((Diaryl- and Dialkylphosphino)alkyl)cyclopentadienyl Ligands and Their Use in the Preparation of Heterobinuclear Ti/Mo and Zr/Mo Complexes

Article abstract of DOI:10.1021/om990279w

The syntheses of a series of bifunctional ligands, in which a cyclopentadienyl and a phosphine group are linked by either a CH₂ or a C₂H₄ fragment (^-C₅H₄(CH₂)_nPR

Full text of DOI:10.1021/om990279w

3490
Organometallics 1999, 18, 3490-3501
((Dia r yl- a n d Dia lk ylp h osp h in o)a lk yl)cyclop en ta d ien yl
Liga n d s a n d Th eir Use in th e P r ep a r a tion of
Heter obin u clea r Ti/Mo a n d Zr /Mo Com p lexes
Todd W. Graham, Angela Llamazares, Robert McDonald,† and Martin Cowie*
Department of Chemistry, University of Alberta, Edmonton, AB, Canada T6G 2G2
Received April 21, 1999
The syntheses of a series of bifunctional ligands, in which a cyclopentadienyl and a
phosphine group are linked by either a CH2 or a C2H4 fragment (-C5H4(CH2)nPR2; n ) 1, 2;
R ) Me, Ph), are reported, as is that of the related ligand, Li2[Me2C(C5H3CMe2PPh2)2], in
which both substituted cyclopentadienyl rings are linked by the CMe2 group. The metallocene
dichloride complexes Cp′2MCl2 (M ) Ti, Zr; Cp′2 ) 2C5H4(CH2)nPPh2 (n ) 1, 2), Me2C-
((C5H3)C(Me)2PPh2)2) have also been prepared, and the X-ray structure of (η5-C5H4(CH2)2-
PPh2)2ZrCl2 (7) has been determined. This complex has the pendent phosphinoalkyl arms
close to eclipsed on the two C5H4 groups and bisecting the ZrCl2 angle, but bent in opposite
directions away from the ZrCl2 plane. Reaction of 7 and its Ti analogue with (COD)Mo(CO)4
yields the heterobinuclear complexes [(µ-η5:η1-C5H4(CH2)2PPh2)2MCl2Mo(CO)4] (M ) Ti, Zr).
Structure determinations of these bimetallic complexes show the expected cis arrangement
of phosphine moieties at Mo, M-Mo separations of 6.895 Å (M ) Ti) and 6.945 Å (M ) Zr),
and MCl2 moieties aimed at right angles to the M-Mo vectors.
In tr od u ction
The interest in early-late heterobimetallic (ELHB)
complexes derives primarily from the goal of utilizing
two metals, having widely divergent properties, to
induce transformations that would not be possible
through the use of either metal alone. When these early
and late transition metals are in close proximity, their
differing properties should result in a polar bifunctional
environment that may be capable of displaying novel
reactivity resulting from some form of cooperative
interaction between the two metals.1,2 Although a
number of routes can be used for the preparation of
ELHB complexes,3-7 one commonly used method in-
volves the use of a bifunctional ligand template that is
capable of binding strongly to both metal types. This
eliminates the need for metal-metal bonds to hold the
metals close enough for a cooperative interaction and
helps maintain the integrity of the complex during
reactions of interest. A variety of bifunctional ligand
types have been used, three of which are shown in
structures A,2a,4 B,5 and C.6 In each type shown, the
phosphine functionality forms favorable interactions
with the late metal, while the anionic functionality binds
to the early metal.
The ubiquity of early-metal cyclopentadienyl and late-
metal phosphine complexes prompted us to use a ligand
similar to the type shown above in C. However, we felt
that using the ligand in C, in which the PR2 moiety is
bound directly to the cyclopentadienyl ring,6 would
result in geometric constraints on the complex that
might inhibit reactivity by holding the metals together
in too rigid an environment. It seemed that the intro-
duction of an alkyl spacer between the cyclopentadienyl
and phosphino groups might alleviate these restrictions,
(3) See for example: (a) Hamilton, D. M.; Willis, W. S.; Stucky, G.
D. J . Am. Chem. Soc. 1981, 103, 4255. (b) Marsella, J . A.; Huffman,
K. G.; Caulton, K. G.; Longato, B.; Norton, J . R. J . Am. Chem. Soc.
1982, 104, 6360. (c) Casey, C. P.; J ordan, R. F.; Rheingold, A. L. J .
Am. Chem. Soc. 1983, 105, 665. (d) Memmler, H.; Walsh, K.; Gade,
H.; Lauher, J . W. Inorg. Chem. 1995, 34, 4062. (e) Katti, K. V.; Cavell,
R. G. Organometallics 1991, 10, 539. (f) Proulx, G.; Bergmann, R. G.
J . Am. Chem. Soc. 1996, 118, 1981. (g) Gelmini, L.; Stephan, D. W.
Organometallics 1988, 7, 849. (h) Baranger, A. M.; Hollander, F. J .;
Bergman, R. G. J . Am. Chem. Soc. 1993, 115, 7890. (i) Bruno, J . W.;
Huffman, J . C.; Green, M. A.; Caulton, K. G. J . Am. Chem. Soc. 1984,
106, 8310. (j) Ozawa, F.; Park, J . W.; Makenzie, P. B.; Schaefer, W. P.;
Henling, L.; Grubbs, P. H. J . Am. Chem. Soc. 1989, 111, 1319. (k)
Baker, R. T.; Fultz, W. C.; Marder, T. B.; Williams, I. D. Organome-
tallics 1990, 9, 2357. (l) Casey, C. P. J . Organomet. Chem. 1990, 400,
205. (m) Lemke, F. R.; Szalda, D. J .; Bullock, R. M. J . Am. Chem. Soc.
1991, 113, 8466. (n) McFarland, J . M.; Churchill, M. R.; See, R. F.;
Lake, C. H.; Atwood, J . D. Organometallics 1991, 10, 3530. (o) Sartain,
W. J .; Selegue, J . P. Organometallics 1989, 8, 2153.
* To whom correspondence should be addressed.
† Faculty Service Officer, Structure Determination Laboratory.
(1) (a) Bullock, R. M.; Casey, C. P. Acc. Chem. Res. 1987, 20, 167.
(b) Tauster, S. J . Acc. Chem. Res. 1987, 20, 389. (c) Stephan, D. W.
Coord. Chem. Rev. 1989, 95, 41. (d) Chetcuti, M. J . Comprehensive
Organometallic Chemistry II, Vol. 10; Abel, E. W., Stone, F. G. A.,
Wilkinson, G., Eds.; Pergamon: Exeter, U.K., 1995.
(2) See for example: (a) Ferguson, G. S.; Wolczanski, P. T.; Pa´rka´nyi,
L.; Zonnevylle, M. C. Organometallics 1988, 7, 1967. (b) Zheng, P. Y.;
Nadsadi, T. T.; Stephan, D. W. Organometallics 1989, 8, 1393. (c)
Baranger, A. M.; Bergman, R. G. J . Am. Chem. Soc. 1994, 116, 3822.
(d) Steffey, B. D.; Vites, J . C.; Cutler, A. R. Organometallics 1991, 10,
3432.
(4) (a) Baxter, S. M.; Ferguson, G. S.; Wolczanski, P. T. J . Am. Chem.
Soc. 1988, 110, 4237. (b) Slaughter, L. M.; Wolczanski, P. T. Chem.
Commun. 1977, 2109.
(5) (a) Choukron, R.; Dahan, F.; Gervais, D.; Rifai, C. Organome-
tallics 1990, 9, 1982. (b) Choukron, R.; Gervais, D.; J aud, J .; Kalck,
R.; Senocq, F. Organometallics 1986, 5, 67.
10.1021/om990279w CCC: $18.00 © 1999 American Chemical Society
Publication on Web 07/30/1999

Heterobinuclear Ti/ Mo and Zr/ Mo Complexes
Organometallics, Vol. 18, No. 17, 1999 3491
P r ep a r a tion of Com p ou n d s. (a ) K[C5H4CH2CH2P P h 2]
(1). A solution of 8.9 g (0.040 mol) of KPPh2 in 220 mL of THF
was added to 4.5 mL (0.044 mol) of spiro[2.4]hepta-4,6-diene
in 450 mL of THF at -80 °C. The solution was warmed to
room temperature and stirred for 3 h, during which time a
white crystalline solid was produced. The solution was con-
centrated in vacuo to ca. two-thirds of the original volume,
and the precipitate was collected on a glass frit. Yield: 9.0 g
(72%) based on KPPh2. The compound was obtained spectro-
scopically pure, but elemental analyses could not be obtained
due to its highly air-sensitive nature.
(b) Li[C5H4CH2CH2P Me2] (2). A suspension of 1.694 g
(0.025 mol) of LiPMe2 in 10 mL of n-pentane was cooled to
-80 °C, and then 40 mL of THF was added. This solution was
added to 2.8 mL (0.027 mol) of spiro[2.4]hepta-4,6-diene in 20
mL of THF at -80 °C, and then the solution was warmed to
room temperature. After stirring for 3 h, the solvent was
removed in vacuo, and then the white solid was triturated in
50 mL of n-pentane, followed by collection on a glass frit and
washing with 3 × 10 mL of n-pentane. The solid was dried in
vacuo to give 2.792 g (70%) of a spectroscopically pure product.
Elemental analyses were not obtained due to the highly air-
sensitive nature of the compound.
allowing greater conformational freedom of the two
metal coordination spheres. Accordingly, we have de-
cided to use (phosphinoalkyl)cyclopentadienyl ligands,
containing a “C1” or “C2” spacer between the Cp and PR2
functionalities as a template upon which to prepare our
ELHB complexes. This approach has been used by a few
others.7
Our strategy for the preparation of the ELHB com-
plexes was to first synthesize early-metal metallocene-
like species, in which the cyclopentadienyl rings are
derivatized with a phosphinoalkyl moiety, and then to
react these early-metal-containing monomers with late-
metal sources. One aspect of interest was the elucidation
of the effect of the different spacer lengths in the C5H4-
(CR2)nPR2 ligands (n ) 1 or 2) on the structures of the
ELHB products and a comparison of these data with
the known ELHB complexes containing the bridging
ligand shown in C above.
Exp er im en ta l Section
(c) K[C5H4CH2P P h 2] (3). A 16.0 g (0.071 mol) sample of
KPPh2 in a mixture of 100 mL of THF and 300 mL of Et2O
was added to a solution of 4.45 g (0.057 mol) of fulvene in 300
mL of Et2O at -80 °C, resulting in the immediate formation
of an off-white precipitate. The mixture was warmed to room
temperature with vigorous stirring, and then the solid was
collected on a frit. Yield: 13.6 g, 63% based on KPPh2. The
compound was obtained spectroscopically pure, but elemental
analyses were not obtained due to its highly air-sensitive
nature.
(d ) [Me2C((C5H3dC(CH3)2)2] (4). This compound was
prepared as previously reported,12 with minor modifications.
A 4.36 mL (0.052 mol) sample of pyrrolidine was added to 1.853
g (0.011 mol) of (C5H5)2CMe2 and 1.6 mL (0.022 mol) of acetone
in 26 mL of a 10:1 mixture of MeOH/Et2O at 5 °C. The solution
was stirred at room temperature for 16.5 h, and then 3.2 mL
(0.0559 mol) of acetic acid was added at -10 °C. A 30 mL
portion of water and 30 mL of Et2O were added in air, and
the organic layer was removed. The aqueous layer was washed
with 3 × 20 mL of Et2O, and then the organic portions were
combined and washed with 3 × 20 mL of water and 1 × 20
mL of brine and then dried over Na2SO4 and molecular sieves.
The solvent was removed under reduced pressure, and the
crude yellow oil eluted through a 600 mL fritted glass filter
containing an 8 cm thick plug of silica. The solvent was
removed under reduced pressure yielding 2.243 g, 82% of a
yellow oil.
(e) Li2[Me2C(C5H3C(CH3)2P P h 2)2] (5). A 2.20 g (8.72
mmol) sample of fulvene 4 in 50 mL of THF was added to a
solution of LiPPh2 (17.2 mmol) in 80 mL of THF at -80 °C
over 5 min. The solution was warmed to room temperature
and stirred for 16 h, and then the solvent was removed in
vacuo. The orange foamy residue was triturated with 100 mL
of n-pentane until a powdery off-white solid was obtained,
which was collected on a glass frit and washed with 3 × 20
mL of n-pentane, followed by drying in vacuo. Yield: 7.69 g,
70% based on LiPPh2. The compound was obtained spectro-
scopically pure, but elemental analyses could not be obtained
due to the highly air-sensitive nature of the compound.
(f) [(η5-C5H4CH2CH2P P h 2)2TiCl2] (6). Meth od 1: A solu-
tion of 0.866 g (2.59 mmol) of TiCl4(THF)2 in 50 mL of THF
was added dropwise to 2.246 g (7.10 mmol) of K[C5H4CH2CH2-
PPh2] (1) in 110 mL of THF over 2 h. After stirring overnight,
a red solution had formed (if the solution was brown at this
point, concentrated HCl could be added dropwise until a red
solution formed), and the solvent was removed in vacuo. The
residue was extracted with 3 × 20 mL of toluene, the extracts
Gen er a l Com m en ts. All reactions were carried out under
an atmosphere of prepurified argon with standard Schlenk
techniques or in a nitrogen-filled Vacuum Atmospheres glove-
box equipped with an HE-493 dri-train. Solvents were dried
and distilled under nitrogen immediately before use. Sodium
benzophenone was used as the drying agent for all solvents
except CH2Cl2, which was distilled from P2O5. Group 4 metal
salts and diphenylphosphine were purchased from Strem or
Aldrich. Diphenylphosphine and TiCl4 were used as received,
and ZrCl4 was sublimed immediately before use. Spiro[2.4]-
hepta-4,6-diene,8 fulvene,9 and [Mo(CO)4(COD)]10 were pre-
pared via literature methods. KPPh2 was prepared by the
reaction of HPPh2 with excess KH in THF solution, and LiPPh2
was prepared via the reaction of HPPh2 with 1 equiv of n-BuLi
(2.5 M in hexanes) in THF. Dimethylphosphine11 was prepared
via the reaction of tetramethyldiphosphine disulfide with
LiAlH4. The product was distilled from the reaction vessel
directly into a flask containing ca. 1 molar equiv of n-BuLi in
50 mL of n-pentane that had been cooled to -80 °C, and the
resulting LiPMe2 was collected on a glass frit.
The 1H, 31P{1H}, and 13C{1H} NMR spectra were recorded
on a Bruker AM-400 spectrometer operating at 400.1, 162.0,
and 100.6 MHz for the respective nuclei. The internal deuter-
ated solvent served as a lock for the spectrometer. Elemental
analyses were performed by the microanalytical service within
the department. NMR spectroscopic data for all compounds
are given in Table 1, while infrared data for the appropriate
compounds are reported together with the details of their
preparation.
(6) (a) Bakhmutov, V. I.; Visseaux, M.; Baudry, D.; Dormond, A.;
Richard, P. Inorg. Chem. 1996, 35, 7316. (b) Tikkanen, W.; Fujita, Y.;
Petersen, J . Organometallics 1986, 5, 888. (c) He, X.-D.; Maisonnat,
A.; Dahan, F.; Poilblanc, R. Organometallics 1989, 8, 2618. (d)
Szymoniak, J .; Kubicki, M. M.; Besancon, J .; Moise, C. Inorg. Chim.
Acta 1991, 180, 153. (e) Casey, C.; Nief, F. Organometallics 1985, 4,
1218. (f) Anderson, G. K.; Lin, M. Organometallics 1988, 7, 2285. (g)
Rausch, M. D.; Edwards, B. H.; Rogers, R. D.; Atwood, J . L. J . Am.
Chem. Soc. 1983, 105, 3882. (h) Baudry, D.; Dormond, A.; Visseaux,
M.; Monot, C.; Chardot, H.; Lin, Y.; Bakhmutov, V. New. J . Chem.
1995, 19, 921. (i) Schore, N. E. J . Am. Chem. Soc. 1979, 101, 7410. (j)
Tueting, D. R.; Iyer, S. R.; Schore, N. E. J . Organomet. Chem. 1987,
320, 349. (k) Kettenbach, R. T.; Bonrath, W.; Butensha¨n, H. Chem.
Ber. 1993, 126, 1657.
(7) (a) Charrier, C.; Mathey, M. J . Organomet. Chem. 1979, 170,
C41. (b) Bosch, B.; Erker, G.; Frohlich, R. Inorg. Chim. Acta 1998, 270,
446. (c) LeBlanc, J . C.; Moise, C.; Maisonnat, A.; Poilblanc, R.; Charrier,
C.; Mathey, F. J . Organomet. Chem. 1982, 231, C43.
(8) Wilcox, C.; Craig, R. J . Am. Chem. Soc. 1961, 83, 3886.
(9) Neuenschwander, M.; Iseli, R. Helv. Chim. Acta 1977, 60, 1061.
(10) King, R. B. J . Organomet. Chem. 1967, 8, 139.
(11) Parshall, G. W. Inorg. Synth. 1968, 11, 157.
(12) Carter, C. A. G. Ph.D. Thesis, University of Alberta, 1998.

3492 Organometallics, Vol. 18, No. 17, 1999
Ta ble 1. NMR Sp ectr oscop ic Da ta for Com p ou n d sa
NMRb
Graham et al.
compound
K[C5H4CH2CH2PPh2] (1)c
δ(31P{1H}) (ppm)
-16.1(s)
δ(1H) (ppm)
3
2.82(dt, 2H, J HH ) 7.0 Hz, J HP ) 11.1 Hz, C5H4CH2CH2PPh2)
3
2.29(t, 2H, J HH ) 7.0 Hz, C5H4CH2CH2PPh2)
5.40(m, 2H, C5H4CH2CH2PPh2)
5.43(m, 2H, C5H4CH2CH2PPh2)
Li[C5H4CH2CH2PMe2] (2)c
-55.9(s)
-16.3(s)
1.61(dt, 2H, J HH ) 7.8 Hz, J HP ) 1.2 Hz, C5H4CH2CH2PMe2)
2.61(dt, 2H, J HH ) 7.8 Hz, J HP ) 10.3 Hz, C5H4CH2CH2PMe2)
5.48(m, 2H, C5H4CH2CH2PMe2)
5.53(m, 2H, C5H4CH2CH2PMe2)
3.41(s, br, 2H, C5H4CH2PPh2)
5.37(s, br, 4H, C5H4CH2PPh2)
1.49(s, 6H, CMe2)
3
3
K[C5H4CH2PPh2] (3)c
[Me2C(C5H3dCMe2)2] (4)d
2.16(s, 6H, dCMe)
2.17(s, 6H, dCMe)
6.22(m, 2H, C5H3)
6.42(m, 2H, C5H3)
6.49(m, 2H, C5H3)
3
Li2[Me2C(C5H3CMe2PPh2)2] (5)c
20.2(s)
1.30(d, 12H, J HP ) 11.8 Hz, CMe2PPh2)
1.61(s, 6H, Me2C(C5H3CMe2PPh2)2)
5.58(m, 2H, C5H3)
5.65(m, 2H, C5H3)
5.77(m, 2H, C5H3)
[(η5-C5H4CH2CH2PPh2)2TiCl2] (6)
[(η5-C5H4CH2CH2PPh2)2ZrCl2] (7)
-16.5(s)
-16.5(s)
2.36(t, 4H, CH2CH2PPh2)
2.81(dt, 4H, J HP ) 9.0 Hz, J HH ) 7.5 Hz, CH2CH2PPh2)
6.26(m, 4H, C5H4)
6.33(m, 4H, C5H4)
2.33(t, 4H, J HH ) 8.0 Hz, CH2CH2PPh2)
2.73(dt, 4H, J HP ) 9.1 Hz, J HH ) 8.0 Hz, CH2CH2PPh2)
6.18(m, 4H, C5H4)
6.25(m, 4H, C5H4)
3.51(s, br, 4H, CH2PPh2)
6.00(m, 4H, C5H4)
6.25(m, 4H, C5H4)
3.41(s, br, 4H, CH2PPh2)
5.94(m, 4H, C5H4)
[(η5-C5H4CH2PPh2)2TiCl2] (8)
-9.6(s)
-9.7(s)
32.7(s)
[(η5-C5H4CH2PPh2)2ZrCl2] (9)
6.17(m, 4H, C5H4)
3
rac-[Me2C(C5H3CMe2PPh2)2ZrCl2] (10a )
1.48(d, 6H, J HP ) 13.0 Hz, CMe2PPh2)
1.58(s, 6H, Me2C(C5H3CMe2PPh2)2
3
1.75(d, 6H, J HP ) 16.2 Hz, CMe2PPh2)
5.07(m, 2H, C5H3)
5.48(m, 2H, C5H3)
5.99(m, 2H, C5H3)
3
meso-[Me2C(C5H3CMe2PPh2)2ZrCl2] (10b)
33.2(s)
1.47(d, 6H, J HP ) 14.9 Hz, CMe2PPh2)
1.53(s, 3H, Me2C(C5H3CMe2PPh2)2
1.73(s, 3H, Me2C(C5H3CMe2PPh2)2
3
1.77(d, 6H, J HP ) 17.6 Hz, CMe2PPh2)
5.20(m, 2H, C5H3)
5.66(m, 4H, C5H3)
6.24(br, 8Η, C5H4)
2.45 (m, 4Η, CH2CH2PPh2)
2.55 (m, 4Η, CH2CH2PPh2)
6.37 (m, 4Η, C5H4)
[(µ-η5:η1-C5H4CH2CH2PPh2)2TiCl2Mo(CO)4]d (11)
[(µ-η5:η1-C5H4CH2CH2PPh2)2ZrCl2Mo(CO)4]d (12)
26.6 (s, Mo-P)
26.6 (s, Mo-P)
6.31 (m, 4Η, C5H4)
2.55 (m, 8H, CH2CH2PPh2)
a
b
Abbreviations used: for NMR (s) singlet, (d) doublet, (dt) doublet of triplets, (m) multiplet, (dd) doublet of doublets, (br) broad. NMR
spectra recorded in CD2Cl2 unless otherwise stated. c NMR spectra recorded in THF-d8. NMR spectra recorded in CDCl3; phosphorus-
d
bound phenyl groups omitted.
were filtered through Celite to remove KCl and polymeric
materials, and then the solvent was removed in vacuo. The
red residue was triturated with three 10 mL portions of Et2O,
yielding 520 mg of a red powder. Yield: 30%. Anal. Calcd for
was extracted with 4 × 10 mL of toluene, then the extracts
were filtered through Celite to remove KCl and polymeric
materials, and the solvent was removed in vacuo. The red
residue was triturated in Et2O (4 × 15 mL), resulting in the
formation of a red powder. Yield: 628 mg, 57%. The product
was spectroscopically identical to that obtained by method 1.
(g) [(η5-C5H4CH2CH2P P h 2)2Zr Cl2] (7). A 815 mg (2.58
mmol) sample of K[C5H4CH2CH2PPh2] (1) in 20 mL of toluene
was added to 300 mg (1.29 mmol) of ZrCl4 in 20 mL of toluene
at -80 °C. The solution was warmed slowly to room temper-
ature and stirred for 18 h. The solvent was removed in vacuo,
and then the residue was dissolved in 40 mL of CH2Cl2 and
filtered through Celite. If the filtrate was yellow, then a small
amount (ca. 100 mg) of activated carbon was added, followed
by stirring for 5 min and filtration through Celite to give a
colorless solution. Alternatively, the yellow impurity could be
C
36H32P2Cl2Ti: C, 67.77; H, 5.39. Found: C, 67.60; H, 5.38.
Meth od 2: A solution of 0.568 g (1.63 mmol) of TiCl3(THF)3
in 60 mL of THF was added to 1.100 g (3.47 mmol) of K[C5H4-
CH2CH2PPh2] (1) in 120 mL of THF at -80 °C over 30 min.
The solution was warmed to room temperature and stirred for
19 h, during which time the color changed to dark green. The
flask was opened to the air, and ca. 3.5 mL of concentrated
HCl was added, resulting in a color change to red over 5 min.
The solution was cooled to -40 °C, and then 2 mL of pyridine
was added to the solution, which was then warmed to room
temperature. After stirring for several minutes at room
temperature, the solvent was removed in vacuo. The residue

Heterobinuclear Ti/ Mo and Zr/ Mo Complexes
Organometallics, Vol. 18, No. 17, 1999 3493
precipitated by the addition of n-pentane to a CH2Cl2 solution
of the compound at 0 °C, followed by solvent removal via
cannula. The solvent from the filtrate was removed in vacuo,
and the colorless residue recrystallized by dissolving in 5 mL
of CH2Cl2, cooling to -20 °C, and slowly adding 30 mL of
n-pentane. The solvent was removed via cannula and the white
microcrystalline solid washed with 3 × 10 mL of n-pentane
and then dried in vacuo. Crystalline material could be obtained
by adding n-pentane to a CH2Cl2 solution of the compound at
0 °C until the solution became slightly turbid, followed by
cooling to -20 °C for 48 h. Yield: 415 mg, 45%. Anal. Calcd
for C38H36P2Cl2Zr: C, 63.68; H, 5.06. Found: C, 63.45; H, 5.01.
(h ) [(η5-C5H4CH2P P h 2)2TiCl2] (8). A 286 mg (0.823 mmol)
sample of TiCl3(THF)3 in 30 mL of THF was added to 548 mg
(1.81 mmol) of K[C5H4CH2PPh2] (3) in 60 mL of THF at -80
°C over 30 min. The solution was warmed to room temperature
and stirred for 18 h. Concentrated HCl (25 drops) was added
in air to the green solution, resulting in an immediate color
change to red. After stirring for 2 min the solvent was removed
in vacuo and the red residue extracted with 4 × 10 mL of
toluene. The extracts were filtered through Celite, and the
solvent was removed in vacuo. The red residue was triturated
with n-pentane (3 × 10 mL) until a red powder was obtained,
which was then dried in vacuo. Yield: 290 mg, 55%. Anal.
Calcd for C36H32P2Cl2Ti: C, 67.00; H, 5.00. Found: C, 66.53;
H, 5.23.
(i) [(η5-C5H4CH2P P h 2)2Zr Cl2] (9). A suspension of 368 mg
(1.22 mmol) of K[C5H4CH2PPh2] (3) in 15 mL of toluene was
added to a suspension of 136 mg (0.583 mmol) of ZrCl4 in 15
mL of toluene at -80 °C over 5 min. The suspension was
warmed slowly to room temperature and stirred for a total of
18 h. The solvent was removed in vacuo and the residue
dissolved in 20 mL of CH2Cl2, followed by filtration through
Celite. The solvent was removed in vacuo and the yellow oily
solid redissolved in 3 mL of CH2Cl2 and cooled to -10 °C. A
30 mL sample of n-pentane was added, resulting in the
formation of an oily yellow solid and milky white suspension.
The suspension was removed via cannula, the solvent volume
was reduced to half in vacuo, and the mixture was cooled to
-20 °C. A 10 mL sample of n-pentane was added, and the
mixture was stirred at -20 °C for 30 min. The solvent was
removed via cannula, and the white solid was washed with 2
× 10 mL of n-pentane and dried in vacuo. Yield: 107 mg of
an analytically pure product. The oily solid was triturated in
3 × 10 mL of n-pentane and dried in vacuo to give 49 mg of a
spectroscopically pure product. Total yield: 156 mg, 40%. Anal.
Calcd for C36H32P2Cl2Zr: C, 62.78; H, 4.68. Found: C, 62.62;
H, 4.66.
(j) [Me2C(C5H3C(CH3)2P P h 2)2Zr Cl2] (10). A 563 mg (0.884
mmol) sample of Li2[Me2C(C5H3C(CH3)2PPh2)2] (5) in 25 mL
of toluene was added to 200 mg (0.858 mmol) of ZrCl4 in 15
mL of toluene at -80 °C over 5 min. The solution was warmed
slowly to room temperature and stirred for 16.5 h. The solvent
was removed in vacuo, then 20 mL of CH2Cl2 was added, and
the yellow solution was filtered through Celite. The solvent
was removed in vacuo, and the residue was recrystallized by
dissolving in 3 mL of CH2Cl2, cooling to - 15 °C, and then
adding 20 mL of n-pentane, resulting in the formation of a
yellow-brown solid. The solution was filtered through Celite
at -15 °C, and then the solvent was removed in vacuo, yielding
a bright yellow solid. Trituration of this solid with 3 × 10 mL
of n-pentane gave a bright yellow powdery solid. Yield: 269
mg, 40%. Anal. Calcd for C43H44P2Cl2Zr: C, 65.80; H, 5.65.
Found: C, 65.17, H, 5.65. Repeated attempts to obtain
satisfactory carbon analyses were unsuccessful, presumably
owing to the difficulties encountered in removing LiCl.
(k ) [(µ-η5:η1-C5H4CH2CH2P P h 2)2TiCl2Mo(CO)4] (11). A
28.6 mg (0.091 mmol) sample of [(COD)Mo(CO)4] in 10 mL of
THF was added via cannula to 61 mg (0.091 mmol) of [(η5-
C5H4CH2CH2PPh2)2TiCl2] (6) in 10 mL of THF. The red
solution was stirred for 1 h, and then the solvent was removed
in vacuo. The red residue was recrystallized by dissolving in
5 mL of CH2Cl2 and adding 20 mL of MeOH dropwise, followed
by stirring for 1 h. The red microcrystalline solid that had
formed was washed with 2 × 10 mL of MeOH and 1 × 10 mL
of Et2O and then dried in vacuo. Yield: 30 mg, 40% of an
analytically pure product. Anal. Calcd for C42H36P2O4Cl2-
TiMo: C, 57.23; H, 4.12. Found: C, 57.11; H, 3.96. IR spectrum
(Nujol): 2023 (med, sh), 1927 (s), 1896 (vs).
(l) [(µ-η5:η1-C5H4CH2CH2P P h 2)2Zr Cl2Mo(CO)4] (12). A 72
mg (0.227 mmol) sample of [(COD)Mo(CO)4] in 30 mL of THF
was added via cannula to 163 mg (0.227 mmol) of [(η5-C5H4-
CH2CH2PPh2)2ZrCl2] (7) in 30 mL of THF. The yellow solution
was stirred for 1.5 h, and then the solvent was removed in
vacuo. The yellow residue was recrystallized by dissolving in
CH2Cl2 and filtering if necessary, followed by the dropwise
addition of MeOH until a white solid formed. The solid was
washed with 2 × 10 mL of MeOH, then 1 × 10 mL Et2O,
followed by drying in vacuo. Yield: 173 mg, 84%. Anal. Calcd
for C42H36P2O4Cl2ZrMo: C, 54.55; H, 3.92. Found: C, 54.56;
H, 3.88. IR spectrum (Nujol): 2022 (med, sh), 1928 (s), 1896
(vs).
X-r a y Da ta Collection a n d Str u ctu r e Solu tion . (a)
Crystals of 7 were obtained by layering n-pentane onto a CH2-
Cl2 solution of the compound. Data were collected to a
maximum 2θ ) 50.0° on a Siemens P4/RA diffractometer using
Mo KR radiation at -60 °C. Unit cell parameters were
obtained from a least-squares refinement of the setting angles
of 40 reflections with 25.4° < 2θ < 28.0°. The systematic
absences indicated the space group to be I2/a (a nonstandard
setting of C2/c [No. 15]). Three reflections were chosen as
intensity standards and were remeasured after every 300
reflections, with no decay evident. The data were corrected
for absorption through use of a semiempirical correction based
on azimuthal (ψ) scans of several reflections. See Table 2 for
a summary of crystal data and X-ray data collection informa-
tion.
The structure was solved by direct methods (SHELXS-8613),
and refinement was completed using the program SHELXL-
93.14 Hydrogen atoms were assigned positions based on the
geometries of their attached carbon atoms and were given
thermal parameters 20% greater than those of the attached
carbons. The final model refined to values of R1(F) ) 0.0279
(for 2707 data with Fo g 2σ(Fo2)) and wR2(F2) ) 0.0676 (for
2
all 3053 independent data).
(b) Crystals of both 11 and 12 were obtained by slow
diffusion of Et2O into CH2Cl2 solutions of the compounds. Data
were collected at -50 °C to a maximum 2θ ) 50.0° on an
Enraf-Nonius CAD4 diffractometer using Mo KR radiation.
Unit cell parameters were obtained from a least-squares
refinement of the setting angles of 24 reflections with 20.0° <
2θ < 24.0°. Both compounds crystallized in the space group
P1h (No. 2). Three reflections were chosen as intensity stan-
dards and were remeasured after every 7200 s of X-ray
exposure time; in both cases no decay was observed. Absorption
corrections were applied to the data for compound 12 using
the method of Walker and Stuart,15 whereas for 11 the crystal
faces were indexed and measured, and a Gaussian integration
was carried out. See Table 2 for a summary of crystal data
and X-ray data collection information.
(13) Sheldrick, G. M. Acta Crystallogr. 1990, A46, 467-473.
(14) Sheldrick, G. M. SHELXL-93. Program for crystal structure
determination; University of Go¨ttingen, Germany, 1993. Refinement
2
2
on Fo for all reflections except for two having Fo < -3σ(Fo2) for
compound 11 and except for three having Fo2 < -3σ(Fo2) for compound
12. Weighted R-factors wR2 and all goodnesses of fit S are based on
Fo2; conventional R-factors R1 are based on Fo, with Fo set to zero for
negative Fo2. The observed criterion of Fo > 2σ(Fo2) is used only for
2
calculating R1 and is not relevant to the choice of reflections for
refinement. R-factors based on Fo2 are statistically about twice as large
as those based on Fo, and R-factors based on ALL data will be even
larger.
(15) Walker, N.; Stuart, D. Acta Crystallogr., Sect. A 1983, 39, 158-
166.

3494 Organometallics, Vol. 18, No. 17, 1999
Graham et al.
Ta ble 2. Cr ysta llogr a p h ic Deta ils for Com p ou n d s 7, 11, a n d 12
11‚0.5CH2Cl2
7
12‚0.5CH2Cl2
A. Crystal Data
formula
fw
C
38H36Cl2P2Zr
C42.5H37Cl3MoO4P2Ti
923.85
C42.5H37Cl3MoO4P2Zr
967.17
716.73
cryst dimens (mm)
cryst system
space group
0.70 × 0.50 × 0.29
monoclinic
I2/a (a nonstandard setting
of C2/c [No. 15])
0.39 × 0.21 × 0.16
0.53 × 0.35 × 0.30
triclinic
triclinic
P1h (No. 2)
P1h (No. 2)
unit cell paramsa
a (Å)
b (Å)
26.974(3)
6.4587(9)
20.671(2)
13.336(2)
13.547(3)
12.330(2)
104.46(2)
107.996(13)
77.340(14)
2026.9(7)
2
13.4643(11)
13.553(2)
12.4283(15)
104.568(11)
108.050(8)
77.489(8)
2063.1(4)
2
c (Å)
R (deg)
â (deg)
γ (deg)
V (Å3)
Z
105.598(8)
3468.5(7)
4
1.373
0.589
Fcalcd (g cm-3
)
1.514
0.825
1.557
0.869
abs coeff (mm-1
)
B. Data Collection and Refinement Conditions
diffractometer
radiation (λ [Å])
Siemens P4/RAc
Enraf-Nonius CAD4b
graphite-monochromated
Mo KR (0.71073)
-50
Enraf-Nonius CAD4b
graphite-monochromated
Mo KR (0.71073)
-50
graphite-monochromated
Mo KR (0.71073)
-60
temperature (°C)
scan type
θ-2θ
θ-2θ
θ-2θ
data collection 2θ limit (deg)
50.0
50.0
50.0
total data collected
6236 (-32 e h e 32,
-7 e k e 7,
-24 e l e 24)d
3053
7410 (0 e h e 15,
-15 e k e 16,
-14 e l e 13)
7073
7517 (-14 e h e 15,
-15 e k e 16,
0 e l e 14)
indep reflns
7072
2
2
2
no. of observations (NO)
range of transmission
factors
2707 (Fo g 2σ(Fo2))
3404 (Fo g 2σ(Fo2))
4563 (Fo g 2σ(Fo2))
0.8765-0.8039
0.8878-0.8367
1.135-0.800
2
2
2
no. of data/restraints/params
goodness-of-fit (S)f
final R indicesg
3053 [Fo g -3σ(Fo2)]/0/195
7071 [Fo g -3σ(Fo2)]/3e/482
7069 [Fo g -3σ(Fo2)]/3e/482
2
2
2
1.077 [Fo g -3σ(Fo2)]
1.007 [Fo g -3σ(Fo2)]
1.023 [Fo g -3σ(Fo2)]
R1[Fo2 > 2σ(Fo2)]
0.0279
0.0676
0.378 and -0.240
0.0760
0.2291
1.053 and -1.095
0.0703
0.2274
1.401 and -1.256
2
wR2[Fo g -3σ(Fo2)]
largest diff peak and hole (e Å-3
)
a
Obtained from least-squares refinement of 40 reflections with 25.4° < 2θ < 28.0° for compound 7, 24 reflections with 20.0° < 2θ <
b
23.9° for compound 11, and 24 reflections with 19.8° < 2θ < 23.8° for compound 12. Programs for diffractometer operation and data
collection were those supplied by Enraf-Nonius. c Programs for diffractometer operation, data collection, data reduction, and absorption
d
correction were those supplied by Siemens. Data were collected from Friedel-opposite quadrants of reciprocal space with indices of the
form (+h - k ( l) and (-h + k ( l). e An idealized geometry was imposed on the inversion-disordered CH2Cl2 solvent molecule by setting
d(Cl(3)-C(99)) ) d(Cl(3′)-C(99)) ) 1.80(1) Å and d(Cl(3)‚‚‚Cl(3′)) ) 2.95(1) Å for compounds 11 and 12. f S ) [∑w(Fo - Fc2)2/(n - p)]1/2
2
(n ) number of data; p ) number of parameters varied; w ) [σ2(Fo2) + (aoP)2 + a1P]-1 where ao ) 0.0342, a1 ) 1.9425 for compound 7,
g
ao ) 0.0988, a1 ) 0 for compound 11 and a0 ) 0.1170 and a1 ) 16.5289 for compound 12; P ) [max(Fo2, 0) + 2Fc2]/3). R1 ) ∑||Fo| -
|Fc||/∑|Fo|; wR2 ) [∑w(Fo - Fc2)2/∑w(Fo4)]1/2
.
2
Structure solution and refinement of 11 and 12 was as
described for 7. The final model for 11 refined to values of R1-
(F) ) 0.0760 (for 3404 data with Fo g 2σ(Fo2)) and wR2(F2) )
2
0.2291 (for all 7073 independent data), while the final refine-
ment indices for 12 were R1(F) ) 0.0703 (for 4563 observed
data) and wR2(F2) ) 0.2274 (for all 7072 independent data).
1
{1H} NMR chemical shift of δ -16.1, and the H NMR
spectrum shows methylene signals at δ 2.29 as a triplet
(3J HH ) 7.0 Hz) and δ 2.82 as a doublet of triplets (J HH
) 7.0 Hz, J HP ) 11.1 Hz). The cyclopentadienyl region
of the 1H NMR spectrum displays two AA′BB′ multiplets
at δ 5.40 and 5.43, the pattern of which appears to be
typical of these compounds. In a similar reaction (start-
ing from LiPMe2), the dimethylphosphino derivative Li-
[C5H4CH2CH2PMe2] (2) is obtained, which displays a
singlet in the 31P{1H} NMR spectrum at δ -55.9. The
methylene and Cp′ resonances are as expected in the
1H NMR spectrum.
The analogous “C1” ligand, having a one-carbon spacer
between the cyclopentadienyl and phosphine units,
K[C5H4CH2PPh2] (3), is obtained by the addition of
potassium diphenylphosphide to a solution of fulvene
at low temperature, as shown in eq 2. The spectroscopy
Resu lts a n d Com p ou n d Ch a r a cter iza tion
P r ep a r a t ion of Cyclop en t a d ien yl P h osp h in e
Ligan ds. The first ligand synthesized, -C5H4(CH2)2PPh2,
was prepared by a modification of an earlier method
involving the reaction of lithium diphenylphosphide
with spiro[2.4]hepta-4,6-diene, followed by hydrolysis
and then purification via chromatography.16 We have
found that this ligand may be obtained more conve-
niently as a spectroscopically pure crystalline solid by
using KPPh2, instead of the lithium salt as shown in
eq 1, and may be used directly without further purifica-
tion. The compound K[C5H4CH2CH2PPh2] (1) has a 31P-
(16) Kauffmann, T.; Bisling, M.; Teuben, J . H.; Konig, R. Angew.
Chem., Int. Ed. Engl. 1980, 19, 328.

Heterobinuclear Ti/ Mo and Zr/ Mo Complexes
Organometallics, Vol. 18, No. 17, 1999 3495
Sch em e 1
transition metals, was to first complex the cyclopenta-
dienyl unit of the heterobifunctional ligands to an early
metal, followed by reaction of these metallocene-like
derivatives with late-metal sources. The compound [(η5-
C5H4CH2CH2PPh2)2TiCl2] (6) was prepared initially by
the reaction of K[C5H4CH2CH2PPh2] (1) with TiCl4-
(THF)2. This reaction proceeds with concomitant reduc-
tion of Ti(IV) to Ti(III), as evidenced by the formation
of a dark green solution. It was found that pure 6 could
be obtained, albeit in low yield, by air oxidation of these
green solutions, followed by subsequent workup. Alter-
natively, 6 can be obtained in higher yield via the
reaction of K[C5H4CH2CH2PPh2] (1) with TiCl3(THF)3,
followed by oxidation with HCl. Although an analyti-
cally pure product can be obtained via this method,
F igu r e 1. Perspective view of [(η5-C5H4C2H4PPh2)2ZrCl2]
(7). Thermal ellipsoids are drawn at the 20% level except
for methylene hydrogens, which are shown arbitrarily
small. Only the ipso carbons of the phenyl rings are shown,
and hydrogens on cyclopentadienyl rings are also not
shown. These rings are numbered starting at the ipso
carbon. Primed atoms are related to unprimed ones by a
crystallographic 2-fold rotation axis (1/4, y, 0) passing
through the Zr atom.
on this compound is similar to that of 1 except for the
single 1H resonance for the methylene group and a
single resonance for the cyclopentadienyl hydrogens,
which are apparently coincidentally degenerate.
1
signals in the H and 31P NMR spectra of the product
are often broadened slightly, presumably due to the
presence of small amounts of paramagnetic Ti(III)-
containing impurities. The 31P{1H} NMR resonance for
compound 6 is very similar to that of the free ligand,
suggesting that the phosphorus is not coordinated to the
titanium center, and the 1H NMR spectrum also re-
sembles that of the free ligand, apart from a shift of
the Cp hydrogen signals to lower field. The zirconocene
dichloride derivative, [(η5-C5H4CH2CH2PPh2)2ZrCl2] (7),
is prepared via the reaction of K[C5H4CH2CH2PPh2] (1)
with either ZrCl4 or ZrCl4(THF)2 and has remarkably
similar spectroscopic properties to compound 6, includ-
ing an identical 31P NMR chemical shift. Single crystals
of 7, suitable for X-ray analysis, were obtained by slowly
cooling a CH2Cl2/n-pentane solution of the complex, and
the resulting structure is shown to consist of a typical
pseudotetrahedral ligand arrangement around zirco-
nium (see Figure 1). A crystallographic 2-fold axis
bisects the Cl-Zr-Cl′ angle. Selected bond lengths and
angles are given in Table 3. The Zr-Cl bond length and
Cl-Zr-Cl′ bond angle of 2.4448(6) Å and 99.69(3)°,
respectively, are typical of zirconocene-type compounds.17
Although one might have anticipated that the bulky
substituents on the Cp′ ring would be staggered with
respect to each other, thereby minimizing mutual repul-
sions, they are twisted only slightly from an eclipsed
The related “C1” ligand, Li2[Me2C(C5H3C(Me)2PPh2)2]
(5), in which each Cp ring is linked to the phosphine
moiety by a C(CH3)2 spacer group, and in which the Cp
rings are also connected to each other by a C(CH3)2
linker, is obtained by addition of lithium diphenylphos-
phide to the difulvene shown in Scheme 1. The presence
of alkyl substituents on the spacer carbon between the
Cp and PPh2 groups results in a large change in the
31P NMR chemical shift, when compared to K[C5H4CH2-
PPh2] (3), with the resonance for the two equivalent
phosphines appearing as a singlet at δ 20.2. The 1H
NMR spectrum shows a doublet for the methyls bound
to the spacer carbon with phosphorus coupling of 11.8
Hz, and this coupling allows for the differentiation of
this signal and that of the methyls bound to the linker
carbon (linking the two Cp groups), which appears as a
singlet. Three separate multiplets are observed for the
cyclopentadienyl hydrogens.
P r ep a r a tion of Meta llocen e Dich lor id es. Our
strategy for the preparation of ELHB complexes, or
analogous compounds containing mid rather than late
(17) (a) Petersen, J . L.; Egan, J . W. Inorg. Chem. 1983, 22, 3571.
(b) Green, J . C.; Green, M. L. H.; Prout, C. K. J . Chem. Soc., Chem.
Commun. 1972, 421, and references therein.

3496 Organometallics, Vol. 18, No. 17, 1999
Ta ble 3. Selected Bon d Len gth s a n d An gles for [(η5-C5H4CH2CH2P P h 2)2Zr Cl2] (7)
(a) Selected Interatomic Angles (Å)
Graham et al.
atom 1
atom 2
distance
atom 1
atom 2
distance
Zr
Zr
Zr
Zr
Zr
Zr
P
Cl
2.4448(6)
2.565(2)
2.509(2)
2.473(2)
2.462(2)
2.534(2)
1.850(2)
1.841(2)
P
C(31)
C(2)
1.837(2)
1.539(3)
1.498(3)
1.424(3)
1.400(3)
1.398(3)
1.408(4)
1.410(3)
C(10)
C(11)
C(12)
C(13)
C(14)
C(1)
C(1)
C(2)
C(10)
C(10)
C(11)
C(12)
C(13)
C(10)
C(11)
C(14)
C(12)
C(13)
C(14)
P
C(21)
(b) Selected Interatomic Angles (deg)
atom 1
atom 2
atom 3
angle
atom 1
atom 2
atom 3
angle
Cl
Zr
P
P
P
Cl′a
99.69(3)
98.73(9)
102.25(10)
99.89(9)
111.54(15)
110.2(2)
125.4(2)
127.6(2)
106.7(2)
108.8(2)
107.7(2)
107.8(2)
108.9(2)
118.4(2)
122.9(2)
C(22)
C(21)
C(22)
C(23)
C(24)
C(21)
P
C(21)
C(22)
C(23)
C(24)
C(25)
C(26)
C(31)
C(31)
C(31)
C(32)
C(33)
C(34)
C(35)
C(36)
C(26)
C(23)
C(24)
C(25)
C(26)
C(25)
C(32)
C(36)
C(36)
C(33)
C(34)
C(35)
C(36)
C(35)
118.7(2)
121.2(3)
119.8(3)
120.5(2)
119.9(3)
119.9(2)
125.0(2)
116.8(2)
118.1(2)
120.9(3)
120.6(3)
119.3(3)
120.8(3)
120.4(3)
C(1)
C(1)
C(21)
P
C(1)
C(2)
C(2)
C(11)
C(10)
C(11)
C(12)
C(10)
P
C(21)
C(31)
C(31)
C(2)
C(1)
C(2)
C(10)
C(10)
C(10)
C(11)
C(12)
C(13)
C(14)
C(21)
C(21)
C(10)
C(11)
C(14)
C(14)
C(12)
C(13)
C(14)
C(13)
C(22)
C(26)
P
C(32)
C(31)
C(32)
C(33)
C(34)
C(31)
P
a
Primed atoms are related to unprimed ones via the 2-fold rotational axis (1/4, y, 0).
conformation, with a C(10)-Cp(c)-Cp(c)′-C(10′) torsion
angle of 27.2° (Cp(c) ) centroid of C5 ring), in which
the substituents essentially bisect the Cl-Zr-Cl′ angle,
giving a C(2)-C(10)-Zr-X torsion angle of 12.9° (X )
bisector of Cl-Zr-Cl′ angle). Despite the almost eclipsed
alignment, the phosphinoalkyl arms are directed away
from each other, with the alkyl substituents on each ring
aimed in opposite directions away from the ZrCl2 plane.
In this orientation the bulky substituents minimize
repulsions between each other. In addition, by occupying
the carbons of the tilted monosubstituted Cp groups that
are furthest from each other on opposite Cp rings, steric
repulsions are further minimized. This orientation of
the alkyl substituents is essentially that desired for
using the (Ph2PC2H4C5H4)2ZrCl2 moiety as a bidentate
metalloligand on a late metal except that the phosphines
are directed in essentially opposite directions (up and
down in Figure 1). However, rotation about the C(10)-
C(2) and C(10′)-C(2′) bonds, bringing the phosphine
moieties closer, is all that is required to allow this unit
to function as a chelating metalloligand.
The presence of a diphenylphosphinoethyl group
bound to each cyclopentadienyl ring of 7 has resulted
in a slight asymmetry in the Zr-C bond lengths, with
these distances decreasing steadily from 2.565(2) Å at
the substituted carbon to 2.468 Å (average) for C(12)
and C(13). In addition, the asymmetry is also evident
in the angles within the Cp′ groups (Cp′ ) C5H4R),
where that of the substituted carbon is 106.7(2)° and
the average for the unsubstituted carbons is 108.3°. The
Cp(c)-Zr-Cp(c)′ angle and the Cp(c)-Zr distances are
remarkably similar to those found in Cp′2ZrCl2, with
values of 130.9° and 2.205 (average) Å, respectively.
The structure of 7 can be compared with four closely
related group 4 metallocenes.7b,18-20 It differs substan-
tially from the structures of rac-[(C5H4CH(CH3)PPh2)2-
ZrCl2], shown in D,7b and a related amine species rac-
[(C5H4CH(n-C4H9)NMe2)2ZrCl2],18 mainly with respect
to the orientation of the pendent groups to each other
and to the chloro ligands. In compound D, these sub-
stituents are oriented with a C-Cp(c)-Cp(c)′-C′ dihe-
dral angle of 178.6° (Cp(c) ) centroid of C5 ring, C and
C′ are the spacer carbons), resulting in a staggered
arrangement, and are also oriented at ca. 90° to the Cl-
Zr-Cl bisector. In this case compound D can function
as a chelating metalloligand by rotation of both Cp′ rings
together by ca. 90°, to give an eclipsed arrangement.
The related amine18 has structural parameters closely
resembling those of D. In [(C5H4C(CH3)2PPh2)2Zr-
(CH3)2]19 a slightly different geometry is observed. The
phosphinoalkyl arms are eclipsed but are again directed
approximately 90° to the bisector of the H3C-Zr-CH3
angle. This latter arrangement differs from 7 only by
an approximate 90° rotation of the Zr(CH3)2 moiety. In
these compounds16,18,19 the pendent arms are oriented
away from the ligands (Cl or CH3) in the equatorial
plane in order to minimize contacts between these
(18) Bertuleit, A.; Fritze, C.; Erker, G.; Fro¨hlich, R. Organometallics
1997, 16, 2891.
(19) Bosch, B. E.; Erker, G.; Fro¨hlich, R.; Meyer, O. Organometallics
1997, 16, 5449.

Heterobinuclear Ti/ Mo and Zr/ Mo Complexes
Organometallics, Vol. 18, No. 17, 1999 3497
ligands and the alkyl substituents on the spacer car-
bons. The [(C5H4CH2CH2N(iPr)2)2TiCl2] complex,20 on
the other hand, having a C2H4 spacer between the C5H4
group and the amine functionality, has an arrangement
of Cp′ groups much like that in 7, in which the
aminoalkyl arms are eclipsed and bisecting the TiCl2
angle, but aimed away from the TiCl2 plane. This
arrangement is possible in which there are no substit-
uents on the carbon linkers, since no unfavorable
contacts involving the equatorial Cl ligands result.
These differences in pendent arm orientation between
these compounds may have important implications in
the formation of ELHB complexes, with these observed
orientations suggesting that oligomeric mixed-metal
complexes may result by binding of the phosphine
substituents to different late metals. A subsequent
paper will describe a series of tetranuclear complexes
formed in this way.21
decrease the rotational freedom of the cyclopentadienyl
units, which may influence the tendency for binuclear
complexes to form instead of oligomers. In addition, the
linker should decrease the Cp(c)-M-Cp(c) angle, re-
sulting in easier substrate access to the metal center.22
Reaction of Li2[Me2C(C5H3CMe2PPh2)2] (5) with ZrCl4
afforded a bright yellow product, the 31P{1H} NMR
spectrum of which showed the presence of two com-
pounds, rac- and meso-[ Me2C(C5H3CMe2PPh2)2ZrCl2]
(10a , 10b) in a 2.8:1 ratio.23 These isomers have
The titanocene dichloride derivative with a dipheny-
phosphinomethyl unit bound to the cyclopentadienyl
ring, [(η5-C5H4CH2PPh2)2TiCl2] (8), was prepared simi-
larly, by the reaction of K[C5H4CH2PPh2] (3) with TiCl3-
(THF)3, followed by oxidation with HCl, as shown in eq
3. 1H NMR spectroscopy of this complex shows the
differing solubility properties, allowing the rac isomer
to be obtained spectroscopically pure, although the meso
isomer was always found to be contaminated with the
rac isomer. The 31P{1H} NMR spectrum of the rac
isomer is similar to that of [(η5-C5H4CMe2PPh2)2ZrCl2],7b
with a 31P NMR chemical shift of δ 32.7. 1H NMR
spectroscopy on this compound shows three cyclopen-
tadienyl hydrogen resonances in the expected region,
as well as doublets at δ 1.48 (J HP ) 13.0 Hz) and 1.75
(J HP ) 16.2 Hz) for the methyl groups adjacent to the
PPh2 groups, and a single resonance for the methyl
groups on the CMe2 moiety linking the two Cp′ groups.
The spectral parameters for the meso isomer are very
similar to those of 10a (see Table 1) except that two of
the C5H3 protons are accidentally equivalent, and two
signals are observed for the methyl groups on the linker
carbon. The key to determining whether the rac or meso
isomer is obtained is the signal for the methyl hydrogens
of the Me2C group linking the two Cp′ groups. In the
meso isomer, these methyl groups are in different
chemical environments and are expected to have dif-
ferent chemical shifts, whereas in the rac isomer, which
contains a C2 rotation axis, the methyl groups are
chemically equivalent. A pure sample of the rac isomer
shows only the single resonance for the two methyls on
the linker carbon.
methylene singlet having no resolvable phosphorus
coupling, as was observed for the free ligand, and the
Cp′ resonances are again shifted slightly downfield
compared to the free ligand. The 31P{1H} NMR spectrum
of compound 8 shows a significant downfield shift, to δ
-9.6, and contrasts with the observations in 6 and 7 of
essentially no change in the 31P NMR spectrum upon
coordination of the Cp′ groups. In compound 8, phos-
phine coordination to the early metal is not expected,
as chelation should result in a significantly strained
ring, although chelation of these ligand types has been
observed in [(η5:η1-C5H4CMe2PPh2)2ZrMe][MeB(C6F5)3],19
presumably to alleviate the electron deficiency at Zn
that would otherwise result. The downfield 31P shift in
8 is likely due to the smaller cyclopentadienyl-PPh2
separation, which results in the phosphine unit being
closer to the metal. The zirconium-containing derivative
[(η5-C5H4CH2PPh2)2ZrCl2] (9) was prepared via reaction
of K[C5H4CH2PPh2] (3) with ZrCl4 or ZrCl4(THF)2 and
has spectroscopic features similar to compound 8. The
similarity of the 31P NMR chemical shift provides
further evidence that the phosphorus is not metal-
bound, since one would expect a significant chemical-
shift difference upon coordination to the two different
metals.
Mixed -Meta l Com p lexes of Mo. We began our
studies of heterobimetallic complexes incorporating the
bifunctional ligands described earlier with the synthesis
of cis-[(µ-η5:η1-C5H4CH2CH2PPh2)2MCl2Mo(CO)4] (M )
Ti (11), Zr (12)), in which the C5H4 and PPh2 groups
are linked by a C2 spacer. It was of interest to compare
the structures of 11 and 12 with the related compounds,
[(µ-η5:η1-C5H4PPh2)2ZrCl2Mo(CO)4] (E)6b and [(µ-η5;η1-
C5Me4PPh2)2TiCl2Mo(CO)4] (F ),6d in which the PPh2
groups are bound directly to the C5H4 or C5Me4 moieties,
Preparation of an ansa-metallocene dichloride was of
interest in hopes that the presence of a one-carbon
linker between the two cyclopentadienyl rings would
(20) J utzi, P.; Redeker, T.; Neumann, B.; Stammler, H.-G. Organo-
metallics 1996, 15, 4153.
(21) Graham, T. W.; Llamazares, A.; McDonald, R.; Cowie, M.
Organometallics 1999, 18, 3502.
(22) Smith, J . A.; Von Seyerl, J .; Huttner, G.; Brintzinger, H. H. J .
Organomet. Chem. 1979, 173, 175.
(23) Halterman, R. L. Chem. Rev. 1992, 92, 965.

3498 Organometallics, Vol. 18, No. 17, 1999
Graham et al.
Ta ble 4. Selected Bon d Len gth s a n d An gles for Com p ou n d 11
(a) Selected Interatomic Distances (Å)
atom 1
atom 2
distance
atom 1
atom 2
distance
Mo
Mo
Mo
Mo
Mo
Mo
Ti
Ti
Ti
Ti
Ti
Ti
Ti
Ti
Ti
Ti
Ti
Ti
P(1)
P(2)
C(1)
C(2)
C(3)
C(4)
Cl(1)
Cl(2)
C(7)
2.535(3)
2.535(3)
P(2)
P(2)
P(2)
O(1)
O(2)
O(3)
O(4)
C(5)
C(6)
C(7)
C(7)
C(8)
C(12)
C(41)
C(51)
C(1)
C(2)
C(3)
C(4)
C(6)
C(7)
C(8)
1.826(10)
1.812(11)
1.831(11)
1.144(14)
1.132(13)
1.165(14)
1.176(13)
1.544(14)
1.50(2)
1.968(12)
2.067(13)
1.978(13)
1.965(11)
2.351(3)
2.325(4)
2.487(11)
2.422(12)
2.393(11)
2.333(12)
2.375(11)
2.431(11)
2.384(12)
2.344(12)
2.371(11)
2.437(12)
1.816(11)
1.835(12)
1.850(11)
C(8)
C(9)
1.407(14)
1.43(2)
1.41(2)
C(11)
C(9)
C(10)
C(11)
C(14)
C(15)
C(16)
C(17)
C(18)
C(5)
C(9)
C(10)
C(11)
C(13)
C(14)
C(15)
C(18)
C(16)
C(17)
C(18)
1.35(2)
C(10)
C(12)
C(13)
C(14)
C(14)
C(15)
C(16)
C(17)
1.425(15)
1.498(14)
1.534(13)
1.388(15)
1.42(2)
1.39(2)
1.42(2)
1.398(15)
P(1)
P(1)
P(1)
C(21)
C(31)
(b) Selected Interatomic Angles (deg)
atom 1
atom 2
atom 3
angle
atom 1
atom 2
atom 3
angle
P(1)
P(1)
P(1)
P(1)
P(1)
P(2)
P(2)
P(2)
P(2)
C(1)
C(1)
C(1)
C(2)
C(2)
C(3)
Cl(1)
Mo
Mo
Mo
Mo
Mo
Mo
Mo
Mo
Mo
Mo
Mo
Mo
Mo
Mo
Mo
Mo
Ti
P(2)
C(1)
C(2)
C(3)
C(4)
C(1)
C(2)
C(3)
C(4)
C(2)
C(3)
C(4)
C(3)
C(4)
C(4)
Cl(2)
C(5)
C(12)
O(1)
O(2)
100.64(10)
90.0(4)
90.9(3)
Mo
Mo
P(1)
C(3)
C(4)
C(5)
C(6)
C(7)
C(7)
C(7)
C(8)
O(3)
O(4)
C(6)
C(7)
173.9(12)
176.9(11)
120.5(8)
82.9(3)
C(5)
C(6)
C(6)
C(8)
C(7)
C(8)
C(9)
C(7)
P(2)
C(12)
C(13)
C(13)
C(15)
C(14)
C(15)
C(16)
C(14)
112.7(10)
128.0(11)
125.6(10)
106.4(10)
108.3(11)
109.3(10)
108.6(11)
107.2(10)
113.3(7)
173.5(4)
169.3(4)
89.2(3)
93.4(4)
85.4(4)
89.0(5)
89.4(6)
84.0(5)
173.6(5)
91.5(5)
C(8)
C(11)
C(11)
C(9)
C(9)
C(10)
C(11)
C(10)
C(13)
C(14)
C(15)
C(18)
C(18)
C(16)
C(17)
C(18)
C(17)
C(10)
C(11)
C(12)
C(13)
C(14)
C(14)
C(14)
C(15)
C(16)
C(17)
C(18)
115.8(9)
126.5(10)
125.5(10)
107.4(10)
110.2(11)
106.4(10)
108.9(10)
107.2(10)
94.4(5)
93.54(14)
112.7(4)
124.7(4)
173.3(11)
179.0(10)
P(1)
P(2)
C(1)
C(2)
Mo
Mo
Mo
to determine the effect on the geometry of having
different length spacers between the C5R4 and PPh2
moieties.
structures were determined by single-crystal X-ray
techniques. Both compounds are isostructural, with only
subtle differences resulting from the two different
metallic radii (Ti ) 1.47 Å, Zr ) 1.60 Å).24 The general
structural features of complexes 11 and 12 are quite
similar, with pseudotetrahedral ligand arrangements
around the group 4 metal and pseudooctahedral geom-
etries about molybdenum, with the phosphines of the
C5H4CH2CH2PPh2 units bound cis to this metal. Figure
2 shows a perspective view of the Zr complex (12), while
Figure 3 shows an alternate view of the same molecule
looking down the Cp-Cp vector. The drawings for the
Ti analogue (11) are not shown since they are essentially
identical and the numbering scheme used is the same
for both compounds. Selected bond lengths and angles
for both compounds are given in Tables 4 and 5,
respectively. The structural results confirm that the
binuclear framework has been achieved in which the
early and mid transition metals are bridged by the
-C5H4CH2CH2PPh2 ligands. Metal-metal separations
Compounds 11 and 12 were prepared via the reaction
of [(η5-C5H4CH2CH2PPh2)2MCl2] (M ) Ti (6), Zr (7)) with
(COD)Mo(CO)4, resulting in facile displacement of the
labile diene group. The 31P chemical shift upon coordi-
nation of the phosphino groups to the molybdenum
center changes from δ -16.5 (for both 6 and 7) to δ 26.6
and 26.5 (both singlets) for 11 and 12, respectively. The
1
Cp′ hydrogens in the H NMR spectrum of compound
11 are apparently coincidentally degenerate and reso-
nate at δ 6.24, whereas the two methylene resonances
of the phosphine arms are complex multiplets at δ 2.55
and 2.45. The 1H NMR spectrum of compound 12 is
similar to that of 11 except that the Cp′ region now
shows the expected AA′BB′ multiplets, centered at δ
6.34. The solution IR spectra of both compounds are
virtually identical, with a carbonyl band pattern similar
to that of (COD)Mo(CO)4,10 indicating that the phos-
phines, like the COD moiety, are bound to the molyb-
denum center in a cis arrangement. To confirm the
structures of compounds 11 and 12 and to compare the
differences due to the different group 4 metals, their
(24) (a) Wells, A. F. Structural Inorganic Chemistry, 5th ed.;
Clarendon Press: Oxford, 1984. (b) Values given are for coordination
number 12; for pseudotetrahedral geometries these values should be
multiplied by 0.88.

Heterobinuclear Ti/ Mo and Zr/ Mo Complexes
Organometallics, Vol. 18, No. 17, 1999 3499
Ta ble 5. Selected Bon d Len gth s a n d An gles for Com p ou n d 12
(a) Selected Interatomic Distances (Å)
atom 1
atom 2
distance
atom 1
atom 2
distance
Mo
Mo
Mo
Mo
Mo
Mo
Zr
Zr
Zr
Zr
Zr
Zr
Zr
Zr
Zr
Zr
Zr
Zr
P(1)
P(2)
C(1)
C(2)
C(3)
C(4)
Cl(1)
Cl(2)
C(7)
2.539(2)
2.540(2)
1.991(12)
2.071(11)
1.997(11)
2.006(10)
2.431(3)
2.412(2)
2.554(9)
2.542(9)
2.513(9)
2.474(9)
2.487(9)
2.541(9)
2.511(9)
2.475(9)
2.502(9)
2.541(8)
1.824(9)
1.823(10)
1.833(9)
P(2)
P(2)
P(2)
O(1)
O(2)
O(3)
O(4)
C(5)
C(6)
C(7)
C(7)
C(8)
C(12)
C(41)
C(51)
C(1)
C(2)
C(3)
C(4)
C(6)
C(7)
C(8)
1.837(8)
1.826(9)
1.838(9)
1.155(13)
1.146(12)
1.148(12)
1.132(11)
1.542(13)
1.497(13)
1.404(13)
1.413(13)
1.382(14)
1.412(15)
1.402(13)
1.536(12)
1.515(12)
1.412(12)
1.408(13)
1.415(13)
1.401(14)
1.389(13)
C(8)
C(9)
C(11)
C(9)
C(10)
C(11)
C(14)
C(15)
C(16)
C(17)
C(18)
C(5)
C(9)
C(10)
C(11)
C(13)
C(14)
C(15)
C(18)
C(16)
C(17)
C(18)
C(10)
C(12)
C(13)
C(14)
C(14)
C(15)
C(16)
C(17)
P(1)
P(1)
P(1)
C(21)
C(31)
(b) Selected Interatomic Angles (deg)
atom 1
atom 2
atom 3
angle
atom 1
atom 2
atom 3
angle
P(1)
P(1)
P(1)
P(1)
P(1)
P(2)
P(2)
P(2)
P(2)
C(1)
C(1)
C(1)
C(2)
C(2)
C(3)
Cl(1)
Mo
Mo
Mo
Mo
Mo
Mo
Mo
Mo
Mo
Mo
Mo
Mo
Mo
Mo
Mo
Mo
Zr
P(2)
C(1)
C(2)
C(3)
C(4)
C(1)
C(2)
C(3)
C(4)
C(2)
C(3)
C(4)
C(3)
C(4)
C(4)
Cl(2)
C(5)
C(12)
O(1)
O(2)
100.87(8)
90.3(3)
90.9(3)
83.2(3)
172.8(3)
168.7(3)
88.9(3)
94.1(3)
85.6(3)
89.3(4)
88.9(4)
83.3(4)
173.8(4)
92.3(4)
93.3(4)
96.30(10)
112.2(3)
125.1(3)
173.9(9)
178.3(9)
Mo
Mo
P(1)
C(3)
C(4)
C(5)
C(6)
C(7)
C(7)
C(7)
C(8)
O(3)
O(4)
C(6)
C(7)
174.9(9)
177.6(9)
119.8(6)
113.1(8)
127.1(9)
126.5(9)
106.3(9)
109.5(9)
108.2(8)
107.0(9)
109.0(9)
112.2(6)
113.6(7)
123.2(8)
128.4(8)
107.7(8)
107.1(8)
108.5(8)
107.9(8)
108.8(8)
C(5)
C(6)
C(6)
C(8)
C(7)
C(8)
C(9)
C(7)
P(2)
C(12)
C(13)
C(13)
C(15)
C(14)
C(15)
C(16)
C(14)
C(8)
C(11)
C(11)
C(9)
C(9)
C(10)
C(11)
C(10)
C(13)
C(14)
C(15)
C(18)
C(18)
C(16)
C(17)
C(18)
C(17)
C(10)
C(11)
C(12)
C(13)
C(14)
C(14)
C(14)
C(15)
C(16)
C(17)
C(18)
P(1)
P(2)
C(1)
C(2)
Mo
Mo
Mo
of 6.895 and 6.945 Å in compounds 11 and 12, respec-
tively, rule out any metal-metal interaction. As shown
in Figure 3, the cyclopentadienyl groups are staggered,
where the angle subtended by the C(13)-C(14) and
C(6)-C(7) vectors is 32.9° for both compounds, indicat-
ing that these groups rotate about the Cp-early-metal
bond to find the most sterically favorable orientation.
This drawing also shows that the molybdenum atom is
located to the side of the pocket created by the canted
cyclopentadienyl rings, with Mo-M-X angles (X is the
bisector of the Cl-M-Cl angle) of 103.1° for compound
11 (M ) Ti) and 104.4° for 12 (M ) Zr), indicating that
the early-metal substituents are directed essentially
perpendicular to the early-metal-Mo vector. Also shown
in Figures 2 and 3, access between the early metals and
Mo is inhibited by the orientations of the C2H4 spacer
between the PPh2 and C5H4 groups, in which the
methylene hydrogens block access to this cavity. Despite
the coordination of the phosphine substituents to the
Mo center, the Cp(centroid)-M-Cp(centroid) angles of
128.9° and 128.3° for these compounds are remarkably
similar to those found in Cp′2TiCl2 (130.2°), in Cp′2ZrCl2
(128.9°),17 and in the Zr precursor (7) (130.9°). Some
asymmetry is noted in the M-Cl bond lengths, with
M-Cl(1)/M-Cl(2) distances of 2.351(3) Å/2.325(4) Å for
M ) Ti and 2.431(3) Å/2.412(2) Å for M ) Zr. The M-Cl-
(1) distances are similar to the corresponding bond
lengths in Cp′2TiCl2 and Cp′2ZrCl2,17 suggesting that
this bond length is normal. Furthermore, it may be
recalled that in the early-metal precursor (7) the Zr-
Cl distances were identical (by symmetry) at 2.4448(6)
Å. We suggest that the somewhat shorter M-Cl(2)
distances may be due to intramolecular nonbonded
contacts, since the intermolecular contacts are similar
for both Cl(1) and Cl(2). However, Cl(2) has additional
intramolecular nonbonded contacts with H(5A), H(6B),
and H(12A), ranging from 2.75 to 2.76 Å, which act in
a direction tending to compress the M-Cl(2) distances.
The M-C bond lengths around the Cp′ rings show some
distortion, presumably due to the presence of the
diphenylphosphinoalkyl substituent, with the longest
bond being to the cyclopentadienyl carbon connected to
this substituent; for M ) Zr these distances are Zr-
C(7) ) 2.554(9) and Zr-C(14) ) 2.541(9) Å, whereas the
Zr-C distance for the unsubstituted carbons ranges
from 2.474(9) to 2.542(9) Å. These metal-carbon sepa-
rations decrease with distance from the alkyl substitu-
ent, suggesting that the phosphinoalkyl group is forcing

3500 Organometallics, Vol. 18, No. 17, 1999
Graham et al.
imposable despite the different group 4 metals, indicate
that the most significant structural differences are
probably not attributable to metal differences. More
likely the differences between E and F arise from the
methyl substituents on the cyclopentadienyl ring in F ,
which put severe restrictions on the favored tilt angles
of the C5Me4 groups. Consistent with this argument, the
relative orientations of the MCl2 group and the M-Mo
vector (M ) Ti, Zr) in compounds 11, 12, and E are all
comparable, whereas for F the TiCl2 group is aimed
away from the Mo center; this TiCl2 orientation is a
direct function of the C5Me4 tilt angles.
Attempts to thermally induce the isomerization of the
phosphine groups from a cis to a trans arrangement
were unsuccessful and resulted in significant amounts
of decomposition. Similarly, a convenient mononuclear
Mo synthon having labile groups in a trans orientation
was not available for the direct synthesis of a M/Mo
compound (M ) Ti, Zr) having a trans phosphine
arrangement.
F igu r e 2. Perspective view of [(µ-η5:η1-C5H4C2H4PPh2)2-
ZrCl2Mo(CO)4] (12). Thermal parameters and numbering
convention as described in Figure 1. Numbering scheme
for the Ti analogue (11) is identical.
Discu ssion
Cyclopentadienyl rings having a variety of functional
groups have been reported in the literature,6,7,16,24
a
number of them containing the desired pendent phos-
phine moieties.6,7,16 However, at the time this study was
initiated few had been reported7a,c,16 in which the C5H4
and PR2 moieties were linked with C1 or C2 spacers.
More recently a number of reports of similar ligands
have appeared.7b,19,25 We began with the synthesis of
the (diphenylphosphino)ethylcyclopentadienyl ligand,
because it had been briefly reported as yielding heter-
obinuclear complexes of the type desired.7c We have
prepared the C2-spaced ligand by the reaction of KPPh2
with spiro[2.4]hepta-4,6-diene (eq 1), a modification of
an earlier procedure,16 and we have used the fulvene
route26 for the synthesis of our C1-spaced ligand since
nucleophiles are known to add readily to the positively
polarized C(6) carbon of fulvenes, as shown in eq 2; an
analogous route was described by Erker19 while our
work was in progress.
We were concerned that in a metallocene-like species
the rotational flexibility of the cyclopentadienyl rings
might hinder the formation of binuclear ELHB com-
plexes, leading instead to oligomeric compounds; con-
sequently we investigated the use of the Me2C-bridged
difulvene (4) to form an ansa-metallocene, based on the
idea that in such a structure the pendent phosphines
would be held in the same general direction, suitable
for binding to a single late metal, unlike the structure
found for D, for example, in which these arms were
directed in opposite directions on each side of the
metallocene complex. We again chose the fulvene route,
analogous to the method used by Little and Stone,27 due
to the ease of preparing the fulvene shown in Scheme
F igu r e 3. Alternate view of compound 12 viewed along
the vector joining the Cp centroids.
the Cp rings apart. The structure of 11 shows similar
effects in the Ti-Cp′ carbon bond lengths, and these
deviations have also been observed in [(µ-η5:η1-C5H4-
PPh2)2ZrCl2Mo(CO)4] (E),6b as well as in the precursor
7. About molybdenum, the geometry in 11 and 12 is
pseudooctahedral, with significant distortions due to the
presence of the two cis phosphine substituents, which
result in a P-Mo-P angle of 100.6(1)° in compound
11 and a similar value of 100.87(8)° for compound 12.
The large P-M-P angles result in a corresponding
decrease of the C(1)-Mo-C(4) bond angle to 84.0(5)°
and 83.3(4)° for the two compounds. These distortions
were not nearly as pronounced in compounds E and F
and are likely due to the preference of the phosphi-
noalkyl arms of the metalloligands 6 and 7 to be as far
apart as possible, as was evidenced by the solid-state
structure of compound 7 (vide supra). Although one
would expect a trans influence in the Mo-C bond
lengths, with the carbonyls opposite the phosphines
differing from those opposite each other, there is no
obvious correlation.
(25) Charrier, C.; Mathey, F. Tetrahedron Lett. 1978, 2407.
(26) Ziegler, K.; Scha¨fer, W. J ustus Liebigs Ann. Chem. 1934, 511,
101. Ziegler, K.; Gellert, H.-G.; Martin, H.; Nagel, K.; Schneider, J . F.
J . Organomet. Chem. 1967, 8, 277. Renaut, P.; Tainturier, G.; Gauth-
eron, B. J . Organomet. Chem. 1978, 148, 43. Brickhouse, M. D.;
Squires, R. R. J . Am. Chem. Soc. 1988, 110, 2706. Okudo, J . Chem.
Ber. 1989, 122, 1075. Squires, R. R. Acc. Chem. Res. 1992, 25, 461.
Clark, T. J .; Killian, C. M.; Luthra, S.; Nile, N. A. J . Organomet. Chem.
1993, 462, 247.
A comparison of the Ti and Zr compounds 11 and 12
with compounds F and E, in which the PPh2 groups are
directly bound to the C5Me4 or C5H4 rings, and which
also involve Ti (F ) and Zr (E), is enlightening. Although
some of the significant differences between E and F
were attributed to the smaller radius of Ti versus Zr,6d
the structures of 11 and 12, which are almost super-
(27) Little, R. D.; Stone, K. J . J . Org. Chem. 1984, 49, 1849.

Heterobinuclear Ti/ Mo and Zr/ Mo Complexes
Organometallics, Vol. 18, No. 17, 1999 3501
1. Subsequent addition of 2 equiv of diphenylphosphide
anion provided the precursor to ansa-metallocenes,
compound 5, in good yield. Reaction of this ligand with
ZrCl4 afforded the ansa-zirconocene dichloride 10.
Preparation of the metallocene dichlorides proceeded
smoothly for all ligands synthesized, and the structure
determination of [(C5H4C2H4PPh2)2ZrCl2], together with
the structure of [(C5H4C(CH3)2PhMe2)2ZrCl2] (D), re-
ported by Erker, clearly shows the potential pitfalls in
attempts to obtain heterobinuclear complexes based on
these early-metal-ligand frameworks. Although these
solid-state structures say little about the conformations
in solution, they clearly indicate that such structures
are accessible in solution and demonstrate two differ-
ent conformations that can give rise to oligomeric
species rather than the targeted heterobinuclear com-
plexes. Even in 7, in which the pendent phosphine arms
are close-to-eclipsed, rotation about the C5H4-CH2CH2-
PPh2 bond orients the arms in opposite directions. In
D these arms are in opposite directions by virtue of
rotation of the Cp′ units themselves. Combinations of
these rotational freedoms should arise in solution,
possibly complicating attempts to obtain heterobinucle-
ar species.
Ti (F );7d R ) H, M ) Zr (E)7b) in which no spacer was
used between the C5R4 and PPh2 units. Inclusion of the
C2H4 spacer between the two functional groups has had
little effect except to increase the metal-metal separa-
tion, as expected on the basis of the greater length of
the resulting bridging ligand. The major difference in
the different structures has resulted from the severe
crowding that results in F from the methyl substituents
on the cyclopentadienyl groups, which suggests that the
flexibility needed in these bifunctional ligands to ac-
commodate different geometries in heterobinuclear
complexes may be inhibited by substituents on the
cyclopentadienyl ring which may result in unfavorable
tilts of the cyclopentadienyl groups.
Although the structures of 11 and 12 confirm that
heterobinuclear complexes can be obtained with the
-C5H5C2H4PPh2 ligand, the cis-phosphine arrangement
at the later metal results in too great a metal-metal
separation and an arrangement of spacer methylene
groups that inhibits access between the metals. In the
next paper21 we present results in which ELHB com-
plexes of the metalloligands 6-9 with Rh and Pd are
described, in which a trans-phosphine arrangement is
expected at the late metals.
It seemed that synthesis of heterobinuclear complexes
had the best chance for success if the phosphine arms
coordinated to the second metal in a cis arrangement;
the approach of two (C5H4C2H4PPh2)2MCl2 units to a
single metal in a cis arrangement should be unfavorable,
so the formation of oligomeric products seemed unlikely.
We therefore started with the synthesis of cis-[(C5H4C2H4-
PPh2)2MCl2Mo(CO)4] (M ) Ti, Zr). These compounds
had been briefly reported previously; however their
characterization had relied solely on IR and NMR
spectroscopy.7c
Ack n ow led gm en t. We thank the Natural Sciences
and Engineering Research Council of Canada (NSERC)
and the University of Alberta for financial support of
the research and NSERC for the funding the P4/RA
diffractometer. We also thank The Ministry of Educa-
tion and Science (Spain) for a postdoctoral fellowship
(to A.L.).
Su p p or tin g In for m a tion Ava ila ble: Tables of X-ray
experimental details, atomic coordinates, interatomic distances
and angles, anisotropic thermal parameters, and hydrogen
parameters for compounds 7, 11, and 12. This material is
available free of charge via the Internet at http://pubs.acs.org.
Our structure determinations of 11 and 12 clearly
confirm the heterobinuclear formulations previously
proposed and offer useful comparisons to two related
compounds [(C5R4PPh2)2MCl2Mo(CO)4] (R ) Me, M )
OM990279W