Group 5 metallocene complexes as models for Metal-Mediated hydroboration: Synthesis of a reactive borane adduct, endo-Cp*2Nb(H2BO2C6H4), via hydroboration of coordinated olefins

Article abstract of DOI:10.1021/ja970639b

The olefin complexes Cp*₂M(CH₂=CH(R))(H) (R = H (1), CH₃ (2); M = Nb (a), Ta (b)) react cleanly with catecholborane (HBCat) and HBO₂C₆H₃-4-(t)Bu (HBCat') to give Cp*₂M(H₂

Full text of DOI:10.1021/ja970639b

J. Am. Chem. Soc. 1997, 119, 9699-9708
9699
Group 5 Metallocene Complexes as Models for Metal-Mediated
Hydroboration: Synthesis of a Reactive Borane Adduct,
endo-Cp*2Nb(H2BO2C6H4), via Hydroboration of Coordinated
Olefins
Dean R. Lantero, Donald L. Ward,† and Milton R. Smith, III*
Contribution from the Department of Chemistry, Michigan State UniVersity,
East Lansing, Michigan 48824
ReceiVed February 27, 1997X
Abstract: The olefin complexes Cp*2M(CH2dCH(R))(H) (R ) H (1), CH3 (2); M ) Nb (a), Ta (b)) react cleanly
with catecholborane (HBCat) and HBO2C6H3-4-tBu (HBCat′) to give Cp*2M(H2BCat) (M ) Nb (5a), Ta (5b)) and
Cp*2M(H2BCat′) (M ) Nb (6a), Ta (6b)) and the anti-Markovnikov hydroboration products CatBCH2CH2R and
Cat′BCH2CH2R. Compounds 2a and 2b react with DBCat′ (DBO2C6H3-4-tBu) to afford the deuterated analogs
Cp*2M(D2BCat′) (M ) Nb (6a), Ta (6b)) where the deuterium label is incorporated exclusively in the metal complex.
The hydride resonances in 6a exhibit large perturbations in chemical shift when deuterium is incorporated. On the
basis of this isotopic labeling experiment, a mechanism is proposed where HBCat reacts with the 16-electron alkyl
intermediates, Cp*2MCH2CH2R (R ) H (3), Me (4)), via σ-bond metathesis or oxidative-addition/reductive-elimination
sequences, to generate the alkylboranes and an intermediate hydride, Cp*2MH, that is trapped by additional borane
to give Cp*2M(H2BCat) (5a and 5b). The solid-state structures for Cp*2Nb(η2-H2BO2C6H3-3-tBu) (16) and Cp2*Nb-
(η2-BH4) (17) were determined by X-ray diffraction. The metal-boron distances in these two compounds are identical
within experimental error. While related group 5 catecholateboryl compounds have pronounced boryl character, the
structural parameters for the hydride and boryl ligands in 16 are consistent with formulation as either a borohydride
complex or a borane adduct of “Cp*2NbH”. In contrast to other group 5 boryl complexes, 6a reacts readily with
various two-electron ligands with elimination of HBCat′. For example, H2 reacts reversibly to form Cp*2NbH3 and
HBCat′, while “BH3” and CO react irreversibly to yield Cp*Nb(BH4) and Cp*2Nb(H)(CO) with elimination of HBCat′,
respectively. Ethylene and propylene react at 40 °C to regenerate 1a and 2a, with elimination of HBCat′. When
excess olefin is present, the liberated borane is converted to CatBCH2CH2R. Solutions of 1a and 2a catalyze olefin
hydroboration under mild conditions. Relationships between the reactivity of 1a and 2a and other early metal and
lanthanide catalysts are discussed.
Introduction
several experimental and theoretical studies. Pathways involv-
ing oxidative addition/reductive elimination sequences were
initially proposed based in part on precedent from hydrogena-
tions by Wilkinson’s catalyst6 and the fact that oxidative addition
of catecholborane to (PPh3)3RhCl gives (PPh3)3RhClHBCat.7
Thorough examinations indicate that the reaction manifold is
more complex.5g-h,l In attempts to resolve mechanistic issues,
The hydroboration reaction is one of the most versatile
methods for reduction of unsaturated carbon bonds. In contrast
to hydrogenation, hydroboration yields products where the
inherent functionality of the parent olefin is preserved and the
resulting boron-carbon bond can serve as a synthon for various
functional groups.1 Importantly, these transformations generally
proceed with retention of stereochemistry at carbon. This
feature has sparked interest in controlling selectivity of B-H
additions to olefins and alkynes. In this regard, metal-mediated
hydroborations are attractive since selectivities dictated by a
metal center may differ from those in uncatalyzed reactions.2
Also, catalyzed additions by chiral complexes potentially offer
enantioselectivity in the hydroboration of prochiral substrates.3
Most of the initial efforts in metal-mediated hydroboration
focused on catalysis by late transition metals.4,5 The mecha-
nism(s) that operate in these systems have been the subject of
(4) For metal-catalyzed additions of carboranes to olefins see: (a) Davan,
T.; Corcoran, E. W.; Sneddon, L. G. Organometallics 1983, 2, 1693-1694.
(b) Hewes, J. D.; Kreimendahl, C. W.; Marder, T. B.; Hawthorne, M. F. J.
Am. Chem. Soc. 1984, 106, 5757-5759. (c) Mirabelli, M. G. L.; Sneddon,
L. G. J. Am. Chem. Soc. 1988, 110, 449-453.
(5) For metal-catalyzed additions of catecholborane to olefins see: (a)
Manning, D.; No¨th, H. Angew. Chem., Int. Ed. Engl. 1985, 24, 878-879.
(b) Evans, D. A.; Fu, G. C.; Hoveyda, A. H. J. Am. Chem. Soc. 1988, 110,
6917-6918. (c) Burgess, K.; Ohlmeyer, M. J. Tetrahedron Lett. 1989, 30,
395-398, 5857-5860, 5861-5864. (d) Burgess, K.; van der Donk, W.
A.; Jarstfer, M. B.; Ohlmeyer, M. J. J. Am. Chem. Soc. 1991, 113, 6139-
6144. (e) Burgess, K.; Cassidy, J.; Ohlmeyer, M. J. J. Org. Chem. 1991,
56, 1020-1027. (f) Evans, D. A.; Fu, G. C.; Hoveyda, A. H. J. Am. Chem.
Soc. 1992, 114, 6671-6679. (g) Evans, D. A.; Fu, G. C.; Anderson, B. A.
J. Am. Chem. Soc. 1992, 114, 6679-6685. (h) Westcott, S. A.; Blom, H.
P.; Marder, T. B.; Baker, R. T. J. Am. Chem. Soc. 1992, 114, 8863-8869.
(i) Burgess, K.; van der Donk, W. A.; Westcott, S. A.; Marder, T. B.; Baker,
R. T.; Calabrese, J. C. J. Am. Chem. Soc. 1992, 114, 9350-9359. (j)
Westcott, S. A.; Marder, T. B.; Baker, R. T. Organometallics 1993, 12,
975-979. (k) Westcott, S. A.; Blom, H. P.; Marder, T. B.; Baker, R. T.;
Calabrese, J. C. Inorg. Chem. 1993, 32, 2175-2182.
† Author to whom crystallographic issues should be addressed.
X Abstract published in AdVance ACS Abstracts, August 15, 1997.
(1) Pelter, A.; Smart, K.; Brown, H. C. Borane Reagents; Academic
Press: London, 1988; p 131.
(2) For a review of transition metal catalyzed hydroboration see: Burgess,
K.; Ohlmeyer, M. J. Chem. ReV. 1991, 91, 1179-1191.
(3) Late transition-metal catalysts have been effective for asymmetric
hydroborations of olefins: (a) Burgess, K.; Ohlmeyer, M. J. J. Org. Chem.
1988, 53, 5178-5179. (b) Hayashi, T.; Matsumoto, Y.; Ito, Y. J. Am. Chem.
Soc. 1989, 111, 3426-3428. (c) Brown, J. M.; Hulmes, D. I.; Layzell, T.
P. J. Chem. Soc., Chem. Commun. 1993, 1673-1674.
(6) Collman, J. P.; Hegedus, L. S.; Norton, J. R.; Finke, R. G. Principles
and Applications of Organotransition Metal Chemistry; University Science
Books: Mill Valley, 1987; Chapter 10.
(7) Kono, H.; Ito, K.; Nagai, Y. Chem. Lett. 1975, 1095-1096.
S0002-7863(97)00639-2 CCC: $14.00 © 1997 American Chemical Society

9700 J. Am. Chem. Soc., Vol. 119, No. 41, 1997
Lantero et al.
potential pathways have been evaluated by computational
methods. Mechanisms involving oxidative addition/reductive
elimination sequences seem to be preferred; however, a clear
consensus has not been reached and mechanisms where σ-bond
metatheses operate in lieu of oxidative addition/reductive
elimination are not prohibitively disfavored.8 While labeling
experiments are consistent with oxidative addition/reductive
elimination mechanisms,5g mechanisms involving σ-bond met-
atheses cannot be categorically excluded given precedent in late
metal systems.9
the coordinated olefin into the metal-hydride bond to form alkyl
intermediates is well-documented. Hence, investigating the
reaction chemistry of 1 and 2 with borane reagents offers an
interesting comparison between relative reactivities of the
olefin-hydride resting states and the alkyl intermediates. Given
well-behaved reactivity, we felt that these systems would be
ideally suited for mechanistic investigations; this paper describes
the results.
Experimental Section
By comparison, forays in early transition-metal and lanthanide
systems have been fewer.10,11 When catecholborane (HBCat)
is employed as the borane reagent, complications can arise from
incompatibility between the catalyst and borane reagent, par-
ticularly when the metal-center and borane are supported by
labile alkoxide ligands.11b,c These problems are not necessarily
general as recent reports indicate that group 4 metallocene
complexes catalyze addition of HBCat to olefins where key
hydroboration steps likely occur at the metal-center.11a,12
Metallocene systems are attractive as cyclopentadienyl ligands
and are relatively inert, and thus provide a well-defined
coordination sphere for the metal.
We have been examining the reactivity of B-H bonds with
metallocene olefin complexes.12 Apart from the reasons outlined
above, the chemistry of these species with borane reagents offers
interesting possibilities since the metal-carbon bonds exhibit
reactivity that is consistent with pronounced metallacyclopro-
pane character in the metal-olefin moiety. Thus, ring-opening
σ-bond metathesis between B-H and M-C bonds offers a
pathway for B-C bond formation that is mechanistically distinct
from those postulated in late-metal and lanthanide systems.
Borylation by these routes can lead to unusual selectivities, as
illustrated by the fact that Cp*2Ti(η2-CH2dCH2) reacts with
catecholborane, and related borane reagents, via a dehydroge-
native pathway where borylation proceeds without bond reduc-
tion.
General Considerations. All manipulations were performed using
glovebox, Schlenk, and vacuum-line techniques. Pentane, heptane,
THF, and toluene were predried over sodium. Toluene was distilled
from NaK/benzophenone ketyl. All other solvents were distilled from
sodium/benzophenone ketyl. Benzene-d6 and toluene-d8 were dried
over activated 3 Å molecular sieves and then vacuum transferred into
a sodium mirrored container. Argon and nitrogen gases were purified
by passage over MnO on silica.
Catecholborane (Aldrich) was vacuum transferred prior to use.
Preparation of the other boranes was accomplished by reacting NaBH4
(or NaBD4) (Aldrich) and BF3‚OEt2 (Mallinckrodt) to produce BH3
(or BD3), which was then bubbled through THF solutions of the
appropriate diol (4-tert-butylcatechol (Aldrich) or 3-tert-butylcatechol).14
The THF was then distilled from the solution at atmospheric pressure,
and the remaining solution was vacuum distilled to give the borane as
a clear, viscous liquid. Styrene and 1-hexene (Aldrich) were dried over
activated 3 Å molecular sieves and vacuum transferred prior to use.
Phenethyl alcohol and sec-phenethyl alcohol (both Aldrich) were used
as received. cis-Cyclooctene (Aldrich) was distilled prior to use.
Ethylene (Matheson) and propylene (Matheson) were subjected to
several freeze-pump-thaw cycles and then transferred in vacuo to
reaction vessels. Carbon monoxide (Matheson) and dihydrogen (AGA
Gas Inc.) were used as received. Cp*2NbH(C3H6),13a Cp2NbBH4,15
Cp*2TaH(C3H6),16 and Cp*2Nb(BH4)17 were prepared by literature
methods.
1H NMR spectra were recorded on Varian Gemini-300 (300.0 MHz),
Varian VXR-300 (299.9 MHz), and Varian VXR-500 (499.9 MHz)
spectrometers and referenced to residual proton solvent signals. 2H
NMR spectra were recorded on a Varian VXR-300 spectrometer
operating at 46.0 MHz, and spectra were referenced to residual
deuterium signals of the proteo solvent. 11B NMR spectra were
recorded on a Varian VXR-300 spectrometer operating at 96.23 MHz
and spectra were referenced to a BF3‚Et2O external standard solution.
Infrared spectra were recorded as Nujol mulls in KBr plates with a
Nicolet IR/42 spectrometer. GC samples were run on a Hewlett Packard
5890A gas chromatograph equipped with a Hewlett Packard 3392
integrator and a Foxboro GB-1 capillary column (25 m × 0.25 mm
i.d. with a 0.25 µm film). Elemental analyses were performed by Desert
Analytics, Tucson, AZ.
This reactivity prompted us to examine the chemistry of the
olefin-hydride complexes Cp*2M(CH2dCH(R))(H) (R ) H (1),
CH3 (2); M ) Nb (a), Ta (b)).13 In these systems, insertion of
(8) For BH3 additions mediated by 14-electron intermediates, computa-
tions favor B-H oxidative addition, olefin insertion into the Rh-H bond,
followed by B-C reductive elimination of the alkylborane product.8a For
HB(OH)2 additions effected by 16-electron intermediates, computations
favor B-H oxidative addition, olefin insertion into the Rh-B bond, and
C-H reductive elimination to generate alkylboronate ester.8b,c In the latter
study, the barrier for a pathway involving σ-bond metathesis was only 1.5
kcal/mol higher than the oxidative addition pathway. While addition of HB-
(OH)2 is more relevant to catalyzed addition of HBCat, pathways involving
14-electron intermediates have not been considered.8c (a) Dorigo, A. E.;
Schleyer, P. v. R. Angew. Chem., Int. Ed. Engl. 1995, 34, 115-118. (b)
Musaev, D. G.; Mebal, A. M.; Morokuma, K. J. Am. Chem. Soc. 1994,
116, 10693-10792. (c) Musaev, D. G.; Matsubara, T.; Mebal, A. M.; Koga,
N.; Morokuma, K. Pure Appl. Chem. 1995, 67, 257-263.
Procedures: endo-Cp*2Nb(H2BO2C6H4) (5a). A 15-mL toluene
solution of Cp*2NbH(C3H6) (219 mg, 0.540 mmol) was cooled to -78
°C. Catecholborane (0.161 mL, 1.35 mmol) was added to this stirred
solution via syringe. The resulting mixture was warmed to room
temperature and stirred for 2 h. During the course of the reaction, the
color of the solution gradually lightened from deep red to orange-red,
and solvent evaporation gave a deep orange solid. The reaction mixture
was dissolved in a minimal volume of toluene and cooled to -80 °C.
The product was isolated as orange microcrystals (207 mg, 79%): mp
(9) (a) Hartwig, J. F.; Bhandari, S.; Rablen, P. R. J. Am. Chem. Soc.
1994, 116, 1839-1844. (b) Baker, R. T.; Calabrese, J. C.; Westcott, S. A.;
Marder, T. B. J. Am. Chem. Soc. 1995, 117, 8777-8784.
(10) Harrison, K. M.; Marks., T. J. J. Am. Chem. Soc. 1992, 114, 9220-
9221.
1
171-3 °C; H NMR (C6D6) δ 7.07 (m, 2H, O2C6H4), 6.85 (m, 2H,
(11) (a) Erker, G.; Noe, R.; Wingbermu¨hle, D.; Petersen, J. L. Angew.
Chem., Int. Ed. Engl. 1993, 32, 1213-1215. (b) Burgess, K.; Jaspars, M.
Tetrahedron Lett. 1993, 34, 6813-6816. (c) Burgess, K.; van der Donk,
W. A. Organometallics 1994, 13, 3616-3620. (d) Bijpost, E. A.; Duchateau,
R.; Teuben, J. H. J. Mol. Catal. A: Chem. 1995, 95, 121-128. (e) Pereira,
S.; Srebnik, M. Organometallics 1995, 14, 3127-3128. (f) Pereira, S.;
Srebnik, M. J. Am. Chem. Soc. 1996, 118, 909-910. (g) He, X.; Hartwig,
J. F. J. Am. Chem. Soc. 1996, 118, 1696-1702.
(12) (a) Motry, D. H.; Smith, M. R., III J. Am. Chem. Soc. 1995, 117,
6615-6616. (b) Motry, D. H.; Brazil, A. G.; Smith, M. R., III J. Am. Chem.
Soc. 1997, 119, 2743-2744.
(13) (a) Doherty, N. M.; Bercaw, J. E. J. Am. Chem. Soc. 1985, 107,
2670-2682. (b) Burger, B. J.; Santarsiero, B. D.; Trimmer, M. S.; Bercaw,
J. E. J. Am. Chem. Soc. 1988, 110, 3134-3146.
O2C6H4), 1.83 (s, 30H C5(CH3)5), -8.27 (br, 2H, NbH2); 11B NMR
(C6D6) δ 60.2 (∆ν1/2 ) 240 Hz); IR (cm-1) 1653 (νNbH). Anal. Calcd
for C26H36BNbO2: C, 64.5; H, 7.44. Found: C, 64.1; H, 7.26.
(14) Hocking, M. B. Can. J. Chem. 1973, 51, 2384-2392.
(15) Lucas, C. R.; Green, M. L. H. J. Chem. Soc., Chem. Commun. 1972,
1005.
(16) Gibson, V. C.; Bercaw, J. E.; Bruton, W. J.; Sanner, R. D.
Organometallics 1986, 5, 976-979.
(17) Bell, R. A.; Cohen, S. A.; Doherty, N. M.; Threlkel, R. S.; Bercaw,
J. E. Organometallics 1986, 5, 972-975.
(18) Brown, H. C.; Gupta, S. K. J. Am. Chem. Soc. 1975, 97, 5249-
5255.

Models for Metal-Mediated Hydroboration
J. Am. Chem. Soc., Vol. 119, No. 41, 1997 9701
endo-Cp*2Nb(H2BO2C6H3-4-tBu) (6a). A 20-mL toluene solution
of Cp*2NbH(C3H6) (400 mg, 0.985 mmol) was cooled to -78 °C.
HBO2C6H3-4-tBu (0.360 mL, 2.07 mmol) was then added via syringe
to the stirred solution, and the mixture was warmed to room temper-
ature. After 2 h, the toluene was removed in vacuo, leaving a viscous
red oil. The oil was dissolved in a minimal volume of heptane and
placed in a -80 °C freezer. Upon standing overnight, deep red
microcrystals were deposited. After the filtrate was removed, the
crystals were washed with cold pentane and dried in vacuo (396 mg,
Phenethyl Alcohol Formation. Cp*2NbH(C3H6) (15.3 mg, 0.038
mmol) was dissolved in ∼0.6 mL of benzene-d6 along with styrene
(63.1 mg, 0.607 mmol) and HBO2C6H3-4-tBu (106 mg, 0.602 mmol).
The solution was placed in an NMR tube and flame sealed. The tube
was then maintained at 75 °C for 4 days in a constant temperature oil
bath. Conversion of free borane to the alkyl borane product was
monitored by 1H NMR. Integration revealed 64% hydroborated
products along with 11% of a boron decomposition product5k,11b and
25% unreacted borane. The contents were then converted to the
corresponding alcohol by a literature method.18 The products of the
oxidative workup were analyzed by gas chromatography, with phenethyl
alcohol and sec-phenethyl alcohol detected in a 3.72:1 ratio.
NMR Tube Reactions: Cp*2NbH(C3H6) + cis-cyclooctene +
HBO2C6H3-4-tBu. Cp*2NbH(C3H6) (11.8 mg, 0.029 mmol) was
dissolved in ∼0.6 mL of benzene-d6 along with cis-cyclooctene (3.2
mg, 0.029 mmol) and HBO2C6H3-4-tBu (4.4 mg, 0.025 mmol). The
resulting solution was transferred to a NMR tube that was flame sealed.
The tube was kept at room temperature for 2.5 h. A 1H NMR spectrum
showed primary hydroboration occurred with the bound propylene
fragment on the niobium complex with some hydroboration of the cis-
cyclooctene. Integration of the appropriate tBu resonances established
the ratio of cis-cyclooctene:propylene hydroboration as 1:5.08. The
tube was held at room temperature overnight, and integration of another
1H NMR spectrum determined the ratio of cis-cyclooctene:propylene
hydroboration to be 1:3.81.
Cp*2NbH(C3H6) + cis-cyclooctene + propylene + HBO2C6H3-
4-tBu. Cp*2NbH(C3H6) (9.8 mg, 0.024 mmol) was dissolved in ∼0.6
mL of benzene-d6. HBO2C6H3-4-tBu (51.0 mg, 0.290 mmol) and cis-
cyclooctene (16.0 mg, 0.145 mmol) were added to the solution which
was then placed in a NMR tube. The tube was then frozen with liquid
nitrogen and evacuated. Propylene (0.145 mmol) was then admitted
with the aid of a calibrated volume vessel. The NMR tube was then
flame sealed, and the contents of the tube were heated at 60 °C for 3
days. The rate of hydroboration for propylene was faster than that of
cis-cyclooctene based on integrated 1H spectra (propylene:cis-cy-
clooctene ) 4.53:1). Repeating the reaction without the use of the
niobium complex required heating the reaction at 120 °C for 3 days
and resulted in nearly identical rates of hydroboration; however, the
rate for cis-cyclooctene was slightly faster (propylene:cis-cyclooctene
) 0.93:1).
Cp*2Nb(H2BO2C6H4) + CO. A 10-mg (0.021 mmol) sample was
dissolved in ∼0.8 mL of benzene-d6. This solution was placed in an
NMR tube and frozen with liquid nitrogen. The NMR tube was then
evacuated and 1 atm of CO gas admitted. The tube was then flame
sealed and allowed to warm to room temperature. 1H and 11B NMR
spectra indicated elimination of HBCat, and Cp*2NbH(CO) was the
only Nb-containing product detected. Integration of the 11B spectrum
indicated 91% catecholborane (δ 28.7, d, JHB ) 184 Hz) and 9% B2-
Cat3 (δ 22.7, s).
5a + B2H6. A 15-mg (0.031 mmol) sample of the niobium complex
was dissolved in ∼0.6 mL of toluene-d8 and placed in a NMR tube.
The solution was frozen with liquid nitrogen and then evacuated.
Diborane generation was performed in the following way: A Schlenk
tube was charged with NaBH4 (130 mg, 3.43 mmol) and DME (0.875
mL). This slurry was cooled in a liquid nitrogen bath, and BF3‚Et2O
(0.580 mL, 3.52 mmol) was added under a counter flow of N2 gas.
The Schlenk tube was fitted with an adapter, and the assembly was
attached to a high vacuum manifold. After the reaction vessel had
been evacuated, the suspension was allowed to slowly warm to room
temperature with generation of B2H6. From a calibrated bulb, ap-
proximately 0.150 mmol of diborane was condensed into a NMR tube.
Once the addition of diborane was completed, the NMR tube was flame
sealed. 1H and 11B NMR spectra were initially taken at -5 °C, and
then the temperature was gradually raised to 20 °C over 3 h. The
reaction cleanly formed Cp*2NbBH4 and catecholborane, with 11B
spectra indicating 92% catecholborane (δ 28.4, d, JHB ) 193 Hz) and
8% boron decomposition product (δ 22.3, s).
1
75%): mp 154-7 °C; H NMR (C6D6) δ 7.30 (d, J ) 2.1 Hz, 1H,
O2C6H3), 7.04 (d, J ) 8.1 Hz, 1H, O2C6H3), 6.90 (dd, J ) 7.8, 1.95
Hz, 1H, O2C6H3), 1.85 (s, 30H C5(CH3)5), 1.30 (s, 9H O2C6H3(4-
C(CH3)3), -8.19 (br, 2H, NbH2); 11B NMR (C6D6) δ 60.0 (∆ν1/2
)
300 Hz); IR (cm-1) 1653 (νNbH). endo-Cp*2Nb(D2BO2C6H3-4-tBu)
was similarly prepared from Cp*2NbH(C3H6) and DBO2C6H3-4-tBu:
1H NMR (C6D6) δ 7.28 (d, J ) 1.8 Hz, 1H, O2C6H3), δ 7.03 (d, J )
8.1 Hz, 1H, O2C6H3), δ 6.89 (dd, J ) 8.1, 1.95 Hz, 1H, O2C6H3), δ
2
1.83 (s, 30H C5(CH3)5), δ 1.30 (s, 9H O2C6H3(4-C(CH3)3); H NMR
(C6H6) δ -9.29 (br, NbD2); 11B NMR (C6D6) δ 57.3 (∆ν1/2 ) 266
Hz).
endo-Cp*2Nb(H2BO2C6H3-3-tBu) (16). A 10-mL toluene solution
of Cp*2NbH(C3H6) (251 mg, 0.618 mmol) was cooled to -78 °C.
HBO2C6H3-3-tBu (229 mg, 1.30 mmol) was dissolved in ∼5 mL of
toluene, and the borane solution was transferred to the propylene-
hydride complex via cannula. Once the addition was complete, the
resulting solution was allowed to warm to room temperature and stirred
for 3 h. The toluene was then evaporated, leaving a tacky, deep orange-
red solid. The crude mixture was dissolved in a minimal volume of
pentane and placed in a -80 °C freezer. Orange-red microcrystals
were deposited overnight. The product was collected, washed with
1
cold pentane, and dried in vacuo (218 mg, 65%): mp 141-2 °C; H
NMR (C6D6) δ 7.04 (m, 1H, O2C6H3), 6.96 (m, 1H, O2C6H3), 6.91 (m,
1H, O2C6H3), 1.86 (s, 30H C5(CH3)5), 1.58 (s, 9H O2C6H3(4-C(CH3)3),
-7.91 (br, 2H, NbH2); 11B NMR (C6D6) δ 60.9 (∆ν1/2 ) 280 Hz); IR
(cm-1) 1649 (νNbH).
endo-Cp*2TaH2(BO2C6H4) (5b). Catecholborane (0.120 mL, 0.973
mmol) was added to a 10-mL toluene solution of Cp*2TaH(C3H6) (193
mg, 0.389 mmol) at room temperature. The solution was then stirred
for 2 h during which time the initial pale yellow color gradually turned
a lime green. Evaporation of the toluene yielded a lime green solid.
Washing this solid with pentane followed by drying under vacuum
1
yielded a pale yellow solid (160 mg, 72%): mp 159-62 °C dec; H
NMR (C6D6) δ 7.19 (m, 2H, O2C6H4), 6.89 (m, 2H, O2C6H4), 1.96 (s,
30H C5(CH3)5), -2.12 (br, 2H, TaH2); 11B NMR (C6D6) δ 73.5 (∆ν1/2
) 390 Hz); IR (cm-1) 1774, 1709 (νTaH).
endo-Cp*2TaH2(BO2C6H3-4-tBu) (6b). Toluene (∼5 mL) was used
to dissolve Cp*2TaH(C3H6) (155 mg, 0.313 mmol). HBO2C6H3-4-tBu
(0.140 mL, 0.784 mmol) was then added via syringe to the stirred
solution. The resulting solution was stirred for 2 h at room temperature,
giving a rust orange solution. Evaporation of the toluene yielded a
viscous, deep orange oil. Dissolving this oil in pentane and concentrat-
ing the resulting orange solution yielded pale yellow microcrystals when
placed in a -80 °C freezer overnight. Separating the filtrate and drying
the microcrystals under vacuum gave the pure product (112 mg,
1
57%): mp 156-8 °C; H NMR (C6D6) δ 7.42 (d, J ) 2.1 Hz, 1H,
O2C6H3), δ 7.19 (d, J ) 11.4 Hz, 1H, O2C6H3), δ 6.95 (dd, J ) 8.4,
1.95 Hz, 1H, O2C6H3), δ 1.98 (s, 30H C5(CH3)5), δ 1.29 (s, 9H O2C6H3-
(4-C(CH3)3), δ -2.05 (s, 2H, TaH2); 11B NMR (C6D6) δ 72.7 (∆ν1/2
)
610 Hz); IR (cm-1) 1767, 1720 (νTaH). Anal. Calcd for C30H44BO2-
Ta: C, 57.4; H, 7.01. Found: C, 57.6; H, 7.29. endo-Cp*2TaD2-
(BO2C6H3-4-tBu) was similarly prepared from Cp*2TaH(C3H6) and
DBO2C6H3-4-tBu: 1H NMR (C6D6) δ 7.41 (d, J ) 1.8 Hz, 1H, O2C6H3),
7.18 (d, J ) 8.1 Hz, 1H, O2C6H3), 6.95 (dd, J ) 8.1, 1.8 Hz, 1H,
O2C6H3), 1.97 (s, 30H, C5(CH3)5), 1.29 (s, 9H, O2C6H3(4-C(CH3)3));
2H NMR (C6H6) δ -2.02 (br, TaD2); 11B NMR (C6D6) δ 72.3 (∆ν1/2
) 660 Hz). 6b-d0-2 was prepared as a mixture of isotopomers by
adding equal volumes of HBO2C6H3-4-tBu and DBO2C6H3-4-tBu to a
toluene solution of Cp*2TaH(C3H6). The workup was analogous to
that described above: 1H NMR (C6D6) δ 7.41 (m 1H, O2C6H3), 7.17
(m, 1H, O2C6H3), 6.82 (m, 1H, O2C6H3), 1.98 (s, 30H C5(CH3)5), 1.29
(s, 9H O2C6H3(4-C(CH3)3)), -2.06 (s, 1H, TaHD); 2H NMR (C6H6) δ
-2.02 (br, TaHD); 11B NMR (C6D6) δ 70.0 (∆ν1/2 ) 650 Hz).
6a + CO. A toluene-d8 (0.75 mL) solution of 6a (8.0 mg, 0.015
mmol) was placed in a NMR tube. The contents of the NMR tube
were frozen with liquid nitrogen, and then the tube was evacuated.
Approximately 1 atm of CO was admitted to the frozen contents of
the tube, after which it was flame sealed. The reaction was followed

9702 J. Am. Chem. Soc., Vol. 119, No. 41, 1997
Lantero et al.
1
at 0 °C. Evaluation of H spectral intensities resulted in a calculated
half-life of 80 min.
difference map peaks. SHELXL-93 was used to refine the structure.
The hydrogen atoms of each Cp* methyl group were calculated as two
sets of half-occupancy positions, rotated initially about the C-C bond
by 60° from each other, and refined with occupancy factors of x and
1 - x and with a common isotropic displacement parameter. Later, a
single set of hydrogen atoms for each Cp* methyl group was placed at
the occupancy-weighted average H positions. Restraints were applied
separately to each methyl group to make all C-H distances the same,
all H-H distances the same, and all C-C-H angles the same and to
use a common isotropic displacement parameter. The hydrogen atoms
of the boryl ring and its tert-butyl group were placed in calculated
positions and refined independently without restraints. The toluene
solvent molecule is disordered about an inversion center with both half-
occupancy molecules coplanar and some atoms (e.g., C32 and C36*,
etc.) nearly overlapping each other. The ring protons were restrained
to lie 0.95 Å from their carbon atoms and equidistant from the two
adjacent carbon atoms. The seven carbon atoms and the five ring
hydrogen atoms of the toluene molecule were restrained to lie in a
plane. The methyl hydrogen atoms were treated as the Cp* methyl
hydrogens above. Full-matrix least-squares refinement on F2 (8548
data, 181, restraints, 462 parameters) converged to give the following
values: R1 ) 0.0502, wR2 ) 0.1220, GOF ) 1.067 for [I > 2σ(I)].
6b + CO. A 9.0-mg (0.014 mmol) sample of tantalum complex
was dissolved in 0.75 mL of toluene-d8 and placed in a NMR tube.
The contents were then frozen with liquid nitrogen, and the tube was
evacuated. CO (approximately 1 atm) was admitted, after which the
tube was flame sealed. The reaction was monitored at 70 °C and
evaluation of 1H spectral intensities resulted in a calculated half-life of
130 min. Integration of a final 11B spectrum showed 73% free borane
(δ 29, d, JHB ) 190 Hz) and 27% B2Cat′3 (δ 23, s). A final 1H spectrum
indicated the formation of Cp*2TaH(CO)17 as the major Cp* product.
6a + C2H4. A 15-mg (0.028 mmol) sample of the niobium complex
was dissolved in ∼0.6 mL of toluene-d8 and placed in a NMR tube.
This tube was then frozen with liquid nitrogen and evacuated. Ethylene
(∼1 atm) was then admitted to the frozen contents of the NMR tube,
which was then flame sealed. 1H and 11B NMR spectra were initially
taken at 10 °C and increased to 40 °C over the course of the 3 h reaction
time. The products of the reaction were Cp*2NbH(C2H4) and free
borane. Integration of the final 11B spectrum showed 18% EtBO2C6H3-
4-tBu (δ 36.1, s), 74% of free borane (δ 28.8, d, JHB ) 190 Hz), and
8% of a boron decomposition product (δ 22.8, s).
6a + C3H6. A 15-mg (0.028 mmol) sample of niobium compound
was dissolved in ∼0.6 mL of toluene-d8. This solution was placed in
a NMR tube and frozen with liquid nitrogen, followed by evacuation
of the gases. Approximately 1 atm of propylene was then admitted to
the frozen contents of the NMR tube, after which it was flame sealed.
Initial 1H and 11B NMR spectra (taken at 0 °C) showed the formation
of Cp*2NbH(C3H6) and free borane. However, as the reaction
proceeded the concentration of the free borane diminished and the
concentration of n-propylborane increased. After the 3 h reaction time,
during which the temperature had been gradually raised to 40 °C and
then lowered to 25 °C, a 11B spectrum indicated clean conversion to
n-propylborane (93%, δ 35.6, s) and a boron-containing decomposition
product (7%, δ 22.2, s). A final 1H spectrum revealed the presence of
Cp*2NbH(C3H6).
6a + H2. A benzene-d6 (∼0.6 mL) solution of 6a (15.0 mg, 0.028
mmol) was placed in a NMR tube. This solution was frozen with liquid
nitrogen. Addition of ∼1 atm of H2 gas was preceeded by evacuation
of the frozen NMR tube. 1H and 11B NMR spectra, initially taken at
20 °C and gradually raised to 50 °C over the 4 h reaction time, indicated
formation of Cp*2NbH3 and free borane. Integration of the final 11B
NMR spectrum showed 80% of free borane (δ 28.5, d, JHB ) 185 Hz)
and 20% of the boron decomposition product, B2Cat3 (δ 22.2, s).
Crystal Structure Determination and Refinement for 16 and 17.
Data were collected at 173 K on a Rigaku AFC6S diffractometer using
Mo KR radiation (λ ) 0.71069 Å). Final unit cell parameters were
obtained by least-squares refinement of 25 reflections that had been
accurately centered. Three standard reflections were recorded after
every 150 scans, with no significant change in intensity noted. The
structures were solved by using SHELXL-86 in the teXsan19 software
package. Atomic coordinates and thermal parameters were refined F2
data, using the full-matrix least-squares program, SHELXL-93.20 All
non-hydrogen atoms were refined by using anisotropic thermal
parameters. All crystallographic computations were performed on
Silicon Graphics Indigo computers.
Orange-red crystals of Cp*2Nb(H2BO2C6H3-3-tBu) (fw 539.7) were
grown from a toluene/heptane solution cooled to -35 °C. The crystals
were coated with Paratone-N, and a suitable crystal was selected and
mounted on a glass fiber. Compound 16 crystallized in a triclinic crystal
system with systematic absences indicating the space group P1h (No.
2) with the following cell parameters: a ) 9.923(3) Å, b ) 12.109(4)
Å, c ) 13.588(4) Å, R ) 97.74(2)°, â ) 100.86(2)°, γ ) 107.09(2)°,
V ) 1500.9(8) Å3, Z ) 2. Density(calculated) ) 1.298 mg/cm3. The
8555 independent reflections were collected between 4.0 e 2θ e 60.0°,
using ω/2θ scans.
2
For all data R1 ) 0.0730 and wR2 ) 0.1402. R1 ) ∑(Fo - Fc2)/
2
2
2
2
∑Fo , wR2 ) [(w(Fo - Fc2)2/∑w(Fo )2)]1/2, and GOF ) (∑(|Fo | -
|Fc2|)/σ)/(n - m), where w ) 1/σ2(Fo ), n is the number of reflections
2
used, and m is the number of variables.
Green crystals of Cp*2NbBH4 (fw 377.7) were grown from a
concentrated solution of heptane cooled to -35 °C. The crystals were
coated with Paratone-N, and a suitable crystal was selected and mounted
on a glass fiber. Compound 17 crystallized in a monoclinic crystal
system with systematic absences indicating the space group P21/c (No.
14) with the following cell parameters: a ) 8.482(2) Å, b ) 25.306(6)
Å, c ) 9.924 (3) Å, â ) 113.84(2)°, V ) 1948.39(9) Å3, Z ) 4.
Density(calculated) ) 1.298 mg/cm3. The 3423 independent reflections
were collected between 4.0 e 2θ e 50.0°, using ω scans. All non-
hydrogen atoms were located in the solution found by SHELXS-86.
Methyl hydrogens for the Cp* ligand were placed in calculated positions
(0.95 Å), and their isotropic thermal parameters were allowed to ride
on the parent carbon atoms. The hydrides were located in the
difference-fourier map and were refined isotropically. Full-matrix least-
squares refinement on F2 (3402 data, 30 restraints, 335 parameters)
converged to give the following values: R1 ) 0.0455, wR2 ) 0.1110,
GOF )1.139 for [I > 2σ(I)]. For all data R1 ) 0.1125 and wR2 )
2
2
0.1566. R1 ) ∑(Fo2 - Fc2)/∑Fo , wR2 ) [(w(Fo2 - Fc2)2/∑w(Fo )2)]1/2
,
and GOF ) (∑(|Fo | - |Fc2|)/σ)/(n - m), where w ) 1/σ2(Fo ), n is
the number of reflections used, and m is the number of variables.
2
2
Results and Discussion
Bercaw and co-workers have shown that olefin-hydride
complexes in the group 5 bent-metallocene family can serve as
models for olefin insertion into metal-hydride bonds. In these
systems equilibrium between 16-electron intermediates, Cp*2-
MCH2CH2R (R ) H (3), CH3 (4)), and the ground state olefin
hydride structures, 1 and 2, has been established by dynamic
NMR experiments.13,21 The fact that 1 and 2 form 18-electron,
alkylcarbonyl complexes, Cp*2M(CO)(CH2CH2R), in trapping
reactions with CO illustrates the important role of the 16-electron
alkyl intermediates to the reactivity of the olefin-hydride
complexes. Although 1 and 2 generally react via these
coordinatively unsaturated intermediates, the formation of a
stable adduct between Cp*2Ta(CH2dCH2)(H) and trimethyla-
luminum shows that the olefin ligand in the ground state
complexes can interact with Lewis acids.22 Hence, borane
interactions with both the ground state olefin complexes (1, 2)
and the alkyl intermediates (3, 4) are not unreasonable.
All non-hydrogen atoms except one Cp* methyl carbon and the
toluene solvent molecule were found by using SHELXS-86. DIRDIF
WFOUR was used to extend and initially “refine” the structure. The
toluene solvent molecule was found and modeled later to fit the
(21) (a) Green, M. L. H.; Sella, A.; Wong, L.-L. Organometallics 1992,
11, 2650-2659. (b) Green, M. L. H.; Wong, L.-L.; Sella, A. Organome-
tallics 1992, 11, 2660-2668.
(22) McDade, C.; Gibson, V. C.; Santersiero, B. D.; Bercaw, J. E.
Organometallics 1988, 7, 1-7.
(19) TEXRAY Structure Analysis Package, SGI Version 1.8, Molecular
Structure Corp., The Woodlands, TX, 1996.
(20) Sheldrick, G. M. University of Go¨ttingen, Germany, 1993.

Models for Metal-Mediated Hydroboration
J. Am. Chem. Soc., Vol. 119, No. 41, 1997 9703
Scheme 1
Compounds 1 and 2 reacted smoothly with 2 equiv of
catecholborane (HBCat) to produce endo-Cp*2M(H2BCat) (M
) Nb (5a), Ta (5b)) and the alkylborane products EtBCat and
nPrBCat (eq 1).23 The analogs endo-Cp*2M(H2BCat′) (M )
HBCat′ could generate 8 directly, circumventing intermediate
7. Pathway B invokes ring-opening of the metallacyclopropane
similar to that proposed in reactions of Ti olefin complexes.11g,12
Reductive elimination of nPrBCat from the dihydride complex
(9) generates the common hydride intermediate. In Pathway
C, metathesis between metal-hydride and B-H bonds generates
a boryl-olefin intermediate 10 with loss of H2.24 Trapping of
the product of olefin insertion into the metal-boron bond 11
with H2 gives a dihydride intermediate 12 that can reductively
eliminate alkylborane to yield 8.
Cp*2M(CH2dCH(R))(H) + 2HBCat f
Cp*2M(H2BCat) + RCH2CH2BCat (1)
R ) H, CH3; M ) Nb, Ta
Nb (6a), Ta (6b)) were similarly prepared from HBO2C6H3-4-
tBu (HBCat′), in which the tBu substituent provides a convenient
NMR handle for monitoring catecholate species in solution.
Since the propylene compounds reacted more readily than their
ethylene counterparts, compounds 5 and 6 were most conve-
niently prepared from 2a and 2b. When the reactions were
monitored by NMR, intermediates were not observed, and
addition of 1 equiv of HBCat′ resulted in a 50% conversion of
2a and 2b to compounds 6a and 6b, respectively. The 11B
chemical shifts for the H2BCat moiety are similar to those
observed in the related cyclopentadienyl counterparts, and high-
These mechanisms all assume that the metal center mediates
B-C bond formation. This assumption can be dangerous since
a common problem in early metal systems is the tendency for
the metal to effect disproportionation of HBCat to BH3 and
B2Cat3.5k,11b Thus, hydroboration by transient BH3 can mas-
querade as hydroboration where B-C bond formation occurs
at the metal center. Although traces of B2Cat3 were observed,
hydroboration by BH3 did not seem likely in these systems
because BnPr3 was not observed in the reaction mixtures.
Nevertheless, participation of BH3 can be definitively addressed
by labeling experiments (vide infra).
1
field H NMR resonances that are typical for group 5 decam-
Efforts to apply a rigorous kinetic analysis were complicated
by irreproducibility in rate constants measured under pseudo-
first-order conditions of high borane concentration. Despite
these inconsistencies, two significant variations in rate were
observed. First, the Nb complexes reacted more rapidly than
their Ta analogs. Second, observed rates for the propylene
complexes (2a,b) were faster than those for their ethylene
counterparts (1a,b). Based on the fact that propylene insertion
into the M-H is faster than insertion of ethylene,13 it is tempting
to argue that this qualitative disparity in rates is consistent with
the proposed mechanism. However, this is misleading and can
ethylmetallocene hydrides are observed.
Some potential mechanisms that could account for the
observed products from the reaction of the propylene complexes
(2) with HBCat′ are shown in Scheme 1. In Pathway A, the
reaction proceeds via the aforementioned alkyl hydride inter-
mediates. After the olefin insertion step, catecholborane oxi-
datively adds to 4 to generate a boryl intermediate, 7. Reductive
elimination of the organic borane from this putative species
generates monohydride 8. Intermediate 8 is common to all three
mechanisms, and compound 6 forms when 8 is trapped with
HBCat′. Alternatively, σ-bond metathesis between 4 and
(23) This approach has recently been reported as a route to preparing
bis(silyl)niobocene complexes: Nikonov, G. I.; Kuzmina, L. G.; Lemen-
ovskii, D. A.; Kotov, V. V. J. Am. Chem. Soc. 1995, 117, 10133-10134.
(24) Boryl synthesis from metathesis of B-H and Nb-H bonds has
precedent: Hartwig, J. F.; De Gala, S. R. J. Am. Chem. Soc. 1994, 116,
3661-3662.

9704 J. Am. Chem. Soc., Vol. 119, No. 41, 1997
Lantero et al.
Scheme 2
the positions where deuterium incorporation is expected based
on this operating mechanism.
As a final test for hydroboration by X2BH species, the reaction
of 2a with HBCat′ was performed in the presence of excess
cyclooctene. The rates for uncatalyzed hydroboration of cy-
clooctene and R-olefins by BH3 are competitive.26 Thus, any
B-C bond forming pathway that does not occur at the metal
should give appreciable quantities of cyclooctene hydroboration
n
products. In this case, 2a is converted to 6a and PrBCat′ is
the primary alkylborane product observed. This result, coupled
with the isotopic labeling experiments, conclusively demon-
strates that B-C bond formation occurs at the metal center.
“Catecholateboryl” complexes of the group 5 transition
elements are rare. Compounds 5 and 6 are the first in this family
supported by pentamethylcyclopentadienyl ligands, and they join
the recently reported examples endo-Cp2Nb(H2BCat) (13),24
endo-Cp2TaH2(BCat) (14), and exo-Cp2TaH2(BCat) (15).27 For
endo-Cp′2MH2BX2 complexes, three limiting forms that can be
used to describe H2BX2 ligation include d0 boryl (A), d2
borohydride (B), and d2 agostic borane (C) structures.28 Chemi-
cal shift perturbations on isotopic substitution (vide infra) and
crystallographic data support interactions between boron and
the hydride coligands in 13 and the central hydride ligand in
15. Compound 14 is the only definitive example of structure
A as B-H interactions are not confirmed by crystallographic
and spectroscopic data.
only be confirmed if kobs approaches a limiting value (the rate
of olefin insertion) at high borane concentration.
Since rate constant variations precluded detailed analyses, our
mechanistic hypotheses were tested by isotopic labeling and
trapping experiments. The reaction of 2a with catecholborane-d
provided considerable information with regard to the pathways
in Scheme 1. When the reactions were monitored by NMR,
scrambling between the deuterated borane and hydride or olefin
protons in the olefin complexes was not observed. Thus,
reversible B-H/M-H exchange by interaction between cat-
echolborane and 1 or 2 does not occur. Some important
conclusions can be drawn from the following observation: The
nascent alkylborane products show no deuterium incorporation.
On the basis of this result, scenarios where the organoborane
products arise from hydroboration of free olefins by DBCat or
BD3, generated from disproportionation of DBCat, can be
excluded. The isotopic labeling experiment also bears strongly
on the mechanistic possibilities in Scheme 1. For pathways B
and C, deuterated borane products would be expected given that
eliminations from d1-dihydride intermediates are common to
both mechanisms.25
Given the steric and electronic differences between Cp and
Cp* ligands, M-B or B-H interactions in 6a could differ
significantly from those in 13. Attempts to grow single crystals
of 5a and 6a were not successful; however, suitable single
crystals of the compound endo-Cp*2Nb(H2BO2C6H3-3-tBu) (16),
derived from reaction of 2a with HBO2C6H3-3-tBu, were grown
from a toluene/heptane solution. The molecular structure for
16 reveals some interesting features (Figure 1a). The structural
parameters for the “boryl” fragment differ significantly from
its Cp counterpart. Most notably, the Nb-B distance of
2.348(4) Å is considerably longer than the Nb-B distance of
2.292(5) Å in 13. This elongation in the Nb-B vector, in
addition to the positions of the hydride ligands (vide infra),
suggests that 16 might be more accurately described as a
borohydride complex. The best benchmark for comparison in
the decamethylmetallocene system is the borohydride complex,
Cp*2Nb(BH4) (17).17,29 Since 17 had not been crystallographi-
cally characterized, we determined its structure. For compound
17 (Figure 1b), the Nb-B bond length of 2.35(1) Å is identical,
within experimental error, to that observed in 16.
2
1
The fate of the deuterium label is revealed by H and H
NMR spectra. For the Nb complex, 6a, deuterium is incorpo-
rated at niobium-deuteride and Cp* methyl positions in a 90:
10 ratio (2H NMR). This is also confirmed by integrating the
1
methyl resonances of the d0, d1, and d2 isotopomers in the H
NMR spectra (vide infra). Label scrambling in the Nb
compound is not surprising as deuterium incorporation at the
Cp* methyl positions has been observed when Cp*2NbH3 is
treated with D2. This scrambling was attributed to reversible
C-H activation of the Cp* methyl group in the intermediate
hydride, Cp*2NbD (8).16 Similar participation of 8 in the present
reaction could account for scrambling of the isotopic label in
6a. When crystalline samples of labeled 6a are stored at room
temperature for extended periods, the isotopic label at the
hydride position gradually diminishes and incorporation at the
cyclopentadienyl methyl positions is enhanced (1H, 2H NMR).
For the Ta propylene-hydride complex, reaction with DBCat′
generates 6b-d2. In this case, deuterium incorporation was only
observed at the metal deuteride positions. The experimental
results from the labeling experiment are most consistent with
the mechanism in pathway A as the deuterium label should be
confined to the metal complexes 6a and 6b. Scheme 2 shows
Geometric parameters for the hydride ligands are crucial for
discriminating between boryl and borohydride structures. Al-
though the uncertainties in hydride positions determined by
X-ray diffraction are large when a hydride is bonded to a heavy
element, useful information can be obtained if the data are of
sufficient quality and proper care is exercised in their interpreta-
(26) Suzuki, A.; Dhillon, R. S. Top. Curr. Chem. 1986, 130, 23-88.
(27) Lantero, D. R.; Motry, D. H.; Ward, D. L.; Smith, M. R., III. J.
Am. Chem. Soc. 1994, 116, 10811-10812.
(28) (a) Calderon, J. L.; Cotton, F. A.; Shaver, A. J. Organomet. Chem.
1972, 42, 419-427. (b) Cotton, F. A.; Jeremic, M.; Shaver, A. Inorg. Chim.
Acta 1972, 6, 543-551.
(29) An Nb-B distance of 2.26 Å has been reported in Cp2Nb(BH4).
The uncertainty in the Nb-B distance from this study is 0.06 Å: Kirillova,
N. I.; Gusev, A. I.; Struchkov, Y. T. J. Struct. Chem. USSR 1974, 15, 622-
624.
(25) If dihydride intermediates are involved, the failure to detect
deuterated borane products would require that kH/kD > 50.

Models for Metal-Mediated Hydroboration
J. Am. Chem. Soc., Vol. 119, No. 41, 1997 9705
symmetrically disposed, the refined positions in 16 are asym-
metric. We hesitate to attach undue significance to this
observation, although contributions from an agostic borane
structure should not be dismissed given the precedent established
by Cotton and co-workers28 and the recently reported structure
of Cp2Ti(η2-HBCat)2.34 In any event, the metrical data for the
hydride ligands and the long Nb-B distance are clearly
consistent with pronounced Nb(III) character in 16 relative to
its Cp counterpart, compound 13. Of the structurally character-
ized compounds with the formulation Cp2Nb(H2BX2), the MH2-
BX2 linkage in 16 most closely resembles the complex that
results from the reaction between Cp2NbH3 and 9-BBN, Cp2-
Nb(H2BC8H14) (18).24
The recently synthesized boryl compounds in these group 5
metallocene systems establish a database for assessing relation-
ships between structural and spectroscopic features for boryl
and borohydride complexes. Relevant NMR and IR data for
Cp′2MH2BX2 derivatives and related hydride complexes are
given in Table 2.15-17,24,27,35-38 At first glance, the 11B chemical
shifts appear to correlate with the boryl and borohydride
character. For example, endo-Cp2TaH2(BCat) and endo-Cp*2-
TaH2(BCat) exhibit low-field chemical shifts and their structural
and spectroscopic data are more consistent with d0 boryl
structures. Meanwhile, in parent and alkyl-substituted d2
borohydride complexes, the 11B nuclei resonate at higher field.
However, the 11B data for Cp2Nb(H2BCat) and Cp*2Nb(H2BCat)
clearly illustrate the limitations of empirical correlations of
chemical shift to boryl and borohydride bonding modes. This
is not surprising as the observed chemical shift results from
several contributing factors. While boryl and borohydride
complexes typically exhibit a 11B resonance between 40 and
75 ppm, subtle structural variations cannot be reliably correlated
to chemical shift.
For compound 13, Hartwig and co-workers attributed large
perturbations in the hydride chemical shifts to rapid chemical
equilibrium between d0 borylhydride (A) and d2 dihydridoborate
(B) structures.24 For compounds 6a-d0-2, we observe perturba-
tions in the hydride resonances of similar magnitude to those
in 13-d0-2, and the hydride/deuteride resonance shifts to higher
field as D is substituted for H. For the tantalum congener, 6b,
1H NMR spectra reveal only minor perturbations in the hydride
resonances to the d0 and d1 isotopomers as was observed for
the Cp analogue, 14.
Figure 1. ORTEP diagrams of (a) endo-Cp*2Nb(H2BO2C6H3-3-tBu)
(16) and (b) Cp*2Nb(H4) (17). Ellipsoids represent 30% probability
for electron density.
tion. After the refinement for the non-hydrogen atom positions
in 16 and 17 had converged, difference maps revealed Nb-H
peaks in chemically reasonable positions. These atom positions
were added to the models, and in both cases they refined
normally. To ensure that these refined positions were not
artifacts of residual electron density at the Nb center, difference
maps were calculated by fixing the non-hydrogen atom positions
and using low-angle data (4 < 2θ < 30).30 Difference Fourier
maps (see Supporting Information) clearly show maxima at the
same positions as found by refining the hydride ligands. Thus,
the hydride positions do not appear to be artifacts of refinement.
Table 1 contains the metrical parameters for the Nb(H2BX2)
moieties in 13, 16, and 17.
The H-Nb-H angle in 16 more closely resembles that in
17 (58(3)°) than Hexo-Nb-Hexo in Cp2NbH3 (126 (3)°),31 and
the B-H distances in 16 lie near the mean found for the
corresponding distances in 13 and 17. The B-H contacts are
longer than those typically found in borohydride complexes or
diborane; however, similar B-H distances have been reported
for hydrides that bridge boron centers bearing N or O substit-
uents.32,33 Thus, 16 may be best described as an HBCat′ adduct
of “Cp*2NbH”. While the hydride ligands in 13 and 17 are
For the sake of the current discussion, it is important to
consider the origins of these abnormally large perturbations and
their bearing on potential solution structures. Isotopic perturba-
tions for 1H resonances in partially deuterated hydrocarbons are
well-documented.39 For example, the methyl resonance in
toluene-dR lies 0.015 ( 0.002 ppm upfield of the same resonance
in toluene.40 The interpretation of the large isotopic perturba-
tions in Cp′2Nb(H2BCat) is more complex. By using the simple
harmonic oscillator approximation, vibrational zero-point ener-
gies can be expressed as a function of the force constant and
reduced mass. By virtue of deuterium’s larger mass, the zero-
point energies in deuterium isotopomers of A and B will be
(34) Hartwig, J. F.; Muhoro, C. N.; He, X. J. Am. Chem. Soc. 1996,
118, 10936-10937.
(35) Antin˜olo, A.; Fajardo, M.; Otero, A.; Royo, P. J. Organomet. Chem.
1983, 246, 269-278.
(30) The scattering factor for hydrogen decays rapidly as sin(θ/λ)
increases. Thus, hydride positions can be more reliably located from low-
angle data: La Placa, S. J.; Ibers, J. A. Acta Crystallogr. 1965, 18, 511-
519.
(36) Green, M. L. H.; Wong, L. L. J. Chem. Soc., Chem. Commun. 1989,
571-573.
(37) Tebbe, F. N.; Parshall, G. W. J. Am. Chem. Soc. 1971, 93, 3793-
3795.
(31) Wilson, R. D.; Koetxle, T. F.; Hart, D. W.; Kvick, Å.; Tipton, D.
L.; Bau, R. J. Am. Chem. Soc. 1977, 99, 1775-1781.
(32) Mu¨ller, G.; Kru¨ger, C. Acta Crystallogr. 1986, C42, 1856-1859.
(33) Yalpani, M.; Ko¨ster, R.; Boese, R. Chem. Ber. 1993, 126, 565-
570.
(38) Green, M. L. H.; McCleverty, J. A.; Pratt, L.; Wilkinson, G. J. Chem.
Soc. 1961, 4854-4859.
(39) Bovey, F. A. Nuclear Magnetic Resonance Spectroscopy, 2nd ed.;
Academic Press: San Diego, 1988; p 141.
(40) Tiers, G. V. D. J. Chem. Phys. 1958, 29, 963-964.

9706 J. Am. Chem. Soc., Vol. 119, No. 41, 1997
Lantero et al.
Table 1. Geometric Parameters for endo-Cp′2Nb(H2BX2) Compounds
compd
Nb-B
Nb-H
B-H
H-Nb-H
H-B-H
endo-Cp2Nb(H2BCat) (13)
2.292(5)
1.62(4)
1.58(5)
1.86(4)
1.63(4)
1.88(7)
1.79(8)
1.69(4)
1.62(5)
1.36(4)
1.46(4)
1.17(7)
1.16(8)
92(2)
endo-Cp*2Nb(H2BO2C6H3-3-tBu) (16)
Cp*2Nb(BH4) (17)
2.348(4)
2.35(1)
73(2)
58(3)
95(2)
100(5)
Table 2. Spectroscopic Data for Cp′2M(H2BX2) and Cp′2MH3 Compounds
b
b
compd
11B NMRa
ν
BHM (cm-1
)
ν )
MH (cm-1
endo-Cp*2Nb(H2BCat) (5a)
endo-Cp*2Nb(H2BO2C6H3-4-tBu) (6a)
endo-Cp*2Nb(H2BO2C6H3-3-tBu) (16)
endo-Cp2Nb(H2BCat)c (13)
endo-Cp*2Ta(H2BCat) (5b)
endo-Cp*2Ta(H2BO2C6H3-4-tBu) (6b)
endo-Cp2Ta(H2BCat)e (14)
exo-Cp2Ta(H2BCat)e (15)
Cp2Nb(H2BC8H14)c (18)
Cp*2Nb(BH4)f (17)
60.2
60.0
60.9
59
73.5
72.7
70.0
64.7
57
1653
1653
1649
1742, 1675d
1774, 1709
1767, 1720
1768
1771, 1703
1605
1728, 1620
1745, 1628
1882-1860
1747, 1621
43.4
40.5
Cp2Nb(BH4)g
Cp2Ta(BH4)h
CpCp*Ta(BH4)i
47.2
Cp*2NbH3f
1747, 1695
1710
1750, 1720
1735
j
Cp2NbH3
k
Cp*2TaH3
l
Cp2TaH3
a 11B spectra were recorded in benzene-d6 and are referenced to an external BF3‚OEt2 standard. b Recorded as Nujol mulls unless otherwise
indicated. c Reference 24. d KBr pellet. e Reference 27. f Reference 17. g Reference 15. h Reference 35. i Reference 36. j Reference 37. k Reference
16. l Reference 38.
Scheme 3
lower than their hydride counterparts. More importantly, if the
force constants for the hydride vibrational modes in A and B
are not identical, the energy separations ∆G°(A-d0) - ∆G°(B-
d0) and ∆G°(A-d1) - ∆G°(B-d1) will be different, and deuteration
will shift the equilibrium between structures A and B. If the
equilibrium is rapid on the NMR time scale, the observed
hydride chemical shift will be a weighted average of the
chemical shifts for the hydride ligands in the limiting structures,
and the isotopic perturbation of the equilibrium between A and
B will lead to a shift in the hydride/deuteride resonance.41 The
fact that the hydride/deuteride resonances shift to higher field
with each deuterium substitution excludes any equilibrium
between degenerate structures. Although equilibrium between
A and B is an appealing explanation for the observed isotopic
perturbation, other nondegenerate equilibria could account for
these spectroscopic features. Some potential alternatives are
shown in Scheme 3.42 Lastly, the equilibria responsible for
in assigning structural types based on spectroscopic similarities,
perturbations in compounds 16 and 13 need not be analogous.
the spectroscopic data for 6b do not support significant B-H
The large chemical shift perturbations provide strong evidence
for equilibrium between structural extremes in these systems;
however, the structures that participate in the equilibria cannot
be unambiguously assigned from these data.
interactions. Thus, 6b most likely adopts a d0 boryl structure
similar to its Cp counterpart, 14.
Vibrational spectroscopy is useful for identifying three-center,
two-electron bonding in borohydride and related agostic
complexes.43-48 With several structurally characterized com-
pounds in hand, the utility of vibrational spectroscopy for
Attempts to crystallographically characterize 6b have not been
successful; however, “normal” isotopic perturbations for the
hydride/deuteride chemical shifts in 6b-d0-2 argue against rapid
structural equilibration. Although structural differences for the
Nb Cp and Cp* compounds clearly illustrate the pitfalls inherent
(43) Nakamoto, K. Infrared and Raman Spectra of Inorganic and
Coordination Compounds; 4th ed.; Wiley-Interscience: New York, 1986.
(44) Marks, T. J.; Kennelly, W. J. J. Am. Chem. Soc. 1975, 97, 1439-
1443.
(41) The difference in chemical shifts between isotopomers is temperature
dependent. At 20 °C, ∆δ between Cp2*Nb(H2BCat′) and Cp2*Nb(HDBCat′)
and Cp2*Nb(HDBCat′) and Cp2*Nb(D2BCat′) is 0.42 and 0.43 ppm,
respectively. At -60 °C, ∆δ between Cp2*Nb(H2BCat′) and Cp2*Nb-
(HDBCat′) and Cp2*Nb(HDBCat′) and Cp2*Nb(D2BCat′) is 0.54 and 0.51
ppm, respectively.
(45) Strong evidence has been presented for three-center, two-electron
bonding interactions in [H2B(3,5-Me2Pz)2](C7H7)(CO)2Mo in solution28a and
in the solid state.28b Perturbations in νBH were not reported; however, it is
interesting to note that only one BH stretch is observed in IR spectra for
these compounds instead of the two stretches normally observed in
bispyrazolylborate complexes: Trofimenko, S. Inorg. Chem. 1970, 9, 2493-
2499.
(46) Kubas, G. J. Acc. Chem. Res. 1988, 21, 120-128.
(47) Brookhart, M.; Green, M. L. H. J. Organomet. Chem. 1983, 250,
395-408.
(42) Intermediacy of Cp*2Nb(η1-H2BCat′) could account for the observed
scrambling between Cp* methyl hydrogen and Nb-D positions in solid
samples of 6a; however, rapid HBCat′ elimination prevents evaluation of
intramolecular processes.

Models for Metal-Mediated Hydroboration
J. Am. Chem. Soc., Vol. 119, No. 41, 1997 9707
Scheme 4
identifying more subtle variations in ligation can be assessed.
From similar stretching frequencies for terminal hydride stretches
in Cp*2NbH3 and bridging borohydride vibrations in Cp*2NbBH4,
it is apparent that boryl/hydride and borohydride structures
cannot be distinguished on the basis of IR data. With respect
to agostic borane structures, νBH can be estimated to lie within
a broad frequency range from 2200 to 1600 cm-1 by considering
decreases in νEH observed upon coordination of other E-H
bonds (E ) H, C, Si) to metal centers. A high-frequency νBH
mode could be correlated to an agostic borane structure;
however, assignments for lower frequency modes observed for
the compounds in this study are ambiguous since this portion
of the window overlaps regions typical for stretching frequencies
in borohydride and metal-hydride complexes.
Unfortunately, structures of endo-Cp2M(H2BX2) complexes
cannot be unambiguously assigned from their NMR and IR
spectra. In some cases spectroscopic data can provide informa-
tion regarding interactions between boron and hydride coligands;
however, we hesitate to attach undue significance to these
observations since the reactivity of compounds where spectro-
scopic similarities are striking can differ dramatically.
In marked contrast to other catecholate-substituted group 5
boryl/borohydride complexes, 6a exhibits clean reaction chem-
istry with various small molecules (Scheme 4) that is reminiscent
of Cp2M(BH4) complexes (M ) Nb,15,49 Ta35). The unifying
feature for the reactions in Scheme 4 is elimination of HBCat′
from 6a under mild conditions. Some details of the reactions
shown in Scheme 4 merit comment. The reaction with ethylene
is interesting, as extrusion of HBCat′ (observed by 11B NMR)
is accompanied by regeneration of the olefin-hydride complex
1a that served as the starting point for synthesis of the boryl
complex. On further heating, the liberated borane is converted
to EtBCat′. As this result would suggest, solutions of 6a
catalyze olefin hydroboration.
Cp*2NbH3 and HBCat′ under mild conditions. A reasonable
mechanism involves B-H elimination to generate “Cp*2NbH”,
which is trapped by H2. For the reaction between 6a and D2,
HBCat′ is the initial elimination product as judged by 11B NMR.
Thus, elimination of the metal-bound hydride is rapid with
respect to deuterium scrambling. As the reaction proceeds,
DBCat′ is produced from the equilibrium between Cp*2NbH3-xDx
and HBCat′. Related reactions of 6a with CO and B2H6
generate the Nb(III) complexes Cp*2NbH(CO) and Cp2*Nb(BH4)
with elimination of HBCat′. For these species, HBCat′ elimina-
tion is irreversible as 6a cannot be regenerated by addition of
HBCat′ to Cp*2NbH(CO) or Cp2*Nb(BH4) under thermal
conditions. In the elimination reactions outlined, traces of
B2Cat3 (typically <10% by 11B integration) are formed in these
systems. We have not been able to pinpoint the source(s)
responsible for this degradation.
Cp*2Nb(H2BCat′) (6a) also eliminates borane more readily
than does Cp*2Nb(BH4) (17). This difference in reactivity
parallels related chemical behavior of BH3 and HBCat. While
HBCat exists as a monomer, the barrier for dissociation of B2H6
into two BH3 fragments is considerable.50 Thus, if the barrier
for borane elimination from 6a and 17 is also sensitive to borane
heteroatom substitution, a lower activation energy would be
expected for 6a relative to 17.
The tantalum analog, compound 6b, reacts sluggishly when
compared to its Nb cousin, as illustrated by the reactions of 6a
and 6b with CO that generate HBCat′ and the carbonyl
complexes Cp*2Nb(H)CO and Cp*2Ta(H)CO, respectively. For
6a, the reaction proceeds with a half-life of 80 min at 0 °C,
while the corresponding reaction for 6b has a half-life of 130
min at 70 °C. The difference in reactivity between 6a and 6b
would be conveniently explained if 6b is a Ta(V) boryl complex.
In this case a higher barrier for reductive elimination of HBCat′
would be expected for 6b relative to the Nb(III) species, 6a,
since borane elimination from a d0 boryl complex requires a
formal two-electron reduction of the metal center. The observa-
tion that other niobocene and tantalocene MH2BCat compounds
Compound 6a also reacts with other two-electron ligands.
In each case, elimination of HBCat′ formally precedes trapping
of “Cp*2NbH” by oxidative-addition or coordination of the
substrate in question. For example, 6a reacts with H2 to generate
(48) Spaltenstein, E.; Palma, P.; Kreutzer, K. A.; Willoughby, C. A.;
Davis, W. M.; Buchwald, S. L. J. Am. Chem. Soc. 1994, 116, 10308-
10309.
(49) Otto, E. E. H.; Brintzinger, H. H. J. Organomet. Chem. 1978, 148,
29-33.
(50) A dissociation enthalpy of 36.0 ( 3.0 kcal/mol has been measured
for diborane.50a Computational methods predict a similar barrier for this
reaction:50b (a) Fehlner, T. P.; Mappes, G. W. J. Phys. Chem. 1969, 73,
873-882. (b) Stanton, J. F.; Bartlett, R. J.; Lipscomb, W. N. Chem. Phys.
Lett. 1987, 138, 525-530.

9708 J. Am. Chem. Soc., Vol. 119, No. 41, 1997
Lantero et al.
with significant M(V) character do not readily eliminate borane
is consistent with this notion.
in the formal oxidation state of the metal. Thus, the disparities
in rates for HBCat elimination from group 5 metallocene boryl/
borohydride species are most likely related to their structures.
The facile elimination of HBCat′ from 6a is the key to its
catalytic activity. While hydrocarbon activation and elimination
have been well studied, considerably less is known for analogous
reactivity of B-H bonds.5aik,7,51,52 Since thermal chemistry is
clean and deuterium isotopomers can be readily prepared, this
system is particularly well-suited for studying the factors that
influence elimination of B-H bonds from transition metal
centers. We are currently investigating the mechanistic details
of borane elimination from these compounds, and are exploring
reactivity of related olefin-hydride complexes.
Conclusions
The olefin complexes 1 and 2 serve as well-defined models
for investigating metal-mediated hydroboration. Isotopic label-
ing and deuterium trapping experiments strongly support a
mechanism where B-C bond formation results from reaction
of borane reagent with the 16-electron alkyl intermediates
Cp*2MCH2CH2R. Significantly, hydroboration by transient BH3
is not observed under stoichiometric or catalytic conditions. In
contrast to the chemistry observed for Cp*2Ti(η2-CH2dCH2)
and related Ti olefin complexes, isotopic labeling experiments
preclude borane attack on the olefin ligands in the ground state
complexes 1 and 2. Despite their disparate olefin borylation
products, the reactivity of Cp*2Ti(CH2dCH2) and Cp*2M-
(CH2dCH(R))(H) with borane reagents is linked by a common
feature. In both cases, interactions of B-H bonds with M-C
bonds in coordinatively unsaturated, 16-electron complexes/
intermediates most likely trigger B-C bond formation. The
postulated mechanism for hydroboration of the olefin ligands
in 1 and 2 is similar to those proposed for lanthanide and
zirconium systems. Although B-H oxidative addition is
precluded in lanthanide systems,10 a borohydride resting struc-
ture for lanthanide catalysts is not unreasonable given the
structure of 16.
Acknowledgment. The NMR equipment was provided by
the National Science Foundation (CHE-8800770) and the
National Institutes of Health (1-S10-RR04750-01). We thank
the administrators of the ACS Petroleum Research Fund, the
Exxon Educational Foundation, and the National Science
Foundation (CHE-9520176) for support of this work.
Supporting Information Available: Tables of atomic
coordinates, thermal parameters, and bond distances and angles
for 16 and 17 and electron density maps used in locating hydride
positions in 16 and 17 (46 pages). See any current masthead
page for ordering and Internet access instructions.
While boryl M(V) structures appear to be major contributors
in other catecholateboryl compounds, endo-Cp*2Nb(η2-H2-
BO2C6H3-3-tBu) is better described as an Nb(III) complex. The
structural parameters are consistent with its formulation as a
borohydride complex, or an HBO2C6H3-3-tBu adduct of
“Cp*2NbH”. In contrast to other group 5, metallocene cat-
echolateboryl complexes, Cp*2Nb(H2BCat′) readily eliminates
HBCat′. The differences in reactivity can be rationalized by
considering the changes in oxidation state that accompany
extrusion of borane for these species. The formal oxidation
state of the metal is unchanged for elimination from the
decamethylniobocene complexes, whereas borane elimination
from M(V) boryl complexes requires a two-electron reduction
JA970639B
(51) For related additions of B-H bonds to Rh see: (a) Baker, R. T.;
Ovenall, D. W.; Harlow, R. L.; Westcott, S. A.; Taylor, N. J.; Marder, T.
B. Organometallics 1990, 9, 3028-3030. (b) Westcott, S. A.; Taylor, N.
J.; Marder, T. B.; Baker, R. T.; Jones, N. L.; Calabrese, J. C. J. Chem.
Soc., Chem. Commun. 1991, 304-305. (c) Baker, R. T.; Calabrese, J. C.;
Westcott, S. A.; Nguyen, P.; Marder, T. B. J. Am. Chem. Soc. 1993, 115,
4367-4368.
(52) For oxidative additions of B-H bonds to Ir see: (a) Baker, R. T.;
Ovenall, D. W.; Calabrese, J. C.; Westcott, S. A.; Taylor, N. J.; Williams,
I. D.; Marder, T. B. J. Am. Chem. Soc. 1990, 112, 9399-9400. (b) Knorr,
J. R.; Merola, J. S. Organometallics 1990, 9, 3008-3010. (c) Westcott, S.
A.; Marder, T. B.; Baker, R. T.; Calabrese, J. C. Can. J. Chem. 1993, 71,
930-936. (d) Cleary, B. P.; Eisenberg, R. Organometallics 1995, 14, 4525-
4534.