Synthesis of Alkylidene-Bridged Cp/Phosphido Group 4 Metal Complexes-Precursors of the (CpCPR)M-Constrained-Geometry Catalyst Family

Article abstract of DOI:10.1021/om034201y

Phosphide [PHR¹Li (R¹ = cyclohexyl, phenyl) addition to the fulvenes (C₅H₄)CMe₂ (1) and (C ₅H₄)CHCMe₃ (2) yields the corresponding phosphinoalkyl-substituted cyclopentadienides. Subsequent deprotonation at phosphorus with LDA followed by transmetalation of the resulting [(C ₅H₄)CR²R³PR¹]Li ₂ dianion equivalents to Cl₂Ti(NMe₂) ₂ or Cl₂Zr(NEt₂)₂(THF)₂ yields the corresponding [ CpCPR¹ MX₂] constrained-geometry systems (eight examples, 11-18). These systems contain a chiral phosphorus center that is characterized by a low inversion barrier (ΔG_inv^? ranging from ca. 7.5 to 10.0 kcal mol^-1). The structural and some spectroscopic features of the (CpCPR)Ti and Zr complexes were investigated by DFT calculations. Treatment with a large excess of methylalumoxane gave ethene/1-octene copolymerization catalysts. The [(C₅H₄)CMe₂PR ¹]ZrX₂ systems gave the most active catalysts in this series.

Full text of DOI:10.1021/om034201y

1836
Organometallics 2004, 23, 1836-1844
Syn th esis of Alk ylid en e-Br id ged Cp /P h osp h id o Gr ou p 4
Meta l Com p lexessP r ecu r sor s of th e
“(Cp CP R)M-Con str a in ed -Geom etr y” Ca ta lyst F a m ily
Ste´phane Bredeau, Gereon Altenhoff, Klaus Kunz, Steve Do¨ring,
Stefan Grimme,† Gerald Kehr, and Gerhard Erker*
Organisch-Chemisches Institut der Universita¨t Mu¨nster,
Corrensstrasse 40, D-48149 Mu¨nster, Germany
Received September 26, 2003
Phosphide [PHR1]Li (R1 ) cyclohexyl, phenyl) addition to the fulvenes (C5H4)CMe2 (1)
and (C5H4)CHCMe3 (2) yields the corresponding phosphinoalkyl-substituted cyclopentadi-
enides. Subsequent deprotonation at phosphorus with LDA followed by transmetalation of
the resulting [(C5H4)CR2R3PR1]Li2 dianion equivalents to Cl2Ti(NMe2)2 or Cl2Zr(NEt2)2(THF)2
yields the corresponding [“CpCPR1”MX2] “constrained-geometry” systems (eight examples,
11-18). These systems contain a chiral phosphorus center that is characterized by a low
inversion barrier (∆Gq ranging from ca. 7.5 to 10.0 kcal mol-1). The structural and some
inv
spectroscopic features of the (CpCPR)Ti and Zr complexes were investigated by DFT
calculations. Treatment with a large excess of methylalumoxane gave ethene/1-octene
copolymerization catalysts. The [(C5H4)CMe2PR1]ZrX2 systems gave the most active catalysts
in this series.
In tr od u ction
Surprisingly, systems where the Cp and amido groups
of the constrained-geometry ligands were connected by
a C1-based bridging unit instead of the ubiquitous Si1-
based bridging unit were not explored until recently.7
Meanwhile, a number of such alkylidene-bridged Cp/
amido group 4 metal catalyst precursors have been
reported.8,9 In addition to these “CpCNR” metal com-
plexes some “CpCO” complexes have also been reported,
and their catalytic features tested.10,11 We have previ-
ously described the structures of two examples of related
[“CpSiPR”ZrX2] systems,4a,12,13 and we have recently
Dimethylsilanediyl-bridged ansa-metallocenes have
found widespread use as catalyst precursors in homo-
geneous Ziegler-Natta chemistry.1 Formal replacement
of one Cp (or indenyl) group in this ligand arrangement
by an alkylimido moiety (e.g. [Me3C-N-]) leads to the
class of (“CpSiNR”)M or “constrained geometry” cata-
lysts, some examples of which have found interesting
applications, especially in copolymerization reactions.
2-4 Due to the high general interest in this class of olefin
polymerization catalysts a number of ligand variations
has been described and tested, including systems with
additional donor substituents5 or longer connecting
chains between the Cp and amido functional groups.6
(5) Du Plooy, K. E.; Moll, U.; Wocadlo, S.; Massa, W.; Okuda, J .
Organometallics 1995, 14, 3129-3131. Okuda, J .; du Plooy, K. E.;
Massa, W.; Kang, H.-C.; Rose, U. Chem. Ber. 1996, 129, 275-277.
Okuda, J .; Verch, S.; Spaniol, T. P.; Stu¨rmer, R. Chem. Ber. 1996, 129,
1429-1431. Amor, F.; Spaniol, T. P.; Okuda, J . Organometallics 1997,
16, 4765-4767. Brown, S. J .; Gao, X.; Harrison, D. G.; Koch, L.; Spence,
R. E. v. H.; Yap, G. P. A. Organometallics 1998, 17, 5445-5447. Amor,
F.; Butt, A.; du Plooy, K. E.; Spaniol, T. P.; Okuda, J . Organometallics
1998, 17, 5836-5849. Feng, S.; Klosin, J .; Kru¨per, W. J ., J r.; McAdon,
M. H.; Neithamer, D. R.; Nickias, P. N.; Patton, J . T.; Wilson, D. R.;
Abboud, K. A.; Stern, C. L. Organometallics 1999, 18, 1159-1167.
Ashe, A. J ., III; Fang, X.; Kampf, J . W.; Organometallics 1999, 18,
1363-1365. Galan-Fereres, M.; Koch, T.; Hey-Hawkins, E.; Eisen, M.
S. J . Organomet. Chem. 1999, 580, 145-155. Okuda, J .; Eberle, T.;
Spaniol, T. P.; Piquet-Faure´, V. J . Organomet. Chem. 1999, 591, 127-
137. J uvaste, H.; Pakkanen, T. T.; Iiskola, E. I. Organometallics 2000,
19, 1729-1733. Gentil, S.; Pirio, N.; Meunier, P.; Gallucci, J . C.;
Schloss, J . D.; Paquette, L. A. Organometallics 2000, 19, 4169-4172.
J ime´nez, G.; Royo, P.; Cuenca, T.; Herdtweck, E. Organometallics 2002,
21, 2189-2195.
† DFT calculations.
(1) Review: Brintzinger, H.-H.; Fischer, D.; Mu¨lhaupt, R.; Rieger,
B.; Waymouth, R. Angew. Chem. 1995, 107, 1255-1283; Angew. Chem.,
Int. Ed. Engl. 1995, 34, 1143-1170.
(2) Piers, W. E.; Shapiro, P. J .; Bunel, E. E.; Bercaw, J . E. Synlett
1990, 2, 74-84. Shapiro, P. J .; Bunel, E. E.; Schaefer, W. P.; Bercaw,
J . E. Organometallics 1990, 9, 867-869. Okuda, J . Chem. Ber. 1990,
123, 1649-1651. Shapiro, P. J .; Cotter, W. D.; Schaefer, W. P.;
Labinger, J . A.; Bercaw, J . E. J . Am. Chem. Soc. 1994, 116, 4623-
4640.
(3) Reviews: McKnight, A. L.; Waymouth, R. M. Chem. Rev. 1998,
98, 2587-2598. Okuda, J .; Eberle, T. Half-Sandwich Complexes as
Metallocene Analogues. In Metallocenes-Synthesis, Reactivity, Ap-
plications; Togni, A., Haltermann, R. L., Eds.; Wiley-VCH: Weinheim,
Germany, 1998; Vol. 1, pp 415-453.
(6) Hughes, A. K.; Meetsma, A.; Teuben, J . H. Organometallics 1993,
12, 1936-1945. Sinnema, P.-J .; van der Veen, L.; Speck, A. L.;
Veldman, N.; Teuben, J . H. Organometallics 1997, 16, 4245-4247.
Witte, P. T.; Meetsma, A.; Hessen, B.; Budzelaar, P. H. M. J . Am.
Chem. Soc. 1997, 119, 10561-10562. Gomes, P. T.; Green, M. L. H.;
Martins, A. M. J . Organomet. Chem. 1998, 551, 133-138. Van Leusen,
D.; Beetstra, D. J .; Hessen, B.; Teuben, J . H. Organometallics 2000,
19, 4084-4089.
(4) (a) Stevens, J . C.; Timmers, F. J .; Wilson, D. R.; Schmidt, G. F.;
Nickias, P. N.; Rosen, R. K.; Knight, G. W.; Lai, S. Eur. Pat. Appl. EP
416815-A2, 1991 (Dow Chemical Co.). Stevens, J . C. Stud. Surf. Sci.
Catal. 1996, 101, 11-20. (b) Canich, J . M. Eur. Pat. Appl. EP 420436-
A1, 1991 (Exxon Chemical Co.). Canich, J . M.; Hlatky, G. G.; Turner,
H. W. PCT Appl. WO 92-00333, 1992. Pannell, R. B.; Canich, J . M.;
Hlatky, G. G. PCT Int. Appl. WO 94/00500, 1994 (Exxon Chemical Co.).
Canich, J . M. PCT Int. Appl. WO 96/00244, 1996 (Exxon Chemical Co.).
Devore, D. D.; Crawford, L. H.; Stevens, J . C.; Timmers, F. J .; Mussell,
R. D.; Wilson, D. R.; Rosen, R. K. PCT Int. Appl. WO 95/00526, 1995
(Dow Chemical Co.). Nickias, P. N.; McAdon, M. H.; Patton, J . T.;
Friedrichsen, B. P.; Soto, J .; Stevens, J . C.; Vanderlende, D. D. PCT
Int. Appl. WO 97/15583, 1997 (Dow Chemical Co.).
(7) Duda, L.; Erker, G.; Fro¨hlich, R.; Zippel, F. Eur. J . Inorg. Chem.
1998, 1153-1162. Kunz, D.; Erker, G.; Fro¨hlich, R.; Kehr, G. Eur. J .
Inorg. Chem. 2000, 409-416.
(8) Preliminary communication: Kunz, K.; Erker, G.; Do¨ring, S.;
Fro¨hlich, R.; Kehr, G. J . Am. Chem. Soc. 2001, 123, 6181-6182.
10.1021/om034201y CCC: $27.50 © 2004 American Chemical Society
Publication on Web 03/19/2004

Cp/ Phosphido Group 4 Metal Complexes
Organometallics, Vol. 23, No. 8, 2004 1837
reported the first example of a (“CpCPR”)MIVX2-derived
constrained-geometry catalyst system in a preliminary
communication.8 Before that report, to the best of our
knowledge, only a few such ligand systems had been
described, e.g. by J utzi,14 Hey-Hawkins,15 and Mu¨ller
et al.,16 but not their attachment to the group 4 metals.
Sch em e 1
It is probably of some interest to study such phosphido
analogues of the usual constrained-geometry group 4
metal catalysts to learn what effect the incorporation
of a stereogenic phosphorus atom and the increased
strain of the ligand framework will have on the catalytic
behavior of such catalyst systems. We will, therefore,
report here on the synthesis of a small series of such
“CpCPR” titanium and zirconium complexes and briefly
describe some of their catalytic features.
Syn th esis a n d Sp ectr oscop ic Ch a r a cter iza tion
of th e (“Cp CP R”)Ti a n d -Zr Com p lexes
Fulvenes have extensively been used as precursors
for substituted Cp ligands17 as well as the ligands of
ansa-metallocenes.18 It is well established that a variety
of nucleophiles can add to the electrophilic fulvene C6
carbon atom to form the respective cyclopentadienides.19
We had used this typical reaction pattern also for the
synthesis of specific examples of the “CpCNR” ligand
systems. In these cases amide anion equivalents [RNH-]
were added to a non-CH-acidic fulvene, such as (C5H4)-
CHCMe3.8,9
(9) Ko¨nemann, M.; Erker, G.; Fro¨hlich, R.; Wu¨rthwein, E.-U. J . Am.
Chem. Soc. 1997, 119, 11155-11164. Bertuleit, A.; Ko¨nemann, M.;
Duda, L.; Erker, G.; Fro¨hlich, R. Top. Catal. 1999, 7, 37-44. Kunz,
K.; Erker, G.; Kehr, G.; Fro¨hlich, R. Organometallics 2001, 20, 392-
400. Kunz, K.; Erker, G.; Do¨ring, S.; Bredeau, S.; Kehr, G.; Fro¨hlich,
R. Organometallics 2002, 21, 1031-1041.
We have now treated 6-tert-butylfulvene (2) in a
similar way with the phosphido nucleophiles Li[PH-
(cyclohexyl)] and Li[PH(phenyl)], respectively. In each
case, clean addition of the [R1PH-] reagent at the
fulvene C6 carbon atom was observed to yield the
corresponding substituted cyclopentadienides [(C5H4)-
CH(CMe3)PR1H]Li (4 (92% isolated), 6 (97%)). Each of
these functionalized cyclopentadienides contains two
persistent chirality centers, one at carbon (C6; see
Scheme 1) and the other at phosphorus. Consequently,
in each case a mixture of diastereoisomers is obtained
(both ca. 3:1). The major isomer of 4 exhibits a pair of
1H NMR Cp signals at δ 6.10 (2-H/5-H) and δ 6.58 (3-
H/4-H). The 6-H resonance occurs at δ 3.00 with a
(10) Kunz, K.; Erker, G.; Kehr, G.; Fro¨hlich, R.; J acobsen, H.; Berke,
H.; Blacque, O. J . Am. Chem. Soc. 2002, 124, 3316-3326.
(11) See also: (a) Rieger, B. J . Organomet. Chem. 1991, 420, C17-
C20. Herrmann, W. A.; Morawietz, M. J . A.; Priermeier, T. Angew.
Chem. 1994, 106, 2025-2028; Angew. Chem., Int. Ed. Engl. 1994, 33,
1946-1949. Trouve´, G.; Laske, D. A.; Meetsma, A.; Teuben, J . H. J .
Organomet. Chem. 1996, 511, 255-262. Chen, Y.; Fu, P.; Stern, C. L.;
Marks, T. J . Organometallics 1997, 16, 5958-5963. Gielens, E. E. C.
G.; Tiesnitsch, J . Y.; Hessen, B.; Teuben, J . H. Organometallics 1998,
17, 1652-1654. Christie, S. D. R.; Man, K. W.; Whitby, R. J .; Slawin,
A. M. Z. Organometallics 1999, 18, 348-359. Baker, R. W.; Wallace,
B. J . J . Chem. Soc., Chem. Commun. 1999, 1405-1406. Rau, A.;
Schmitz, S.; Luft, G. J . Organomet. Chem. 2000, 608, 71-75. (b)
Ciruelos, S.; Cuenca, T.; Go´mez-Sal, P.; Manzanero, A.; Royo, P.
Organometallics 1995, 14, 177-185. Ciruelos, S.; Cuenca, T.; Go´mez,
R.; Go´mez-Sal, P.; Manzanero, A.; Royo, P. Organometallics 1996, 15,
5577-5585. Park, J . T.; Yoon, S. C.; Bae, B.-J .; Seo, W. S.; Suh, I.-H.;
Han, T. K.; Park, J . R. Organometallics 2000, 19, 1269-1276.
(12) Altenhoff, G.; Bredeau, S.; Erker, G.; Kehr, G.; Kataeva, O.;
Fro¨hlich, R. Organometallics 2002, 21, 4084-4089.
2
coupling constant to phosphorus of J PH ) 4 Hz. The
1H NMR signal of the corresponding P-H moiety is
observed at δ 3.23 (1J PH ) 204 Hz). The 31P NMR
resonance occurs at δ -16.4. The minor diastereoisomer
of the reagent 4 exhibits its 31P NMR signal at δ -47.6
(1J PH ) 183 Hz). In d8-THF solution both isomers show
only a single 7Li NMR signal at δ -4.9. Complex 6
shows similar NMR features (see the Experimental
Section).
(13) See also: Tardif, O.; Hou, Z.; Nishiura, M.; Koizumi, T.;
Wakatsuki, Y. Organometallics 2001, 20, 4565-4573. Kotov, V. V.;
Avtomonov, E. V.; Sundermeyer, J .; Harms, K.; Lemenovskii, D. A.
Eur. J . Inorg. Chem. 2002, 678-691.
(14) Heidemann, T.; J utzi, P. Synthesis 1994, 777-778.
(15) (a) Koch, T.; Blaurock, S.; Somoza, F. B., J r.; Voigt, A.; Kirmse,
R.; Hey-Hawkins, E. Organometallics 2000, 19, 2556-2563. (b) Koch,
T.; Hey-Hawkins, E. Polyhedron 1999, 18, 2113-2116. Ho¨cher, T.;
Blaurock, S.; Hey-Hawkins, E. Eur. J . Inorg. Chem. 2002, 1174-1180.
(16) Bildmann, U. J .; Mu¨ller, G. Z. Naturforsch. 2000, 55, 895-900.
(17) Sullivan, M. F.; Little, W. F. J . Organomet. Chem. 1967, 8, 277-
285. Renaut, P.; Tainturier, G.; Gautheron, B. J . Organomet. Chem.
1978, 148, 35-42. Bensley, D. M., J r.; Mintz, E. A. J . Organomet.
Chem. 1988, 353, 93-102. Kettenbach, R. T.; Bonrath, W.; Butenscho¨n,
H. Chem. Ber. 1993, 126, 1657-1669. Bosch, B. E.; Erker, G.; Fro¨hlich,
R.; Meyer, O. Organometallics 1997, 16, 5449-5456.
The Li-phosphido reagents Li[PHR1] are much less
basic than their Li[NHR] counterparts. We made use
of this characteristic feature which allowed us to add
the reagents Li[PH(cyclohexyl)] and Li[PH(phenyl)],
respectively, to 6,6-dimethylfulvene (1). Although a
small fraction of the competing deprotonation reaction
took place, the major reaction pathway was addition of
the phosphido anion reagents to the electrophilic fulvene
carbon center C6 to give the corresponding (phosphi-
(18) Ewen, J . A.; J ones, R. L.; Razavi, A.; Ferrara, J . D. J . Am. Chem.
Soc. 1988, 110, 6255-6256.
(19) Ziegler, K.; Scha¨fer, W. Liebigs Ann. Chem. 1934, 511, 101-
109. Ziegler, K.; Gellert, H.-G.; Martin, H.; Nagel, K.; Schneider, J .
Liebigs Ann. Chem. 1954, 589, 91-121.

1838 Organometallics, Vol. 23, No. 8, 2004
Bredeau et al.
F igu r e 1. Decoalescence of the 1H NMR (C5H4) signals of 11 (in CDFCl2/CDF2Cl/CD2Cl2 at 600 MHz; S denotes a residual
CDHCl2 solvent signal).
noalkyl)-functionalized cyclopentadienides [(C5H4)CMe2-
PHR1]Li (3, R1 ) cyclohexyl; 5, R1 ) phenyl), each in
ca. 70% yield. Compound 5 exhibits a 31P NMR signal
(NMe2)2 to yield the corresponding [CpCPR1]TiX2 com-
plexes 11 (65%) and 13 (56%).
At ambient temperature the NMR spectra are in
accord with a symmetry-averaged apparent Cs-sym-
metric structure of these complexes. However, this is
due to an equilibration of two enantiomeric structures
on the NMR time scale caused by a rapid inversion at
the stereogenic phosphorus atom. In a CDFCl2/CDF2-
Cl/CD2Cl2 solvent mixture21 lowering the temperature
leads to a broadening and decoalescence of the pair of
C5H4 1H NMR signals into two pairs (see Figure 1). At
the same time we observe the decoalescence of the single
NMe2 1H NMR resonance to a 1:1 pair of signals below
213 K (∆ν ) 60 Hz) and also the decoalescence of the
signal of the pair of methyl groups at carbon atom C6
(Tc ) 223, ∆ν ) 150 Hz). In addition a complex pattern
evolves of the 1H NMR P-cyclohexyl resonances that was
not further analyzed. From the Cp, NMe2, and CMe2
1H NMR coalescences an activation energy of ∆Gq(213
K) ) 10.0 ( 0.5 kcal mol-1 was estimated22 for the
phosphorus inversion of complex 11 and the concomitant
conformational equilibration of the chiral ligand back-
bone of the [“CpCP(cyclohexyl)”]Ti(NMe2)2 system.
1
at δ -1.1 with a corresponding H NMR PH resonance
at δ 4.40 (1J PH ) 209 Hz). The methyl groups at C6 are
diastereotopic, due to the adjacent phosphorus chirality
center. They give rise to 1H NMR features at δ 1.67 (3J PH
) 7 Hz) and δ 1.61 (3J PH ) 14 Hz). The C6 carbon atom
2
of 5 shows a 13C NMR resonance at δ 28.4 (with J PC
)
20 Hz), and the 7Li NMR resonance (in d8-THF/d8-
toluene 1:10) of 5 is observed at δ -7.5. Compound 3
shows similar NMR features (for details see the Experi-
mental Section).
Treatment of 4 with LDA in THF at 0 °C resulted in
deprotonation at phosphorus. The resulting dilithio salt
of the “CpCPR” ligand system (8) was in this case
isolated in ca. 50% yield. The reagent 8 was then reacted
with 1 molar equiv of Cl2Ti(NMe2)2.20a After removal of
lithium chloride the (“CpCPR”)TiX2 complex 12 (see
Scheme 1) was isolated as a red-brown oil in 56% yield.
The titanium complex 12 contains a carbon chirality
1
center (C6) that leads to the observation of four H NMR
Cp signals at δ 5.71, 5.54, 5.43, and 4.82 (13C NMR δ
112.0, 111.0, 110.8, 108.7) and a pair of signals of the
diastereotopic -NMe2 σ ligands at titanium (1H NMR
δ 3.11, 2.91; 13C NMR δ 48.6, 45.1). The 13C NMR
resonance of the bridging carbon atom C6 occurs at δ
The corresponding [“CpCP(phenyl)”]Ti(NMe2)2 system
(13) exhibits an analogous chiral structure that is
caused by the presence of a stereogenic nonplanar
tricoordinate phosphorus center. Consequently at low
temperature (<173 K) complex 13 features four separate
C5H4 1H NMR signals, a pair of N(CH3)2 resonances,
and a pair of C(CH3)2 signals (frozen spectrum in
CDFCl2/CDF2Cl/CD2Cl2). Pairwise coalescence is ob-
served upon raising the monitoring temperature. From
1
47.9 with J PC ) 30 Hz. The 31P NMR resonance of
complex 12 is strongly temperature dependent. It is
almost linearly shifted from δ -25.5 at 353 K to δ -65.5
at 193 K. A discussion and possible explanation will be
given in DFT Calculations.
1
For the preparation of the corresponding (“CpCPPh”)-
Ti(NMe2)2 complex (14), the reagent 10 was generated
in situ by LDA deprotonation of 6 followed by treatment
with the Cl2Ti(NMe2)2 reagent. Complex 14 was isolated
in 79% yield as a red-brown oil. It also shows a
temperature dependency of the 31P NMR shift, although
not nearly as pronounced as in 12 (14: 31P NMR δ -54.4
at 298 K, δ -61.3 at 193 K).
the temperature-dependent 600 MHz H NMR spectra
in the Freon solvent mixture a Gibbs activation energy
of ∆Gq(213 K) ) 9.4 ( 0.5 kcal mol-1 was obtained22 for
the phosphorus inversion process of complex 13. At 298
K only symmetry-averaged NMR spectra of complex 13
are observed (1H NMR in d8-toluene: δ 5.71/5.28 (C5H4),
δ 3.00 (NMe2), δ 1.46 (3J PH ) 11 Hz, CMe2)), due to this
rapid equilibration on the NMR time scale. In addition
the 31P NMR resonance of 13 shows a marked temper-
ature-dependent chemical shift (δ 3.0 at 298 K, δ -13.0
at 193 K).
Similarly, deprotonation of 3 or 5 with LDA gave the
Li2[CpCPR1] reagents 7 (R1 ) cyclohexyl) and 9 (R1 )
phenyl), respectively. Both were treated with Cl2Ti-
(20) (a) Benzing, E.; Kornicker, W. Chem. Ber. 1961, 94, 2263-2267.
(b) Brenner, S.; Kempe, P.; Arndt, P. Z. Anorg. Allg. Chem. 1995, 621,
2021-2024.
(21) Siegel, J . S.; Anet, F. A. L. J . Org. Chem. 1988, 53, 2629-2630.
(22) Green, M. L. H.; Wong, L.-L.; Seela, A. Organometallics 1992,
11, 2660-2668 and references therein.

Cp/ Phosphido Group 4 Metal Complexes
Organometallics, Vol. 23, No. 8, 2004 1839
Sch em e 2
Sch em e 3
mental Section). The 31P NMR chemical shift of complex
18 (δ -80.7 at 298 K) is almost temperature invariant.
Treatment of the dilithiated reagent Li2[(C5H4)CMe2P-
(phenyl)] (9) with Cl2Zr(NEt2)2(THF)2 furnished the
product 17 (orange oil, 65% isolated). Again, the rapid
dynamic conformational equilibration leads to the ob-
servation of apparently symmetry-equivalent pairs of
groups and substituents for the actual chiral complex
17. At ambient temperature it features 13C NMR signals
of the NCH2CH3 ligands at δ 42.7 (CH2) and δ 15.8
Inversion at phosphorus in phosphanes is usually very
slow, which makes trivalent phosphorus in most ligand
environments a stable center of chirality.23 There are,
however, a number of cases known where a σ-bonded
metal electrophile strongly interacts with the phospho-
rus lone pair, which may result in a partial double-bond
character of the ensuing metal-P interaction.24
A
1
(CH3). However, the H NMR spectrum of the N(CH2-
noteworthy example is the bis(phosphido)hafnocene
complex 19, described by Baker et al.,25 which actually
exhibits two P atoms in a significantly different bonding
situation in the solid state as well as in solution (Scheme
2). The 19 h 19′ interconversion probably represents a
process that is closely related to the dynamic process
which is observed here for the complexes 11 and 13. It
is likely that the observed P-inversion process involves
a π-bonded P-Ti-type structure. DFT calculations of
this process (see below), however, indicate that in our
case the π-bonded structure (20‡) represents a transition
state rather than a high-lying intermediate.
CH3)2 system in 17 is more complicated, due to the
diastereotopism of its N-CH2 nuclei (δ 3.29/3.19),
whereas the system only gives rise to a single N(CH2C-
H3)2 1H NMR methyl resonance at δ 0.92. The 13C NMR
C6 signal of 17 appears at δ 34.2 (1J PC ) 22 Hz) with
the adjacent C(CH3)2 resonance at δ 31.4 (2J PC ) 14 Hz,
3
1H NMR δ 1.60 with J PH ) 11 Hz). The Cp-methine
hydrogen atoms of complex 17 are observed in d8-
toluene at ambient temperature at δ 5.98 (3-H/4-H) and
δ 5.54 (2-H/5-H), respectively. These signals decoalesce
at very low temperature (Tcoal in CDFCl2/CDF2Cl/CD2-
Cl2 at ca. 173 K) to two pairs of C5H4 1H NMR
1
The corresponding zirconium complexes were pre-
pared in a similar way. They show analogous features.
resonances. From these dynamic H NMR (600 MHz)
spectra a Gibbs activation energy of ∆Gq(173 K) ) 7.5
( 0.5 kcal mol-1 was estimated for the conformational
equilibration process of complex 17, which includes
stereochemical inversion at the phosphorus atom of the
ligand system (see Scheme 3). We note that this barrier
is markedly lower than that of the corresponding
titanium complex 13 (see above).
Treatment of the reagent Li2[(C5H4)CH(CMe3)P(cyclo-
20b
hexyl)] (8) with Cl2Zr(NEt2)2(THF)2
gave the corre-
sponding [“CpCP(cyclohexyl)”Zr(NEt2)2] complex 16 as
an orange oil in ca. 60% yield. The chiral complex
features four separate NMR signals of the C5H4 methine
units (1H at δ 6.02, 5.71, 5.56, 5.06; 13C at δ 110.3, 108.6,
108.1, 107.8) and a 31P NMR signal at δ -69.2 (298 K).
The two -NEt2 groups are different (cis and trans to
the CMe3 substituent at C6). This gives rise to pairs of
13C NMR NCH2 resonances (δ 43.5, 41.4) and NCH2CH3
signals (δ 15.9, 15.6). Due to the diastereotopism of the
-NCH2 hydrogen atoms in 16, the corresponding 1H
NMR spectrum is more complicated, giving rise to (in
principle) four groups of NCH2 1H NMR signals (found
at δ 3.30 (2H), 3.22 (4H), and 3.15 (2H)) and two
corresponding NCH2CH3 triplets (observed at δ 0.95 and
0.94).
For the preparation of the P-phenyl derivative 18 (see
Scheme 1) the ligand system Li2[(C5H4)CH(CMe3)P-
(phenyl)] (10) was again generated in situ by treatment
of 6 with LDA. The subsequent reaction with Cl2Zr-
(NEt2)2(THF)2 gave the [“CpCP(phenyl)”Zr(NEt2)2] com-
plex 18 again as an orange oil (59% isolated yield). It
shows similar spectroscopic features (see the Experi-
The complex [“CpCP(cyclohexyl)”Zr(NEt2)2] (15) was
prepared analogously by treatment of 7 with the reagent
Cl2Zr(NEt2)2(THF)2. Complex 15 was isolated in close
to 70% yield. Again, it shows dynamic behavior in
solution, featuring pairwise decoalescence of its C5H4
1H NMR signals at very low temperature in CDFCl2/
CDF2Cl/CD2Cl2 solution. The Gibbs activation energy
of this P-inversion process was determined to be ∆Gq-
(193 K) ) 8.6 ( 0.5 kcal mol-1
.
DF T Ca lcu la tion s
Because X-ray data are unfortunately not available
for compounds 11-18, their structures, the inversion
barriers, and the 31P NMR chemical shifts were inves-
tigated by quantum-chemical methods. To test the
quantum-chemical methodology, a similar compound,
where experimental structural data were available, was
considered first. In the zirconium compound 21 (Chart
1), C6 has formally been replaced by a silicon atom and
(23) Phosphorus-31 NMR Spectral Properties in Compound Char-
acterization and Structural Analysis; Quin, L. D., Verkade, J . G., Eds.;
VCH: Weinheim, Germany, 1994.
(24) Weber, L.; Meine, G.; Boese, R.; Augart, N. Organometallics
1987, 6, 2484-2488.
(25) Baker, R. T.; Whitney, J . F.; Wreford, S. S. Organometallics
1983, 2, 1049-1051.
the C5Me4 ligand instead of C5H4 is present.4a,12,13
A
detailed comparison between computed and experimen-
tal structural data is given in the Supporting Informa-
tion.

1840 Organometallics, Vol. 23, No. 8, 2004
Bredeau et al.
Ch a r t 1
Ta ble 1. Com p u ted (DF T-BP /TZVP ) In ver sion
Ba r r ier s, M-P Bon d Len gth s, a n d P h osp h or u s
P yr a m id a liza tion An gles for Com p ou n d s 11a , 13a ,
15a , a n d 17a (See Text) a n d th e Cor r esp on d in g
Tr a n sition Sta tes for Th eir In ver sion a t
P h osp h or u sa
b
∆G°rel
∆Gq
d(P-M) (C-P-M)-R1
inv
compd R1
M
(kcal/mol) (kcal/mol) (pm)
(deg)
11a
11a (TS)
13a
13a (TS)
15a
15a (TS)
17a
Me Ti
Ph Ti
Me Zr
Ph Zr
0
7.8
0
5.9
0
6.7
0
257.8
241.2
255.7
242.6
270.9
254.8
267.6
256.5
70.2
4.9
57.0
3.8
72.2
4.5
56.9
3.4
10.0
9.4
F igu r e 2. DFT-BP/TZVP calculated ground-state struc-
tures of 11a and 17a and their corresponding transition
states (TS) of the rapid inversion process at phosphorus.
8.6
17a (TS)
5.2
7.5
a
ment is observed for 13a , where the barrier deviates
by 3.5 kcal/mol. The most striking feature of all TS
structures is the shortened M-P bond length. It amounts
to about 16 pm for the two compounds with R1 ) Me
and 13 pm (Ti) and 11 pm (Zr), respectively, for the two
phenyl-substituted systems. This indicates the better
ability of the phosphorus lone-pair orbital to interact
with the metal fragment in the transition states and
explains the small barrier for the inversion process.
These findings, furthermore, allow us to gain a
qualitative understanding of the observed temperature
dependence of the 31P NMR chemical shifts. The 31P
NMR shielding constants strongly depend on the lengths
(and consequently the strengths) of the M-P bond. For
example, for 13a we have calculated a change of about
156 ppm when going from the minimum structure to
the TS. Because of the asymmetry of the double-
minimum potential energy curve for the inversion
process, thermally populated higher vibrational levels
have more probability closer to the TS region (i.e. at
shorter M-P distances), which thus would strongly
affect the observed NMR chemical shift. A quantitative
description of this process would require a combined
quantum-mechanical/Boltzmann averaging of the com-
puted shielding constants along complex hypersurfaces,
which is currently out of reach computationally.
See Figure 2 for a depiction of the calculated structures.
b
Experimental values ((0.5 kcal/mol).
Density functional theory (DFT) with a gradient-
corrected BP86 functional and AO basis set of valence
triple-ú quality was employed as outlined in more detail
in the Experimental Section. The agreement between
experiment and theory in the case of 21 was generally
good. The bond distances were systematically overesti-
mated by 4-5 pm, and angles deviated only by about
1-2° from experiment. The systematic overestimation
of the bond distances between second- and third-row
atoms by almost all current density functionals is well-
known but is not expected to influence our general
conclusions. To simplify matters, the experimentally
investigated compounds 11, 13, 15, and 17 have been
slightly modified in the computational procedure; i.e.,
the substituents on the nitrogen atoms are always
methyl and the cyclohexyl moiety in 11 and 15 has also
been replaced by a methyl group (11a -17a ). The results
for these systems are summarized in Table 1. The
structures of 11a and 17a and their corresponding
transition states (TS) of inversion at phosphorus are
shown graphically in Figure 2. For both sets of mol-
ecules (11a and 13a with Ti and 15a and 17a with Zr)
the effect of the substituent at the phosphorus atom is
clearly evident. Replacing Me by Ph leads to shorter
M-P bond distances (about 2 pm) and smaller pyrami-
dalization angles (by 13-16°) due to conjugative inter-
actions of the lone-pair orbital with the phenyl ring. The
effect is slightly more pronounced for the Zr compounds
15 and 17 than for their Ti analogues. As expected,
these structural features furthermore correlate with the
barrier heights for inversion at the phosphorus atom.
The computed free enthalpies of activation are, in
general, in good agreement with those from experiment
(about 1-2 kcal too low). The effect of the metal (smaller
barrier for the Zr compounds relative to those with Ti)
and the substituent effect (about 1 kcal/mol smaller
barrier for R1 ) Ph) are computed almost quantitatively.
The only notable difference between theory and experi-
Olefin P olym er iza tion Rea ction s
The “CpCPR” titanium and zirconium complexes 11-
18 were used as group 4 metal components for the
generation of homogeneous constrained-geometry Zie-
gler-Natta catalysts. In each case the respective metal
diamide complex was activated by treatment with a
large excess of methylalumoxane in toluene solution (Al:
Zr ratios ranging from ca. 825 to ca. 1450).
All catalyst systems polymerize ethene at 60 °C to
give linear polyethylene, but in most cases the catalyst
activities were rather low (see Table 2). Noteworthy
exceptions are the zirconium catalysts derived from the
isopropylidene-bridged Cp/phosphido ligands (15 and

Cp/ Phosphido Group 4 Metal Complexes
Organometallics, Vol. 23, No. 8, 2004 1841
Ta ble 2. Eth en e P olym er iza tion Rea ction s w ith
th e “Cp CP R”Ti/Zr X2 Com p lexesa
quite acceptable polymerization activities upon MAO
activation. The most active system in this series, 15/
MAO, was found to be more than 50 times more active
than its titanium analogue 11/MAO (see Table 3), and
the related zirconium system 17/MAO is still a >15
times more active copolymerization catalyst, as com-
pared to its titanium relative 13/MAO.
However, not all the “CpCPR”Zr systems give very
active catalysts for ethene/1-octene copolymerization.
This activity depends critically on the structure and
substitution pattern of the respective ligand backbone.
This is illustrated by the last two entries in Table 3:
the (CpCPR)Zr systems 16 and 18 upon MAO activation
give only low-activity polymerization catalysts.
We conclude that the alkylidene-bridged (CpCPR)MX2
catalyst precursorssrelatives of the ubiquitous (CpSiN-
R)MX2 systemsare readily available by means of a
fulvene route. The systems can be activated by treat-
ment with MAO to give active olefin polymerization and
copolymerization catalysts. The titanium complexes
seem to be markedly less active than the zirconium
systems, and the catalyst activities generally seem to
depend critically on the substitution pattern of the
backbone, which will leave some room for further
catalyst developments in this area. The Me2C-bridged
zirconium complexes (15, 17) give rather active copo-
lymerization catalysts. The molecular weights of the
obtained products are rather low (see Table 3), whereas
typical 1-octene incorporation ratios were achieved. The
study shows that the exchange of groups and atoms at
the backbones of the constrained-geometry complex
systems has a profound influence on the actual catalyst
performance and characteristics, which may open up
routes to interesting new catalyst systems and product
targets.
amt of
amt of mp
compd
M
CR1R2 b PRc cat. (mg)f Al/M PE (g) (°C) actg
11
13
12
14
15
17
16
18
Ti CMe2
Ti CMe2
Ti CHRd
Ti CHRd
Zr CMe2
Zr CMe2
Zr CHRd
Zr CHRd
Cye
Ph
Cy
Ph
Cy
Ph
Cy
Ph
15
14
11
13
10
20
11
17
825
840
1200
1000
1450
825
1.8
4.4
1.0
0.6
8.5
6.9
0.3
0.4
129
130
128
129
21
56
18
9
130 193
130
126
129
78
6
1400
950
6
a
For each complex a representative example out of a series of
similar experiments is listed. Reactions were carried out in toluene
b
solution, for 1 h at 60 °C, under 2 bar of ethene. Bridging
d
alkylidene unit. c Hydrocarbyl group at phosphorus. R ) CMe3.
g
e Cy ) cyclohexyl. f Group 4 metal component. Activity in units
of g of PE/((mmol of cat.)(bar of ethene) h).
Ta ble 3. Eth en e/1-Octen e Cop olym er iza tion w ith
th e (Cp CP R)MX2 Com p lexes a t 90 °Ca
amt of
Mw/
Mn
compd
M
CR1R2 PR cat. (mg) Al/M actc a:bd
Mw
11
13
12
14
15
17
16
18
Ti CMe2 Cy
Ti CMe2 Ph
Ti CHRb Cy
Ti CHRb Ph
Zr CMe2 Cy
Zr CMe2 Ph
Zr CHRb Cy
Zr CHRb Ph
13
15
13
12
10
10
11
11
960 15 6:1 12000 2.6
825 25 5:1
960 18 6:1
e
1000
5 7:1 14000 1.9
1650 860 6:1 9000f 1.9f
1500 400 6:1 8000 1.9
1400 40 6:1 5000 1.9
1500 12 6:1 13000 2.1
a
In toluene solution, for 1 h reaction time, under 2 bar of
ethene. Representative examples out of larger series of similar
b
experiments are listed. R ) CMe3. c Activity in units of (g of
d
copolymer)/((mmol of cat.)(bar of ethene) h). Ethene/1-octene ratio
in the copolymer. e Not determined. f Values from a copolymer
obtained at 1 bar of ethene pressure.
17), which show slightly higher ethene polymerization
activities under the applied reaction conditions.
More importantly, the 11-18/MAO systems provide
active catalysts for the random copolymerization of
ethene with the R-olefin 1-octene. Typical reaction
conditions and polymerization results are listed in Table
3. In all cases a reasonable incorporation of the long-
chain 1-alkene component was achieved under the
selected reaction conditions (90 °C in toluene solution
at 2 bar ethene pressure). The characteristic 13C NMR
spectrum26 of a typical copolymer example, namely the
ethene/1-octene copolymer obtained at the most active
catalyst system (15/MAO, entry 5 in Table 3), is depicted
and described in Figure 3.
In the “normal” “CpSiNR”M series the titanium
complexes usually seem to give the best catalyst per-
formances.27 In this “CpCPR”M series of complexes the
situation seems to be rather different. The catalyst
activities depend strongly on the metal and apparently
also on the specific substitution pattern of the ligand
backbone. All four titanium catalyst systems give rather
low activities in ethene/1-octene copolymerization (see
Table 3). Generally, their direct zirconium counterparts
all seem to be more active catalysts. This is drastically
illustrated with the Zr systems [Cp(CMe2)PCy]ZrX2 (15)
and [Cp(CMe2)PPh]ZrX2 (17), which both give in general
Exp er im en ta l Section
All reactions were carried out under argon using Schlenk-
type glassware or in a glovebox. Solvents, including deuterated
solvents used for NMR spectroscopy, were dried and distilled
prior to use. NMR spectra were measured using a Bruker AC
200 P or Varian Unity Plus 600 NMR spectrometer. Most
assignments were based on a series of 2D NMR experiments.28
Melting points were determined by differential scanning
calorimetry (2010 DSC, Du Pont/STA Instruments). GPC was
performed using a Agilant 1100 chromatograph with UV/vis
absorbance detector and RI detector. Pentafulvenes 1 and 2,29
LiPHR (R ) Cy, Ph),30 and the reagents Cl2Zr(NEt2)2(THF)2,20b
20a
and Cl2Ti(NMe2)2
were prepared according to literature
procedures. Some of the products were obtained as viscous oils,
which were difficult to get analytically pure.
P r ep a r a tion of Lith iu m [1-Meth yl-1-(cycloh exylp h os-
p h id o)et h yl]cyclop en t a d ien id e (3). Lithium cyclohexyl-
phosphide (920 mg, 7.5 mmol) was suspended in 70 mL of
pentane and reacted with 850 mg (8.0 mmol) of 6,6-dimeth-
ylfulvene (1) at -78 °C. Over 12 h a white solid precipitated.
The solid product was collected by filtration, washed twice with
10 mL of pentane, and dried in vacuo to give 1.21 g (71%) of
1
3. H NMR (200.1 MHz, THF-d8): δ 5.62, 5.56 (each m, each
2H, Cp H), 3.00 (dd, 1J PH ) 194 Hz, 3J HH ) 2.1 Hz, 1H, P-H),
3
1.47 (d, J PH ) 11.2 Hz, 6H, CH3), 1.73-0.81 (m, 11H, Cy H).
(26) Randall, J . C. J MS-Rev. Macromol. Chem. Phys. 1989,
C29(2&3), 201-317.
(27) Okuda, J .; Eberle, T.; Spaniol, T. P. Chem. Ber. 1997, 130, 209-
215. Eberle, T.; Spaniol, T. P.; Okuda, J . Eur. J . Inorg. Chem. 1998,
237-244 and references therein.
(28) Braun, S.; Kalinowski, H. O.; Berger, S. 150 and More Basic
NMR Experiments; VCH: Weinheim, Germany, 1998, and references
therein.
(29) Stone, K. L.; Little, R. D. J . Org. Chem. 1984, 49, 1849-1853.
(30) Kunz, K. Doctoral Dissertation, Universita¨t Mu¨nster, 2001.

1842 Organometallics, Vol. 23, No. 8, 2004
Bredeau et al.
F igu r e 3. 13C{1H} NMR spectrum (C6D6, 300 K) of the ethene/1-octene ca. 6:1 copolymer obtained with the 15/MAO
catalyst at 90 °C in toluene (1 bar of ethene pressure).
1
1
3
3
31P NMR (80.1 MHz, THF-d8): δ 7.42 (dm, J PH ) 194 Hz,
4.99 (dd, J PH ) 220 Hz, J HH ) 5 Hz, 1H, P-H), 3.45 (d, J HH
) 5 Hz, 1H, 6-H), 1.19 (s, 9H, C(CH3)3). 31P NMR (81.0 MHz,
7
P-H). Li NMR (232.8 MHz, THF-d8, 298 K): δ -4.9.
1
benzene-d6/THF-d8 (4:1)): diastereomer A, δ -30.5 (d, J PH
)
P r ep a r a tion of Lith iu m [1-(Cycloh exylp h osp h id o)-2,2-
dim eth ylpr opyl]cyclopen tadien ide (4). Lithium cyclohexyl-
phosphide (920 mg, 7.5 mmol) was suspended in 70 mL of
pentane and reacted with 1.07 g (8.0 mmol) of 6-tert-butylful-
vene (2) at -78 °C. Over 12 h a white solid precipitated. The
solid was collected by filtration, washed twice with 10 mL of
pentane, and dried in vacuo to give 1.58 g (6.9 mmol, 92%) of
4 as a 3:1 mixture of two diastereomers. 1H NMR (200.1 MHz,
THF-d8/benzene-d6 (1:4)): diastereomer A (major), δ 6.58, 6.10
220 Hz); diastereomer B, δ -58.2 (d, 1J PH ) 189 Hz). 7Li NMR
(232.8 MHz, THF-d8/toluene-d8 (1:10), 289-193 K): δ -7.7.
Dep r oton a tion of 3: F or m a tion of 7. A sample of 228
mg (1.0 mmol) of 3 was dissolved in 5 mL of THF and treated
with 107 mg (1.0 mmol) of LDA in 3 mL of THF at 0 °C for 1
h. The solvent was removed, and subsequent recrystallization
by diffusion of pentane into a concentrated THF solution gave
187 mg (80%) of 7 as pale yellow crystals. 1H NMR (200.1 MHz,
3
1
3
THF-d8): δ 5.62, 5.56 (each m, each 2H, Cp H), 1.47 (d, J PH
(each m, each 2H, Cp H), 3.23 (ddd, J PH ) 204 Hz, J HH
)
) 11.2 Hz, 6H, 7-H), 1.73-0.81 (m, 11H, Cy H). 31P NMR (81.0
MHz, THF-d8): δ 33.3 (s).
3J HH ) 6 Hz,3J HH ) 4 Hz, 1H, P-H), 3.00 (dd, J PH ) 4 Hz,
3J HH ) 6 Hz, 1H, 6-H), 1.27 (s, 9H, C(CH3)3), 1.8-1.17 (11H,
Cy H); diastereomer B, δ 6.58, 6.10 (each m, each 2H, Cp H),
3.55 (ddd, 1J PH ) 183 Hz, 3J HH ) 6 Hz, 3J HH ) 4 Hz, 1H, P-H),
2
Rea ction of 4 w ith LDA: F or m a tion of 8. A sample of
270 mg (1.0 mmol) of 4 was dissolved in 5 mL of THF and
treated with 107 mg (1.0 mmol) of LDA in 3 mL of THF at 0
°C for 1 h. The solvent was then removed, and subsequent
recrystallization by diffusion of pentane into a concentrated
THF solution yielded 135 mg (50%) of 8 as pale yellow crystals.
1H NMR (200.1 MHz, THF-d8/benzene-d6 (4:1)): δ 5.92, 5.83
(each m, each 2H, Cp H), 1.47 (9H, C(CH3)3), 2.42-1.02 (bm,
11H, Cy H). 31P NMR (81.0 MHz, THF-d8/benzene-d6): δ -24.3
(s).
2
3
3.39 (t, J PH ) 6 Hz, J HH ) 6 Hz, 1H, 6-H), 1.24 (s, 9H,
C(CH3)3), 1.8-1.17 (11H, Cy-H). 31P NMR (81.0 MHz, THF-
d8): diastereomer A, δ -16.4 (1J PH ) 204 Hz); diastereomer
B, δ -47.6 (1J PH ) 183 Hz). 7Li NMR (232.8 MHz, THF-d8,
298 K): δ -4.9.
P r ep a r a tion of Lith iu m [1-Meth yl-1-(p h en ylp h osp h i-
d o)eth yl]cyclop en ta d ien id e (5). Lithium phenylphosphide
(870 mg, 7.5 mmol) was suspended in 70 mL of pentane and
reacted with 850 mg (8.0 mmol) of 6,6-dimethylfulvene (1) at
-78 °C. Over 12 h a white solid precipitated. The precipitate
was collected by filtration, washed twice with 10 mL of
pentane, and dried in vacuo to give 1.21 g (71%) of 5. 1H NMR
(200.1 MHz, THF-d8/benzene-d6 (1:4)): δ 7.24 (m, 2H, Ph H),
Tr ea tm en t of 5 w ith LDA: F or m a tion of 9. A sample of
228 mg (1.0 mmol) of 5 was dissolved in 5 mL of THF and
treated with 107 mg (1.0 mmol) of LDA in 3 mL of THF at 0
°C for 1 h. The solvent was removed, and subsequent recrys-
tallization by diffusion of pentane in a concentrated THF
solution gave 126 mg (54%) of 9 as pale yellow crystals. 1H
NMR (200.1 MHz, THF-d8): δ 7.24 (m, 2H, Ph H), 7.12 (m,
3H, Ph H), 6.06 (m, 4H, Cp H), 1.67 (d, 3J PH ) 7 Hz, 3H, 7-H),
1
7.12 (m, 3H, Ph H), 6.06 (m, 4H, Cp H), 4.40 (d, J PH ) 209
3
3
Hz, 1H, P-H), 1.67 (d, J PH ) 7 Hz, 3H, CH3), 1.61 (d, J PH
)
14 Hz, 3H, CH3). 31P NMR (200.1 MHz, THF-d8/benzene-d6
(1:10)): δ -1.1 (d, 1J PH ) 209 Hz). 7Li NMR (232.8 MHz, THF-
d8/toluene-d8 (1:10), 289-193 K): δ -7.5.
3
1.61 (d, J PH ) 14 Hz, 3H, 7-H′). 31P{1H} NMR (81.0 MHz,
THF-d8/benzene-d6 (1:4)): δ 28.7 (s).
P r ep a r a t ion of Lit h iu m [1-(P h en ylp h osp h id o)-2,2-
dim eth ylpr opyl]cyclopen tadien ide (6). Lithium phenylphos-
phide (920 mg, 7.5 mmol) was suspended in 70 mL of pentane
and reacted with 1.07 g (8.0 mmol) of 6-tert-butylfulvene (2)
at -78 °C. Over 12 h a white solid precipitated. The solid was
collected by filtration, washed twice with 10 mL of pentane,
and dried in vacuo to give 1.85 g (97%) of 6 as a mixture of
P r ep a r a tion of 11. A solution of 117 mg (0.50 mmol) of 7
in 7 mL of THF was added dropwise to 97 mg (0.47 mmol) of
Cl2Ti(NMe2)2 in 3 mL of THF at 0 °C. Over 2 h the reaction
mixture was stirred at 0 °C. The solvent was removed in vacuo,
and the oily residue was treated with 10 mL of pentane. After
removal of LiCl by filtration and evaporation of the solvent,
1
106 mg (65%) of 11 was obtained as a red-brown oil. H NMR
1
two diastereoisomers (1:3). H NMR (200.1 MHz, benzene-d6/
(599.8 MHz, toluene-d8, 298 K): δ 5.69 (m, 2H, 3-H, 4-H), 5.18
(m, 2H, 2-H, 5-H), 2.99 (s, 12H, NCH3), 1.92 (m, 2H, 2′-H),
THF-d8 (4:1)): diastereomer A, δ 7.80, 7.60, 7.05 (each m, 5H,
1
3
Ph H), 6.09 u. 6.05 (each m, each 2H, Cp H), 4.26 (dd, J PH
)
1.76 (m, 2H, 3′-H), 1.60 (m, 1H, 4′-H), 1.53 (d, J PH ) 10 Hz,
3
3
220 Hz, J HH ) 5 Hz, 1H, P-H), 3.51 (d, J HH ) 5 Hz, 1H,
6-H), 1.25 (s, 9H, C(CH3)3); diastereomer B, δ 7.80, 7.60, 7.05
(each m, 5H, Ph-H), 6.09 and 6.05 (each m, each 2H, Cp H),
6H, C(CH3)2), 1.40 (m, 2H, 2′-H′), 1.26 (m, 2H, 3′-H′), 1.22 (m,
1H, 4′-H′), 1.21 (m, 1H, 1′-H). 13C{1H} NMR (150.8 MHz,
toluene-d8, 298 K): δ 120.0 (C-1), 110.9 (C-3, C-4), 107.2 (C-2,

Cp/ Phosphido Group 4 Metal Complexes
Organometallics, Vol. 23, No. 8, 2004 1843
3
1
C-5), 46.6 (d, J PC ) 5.2 Hz, NCH3), 36.7 (d, J PC ) 38.3 Hz,
K). Anal. Calcd for C20H31N2PTi (378.4): C, 63.49; H, 8.26; N,
7.40. Found: C, 62.45; H, 8.52; N, 6.38.
2
2
C-1′), 34.9 (d, J PC ) 15.1 Hz, C-2′), 30.5 (C-6), 28.8 (d, J PC
)
3
14.6 Hz, C(CH3)2), 27.7 (d, J PC ) 9.3 Hz, C-3′), 27.0 (C-4′).
31P{1H} NMR (202.6 MHz, toluene-d8): δ 14.2 (298 K). Anal.
Calcd for C18H33N2PTi (356.3): C, 60.67; H, 9.33; N, 7.86.
Found: C, 61.58; H, 9.15; N, 8.13.
P r ep a r a tion of th e Zir con iu m Com p lex 15. A solution
of 117 mg (0.50 mmol) of 7 in 7 mL of THF was added dropwise
to 214 mg (0.47 mmol) of Cl2Zr(NEt2)2(THF)2 in 3 mL of THF
at 0 °C. Over 2 h the reaction mixture was stirred at 0 °C.
The solvent was removed in vacuo, and the oily residue was
treated with 10 mL of pentane. After removal of LiCl and
evaporation of the solvent 148 mg (69%) of 15 was obtained
Syn th esis of th e Tita n iu m Com p lex 12. A solution of 120
mg (0.50 mmol) of 8 in 7 mL of THF was added dropwise to
98 mg (0.48 mmol) of Cl2Ti(NMe2)2 in 3 mL of THF at 0 °C.
Over 2 h the reaction mixture was stirred at 0 °C. The solvent
was removed in vacuo, and the oily residue was treated with
10 mL of pentane. After filtration of LiCl and evaporation of
the solvent, 108 mg (56%) of 12 was obtained as a red-brown
oil. 1H NMR (599.8 MHz, toluene-d8, 298 K): δ 5.71, 5.54 (each
m, each 1H, 3-H and 4-H), 5.43 (m, 1H, 2-H), 4.82 (m, 1H,
5-H), 3.11, 2.91 (each s, each 6H, NCH3), 3.08 (s, 2H, 6-H and
1′-H), 1.82 and 1.27 (bm, each 2H, 2′-H), 1.75, 1.18 (bm, each
2H, 3′-H), 1.62 (bm, 2H, 4′-H), 1.22 (s, 9H, C(CH3)3). 13C{1H}
NMR (150.8 MHz, toluene-d8, 298 K): δ 114.3 (C-1), 112.0
(1J CH ) 173 Hz, C-5), 111.0 and 110.8 (1J CH ) 175 Hz, 174 Hz,
1
as an orange oil. H NMR (599.8 MHz, toluene-d8, 298 K): δ
5.92 (m, 2H, 3-H and 4-H), 5.47 (m, 2H, 2-H and 5-H), 3.26,
3.15 (each m, each 4H, NCH2), 1.99 (m, 2H, 2′-H), 1.76 (m,
3
2H, 3′-H), 1.62 (d, J PH ) 10 Hz, 6H, C(CH3)2), 1.60 (m, 1H,
1′-H), 1.58 (m, 1H, 4′-H), 1.47 (m, 2H, 2′-H′), 1.26 (m, 2H, 3′-
H′), 1.23 (m, 1H, 4′-H′), 0.93 (t, 3J HH ) 6.0 Hz, 6H, NCH2CH3).
13C{1H} NMR (150.8 MHz, toluene-d8, 298 K): δ 124.0 (C-1),
3
108.1 (C-3 and C-4), 106.4 (d, J PC ) 4 Hz, C-2 and C-5), 42.5
3
2
(d, J PC < 2 Hz, NCH2), 36.0 (d, J PC ) 16 Hz, C-2′), 35.6 (d,
1J PC ) 38 Hz, C-1′), 33.2 (C-6), 30.9 (d, 2J PC ) 13 Hz, C(CH3)2),
27.6 (d, J PC ) 11 Hz, C-3′), 26.9 (C-4′). 31P{1H} NMR (81.0
3
1
3
MHz, toluene-d8): δ -11.5 (298 K). Anal. Calcd for C22H41N2-
PZr (455.8): C, 57.98; H, 9.07; N, 6.15. Found: C, 58.27; H,
8.61; N, 4.91.
C-3 and C-4), 108.7 (d, J CH ) 175 Hz, J CP ) 4 Hz, C-2), 48.6
1
1
1
(d, J CH ) 131 Hz, NCRH3), 47.9 (d, J CH ) 129 Hz, J CP ) 30
Hz, C-6), 45.1 (s, 1J CH ) 133 Hz, NCâH3), 44.8 (1J CH ) 132 Hz,
C-1′), 34.6 (d, 1J CP ) 9 Hz, CMe3), 29.1 (d, 1J CH ) 124 Hz, 3J CP
Syn th esis of th e Zir con iu m Com p lex 16. A solution of
120 mg (0.50 mmol) of 8 in 7 mL of THF was added dropwise
to 214 mg (0.47 mmol) of Cl2Zr(NEt2)2(THF)2 in 3 mL of THF
at 0 °C. Over 2 h the reaction mixture was stirred at 0 °C.
The solvent was removed in vacuo, and the oily residue was
treated with 10 mL of pentane. After filtration of LiCl and
evaporation of the solvent 143 mg (59%) of 16 was obtained
as a deep orange oil. 1H NMR (599.8 MHz, toluene-d8, 298 K):
δ 6.02, 5.56 (each m, each 1H, 3-H and 4-H), 5.71, 5.06 (each
m, each 1H, 5-H and 2-H), 3.31 (s, 1H, 6-H), 3.30 (ABX3, m,
2H, NRCH2CH3), 3.22 (ABX3, 4H, NâCH2CH3), 3.15 (ABX3, 2H,
NRCH2CH3), 1.22 (s, 9H, C(CH3)3), 0.95 (t, 6H, NRCH2CH3), 0.94
(t, 6H, NâCH2CH3) 1.8-0.9 (br, Cy H, 11H). 13C{1H} NMR
(150.8 MHz, toluene-d8, 298 K): δ 117.5 (C-1), 110.3 and 108.6
(C-2 and C-5), 108.1 and 107.8 (C-3 and C-4), 57.2 (C-6), 43.5
(NâCH2CH3), 41.4 (NRCH2CH3), 28.7 (C(CH3)3), 15.9 (Nâ-
CH2CH3), 15.6 (NRCH2CH3). Cy carbon signals were not
resolved. 31P{1H} NMR (81.0 MHz, toluene-d8): δ -69.2. Anal.
Calcd for C24H45N2PZr (483.8): C, 59.58; H, 9.37; N, 5.79.
Found: C, 58.35; H, 9.90; N, 5.50.
1
3
) 6 Hz, C(CH3)3), 28.3 (d, J CH ) 125 Hz, J CP ) 11 Hz, C-3′),
27.2 (C-4′). 31P{1H} NMR (202.6 MHz, toluene-d8): δ -25.5
(353 K), -28.7 (333 K), -32.7 (313 K), -36.3 (298 K), -41.0
(b, 273 K), -46.9 (253 K), -53.0 (233 K), -59.3 (213 K), -65.5
(193 K), -68.5 (183 K). Anal. Calcd for C20H37N2PTi (384.4):
C, 62.49; H, 9.70. Found: C, 62.27; H, 9.66.
P r ep a r a tion of th e Tita n iu m Com p lex 13. A solution of
114 mg (0.50 mmol) of 9 in 7 mL of THF was added dropwise
to 98 mg (0.48 mmol) of Cl2Ti(NMe2)2 in 3 mL of THF at 0 °C.
Over 2 h the reaction mixture was stirred at 0 °C. The solvent
was removed in vacuo, and the oily residue was treated with
10 mL of pentane. After filtration of LiCl and evaporation of
the solvent 108 mg (56%) of 13 was obtained as a red-brown
1
oil. H NMR (599.8 MHz, toluene-d8, 298 K): δ 7.31 (m, 2H,
3′-H), 7.04 (m, 2H, 2′-H), 6.96 (m, 1H, 4′-H), 5.71 (m, 2H, 3-H
and 4-H), 5.28 (m, 2H, 2-H and 5-H), 3.00 (s, 12H, NCH3), 1.46
(d, J PH ) 11 Hz, 6H, C(CH3)2). 13C{1H} NMR (599.8 MHz,
3
1
toluene-d8, 298 K): δ 143.0 (d, J PC ) 51 Hz, C-1′), 136.1 (d,
1J CH ) 155 Hz, J PC ) 15 Hz, C-3′), 126.7 (d, J CH ) 155 Hz,
2J PC ) 4.5 Hz, C-2′), 126.2 (1J CH ) 160 Hz, C-4′), 123.1 (C-1),
111.3 (1J CH ) 175 Hz, C-3 and C-4), 107.7 (d,1J CH ) 172 Hz,
3J PC ) 3 Hz, C-2 and C-5), 46.7 (d, 1J CH ) 134 Hz,3J PC ) 5 Hz,
3
1
P r ep a r a tion of th e Zir con iu m Com p lex 17. A solution
of 114 mg (0.50 mmol) of 9 in 7 mL of THF was added dropwise
to 208 mg (0.48 mmol) of Cl2Zr(NEt2)2(THF)2 in 3 mL of THF
at 0 °C. Over 2 h the reaction mixture was stirred at 0 °C.
The solvent was removed in vacuo, and the oily residue was
treated with 10 mL of pentane. After filtration of LiCl and
evaporation of the solvent 146 mg (65%) of 17 was obtained
1
1
NCH3), 32.2 (d, J PC ) 36 Hz, C-6), 29.7 (d, J CH ) 124 Hz,
2J PC ) 15 Hz, CH3). 31P{1H} NMR (202.6 MHz, toluene-d8): δ
3.0 (ν1/2 ) 5 Hz, 298 K), -0.7 (ν1/2 ) 17 Hz, 273 K), -6.9 (ν1/2
) 30 Hz, 233 K), -10.0 (ν1/2 ) 27 Hz, 213 K), -13.0 (ν1/2 ) 45
Hz, 193 K). Anal. Calcd for C18H27N2PTi (350.3): C, 61.72; H,
7.77. Found: C, 62.17; H, 7.55.
1
as an orange oil. H NMR (599.8 MHz, toluene-d8, 298 K): δ
7.56 (m, 2H, 2′-H), 7.06 (m, 2H, 3′-H), 6.96 (m, 1H, 4′-H), 5.98
(m, 2H, 3-H and 4-H), 5.54 (m, 2H, 2-H and 5-H), 3.29, 3.19
Syn th esis of th e Tita n iu m Com p lex 14. A sample of 125
mg (0.50 mmol) of 6 in 5 mL of THF was treated with 53 mg
(0.50 mmol) of LDA in 3 mL of THF at 0 °C, and the mixture
was stirred for 1 h. The solution of the in situ generated
dilithium salt 10 was then added dropwise to a solution of 98
mg (0.48 mmol) of Cl2Ti(NMe2)2 in 3 mL of THF at 0 °C. Over
2 h the reaction mixture was stirred at 0 °C. The solvent was
removed in vacuo, and the oily residue was treated with 10
mL of pentane. After filtration of LiCl and evaporation of the
solvent 152 mg (79%) of 14 was obtained as a brown oil. 1H
NMR (599.8 MHz, toluene-d8, 253 K): δ 7.28, 7.03 (each m,
each 2H, 2′-H and 3′-H), 6.93 (m, 1H, 4′-H), 5.69 (m, 1H, Cp
H), 5.21 (m, 1H, Cp H), 5.38 (m, 1H, Cp H), 5.10 (m, 1H, Cp-
H), 3.47 (d, 2J PH ) 3 Hz, 1H, 6-H), 2.96, 2.63 (each s, each 6H,
NCH3), 1.26 (s, 9H, C(CH3)3). 13C{1H} NMR (100.1 MHz,
benzene-d6, 298 K): δ 131.5 (C-2′), 127.5 (C-3′), 125.0 (C-4′),
112.9, 112.5, 112.0 (each Cp C), 49.0 (C-6), 46.0 (NCH3), 29.0
(C(CH3)3). 31P{1H} NMR (202.6 MHz, toluene-d8): δ -54.4 (298
K), -55.2 (288 K), -57.5 (253 K), -60.0 (213 K), -61.3 (193
3
(ABX3, each m, each 4H, NCH2CH3), 1.60 (d, J PH ) 11 Hz,
3
6H, C(CH3)2), 0.92 (t, J HH ) 6.9 Hz, 12H, NCH2CH3). 13C-
1
{1H} NMR (150.8 MHz, toluene-d8, 298 K): δ 142.5 (d, J PC
)
45 Hz, C-1′), 136.3 (d, 1J CH ) 158 Hz, 2J PC ) 14 Hz, C-2′), 129.2
(1J CH ) 157 Hz, C-4′), 127.6 (d, J CH ) 162 Hz, J PC ) 5 Hz,
C-3′), 125.8 (C-1), 108.6 (1J CH ) 173 Hz, C-3 and C-4), 106.7
(d, 1J CH ) 171 Hz, 3J PC ) 4 Hz, C-2 and C-5), 42.7 (1J CH ) 131
1
3
1
1
Hz, NCH2CH3), 34.2 (d, J PC ) 22 Hz, C-6), 31.4 (d, J CH
)
124 Hz, J PC ) 14 Hz, C(CH3)2), 15.8 (1J CH ) 127 Hz,
NCH2CH3). 31P{1H} NMR (81.0 MHz, toluene-d8): δ -22.5 (298
K). Anal. Calcd for C22H35N2PZr (449.7): C, 58.75; H, 7.84; N,
6.23. Found: C, 57.10; H, 7.71; N, 5.81.
2
P r ep a r a tion of th e Zir con iu m Com p lex 18. A sample
of 125 mg (0.50 mmol) of 6 in 5 mL of THF was treated with
53 mg (0.50 mmol) of LDA in 3 mL of THF at 0 °C, and the
mixture was stirred for 1 h. The solution of 10 was then added
dropwise to 214 mg (0.47 mmol) of Cl2Zr(NEt2)2(THF)2 in 3
mL of THF at 0 °C. Over 2 h the reaction mixture was stirred

1844 Organometallics, Vol. 23, No. 8, 2004
Bredeau et al.
at 0 °C. The solvent was removed in vacuo, and the oily residue
was treated with 10 mL of pentane. After filtration of LiCl
and evaporation of the solvent 174 mg (59%) of 18 was
carried out, yielding only real vibrational frequencies for the
minima and exactly one imaginary normal mode (in the range
80-120i cm-1) for the transition states. The calculations of
the NMR chemical shifts have been performed also at the DFT-
BP86/TZVP level using the GIAO method.
P olym er iza tion Rea ction s. (a ) Eth en e P olym er iza tion .
A glass autoclave was charged with 200 mL of toluene and 20
mL of a 10% solution of MAO in toluene and then thermostated
for 1 h; the solution was then saturated with ethene (at 1 or
2 bar) for 45 min. The catalyst was dissolved in toluene and
directly injected in the autoclave. The polymerization reaction
was stopped by quenching with 20 mL of aqueous HCl/
methanol (1:1 v/v). The resulting polymer was collected by
filtration, washed subsequently with HCl and water, and dried
at 80 °C in vacuo overnight. Melting points of the polyethylene
samples were measured by DSC.
(b) Eth en e-1-Octen e Cop olym er iza tion . A glass auto-
clave was charged with 30 mL of toluene, 50 mL of 1-octene,
and 20 mL of a 10% solution of MAO in toluene and then
thermostated for 1 h and the solution saturated with ethene
(at 1 or 2 bar) for 45 min. The catalyst was dissolved in toluene
and directly injected into the autoclave. The polymerization
reaction was stopped by quenching with 20 mL of aqueous HCl/
methanol (1:1 v/v). The reaction mixture was filtered and
washed with water, the solvent was removed in vacuo, and
the copolymer was dried at 80 °C in vacuo overnight.
(c) P olym er Ch a r a cter iza tion s. 13C NMR spectra of the
copolymer samples (100 mg) were obtained in benzene-d6 (0.5
mL) at 25 °C (oils) or tetrachloroethane-d2 (0.5 mL) at 80 °C
(solids). The ethene to 1-octene ratio was determined by
integration of 13C NMR signals (see Figure 3) using the
equations
1
obtained as a deep orange oil. H NMR (599.8 MHz, toluene-
d8, 298 K): δ 7.36, 7.18 (each m, each 2H, 2′-H and 3′-H), 6.82
(m, 1H, 4′-H), 5.90 (m, 1H, Cp H), 5.81 (m, 1H, Cp H), 5.79
2
(m, 1H, Cp H), 5.27 (m, 1H, Cp-H), 3.62 (d, J PH ) 4 Hz, 1H,
6-H), 3.19 (ABX3, m, 4H, NRCH2CH3), 2.94 (ABX3, 4H, NâCH2-
3
CH3), 1.27 (s, 9H, C(CH3)3), 0.98 (t, J CH ) 7 Hz, 6H,
3
NRCH2CH3), 0.56 (t, J CH ) 7 Hz, 6H, NâCH2CH3). 13C{1H}
NMR (100.1 MHz, benzene-d6, 298 K): δ 130.0 (C-2′), 127.0
(C-3′), 116.0 (C-1), 110.5, 110.0, 108.5, 107.0 (each Cp C), 47.0
(C-6), 42.0 (NâCH2CH3), 34.9 (CMe3), 34.0 (NRCH2CH3), 28.0
(C(CH3)3), 16.0 (NâCH2CH3), 15.5 (NRCH2CH3). C-4′ was not
observed. 31P{1H} NMR (202.6 MHz, toluene-d8): δ -80.7 (298
K), -81.2 (253 K), -80.7 (213 K), -80.3 (193 K). Anal. Calcd
for C24H39N2PZr (477.8): C, 60.33; H, 8.23; N, 5.86. Found:
C, 58.07; H, 8.90; N, 5.67.
Tech n ica l Deta ils of th e Qu a n tu m -Ch em ica l Ca lcu la -
tion s. The calculations have been performed with the TUR-
BOMOLE suite of programs.31 All structures have been fully
optimized without any symmetry restrictions at the density
functional (DFT) level, employing the BP86 functional,32
a
large Gaussian AO basis of valence-triple-ú quality with one
set (two for P) of polarization functions (TZVP: C, N [5s3p1d],
H [3s1p], P [5s5p2d], Ti [6s4p3d]),33,34 and the resolution-of-
the-identity (RI) approximation to represent the Coulomb
operator.34,35 For zirconium
a [5s3p3d] basis set and an
effective core potential with 28 core electrons36 has been used.
Further calculations of the Cartesian second derivatives were
(31) Ahlrichs, R.; Ba¨r, M.; Baron, H.-P.; Bauernschmitt, R.; Bo¨cker,
S.; Ehrig, M.; Eichkorn, K.; Elliott, S.; Furche, F.; Haase, F.; Ha¨ser,
M.; Horn, H.; Huber, C.; Huniar, U.; Kattannek, M.; Ko¨lmel, C.;
Kollwitz, M.; May, K.; Ochsenfeld, C.; O¨ hm, H.; Scha¨fer, A.; Schneider,
U.; Treutler, O.; von Arnim, M.; Weigend, F.; Weis, P.; Weiss, H.
TURBOMOLE (version 5.5); Universita¨t Karlsruhe, Karlsruhe, Ger-
many, 2002.
R ) I(Cδδ+,C3B)/I(CRδ+,C6B) ) I(Cδδ+,C3B)/I(Câδ+,C5B
ethene/1-octene ) [(3R - 1)/2] + 2
)
(1)
(2)
Ackn owledgm en t. Financial support from the Fonds
der Chemischen Industrie and the Deutsche For-
schungsgemeinschaft is gratefully acknowledged.
(32) Becke, A. D. Phys. Rev. A 1988, 38, 3098-3100. Perdew, J . P.
Phys. Rev. B 1986, 33, 8822-8824.
(33) Scha¨fer, A.; Horn, H.; Ahlrichs, R. J . Chem. Phys. 1992, 97,
2571-2577.
(34) All basis sets are available from the TURBOMOLE homepage,
http://www.turbomole.com, via FTP Server Button (in the subdirecto-
ries basen (AO basis sets) and jbasen (RI basis sets)).
(35) Eichkorn, K.; Treutler, O.; O¨ hm, H.; Ha¨ser, M.; Ahlrichs, R.
Chem. Phys. Lett. 1995, 240, 283-290.
(36) Andrae, D.; Haeussermann, U.; Dolg, M.; Stoll, H.; Preuss, H.
Theor. Chim. Acta 1990, 77, 123-141.
Su p p or tin g In for m a tion Ava ila ble: Text and a table
giving details of the DFT calculation of 21 and 13C NMR data
of the compounds 3-9. This material is available free of charge
via the Internet at http://pubs.acs.org.
OM034201Y