(CH₂CHMe₂)₂. To a stirred suspension of 0.84 g (1.5

mmol) of [1a ]HfCl₂in 30 mL of ether was added 1.5 mL of a

2.0 M solution of ClMgCH₂CHMe₂in THF. After 1 h the

mixture was filtered, and the ether was removed from the

filtrate in vacuo to give 895 mg (99%) of [1a ]Hf(CH₂CHMe₂)₂

as a colorless powder pure by NMR: ¹H NMR δ 0.51 (d, 4,

HfCH₂), 0.95 (d, 12, CHMe₂), 2.01 (heptet, 2, CHMe₂), 2.56 (s,

12, ArMe), 3.36 (m, 8, OCH₂CH₂N), [7.01 (m, 2), 7.16 (m, 4)]

(H_aryl); ¹³C{¹H} NMR δ 19.29 (ArMe), 29.08 (CHMe₂), 30.16

(CHMe₂), 55.18 (CH₂N), 73.12 (OCH₂), 88.31 (HfCH₂), [124.93,

129.04, 135.44, 150.06] (C_aryl).

[1b]TiCl₂. Preparation of this complex was completely

analogous to that of [1a ]TiCl₂from 758 mg of the 90% (by

mass) pure H₂[1b] (1.86 mmol), 1.49 mL of 2.5 M solution of

BuLi (3.73 mmol), and 623 mg (1.86 mmol) of TiCl₄(THF)₂.

The isolated yield of red [1b]TiCl₂was 346 mg (38%): ¹H NMR

δ 1.32 (t, 12, ArCH₂Me), [2.75 (A wing of AB q of q), 3.05 (B

wing of AB q of q), 8] (ArCH₂Me), 3.35 (t, 4, CH₂N), 3.70 (t, 4,

OCH₂), 7.09 (pseudo s, 6, H_aryl); ¹³C{¹H} NMR δ 15.22

7.56; N, 6.46.

[1a]Zr (CH₂CMe₃)Cl. Achilled (-30 °C)solution ofBrMgCH₂-

CMe₃(3.16 M in ether, 0.236 mL, 0.74 mmol) was added to a

suspension of [1a ]ZrCl₂(160 mg, 0.34 mmol) in CH₂Cl₂(8 mL)

at -30 °C. A fine precipitate formed, and after the mixture

was stirred for 30 min at room temperature, dioxane (70 mg,

1.80 mmol) was added. After 20 min the reaction mixture was

filtered. All volatile components were removed from the

filtrate, and the residue was dissolved in ether and filtered

again. The ether solution was cooled to -30 °C to yield 42

mg (24%) of colorless crystals in two crops: ¹H NMR δ 1.11

(s, 9, CH₂CMe₃), 1.23 (s, 2, ZrCH₂), 2.46 (s, 6, ArMe), 2.50 (s,

6, ArMe), 3.16 (m, 4, CH₂N), 3.37 (m, 2, OCH₂C), 3.50 (m, 2,

OCH₂C), [6.97 (t, 2), 7.08 (d, 4)] (H_aryl); ¹³C{¹H} NMR δ [19.8,

20.0] (ArMe), 35.0 (CH₂CMe₃), 35.8 (CH₂CMe₃), 55.5 (CH₂N),

73.4 (OCH₂), 81.4 (ZrCH₂), [126.1, 129.52, 129.55, 135.05,

135.14, 149.8] (C_aryl). Anal. Calcd for C₂₅H₃₇ClN₂OZr: C,

59.08; H, 7.34; N, 5.51. Found: C, 59.14; H, 7.42; N, 5.46.

Group 4 Complexes with Diamido/ Donor Ligands

Organometallics, Vol. 17, No. 22, 1998 4809

(ArCH₂Me), 24.27 (ArCH₂Me), 61.06 (CH₂N), 72.17 (OCH₂),

[126.45, 127.15, 136.64, 155.10] (C_aryl).

mmol) in pentane (14 mL) at room temperature. Crystals

formed immediately. After the mixture stood overnight, the

crystals were collected and the mother liquor was cooled to

-30 °C to yield a second crop; total yield 3.85 g (68%): ¹H

NMR δ 1.28 (d, 12, CHMe₂), 1.31 (d, 12, CHMe₂), 2.56 (s, 12,

ZrNMe₂), 3.33 (t, 4, CH₂N), 3.56 (t, 4, OCH₂), 3.71 (heptet, 4,

CHMe₂), 7.05-7.18 (m, 6, H_aryl); ¹³C{¹H} NMR δ 25.0 (CHMe₂),

27.0 (CHMe₂), 28.6 (CHMe₂), 42.7 (ZrNMe₂), 57.7 (CH₂N), 72.8

(OCH₂), [124.2, 125.2, 146.2, 150.2] (C_aryl). Anal. Calcd for

[1b]TiMe₂. To a stirred suspension of 243 mg (0.5 mmol)

of [1b]TiCl₂in 9 mL of ether was added 0.34 mL of a 3.0 M

solution of MeMgCl in THF. The greenish mixture was stirred

for 1 h at room temperature. The solvent was removed in

vacuo. Pentane (10 mL) was added to the residue, and the

mixture stood at -30 °C. After 2 h the precipitate was isolated

and dried in vacuo to yield 85 mg (38%) of pure [1b]TiMe₂in

a form of a yellow-brown powder: ¹H NMR δ 0.89 (s, 6, TiMe),

1.32 (t, 12, ArCH₂Me), 3.01 (q, 8, ArCH₂Me), 3.34 (t, 4, CH₂N),

3.45 (t, 4, OCH₂), 7.10-7.21 (m, 6, H_aryl); ¹³C{¹H} NMR δ 15.81

(ArCH₂Me), 24.46 (ArCH₂Me), 58.82 (CH₂N), 61.14 (TiMe),

70.64 (OCH₂), [125.83, 126.56, 139.32, 151.92] (C_aryl).

[1b]Zr (NMe₂)₂a n d [1b]Hf(NMe₂)₂. These complexes were

prepared on a 2 mmol scale from a solution of H₂[1b] in 4 mL

of pentane and M(NMe₂)₄(M ) Zr, Hf) in 4 mL of pentane.

After 1 h all volatile components were removed in vacuo

leaving the desired complexes in quantitative yields and of

sufficient purity for further purposes. The Zr compound was

isolated as a slightly yellowish oil which solidified in the

freezer. The Hf analogue was formed as a colorless solid.

[1b]Zr(NMe₂)₂: ¹H NMR δ 1.32 (t, 12, ArCH₂Me), 2.88 (m,

8, ArCH₂Me), 2.61 (s, 12, ZrNMe₂), 3.25 (t, 4, CH₂N), 3.57 (t,

4, OCH₂), [7.04 (dd, 2), 7.15 (d), 4] (H_aryl); ¹³C{¹H} NMR δ 15.81

(ArCH₂Me), 24.51 (ArCH₂Me), 41.75 (ZrNMe₂), 55.60 (CH₂N),

72.26 (OCH₂), [123.91, 126.00, 140.41, 151.24] (C_aryl).

[1b]Hf(NMe₂)₂: ¹H NMR δ 1.32 (t, 12, ArCH₂Me), [2.85 (A

wing of AB q of q), 2.99 (B wing of AB q of q), 8] (ArCH₂Me),

2.66 (s, 12, HfNMe₂), 3.28 (t, 4, CH₂N), 3.52 (t, 4, OCH₂), [7.04

(dd, 2), 7.16 (d), 4] (H_aryl); ¹³C{¹H} NMR δ 15.83 (ArCH₂Me),

24.44 (ArCH₂Me), 41.60 (HfNMe₂), 55.66 (CH₂N), 72.44 (OCH₂),

[123.92, 125.94, 140.67, 151.44] (C_aryl).

C

₃₂H₅₄N₄OZr: C, 63.84; H, 9.04; N, 9.31. Found: C, 63.96; H,

9.16; N, 9.20.

[1c]Zr Cl₂. Neat Me₃SiCl (578 mg, 5.3 mmol) was added to

a solution of [1c]Zr(NMe₂)₂(400 mg, 0.664 mmol) in 10 mL of

diethyl ether at room temperature. The solutions were mixed

thoroughly by shaking, and the reaction mixture was allowed

to stand overnight at room temperature to yield colorless

crystals (285 mg; 73% yield). If the reaction mixture is too

concentrated, [1c]Zr(NMe₂)Cl will cocrystallize with [1c]ZrCl₂:

¹H NMR δ 1.26 (d, 12, CHMe₂), 1.51 (d, 12, CHMe₂), 3.35 (t,

4, CH₂N), 3.66 (t, 4, OCH₂), 3.73 (heptet, 4, CHMe₂), 7.17 (br,

6, H_aryl); ¹³C{¹H} NMR δ 25.4 (CHMe₂), 26.9 (CHMe₂), 29.0

(CHMe₂), 59.3 (CH₂N), 73.6 (OCH₂), [125.1, 127.6, 145.1, 146.4]

(C_aryl). Anal. Calcd for C₂₈H₄₂Cl₂N₂OZr: C, 57.51; H, 7.24;

N, 4.79. Found: C, 57.10; H, 7.28; N, 4.46.

[1c]Zr Me₂. A chilled (-30 °C) solution of MeMgBr (4.1 M

in ether, 0.428 mL, 1.75 mmol) was added to a suspension of

[1c]ZrCl₂(500 mg, 0.85 mmol) in diethyl ether (20 mL) at -30

°C. A fine precipitate slowly replaced the suspension of

crystals, and after the mixture was stirred for 2 h at room

temperature, dioxane (154 mg, 1.75 mmol) was added. After

an additional 20 min all volatile components were removed

and the residue was extracted with pentane. Recrystallization

from pentane yielded 280 mg (61%) of colorless crystals: ¹H

NMR δ 0.30 (s, 6, ZrMe), 1.23 (d, 12, CHMe₂), 1.38 (d, 12,

CHMe₂), 3.41 (br, 8, OCH₂CH₂N), 3.84 (heptet, 4, CHMe₂), 7.12

(br, 6, H_aryl); ¹³C{¹H} NMR δ 24.9 (CHMe₂), 27.3 (CHMe₂), 28.9

(CHMe₂), 43.6 (ZrMe), 58.6 (CH₂N), 73.6 (OCH₂), [124.7, 126.5,

146.5, 147.1] (C_aryl). Anal. Calcd for C₃₀H₄₈N₂OZr: C, 66.24;

H, 8.89; N, 5.15. Found: C, 66.32; H, 8.87; N, 5.12.

[1b]Zr Cl₂a n d [1b]HfCl₂. These two compounds were

prepared on a 2 mmol scale from [1b]M(NMe₂)₂(M ) Zr, Hf)

in 7 mL of ether and 20 equiv of Me₃SiCl. After 14 h the

crystalline products were filtered off, washed with two portions

of pentane (3 mL each), and dried in vacuo. The isolated yields

were 88% for [1b]ZrCl₂and 83% for [1b]HfCl₂.

[1b]ZrCl₂: ¹H NMR δ 1.33 (t, 12, ArCH₂Me), [2.81 (A wing

of AB q of q), 3.02 (B wing of AB q of q), 8] (ArCH₂Me), 3.17 (t,

4, CH₂N), 3.57 (t, 4, OCH₂), 7.10 (pseudo s, 6, H_aryl); ¹³C{¹H}

NMR δ 15.54 (ArCH₂Me), 24.48 (ArCH₂Me), 57.26 (CH₂N),

73.28 (OCH₂), [126.70, 126.73, 139.70, 147.28] (C_aryl).

[1b]HfCl₂: ¹H NMR δ 1.34 (t, 12, ArCH₂Me), [2.86 (A wing

of AB q of q), 3.04 (B wing of AB q of q), 8] (ArCH₂Me), 3.25 (t,

4, CH₂N), 3.49 (t, 4, OCH₂), 7.04-7.14 (m, 6, H_aryl); ¹³C{¹H}

NMR δ 15.63 (ArCH₂Me), 24.42 (ArCH₂Me), 55.65 (CH₂N),

73.57 (OCH₂), [126.24, 126.65, 140.04, 148.15] (C_aryl).

[1b]Zr Me₂a n d [1b]HfMe₂. The procedures for the syn-

theses of these two complexes were identical. To a suspension

of 1.4 mmol of [1b]MCl₂(M ) Zr, Hf) in 12 mL of ether was

added 0.95 mL of a 3.0 M solution of MeMgCl in THF. After

1 h the solvent was removed in vacuo and the resulting solid

was washed with 3 mL of cold (-30 °C) pentane and dried to

yield [1b]ZrMe₂(88%, pale beige powder) or [1b]HfMe₂(93%,

colorless powder), respectively.

[1b]ZrMe₂: ¹H NMR δ 0.21 (s, 6, ZrMe), 1.27 (t, 12,

ArCH₂Me), 2.93 (symm 10-line m, 8, ArCH₂Me), 3.34 (m, 8,

OCH₂CH₂N), 7.09-7.19 (m, 6, H_aryl); ¹³C{¹H} NMR δ 16.11

(ArCH₂Me), 24.65 (ArCH₂Me), 42.96 (ZrMe), 56.88 (CH₂N),

74.01 (OCH₂), [125.80, 126.64, 141.83, 147.07] (C_aryl).

[1b]HfMe₂: ¹H NMR δ 0.05 (s, 6, HfMe), 1.29 (t, 12,

ArCH₂Me), 2.96 (m, 8, ArCH₂Me), 3.34 (pseudo s, 8, OCH₂-

CH₂N), 7.08-7.18 (m, 6, H_aryl); ¹³C{¹H} NMR δ 16.05 (ArCH₂-

Me), 24.52 (ArCH₂Me), 53.60 (HfMe), 56.86 (CH₂N), 73.55

(OCH₂), [125.50, 126.59, 141.57, 147.72] (C_aryl). Anal. Calcd

for C₂₆H₄₀N₂OHf: C, 54.30; H, 7.01; N, 4.87. Found: C, 53.90;

H, 7.38; N, 4.87.

[1c]Zr (CH ₂CH Me₂)₂. A chilled (-30 °C) solution of

BrMgCH₂CHMe₂(2.51 M in ether, 0.286 mL, 0.72 mmol) was

added to a suspension of [1c]ZrCl₂(205 mg, 0.35 mmol) in

diethyl ether (10 mL) at -30 °C. A fine precipitate slowly

replaced the suspension of crystals, and after the solution was

stirred for 1.5 h at room temperature, dioxane (63 mg, 0.72

mmol) was added. After 20 min all volatile components were

removed in vacuo and the residue was recrystallized from

pentane to yield 158 mg (72%) of colorless crystals: ¹H NMR

δ 0.70 (d, 4, ZrCH₂), 0.85 (d, 12, CH₂CHMe₂), 1.23 (d, 12,

CHMe₂), 1.45 (d, 12, CHMe₂), 1.92 (heptet, 2, CH₂CHMe₂), 3.42

(br, 8, OCH₂CH₂N), 3.91 (heptet, 4, CHMe₂), 7.12-7.15 (br, 6,

H

_aryl); ¹³C{¹H} NMR δ 24.6 (CHMe₂), 27.4 (CHMe₂), 28.4 (CH₂-

CHMe₂), 28.9 (CHMe₂), 29.7 (CH₂CHMe₂), 58.3 (CH₂N), 74.5

(OCH₂), 78.1 (ZrCH₂), [124.6, 126.2, 146.0, 149.2] (C_aryl). Anal.

Calcd for C₃₆H₆₀N₂OZr: C, 68.84; H, 9.63; N, 4.46. Found: C,

68.96; H, 9.59; N, 4.40.

[1c]Hf(NMe₂)₂. A solution of Hf(NMe₂)₄(833 mg, 2.35

mmol) in 3.5 mL of pentane was added to a solution of 1.0 g

(2.35 mmol) of H₂[1c] in 3.5 mL of pentane at room temper-

ature. After 20 min crystallization of the product began. The

reaction was allowed to proceed for 12 h, after which colorless

blocks were isolated and washed with cold pentane. Cooling

of the mother liquor to -30 °C yielded a second crop; total yield

1.21 g (74.7%): ¹H NMR δ 1.31 (d, 12, CHMe₂), 1.35 (d, 12,

CHMe₂), 2.65 (s, 12, HfNMe₂), 3.39 (t, 4, CH₂N), 3.53 (t, 4,

OCH₂), 3.78 (heptet, 4, CHMe₂), 7.09-7.21 (m, 6, H_aryl); ¹³C-

{¹H} NMR δ 24.64 (CHMe₂), 26.67 (CHMe₂), 28.09 (CHMe₂),

42.26 (HfNMe₂), 57.39 (CH₂N), 72.49 (OCH₂), [123.79, 124.76,

145.99, 150.12] (C_aryl). Anal. Calcd for C₃₂H₅₄N₄OHf: C, 55.76;

H, 7.90; N, 8.13. Found: C, 55.33; H, 7.90; N, 7.87.

[1c]Zr (NMe₂)₂. A solution of Zr(NMe₂)₄(2.5 g, 9.4 mmol)

in pentane (4 mL) was added to a solution of H₂[1c] (4.0 g, 9.4

[1c]HfCl₂. Me₃SiCl (1.67 g, 15.4 mmol) was added to a

4810 Organometallics, Vol. 17, No. 22, 1998

Aizenberg et al.

solution of 1.06 g (1.54 mmol) of [1c]Hf(NMe₂)₂in 25 mL of

ether. After 1 day the vial was shaken to induce crystalliza-

tion. A fine white precipitate formed immediately. After

several hours it was isolated by filtration and washed with

pentane. The mother liquor was cooled to -30 °C to yield a

second crop; total yield 940 mg (90.5%): ¹H NMR δ 1.28 (d,

12, CHMe₂), 1.49 (d, 12, CHMe₂), 3.41 (t, 4, CH₂N), 3.57 (t, 4,

OCH₂), 3.77 (heptet, 4, CHMe₂), 7.15 (br s, 6, H_aryl); ¹³C{¹H}

NMR δ 25.28 (CHMe₂), 26.65 (CHMe₂), 28.67 (CHMe₂), 58.51

(CH₂N), 73.62 (OCH₂), [124.66, 126.76, 145.08, 146.61] (C_aryl).

Satisfactory elemental analyses could not be obtained on this

compound (low in C).

[1c]HfMe₂. MeMgCl in THF (0.78 mL of a 3.0 M solution)

was added to a stirred suspension of 753 mg (1.12 mmol) of

[1c]HfCl₂in 12 mL of ether. The reaction mixture was stirred

for an additional 4 h and filtered. The solvent was removed

from the filtrate in vacuo, and the residue was extracted with

minimal ether. The extract was filtered, and the ether was

removed in vacuo to give 575 mg (81.3%) of pure [1c]HfMe₂in

the form of a colorless powder: ¹H NMR δ 0.17 (s, 6, HfMe),

1.28 (d, 12, CHMe₂), 1.42 (d, 12, CHMe₂), 3.43 (m, 8,

OCH₂CH₂N), 3.90 (heptet, 4, CHMe₂), 7.18 (br s, 6, H_aryl); ¹³C-

{¹H} NMR δ 24.95 (CHMe₂), 26.98 (CHMe₂), 28.58 (CHMe₂),

53.90 (HfMe), 58.33 (CH₂N), 73.55 (OCH₂), [124.29, 125.90,

146.05, 147.18] (C_aryl). Satisfactory elemental analyses could

not be obtained on this compound (low in C).

[1c]HfEt₂. EtMgBr (0.117 mL of a 3.0 M ether solution,

0.35 mmol, 2.05 equiv) was added to a suspension of 115 mg

(0.171 mmol) of [1c]HfCl₂in 7 mL of ether. A colorless solution

formed almost immediately. After 1.5 h all the volatile

components were removed in vacuo and the resulting residue

was extracted with pentane. The extract was filtered and

evacuated to dryness to yield crude [1c]HfEt₂. Recrystalliza-

tion from pentane at -30 °C gave colorless needles (80 mg,

74%) suitable for X-ray analysis: ¹H NMR δ 0.50 (q, 4, HfCH₂),

1.28 (t, 6, CH₂Me), 1.28 (d, 12, CHMe₂), 1.44 (d, 12, CHMe₂),

3.40 (m, 4, CH₂N), 3.49 (m, 4, OCH₂), 3.89 (heptet, 4, CHMe₂),

7.16-7.22 (m, 6, H_aryl); ¹³C{¹H} NMR δ 12.04 (CH₂Me), 24.29

(CHMe₂), 27.11 (CHMe₂), 28.44 (CHMe₂), 58.16 (CH₂N), 66.33

(HfCH₂), 73.52 (OCH₂), [124.16, 125.76, 146.21, 147.71] (C_aryl).

(TsOCH₂CH₂)₂S. A slurry of pulverized KOH flakes (3.96

g, 70.6 mmol) in THF (15 mL) was added to a solution of

(HOCH₂CH₂)₂S (2 mL, 20 mmol) in THF (10 mL). The reaction

mixture was stirred vigorously at room temperature for 1 h

and then cooled to 0 °C. A solution of tosyl chloride (7.63 g,

40 mmol) in THF (25 mL) was added dropwise to the chilled

mixture over the course of 1 h. The reaction mixture was

stirred vigorously at 0 °C for 24 h and filtered. The filtrate

was then dried over anhydrous MgSO₄and the solvent was

removed in vacuo to give a viscous pale yellow oil in essentially

quantitative yield which would sometimes crystallize upon

standing. No further characterization or purification of this

compound has been undertaken due to its apparent decom-

position in the solid state, even at 0 °C: ¹H NMR (CDCl₃) δ

7.80 (d, 4, H_aryl), 7.37 (d, 4, H_aryl), 4.10 (t, 4, OCH₂), 2.74 (t, 4,

CH₂S), 2.47 (s, 6, CH₃).

[(2,6-Me₂C₆H₃)NHCH₂CH₂]₂S (H₂[2a ]). Neat H₂N(2,6-

Me₂C₆H₃) (14.78 mL, 120 mmol) was added to (TsOCH₂CH₂)₂S

(approximately 20 mmol), and the resulting viscous yellow

solution was stirred at 50 °C for 24 h. After several hours at

50 °C, a white precipitate was observed. After 24 h, a solution

of NaOH (10 g, 250 mmol) in water (30 mL) was added to the

reaction mixture. After 30 min the mixture was extracted with

several portions of diethyl ether. The ether layers were

combined and dried over anhydrous MgSO₄, and the ether was

removed in vacuo. The resulting yellow oil was heated in

vacuo (80 °C, 300 mTorr) in order to remove any excess H₂N-

(2,6-Me₂C₆H₃). The remaining viscous orange oil was dissolved

in pentane and the solution stood at 0 °C to yield pale peach

crystals in 70% yield overall (from the diol): ¹H NMR δ 6.96

(d, 4, H_m), 6.86 (t, 2, H_p), 3.21 (br s, 2, NH), 2.90 (t, 4, NCH₂),

2.31 (t, 4, CH₂S), 2.18 (s, 12, ArMe); ¹³C{¹H} NMR δ 146.3

(C_ipso), 130.0 (C_o), 129.6 (C_m), 122.8 (C_p), 47.6 (NHCH₂), 33.3

(CH₂S), 19.3 (ArMe). Anal. Calcd for C₂₀H₂₈N₂S: C, 73.12;

H, 8.59; N, 8.53. Found: C, 73.19; H, 8.55; N, 8.48.

[(2,6-i-P r ₂C₆H₃)NHCH₂CH₂]₂S (H₂[2c]). Neat 2,6-i-Pr₂C₆H₃-

NH₂(22.63 mL, 120 mmol) was added to (TsOCH₂CH₂)₂S

(approximately 20 mmol), and the resulting viscous yellow

solution was stirred at 50 °C for 24 h. After several hours at

50 °C, a white precipitate was observed. After 24 h a solution

of NaOH (10 g, 250 mmol) in water (30 mL) was added to the

reaction mixture. After it was stirred for an additional 30 min

at room temperature, the mixture was extracted with several

portions of diethyl ether. The organic layers were combined

and dried over anhydrous MgSO₄, and solvent was removed

in vacuo. The resulting yellow oil was heated in vacuo (90

°C, 300 mTorr) in order to remove any excess H₂N(2,6-i-

Pr₂C₆H₃). The remaining viscous orange oil was essentially

pure product (7.26 g, 82%); no further purification was

necessary: ¹H NMR δ 7.10 (m, 6, H_aryl), 3.42 (sept, 4, CHMe₂),

2.98 (t, 4, NCH₂), 2.64 (t, 4, CH₂S), 1.23 (d, 24, CHMe₂).

[2a ]Zr (NMe₂)₂. A solution of Zr(NMe₂)₄(0.81 g, 3.04 mmol)

in pentane (4 mL) and a solution of H₂[2a ] (1.0 g, 3.04 mmol)

in pentane (14 mL) and toluene (2 mL) were combined and

allowed to stand for 20 h at room temperature. The solvent

was removed in vacuo, and the remaining oily yellow residue

was washed several times with pentane to remove all remain-

ing traces of toluene. Recrystallization from pentane at -35

°C gave 1.21 g (79%) of off-white crystals: ¹H NMR δ 7.10 (d,

1

4, H_m), 6.90 (t, 2, H_p), 3.31 (br s), 2.54 (t), 2.40 (s); H NMR

(toluene-d₈, 100 °C) δ 6.98 (d, 4, H_m), 6.79 (t, 2, H_p), 3.39 (t, 4,

NCH₂), 2.73 (t, 4, CH₂S), 2.58 (br s, 12, ZrNMe₂), 2.37 (s, 12,

ArMe); ¹³C{¹H} NMR δ 152.9 (C_ipso), 135.1 (C_o), 129.0 (C_m),

123.9 (C_p), 56.7 (NCH₂), 43.3 (br, ZrNMe₂), 36.2 (CH₂S), 19.7

(ArMe). Anal. Calcd for C₂₄H₃₈N₄SZr: C, 56.98; H, 7.57; N,

11.08. Found: C, 57.02; H, 7.50; N, 10.85.

[2a ]Zr Cl₂. Neat Me₃SiCl (6.1 mL, 48 mmol) was added to

a solution of [2a ]Zr(NMe₂)₂(1.21 g, 2.39 mmol) in 80 mL of

diethyl ether at room temperature, and the mixture was

shaken vigorously. After approximately 48 h pale yellow

crystals were isolated in 97% yield: ¹H NMR (CD₂Cl₂) 7.08

(m, 6, H_aromat), 4.23 (m, 2), 3.60-3.38 (m, 6), 2.40 (s, 6, ArMe),

2.42 (s, 6, ArMe); ¹³C{¹H} NMR (CD₂Cl₂) δ 146.2 (C_ipso), 136.1

(C_o), 135.7 (C_o), 129.6 (C_m), 129.4 (C_m), 127.2 (C_p), 59.2 (NCH₂),

34.3 (CH₂S), 19.32 (ArMe), 19.28 (ArMe). Anal. Calcd for

C

₂₀H₂₆Cl₂N₂SZr: C, 49.16; H, 5.36; N, 5.73. Found: C, 48.89;

H, 5.40; N, 5.63.

[2a ]Zr Me₂. A chilled (-30 °C) solution of MeMgBr (3.0 M

in ether, 0.35 mL, 1.05 mmol) was added to a suspension of

[2a ]ZrCl₂(0.250 g, 0.512 mmol) in diethyl ether (6 mL) at -30

°C. The reaction mixture was warmed to room temperature

and was stirred for 20 min. 1,4-Dioxane (92 µL, 1.08 mmol)

was added, and the mixture was stirred at room temperature

for an additional 10 min. The mixture was then filtered

through Celite, and the solvent was removed from the filtrate

in vacuo. The white residue was recrystallized from pentane

at -35 °C to give the product in 71% yield: ¹H NMR δ 7.01

1

(d, 4, H_m), 6.98 (t, 2, H_p), 3.35 (br s), 2.44 (m), 0.29 (br s); H

NMR (toluene-d₈, 50 °C) δ 7.02 (d, 4, H_m), 6.93 (t, 2, H_p), 3.38

(t, 4, NCH₂), 2.57 (t, 4, CH₂S), 2.40 (s, 12, ArMe), 0.02 (br s, 6,

ZrMe₂); ¹³C{¹H} NMR δ 146.3 (C_ipso), 137.0 (C_o), 129.5 (C_m),

126.4 (C_p), 57.6 (NCH₂), 33.5 (CH₂S), 19.5 (ArMe).

¹³CH₃MgI (1.1M in diethyl ether) was used to prepare [2a ]-

Zr(¹³CH₃)₂: ¹³C NMR (CD₂Cl₂, -40 °C) δ 41.0, 37.5.

[2c]Zr (NMe₂)₂. A solution of Zr(NMe₂)₄(3.22 g, 7.31 mmol)

in toluene (15 mL) was added to a solution of H₂[2c] (1.95 g,

7.31 mmol) in toluene (35 mL) at room temperature, and the

mixture was heated to 110 °C overnight. After approximately

20 h at 110 °C the solvent was removed in vacuo. The

remaining orange oil was washed with several portions of

pentane to ensure removal of all traces of toluene. The oily

yellow solid was recrystallized from pentane at -35 °C to

Group 4 Complexes with Diamido/ Donor Ligands

Organometallics, Vol. 17, No. 22, 1998 4811

afford large yellow crystals in 66% yield: ¹H NMR δ 7.15 (m,

6, PhNMe₂), 3.23 (heptet, 4, CHMe₂), 3.33 (m, 2, CH₂N), 3.73

(m, 2, CH₂N), 3.98 (m, 2, OCH₂), 4.22 (m, 2, OCH₂), [5.85 (m,

4), 6.33 (br t, 1H)] (PhNMe₂), 7.11-7.27 (m, 6, H_aryl).

H

_aryl), 3.71 (br m), 3.50 (br s), 2.54 (br m), 1.36 (d, 12, CHMe₂),

1.24 (d, 12, CHMe₂); ¹H NMR (toluene-d₈, 100 °C) δ 7.17-6.92

(m, 6, H_aryl), 3.69 (sept, 4, CHMe₂), 3.56 (t, 4, NCH₂), 2.71 (t,

4, CH₂S), 2.56 (br s, 12, ZrNMe₂), 1.31 (d, 12, CHMe₂), 1.22 (d,

12, CHMe₂); ¹³C{¹H} NMR δ 150.3 (C_ipso), 146.4 (C_o), 125.5 (C_m),

124.4 (C_p), 59.8 (NCH₂), 44.5 (br, ZrNMe₂), 36.0 (CH₂S), 28.8

(CHMe₂), 27.1 (CHMe₂), 24.9 (CHMe₂).

Gen er a tion of {[1c]HfMe(P h NMe₂)}[B(C₆F ₅)₄]. [1c]-

HfMe₂(16 mg, 0.025 mmol) and [PhNMe₂H][B(C₆F₅)₄] (20 mg,

0.025 mmol) were combined in 0.8 mL of C₆D₅Br at -30 °C.

The reaction mixture was stirred for 10 min while being

warmed to room temperature, after which it was chilled to -20

°C: ¹H NMR (C₆D₅Br, -20 °C) δ -0.17 (s, 3, HfMe), 1.07 (d,

6, CHMe₂), 1.11 (d, 6, CHMe₂), 1.20 (d, 6, CHMe₂), 1.33 (d, 6,

CHMe₂), 2.62 (s, 6, PhNMe₂), 3.01 (heptet, 2, CHMe₂), 3.15

(heptet, 2, CHMe₂), 3.20 (dt, 2, CH₂N), 3.55 (br, 2, CH₂N), 3.78

(br, 2, OCH₂), 4.01 (br, 2, OCH₂), [5.62 (m, 2), 5.74 (t, 2), 6.06

(t, 1H)] (PhNMe₂), 7.05-7.22 (m, 6, H_aryl).

Gen er a tion of {[1a ]Zr Me(eth er )}[B(C₆F ₅)₄]. To a chilled

(-30 °C) suspension of [PhNMe₂H][B(C₆F₅)₄] (80 mg, 0.1 mmol)

in 2.5 mL of C₆H₅Cl was added a similarly precooled solution

of [1a ]ZrMe₂(43 mg, 0.1 mmol) in 2.5 mL of C₆H₅Cl. The

reaction mixture was periodically mixed by shaking and was

warmed slowly to room temperature. After 30 min all the

[PhNMe₂H][B(C₆F₅)₄] dissolved. Ether (1 mL) and then 15 mL

of pentane were then added, which led to bleaching of the color,

formation of an opaque solution, and slow precipitation of an

oily product. When the reaction mixture was cooled to -30

°C, the oil was transformed into tiny dendritic pale beige

microcrystals. These were separated, washed with two 5 mL

portions of pentane/ether (1:1), and dried in vacuo to yield 98

mg (84%) of light white powder. The compound is thermally

unstable both in solution and in the solid state and should be

stored at -30 °C: ¹H NMR (C₆D₅Br) δ 0.21 (t, 6, O(CH₂Me)₂),

0.40 (s, 3, ZrMe), 1.93 (s, 6, ArMe), 2.22 (s, 6, ArMe), 2.80 (q,

4, O(CH₂Me)₂), 3.15 (dt, 2, CH₂N), 3.48 (m, 2, CH₂N), 3.87 (m,

2, OCH₂), 4.05 (m, 2, OCH₂), 6.88-7.00 (m, 6, H_aryl); ¹³C{¹H}

NMR (C₆D₅Br) δ 11.74 (O(CH₂Me)₂), 17.98 (ArMe), 18.10

(ArMe), 40.57 (ZrMe), 54.81 (CH₂N), 68.22 (O(CH₂Me)₂), 74.32

(OCH₂), [127.78, 129.99, 134.28, 136.28, 140.90] (C_aryl), [136.4

(dm), 138.3 (dm), 148.5 (dm)] (C₆F₅) (one resonance belonging

to an aryl carbon and one belonging to a C₆F₅carbon were not

found, probably because of the overlap with the resonances of

the solvent); ¹⁹F{¹H} NMR (C₆D₅Br) δ -166.31 (m, F_m),

-162.44 (t, F_p), -132.17 (br, F_o).

Gen er a tion of {[1a ]HfMe(eth er )}[B(C₆F ₅)₄]. The pro-

cedure, molar amounts of the reactants, and observations were

similar to those described for the preparation of {[1a ]ZrMe-

(ether)}[B(C₆F₅)₄], but due to much higher thermal stability

of the Hf complexes involved, no cooling was necessary. {[1a ]-

HfMe(Et₂O)}[B(C₆F₅)₄] was formed in 82% yield as colorless

microcrystals: ¹H NMR (C₆D₅Br) δ 0.17 (t, 6, O(CH₂Me)₂), 0.36

(s, 3, HfMe), 2.01 (s, 6, ArMe), 2.22 (s, 6, ArMe), 2.80 (q, 4,

O(CH₂Me)₂), 3.20 (dt, 2, CH₂N), 3.46 (m, 2, CH₂N), 3.86 (m, 2,

OCH₂), 4.02 (m, 2, OCH₂), 6.88-6.98 (m, 6, H_aryl); ¹³C{¹H}

NMR (C₆D₅Br) δ 11.32 (O(CH₂Me)₂), 17.89 (ArMe), 18.28

(ArMe), 45.49 (HfMe), 54.31 (CH₂N), 68.59 (O(CH₂Me)₂), 75.04

(OCH₂), [127.07, 134.10, 135.49, 142.57] (C_aryl), [136.4 (dm),

138.4 (dm), 148.5 (dm)] (C₆F₅); two resonances belonging to

an aryl carbon and one belonging to a C₆F₅carbon were not

found, probably because of the overlap with the resonances of

the solvent; ¹⁹F{¹H} NMR (C₆D₅Br) δ -165.95 (t, 2F_meta),

-162.06 (t, 1F_para), -131.83 (br, 2F_ortho).

Gen er a tion of {[2a ]Zr Me(THF )}[B(C₆F ₅)₄]. Solid [2a ]-

ZrMe₂(17 mg, 37 µmol) was added to a solution of [PhNMe₂-

H]]B(C₆F₅)₄] (29 mg, 37 µmol) in C₆D₅Br (0.7 mL) at -35 °C.

The yellow solution was stirred at room temperature for 5 min,

during which time the color changed to brown. A drop of THF

was added, and the solution changed color to bright yellow:

¹H NMR (C₆D₅Br) δ 7.2 (t, 2, free PhNMe₂), 6.94 (m, H_aryl),

6.72 (t, 1, free PhNMe₂), 6.60 (d, 2, free PhNMe₂), 3.63 (m, 2,

NCH₂CH₂S), 3.42 (m, THF), 3.35 (m, 2, NCH₂CH₂S), 2.96 (m,

4, NCH₂CH₂S), 2.61 (s, 6, free PhNMe₂), 2.20 (s, 6, PhCH₃),

1.89 (s, 6, PhCH₃), 1.48 (m, THF), 0.20 (s, 3, ZrMe).

[2c]Zr Cl₂. Neat TMSCl (2.37 mL, 18.71 mmol) was added

to a room-temperature solution of [2c]Zr(NMe₂)₂(0.410 g, 0.66

mmol) in diethyl ether (10 mL). The resulting yellow solution

was stirred at room temperature for approximately 48 h. A

white precipitate was observed after several hours. After the

mixture was stirred at room temperature over two nights,

solvent was removed in vacuo, affording 0.317 g (79%) of white

powder: ¹H NMR (CDCl₃, room temperature) δ 7.26 (m, 6,

H

_aromat), 4.35 (m, 2), 3.70 to 3.40 (m, 10), 1.38 (d, 6, CHMe₂),

1.35 (d, 6, CHMe₂), 1.25 (d, 6, CHMe₂), 1.20 (d, 6, CHMe₂); ¹³C-

{¹H} NMR (CDCl₃) δ 145.5 (C_ipso), 144.6 (C_o), 127.4 (C_m), 124.6

(C_p), 61.5 (NCH₂), 34.0 (CH₂S), 29.0 (CHMe₂), 28.3 (CHMe₂),

27.2 (CHMe₂), 27.0 (CHMe₂), 24.6 (CHMe₂), 23.8 (CHMe₂).

Anal. Calcd for C₂₈H₄₂N₂SZrCl₂: Cl, 11.80. Found: Cl, 12.07.

[2c]Zr Me₂. A chilled solution of MeMgBr (3.0 M in ether,

0.352 mL, 1.06 mmol) was added to a suspension of [2c]ZrCl₂

(0.317 g, 0.527 mmol) in diethyl ether (15 mL) at -35 °C. The

reaction mixture was stirred at room temperature for 45 min.

1,4-Dioxane (0.09 mL, 1.06 mmol) was added, and after 15 min

the mixture was filtered through Celite. The solvent was

removed from the filtrate in vacuo, and the residue was

recrystallized from diethyl ether at -30 °C to yield white

crystals in 73% yield: ¹H NMR δ 7.18 (m, H_aryl), 3.8 (br s),

3.55 (br s), 2.51 (t), 1.40 (d, 12, CHMe₂), 1.20 (d, 12, CHMe₂),

0.35 (br s); ¹H NMR (toluene-d₈, 60 °C) δ 7.15-6.94 (m, 6,

H

_aryl), 3.80 (br, 4, CHMe₂), 3.60 (br, 4, NCH₂), 2.65 (br, 4,

CH₂S), 1.38 (d, 12, CHMe₂), 1.21 (d, 12, CHMe₂), 0.16 (br s, 6,

ZrMe₂); ¹³C{¹H} NMR δ 147.1 (C_ipso), 145.3 (C_o), 127.1 (C_m),

124.9 (C_p), 60.7 (NCH₂), 33.9 (CH₂S), 29.1 (CHMe₂), 27.4

(CHMe₂), 24.8 (CHMe₂). Anal. Calcd for C₃₀H₄₈N₂SZr: C,

64.34; H, 8.64; N, 5.00. Found: C, 64.46; H, 8.57; N, 4.91.

¹³CH₃MgI (1.1 M in diethyl ether) was used to prepare [2c]-

Zr(¹³CH₂)₂: ¹³C{¹H} NMR (C₆D₆, 0 °C) δ 45.59, 39.39.

[2c]Zr (i-Bu )₂. A chilled solution of ClMgCH₂CHMe₂(2.0

M in ether, 0.62 mL, 1.24 mmol) was added to a solution of

[2c]ZrCl₂(0.373 g, 0.62 mmol) in diethyl ether (15 mL) at -35

°C. The reaction was carried out as outlined for [2c]ZrMe₂to

give pale yellow crystals of the product in 46% yield: ¹H NMR

δ 7.18 (m, H_aromat), 3.82 (br s), 3.77 (br s), 2.50 (br s), 1.46 (d),

1.20 (d), 0.81 (br s); ¹H NMR (toluene-d₈, 80 °C) δ 7.19 to 6.91

(m, 6, H_aromat), 3.84 (br, 4, PhCHMe₂), 3.63 (br, 4, NCH₂), 2.66

(br, 4, CH₂S), 1.84 (br, 2, CH₂CHMe₂), 1.44 (d, 12, PhCHMe₂),

1.21 (d, 12, PhCHMe₂), 0.83 (m, 12, CH₂CHMe₂), 0.59 (br, 4,

ZrCH₂); ¹³C{¹H} NMR δ 147.9 (C_ipso), 146.6 (C_o), 127.3 (C_p),

125.1 (C_m), 81.7 (br, CH₂CHMe₂), 61.4 (CH₂N), 34.7 (SCH₂),

31.9 (br, CH₂CHMe₂), 29.0 (CHMe₂), 28.4 (CH₂CHMe₂), 27.8

(CHMe₂), 24.6 (CHMe₂).

Gen er a tion of {[1a ]Zr Me(P h NMe₂)}[B(C₆F ₅)₄]. Solid

[1a ]ZrMe₂(13 mg, 0.030 mmol) was added to a chilled (-30

°C) suspension of [PhNHMe₂][B(C₆F₅)₄] (24 mg, 0.030 mmol)

in C₆D₅Cl. The reaction mixture was stirred for 5 min while

it was warmed to room temperature and then chilled to 0 °C:

¹H NMR (C₆D₅Cl) δ 0.26 (s, 3, ZrMe), 2.22 (s, 12, ArMe), 2.67

(s, 6, PhNMe₂), 3.10 (m, 2, CH₂N), 3.41 (m, 2, CH₂N), 3.95 (m,

2, OCH₂), 4.12 (m, 2, OCH₂), [5.79 (d, 2), 6.03 (br, 3H)]

(PhNMe₂), 7.13 (m, 6, H_aryl).

Gen er a tion of {[1c]Zr Me(P h NMe₂)}[B(C₆F ₅)₄]. Solid

[1c]ZrMe₂(20 mg, 0.037 mmol) was added to a chilled (-30

°C) suspension of [PhNHMe₂][B(C₆F₅)₄] (29 mg, 0.037 mmol)

in C₆D₅Cl. The reaction mixture was stirred for 5 min while

it was warmed to room temperature and then chilled to 0 °C:

¹H NMR (C₆D₅Cl) δ 0.22 (s, 3, ZrMe), 1.16 (d, 6, CHMe₂), 1.24

(d, 6, CHMe₂), 1.35 (d, 6, CHMe₂), 1.45 (d, 6, CHMe₂), 2.80 (s,

P olym er iza tion of 1-Hexen e u sin g [2a ,c]Zr R ₂Com -

4812 Organometallics, Vol. 17, No. 22, 1998

Aizenberg et al.

p lexes. In a typical experiment, 10 mmol of 1-hexene (1.25

mL) was added to a solution of 50 µmol of [2a ]ZrMe₂(0.022 g)

and 50 µmol of [PhNMe₂H][B(C₆F₅)₄] (0.039 g) in chlorobenzene

(13 mL) was added at 0 °C. After 3 h the reaction was

quenched by addition of HCl in diethyl ether (4 mL, 1.0 M).

Most solvent was removed at 15 Torr (water aspirator) at 50

°C, and the viscous oil was dried at approximately 200 mTorr

at 60 °C for 18-20 h.

P olym er iza tion Exp er im en ts. (a ) P r ep a r a tive Sca le

P olym er iza tion s of 1-Hexen e u sin g Zr a n d Hf Ca ta lysts.

To a chilled (-30 °C) solution of [1a -d ]MMe₂(M ) Zr, Hf)

(0.040 mmol) in chlorobenzene (4 mL) was added a solution of

[Ph₃C][B(C₆F₅)₄] (40 mmol) in chlorobenzene (6 mL) at -30

°C, and the reaction mixture was stirred for 5 min. The

reaction mixture was allowed to equilibrate at 0 °C, and an

appropriate amount of 1-hexene was added in one shot. After

the mixture was stirred for 1 h at 0 °C, the reaction was

quenched with HCl (1.0 M in ether, 4 mL). All volatile

components were removed in vacuo (100 mTorr) at 120 °C.

Experiments with [PhNMe₂H][B(C₆F₅)₄] as an initiator were

carried out similarly, but prior to the addition of the monomer

the reaction mixture was stirred for 15 min while being

warmed to room temperature to ensure complete dissolution

of the initiator and its reaction with the metal complex. The

GPC and polymer yield data are compiled in Table 3.

the reaction flask was weighed again to show a negligible

increase in mass (less than 10 mg).

(c) NMR Sca le P olym er iza tion of 1-Hexen e u sin g Ti

Ca ta lysts. To a solution of [1a ]TiMe₂(19 mg, 0.049 mmol)

in 0.6 mL of C₆D₅Br was added at -30 °C a suspension of

[PhNMe₂H][B(C₆F₅)₄] (40 mg, 0.05 mmol) in 0.15 mL of C₆D₅-

Br. The mixture was stirred at room temperature for 15 min,

over which time the color changed from yellow to dark red.

The solution was precooled again to -30 °C, and to it was

added 155 µL (105 mg, 25 equiv) of 1-hexene. The reaction

mixture was transferred to an NMR tube and was kept in an

1

ice bath at 0 °C. Periodic monitoring by H NMR allowed us

to follow the course of the polymerization reaction. After 3.5

h the yield of the polymer was roughly estimated by ¹³C{¹H}

NMR to be ∼75%. After overnight the monomer was still not

fully consumed. The experiment utilizing [1a ]Ti(CH₂Ph)₂was

carried out completely analogously and yielded similar results.

Ack n ow led gm en t. R.R.S. thanks the Department

of Energy (DE-FG02-86ER13564) and Exxon for sup-

porting this research, and M.A. is grateful for a one-

year Rueff-Wormser Postdoctoral Fellowship.

Su p p or tin g In for m a tion Ava ila ble: Figures giving ad-

ditional views and tables giving crystal data and structure

refinement details, atomic coordinates, bond distances and

angles, and thermal parameters for [1a ]Ti(CH₂Ph)₂, [1a ]ZrMe₂,

[1c]HfEt₂, and [2a ]ZrMe₂(23 pages). Ordering information

is given on any current masthead page.

(b) Attem p ted P olym er iza tion of Eth ylen e u sin g {[1a ]-

MMe(Et₂O)}[B(C₆F ₅)₄] (M ) Zr , Hf). A solution of 0.01

mmol of a metal complex in 10 mL of chlorobenzene was

stirred at 0 °C under 1 atm of ethylene for 10 min in a weighed

flask. The reaction was quenched with HCl (1.0 M in ether, 4

mL). After all volatile components were removed in vacuo,

OM980512A

Article Doi

Synthesis of group 4 complexes that contain the tridentate diamido/donor ligands [(ArylNCH2CH2)2O]2- and zirconium complexes that contain [(ArylNCH2CH2)2S]2- and an evaluation of their activity for the polymerization of 1-hexene

DOI: 10.1021/om980512a

Source and publish data:

Authors:

Article abstract of DOI:10.1021/om980512a

Full text of DOI:10.1021/om980512a

Products guided by the article

R&D Labs maybe for 215383-17-8

Relevant to this article

Hot Product