Novel zirconium and hafnium complexes of monoanionic Di-N,N′-chelating pyridyl- and quinolyl-1-azaallyl ligands and their activity in olefin polymerization catalysis

Article abstract of DOI:10.1021/om980850b

The lithium complexes [Li{N(SiMe₃)C(R¹)C(R²)(C₅H ₄N-2)}]₂ (1a, 2a, and 3a) were each treated with MCl₄ to afford the racemic complexes [M{N(SiMe₃)C(R¹)C(R

Full text of DOI:10.1021/om980850b

1444
Organometallics 1999, 18, 1444-1452
Novel Zir con iu m a n d Ha fn iu m Com p lexes of
Mon oa n ion ic Di-N,N′-ch ela tin g P yr id yl- a n d
Qu in olyl-1-a za a llyl Liga n d s a n d Th eir Activity in Olefin
P olym er iza tion Ca ta lysis
Berth-J an Deelman,† Peter B. Hitchcock,† Michael F. Lappert,*,†
Wing-Por Leung,‡ Hung-Kay Lee,‡ and Thomas C. W. Mak‡
The Chemistry Laboratory, University of Sussex, Brighton, BN1 9QJ , U.K., and Department of
Chemistry, The Chinese University of Hong Kong, Shatin, New Territories, Hong Kong
Received October 13, 1998
The lithium complexes [Li{N(SiMe3)C(R1)C(R2)(C5H4N-2)}]2 (1a , 2a , and 3a ) were each
treated with MCl4 to afford the racemic complexes [M{N(SiMe3)C(R1)C(R2)(C5H4N-2)}2Cl2]
(M ) Zr, R1 ) Ph, R2 ) H (1b); M ) Zr, R1 ) But, R2 ) H (2b); M ) Hf, R1 ) But, R2 ) H
(2c); M ) Zr, R1 ) Ph, R2 ) SiMe3 (3b)). Similarly, Li{N(SiMe3)C(Ph)C(R)(C9H6N-2)} (4a
and 5a ) afforded the racemic complexes [Zr{N(SiMe3)C(Ph)C(R)(C9H6N-2)}2Cl2] (R ) H (4b);
R ) SiMe3 (5b)). X-ray structural analysis of 2b, 2c, and 3b revealed that these complexes
have C2 octahedral geometries with their chloride ligands in cis positions. Molecular orbital
calculations on model systems of the bis{3-(2-pyridyl)-1-azaallyl}zirconium system [Zr(LL)2]2+
(LL ) [N(H)C(H)C(H)(2-C5H4N]) demonstrate that (i) the frontier orbitals are similar to
those of [Zr(η5-C5H5)2]2+ and (ii) the bis{3-(2-pyridyl)-1-azaallyl} ligand environment is more
electron-donating, making the zirconium system less electrophilic. Conproportionation of
ZrCl4 with [Zr{N(SiMe3)C(R1)C(R2)(C5H4N-2)}2Cl2] or [Zr{N(SiMe3)C(Ph)C(SiMe3)(C9H6N-
2)}2Cl2] afforded the mono(1-azaallyl)zirconium complexes Zr{N(SiMe3)C(R1)C(R2)(C5H4N-
2)}Cl3 (R1 ) But, R2 ) H (2d ); R1 ) Ph, R2 ) SiMe3 (3d )) and Zr{N(SiMe3)C(Ph)C(SiMe3)-
(C9H6N-2)}Cl3 (5d ), respectively. When activated with methylaluminoxane (MAO), these
compounds were highly active in ethylene polymerization. Compound 3d also showed modest
activity in the polymerization of 1-hexene and the copolymerization of ethylene and 1-hexene.
In tr od u ction
b in Figure 1) on zirconium8 and the observation that
some of the derived [Zr{N,N′-bis(trimethylsilyl)-â-
diketiminato}Cl3] complexes show high activity in cata-
lytic R-olefin polymerization when combined with meth-
ylaluminoxane (MAO) as cocatalyst,9 we designed novel
The continuing development of new spectator ligands
in early transition metal chemistry, especially for group
4 metals, is stimulated by the search for new noncyclo-
pentadienyl catalysts for R-olefin polymerization.1 As a
result, a range of new ligands has become available.
Examples include polydentate Schiff bases,2 benza- and
alkylamidinates,3 multidentate amides,4 macrocyclic
nitrogen ligands,5 porphyrins, porphyrinogens,6 and
biphenoxy7 ligands. However, relatively few of their
derived metal complexes have had catalytic activities
that can match those of the bis(cyclopentadienyl) sys-
tems.
(3) (a) Herskovics-Korine, D.; Eisen, M. S. J . Organomet. Chem.
1995, 503, 307. (b) Flores, J . C.; Chien, J . C. W.; Rausch, M. D.
Organometallics 1995, 14, 1827. (c) Go´mez, R.; Duchateau, R.; Cher-
nega, A. N.; Teuben, J . H.; Edelmann F. T.; Green, M. L. H. J .
Organomet. Chem. 1995, 491, 153. (d) Roesky, H. W.; Meller, B.;
Noltemeyer, M.; Schmidt, H.-G.; Scholz, U.; Sheldrick, G. M. Chem.
Ber. 1988, 121, 1403. (e) Volkis, V.; Shmulinson, M.; Averbuj, C.;
Lisovskii, A.; Edelmann, F. T.; Eisen, M. S. Organometallics 1998, 17,
3155. (f) Littke, A.; Sleiman, N.; Bensimon, C.; Richeson; D. S.; Yap,
G. P. A.; Brown, S. J . Organometallics 1998, 17, 446.
(4) (a) Cloke, F. G. N.; Hitchcock, P. B.; Love, J . B. J . Chem. Soc.,
Dalton Trans. 1995, 25. (b) Kol, M.; Schrock, R. R.; Kempe, R.; Davis,
W. M. J . Am. Chem. Soc. 1994, 116, 4382. (c) Scollard, J . D.;
McConville, D. H.; Rettig, S. J . Organometallics 1997, 16, 1810. (d)
Tsuie, B.; Swenson, D. C.; J ordan, R. J .; Petersen, J . L. Organometallics
1997, 16, 1392. (e) Baumann, R.; Davis, W. M.; Schrock, R. R. J . Am.
Chem. Soc. 1997, 119, 3830. (f) Gibson, V. C.; Kimberley, B. S.; White,
A. J . P.; Williams, D. J .; Howard, P. Chem. Commun. 1998, 313.
(5) (a) Giannini, L.; Solari, E.; Floriani, C.; Chiesi-Villa, A.; Rizzoli,
C. Angew. Chem., Int. Ed. Engl. 1994, 33, 2204. (b) Uhrhammer, R.;
Black, D. G.; Gardner, T. G.; Olsen, J . D.; J ordan, R. F. J . Am. Chem.
Soc. 1993, 115, 8493, and references therein.
Stimulated by the successful introduction of bulky
N,N′-bis(trimethylsilyl)-â-diketiminato ligands (a and
† University of Sussex.
‡ The Chinese University of Hong Kong.
(1) For some recent reviews see: (a) Brintzinger, H. H.; Fischer,
D.; Mu¨lhaupt, R.; Rieger, B.; Waymouth, R. M. Angew. Chem., Int.
Ed. Engl. 1995, 34, 1143. (b) Mo¨hring, P. C.; Coville, N. J . J .
Organomet. Chem. 1994, 479, 1. (c) Horton, A. D. Trends Polym. Sci.
1994, 2, 158. (d) J aniak, C. J . in Metallocenes; Synthesis, Reactions
and Applications; Togni, A., Halteman, R. L., Eds.; Wiley-VCH:
Weinheim, 1998; Vol. 2, Chapter 9. (e) Britovsek, G. J . P.; Gibson, V.
C.; Wass, D. F. Angew. Chem., Int. Ed. Engl. 1999, 38, 429.
(2) (a) Tjaden, E. B.; Swenson, D. C.; J ordan, R. F.; Petersen, J . L.
Organometallics 1995, 14, 371. (b) Corazza, F.; Solari, E.; Floriani,
C.; Chiesi-Villa, A.; Guastini, C. J . Chem. Soc., Dalton Trans. 1990,
1335. (c) Cozzi, P. G.; Floriani, C.; Chiesi-Villa, A.; Rizzoli, C. Inorg.
Chem. 1995, 34, 2921, and references therein.
(6) (a) Solari, E.; Musso, F.; Floriani, C.; Chiesi-Villa, A.; Rizzoli, C.
J . Chem. Soc., Dalton Trans. 1994, 2015. (b) Brand, H.; Arnold, J .
Angew. Chem., Int. Ed. Engl. 1994, 33, 95, and references therein.
(7) van der Linden, A.; Schaverien, C. J .; Meijboom, N.; Ganter, C.;
Orpen, A. G. J . Am. Chem. Soc. 1995, 117, 3008.
(8) Hitchcock, P. B.; Lappert, M. F.; Liu, D.-S. J . Chem. Soc., Chem.
Commun. 1994, 2637.
10.1021/om980850b CCC: $18.00 © 1999 American Chemical Society
Publication on Web 03/19/1999

Novel Zirconium and Hafnium Complexes
Organometallics, Vol. 18, No. 8, 1999 1445
Sch em e 1
F igu r e 1. â-Diketiminato ligands a and b and the newly
developed 2-pyridyl- (c) and 2-quinolyl- (d ) substituted
1-azaallyl ligands.
â-diketiminate-based ligands that possess a nitrogen
heterocycle (pyridine or quinoline) as part of the ligand
backbone (c and d in Figure 1). These monoanionic
N,N′-bidentate ligands, which were readily obtained as
lithium derivatives from lithiated 2-methylpyridine or
2-methylquinoline by a 1,2 insertion of a suitable nitrile
R1CN (R1 ) Ph or But) into the Li-C bond followed by
a 1,3 trimethylsilyl shift,10 are less bulky than the above
N,N′-bis(trimethylsilyl)-â-diketiminates (a and b in
Figure 1) and were expected to have both good σ- and
π-donating properties. Apart from being a variation on
the â-diketiminate theme, we anticipated that, because
of the less sterically demanding nature of the ligands,
formation of bis(ligand)zirconium dichloride derivatives
might be feasible, which would place the zirconium
center into a chiral C2 symmetrical coordination envi-
ronment, forcing the chlorine atoms into cis positions.
In a preliminary communication we reported on some
lithium and zirconium complexes of these pyridyl- and
quinolyl-1-azaallyl ligands11 and noted that the bis-
(ligand)zirconium(IV) chlorides with cis chloride ligands
were indeed chiral, having either C1 or C2 symmetries.
Details on the synthesis and characterization of the
lithium complexes were reported in a previous paper.10
Herein we provide full details on (i) the synthesis and
characterization of the new zirconium(IV) and haf-
nium(IV) bis(1-azaallyl) chlorides and zirconium(IV)
mono(1-azaallyl) chlorides and (ii) their catalytic activity
in olefin polymerization when activated with MAO.
the highly substituted derivative 3b, two sets of 1H
resonances for the ortho-protons of the phenyl groups
(δ ) 7.68 and 7.63, intensity ratio ) 1:1) and two 13C
signals for the CSiMe3 fragment (δ ) 118.6 and 118.3)
were observed, which is attributed to different pucker-
ing modes of the highly substituted ZrNCCCN metalla-
cycles and hindered rotation of the phenyl group (vide
infra). It should be noted that no other isomers were
observed according to the NMR spectra of each of the
purified complexes 1b-5b. The hafnium congener of 2b,
i.e., 2c, was prepared similarly from 2a and HfCl4. The
elemental analytical data for 1b, 2b, and 5b were poor,
believed to be due to occluded LiCl.
The molecular structures of each of the crystalline
complexes 2b, 2c, and 3b were established by X-ray
crystallography and are illustrated in Figures 2, 3, and
4, respectively; selected structural parameters for 2b
and 2c are listed in Table 3. A detailed discussion of
the molecular structure of complex 3b has already been
communicated.11 Important structural features of all
three compounds are (i) the distorted octahedral coor-
dination geometry about the metal center and (ii) the
cis-disposition of the chloride ligands. This cis-arrange-
ment is also observed in zirconocene dichlorides and is
crucial from a catalytic standpoint. In complex 2b the
2-pyridylazaallyl ligand is coordinated in a η1,η3-fashion
with a η3-bonded azaallyl part of the ligand supple-
mented by a σ-bonded pyridyl moiety. For the isoleptic
hafnium complex 2c it was found that, although the
overall structure is rather similar, the long M-C
distances (all above 2.9 Å) indicate that the η3-interac-
tion is no longer present, which may well be the result
of smaller covalent contributions in the ligand-to-metal
bonding for hafnium. The “wrapping” of the ligand
Resu lts a n d Discu ssion
P r ep a r a tion a n d Str u ctu r a l Ch a r a cter iza tion of
th e Com p lexes. Each of the lithium derivatives of the
2-pyridyl- and 2-quinolyl substituted 1-azaallyl ligands
(1a -5a )10 reacted smoothly with ZrCl4 in Et2O or THF
to form racemic mixtures of bis(1-azaallyl)zirconium
dichlorides 1b-5b (Scheme 1). According to their NMR
spectral data (Tables 1 and 2), these bis(1-azaallyl)-
zirconium dichlorides adopt C2 symmetrical geometries
except for 4b, which has no symmetry element (C1). For
(9) Lappert, M. F.; Liu, D.-S. Neth. Pat. Appl. 9500085, 1995.
(10) Deelman, B.-J .; Lappert, M. F.; Lee H.-K.; Mak, T. C. W.; Leung,
W.-P.; Wei, P.-R. Organometallics 1997, 16, 1247.
(11) (a) Deelman, B.-J .; Hitchcock, P. B.; Lappert, M. F.; Lee, H.-
K.; Leung, W.-P. J . Organomet. Chem. 1996, 513, 281. (b) Kristen, M.
O.; Go¨rtz, H. H.; Deelman, B. J .; Lappert, M. F.; Lee, H.-K.; Leung,
W. P. Germ. Pat. Appl. 97105878.9-2109, 1997.

1446 Organometallics, Vol. 18, No. 8, 1999
Deelman et al.
Ta ble 1. 1H NMR Da ta
ates [M{N(SiMe3)C(Ph)N(SiMe3)}2Cl2] (M ) Ti, Zr; Cl-
Ti-Cl ) 98.6(1)°, Cl-Zr-Cl ) 103.8(1)°),3a,d which also
have their chloride ligands cis-dispositioned in distorted
octahedral metal coordination environments.
com-
pound
aryl
CH But SiMe3
1ba,c 9.04 (d, J ) 6.0 Hz, 2H, 6-py)
7.66 (d, J ) 7.2 Hz, 4H, o-Ph)
7.1-7.0 (m, 6H, m-Ph, p-Ph)
6.77 (t, J ) 7.7 Hz, 2H, 4-py)
6.35 (d, J ) 7.9 Hz, 2H, 3-py)
6.30 (t, J ) 6.6 Hz, 2H, 5-py)
2bb,c 10.18-10.16 (d, 1H, 6-py)
6.98-6.92 (dt, 1H, 3-py)
6.05
0.21
The mono(1-azaallyl)zirconium(IV) trichlorides 2d ,
3d , and 5d were readily obtained by conproportionation
of complex 2b, 3b, or 5b with ZrCl4 in either dichloro-
methane or toluene as solvent (eq 1). All three com-
pounds gave satisfactory NMR spectral data (Tables 1
and 2) and for 3d and 5d MS and elemental analyses
data. It is noteworthy that 5d showed two sets of ligand
resonances in both the 1H and 13C NMR spectra,
indicating that this compound is probably dimeric in
solution. It is likely that compounds 2d and 3d are also
dinuclear, but this is not firmly established. Mass
spectra show that compounds 3d and 5d easily form
monomeric molecular ions in the gas phase under EI
conditions.
5.81 0.79
5.74 0.76
5.21 1.13
0.43
0.39
0.24
6.57-6.49 (m, 2H, 4,5-py)
2cb,c 10.14-10.12 (d, 1H, 6-py)
6.92-6.85 (d, 1H, 3-py)
6.49-6.44 (m, 2H, 4,5-py)
2d b,c 10.48 (m, 1H, 6-py)
8.30-8.28 (d, 1H, 3-py)
6.52-6.47 (m, 2H, 4,5-py)
3ba,c 8.58 (d, J ) 4.7 Hz, 2H, 6-py)
7.69 (d, J ) 5.3 Hz, 2H, o-Ph)
7.63 (d, J ) 6.8 Hz, 2H, o-Ph)
7.28 (d, J ) 8.0 Hz, 2H, 3-py)
7.12 (m, 6H, m-Ph, p-Ph)
0.03
0.01
(1)
6.87 (t, J ) 6.9 Hz, 2H, 4-py)
6.00 (t, J ) 6.8 Hz, 2H, 5-py)
3d b,e 8.88 (d, J ) 5.4 Hz, 1H, py)
8.36 (t, J ) 8.1 Hz, 1H, py)
7.79 (d, J ) 8.1 Hz, 1H)
-0.09
-0.28
7.73-7.42 (6H, Ph and py)
4bb,c 9.30 (d br, J ) 8.3 Hz, 1H, 9-qui)
9.04 (m br, 1H, 9-qui)
5.63
5.48
0.36
0.18
Molecu la r Or b it a l Descr ip t ion of t h e Bis(p yr -
id yl-1-a za a llyl)zir con iu m F r a gm en ts. To investigate
the electronic properties of the bis(pyridyl-1-azaallyl)
group 4 metal complexes and to compare them to the
well-known bis(cyclopentadienyl)metal complexes, we
performed simple EHMO calculations on the idealized
C2 symmetrical model systems [Zr(LL)2]2+ (LL ) {N(H)C-
(H)C(H)(C5H4N-2)}). To compare the effect of the dif-
fering coordination modes of the pyridyl-1-azaallyl
ligand in 2b and 3b, both geometries were considered
in the calculations, i.e., structures A and B, respectively
(Figure 6). Throughout this discussion the labeling
scheme of Figure 5 is adopted.
From ligand-to-metal overlap populations it is con-
cluded that the bis(pyridylazaallyl) ligand environment
is more electron-donating than that of bis(cyclopenta-
dienyl) (Table 5). Hence, the metal center in the bis-
(pyridyl-1-azaallyl)metal complexes is less electrophilic
than in the corresponding bis(cyclopentadienyl) systems.
The interaction of the pyridyl-1-azaallyl ligand with the
metal center in A (overlap population ) 1.73) is slightly
stronger than that for B (overlap population ) 1.67),
which corresponds to the higher substituted and less
flexible ligand system of complex 3b. This is not due to
weaker Zr-N overlap populations in B, since in B the
Zr-N(amide) interactions are markedly stronger due to
the π-NfZr overlap with the Zr dyz orbital (overlap
population 0.10). Such a π-contribution is almost absent
in A (overlap population 0.02). However, in A there are
stabilizing interactions with the carbon atoms C2
(overlap population 0.03) and C3 (overlap population
0.09) of the 1-azaallyl ligand. In B, interaction of the
metal center with these atoms is destabilizing, resulting
in a less favorable interaction. In conclusion, the higher
steric congestion in 3b than 2b results in a weaker
interaction of the pyridyl-1-azaallyl ligand with the
metal center for 3b.
8.21 (s br, 2H, qui)
7.69 (d br, J ) 5.1 Hz, 2H, o-Ph)
7.31-6.82 (14H, p-Ph, m-Ph, qui)
6.20 (d, J ) 8.6 Hz, 1H, qui)
5.65 (d, J ) 8.7 Hz, 1H, qui)
5ba,c 8.82 (d, J ) 8.3 Hz, 2H, 9-qui)
8.28 (d, J ) 8.4 Hz, 2H, qui)
7.80 (d, J ) 6.9 Hz, 2H, o-Ph)
7.54 (d, J ) 8.6 Hz, 2H, o-Ph)
7.26-7.17 (m, 8H, p-Ph, m-Ph, qui)
6.77-6.75 (m, 2H, qui)
0.09
0.08
6.60-6.57 (m, 4H, qui)
5d a,e 8.80 (d, J ) 8.7 Hz, 1H, qui)
8.75 (d, J ) 8.4 Hz, 1H, qui)
8.0-7.5 (14H, Ph and qui)
7.40 (t, J ) 7.5 Hz, 1H)
0.26
-0.10
-0.20
-0.31
7.25 (t, J ) 7.7 Hz, 1H)
6.95 (t, J ) 7.3 Hz, 1H)
6.86 (t, J ) 7.6 Hz, 1H)
6.65 (d, J ) 8.4 Hz, 1H, qui)
6.34 (d, J ) 8.6 Hz, 1H, qui)
a
d
360 MHz. b 250 MHz. c In benzene-d6. In THF-d8. e In dichlo-
romethane-d2.
around the metal center in 2b and 2c is markedly
different from the η2-bonding mode observed in the
solid-state structure of 3b (Figure 5). This is most
probably caused by the higher degree of substitution of
the 3-(2-pyridyl)-1-azaallyl, preventing π-coordination
of the azaallyl part of the ligand system. A comparison
of selected structural parameters for the three com-
plexes is presented in Table 4.
The above results indicate that the 2-pyridylazaallyl
ligand can adopt different bonding modes. Depending
on the steric requirements of the substituents on the
ligand backbone and the electronic requirements at the
metal center, the ligand is able to behave as either an
η2- or η1,η3-donor.
Complexes 2b, 2c, and 3b are structurally similar to
(i) the Schiff-base complexes [ZrL2Cl2] (L ) a norephe-
drine-derived ligand, Cl-Zr-Cl ) 93.7(1)°),20 (ii) [Zr-
(msal)2Cl2] (msal ) N-methylsalicylideneiminate, Cl-
Zr-Cl ) 96.9(1)° and 98.8(1)°),2b and (iii) the benzamidin-
The empty frontier orbitals of A and B that can
participate in bonding of two additional σ-donating

Novel Zirconium and Hafnium Complexes
Organometallics, Vol. 18, No. 8, 1999 1447
Ta ble 2. 13C NMR Da ta
compound
aryl
CH or CSiMe3
NCPh or NCBut
CMe3
SiMe3
3.1
1ba,d
154.2, 149.8, 141.2, 138.0, 129.2,
128.8, 127.9, 123.3, 120.2
156.4, 151.4, 138.1, 123.7, 120.9
111.2
111.5
112.3
89.9
155.0
2ba,d
2ca,d
2d a,d
3ba,d
170.6
171.1
161.4
165.0
38.2
30.5
38.5
30.8
35.8
29.2
4.0
4.3
1.3
157.1, 151.2, 138.0, 123.9, 120.5
150.1, 147.3, 135.3, 122.0, 116.3
162.9, 149.8, 143.4, 137.9, 134.4,
130.4, 129.3, 128.6, 128.4, 128.3,
127.9, 127.0, 125.8
151.8, 148.3, 143.0, 142.1, 131.0,
130.7, 129.5, 129.1, 128.9, 127.2,
126.9
155.2, 152.5, 145.8, 145.4, 142.3,
138.4, 138.2, 138.1, 137.3, 130.8,
130.5, 130.2, 127.0, 126.9, 126.9,
126.4, 125.3, 124.9, 121.9, 120.4
163.5, 145.8, 142.8, 138.7, 135.1,
131.5, 130.7, 129.8, 127.3, 126.9,
126.5, 125.3, 124.9, 119.5, 118.4
161.9, 145.2, 144.1, 144.0, 142.6,
141.9, 140.5, 133.1, 132.0, 131.6,
131.2, 130.9, 129.7, 129.3, 129.0,
128.9, 128.5, 127.2, 126.5, 125.0,
124.0, 123.8, 123.3
118.6,
118.3
4.3,
2.3
3d b,f
4ba,g
123.2
160.0
165.1
1.83
0.95
105.3,
91.4
4.3
3.7
5ba,c
104.9
112.9
166.5
163.6
3.9
2.6
5d b,f,h
3.3
3.0
2.8
0.2
a
b
d
g
h
62.9 MHz. 75.5 MHz. c Benzene/benzene-d6. Benzene-d6. e THF-d8. f Dichloromethane-d2. Toluene/toluene-d8. Low signal/noise
ratio due to low solubility.
F igu r e 3. Molecular structure of [Hf{N(SiMe3)C(But)C-
(H)(C5H4N-2)}2Cl2] (2c) showing atom-labeling scheme. The
thermal ellipsoids are drawn at the 20% probability level.
F igu r e 2. Molecular structure of [Zr{N(SiMe3)C(But)C-
(H)(C5H4N-2)}2Cl2] (2b) showing atom-labeling scheme.
The thermal ellipsoids are drawn at the 20% probability
level.
acceptor orbitals for [Zr(LL)2]2+ are at higher energy
than in C, this does not significantly affect the Zr-H
bond (Table 5). Mixing-in of lower lying filled orbitals
is responsible for this effect. Hence, in comparing
different metal-ligand combinations, shape and sym-
metry of the fragment frontier orbitals are more infor-
mative than the exact energies of those orbitals. For a
better comparison, the actual overlap populations with
additional ligands entering the metal coordinaton sphere
should be considered.
Olefin P olym er iza t ion Act ivit y of t h e Com -
p lexes. Each of the bis(1-azaallyl)zirconium(IV) chlo-
rides 1b-5b, combined with a large excess of MAO, was
tested for catalytic activity in ethylene polymerization.
However, only complex 2b was found to display modest
activity.13 In contrast, mono(1-azaallyl)zirconium(IV)
ligands (e.g., H-, Cl-, or alkyl-) are represented in
Figure 6 along with those of [Zr(η5-C5H5)2]2+ (C) for
comparison.12 It should be noted that these relevant
frontier orbitals are similar in symmetry and shape for
all three fragments A, B, and C. However, due to the
strong σ-donation of the pyridyl groups to the dxz orbital
of zirconium, molecular orbital 24b of fragment A is
substantially raised in energy compared to the corre-
sponding orbital (6b2) of the biscyclopentadienylzirco-
nium fragment C. For B this effect is even more
pronounced, and in addition the acceptor orbital of “a”
symmetry (27a) is also significantly raised in energy.
To probe the bonding of additional σ-donors, hypo-
thetical dihydrides [Zr(LL)2H2] were constructed for
both ligand arrangements. From Zr-H overlap popula-
tions it follows that, despite the fact that the metal
(13) 1.4 kg of PE mol-1 Zr h-1 ([Zr] ) 0.70 mM, Al/Zr ) 200, T ) 25
°C, p ) 1 bar).
(12) Lauher, J . W.; Hoffmann, R. J . Am. Chem. Soc. 1976, 98, 1729.

1448 Organometallics, Vol. 18, No. 8, 1999
Deelman et al.
F igu r e 4. Molecular structure of [Zr{N(SiMe3)C(Ph)C-
(SiMe3)(C5H4N-2)}2Cl2] (3b) showing atom-labeling scheme.
The thermal ellipsoids are drawn at the 50% probability
level.
Ta ble 3. Selected Bon d Dista n ces (Å) a n d An gles
(d eg) for [M{N(SiMe3)C(Bu t)C(H)(C5H4N-2)}2Cl2] (M
) Zr (2b); M ) Hf (2c))
2b
2c
Zr(1)-Cl(1)
Zr(1)-Cl(2)
Zr(1)-N(1)
Zr(1)-N(2)
Zr(1)-N(3)
Zr(1)-N(4)
Zr(1)-C(6)
Zr(1)-C(7)
Zr(1)-C(20)
Zr(1)-C(21)
N(1)-C(5)
N(2)-C(7)
N(3)-C(19)
N(4)-C(21)
C(5)-C(6)
2.5015(12) Hf(1)-Cl(1)
2.5252(12) Hf(1)-Cl(2)
2.429(1)
2.419(1)
2.335(1)
2.142(1)
2.326(1)
2.158(1)
3.172(2)
2.909(2)
3.199(2)
2.922(2)
1.325(1)
1.397(1)
1.343(2)
1.391(2)
1.461(1)
1.353(2)
1.455(1)
1.326(1)
83.4(1)
2.322(3)
2.191(3)
2.311(3)
2.231(3)
2.756(4)
2.655(4)
2.708(4)
2.678(4)
1.349(5)
1.359(5)
1.337(5)
1.357(5)
1.439(5)
1.396(5)
1.457(5)
1.384(5)
81.38(4)
128.57(11)
178.14(12)
Hf(1)-N(1)
Hf(1)-N(2)
Hf(1)-N(3)
Hf(1)-N(4)
Hf(1)-C(6)
Hf(1)-C(7)
Hf(1)-C(20)
Hf(1)-C(21)
N(1)-C(5)
N(2)-C(7)
N(3)-C(19)
N(4)-C(21)
C(5)-C(6)
F igu r e 5. Comparison of the two bonding modes of the
3-(2-pyridyl)-1-azaallyl ligands in [M{N(SiMe3)C(R1)C(R2)-
(C5H4N-2)}2Cl2] complexes (M ) Zr, R1 ) But, R2 ) H (2b);
M ) Hf, R1 ) But, R2 ) H (2c); R1 ) Ph, R2 ) SiMe3 (3b)).
Hydrogen atoms and ligand substituents have been omitted
for clarity. For better comparison the structural drawing
corresponding to 2b and 2c (top) is drawn in its opposite
enantiomeric representation.
Ta ble 4. Com p a r ison of Selected Bon d Dista n ces
a n d An gles in Com p lexes 2b, 2c, a n d 3b Accor d in g
to La belin g Sch em e of F igu r e 5
C(6)-C(7)
C(6)-C(7)
C(19)-C(20)
C(20)-C(21)
Cl(1)-Zr-Cl(2)
N(1)-Zr-N(3)
N(2)-Zr-N(4)
C(19)-C(20)
C(20)-C(21)
Cl(1)-Hf-Cl(2)
N(1)-Hf-N(3)
N(2)-Hf-N(4)
bond distance (Å)
or angle (deg)
2b
(M ) Zr)
2c
3b
(M ) Zr)
(M ) Hf)
115.0(1)
157.5(1)
M-Cl(1)
2.5252(12)
2.5015(12)
2.191(3)
2.322(3)
2.231(3)
2.311(3)
2.655(4)
2.756(4)
2.678(4)
2.708(4)
81.38(4)
178.14(12)
128.57(11)
2.429(1)
2.419(1)
2.142(1)
2.335(1)
2.158(1)
2.326(1)
2.909(2)
3.172(2)
2.922(2)
3.199(2)
83.4(1)
2.434(1)
2.434(1)
2.141(3)
2.354(3)
2.141(3)
2.354(3)
2.868(4)
3.277(4)
2.868(4)
3.277(4)
95.06(7)
165.0(2)
76.7(2)
M-Cl(2)
M-N(1)
chlorides 2d , 3d , and 5d displayed activities in the
range 6-75 kg of PE mol-1 Zr h-1 with the activity order
5d < 3d < 2d ([Zr] ) 1.5 mM, Al/Zr ) 100, T ) 25 °C,
p ) 1 bar). Under these conditions, 3d (20 kg mol-1 Zr
h-1) showed activity similar to [Zr(η5-C5H5)2Cl2]/MAO
(38 kg mol-1 Zr h-1), which is known to be one of the
most active systems in ethylene polymerization.1b The
activities found should be regarded as lower limits to
the intrinsic activity of these complexes since we did
not make attempts to optimize [Zr] and Al/Zr ratios.
Also for the most active catalysts (2d and [Cp2ZrCl2])
mass transport limitation cannot be excluded. Although
GPC analysis of the PE produced was hindered by the
extremely low solubility of the polymer samples in 1,2-
dichlorobenzene even at high temperature, intrinsic
viscosity ([η]) measurements were indicative of high
molecular weight.14
M-N(2)
M-N(3)
M-N(4)
M-C(1)
M-C(2)
M-C(4)
M-C(5)
Cl(1)-M-Cl(2)
N(1)-M-N(3)
N(2)-M-N(4)
157.5(1)
115.0(1)
ization (0.133 g mol-1 Zr h-1) was observed, which is
probably a concentration effect.15 End-group analysis of
the oligohexenes formed (1H NMR) indicated that the
material consisted of substantial amounts (68%) of
oligomers terminated by internal double bonds (δ )
5.37-5.34); the signals observed at δ ) 133.4-125.3 in
the 13C NMR spectrum are also typical for internal
double bonds.7 In addition, vinylidene (H2CdC(R)R′,
27%) and allyl (H2CdCHCH2R, 5%) terminated oligo-
Complex 3d was tested in the polymerization of
propylene and styrene but was found to be inactive
under ambient conditions. For 1-hexene some oligomer-
(15) The solubility of propene in toluene under the conditions
employed is only 0.9 M (Plo¨cker, U.; Knapp, H.; Prausnitz, J . Ind. Eng.
Chem. Process Des. Rev. 1978, 17 (3), 324), whereas the concentration
of 1-hexene was 5.1 M.
(14) Intrinsic viscosity values ([η]) of 9.1 (3d ) and 7.3 dL/g (5d ) were
measured.

Novel Zirconium and Hafnium Complexes
Organometallics, Vol. 18, No. 8, 1999 1449
elimination, whereas allylic end-groups most probably
originate from final ethylene insertions followed by â-H
elimination. However, the presence of RCHdCHR′ end-
groups in both the oligohexenes and the 1-hexene/
ethylene copolymer implies that a favorable termination
pathway involves secondary 2,1-insertion of 1-hexene
followed by â-hydrogen elimination. Brintzinger et al.16
have shown that for metallocene/MAO systems these
2,1-misinsertions result in the formation of dormant
sites for which further insertion of monomer is sterically
substantially hindered. The formation of dormant sites
resulting from 2,1-insertion of olefin could also account
for the inactivity of 3d in propene polymerization and
the low activity of 3d in 1-hexene polymerization.
Another noteworthy observation is the presence of allyl
end-groups in the oligohexenes, possibly due to chain
termination by â-alkyl elimination.17
The low activity of the bis(1-azaallyl) complexes in
ethylene polymerization cannot be explained a priori by
pure steric considerations. Molecular mechanics studies
indicated that there is enough space for a coordinating
alkyl chain (R-) and an ethylene molecule to occupy cis-
positions in the coordination gap of a [Zr(LL)2]2+ frag-
ment (LL ) 3-(2-pyridyl)-1-azaallyl ligand). The result-
ing hypothetical [Zr(LL)2(R)(CH2CH2)]+ species would
be the 1-azaallyl analogue of the presumed active
species in metallocene-based Ziegler-Natta polymeri-
zation, i.e., [Zr(η5-C5H5)2(R)(CH2CH2)]+.1
F igu r e 6. Energy level diagram of frontier acceptor
orbitals of [Zr(LL)2]2+ and [Zr(η5-C5H5)2]2+ fragments.
One feasible explanation for the inactivity of the bis-
(1-azaallyl)zirconium complexes in ethylene polymeri-
zation might be that the [Zr(LL)2]2+ fragment, unlike
its cyclopentadienyl counterpart, possesses empty low
lying pyridyl (or quinolyl) π* levels, which renders the
1-azaallyl ligands susceptible to nucleophilic attack by
incoming alkyl nucleophiles. This hypothesis is sup-
ported by the reaction of 5b with PhCH2MgCl (2 equiv
in Et2O), which resulted in nucleophilic attack by the
benzyl group at the 4-position of the quinoline function-
Ta ble 5. Over la p P op u la tion s Ca lcu la ted for
Mod el F r a gm en ts [Zr (LL)2]2+ (LL )
N(H)C(H)C(H)(C5H4N-2)) a n d [Zr Cp 2]2+ (Cp )
η5-C5H5)
[Zr(LL)2]2+ [Zr(LL)2]2+
overlap population
(A)
(B)
[ZrCp2]2+
1.42
Zr-{(LL)2} or Zr-{Cp2}
Zr-{N(amido)}
Zr-{N(py)}
Zr-H (in Zr(LL)2H2 or
Zr(Cp)2H2)
1.73
0.49
0.41
0.61
1.67
0.59
0.43
0.62
1
ality to form 5e (eq 2). The H NMR spectrum of this
0.62
compound displayed a typical ABX pattern for the C(H)-
(CH2Ph) moiety and a doublet (δ ) 4.84) which was
assigned to the NCdC(H)-CH fragment. The 13C NMR
spectrum displayed a characteristic C(H)(CH2Ph) reso-
nance (δ ) 49.4). Consequently, it seems feasible that
alkylation by a large excess of MAO, which can also
contain substantial amounts of AlMe3, could lead to
nucleophilic attack on one (or both) of the 1-azaallyl
ligands, producing an inactive complex.
mers were observed (1H NMR: δ ) 5.80 (m, 1H,
CH2dC(H)CH2R), 5.01-4.90 (m, 2H, CH2dC(H)CH2R)
and 4.72-4.56 (4 × s, CH2dC(R)R′)).17 A value of Mn )
950 was deduced from the relative intensities of end-
group signals and aliphatic hydrogen signals.
Complex 3d was also found to be active in copoly-
merization of ethylene and 1-hexene (2.7 kg of polymer
mol-1 Zr h-1). End-group analysis of the toluene-soluble
fraction again indicated the presence of internal double
bonds (59%), vinylidene groups (26%), and allyl groups
(15%). From relative intensities of the methyl region and
the methylene/methine region of the 1H NMR spectrum,
the ratio of 1-hexene to ethylene units in the toluene-
soluble fraction of the copolymer was calculated to be
ca. 1.5. Relative intensities of the aliphatic region and
the end-group resonances indicated that Mn ) 1900.
Vinylidene end-groups in the copolymer are attributed
to a primary 1,2-insertion of 1-hexene followed by â-H
Although we have not investigated this option in
detail we do not think that ligand transfer to aluminum
could explain the low activity of the bis(pyridyl-1-
azaallyl)zirconium complexes since such a process would,
at least in the first step, lead to formation of mono-
(pyridyl-1-azallyl)zirconium species which are expected
to be very similar in nature to the species resulting from
reaction of the mono(pyridyl-1-azaallyl)zirconium tri-
chlorides with MAO. However, for the mono(pyridyl-1-
azaallyl)zirconium trichlorides themselves such a proc-
ess remains a feasible option. Since our primary interest
in these group 4 metal complexes was to investigate
whether they could serve as catalysts for stereoregular
polymerization of propene and higher 1-alkenes, we did
not investigate this aspect any further once it had
(16) (a) Stehling, U.; Diebold, J .; Kirsten, R., Ro¨ll, W.; Brintzinger,
H.-H.; J u¨ngling, S.; Mu¨lhaupt, R.; Langhauser, F. Organometallics
1994, 13, 964. (b) J u¨ngling, S.; Mu¨lhaupt, R.; Stehling, U.; Brintzinger,
H.-H.; Fischer, D.; Langhauser, F. Macromol. Symp. 1995, 97, 205.
(17) Resconi, L.; Piemontesi, F.; Franciscono, G.; Abis, L.; Fiorani,
T. J . Am. Chem. Soc. 1992, 114, 1025.

1450 Organometallics, Vol. 18, No. 8, 1999
Deelman et al.
tions were performed using the Hyperchem software package18
with the standard MM+ parameter set.
become clear to us that none of the complexes prepared
displayed the desired property.
[Zr {N(SiMe3)C(P h )C(H)(C5H4N-2)}2Cl2] (1b). A solution
of 1a (2.38, 4.34 mmol) in THF (12 mL) was added to a solution
of ZrCl4 (0.99 g, 4.25 mmol) in THF (60 mL) at 0 °C. The
reaction mixture was allowed to reach room temperature,
stirred for 15 h, and finally refluxed during 1.5 h, resulting in
an orange-red solution. Volatiles were removed in a vacuum,
and the residue was stripped with hexane (10 mL) and
extracted with Et2O (50 mL) and CH2Cl2 (2 × 10 mL). The
combined extracts were concentrated to ca. 50 mL, and cooling
to -30 °C afforded yellow crystals, which were washed with
hexane (5 mL). Yield: 1.58 g (2.27 mmol, 53%). Anal. Calcd
for C32H38N4Cl2Si2Zr: C, 55.1; H, 5.50; N, 8.04. Found: C, 53.1;
H, 5.42; N, 7.42. MS (EI, 70 eV): m/z 695 (31, (M - 1)+), 681 (9,
(M - Me)+), 659 (9, (M - Cl)+), 623 (6, (M - 2Cl)+).
(2)
Con clu sion s
[Zr {N(SiMe3)C(Bu t)C(H)(C5H4N-2)}2Cl2] (2b). A solution
of 1b (1.74 g, 3.42 mmol) in Et2O (40 mL) was added to a
stirring suspension of ZrCl4 (0.80 g, 3.43 mmol) in Et2O (10
mL) at 0 °C. The yellow reaction mixture was allowed to reach
room temperature and stirred for 20 h, and then volatiles were
removed in a vacuum. The residue was extracted with toluene
and filtered through Celite. Concentration to ca. 5 mL afforded
bright yellow crystals of the product. Yield: 0.63 g (0.96 mmol,
28%). Anal. Calcd for C28H46N4Cl2Si2Zr: C, 51.2; H, 7.06; N,
8.53. Found: C, 50.4; H, 6.93; N, 6.93.
[Hf{N(SiMe3)C(Bu t)C(H)(C5H4N-2)}2Cl2] (2c). A proce-
dure similar to that used for 2b was used, starting from HfCl4
(0.38 g, 1.2 mmol). Yield: 0.44 g (0.59 mmol, 49%). Anal. Calcd
for C28H46N4Cl2Si2Hf: C, 45.2; H, 6.23; N, 7.53. Found: C, 45.1;
H, 6.09; N, 6.97.
We have shown that the newly developed pyridyl- and
quinolyl-susbtituted 1-azaallyl ligands can be success-
fully introduced on group 4 metals, stabilizing these
highly electrophilic metal centers and providing an easy
access to racemic mixtures of nonsymmetrical and C2
symmetrical complexes. The absence of meso forms is a
major synthetic advantage when compared to the syn-
thesis of chiral ansa-metallocene derivatives. The 1-aza-
allyl ligands are highly electron-donating, and the metal
frontier orbitals of bis(1-azaallyl)M species are compa-
rable in shape to those of the M(η5-C5H5)2 fragment of
bent metallocenes. Sterically these novel ligands are
very demanding and perhaps even more so than bulky
cyclopentadienyls.
[Zr {N(SiMe3)C(P h )C(SiMe3)(C5H4N-2)}2Cl2] (3b). Com-
pound 3a (3.86 g, 5.70 mmol) was added to ZrCl4 (1.33 g, 5.71
mmol) in Et2O (60 mL). The reaction mixture was stirred for
16 h, resulting in a yellow suspension. Volatiles were removed
in a vacuum, and the residue was extracted with CH2Cl2 (2 ×
50 mL) and filtered twice. Volatiles were removed from the
filtrate in a vacuum, and crystallization from toluene (50 mL)
at -30 °C afforded yellow crystals of 3b (1.32 g, 1.57 mmol,
28%). Concentration of the mother liquor gave another 1.13
g. Total yield of 3b: 2.45 g (3.1 mmol, 54%). Anal. Calcd for
C38H54Cl2N4Si4Zr: C, 54.2; H, 6.47; N, 6.66. Found: C, 54.2;
H, 6.38; N, 6.56. MS (EI, 70 eV): m/z 840 (15, M+), 825 (5, (M
- Me)+), 805 (11, (M - Cl)+).
X-r a y Str u ctu r e Deter m in a tion s of [M{N(SiMe3)C-
(Bu t)C(H)(C5H4N-2)}2Cl2] (M ) Zr (2b); M ) Hf (2c)) a n d
[Zr {N(SiMe3)C(P h )C(SiMe3)(C5H4N-2)}2Cl2] (3b). Selected
single crystals of 2b, 2c, and 3b were sealed in Lindemann
glass capillaries under dinitrogen. For 2b raw intensities were
collected in the variable ω-scan mode on a four-circle diffrac-
tometer (Siemens P4 or Rigaku AFC7R) using Mo KR radiation
(λ ) 0.710 73 Å) at 294 K. Determination of the crystal class
and orientation matrix was performed according to established
procedures. Unit-cell parameters were calculated from least-
squares fitting of 2θ angles for 20 (or 25) selected reflections.
Crystal stability was monitored by recording two (or three)
check reflections at intervals of 100/150 data measurements,
and no significant variation was detected. The raw data were
processed with a learn-profile procedure,19 and empirical
absorption corrections were applied by fitting a pseusoellipsoid
to the ψ-scan data of selected strong reflections over a range
of 2θ angles. For 2c, the intensity data were collected at 294
K on a MSC/Rigaku RAXIS-IIC imaging plate diffractometer
using Mo KR radiation (λ ) 0.710 73 Å) from a Rigaku RU-
200 rotating-anode generator operating at 50 kV and 90 mA
(2θmin ) 2°, 2θmin ) 55° oscillation frames in the range 0-180°,
exposure 8 min per frame). A self-consistent semiempirical
absorption correction based on Fourier coefficient fitting of
Despite the electronic analogy between them, the bis-
(1-azaallyl)zirconium dichloride complexes are inactive
in ethylene polymerization, whereas zirconocene dichlo-
rides belong to the most active precatalysts known to
date. One feasible explanation for this inactivity is
nucleophilic alkylation of the 1-azaallyl ligands. Mono-
(1-azaallyl)zirconium trichlorides on the other hand
were found to be highly active precatalysts with activi-
ties of the same order of magnitude as observed for
[Zr(η5-C5H5)2Cl2]. Activity is mainly limited to ethylene
however. Although modest activity in copolymerization
of ethylene with 1-hexene was observed, homopoly-
merization of higher olefins was much slower (1-hexene)
or not taking place at all (propylene and styrene). For
propene and 1-hexene this is presumably due to facile
formation of dormant sites by 2,1-misinsertions, which
is most likely the result of the open coordination sphere
of the zirconium center in the catalytically active
species.
Exp er im en ta l Section
Gen er a l Con sid er a tion s. All experiments were performed
under argon using standard Schlenk and vacuum line tech-
niques. Pentane, Et2O, THF, THF-d8, cyclohexane, benzene,
benzene-d6, toluene, and toluene-d8 were distilled from Na/K
alloy and degassed prior to use. Polymerization grade ethylene
and propylene (BOC) were purified by passing through BASF
Cu catalyst and molecular sieves (4 Å) columns. MAO was
obtained from Witco as a 10 wt % solution in toluene (density
≈ 0.9 g/mL). Other compounds were purchased and purified
by standard procedures. NMR spectra were recorded on Bruker
WM250, WM300, and WM360 spectrometers at ambient
temperatures. MS spectra (EI) were recorded on a VG Auto-
spec intrument. Elemental analyses were carried out by Medac
Ltd. (Brunel University). Viscosity measurements were per-
formed at BASF, Ludwigshaven. Molecular mechanics calcula-
(18) HyperChem, release 3 for Windows 3.1; Hypercube Inc.
(19) R. Diamond Acta Crystallogr., Sect. A 1968, 25, 43.

Novel Zirconium and Hafnium Complexes
Organometallics, Vol. 18, No. 8, 1999 1451
Ta ble 6. Deta ils of th e X-r a y Str u ctu r e Deter m in a tion s of [M{N(SiMe3)C(R1)C(R2)(C5H4N-2)}2Cl2] (M ) Zr ,
R1 ) Bu t, R2 ) H (2b); M ) Hf, R1 ) Bu t, R2 ) H (2c); R1 ) P h , R2 ) SiMe3 (3b))
empirical formula
fw
C28H46Cl2N4Si2Zr (2b)
657
C28H46Cl2N4Si2Hf (2c)
744.3
C39H56Cl4N4Si4Zr (3b)
926.3
cryst size (mm)
cryst syst
space group
a (Å)
b (Å)
c (Å)
0.20 × 0.20 × 0.70
orthorhombic
Pbca (No. 61)
10.219(2)
15.543(3)
41.308(8)
90
90
90
6561(3)
1.330
0.12 × 0.22 × 0.24
0.50× 0.30× 0.25
monoclinic
P2/n (No. 13)
15.779(4)
9.331(4)
17.488(3)
90
112.46(2)
90
monoclinic
P21/c (No. 14)
21.407(4)
9.999(2)
15.823(3)
R (deg)
90
â (deg)
104.17(3)
90
3283.8(2)
1.505
γ (deg)
V (Å3)
2379.5(13)
1.29
2
Dcalcd (mg/m3)
Z
8
4
F(000)
2752
0.594
1504
3.436
964
0.59
abs coeff (mm-1
)
θ range for data collcn (deg)
2.58-27.5
0 e h e 13, 0 e k e 20,
0 e l e 53
7504
2-55
2-25
index ranges
-27 e h e 26, -11 e k e 11,
0 e l e 19
0 e h e 18, 0 e k e 11,
-20 e l e 19
4329
no. of reflns collcd
6787
no. of indept reflns
no. of reflns with I > 2σ(I)
structure solution
7504
3636 [I > 4σ(I)]
direct methods
6526 [R(int) ) 0.038]
5616 [I > 8σ(I)]
direct methods
4176 [R(int) ) 0.0229]
2914 [I > 2σ(I)]
direct methods
refinement method
no. of params refined
goodness-of fit
full-matrix least-squares on F2 full-matrix least-squares on F2 full-matrix least-squares on all F2
334
1.39
335
1.77
240
1.078 (on F2)
R1 ) 0.050, wR2 ) 0.123
R1 ) 0.081, wR2 ) 0.140
0.78, -0.50
final R indices (obs data)a
R indices (all data)
R1 ) 0.054, wR2 ) 0.1315
R1 ) 0.1511, wR2 ) 0.1621
R1 ) 0.044, wR ) 0.065
largest diff peak and hole, e A-3 0.50, -0.43
1.1, -0.98
abs corr from ψ scans
Tmax ) 1.00
Tmin ) 0.985
0.002
Tmax ) 1.00
Tmin ) 0.96
maximum shift/esd
0.002
0.002
a
2
2
2
2
R1 ) ∑||F0| - |Fc||/∑|F0|, wR ) [∑w(|F0| - |Fc|)2/∑w|F0| ]-1, wR2 ) [∑w(|F0| - |Fc| )2/∑w(|F0| )2]-1
.
symmetry-equivalent reflections was applied by using the
ABSCOR program.20 The crystal structures of 2b and 2c were
determined by the direct method, which yielded the positions
of all non-hydrogen atoms, and all non-hydrogen atoms were
subjected to anistropic refinement. The hydrogen atoms were
generated geometrically (C-H bonds fixed at 0.96 Å and
allowed to ride on their respective parent atoms); they were
assigned appropriate isotropic temperature factors and in-
cluded in the structure factor calculations. All computation
were performed on an IBM-compatible PC with the Siemens
SHELXTL PLUS program package.21 Analytical expressions
of neutral-atom scattering factors were employed, and an-
omalous dispersion corrections were incorporated. For 3b, data
were collected on an Enraf-Nonius CAD-4 diffractometer using
monochromated Mo KR (λ ) 0.710 73 Å) radiation and cor-
rected for Lorentz, polarization, and absorption effects. The
structure was refined on F2 using SHELXL-93.22 Hydrogen
atoms were included in riding mode. There is a molecule of
CH2Cl2 solvate for which the C atom could not be located.
Crystallographic data and further experimental details of the
structure determinations for 2b, 2c, and 3b are listed in Table
6.
[Zr {N(SiMe3)C(P h )C(H)(C9H6N-2)}2Cl2] (4b). To a stirred
suspension of ZrCl4 (1.15 g, 4.94 mmol) in Et2O (100 mL) was
added 4a (3.13 g, 9.65 mmol), resulting in a yellow suspension.
The reaction mixture was stirred for 15 h, and volatiles were
removed in a vacuum. The residue was extracted with CH2Cl2
(1 × 30 mL, 1 × 10 mL). The filtrate was evaporated to dryness
and suspended in hexane (20 mL). The solid was collected on
a frit and washed with hexane (20 mL), affording a yellow
powder (3.70 g). Recrystallization from CH2Cl2/hexane and hot
toluene gave yellow needles of 4b (2.76 g, 3.10 mmol, 65%).
Anal. Calcd for C40H42Cl2N4Si2Zr(C7H8): C, 63.5; H, 5.67; N,
6.30. Found: C, 63.9; H, 5.67; N, 6.22. MS (EI, 70 eV): m/z
686 (26, (M - SiMe3Cl)+).
[Zr {N(SiMe3)C(P h )C(SiMe3)(C9H6N-2)}2Cl2] (5b). To a
stirred suspension of ZrCl4 (0.50 g, 2.2 mmol) in Et2O (50 mL)
was added 5a (1.70 g, 4.3 mmol), resulting in an orange
suspension. The reaction mixture was stirred for 3 days, and
volatiles were removed in a vacuum. The residue was extracted
with warm toluene (45 mL). The filtrate was evaporated to
dryness and the residue suspended in hexane (25 mL). The
orange solid was collected and washed with hexane (20 mL).
Recrystallization from CH2Cl2/Et2O (1:1) afforded the orange
microcrystalline solid 5b (0.78 g, 0.83 mmol, 37%). Anal. Calcd
for C46H58Cl2N4Si4Zr: C, 58.7; H, 6.21; N, 5.95. Found: C, 56.8;
H, 6.10; N, 4.84. MS (EI, 70 eV): m/z 834 (3, ((M - SiMe3-
Cl)+). In another experiment starting from 5a (12.7 g, 32.0
mmol), 8.14 g (8.65 mmol, 54%) of 5b was obtained. Extraction
was carried out with CH2Cl2 (200 mL) instead of toluene.
[Zr {N(SiMe3)C(Bu t)C(H)(C4H5N-2)}Cl3] (2d ). ZrCl4 (0.11
g, 0.49 mmol) was added to a solution of 2b (0.32 g, 0.49 mmol)
in toluene (15 mL). After stirring for 10 h at 80 °C, volatiles
were removed in a vacuum. Extraction with CH2Cl2, evapora-
tion of the solvent from the extract, and washing with hot
hexane (20 mL) afforded the orange solid 2d (0.41 g, 95%).
[Zr {N(SiMe3)C(P h )C(SiMe3)(C5H4N-2)}Cl3] (3d ). To a
suspension of ZrCl4 (0.63 g, 2.70 mmol) in toluene (35 mL) was
added 3b (2.16 g, 2.57 mmol). After stirring for 16 h a yellow
oil had formed from which volatiles were removed in a vacuum.
The residue was stripped and washed with hexane (20 mL),
resulting in a yellow powder identified as 3d . Yield: 1.73 g
(3.22 mmol, 63%). MS (EI, 70 eV): m/z 536 (13, M+), 521 (39,
(M - Me)+), 501 (45, (M - Cl)+). Recrystallization from Et2O/
CH2Cl2 (1:1) afforded the mono ether adduct in essentially
quantitative yield (1H NMR). Anal. Calcd for C23H37N2Si2-
ZrCl3O: C, 45.2; H, 6.10; N, 4.58. Found: C, 45.3; H, 6.00; N,
4.91. A better yield (2.95 g, 89%) was obtained when using
CH2Cl2 as solvent.
(20) Higashi, T. ABSCOR-An Empirical Absorption Correction
Based on Fourier Coefficient Fitting; Rigaku Corporation: Tokyo, 1995.
(21) Sheldrick G. M. SHELXL PC Manual; Siemens Analytical X-ray
Instruments, Inc.: Madison, WI, 1990.
(22) Sheldrick, G. M. SHELXL-93, A program for crystal structure
refinement; University of Go¨ttingen, 1993.

1452 Organometallics, Vol. 18, No. 8, 1999
Deelman et al.
[Zr {N(SiMe3)C(P h )C(SiMe3)(C9H6N-2)}Cl3] (5d ). To a
suspension of ZrCl4 (0.39 g, 1.67 mmol) in CH2Cl2 (25 mL) was
added 5b (1.53, 1.63 mmol). After stirring for 16 h the orange
mixture was filtered and the filtrate concentrated to ca. 10
mL. Cooling to -30 °C afforded yellow crystals of 5d (0.81 g).
Concentration of the mother liquor gave a second crop. Total
yield: 1.10 g (1.87 mmol, 57%). Anal. Calcd for C23H29N2Si2-
ZrCl3: C, 47.0; H, 4.98; N, 4.77. Found: C, 46.8; H, 4.86; N,
4.86. MS (EI, 70 eV): m/z 586 (16, M+), 571 (34, (M - Me)+),
551 (8, (M - Cl)+).
[Zr {N(SiMe3)C(P h )C(SiMe3)(C9H6N-2)}{N(SiMe3)C(P h )C-
(SiMe3)(P h CH2-4-C9H6N-2)}Cl] (5e). To a solution of 5b (0.50
g, 0.53 mmol) in Et2O (30 mL) at -70 °C was added PhCH2-
MgCl (0.95 mL, 0.52 mmol of a 0.547 M solution in Et2O). The
reaction mixture was allowed to reach room temperature; a
further portion of the Grignard solution (1.0 mL, 0.55 mmol)
was added. The mixture was stirred for another 16 h. Volatiles
were removed in a vacuum, and the residue was extracted with
hexane (2 × 35 mL). Crystallization at -30 °C afforded 5e
(0.26 g, 0.26 mmol, 50%) as an orange material. 1H NMR (360
MHz, benzene-d6): δ 8.65 (d, J ) 8.3 Hz, 1H, 9-qui), 8.03 (d,
J ) 6.6 Hz, 1H, 9-qui), 7.57 (m, 2H), 7.33-7.10 (m and C6D5H),
6.89 (t, J ) 7.3 Hz, 1H), 6.49 (d, J ) 7.8 Hz, 1H), 6.38 (t, J )
7.3 Hz, 1H), 6.06 (t, J ) 6.8 Hz, 1H), 5.83 (d, J ) 8.1 Hz, 1H),
4.84 (d, J ) 5.0 Hz, 1H, NCdC(H)), 3.99 (X part of ABX
pattern, 1H, C(H)(CH2Ph)), 3.01 (AB part of ABX pattern, J AB
) 13 Hz, 2H, C(H)(CH2Ph)), 0.42 (s, SiMe3), 0.25 (s, SiMe3),
-0.03 (s, SiMe3), -0.12 (SiMe3). 13C NMR (75.4 MHz, benzene/
benzene-d6): δ 49.4 (C(H)(CH2Ph), 41.9 (C(H)(CH2Ph), 3.9
(SiMe3), 2.7 (SiMe3), 2.5 (SiMe3), 1.1 (SiMe3). Anal. Calcd for
isolated, washed succesively with HCl (1 M), water, and
MeOH, and dried at 80 °C to constant weight (usually 24 h).
1-Hexen e P olym er iza tion . To 3d (32.0 mg, 59.5 µmol),
dissolved in a solution of MAO (10% (w/w) in toluene, 5.60 mL,
8.7 mmol of Al), was added 1-hexene (10.0 mL, 80.0 mmol).
The red solution, which turned brown slowly, was stirred for
25 h and then quenched with MeOH/HCl (1 M). The water
layer was separated and extracted with toluene (35 mL). The
combined toluene layers were washed with water (2 × 25 mL)
and dried (MgSO4). Volatiles were removed in a vacuum (40
°C, 0.001 mbar), affording a yellow oil. Yield: 0.20 g. 1H NMR
end-group analysis (300 MHz, chloroform-d1): δ 5.80 (m, 1H,
CH2dC(H)CH2R), 5.37-5.34 (RC(H)dC(H)R′), 5.01-4.90 (m,
2H, CH2dC(H)CH2R), 4.72-4.56 (4 × s, CH2dC(R)R′). 13C
NMR (75.4 MHz, chloroform-d1): δ 133.4-125.3 (internal
double bond end-groups), 40.21, 34.59, 32.35, 28.70, 23.23,
14.19; resonances due to vinylidene and allyl end-groups not
observed due to low intensity.
Eth ylen e/1-Hexen e Cop olym er iza tion . 1-Hexene (10.0
mL, 80.0 mmol) dissolved in a solution of MAO ((10% (w/w) in
toluene, 3.0 mL, 4.7 mmol of Al) was added to a mixture of 3d
(4.6 mg, 8.5 µmol) and MAO (10% (w/w) in toluene, 0.50 mL,
0.76 mmol of Al) and then pressurized with ethylene (p ) 1.5
bar). After ca.1 h the polymerization reaction was quenched
with a 10:1 mixture of MeOH in aqueous HCl (1 M), and solid
polymer was isolated and washed with successively HCl (1 M),
water, and MeOH. Drying (80 °C, 24 h) afforded the polymer
(23 mg). The toluene layer was washed with water (2 × 10
mL) and dried (MgSO4). Volatiles were removed in a vacuum
(40 °C, 0.001 mbar), affording a colorless wax (10 mg). 1H NMR
end-group analysis of the toluene-soluble fraction (300 MHz,
chloroform-d1): δ 5.70 (m, 1H, CH2dC(H)CH2R), 5.32-5.28
(RC(H)dC(H)R′), 4.95-4.84 (m, 2H, CH2dC(H)CH2R), 4.66-
4.59 (3 × s, CH2dC(R)R′).
C
53H65ClN4Si4Zr: C, 63.8; H, 6.57; N, 5.62. Found: C, 64.0;
H, 6.68; N, 5.74. MS (EI, 70 eV): m/z 995 (0.9, M+), 960 (0.2,
(M - Cl)+), 905 (13, (M - CH2Ph)+), 481 (37, [{N(SiMe3)C-
(Ph)C(SiMe3)(PhCH2-4-C6H9N-2)}H]+), 389 (58, [N(SiMe3)C-
(Ph)C(SiMe3)(C6H9N-2)]+).
EHMO Ca lcu la tion s. Extended Hu¨ckel calculations were
carried out using the CACAO software package with the
standard parameter set as supplied with the program (weighted
Ack n ow led gm en t. We thank the European Com-
munity for a Human Capital and Mobility Grant for B.-
J .D., Hong Kong Research Grant Council for a Ear-
marked Grant CUHK 317/96P, Dr. Ali Abdul-Sada for
MS measurements, and BASF for the generous supply
of high-purity ethylene and propylene and for analysis
of polymer samples.
H
ij formula).23 Idealized C2 symmetrical models A and B were
constructed using the (averaged) bonding parameters of the
X-ray structures of 2b and 3b, respectively. Other parameters
used were Zr-H ) 1.9 Å, N-H ) 1.07 Å, C-H ) 1.09 Å, Zr-
Cp ) 2.2 Å, Cp-Zr-Cp ) 130°, and H-Zr-H ) 95°.
Eth ylen e P olym er iza tion . In a typical experiment 5d (31
mg, 0.053 mmol) was dissolved in a solution of MAO in toluene
(10% (w/w), 3.5 mL, 5.4 mmol of Al). The solution was then
diluted with toluene (31.5 mL) and degassed, and ethylene (1
bar) was admitted to the stirred reaction mixture. After 30-
180 min, the polymerization reaction was quenched with a 10:1
mixture of MeOH and HCl (1 M). The solid polymer was
Su p p or tin g In for m a tion Ava ila ble: Listings of all final
refined and all calculated atomic coordinates, anisotropic
thermal parameters, and a complete list of bond lengths and
angles for the X-ray crystallographic structure determinations
of 2b, 2c, and 3b can be found in the Supporting Information.
This material is available free of charge via the Internet at
http://pubs.acs.org.
(23) CACAO for the PC, version 4.0: Mealli, C.; Proserpio, D. M. J .
Chem. Educ. 1990, 67, 399.
OM980850B