Binding preferences of the tribenzylidenemethane ligand in high- oxidation-state tantalum complexes

Article abstract of DOI:10.1021/ja981640t

Reaction of Li₂(TBM)(TMEDA)₂ (TBM = tribenzylidenemethane) with TaMe₃Cl₂ gives (TBM)TaMe₃ (1) in 44% yield. Structural characterization of the (tert-Bu-TBM)TaMe₃ derivative 2 shows an eclipsed orientation of the TaMe₃ tripod relative to the inner core of the TBM ligand. Treatment of (TBM)TaMe₃ with ZnCl₂ cleanly replaces one methyl ligand for chloride to give (TBM)TaMe₂Cl (3) which is a versatile precursor to (TBM)Ta-based complexes. Addition of LiNPh₂ to 3 gives (TBM)TaMe₂(NPh₂) (4). Structural characterization reveals that both 3 and 4 have eclipsed frameworks. Metallocene-mimics are accessible by reacting LiCp (Cp = C₅H₅), LiCp* (Cp* = C₅Me₅), LiCp' (Cp' = C₅H₄Me), or LiFlu (Flu = fluorenyl) with 3 to give Cp(TBM)TaMe₂ (5), Cp*(TBM)TaMe₂ (6), Cp'(TBM)TaMe₂ (7), and Flu(TBM)TaMe₂ (8), respectively. The solid-state structures of 5, 7, and 8 display gross molecular geometries similar to those of group 4 metallocenes. Complex 3 reacts with tris(pyrazolylborate) or bis(pyrazolylborate) salts. Thus, [HB(pz)₃](TBM)TaMe₂ (9), [HB(3,5-Me₂-1-pz)₃](TBM)TaMe₂ (10), and [H₂B(pz)₂](TBM)TaMe₂ (11) are obtained from Na[HB(pz)₃], K[HB(3,5-Me₂- 1-pz)₃], and K[H₂B(pz)₂], respectively. Structural characterization of 9, 10, and 11 shows that TBM can adopt a continuum of bonding modes, from η⁴ to η², depending on the steric hindrance around the metal center. The TMM ligand participates in hydrogenation and insertion reactions, indicating that TMM is a weak ancillary ligand. Combining 6, 7, or 8 with MAO results in short-lived ethylene polymerization catalysts. Finally an electronic description of the model complex (TMM)TaMe₃ is developed to account for the eclipsed molecular structures of 1-4. A comparison against the orbital description of staggered (TMM)Fe(CO)₃ is also made.

Full text of DOI:10.1021/ja981640t

12512
J. Am. Chem. Soc. 1998, 120, 12512-12523
Binding Preferences of the Tribenzylidenemethane Ligand in
High-Oxidation-State Tantalum Complexes
George Rodriguez,† John P. Graham,§ W. Donald Cotter,‡ Caroline K. Sperry,†
Guillermo C. Bazan,*,| and Bruce E. Bursten*,§
Contribution from the Department of Chemistry, UniVersity of Rochester, Rochester, New York,
14627-0216, Department of Chemistry, Mount Holyoke College, South Hadley, Massachusetts 01075,
Department of Chemistry, The Ohio State UniVersity, Columbus, Ohio 43210, and Department of
Chemistry, The UniVersity of California, Santa Barbara, California 93106
ReceiVed May 11, 1998
Abstract: Reaction of Li2(TBM)(TMEDA)2 (TBM ) tribenzylidenemethane) with TaMe3Cl2 gives (TBM)-
TaMe3 (1) in 44% yield. Structural characterization of the (tert-Bu-TBM)TaMe3 derivative 2 shows an eclipsed
orientation of the TaMe3 tripod relative to the inner core of the TBM ligand. Treatment of (TBM)TaMe3 with
ZnCl2 cleanly replaces one methyl ligand for chloride to give (TBM)TaMe2Cl (3) which is a versatile precursor
to (TBM)Ta-based complexes. Addition of LiNPh2 to 3 gives (TBM)TaMe2(NPh2) (4). Structural
characterization reveals that both 3 and 4 have eclipsed frameworks. Metallocene-mimics are accessible by
reacting LiCp (Cp ) C5H5), LiCp* (Cp* ) C5Me5), LiCp′ (Cp′ ) C5H4Me), or LiFlu (Flu ) fluorenyl) with
3 to give Cp(TBM)TaMe2 (5), Cp*(TBM)TaMe2 (6), Cp′(TBM)TaMe2 (7), and Flu(TBM)TaMe2 (8),
respectively. The solid-state structures of 5, 7, and 8 display gross molecular geometries similar to those of
group 4 metallocenes. Complex 3 reacts with tris(pyrazolylborate) or bis(pyrazolylborate) salts. Thus, [HB-
(pz)3](TBM)TaMe2 (9), [HB(3,5-Me2-1-pz)3](TBM)TaMe2 (10), and [H2B(pz)2](TBM)TaMe2 (11) are obtained
from Na[HB(pz)3], K[HB(3,5-Me2-1-pz)3], and K[H2B(pz)2], respectively. Structural characterization of 9,
10, and 11 shows that TBM can adopt a continuum of bonding modes, from η4 to η2, depending on the steric
hindrance around the metal center. The TMM ligand participates in hydrogenation and insertion reactions,
indicating that TMM is a weak ancillary ligand. Combining 6, 7, or 8 with MAO results in short-lived ethylene
polymerization catalysts. Finally, an electronic description of the model complex (TMM)TaMe3 is developed
to account for the eclipsed molecular structures of 1-4. A comparison against the orbital description of staggered
(TMM)Fe(CO)3 is also made.
Introduction
Group 5 metallocene mimics that can be used as Ziegler-
Natta catalyst precursors have been sought for some time, since
classical vanadium catalysts are known to have excellent
activities.6 One strategy takes advantage of dianionic 6 π
electron donors to create a metallocene-like molecular structure
and maintain a formally d0 metal center:
There is considerable current interest in finding ligands that
serve as cyclopentadienyl substitutes for early transition-metal
complexes.1 This effort is promoted, in part, by the expectation
of developing new homogeneous olefin polymerization catalysts.
In this respect, group 4 metallocenes remain the most thoroughly
studied class of complexes (i.e., Cp2ZrMe2, Cp ) C5H5; or Cp*2-
ZrCl2, Cp* ) C5Me5).2,3 Non-Cp ligands can be used in the
design of unique catalysts. For example, activated forms of
[(η5-C5Me4)SiMe2(η1-NCMe3)]TiMe2 can incorporate substan-
tial amounts of R-olefins into growing polyethylene chains and
[(Ar)N(CH2)3N(Ar)]TiMe2 with B(C6F5)3 can polymerize R-ole-
fins in a living fashion.4,5
† University of Rochester.
‡ Mount Holyoke College.
Despite structural and electronic similarity to well-known
group 4 catalysts, tantalum compounds have thus far proven
less effective. Dianionic boron cages serve as 6π donor ligands
in (C5H4Me)[C2B9H11]TaMe2 and the benzyne complex Cp[Et2-
§ The Ohio State University.
| The University of CaliforniasSanta Barbara.
(1) Recent overviews: (a) Bochmann, M. J. Chem. Soc., Dalton Trans.
1996, 255. (b) Sperry, C. K.; Rodriguez, G.; Bazan, G. C. J. Organomet.
Chem. 1997, 548, 1.
(2) (a) Brintzinger, H. H.; Fischer, D.; Mu¨lhaupt, R.; Rieger, B.;
Waymouth, R. M. Angew. Chem., Int. Ed. Engl. 1995, 34, 1143. (b) Ziegler
Catalysts; Fink, G., Mu¨lhaupt, R., Brintzinger, H. H., Eds.; Springer-
Verlag: Berlin, 1995.
(3) Jordan, R. F. AdV. Organomet. Chem. 1991, 32, 325.
(4) (a) Lai, S.; Wilson, J. R.; Knight, G. W.; Stevens, J. C.; Chum, P. S.
U.S. Patent 5,272,236. (b) Shapiro, P. J.; Cotter, W. D.; Schaefer, W. P.;
Labinger, J. A.; Bercaw, J. E. J. Am. Chem. Soc. 1994, 116, 4623.
(5) Scollard, J. D.; McConville, D. H. J. Am. Chem. Soc. 1996, 118,
10008.
(6) (a) Haszeline, R. N.; Hyde, T. G.; Tait, P. J. T. Polymer 1973, 14,
215. (b) Haszeline, R. N.; Hyde, T. G.; Tait, P. J. T. Polymer 1973, 14,
221. (c) Haszeline, R. N.; Hyde, T. G.; Tait, P. J. T. Polymer 1973, 14,
224. (d) Doi, Y.; Tokuhiro, N.; Suzuki, S.; Soga, K. Makromol. Chem.
Rapid Commun. 1987, 8, 285. (e) Fernandez, D.; Fernandez, C.; Rodriguez,
J. D.; Arberola, A. Eur. Polym. J. 1989, 25, 1309.
10.1021/ja981640t CCC: $15.00 © 1998 American Chemical Society
Published on Web 11/20/1998

Tribenzylidenemethane Ligand in Tantalum Complexes
J. Am. Chem. Soc., Vol. 120, No. 48, 1998 12513
C2B4H4]Ta(η2-C6H4)(PMe3). Transformation of these precur-
sors into active catalysts has not been reported.7,8 Polymeri-
zation is observed when the borollide complex Cp*[C4H4B-
N(CHMe2)2]TaMe2 is treated with MAO and ethylene, but the
propagating species decompose rapidly at polymerization tem-
peratures.9 Living olefin polymerization has been reported using
complexes of the type Cp*(η4-butadiene)M(Me)2 (M ) Ta, Nb)
with MAO at low temperatures.10 We recently reported
zirconium and tantalum complexes containing the tribenzyli-
denemethane (TBM) ligand. Like its simpler relative, trimeth-
ylenemethane (TMM), TBM can donate up to 6 π electrons,
and for this reason it may be viewed as a dianionic analogue of
the cyclopentadienyl ligand.11,12
ligand (i.e., δ 4.41 for the methylene protons in 1 relative to δ
5.21 for Li2(TBM)(TMEDA)2. In 2 the three benzylic arms
are chemically inequivalent, but a single resonance is observed
(δ 0.7) for the three Me groups as a result of rapid TBM rotation
on the NMR time scale at room temperature. The reactions in
eq 1 are remarkable, considering the ease of reduction of Me3-
TaCl2.
Treating 1 with ZnCl2 in THF/C6H6 results in an orange to
1
red color change. Analysis by H NMR spectroscopy reveals
that the product contains a π-bound TBM and two inequivalent
Ta-Me groups (δ 0.91 and 0.82). These spectroscopic features
together with elemental analysis are consistent with the forma-
tion of (TBM)TaMe2Cl (3) (eq 2).11 Other Lewis acids, such
as AlCl3 or B(C6F5)3 in CH2Cl2, are also effective in forming 3
from 1, but do so at slower rates. Compound 3 is a versatile
starting material for preparing a variety of TBM-containing Ta
complexes (vide infra).
In this contribution we present a detailed study of TBM
complexes of tantalum. The primary aim is to obtain a clearer
picture of the Ta-TBM bonding interaction. We show that
piano-stool complexes adopt the eclipsed conformation and that
this preference has an electronic origin. Also, a range of
pyrazolylborate complexes is reported in which TBM assumes
a continuum of coordination modes. Thus, TBM is able to
accommodate the electronic and steric requirements of the metal
through subtle but significant changes in its coordination
geometry. Reactivity studies show that, at least in the case of
tantalum, the TBM ligand is more reactive than Cp and less
effective at stabilizing catalytic species.
The solid-state structures of 2 and 3 were previously
determined by X-ray crystallography.11 Tantalum is seven-
coordinate with a “domed” TBM framework. The three phenyl
rings are in a propeller-like arrangement. The overall molecular
geometry of 2 and 3 is analogous to that of piano-stool
compounds (CpMX3),13 where TBM replaces Cp as the “seat”
of the stool. An unexpected molecular characteristic is the
eclipsed, trigonal prismatic-like arrangement of the TBM
framework (neglecting TBM's inner carbon) relative to the
TaMe3 and TaMe2Cl tripods. By contrast, the late metal
counterpart (TMM)Fe(CO)3 prefers a staggered geometry.14
To probe further the relative energies of the different
conformations, we introduced a sterically demanding π donor
ligand which might enforce a staggered (quasi-octahedral)
geometry. (TBM)Ta(NPh2)Me2 (4) was therefore prepared, via
Results and Discussion
Synthesis and Structural Characterization. The slow
addition of Li2(TBM)(TMEDA)2 or Li2(tert-Bu-TBM)(TME-
DA)2 to TaMe3Cl2 results in the formation of (TBM)TaMe3 (1)
and (tert-Bu-TBM)TaMe3 (2), respectively (eq 1).11 Isolation
of 1 requires efficient removal of TMEDA by xylene condensa-
tion-evaporation cycles on the crude reaction mixture. Extrac-
tion with benzene affords 1 as a thermally stable orange solid.
Compound 2 may be isolated in a similar manner by performing
the condensation-evaporation cycles with octane. Removal of
TMEDA is important for separating LiCl(TMEDA)x byproducts.
Higher field chemical shifts are characteristic of the coordinated
(7) (a) Uhrhammer, R.; Crowther, D. J.; Olsen, J. D.; Swenson, D. C.;
Jordan, R. F. Organometallics 1992, 11, 3098. (b) Uhrhammer, R.; Su, Y.
X.; Swenson, D. C.; Jordan, R. F. Inorg. Chem. 1994, 33, 4398.
(8) Hausknecht, K. L.; Stockman, K.; Sabat, M.; Finn, M. G.; Grimes,
R. N. J. Am. Chem. Soc. 1995, 117, 1163.
(9) Bazan, G. C.; Donnelly, S.; Rodriguez, G. J. Am. Chem. Soc. 1995,
117, 2671.
(10) (a) Mashima, K.; Nakamura, A. J. Organomet. Chem. 1995, 500,
261. (b) Mashima, K.; Fujikawa, S.; Tanaka, Y.; Urata, H.; Toshiyuki, O.;
Tanaka, E.; Nakamura, A. Organometallics 1995, 14, 2633. (c) Okamoto,
T.; Yasuda, H.; Nakamura, A.; Kai, Y.; Kanehisa, N.; Kasai, N. Organo-
metallics 1988, 7, 2266.
(11) Rodriguez, G.; Bazan, G. C. J. Am. Chem. Soc. 1995, 117, 10155.
(12) (a) Bazan, G. C.; Rodriguez, G.; Cleary, B. P. J. Am. Chem. Soc.
1994, 116, 2177. (b) Bazan, G. C.; Rodriguez, G. J. Am. Chem. Soc. 1997,
119, 343.
(13) Examples of piano stool compounds can be found in general
organometallic texts: (a) Elschenbroich, C., Salzer, A. Organometallics;
VCH: New York, 1989. (b) Collman, J. P.; Hegadus, L. S.; Norton, J. R.;
Finke, R. G. Principles and Applications of Organotransition Metal
Chemistry; University Science Books: Mill Valley, CA, 1987. (c) Com-
prehensiVe Organometallic Chemistry II; Abel, E. W., Stone, F. G.,
Wilkinson, G., Eds.; Pergamon: Oxford, 1995; Vol. 2.
(14) (a) Girard, L. Baird, M. C. J. Organomet. Chem. 1993, 444, 143.
(b) Albright, T. A.; Hofmann, P.; Hoffmann, R. J. Am. Chem. Soc. 1977,
99, 7546. (c) Almenningen, A.; Haaland, A.; Wahl, K. Acta Chem. Scand.
1969, 23, 1145. (d) Churchill, M. R.; DeBoer, B. G. Inorg. Chem. 1973,
12, 525. (e) O’Callaghan, K. A. E.; Brown, S. J.; Page, J. A.; Baird, M. C.;
Richards, T. C.; Geiger, W. E. Organometallics 1991, 10, 3119. (f)
Churchill, M. R.; Gold, K. Inorg. Chem. 1969, 8, 401.

12514 J. Am. Chem. Soc., Vol. 120, No. 48, 1998
Rodriguez et al.
which is difficult to crystallize. In contrast, Cp(TBM)TaMe2
has limited solubility and is isolated by performing the reaction
in THF and collecting the resulting air-stable red precipitate.
The molecular structures of 5 and 6 were sought to elucidate
the Ta-TBM relationship within a metallocene-like environ-
ment. In the case of 6, suitable crystals could not be obtained.
Crystals of 5 were obtained, but its solid-state structural
characterization revealed two independent molecules on a 3-fold
axis with a disorder involving the methyl and Cp ligands.
Anisotropic refinement of the Cp and methyl carbons was not
possible, and detailed intramolecular structural parameters could
not be obtained.17
To resolve the uncertainties raised above, two derivatives
were prepared and structurally characterized: Cp′(TBM)TaMe2
(7 in eq 4) by reaction of 3 with LiCp′ (Cp′ ) C5H4Me), and
Flu(TBM)TaMe2 (8, in eq 5) by reaction of 3 with LiFlu (Flu
) fluorenyl).
Figure 1. ORTEP drawing of 4. Ellipsoids are drawn at 30%
probability. Hydrogen atoms are omitted for clarity.
Table 1. Selected Bond Distances and Angles for 4
atoms
distance (Å)
atoms
angles (deg)
Ta-C(1)
Ta-C(2)
Ta-C(3)
Ta-C(4)
Ta-C(23)
Ta-C(24)
Ta-N
2.249(5)
2.414(6)
2.325(6)
2.405(5)
2.139(6)
2.181(6)
2.006(5)
C(23)-Ta-C(24)
C(2)-C(1)-C(3)
C(2)-C(1)-C(4)
C(3)-C(1)-C(4)
91.3(3)
115.2(5)
117.3(5)
112.7(5)
treatment of 3 with 1 equiv of LiNPh2 (eq 3). Compound 4 is
a moderately air-sensitive, thermally stable, red-orange crystal-
line solid.
Figures 2 and 3 show ORTEP illustrations of 7 and 8,
respectively, with selected bond distances and angles given in
Table 2. The two compounds resemble a bent group 4
metallocene with syn-η4 bound TBM ligands. The CI-Ta-
Cpcent angles in 7 (134.8(8) °) and 8 (135.7(6) °) are very similar
to the Cpcent-Zr-Cpcent angle in Cp2ZrCl2 (132.5°)18 (see Table
2 for angles and definition of CI). The bonding of the fluorenyl
ligand to Ta in 8 is of particular interest. The Ta-C(23)
distance (2.336(4) Å) is substantially shorter than those of Ta-
C(29) (2.717(4) Å) and Ta-C(30) (2.650(3) Å), implying that
the fluorenyl ligand in 8, at least in the solid state, is close to
assuming an η3 geometry.19
Single crystals suitable for crystallography were obtained by
allowing a diethyl ether solution of 4 to evaporate slowly. The
result of this study is shown in Figure 1, and selected bond
lengths and angles are given in Table 1. The short Ta-N
distance (2.006(5) Å)15 and sp2-hybridized N atom suggest a
Ta-N π bond. As shown in Figure 1, the eclipsed conformation
is maintained. Compound 4 is a structural analogue of
CpZrCl2(NR2).16 Note that in 4, the phenyl groups of the amido
ligand are rotated away from the benzylidene arms of TBM to
avoid steric contacts.
The electronic properties of Cp′ and Cp are similar, and the
structure of 7 may be used as a model for that of 5. The
symmetrical binding of the TBM ligand in 7 (Ta-C distances
of 2.468(7), 2.336(6), and 2.477(6) Å) may be compared to the
Treating 3 with CpLi or Cp*Li affords Cp(TBM)TaMe2 (5)
and Cp*(TBM)TaMe2 (6), respectively (eq 4). These reactions
1
are nearly quantitative by H NMR spectroscopy. However,
(17) Schaefer, W. P.; Marsh, R. E.; Rodriguez. G.; Bazan, G. C. Acta
Crystallogr. 1996, B52, 465.
(18) Hunter, W. E.; Hrncir, D. C.; Bynum, R. V.; Penttila, R. A.; Atwood,
J. L. Organometallics 1983, 2, 750.
(19) Kowala, C.; Wunderlich, J. A. Acta Crystallogr. (Sect. B.) 1976,
32, 820.
the Cp* derivative is isolated in only 50% yield as a red oil
(15) Schaller, C. P.; Wolczanski, P. T. Inorg. Chem. 1993, 32, 131.
(16) Pupi, R. M.; Coalter, J. N.; Petersen, J. L.; J. Organomet. Chem.
1995, 497, 17.

Tribenzylidenemethane Ligand in Tantalum Complexes
J. Am. Chem. Soc., Vol. 120, No. 48, 1998 12515
Table 2. Selected Bond Distances (Å) and Angles (deg) for 5,a 7, and 8b
complex
MeL
MeR
CI
CT
CL
CR
LCp-Ta-CI
Cp(TBM)TaMe2 (5)
Cp′(TBM)TaMe2 (7)
Flu(TBM)TaMe2 (8)
2.2445(31)
2.190(7)
2.174(4)
2.2445(31)
2.220(7)
2.168(4)
2.2943(14)
2.259(6)
2.217(3)
2.429(8)
2.468(7)
2.431(3)
2.429(8)
2.336(6)
2.364(4)
2.429(8)
2.477(6)
2.525(3)
131.5(9)
134.8(8)
135.7(6)
a Previously reported (see ref 11). b The atom designators are defined by the following figure (LCp ) Cp, Cp; or Flu).
Figure 2. ORTEP drawing of 7. Ellipsoids are drawn at 30%
probability. Hydrogen atoms are omitted for clarity.
Figure 4. ORTEP drawing of 9. Ellipsoids are drawn at 30%
probability. Hydrogen atoms are omitted for clarity. Important distances
(in Å): Ta-C(1), 2.345(9); Ta-C(2), 2.237(8); Ta-C(3), 2.28(1); Ta-
C(4), 2.896(4); C(1)-C(2), 1.47(1); C(1)-C(3), 1.45(1); C(1)-C(4),
1.39(1).
progressively accentuates the deviation from symmetric η4
binding to a strictly η2-bidentate form.
One equivalent of 3 reacts with Na[HB(pz)3] (HB(pz)3 ) tris-
(pyrazolylborate)) in tetrahydrofuran to give [HB(pz)3](TBM)-
TaMe2 (9) in 60% yield (eq 6). The solid-state structure (Figure
Figure 3. ORTEP drawing of 8. Ellipsoids are drawn at 30%
probability. Hydrogen atoms are omitted for clarity.
binding of TBM in the isoelectronic zirconium complex
[Cp*(TBM)ZrCl2]-.11 In the anion the two benzyl carbons syn
to the Cp* ligand show markedly tighter binding than the single
anti carbon (2.48(1) Å and 2.45(1) Å vs 2.78(1) Å). The trans
effect of the Cp* ligand creates a ground state in which the
TBM ligand approximates an ene-diyl fragment. The internal
C-C bond lengths show a long-long-short pattern as well
(1.47(1) Å and 1.42(1) Å vs 1.38(1) Å). As we will show
below, the presence of bulky tris(pyrazolylborate) ligands
4) reveals that 9 contains a three-coordinate HB(pz)3 ligand and
a distorted TBM ligand (d(Ta-C(4)) ) 2.895(4) Å vs d(Ta-

12516 J. Am. Chem. Soc., Vol. 120, No. 48, 1998
Rodriguez et al.
Figure 6. ORTEP drawing of 11. Ellipsoids are drawn at 30%
probability. Hydrogen atoms are omitted for clarity. Important distances
(in Å): Ta(1)-C(1), 2.282(4); Ta(1)-C(2), 2.509(5); Ta(1)-C(3),
2.365(5); Ta(1)-C(4), 2.446(5); C(1)-C(2), 1.412(7); C(1)-C(3),
1.462(7); C(1)-C(4); 1.418(7).
Figure 5. ORTEP drawing of 10. Ellipsoids are drawn at 30%
probability. Hydrogen atoms are omitted for clarity. Important distances
(in Å): Ta-C(1), 2.675(4); Ta-C(2), 2.251(8); Ta-C(3), 2.287(8);
Ta-C(4), 3.842(5); C(1)-C(2), 1.47(1); C(1)-C(3), 1.49(1); C(1)-
C(4), 1.36(1).
C(1)) ) 2.345(9) Å, d(Ta-C(2)) ) 2.357(8) Å and d(Ta-C(3))
) 2.28(1) Å). For 9 in the solid state the TBM ligand is η3-
bound.
Compound [HB(3,5-Me2-1-pz)3](TBM)TaMe2 (10) is ob-
tained from 3 and K[HB(3,5-Me2-1-pz)3] in 73% yield. The
X-ray structure of 10 is shown in Figure 5. The larger size of
the [HB(3,5-Me2-1-pz)3] ligand relative to [HB(pz)3] changes
the Ta-TBM relationship. Relative to compound 9, there is a
reduction in the Ta-C(2) (2.251(8) Å) and Ta-C(3) (2.287(8)
Å) bond distances, while the Ta-C(1) (2.675(4) Å) and Ta-
C(4) (3.842(5) Å) distances are significantly longer. Orbital
overlap between Ta and C(1) and C(4) can be ruled out based
on these metrical parameters, and as a result TBM is best
described as η2-bound. The contraction of the C(2)-C(1)-
C(3) angle from 110.8(8)° in 9 to 104.5(7)° in 10 is also notable.
Furthermore, the sum of angles about C(1) (359.2(9) °) in 10 is
consistent with sp2 hybridization. This Ta-TBM interaction
is best described as an exo-benzylidene(R,R′-diphenyl)tantala-
cyclobutane fragment.20
Figure 7. Coordination sphere of Tantalum in 10, 9, and 11.
infrared spectrum of 11 in tetrahydrofuran contains absorptions
at 2484 cm-1 (terminal B-H) and 2139 cm-1 (bridging B-H-
Ta). These bands confirm that the Ta-H-B interaction is also
present in solution and that tantalum is nine-coordinate as shown
below.
The reaction of 3 and K[H2B(pz)2] affords [H2B(pz)2](TBM)-
TaMe2 (11). Figure 6 shows the crystallographically determined
coordination environment of tantalum. The TBM core in 11 is
tightly bound as shown by the Ta-C(1) (2.282(5) Å), Ta-C(2)
(2.509(5) Å), Ta-C(3) (2.365(5) Å), and Ta-C(4) (2.446(5)
Å) bond distances and adopts the typical “domed” shape with
the inner carbon closest to the metal. The TBM framework is
pseudoaxial to one of the pyrazolyl arms (C(1)-Ta-N(1) )
177.8(2)°) and staggered relative to the other three ligands. The
boatlike puckering of the six-membered chelating ring [Ta(1),
B(1), and N(1)-N(4)], and the short Ta(1)-B(1) distance
(2.900(5) Å) suggest a three-center, two-electron B-H-Ta
bond. These structural features are nearly identical to those
found in (H(µ-H)B(1-pz)2)TaMe3Cl (Ta-B ) 2.897(12) Å)21
and Cp(H(µ-H)B(1-pz)2)ZrCl2 (Zr-B ) 2.957 (5) Å).22 The
The relative Ta-TBM interactions in 9-11 are easily
compared by inspecting the coordination sphere at tantalum
(Figure 7). Replacing [HB(3,5-Me2-1-pz)3] (cone angle of 224°)
with [HB(pz)3] (cone angle of 184°)23 reduces the steric
envelope around Ta. The more open environment around 9
allows TBM to more closely approach the metal. The η3
bonding mode of TBM, coupled with the π overlap between
C(1) and C(4) observed in 9 resists a simple valence-bond
description. We propose the combination of resonance struc-
tures A and B in Scheme 1 to account for the observed molecular
(20) For an example of an η2 complex of TMM, Cp*2Zr(η2-TMM), see:
Herberich, G. E.; Kreuden, C.; Englert, U. Angew. Chem., Int. Ed. Engl.
1994, 33, 2465.
(21) Reger, D, L.; Swift, C. A.; Lebioda, L. J. Am. Chem. Soc. 1983,
105, 5343.
(22) Reger, D. L.; Mahtab, R.; Baxter, J. C.; Lebioda, L. Inorg. Chem.
1986, 25, 2046.
(23) For a recent review of poly(pyrazolyl)borate chemistry refer to:
Trofimenko S. Chem. ReV. 1994, 93, 943.

Tribenzylidenemethane Ligand in Tantalum Complexes
J. Am. Chem. Soc., Vol. 120, No. 48, 1998 12517
Scheme 1
geometry. Exchanging [HB(pz)3] for [H2B(pz)2] increases
electron demand at the metal and opens a coordination site. As
a result the TBM ligand in 11 can coordinate via four carbons,
donating a total of six π electrons to the metal.
Figure 8. ORTEP drawing of 13. Ellipsoids are drawn at 30%
probability. Hydrogen atoms are omitted for clarity.
Solution characterization by 1H NMR spectroscopy provides
information complementary to the solid-state data. The tem-
perature dependence of the signals for the three benzylidene
arms can be used to estimate the rate of TBM rotation. In the
case of 9, the three arms are equivalent and there is no evidence
of signal broadening, even at temperatures as low as -90 °C.
For 10, these three signals remain inequivalent (δ7.34, 2.13,
2.00) and do not merge, even after heating to 80 °C. Thus, a
static structure is observed for 10 relative to the 1H NMR time
scale.24 In the case of 11, three distinct benzylidene signals (δ
5.75, 5.06, 4.38) coalesce into a sharp singlet at temperatures
above 5 °C. The fluxional exchange in 11 was modeled by
line shape analysis giving ∆H ) 11(1) kcal/mol and ∆S )
4.0(0.6) eu.25
comparison, it is known that hydrogenation of Cp*2ZrMe2
produces the hydride species Cp*2Zr(H)2 cleanly.30
Addition of 1 equiv of 2,6-Me2C6H3NC to 12 gives a complex
1
mixture of products. When 2 equiv of isocyanide are added,
The different dynamic H NMR spectra of 9 and 10, which
1
the reaction is nearly quantitative by H NMR spectroscopy.31
have similar geometries except for the TBM binding mode, may
be understood in the following manner. In order for TBM
rotation to occur in 10, the exo-benzylidene arm needs to
coordinate forming an η4-TBM intermediate. The steric hin-
drance of the [HB(3,5-Me2-1-pz)3] ligand raises the energy of
this η4 intermediate. Inspection of the molecular frameworks
in Figure 7 suggests that formation of the required η4-TBM
intermediate would be facile for 9, resulting in faster rates of
exchange.
The product 13 is characterized by one Cp* signal and six
methyl groups. In addition, there are two signals in the olefin
region, and four sets of doublets (J ) 15 Hz for two sets, J )
8 Hz for the other two sets). These spectroscopic data, and
elemental analysis, are consistent with double isocyanide
insertion. All attempts to incorporate only one 2,6-Me2C6H3-
NC molecule into 12 were frustrated by the fast incorporation
of the second equivalent.
Single crystals of 13 suitable for X-ray crystallography were
obtained by slowly cooling a concentrated diethyl ether solution.
The resulting solid-state structure is shown in Figure 8, while
selected bond distances and angles are given in Table 3. Upon
insertion into a Ta-Me bond, one of the isocyanides forms a
Reactivity. In this section we report on the reactivity of
TBM-Ta metallocene mimics. Of special interest are the
differences between complexes containing the TBM-Ta frame-
work and the isoelectronic Cp-Zr counterpart.
Bercaw and co-workers originally reported the synthesis of
Cp*(TMM)TaMe2 (12).26 Activation of 12 to a reactive 14-
electron species, [Cp*(TMM)TaMe]+, was initially attempted.
No evidence of polyethylene formation was observed when a
solution of 12 and MAO ([Al]/[Ta] ) 800) was exposed to 1
atm of ethylene. Treatment of 12 with stoichiometric activators
B(C6F5)3 and [HMe2NPh][B(C6F5)4] led to decomposition.27,28
Addition of H2 to 12 in the presence of excess PMe3 results in
clean formation of methane, isobutylene, and Cp*TaH4(PMe3)2
(eq 7). The tetrahydride product has been previously character-
ized from the hydrogenation of Cp*TaMe4 with PMe3.29 For
(28) For reactions of [HMe2NPh][B(C6F5)4] with dimethyl zirconocenes
and zirconium alkyls see: (a) Hlatky, G. G., Turner, H. W.; Eckman, R. R.
J. Am. Chem. Soc. 1989, 111, 2728. (b) Bochman, M.; Lancaster, S. J.;
Hursthouse, M. B.; Abdul Malik, K. M. Organometallics 1994, 13, 2235.
(c) Bochman, M.; Lancaster, S. J. Organometallics 1993, 12, 633. (d)
Bochman, M. Angew. Chem., Int. Ed. Engl. 1992, 31, 1181. (e) Bochman,
M.; Jaggar, A. J.; Nicholls, J. C. Angew. Chem., Int. Ed. Engl. 1990, 29,
780. (f) Bochman, M.; Karger, G.; Jaggar, A. J. J. Chem. Soc., Chem.
Commun. 1990, 1038.
(29) Mayer, J. M.; Wolczanski, P. T.; Santarsiero, B. D.; Olson, W. A.;
Bercaw, J. E. Inorg. Chem. 1983, 22, 1149.
(30) (a) Miller, F. D.; Sanner, R. D. Organometallics 1988, 7, 818. (b)
Lin, Z.; Marks, T. J. J. Am. Chem. Soc. 1987, 109, 7979. (c) Jordan, R. F.;
Bajgur, C. S.; Dasher, W. E.; Rheingold, A. L. Organometallics 1987, 6,
1041.
(24) Sandstrom, J. Dynamic NMR Spectroscopy; Academic Press: New
York, 1982.
(31) For other examples of 2,6-Me2C6H3NC insertion into Ta-Me bonds
see: (a) Galakhov, M. V.; Gomez, M.; Jimenez, G.; Royo, P. Organome-
tallics 1995, 14, 2843. (b) Galakhov, M. V.; Gomez, M.; Jimenez, G.;
Penlinghelli, M. A.; Royo, P. Organometallics 1995, 14, 1901. For reactions
of isocyanide insertion into Cp2ZrR2, see (c) Bristow, G. S.; Lappert, M.
F.; Atwood, J. L.; Hunter, W. E. in ComprehensiVe Organometallic
Chemistry; Wilkinson, G. W., Stone, F. G. A., Abel, K. W., Eds.;
Pergamon: Oxford, 1982; Chapter 23, p 600.
(25) Modeled using the Dynmac software package written for the
Macintosh computer. See: Heinekey; M. D.; Hinkle, A. S.; Close, J. D. J.
Am. Chem. Soc. 1996, 118, 5353.
(26) Mayer, J. M.; Curtis, C. J.; Bercaw, J. E. J. Am. Chem. Soc. 1983,
105, 2651.
(27) For reactions of B(C6F5)3 with dimethyl zirconocenes see: Yang,
X.; Stern, C. L.; Marks, T. J. J. Am. Chem. Soc. 1994, 116, 10015.

12518 J. Am. Chem. Soc., Vol. 120, No. 48, 1998
Table 3. Selected Bond Distances and Angles for 13
Rodriguez et al.
atoms
distance (Å)
atoms
angles (deg)
Ta-N(1)
2.160(5)
2.150(6)
1.973(5)
2.216(7)
2.219(7)
1.426(8)
1.268(8)
1.47(1)
Ta-C(33)-C(32)
Ta-N(2)-C(29)
Ta-C(29)-C(30)
C(31)-C(32)-C(33)
C(33)-C(32)-C(34)
C(34)-C(32)-C(31)
105.0(5)
79.6(4)
131.2(5)
110.4(6)
124.8(8)
124.6(8)
Ta-C(19)
Ta-N(2)
Ta-C(29)
Ta-C(33)
N(2)-C(29)
N(1)-C(19)
C(32)-C(33)
C(32)-C(31)
C(32)-C(34)
1.50(1)
1.33(1)
typical iminoacyl ligand containing N(1) and C(19). The second
isocyanide is part of an unusual metallabicyclic unit containing
an amido moiety (Ta-N(2) ) 1.973(5) Å). The distances
within this fragment are consistent with two single bonds
(C(31)-C(32) ) 1.50(1), C(32)-C(33) ) 1.47(1) Å) and one
double bond (C(32)-C(34) ) 1.33(1) Å). Based on the similar
coupling constants, it is apparent that the four sets of doublets
are due to the H atoms on C(31) and C(33), while the olefinic
signals originate from the two H atoms on C(34). The structure
of 13, as shown in eq 8, is consistent with both solid state and
solution data.
Figure 9. Projection representations of the frontier orbitals of e
symmetry for a pyramidal ML3 fragment and a TMM ligand. Sym and
anti indicate symmetry or antisymmetry with respect to the xz plane.
approximate MO calculations33 and density functional theory
via the Amsterdam Density Functional (ADF) implementation34
to explore these observations further. Specifically, we will
compare the bonding and geometric preference in the known
Fe(II) complex (TMM)Fe(CO)3 (A) to those in the model
Ta(V) complex (TMM)TaMe3 (B).
(TMM)ML3 Frontier Orbitals. Because the TBM ligand
in the structures of 2-4 are “domed”, we will consider first
the relevant orbitals of a domed (C3V) TMM ligand. As
expected, the primary interactions between a TMM ligand and
a metal atom are via the π orbitals of the ligand. The π
molecular orbitals of a C3V TMM ligand are the 1a1, 1e, and
2a1 MOs; if the ligand is considered to be a six-π-electron
(TMM)2- ligand, then the nondegenerate 1a1 and doubly
degenerate 1e MOs are filled, the latter being the HOMO of
the dianionic ligand. The filled 1a1 MO and the 2a1 LUMO
are axially symmetric and cannot contribute to a rotational
preference for a staggered or eclipsed geometry. Thus, our
discussion will focus on the interactions between the 1e MO of
the (TMM)2- and a d6 [Fe(CO)3]2+ or a d0 [TaMe3]2+ fragment.
The frontier orbitals of a trigonal pyramidal ML3 fragment have
been previously described by many groups.35 The d-based
orbitals of a pyramidal ML3 fragment adopt a familiar “three-
below-two” splitting pattern, consistent with its octahedral
parentage. The lower set of orbitals consists of a1 and e orbitals,
the lobes of which point between the extensions of the M-L
bonds to the “top” of the fragment. The upper set of orbitals is
an e set that has lobes that lie along the M-L bond extensions.
A qualitative picture of the lower (1e) and upper (2e) sets of
d-based e orbitals, projected along the 3-fold axis of the ML3
fragment, is shown in Figure 9. These orbitals are labeled as
sym or anti to indicate symmetry or antisymmetry with respect
to a vertical mirror plane that contains one of the M-L bonds
(the xz plane in the coordinate system of the figure).
The metallabicyclic fragment probably forms by initial
insertion of isocyanide to give an iminoacyl group, followed
by nucleophilic attack by one of the methylene carbons of TMM
onto the iminoacyl CR. In the case of Cp2ZrR231c and Cp*[C4-
H4B-N(CHMe2)2]TaMe2,9 isocyanide insertion occurs only into
the σ ligands within the metallocene wedge.
Addition of ethylene to solutions of 6, 7, or 8 with MAO
results in a sudden temperature increase and the formation of
polyethylene. Complex 5 could not be used in these test
reactions because it is insoluble in hydrocarbon solvents. Long-
term stability of the active species is an important concern since
many metallocene-based catalysts show thermal degradation.32
Gas uptake measurements show that catalysts derived from 6-8
decompose within minutes after activation. Well-defined cat-
ionic species could not be observed using stoichiometric
activators such as B(C6F5)3 and [HMe2NPh][B(C6F5)4]. Com-
bining 5-8 with either of these reagents did not provide
1
polymerization catalysts. Analysis of these reactions by H
NMR spectroscopy fails since oily mixtures form which phase
separate from the solvent.
Figure 9 also shows representations of the 1e set of π orbitals
of the TMM ligand, in eclipsed and staggered conformations
relative to the ML3 fragment. The comparison of the “match”
of the TMM 1e π orbitals with the metal-based e orbitals is
Electronic Structure Calculations. As noted earlier, the
TBM framework eclipses the monodentate ligands in (t-Bu-
TBM)TaMe3 (2), (TBM)TaMe2Cl (3), and (TBM)Ta(NPh2)Me2
(4). The eclipsed conformation seems counterintuitive on steric
grounds and is in marked contrast to the staggered conformation
observed for (TMM)Fe(CO)3. The eclipsed conformation of
the TBM complexes of d0 Ta(V) complexes, even in the
presence of a bulky diphenylamido ligand, suggests an electronic
origin for the conformation. We have used Fenske-Hall
(33) Hall, M. B.; Fenske, R. F. Inorg. Chem. 1972, 11, 768.
(34) (a) Amsterdam Density Functional program, versions 1.1, 2.0, and
2.1, Theoretical Chemistry, Vrije Universiteit, Amsterdam. (b) Baerends,
E. J.; Ellis, D. E.; Ros, P. Chem. Phys. 1973, 2, 41. (c) te Velde, G.;
Baerends, E. J. J. Comput. Phys. 1992, 99, 84.
(35) See, for example: (a) Elian, M.; Chen, M. M. L.; Mingos, D. M.
P.; Hoffmann, R. Inorg. Chem. 1976, 15, 1148. (b) Calabro, D. C.; Hubbard,
J. L.; Blevins, C. H., II.; Campbell, A. C.; Lichtenberger, D. L. J. Am.
Chem. Soc. 1981, 103, 6839. (c) Albright, T. A.; Burdett, J. K.; Whangbo,
M. H. Orbital Interactions in Chemistry; Wiley: New York, 1985.
(32) Reddy, S. S.; Sivaram, S. Prog. Polym. Sci. 1995, 20, 309.

Tribenzylidenemethane Ligand in Tantalum Complexes
J. Am. Chem. Soc., Vol. 120, No. 48, 1998 12519
Figure 11. Frontier molecular orbital interactions in the eclipsed and
staggered geometries of (TMM)TaMe3 (B).
stabilization of the formally Huckel antiaromatic (4 π electron)
TMM ligand. This situation is reminiscent of the stabilization
of cyclobutadiene in (η4-C4H4)Fe(CO)3, an effect we have called
“metalloaromaticity.”37
Figure 10. Frontier molecular orbital interactions in the eclipsed and
staggered geometries of (TMM)Fe(CO)3 (A).
striking and provides the first indication of the electronic source
of differing rotational preference. In the eclipsed conformation,
the TMM 1e π sym and anti orbitals are oriented to overlap
favorably with their counterparts in the lower 1e orbitals of the
ML3 fragment. In the staggered conformation, the TMM 1e
orbitals overlap better with the upper 2e orbitals of ML3. We
will see that these differences in the optimum TMM-ML3 e
interactions are key in understanding the rotational preferences
in complexes A and B.
Orbital Interactions in (TMM)TaMe3. The bonding in
(TBM)TaMe3, 1, was investigated using the model complex
(TMM)TaMe3, B. A Fenske-Hall MO diagram for complex
B in eclipsed and staggered geometries is given in Figure 11.
The most obvious difference in the electronic structures of A
and B is the d0 electron count for the [TaMe3]2+ fragment of
B. As a consequence, both the 1e and 2e orbitals of the metal
fragment are empty and available for bonding interactions with
the filled 1e orbitals of (TMM)2-
.
In eclipsed B, the (TMM)2- 1e orbitals interact significantly
with the empty 1e, and, to a lesser extent, the 2e orbitals of
[TaMe3]2+, yielding the strongly bonding 1e HOMO of the
complex (Table 4). In staggered B, the 1e orbitals of [TaMe3]2+
have almost zero overlap with the TMM ligand, and the primary
interaction in the 1e HOMO of the complex is between the 1e
orbitals of (TMM)2- and the 2e orbitals of [TaMe3]2+. The 2e
orbitals of [TaMe3]2+ are energetically farther from the (TMM)2-
1e orbitals than are the 1e orbitals of [TaMe3]2+. As a result,
the stabilization of the HOMO is considerably diminished
relative to the eclipsed conformation. The greater stabilization
of the HOMO is the electronic driving force for the seemingly
nonintuitive eclipsed conformation.
Orbital Interactions in (TMM)Fe(CO)3. The bonding and
rotational preference of known complex A have been extensively
studied.36 Our Fenske-Hall molecular orbital description of
eclipsed and staggered A, given in Figure 10, reinforces the
conclusions of these earlier studies. Consistent with the above
discussion, only interactions between the fully occupied 1e
HOMO of (TMM)2- and the d-based 1e and 2e orbitals of [Fe-
(CO)3]2+ are considered. In the d6 [Fe(CO)3]2+ fragment, the
1e orbital is filled whereas the 2e orbital is unoccupied. In the
eclipsed conformation of A, the TMM 1e orbital interacts
primarily with the 1e orbital of [Fe(CO)3]2+. This unfavorable
filled-filled interaction results in destabilization of the 2e HOMO
of the complex.
Rotation of the TMM ligand to a staggered conformation
relative to the [Fe(CO)3]2+ fragment switches the primary
interaction of the TMM 1e orbital from the 1e to the 2e orbital
of [Fe(CO)3]2+, as is evident from the percent compositions of
the 1e and 2e MOs of A in the two conformations (Table 4).
Whereas the 2e HOMO of eclipsed A is antibonding because
of the filled-filled interaction, the 2e HOMO of staggered A is
essentially nonbonding and is therefore at lower energy. The
stabilization of the HOMO of staggered A relative to eclipsed
A provides an electronic preference for the staggered conforma-
tion in addition to the evident steric preference.
The electronic driving force for the preferred eclipsed
conformer of B is similar to that which leads to an eclipsed
conformation in the d0 complex WMe6. In an elegant treatment
of Group VI MX6 complexes, Albright et al. invoked the second-
order Jahn-Teller theorem to propose that WMe6 adopts a
trigonal prismatic geometry in preference to a more intuitive
octahedral geometry.38 Our orbital explanation here for B is
analogous to the one that they used for WMe6: By adopting an
eclipsed geometry, the lower set of empty metal-based e orbitals
2+
in the MMe3 fragment can be utilized more effectively as
acceptors of ligand charge than in an octahedral geometry and
are thus destabilized farther in energy from the filled molecular
orbitals.
The strong bonding interaction of the 1e orbital of (TMM)2-
with the 2e orbital of [Fe(CO)3]2+ in staggered A leads to
Density Functional Calculations. Although the Fenske-
(36) (a) Albright, T. A. Trans. Am. Crystallogr. Assoc. 1980, 16, 35. (b)
Albright, T. A.; Hofmann, P.; Hoffmann, R. J. Am. Chem. Soc. 1977, 99,
7546. (c) Branchadell, V.; Deng, L.; Ziegler, T. Organometallics 1994, 13,
3115.
(37) Bursten, B. E.; Fenske, R. F. Inorg. Chem. 1979, 18, 1760.
(38) Kang, S. K.; Tang, H.; Albright, T. A. J. Am. Chem. Soc. 1993,
115, 1971.

12520 J. Am. Chem. Soc., Vol. 120, No. 48, 1998
Rodriguez et al.
Table 4. Fenske-Hall Energies and Percent Contributions for the 1e and 2e Frontier Molecular Orbitals of Staggered and Eclipsed
(TMM)Fe(CO)3 and (TMM)TaMe3
(TMM)Fe(CO)3
staggered
(TMM)Fe(CO)3
eclipsed
(TMM)TaMe3
eclipsed
(TMM)TaMe3
staggered
1e MO
2e MO
1e MO
2e MO
1e MO
2e MO
1e MO
2e MO
MO energy, eV
% ML3 1e
% ML3 2e
-11.02
1.6
36.8
-8.89
96.6
0.3
-11.11
31.2
15.0
-8.28
57.9
33.0
6.7
-13.80
25.9
8.0
64.6
-6.40
53.2
38.8
6.5
-12.89
2.4
29.9
-9.28
94.0
4.0
% TMM 1e
59.3
1.5
47.0
56.9
0.3
Table 5. Relative Energies and Selected Calculated Geometric Parameters of the Staggered and Eclipsed Conformations of (TMM)Fe(CO)3
(A) and (TMM)TaMe3 (B)
(TMM)Fe(CO)3
staggereda
(TMM)Fe(CO)3
eclipsed
(TMM)TaMe3
eclipsedb
(TMM)TaMe3
staggered
rel energy, kJ/mol
M-L, C
0
64
0
26
2.19
2.28
2.43
116
1.74 (1.81)
1.93 (1.94)
2.11 (2.13)
100 (99)
1.72
1.94
2.15
99
2.16 (2.19)
2.23 (2.20)
2.33 (2.33)
102 (97)
M-Cc, Cc
M-Ct, Cc
L-M-L, deg
a Averaged experimental values from ref 14c are given in parentheses. b Averaged experimental values from the structure of 2 (ref 11) are given
in parentheses. c Cc is the central carbon atom of the TMM ligand; Ct is one of the terminal carbon atoms of the TMM ligand.
Hall method yields useful molecular orbitals and charge
distributions, it cannot be used reliably to calculate the
geometries of molecules.39 Therefore, we used density func-
tional calculations to estimate the rotational barriers in com-
plexes A and B.40 The rotational barriers were determined as
the difference in total energy between the C3V-optimized
geometries of the eclipsed and staggered conformers of each
molecule. Table 5 gives some of the geometric parameters from
the optimized structures of A and B in staggered and eclipsed
conformations. For staggered A and eclipsed B, the calculated
geometries are compared to the experimental structures of
(TMM)Fe(CO)3 and 2.11
14c
The calculated rotational barrier for A is 64 kJ/mol, in
reasonable accord with the experimental value for derivatives
of A (∆GI ∼ 70-75 kJ/mol)41 and consistent with previously
calculated values.36 The calculated geometric parameters are
in good agreement with those from the experimental structure.42
Notably, the geometry of the Fe(CO)3 fragment and the Fe-
TMM distances are nearly invariant between the two conform-
ers; the rotation of the TMM ligand in A is essentially a pure
rotation of the ligand with minimal geometric reorganization
of the molecule.
Figure 12. Variation of the total energy of the d6 [Fe(CO)3]2+ and d0
[TaMe3]2+ ML3 fragments as a function of θ, the L-M-L angle.
is a minimum in the competition between the CO ligands for
back-bonding interactions with the Fe 3dπ orbitals.35a The
stabilization due to back-bonding is a more important factor in
the geometry of the fragment than is the steric interaction
between the CO ligands.
The rotational barrier for B was found to be 26 kJ/mol,
considerably less than that for A. Significantly, the geometry
of the TaMe3 fragment changes drastically between the two
conformers. In particular, the calculated Me-Ta-Me angles
in B increase markedly (from 102° to 116°) as the TMM ligand
is rotated from the favorable eclipsed conformation to the
unfavorable staggered conformation. It is the ability of the
TaMe3 fragment to respond and reorient that leads to the much
lower rotational barrier in B relative to A, as we will now briefly
discuss.
Figure 12 shows the variation of the relative density functional
energies of the d6 [Fe(CO)3]2+ and d0 [TaMe3]2+ fragments as
a function of the L-M-L angle, denoted θ. The [Fe(CO)3]2+
fragment has minimum energy at θ near 90°, and the energy
rises rapidly as θ increases above 100°. At the 90° angle, there
On the basis of VSEPR considerations, the d0 [TaMe3]2+
fragment might be expected to adopt a trigonal planar confor-
mation with θ ) 120°. Instead, Figure 12 indicates that this
fragment has a shallow potential surface with a minimum energy
at θ ) 108°. A similar preference for pyramidalization has
been seen in theoretical studies of other d0 MX3 fragments, such
as ScH3, [TiH3]+, and [TiMe3]+.43 In fact, Jolly and Marynick
found the optimized Me-Ti-Me angle in [TiMe3]+ to be
109°,43a virtually the same as we find here for [TaMe3]2+. This
nonintuitive distortion in these d0 fragments can be viewed as
a result of attempting to maximize the donation from the methyl
ligands to the metal d orbitals; in particular, a reduction of
symmetry from trigonal planar to trigonal pyramidal facilitates
donation into the doubly degenerate (dxz, dyz) metal orbitals,
increasing the overall ligand-to-metal donation.
(39) Bursten, B. E. Pure Appl. Chem. 1991, 63, 839.
(40) For a review of density functional theory applied to organometallic
molecules, see: Ziegler, T. Chem. ReV. 1991, 91, 651.
(41) Magyar, E. S.; Lillya, C. P. J. Organomet. Chem. 1976, 116, 99.
(42) Our calculated geometric parameters vary slightly from those
reported by Ziegler et al. (ref 36c) due to slight differences in basis sets
and levels of optimization.
The pyramidalization of the [TaMe3]2+ fragment leads to
2
2
mixing of the (dx -y , dxy) and (dxz, dyz), which generates the
(43) (a) Jolly, C. A.; Marynick, D. S. Inorg. Chem. 1989, 28, 2893. (b)
Kaupp, M.; Schleyer, P. v. R. J. Phys. Chem. 1992, 96, 7316. (c) Siegbahn,
P. E. M. J. Phys. Chem. 1993, 97, 9096.

Tribenzylidenemethane Ligand in Tantalum Complexes
J. Am. Chem. Soc., Vol. 120, No. 48, 1998 12521
1H and 13C NMR spectra were recorded using a Bruker AMX-400 NMR
spectrometer at 400.1 and 100.6 MHz, respectively. Toluene, pentane,
diethyl ether, and tetrahydrofuran were distilled from benzophenone
ketyl. Elemental analyses were carried out by Desert Analytics. The
preparations of Cp*(TMM)TaMe2,26 Li2(TBM)(TMEDA)2,46 Li2(t-
BuTBM)(TMEDA)2, and Li2(TMM)(TMEDA)2,47 are available in the
literature. MAO was purchased from AKZO Chemicals (toluene, 6.4
wt % Al, type 4).
hybridized 1e and 2e sets depicted in Figure 9. In the
unfavorable staggered conformation of B, the TMM ligand
cannot effectively utilize these hybrids, so the [TaMe3]2+
fragment responds by, in essence, “unhybridizing” itself via
flattening out. By so doing, the TMM ligand can interact
preferentially with the vertically oriented (dxz, dyz) orbitals, which
is preferable to interacting with the hybrids for the unfavorable
conformer. The effect of this response of the metal fragment
to the rotational barrier of the molecule is striking: A calculation
of the rotational barrier of B with the [TaMe3]2+ fragment fixed
at the optimized geometry in the eclipsed conformation yields
a value of 160 kJ/mol, more than six times greater than the
value obtained when both conformers are optimized.
In summary, there is clearly an electronic source for the
preference of d0 (TBM)TaX3 complexes to adopt an eclipsed,
rather than staggered, conformation. However, the relative
softness of the potential surface for distortion of the molecule
leads to much lower rotational barriers than those observed in
more electron-rich systems, such as (TMM)Fe(CO)3.
(TBM)TaMe3 (1). To a benzene solution of Ta(CH3)3Cl2 (5.42 g,
18.2 mmol) is added dropwise Li2(TBM)(TMEDA)2 (9.65 g, 18.2
mmol) in benzene and allowed to react for 1 h. The reaction mixture
is filtered through Celite, and the black solids are washed with a
minimum amount of benzene. The filtrate is brought to dryness and
redissolved in 30 mL of xylenes. The solvent is then evaporated under
vacuum, and the remaining solids are warmed to 60 °C while under at
full vacuum for an additional 1h. The remaining solids were extracted
with benzene, and the xylene dissolution and evaporation are repeated.
This procedure is followed with a second benzene extraction. The
product purification is achieved by solvent removal, suspension over
1
pentane, and filtration to provide an orange solid in 44% (4.12 g). H
NMR δ (CDCl3) 7.45 (d, 6H, o-Ph), 7.31 (t, 6H, m-Ph), 7.08 (t, 3H,
p-Ph), 4.41 (s, 3H, PhCH-), 0.79 (s, 9H, Ta(CH3)3); 13C NMR δ (CDCl3)
139.4, 128.4, 127.8, 125.8, 116.2, 102.1 (TBM), 73.3 Ta(CH3)3. Anal.
Calcd for C25H27Ta: C, 59.05; H, 5.36. Found: C, 59.50; H, 5.41.
(tert-Bu-TBM)TaMe3 (2). To a diethyl ether solution of Ta(CH3)3-
Cl2 (2.54 g, 8.50 mmol) is added dropwise Li2(tert-Bu-TBM)(TMEDA)2
(5.00 g, 8.50 mmol) in diethylether-THF (9:1), and the resulting
mixture is allowed to stir for 1 h. The solvent is then removed and
the resulting solid extracted with pentane. Residual TMEDA can be
removed by dissolving the product in octane followed by reevaporation.
It may be necessary to repeat this step one more time if retention of
TMEDA persists. This procedure affords (tert-Bu-TBM)TaMe3 as an
orange solid in 25% yield (1.19 g). 1H NMR δ (CDCl3) 7.36 (d, 6H,
o-Ph), 7.23 (t, 6H, m-Ph), 7.00 (t, 2H, p-Ph), 4.34 (s, 2H, PhCH), 4.26
(s, 1H, t-Bu-PhCH), 1.27 (s, 9H, t-Bu), 0.70 (s, 9H, Ta(CH3)3); 13C
NMR δ (CDCl3) 148.0, 139.5, 136.3, 128.4, 128.1, 127.7,125.8, 124.6,
116.1 (TBM), 102.4, 102.1, 102.0 (PhCH) 73.0 Ta(CH3)3, 34.3
(C(CH3)3), 31.5 (C(CH3)3). Anal. Calcd for C25H27Ta: C, 61.70; H,
6.24. Found: C, 61.54; H, 6.32.
(TBM)TaClMe2 (3). ZnCl2 in THF (0.536 g, 3.93 mmol) is added
to a stirring benzene solution of 1 (0.500 g, 0.98 mmol). The volatiles
are removed in vacuo, and the product is extracted with a minimum
amount of toluene. The solvent is then evaporated and the residue
suspended in 1,2-dimethoxyethane. Filtration of the resulting orange
precipitate provides the (TBM)TaClMe2 in 63% yield (0.328 g). 1H
NMR δ (C6D6) 7.36 (d, 6H, o-Ph), 7.12 (t, 6H, m-Ph), 6.88 (t, 3H,
p-Ph), 4.50 (s, 3H, PhCH-), 0.91 (s, 3H, Ta(CH3)2), 0.82 (s, 3H, Ta-
(CH3)2) 13C NMR δ (C6D6) 138.8, 129.3, 128.1, 127.2, 107.1, 102.8
(TBM); 72.57, 70.63 Ta(CH3). Anal. Calcd for C24H24TaCl: C, 54.50;
H, 4.58. Found: C, 54.14; H, 4.47.
Summary and Conclusion
We have demonstrated that TBM-Ta compounds are easily
prepared via salt metathesis routes. The piano stool compounds
1-4 prefer eclipsed geometries. This structural arrangement
is favored over the staggered isomer even when a large,
π-donating ligand is present, as in the case of 4. Metallocene
mimics, as shown by the structures of 7 and 8, show the
expected bent-sandwich geometry with TBM replacing one of
the Cp ligands. Despite these structural similarities, the TBM
ligand is more reactive than Cp and participates in hydrogenation
and insertion reactions at rates competitive with the Ta-CH3
bond. We attribute this reactivity to the greater σ character of
the bond between Ta and the outer carbons in TMM (or TBM).
The same σ character relates the eclipsed conformation of
complexes such as (TBM)TaMe3 to the trigonal prismatic
structure of WMe6. In the case of Cp-Ta the bonding
interaction contains more π character and thus it is expected to
be more inert. For these reasons, TMM and TBM are poor
dianionic surrogates for Cp, at least with d0 metals.
Compounds 9-11 demonstrate that TBM can accommodate
a continuum of bonding modes. This property allows it to
donate between four to six electrons and gives the ligand the
ability to respond to the steric and electronic demands of the
metal. Such bonding flexibility is potentially useful in catalytic
applications where vacant sites need to be generated for the
substrate to coordinate to the metal and is analogous to the η5-
η3 ring slip distortion observed for the cyclopentadienyl ligand.
Finally, the electronic structure calculations on the model
complex (TMM)TaMe3 demonstrate that the eclipsed conforma-
tion adopted by compounds 2-4 is a consequence of preferential
interactions between the filled 1e π orbitals of the (TMM)2-
with the empty 1e orbitals of the [TaMe3]2+ fragment. The
rotational preference of this d0 complex differs from that of the
d6 complex (TMM)Fe(CO)3 because in the latter, the 1e orbital
of the [Fe(CO)3]2+ fragment is filled and is unavailable as an
acceptor orbital. The pyramidal44 flexibility of the [TaMe3]2+
fragment leads to a low calculated rotational barrier (26 kJ/
mol) in (TMM)TaMe3.
(TBM)TaMe2(NPh2) (4). To a THF solution of 3 (0.20 g, 0.38
mmol) is added LiNPh2 (66 mg 0.38 mmol) in THF and stirred for 1
h. Solvent removal, extraction with toluene, and a pentane wash affords
the product as a red solid (0.208 g, 83%). 1H NMR: δ (C6D6) 7.41
(d, 6H, Ph), 7.17 (m, 8H, Ph), 6.91(m, 8H, Ph), 6.61 (t, 3H, p-Ph),
4.86 (s, 3H, (PhCH)3C2-), 1.14 (s, 3H, Ta-CH3), 0.23 (s, 3H, Ta-
CH3); 13C NMR: δ (C6D6) 144.6, 143.6 (Ph), 139.9 (PhCH)3C2-), 130.2,
129.5, 128.6, 128.1, 126.1, 125.4, 124.5, 121.5, 121.1, 118.1 (Ph), 100.0
(PhCH)3C2-), 68.5 (Ta-CH3), 62.8 (Ta-CH3). Anal. Calcd for TaC36-
H34N: C, 65.35; H, 5.19; N, 2.11. Found: C, 64.90; H, 5.01; N, 1.92.
Cp(TBM)TaMe2 (5). To a stirring solution of 3 (0.350 g, 0.663
mmol) in THF, is added LiCp (0.050 g, 0.694 mmol) also in THF, and
(45) Burger, B. J.; Bercaw, J. E. In Experimental Organometallic
Chemistry; Wayda, A. L., Darensbourg, M. Y., Eds.; ACS Symp. Ser. 353;
American Chemistry Society, Washington, D.C., 1987.
(46) (a) Wilhelm, D.; Clark, T.; Schleyer, P. v. R.; Buckl, K.; Boche, G.
Chem. Ber. 1983, 116, 1669. (b) Wilhelm, D.; Dietrich, H.; Clark, T.; Mahdi,
W.; Kos, A. J.; Schleyer, P. v. R. J. Am. Chem. Soc. 1984, 106, 7279. (c)
Wilhelm, D.; Clark, T.; Schleyer, P. von. R. J. Chem. Soc., Perkin Trans.
2 1984, 915.
Experimental Section
General Considerations. All manipulations were carried out using
either high-vacuum or glovebox techniques as described previously.45
(44) Collman, J. P.; Hegedus, L. S.; Norton, J. R.; Finke, R. G. Principles
and Applications of Organotransition Metal Chemistry; University Science
Books, Mill Valley, CA, 1987; Chapter 4.
(47) Jones, M. D.; Kemmett, R. D. W. AdV. Organomet. Chem. 1987,
27, 279.

12522 J. Am. Chem. Soc., Vol. 120, No. 48, 1998
Rodriguez et al.
the resulting solution is stirred for 2 h. During this period a red
precipitate appears. The volume is then reduced by one-half and cooled
for 1 h at -30 °C to induce further precipitation. The product is filtered
5.50, 5.47, 5,46 (all s, 3H ea., 4-Hpz), 2.25, 2.10, 2.05, 1.96, 1.88,
1.86 (s, 3H ea., CH3-pz) 2.13, 2.00 (s, 1H ea., PhHCdC(CHPh)2), 1.42,
1.41 (s, 3H ea., Ta-CH3). 13C NMR δ (C6D6) 151.9, 150.9, 144.4,
143.4, 142.5, 141.4 (3-pzquat and 5-pzquat), 129.4-124.7 (Ph), 109.2,
109.0, 108.9 (4-pz), 92.8, 70.8 (Ta-CH3), 85.9, 83.8 (PhCHdC(CH2-
Ph)2), 302, 27.2, 16.1, 15.4, 12.8, 12.1 (CH3-pz). Anal. Calcd for
TaC39H46BN6: C, 59.24; H, 5.88; N, 10.62. Found: C, 59.30; H, 5.90;
N, 10.52.
(H2Bpz2)TBMTaMe2 (11). A THF solution of KHBpz3 (17 mg,
0.095 mmol) was added dropwise to 3 (50 mg, 0.095 mmol) in THF.
After stirring for 1 h, the solvent was removed in vacuo, and the
resulting residue was dissolved with benzene. Filtration followed by
solvent evaporation affords the product in 65% yield (40 mg). Crystal
suitable for X-ray and elemental analysis were grown by slow diffusion
of pentane into a benzene solution of the product. 1H NMR _ (d8-
THF/C6D6, T g 25 °C) 7.50, 7.19 (bd ea., 2H ea., 3-Hpz and 5-Hpz),
7.27 (d, 6H, o-Ph), 7.10 (t, 6H, m-Ph), 6.83 (t, 3H, p-Ph), 5.65 (bt, 2H,
4-Hpz), 4.97 (s, 3H, (PhCH)3C2-), 0.51, 0.36 (Ta-CH3); 13C NMR _
(d8-THF/C6D6, T g 25 °C) 144.1 (PhCH)3C2-), 137.5, 131.8, 130.0,
129.0 (3-pz and 5-pz), 130.9, 129.0, 125.4, 119.2 (Ph), 107.9, 107.4,
107.2 (4-pz), 95.4 (center of a very broad lump, (PhCH)3C2-), 44.0,
36.7 (Ta-CH3). Anal. Calcd for TaC30H32BN4: C, 56.22; H, 5.05; N,
8.74. Found: C, 56.09; H, 4.80; N, 8.50.
Cp*Ta[(η2-C(Me)NAr′][η3-CH2C)(CH2)CH2C(Me)NAr′] (13).
A benzene solution of Cp*(η4-TMM)TaMe2 (0.100 g, 0.250 mmol) is
treated with a benzene solution of 2,6-dimethylphenyl isocyanide (65
mg, 0.500 mmol) and stirred for 10 min. The solvent is replaced with
a minimum amount of pentane and then filtered to remove any unreacted
materials. Pentane removal provides the product in 95% (0.157 g).
X-ray quality crystals were obtained by cooling a diethyl ether solution.
Analytically pure crystals were obtained from chilling a pentane/toluene
(9/1) solution for 4 days (-30 °C). 1H NMR δ (C6D6) 7.15-6.84 (m,
6H, aromatic H's), 4.25, 4.19 (s, 1H each, CdCH2), 4.03, 3.99, 2.89,
2.85, 2.28, 2.26, 2.09, 2.07 (all are s, CH2), 2.38 (s, 6H, CH3), 2.09,
1.89, 1.73, 1.50 (s, CH3), 1.79 (s, 15H, C5(CH3)5). 13C NMR δ (C6D6)
244.9, 158.8, 139.1, 136.1, 131.6, 131.5, 129.7, 128.9, 126.2, 122.4,
113.3, 102.0 (C5(CH3)5), 62.5, 58.5, 51.2, 24.3, 22.5, 21.4, 20.3, 19.5,
19.0, 11.7 (C5(CH3)5). Anal. Calcd for TaC34H45N2: C, 61.61; H, 6.86;
N, 4.22. Found: C, 61.72; H, 6.72; N, 4.17.
Typical Polymerization Procedure. A 10 mg sample of 5 is
dissolved in a toluene solution containing 18% MAO by weight. After
degassing the mixture, 1 atm of ethylene is placed above the catalyst
solution for 45 min at which point the reaction is quenched with 20
mL of 10% HCl in methanol followed by 20 mL of water. This mixture
is stirred for 10 h, and the salt-free polymer is gravity-filtered and
washed with methanol.
1
as a red solid in 54% yield (0.200 g) through a fine fritted funnel. H
NMR δ (THF-d8) 7.25 (d, 6H, o-Ph), 7.16 (t, 6H, m-Ph), 6.95 (t, 3H,
p-Ph), 5.31 (s, 5H, C5H5), 4.41 (s, 3H, PhCH-), 0.186 (s, 6H, Ta(CH3)2);
13C NMR δ (THF-d8) 143.1, 129.5, 128.6, 125.0, 112.8, 99.1 (TBM),
119.8 (C5H5), 48.3, 45.9 (Ta(CH3)2). Anal. Calcd for C29H29Ta: C,
62.35; H, 5.24. Found: C, 62.20; H, 5.26.
Cp*(TBM)TaMe2 (6). To a stirring solution of 3 (0.200 g, 0.378
mmol) in THF, is added LiCp* (0.054 g, 0.378 mmol) in THF and
stirred for 2 h. The THF is removed, and the resulting solids are
extracted with benzene. The solvent is then removed and the residue
suspended in pentane. Filtration of the orange-brown solids provides
1
the final product in 51% (0.121 g) and approximately 98% purity. H
NMR δ (C6D6) 7.38 (s, 6H, o-Ph), 7.17 (t, 6H, m-Ph), 6.94 (t, 3H,
p-Ph), 1.45 (s, 15H, C5(CH3)5), 0.11, -0.44 (s, 6H, Ta(CH3)2), the
-HCPh are broadened into the baseline; 13C NMR δ (CDCl3) 141.3,
129.2, 128.0, 127.7, 124.3, 117.2 (TBM), 121.8 (C5(CH3)5) 51.3, 46.6
(Ta(CH3)2), 11.1 (C5(CH3)5). Anal. Calcd for C32H33Ta: C, 63.23;
H, 6.02. Found: C, 63.40; H, 6.24.
Cp′(TBM)TaMe2 (7). To a stirring solution of 3 (0.250 g, 0.47
mmol) in tetrahydrofuran is added LiCp′ (0.041 g, 0.47 mmol in THF)
and stirred for 1 h. The THF is removed in vacuo followed byproduct
extraction with toluene. The solvent is removed, and the product is
washed with a minimum amount of pentane (0.175 g, 65%). 1H
NMR: δ (C6D6) 7.27 (d, 6H, o-Ph), 7.08 (t, 6H, m-Ph), 6.85 (t, 3H,
p-Ph), 5.24 (d, 1H, R-Cp′), 5.11 (dd, 1H, â-Cp′), 5.09 (d, 1H, R-Cp′),
4.69 (dd, 1H, â-Cp′), 4.37 (bs, 3H, (PhCH)3C2-), 1.75 (s, 3H, Cp-
CH3), -0.23 (s, 3H, Ta-CH3), -0.30 (s, 3H, Ta-CH3); 13C NMR: δ
(C6D6) 142.50 (PhCH)3C2-), 129.2, 124.8, 124.8, 123.7 (Ph) 112.9,
112.1, 111.5, 110.3, 108.5 (Cp′), 86.7 (PhCH)3C2-), 48.5 (Ta-CH3),
46.4 (Ta-CH3), 14.4 Cp-CH3). Anal. Calcd for C30H31Ta: C, 62.93;
H, 5.47. Found: C, 62.54; H, 5.46.
(Flu)(TBM)TaMe2 (8). A THF solution of LiFlu (33 mg, 0.189
mmol) is added to a THF solution of 3 (0.10 g, 0.189 mmol) to afford
an intensely red solution. This mixture is stirred for 30 min at which
point the solvent is removed in vacuo. The product is extracted with
30 mL of a diethyl ether/toluene mixture (1:1). The solvents are
removed to provide FluTaTBM(CH3)2 as a red solid (86 mg, 69%). 1H
NMR: δ (C6D6) 7.60-6.5 (23H, aromatic H's), 5.28 (s, 1H, (C6H4-
CHC6H4), 3.99 (s, 3H, (PhCH)3C2-), -0.83 (Ta-CH3), -1.12 (Ta-CH3).
13C NMR: δ (C6D6) 141.8 (PhCH)3C2-), 139.8, 137.8, 129.4-119.3
(aromatic C's), 115.7 (C6H4CHC6H4), 88.8 (PhCH)3C2-), 60.8 (Ta-CH3),
52.3 (Ta-CH3). Anal. Calcd for TaC37H33: C, 67.47; H, 5.06.
Found: C, 67.96; H, 5.16.
(HBpz3)TBMTaMe2 (9): A THF (THF ) tetrahydrofuran) solution
of NaHBpz3 (22 mg, 0.095 mmol) was added dropwise to 3 (50 mg,
0.095 mmol) in THF. After stirring for 1 h, the solvent was removed
in vacuo, and the resulting residue was dissolved with benzene.
Filtration followed by solvent evaporation affords the product in 60%
yield (40 mg). Crystals suitable for X-ray and elemental analysis were
grown by slow diffusion of pentane into a benzene solution of the
product. 1H NMR δ (C6D6) 7.97, 7.89, 7.87, 7.78, 7.75, 7.72 (all d,
1H ea., 3-Hpz and 5-Hpz), 7.14 (t, 6H, m-Ph), 7.01 (d, 6H, o-Ph) (6.91
(t, 3H, p-Ph), 6.27, 6.15, 6.08 (all t, 1H ea., 4-Hpz) 4.98 (s, 3H,
(PhCH)3C2-), 0.63, 0.19 (s, 3H, Ta-CH3). 13C NMR δ (C6D6) 148.8
(PhCH)3C2-), 143.9, 143.4, 139.8, 136.8, 134.9, 134.4 (3-pz and 5-pz),
131.1, 128.7, 128.3, 121.2 (Ph), 107.4, 106.8, 106.6 (4-pz), 104.3
(PhCH)3C2-), 80.6, 65.4 (Ta-CH3). Anal. Calcd for TaC33H34BN6:
C, 56.16; H, 4.87; N, 11.90. Found: C, 56.03; H, 4.85; N, 11.84.
[HB(3,5-Me2-1-pz)3]Ta[η2-(CHPh)2CdCHPh]Me2 (10). A THF
solution of KHB(1,5-Me2-1-pz)3 (32 mg, 0.095 mmol) was added
dropwise to 3 (50 mg, 0.095 mmol) in THF. After stirring for 1 h, the
solvent was removed in vacuo, and the resulting residue was dissolved
with benzene. Filtration followed by solvent evaporation affords the
product in 73% yield (55 mg). Crystals suitable for X-ray and elemental
analysis were grown by slow diffusion of pentane into a benzene
solution of the product. 1H NMR δ (C6D6) 7.65 (bd, 2H, Ph), 7.40
(bd, 2H, Ph), 7.34 (s, 1H, PhC(H)dC(CHPh)2), 7.33 (bd, 2H, Ph), 7.13
(bd, 2H, Ph), 6.98 (bm, 4H, Ph), 6.84 (t, 1H, p-Ph), 6.70 (t, 1H, p-Ph),
Electronic Structure Calculations. The Fenske-Hall approximate
molecular orbital method33 and Amsterdam Density Functional (ADF)
program34 were used for all calculations. For Fenske-Hall calcula-
tions: All atomic basis functions were generated by a least-squares fit
of Slater-type orbitals to the atomic orbitals from Herman-Skillman
atomic calculations.48 Contracted double-ú representations were used
for the Ta 5d, Fe 3d, C 2p, and O 2p atomic orbitals. An exponent of
1.16 was used for the hydrogen 1s AOs.49 The basis functions for Ta
were derived for the +2 oxidation state with fixed 6s and 6p exponents
of 2.0. The basis functions for Fe were derived from the +1 oxidation
state with fixed 4s and 4p exponents of 2.0 and 1.8, respectively. The
basis functions for C and O were derived from the 0 oxidation state.
The 3σ and 6σ orbitals of CO were deleted from the variational
orbitals.50 The ML3 fragments were given idealized L-M-L angles
of 90°. For ADF calculations: The geometries of the staggered and
eclipsed conformers of (TMM)Fe(CO)3 were optimized using a
procedure analogous to that used by Ziegler et al.,36c i.e., the geometries
were optimized under C3V symmetry at the local density approximation
(LDA) level.51 The rotation barrier was calculated with the incorpora-
(48) Bursten, B. E.; Jensen, J. R.; Fenske, R. F. J. Chem. Phys. 1978,
68, 3320.
(49) Hehre, W. J.; Stewart, R. F.; Pople, J. A. J. Chem. Phys. 1969, 51,
2657.
(50) Lichtenberger, D. L.; Fenske, R. F. J. Chem. Phys. 1969, 51, 4274.
(51) Vosko, S. H.; Wilk, L.; Nusair, M. Can. J. Phys. 1980, 58, 1200.

Tribenzylidenemethane Ligand in Tantalum Complexes
J. Am. Chem. Soc., Vol. 120, No. 48, 1998 12523
tion of Becke’s nonlocal exchange52 and Perdew’s nonlocal correlation53
corrections (BP). The geometry of the staggered and eclipsed structures
of (TMM)TaMe3 were fully optimized under C3V symmetry at the BP
level of calculation. The energies of the [Fe(CO)3]2+ and [TaMe3]2+
fragments at various values of θ were calculated at the BP level. A
double-ú basis set was chosen for C, O, and H, and a triple-ú basis
was used for Ta and Fe. The 1s2 configuration of C and O, the 1s2,
2s2, 2p6 for Fe, and the all atomic subshells for n ) 1 through n ) 4
and the 5s and 5p of Ta were assigned the core and treated by the
frozen core approximation.
Crystallography. Crystals of 4, 8 and 13 were mounted under a
flow of nitrogen onto glass fibers with epoxy. These were then placed
on the diffractometer (Enraf-Nonius-CAD4) in a cold nitrogen stream.
Data were collected at -60 °C using a low-temperature device.
An absorption correction using the program DIFABS54 was applied
to 4, 8, 9, 10, 11, and 13. The structures were determined by heavy-
atom Patterson methods (DIRDIF-Patty: teXsan).55 The structures
were refined using standard least squares routines. Non-hydrogen atoms
were refined with anisotropic thermal parameters. Hydrogen atoms
were included in idealized positions.
stream supplied by a Siemens LT-2A low-temperature device. The
X-ray intensity data were collected on a standard Siemens SMART-
CCD Area Detector System equipped with a normal focus molybdenum-
target X-ray tube operated at 1.5 kW (50 kV, 30 mA). A total of 1.3
hemispheres of data were collected using a narrow frame method with
scan widths of 0.3° in ω, and an exposure time of 10 s/frame. Frames
were integrated to 0.75 with the Siemens SAINT program. The space
groups were assigned on the basis of systematic absences and intensity
statistics by using the XPREP program.56
The structure of 7 was solved by direct methods using SHELXS-86
and refined by full-matrix least-squares on F2. Non-hydrogen atoms
were refined with anisotropic thermal parameters. Hydrogen atoms
were included in idealized positions.
Acknowledgment. G.C.B. is an Alfred Sloan Fellow and a
Henry and Camille Dreyfus Teacher Scholar. The authors are
grateful to these agencies and to the Petroleum Research
Foundation (ACS) for financial assistance. J.P.G. and B.E.B.
acknowledge the Ohio Supercomputer Center for use of their
facilities.
Crystals of 7, 9, 10, and 11 were mounted on glass fibers under
Paratone-8277 and placed on the X-ray diffractometer in a cold nitrogen
Supporting Information Available: Complete details for
the crystallographic studies of compounds 4, 7, 8, 9, 10, 11,
and 13 (72 pages, print/PDF). See any current masthead page
for ordering information and Web access instructions.
(52) Becke, A. D. Phys. ReV. 1988, A38, 2398.
(53) Perdew, J. P. Phys ReV. 1986, B33, 8822. Ibid. 1986, B34, 7406
(erratum).
(54) DIFABS: Walker, N.; Stuart. Acta Crystallogr. 1983, A39, 158-
166. An empirical absorption correction program.
(55) TEXSAN: Single-Crystal Structure Analysis Software. Molecular
Structure Corporation, The Woodlands, TX, 1993.
JA981640T
(56) SHELXTL: Structure Analysis Program. 5.04; Siemens Industrial
Automation, Inc.; Madison, WI, 1995.