Synthesis and properties of the 17-electron, tantalum-centered radical Ta(CO)4(Ph2PCH2CH2PPb2)

Article abstract of DOI:10.1021/om960133m

The novel 17-electron compound Ta(CO)₄dppe (dppe = Ph₂PCH₂CH₂PPh₂) is formed via hydride hydrogen atom abstraction from TaH(CO)₄dppe by the tris(p-fert-butylphenyl)methyl radical. The compound exists in solution as an equilibrium mixture of monomer (the dominant species) and carbonyl-bridged dimer [Ta(CO)₄dppe]₂ but solely as the latter in the solid state. It is very labile but was characterized electrochemically, IR, Raman, and ¹H and ³¹P{¹H} NMR spectroscopically, and chemically. Typical of metal-centered radicals, Ta(CO)₄dppe abstracts halogen atoms from organic halides RX to give the halotantalum compounds TaX(CO)₄dppe and, in some cases, TaR(CO)₄dppe. Cyclic voltammetry experiments show that the oxidation of [Ta(CO)₄(dppe)]^- in CH₂Cl₂/0.1 M [Bu₄N][PF₆] is a one-electron process, irreversible at scan rates below 50 V/s. The formal potential of [Ta(CO)₄(dppe)]^0/- is estimated as -1.2 V vs ferrocene. The reduced lifetime of the radical is ascribed to reactions with the ionic medium, the main electrochemical oxidation products being TaH(CO)₄(dppe) and TaCl(CO)₂(dppe)₂.

Full text of DOI:10.1021/om960133m

Organometallics 1996, 15, 3289-3302
3289
Syn th esis a n d P r op er ties of th e 17-Electr on ,
Ta n ta lu m -Cen ter ed Ra d ica l Ta (CO)4(P h 2P CH2CH2P P h 2)
Mary Ann Koeslag and Michael C. Baird*
Department of Chemistry, Queen’s University, Kingston, ON, Canada K7L 3N6
Sherri Lovelace and William E. Geiger*
Department of Chemistry, University of Vermont, Burlington, Vermont 05445
Received February 23, 1996X
The novel 17-electron compound Ta(CO)4dppe (dppe ) Ph2PCH2CH2PPh2) is formed via
hydride hydrogen atom abstraction from TaH(CO)4dppe by the tris(p-tert-butylphenyl)methyl
radical. The compound exists in solution as an equilibrium mixture of monomer (the
dominant species) and carbonyl-bridged dimer [Ta(CO)4dppe]2 but solely as the latter in
the solid state. It is very labile but was characterized electrochemically, IR, Raman, and
1H and 31P{1H} NMR spectroscopically, and chemically. Typical of metal-centered radicals,
Ta(CO)4dppe abstracts halogen atoms from organic halides RX to give the halotantalum
compounds TaX(CO)4dppe and, in some cases, TaR(CO)4dppe. Cyclic voltammetry experi-
ments show that the oxidation of [Ta(CO)4(dppe)]- in CH2Cl2/0.1 M [Bu4N][PF6] is a one-
electron process, irreversible at scan rates below 50 V/s. The formal potential of
[Ta(CO)4(dppe)]0/- is estimated as -1.2 V vs ferrocene. The reduced lifetime of the radical
is ascribed to reactions with the ionic medium, the main electrochemical oxidation products
being TaH(CO)4(dppe) and TaCl(CO)2(dppe)2.
Although 17-electron, metal-centered radicals have
been characterized for many transition metals,1 such
complexes of the group 5 metals, vanadium, niobium,
and tantalum, have been isolated only in the case of
vanadium. Indeed, V(CO)6 was reported over 30 years
ago2 and remains unique among this class of compounds
because it is a stable radical and its derivatives were
characterized long before metal-centered radicals were
recognized as possibly being common and of general
interest. Furthermore, even now, V(CO)6 is the only
simple binary metal carbonyl complex to be isolated as
a radical monomer which is thermally stable to dimer-
ization,2 and its electronic structure has been the subject
of intense study.2-5 V(CO)6 readily undergoes carbonyl
substitution with neutral, soft ligands (L) such as
phosphines, forming 17-electron monomers of the gen-
eral formula V(CO)4-nLn (n ) 1-4).6 In general, the
substituted carbonyl complexes are more thermally
stable than is V(CO)6, and many have been isolated at
room temperature. In contrast, treatment of V(CO)6
with hard Lewis bases B results in disproportionation
to vanadium(II) and -(-I) species [VB6]2+ and [V(CO)6]-.6b
Although vanadium(0) complexes are thus well-
documented, the corresponding monomeric complexes
of niobium(0) and tantalum(0) have not been isolated.
The neutral hexacarbonyl of tantalum has been tenta-
tively identified in argon matrices by comparison with
the infrared spectrum of V(CO)6,7a but neither Nb(CO)6
nor Ta(CO)6 has been observed under ambient condi-
tions. In 1989, Calderazzo et al.8a reported the synthesis
of cluster complexes, [AgM(CO)4dmpe]3 (dmpe ) 1,2-
bis(dimethylphosphino)ethane), in which niobium and
tantalum were proposed to be present in the zero
oxidation state (eq 1). These complexes were character-
3[Et4N][M(CO)4(dmpe)] + 3AgX f
[AgM(CO)4(dmpe)]3 + 3[Et4N]X (1)
M ) Nb, Ta; X ) NO2, BF4
ized by single-crystal X-ray diffraction,8 which showed
that the structures consist of heterometallic rafts in
which the metal atoms are arranged in almost equi-
lateral triangles with tantalum atoms at the vertices
of a large triangle and silver atoms at the midpoints of
each side. On the basis of the short Ag-Ag bonds (2.842
X Abstract published in Advance ACS Abstracts, J uly 1, 1996.
(1) (a) Baird, M. C. Chem. Rev. 1988, 88, 1217. (b) Trogler, W. C.,
Ed.; Organometallic Radical Processes, J . Organomet. Chem. Library
22; Elsevier Science Pub.: Amsterdam, 1990.
(2) (a) Ercoli, R.; Calderazzo, F.; Alberola, A. J . Am. Chem. Soc.
1960, 82, 2966. (b) Calderazzo, F.; Cini, R.; Ercoli, R. Chem. Ind. 1960,
934. (c) DeVore, T. C.; Franzen, H. F. Inorg. Chem. 1976, 15, 1318.
(3) (a) Haas, H.; Sheline, R. K. J . Am. Chem. Soc. 1966, 88, 3219.
(b) Bernier, J . C.; Kahn, O. Chem. Phys. Lett. 1973, 19, 414.
(4) Bellard, S.; Rubinson, K. A.; Sheldrick, G. M. Acta Crystallogr.
B 1979, 35, 271.
(5) (a) Pratt, D. W.; Myers, R. J . J . Am. Chem. Soc. 1967, 89, 6470.
(b) Rubinson, K. A. J . Am. Chem. Soc. 1976, 98, 5188. (c) Bratt, S. W.;
Kassyk, A.; Perutz, R. N.; Symons, M. C. R. J . Am. Chem. Soc. 1982,
104, 490. (d) Boyer, M. P.; Le Page, Y.; Morton, J . R.; Preston, K. F.;
Vuolle, M. J . Can. J . Spec. 1981, 26, 181. (e) McCall, J . M.; Morton, J .
R.; Preston, K. F. Organometallics 1985, 4, 1272. (f) Pruett, R. L.;
Wyman, J . E. Chem. Ind. 1960, 119. (g) McCall, J . M.; Morton, J . R.;
Preston, K. F. J . Magn. Reson. 1985, 64, 414.
(6) (a) Shi, Q.-Z.; Richmond, T. G.; Trogler, W. C.; Basolo, F. J . Am.
Chem. Soc. 1984, 106, 71. (b) Richmond, T. G.; Shi, Q.-Z.; Trogler, W.
C.; Basolo, F. J . Am. Chem. Soc. 1984, 106, 76. (c) Ellis, J . E.; Faltynek,
R. A.; Rochfort, G. L.; Stevens, R. E.; Zank, G. A. Inorg. Chem. 1980,
19, 1082. (d) Davison, A.; Ellis, J . E. J . Organomet. Chem. 1972, 36,
131. (e) Behrens, H.; Lutz, K. Z. Anorg. Allg. Chem. 1968, 356, 225.
(7) (a) DeKock, R. L. Inorg. Chem. 1971, 10, 1205. (b) Adams, D. M.
Metal-Ligand and Related Vibrations; Edward Arnold Publishers:
London, 1967; p 122.
(8) (a) Calderazzo, F.; Pampaloni, G.; Englert, U.; Stra¨hle, J . Angew.
Chem., Int. Ed. Engl. 1989, 28, 471. (b) Calderazzo, F.; Pampaloni, G.
J . Organomet. Chem. 1992, 423, 307. (c) Calderazzo, F.; Pampaloni,
G.; Englert, U.; Stra¨hle, J . J . Organomet. Chem. 1990, 383, 45.
S0276-7333(96)00133-1 CCC: $12.00 © 1996 American Chemical Society
+
+

3290 Organometallics, Vol. 15, No. 15, 1996
Koeslag et al.
Å for the Nb and Ta complexes, compared to 2.883 Å in
silver metal) and the Ag-M bond lengths (2.888 Å)
which compare favorably to the sums of the metallic
radii (2.91 Å), these clusters were characterized as
complexes of niobium(0) and tantalum(0), although no
evidence for dissociation to odd-electron Nb(0) or Ta(0)
species has been noted.
reacted with protonic acids, anionic dimeric complexes
were obtained (eq 4).11a-c
2[M(CO)6]- + 3HX f [M2(µ-X)3(CO)8]- +
4CO + 1.5H2 (4)
M ) Nb, Ta; X ) Cl, Br, I, MeCO2
Attempts to synthesize simple complexes of nio-
bium(0) or tantalum(0) have begun with diamagnetic,
18-electron complexes in the -1 (d6) and +1 (d4)
oxidation states. By effecting one-electron oxidations
or reductions, respectively, of these types of complexes,
one may in principle obtain the desired 17-electron
radicals, and many such attempts have been made. The
18-electron, octahedral carbonylate ions [M(CO)6]- (M
) V, Nb, Ta) have been prepared via reductive car-
bonylation of the chloro compounds VCl3,9a Nb2Cl10,9b
and TaCl5,9b respectively, in high yield, atmospheric
pressure syntheses of [Nb(CO)6]- and [Ta(CO)6]- being
first reported some years ago.10 Thus Calderazzo et
al.10a described the synthesis of [Nb(CO)6]- using mag-
nesium and zinc as reducing agents (eq 2), while Ellis
et al.10b,e reported the synthesis of both hexacarbonyl-
ates using alkali metal naphthalenides (eq 3).
The same complexes have also been isolated from the
reactions of [Nb(CO)6]- with a variety of metal halides
and protonated bases.11b-d
In view of the greater stability of substituted vana-
dium(0) carbonyl complexes, attempts have also been
made to generate M(CO)6-nLn (M ) Nb, Ta; L ) tertiary
phosphines) by either reaction of the hexacarbonylate
anions in the presence of stabilizing ligands or by
reaction of the corresponding substituted anionic com-
plex. The reaction of [Nb(CO)6]- and [Ta(CO)6]- with
protonated bases was carried out with dppe added to
the reaction mixture (eq 5).11c The product of the
[M(CO)6]- + 2[HB]X + dppe f
MX(CO)4(dppe) + X- + H2 + 2CO + 2B (5)
reaction was not the 17-electron radical, M(CO)4(dppe),
but rather a substituted halo complex. Again two-
electron oxidation processes had occurred, although the
product was monomeric in the presence of the phos-
phine. By reaction of the anionic dimer with dppe, the
same monomeric halo complex was obtained (eq 6).11c
Nb2Cl10 + 6A + 12CO f 5ACl2 + A[Nb(CO)6]2 (2)
A ) Mg/Zn; yield ) 39%
MCl5 + 6A[C10H8] + 6CO f
[M2(µ-X)3(CO)8]- + 2dppe f 2MX(CO)4(dppe) + X-
A[M(CO)6] + 6C10H8 + 5ACl (3)
(6)
M ) Nb, Ta; A ) Na, Li; yields ) 30-50%
X ) Cl, I
A similar reaction was also attempted with [Ta(CO)6]-
and TaCl5, in the presence of PMe3 (eq 7).12 In this case
As salts of bulky cations, the hexacarbonylates are
thermally and chemically quite stable.10b-d The solid
+
materials Z[Ta(CO)6] (Z ) n-Bu4N+, Et4N+, or PPh4
)
[Ta(CO)6]- + TaCl5 + excess PMe3 f
TaCl(CO)3(PMe3)3 + Ta(III) (7)
can be handled in air for short periods of time, but
solutions of all such complexes are very air sensitive.10b
The salts Z[V(CO)6] are reported to be inert to thermal
CO substitution,10d but synthesis of substituted car-
bonylates can be achieved under photolytic conditions,
generating [M(CO)6-nLn]- (n ) 1-3). Phosphine- and
arsine-substituted carbonylates were reported to be
rather reactive under various conditions,10c but substi-
tuted complexes formed with chelating ligands are
generally substantially more stable thermally and less
air-sensitive than complexes containing two mono-
dentate phosphines.10c
the product was the substituted tricarbonyl halo-
complex, which was obtained in ∼75% yield (based on
the hexacarbonylate) although no tantalum-containing
byproduct was identified.12 The authors argued that,
since [Ta(CO)6]- does not react with PMe3 at room
temperature, PMe3 substitution in these reactions must
occur via a 17-electron radical, Ta(CO)6, or a halogeno
species, {Ta(CO)6X}. Since the reaction occurred under
mild reaction conditions, substitution via a 17-electron
radical was postulated to be more likely.12
Reactions of niobium(I) and tantalum(I) complexes
with reducing agents have been no more successful at
producing zerovalent complexes than have the at-
tempted oxidations of the anionic complexes; two-
electron reductions have always being observed. Thus
several attempts have been made to reduce complexes
of the general formulas MX(CO)4L2 and MX(CO)2L4 (M
The possibility of effecting a one-electron oxidation
of the hexacarbonylate anions of niobium and tantalum
to form the neutral hexacarbonyls has proven to be a
tantalizing but, as yet, unattainable goal. Instead, two-
electron oxidations have always been observed to occur
although a variety of oxidizing agents has been used.
For instance, when [Nb(CO)6]- and [Ta(CO)6]- were
(9) (a) Calderazzo, F.; Pampaloni, G. J . Organomet. Chem. 1983,
250, C33. (b) Ellis, J . E.; Davison, A. Inorg. Synth. 1976, 16, 68.
(10) (a) Calderazzo, F.; Englert, U.; Pampaloni, G.; Pelizzi, G. Inorg.
Chem. 1983, 22, 1865. (b) Dewey, C. G.; Ellis, J . E.; Fjare, K. L.; Pfahl,
K. M.; Warnock, G. F. P. Organometallics 1983, 2, 388. (c) Rehder, D.;
Dahlenburg, L.; Mu¨ller, I. J . Organomet. Chem. 1976, 122, 53. (d)
Davison, A.; Ellis, J . E. J . Organomet. Chem. 1971, 31, 239. (e) Ellis,
J . E.; Warnock, G. F.; Barybin, M. V.; Pomije, M. K. Chem. Eur. J .
1995, 1, 521.
(11) (a) Calderazzo, F.; Pampaloni, G.; Zanazzi, P. F. Chem. Com-
mun. 1982, 1304. (b) Calderazzo, F.; Castellani, M.; Pampaloni, G.;
Zanazzi, P. F. J . Chem. Soc., Dalton Trans. 1985, 1989. (c) Calderazzo,
F.; Pampaloni, G.; Pelizzi, G.; Vitali, F. Organometallics 1988, 7, 1083.
(d) Calderazzo, F.; Pampaloni, G.; Zanazzi, P. F. Chem. Ber. 1986, 119,
2796.
(12) Luetkens, J . L.; Santure, D. J .; Huffman, J . C.; Sattelberger,
A. P. Chem. Commun. 1985, 552.
+
+

Ta(CO)4(Ph2PCH2CH2PPh2)
Organometallics, Vol. 15, No. 15, 1996 3291
) Nb, Ta; X ) Cl, Br, I) but to no avail.8c,13 For instance,
Lippard et al. attempted the reaction of equimolar
quantities of TaCl(CO)2(dmpe)2 and sodium naphtha-
lenide, but a mixture of the starting material and
[Ta(CO)2(dmpe)2]-, resulting from a two-electron reduc-
tion, was obtained.13a In those cases where two or more
equivalents of a reducing agent (sodium or sodium
naphthalenide) were used, [Ta(CO)2(dmpe)2]- was the
only product.13 Reduction of the anionic complexes
discussed above also results in formation of the hexa-
carbonylates (eq 8).11b
this radical from the bromo precursor, with yields of the
radical approaching 100%, and therefore we decided to
work with the the tris(p-tert-butylphenyl)methyl radical.
The tantalum complex TaH(CO)4(dppe) (dppe ) 1,2-
bis(diphenylphosphino)ethane) was chosen as a suitable
starting material for our initial study of hydrogen
abstraction reactions with tris(p-tert-butylphenyl)-
methyl because it has been reported to be stable under
nitrogen at low temperatures for some weeks.
A
preliminary account of this work has appeared,14c and
we now describe this investigation in detail, providing
as well improved synthetic procedures for the precursors
Et4N[Ta(CO)4(dppe)] and TaH(CO)4(dppe). We also
report details of of an investigation of the electro-
oxidation of the former.
[M2(µ-X)3(CO)8]- + 4Na + 4CO f
2Na[M(CO)6] + 2NaX + X- (8)
From the above discussion, it is apparent that a great
number of oxidation and reduction reactions have been
attempted for the synthesis of 17-electron radicals of
niobium(0) and tantalum(0) without any success. The
+1 and -1 oxidation states are apparently too readily
accessible to allow one-electron processes to occur, and
therefore a different approach must be utilized if
complexes of Nb(0) or Ta(0) are to be synthesized.
As part of the development of new methodologies for
the synthesis of novel 17-electron species, we14ab and
others15 have demonstrated the utility of persistent
triarylmethyl radicals Ar3C• to abstract hydrogen atoms
from 18-electron, transition metal hydride complexes
MHLn and form the corresponding 17-electron com-
pounds MLn (eq 9). In the absence of suitable radical
Exp er im en ta l Section
Unless otherwise stated, all experiments were performed
under argon (passed through
a heated column of BASF
catalyst to remove traces of oxygen and through a column of
molecular sieves to remove traces of water) using standard
Schlenk line techniques or in a Vacuum Atmospheres glove
box under a nitrogen atmosphere. All solvents were dried
under nitrogen: ethyl ether, toluene, benzene, and pentane
over sodium; hexanes, tetrahydrofuran (THF), and 1,2-
dimethoxyethane (DME) over potassium; acetonitrile and
dichloromethane over calcium hydride. Benzene used for
reactions of (t-BuC6H4)3C• was freeze-thaw degassed at least
three times prior to use, as were all deuterated solvents for
NMR. Other solvents were deoxygenated by bubbling with
argon for a period of at least 30 min. Carbon monoxide was
used as received; the cylinders were not emptied below 500
psig since water was found to contaminate the last portion of
the cylinder gas. Chemicals were purchased from Aldrich,
Strem, and BDH and were used as received.
MHLn + Ar3C• f MLn + Ar3CH
(9)
traps, the species MLn may couple and the organo-
metallic products obtained are often the 18-electron,
diamagnetic dimers [MLn]2. On the other hand, in those
cases where radical coupling is not feasible for steric
reasons, the chemistry of eq 9 provides a very conve-
nient route to new 17-electron compounds. Since a
number of sterically hindered metal hydrides of the
types MH(CO)4L2 and MH(CO)2L4 (M ) Nb, Ta) are
available, it seemed possible that the corresponding
metal-centered radicals M(CO)4L2 and M(CO)2L4 (or
their dimers) might also be synthesized via abstraction
of the hydridic hydrogen atoms with triarylmethyl
radicals.
IR spectra were collected on Bruker IFS-85 or IFS-25 FT-
IR spectrometers, on samples in solution or as Nujol mulls.
Raman spectra were obtained using a Bruker RFS-100 FT-
Raman spectrometer, in most cases on finely-powdered solid
samples which had been placed in 5 mm glass tubes. For
Ta(CO)4(dppe), a spectrum could not be obtained in this way
because the solid was overheated by the YAG laser at all power
settings; solid samples were therefore mixed with Nujol, which
acted as a heat sink and thus allowed collection of spectra
without thermal degradation of the samples. IR and Raman
spectral data are given in Tables 1-4. UV-vis spectra were
recorded on a Hewlett Packard 8452A diode array spectro-
photometer.
Solution NMR spectra were recorded on Bruker AM-400 and
ACF-200 FT spectrometers; the chemical shift references used
While we have previously utilized the triphenylmethyl
(trityl) radical,14a,b the tris(p-tert-butylphenyl)methyl
radical has been used in studies of the kinetics of
hydrogen abstraction from organometallic hydride
complexes,15c and the tris(p-tert-butylphenyl)methyl
1
for solution spectra were TMS for H NMR and external H3PO4
1
for 31P{1H} NMR. For H NMR, internal references in the form
of TMS or the chemical shift of the residual proton of the
deuterated solvent relative to TMS were used. Solid-state
NMR spectra were recorded on a Bruker CXP-200 FT spec-
trometer on samples tightly packed into a MAS sample tube.
For 13C, the reference used was adamantane (external); for
31P, spectra were referenced to dppe (internal, δ -12.0). EPR
spectra were recorded at the National Research Council,
Ottawa, by Drs. K. Preston and P. Kaiser on a Varian E12
spectrometer equipped with a Bruker ER035M gaussmeter
and a Systron-Donner microwave frequency counter. The
magnetic susceptibilities of solid samples were measured on
a J ohnson Matthey magnetic susceptibility balance, and those
of samples in solution, using the Evans NMR method.16a-c
1
group has the advantage that the H NMR spectra of
derivatives exhibit sharp, readily identified and inte-
grated methyl resonances. Furthermore, Norton et al.15c
have published several methods for the generation of
(13) (a) Vrtis, R. N.; Liu, S.; Rao, Ch. P.; Bott, S. G.; Lippard, S. J .
Organometallics 1991, 10, 275. (b) Datta, S.; Wreford, S. S. Inorg.
Chem. 1977, 16, 1134.
(14) (a) Drake, P. R.; Baird, M. C. J . Organomet. Chem. 1989, 363,
131. (b) Kuksis, I.; Baird, M. C. Organometallics 1994, 13, 1551. (c)
Koeslag, M. D.; Baird, M. C. Organometallics 1994, 13, 11.
(15) (a) Bullock, R. M. Comments Inorg. Chem. 1991, 12, 1. (b)
Ungva´ry, F.; Marko´, L. J . Organomet. Chem. 1980, 193, 383. (c) Turaki,
N. N.; Huggins, J . M. Organometallics 1986, 5, 1703. (d) Eisenberg,
D. C.; Norton, J . R. Isr. J . Chem. 1991, 31, 55. (e) Eisenberg, D. C.;
Lawrie, C. J . C.; Moody, A. E.; Norton, J . R. J . Am. Chem. Soc. 1991,
113, 4888.
(16) (a) Evans, D. F. J . Chem. Soc. 1959, 2003. (b) Crawford, T. H.;
Swanson, J . J . Chem. Educ. 1971, 48, 382. (c) Sur, S. K. J . Magn.
Reson. 1989, 82, 169. (d) Int. Crit. Tables Numerical Data Phys., Chem.,
Technol. 1928, 3, 27-29.
+
+

3292 Organometallics, Vol. 15, No. 15, 1996
Koeslag et al.
Electrochemical measurements were carried out in a drybox
under N2. Reagent grade solvents were distilled from drying
agents (CaH2 for CH2Cl2, K for THF) under vacuum. The
supporting electrolye [Bu4N][PF6] was vacuum dried at 373 K
after recrystallization from ethanol. Platinum electrodes were
pretreated by polishing with diamond paste. Platinum disks
constructed from wire of diameter 125 or 495 mm (Goodfellow,
Inc.) were used in rapid scan experiments to minimize ohmic
loss. Virtually identical responses were also obtained at gold
and glassy carbon electrodes. PARC 173 or 273 potentiostats
were used for all electrochemical measurements.
and DME (150 mL) was added ([sodium naphthalenide] ) 0.43
M). The mixture immediately turned green-black and was
stirred at room temperature for 3 h to allow complete forma-
tion of the naphthalenide.
While the sodium naphthalenide was being formed, TaCl5
(3.50 g, 9.77 mmol) was added slowly to 150 mL of rapidly
stirred, cold (195 K) toluene. The sodium naphthalenide
solution was cooled to 195 K and added to the white slurry of
TaCl5 in toluene to produce a dark-brown mixture. The flask
was stirred at low temperature under CO for 6 to 8 h and then
allowed to warm to room temperature overnight.
Potentials in this paper are referred to the ferrocene/
ferrocenium couple, although the experimental reference
electrode was a silver wire anodized in HCl media to produce
a Ag/AgCl electrode. Our quoted potentials may be converted
to the aqueous SCE potential scale by addition of +0.46 V for
CH2Cl2 or +0.56 V for THF. Solutions of [Et4N][Ta(CO)4-
(dppe)] were always kept at 273 K or lower in order to
minimize decomposition.
Tr is(p -ter t-b u t ylp h en yl)m et h yl Br om id e, (t-Bu C6-
H4)3CBr . This was synthesized by the method of Colle and
Lewis for the chloro analogue.17 Recrystallization from toluene
yielded (t-BuC6H4)3CBr as a white powder (11.2 g, 33%). 1H
NMR in C6D6: δ 7.48 (d, J ) 8.6 Hz, 6H, phenyl H), 7.09 (d,
J ) 8.6 Hz, 6H, phenyl H), 1.17 (s, 27H, Me). 1H NMR in
CDCl3: δ 7.16 (d, J ) 8.7 Hz, 6H, phenyl H), 7.07 (d, J ) 8.6
Hz, 6H, phenyl H), 1.20 (s, 27H, Me).
Tr is(p -ter t-b u t ylp h en yl)m et h yl, (t-Bu C6H 4)3C•.15c (t-
BuC6H4)3CBr (100 mg, 0.203 mmol) and Zn powder (500 mg,
excess) were weighed into a flask, which was wrapped in
aluminum foil because of the light sensitivity of the radical.
Indeed, all reaction vessels, syringes, and cuvettes were
wrapped in foil. Benzene (25.0 mL) was added and the
reaction was stirred at room temperature for 3 h. After the
solids were allowed to settle (10-15 min), the required volume
of radical solution was removed by syringe or cannula.
Initially, the yield of the radical was determined by UV-vis
spectroscopy (for the radical, λmax ) 524 nm and ꢀ ) 825 M-1
cm-1).15c By using this method, the yield of (t-BuC6H4)3C• was
consistently ca. 80%. This reaction could be scaled up to 500
mg of (t-BuC6H4)3CBr, the yield remaining approximately the
same.
All subsequent manipulations were done under argon.
[Et4N]Br (3.00 g, 14.3 mmol) in water (50 mL) was slowly
added to the reaction mixture over 30 min. Filtration of the
brown slurry resulted in a bright orange-yellow solution and
a brown tarry residue which were separated by filtration. The
tar was washed with DME (5 × 20 mL), and the combined
washings and filtrate were pumped down to remove DME and
toluene. After most of the solvent was removed, 100 mL of
water was added and the bright yellow precipitate was filtered
out and dried. All naphthalene was removed by sublimation
to give 1.2 g of product (26%). IR (THF): νCO 1888 (w, sh),
1859 cm-1 (vs). Lit.10a (THF, [(Ph3P)2N]+ salt): νCO 1882 (sh),
1854 cm-1 (vs). Raman (powder): νCO 2024 (w), 1885 (vs), 1881
cm-1 (vs).
[Et4N][Ta (CO)4(d p p e)]. Meth od 1. [Et4N][Ta(CO)6] was
substituted with dppe photolytically, according to the method
of Davison and Ellis.10d Photolysis was carried out using a
450 W Hanovia mercury vapor lamp, which was placed in a
water-jacketed well in the reaction vessel. Dppe (1.70 g, 4.27
mmol) was weighed into a reaction vessel, and a solution of
[Et4N][Ta(CO)6] (2.00 g, 4.17 mmol) in THF (100 mL) was
added. The reaction vessel was placed on ice, and the sample
was photolyzed for 4-6 h or until the carbonyl bands for the
starting material had become very weak or had disappeared.
The photolysis was monitored carefully since Davison and
Ellis10d report that the substituted anions of tantalum are
easily decomposed by long exposure to UV radiation. Initially,
carbonyl bands at 1968 (m) and 1822 (vs) cm-1 were observed;
these were tentatively attributed to the intermediate
[Ta(CO)5(dppe)]- on the basis of literature data for
[Ta(CO)5PPh3]-.10d (A third band at ∼1860 cm-1 would be
obscured by a band of the starting material.)
Tr is(p-ter t-bu tylp h en yl)m eth a n e, (t-Bu C6H4)3CH.
A
suspension of (t-BuC6H4)3CBr (300 mg, 0.610 mmol) in ethyl
ether was treated with excess LiAlH4. The original white
suspension turned yellow immediately, but the color slowly
disappeared and the solid dissolved. The reaction was stirred
at room temperature for 30 min, quenched with ethanol, and
filtered. Water was added to precipitate the product from the
solution. The pale yellow solid was air-dried. Yield: 250 mg
(99%). 1H NMR in C6D6: δ 7.19 (m, 12H, phenyl H), 5.54 (s,
1H, CH), 1.21 (s, 27H, Me). Lit. (CDCl3):17 δ 7.11 (m, 12H,
phenyl H), 5.38 (s, 1H, CH), 1.27 (s, 27H, Me).
W(CO)4(d p p e). W(CO)4(dppe), was synthesized using the
method of Chatt et al.18 Recrystallization from toluene yielded
W(CO)4(dppe) as a pale yellow solid (3.2 g, 52%). IR (tolu-
ene): νCO 2017 (m), 1917 (s), 1904 (vs), 1887 cm-1 (s). Lit.18
(CHCl3): νCO 2020 (m), 1920 (sh), 1903 (s), 1885 cm-1 (sh).
Raman (powder): νCO 2013 (s), 1901 (vs), 1854 (vs), 1810 cm-1
(vw). 1H NMR in toluene-d8: δ 7.50 (m, 8H, phenyl H), 6.98
(m, 12H, phenyl H), 2.13 (m, 4H, CH2). 31P NMR in CDCl3: δ
40.9 (s).
The reaction solution was filtered into a flask, and the
precipitated orange-red solid was dried under vacuum and
found to be pure [Et4N][Ta(CO)4(dppe)] by spectroscopic
analysis. The photolysis apparatus was rinsed with aceto-
nitrile (4 × 25 mL) to remove red solid adhering to the sides,
and these acetonitrile washes were added to the filtrate. This
solution was pumped dry at or below room temperature.
The solid recovered from the removal of the solvent consisted
of [Et4N][Ta(CO)4(dppe)] plus small amounts of [Et4N][Ta-
(CO)5(dppe)] and free dppe. The solid was extracted into
acetonitrile (3 × 10 mL) to remove a quantity of beige solid
(dppe), and the extracts were pumped dry to yield a red-black,
sticky material. Scraping this vigorously in the presence of
pentane (discarding the pentane washes) allows a red solid to
be recovered. The solid can be recrystallized from acetone,
DME/hexanes, or from benzene/hexanes to yield red crystals
of the product. Total yield: 1.75 g (51%).
Meth od 2. In an attempt to improve the overall yield of
[Et4N][Ta(CO)4(dppe)], an adaptation of the method for the
synthesis of [Et4N][Ta(CO)6] was attempted. A solution of
dppe (3.90 g, 9.79 mmol) in 150 mL toluene was cooled to 195
K, and TaCl5 (3.50 g, 9.77 mmol) was added slowly, producing
a bright yellow suspension. A cold solution of sodium naph-
thalenide, prepared as above, was added to the suspension,
producing a dark-brown mixture. The reaction was continued
as described above for [Et4N][Ta(CO)6], warming to room
temperature overnight.
[E t 4N][Ta (CO)6]. [Et4N][Ta(CO)6] was synthesized by a
variation of the method of Ellis et al.,10a including the improve-
ments outlined for the synthesis of [M(CO)6]2- (M ) Ti, Zr,
Hf).19 Naphthalene (13.0 g, 0.10 mol, excess) and sodium (1.5
g, 0.065 mol, cut into small pieces) were weighed into a flask,
(17) Colle, T. H.; Lewis, E. S. J . Am. Chem. Soc. 1979, 101, 1810.
(18) Chatt, J .; Leigh, G. J .; Thankarajan, N. J . Organomet. Chem.
1971, 29, 105.
(19) Chi, K. M.; Frerichs, S. R.; Stein, B. K.; Blackburn, D. W.; Ellis,
J . E. J . Am. Chem. Soc. 1988, 110, 163.
All subsequent manipulations were done under argon.
[Et4N]Br (3.00 g, 14.3 mmol) was added to the dark slurry,
+
+

Ta(CO)4(Ph2PCH2CH2PPh2)
Organometallics, Vol. 15, No. 15, 1996 3293
which was stirred for 1 h and then filtered. (Water was not
added to this reaction when the anion was the desired product
because reaction results in formation of the hydride, TaH(CO)4-
(dppe).) The brown, tarry residue was washed repeatedly with
DME to give a dark red solution which was evaporated, at or
below room temperature, to give crude, red product. Naph-
thalene was removed by sublimation.
In some cases the solid was found to consist of a mixture of
the anions [Ta(CO)6]-, [Ta(CO)5(dppe)]-, and [Ta(CO)4(dppe)]-;
in general, photolysis of the solid (after sublimation of naph-
thalene) according to method 1 resulted in conversion to
[Ta(CO)4(dppe)]- as the major product. In most cases, how-
ever, [Ta(CO)4(dppe)]- was essentially the only product, and
purification as in method 1 was sufficient. Total yield: 2.4 g
(30%). IR data in various solvents are given in Table 1.
Raman (powder): νCO 1893 (vs), 1790 (w), 1779 (w), 1742 cm-1
(m). 1H NMR in CD3CN: δ 7.69 (m, 8H, phenyl H), 7.27 (m,
12H, phenyl H), 3.14 (q, J ) 7.3 Hz, 8H, CH2 of [Et4N]+), 2.37
(m, J ) 17.3 Hz, 4H, CH2 of dppe ligand), 1.19 (tt, J ) 1.8, 7.2
Hz, 12H, Me of [Et4N]+). 31P{1H} NMR in CD3CN: δ 49.2 (s).
31P{1H} NMR in C6D6: δ 51.1 (s). Lit.20 in THF/THF-d8: δ
46.8 (s). 31P NMR of solid: δ 53.0 (s), 49.0 (s), 43.6 (s). 13C
NMR of solid: δ 143.7 (s), 140.8 (s), 129.7 (s), 52.3 (s), 31.3 (br
m), 7.9 (s).
Ta (CO)4(d p p e). In a typical synthesis, 100-150 mg of
TaH(CO)4(dppe) was dissolved in benzene (5 mL) and the
solution was cooled to ∼285 K. A solution of (t-BuC6H4)3C•
(∼1.0 molar equiv, based on 80% conversion from (t-BuC6H4)3-
CBr) was added. There was an immediate color change from
red to dark brown. After the solution was stirred for 15 min,
the solvent was removed under vacuum. The resulting brown
solid was washed with hexanes (3 × 10 mL) to remove organic
compounds and dried under vacuum. When the reaction was
carried out at temperatures greater than 285 K, the major
product was TaBr(CO)2(dppe)2. The reaction was worked up
under fluorescent lighting in order to decompose any excess
(t-BuC6H4)3C• that might still be present.
Repeated attempts were made to recrystallize or to chro-
matograph the solid, but these were unsuccessful. The major
product collected from the chromatography of Ta(CO)4(dppe)
on silica gel with toluene was TaH(CO)4(dppe). The effects of
a variety of solvents on Ta(CO)4(dppe) are outlined below (with
IR data). All attempted recrystallizations in benzene/hexanes
or toluene/hexanes resulted in decomposition to a non-
carbonyl-containing species and dppe. By following the pro-
cedure outlined above, Ta(CO)4(dppe) was obtained relatively
pure but was always contaminated with small amounts of dppe
(from decomposition) and some organic species. Further
purification was not possible.
Ta H(CO)4(d p p e). TaH(CO)4(dppe) was synthesized by
variations of the method described by Rehder et al.20
Solid Ta(CO)4(dppe) may be kept at 243 K for about 3 weeks.
After this time, the complex shows signs of significant decom-
position, as indicated by IR spectroscopy. Ta(CO)4(dppe) was,
therefore, only prepared as required and was used within 2
weeks of synthesis.
Meth od 1. A column of silica gel (Kieselgel 60, Merck) on
a glass frit (15 cm × 2.5 cm) was placed under dynamic
vacuum for 1 to 2 h and then filled with first argon and then
with acetonitrile. A saturated solution of [Et4N][Ta(CO)4-
(dppe)] (2.10 g, 2.56 mmol) in acetonitrile was then passed
slowly through the column, and the colored solution which
eluted was collected and pumped dry to yield a sticky red solid.
Scraping the solid vigorously in the presence of pentane
allowed isolation of TaH(CO)4(dppe) as a powdery, pentane-
insoluble, reddish-brown solid. Yield: 1.20 g (68%).
Meth od 2. Method 2 for the synthesis of [Et4N][Ta(CO)4-
(dppe)] was followed through to warming to room temperature
overnight under CO. Water (100 mL) was then added slowly
to the solution, which was stirred at room temperature for 1
h. The resulting red solution was separated from a brown,
tarry precipitate by filtering through a column of silica gel (∼4
cm × 6 cm diameter). The column was thoroughly eluted with
DME, and most of the solvent was removed under reduced
pressure. More water (100 mL) was added, the precipitated
solid was filtered out and pumped dry, and the reddish solid
was chromatographed on silica gel (10-12 cm × 2 cm
diameter) using hexanes, toluene, THF, and acetonitrile in that
order. The initial colorless solution was discarded; all of the
colored fraction was collected and pumped dry to yield red-
brown, solid product.
The solid was extracted into acetonitrile (3 × 10 mL) and
filtered to remove a quantity of yellow-beige solid. The deep-
red extracts were pumped dry, and naphthalene was removed
by sublimation. The acetonitrile extractions were repeated to
remove more beige solid (3 × 5 mL), and the sticky red solid
was stirred vigorously under pentane to obtain TaH(CO)4-
(dppe) as a red-brown powder in 20-25% yield (1.35-1.70 g).
IR data in various solvents are given in Table 2. Raman
(powder): νCO 1986 (m), 1890 (m), 1862 (vs), 1694 cm-1 (w).
1H NMR in C6D6: δ 7.51 (m, 8H, phenyl H), 6.98 (m, 12H,
phenyl H), 2.12 (m, J ) 18.2 Hz, 4H, CH2 of dppe), -2.61 (t,
J ) 21.5 Hz, 1H, TaH). 1H NMR in CD3CN: δ 7.43 (m, 12H,
phenyl H), 7.23 (m, 8H, phenyl H), 2.67 (m, J ) 18.4 Hz, 4H,
CH2 of dppe), -3.51 (t, J ) 19.9 Hz, 1H, TaH). Lit.11c in THF-
d8: δ -3.35 (t, 1H, J ) 21 Hz). 31P{1H} NMR in C6D6: δ 45.5
(s). 31P{1H} NMR in CD3CN: δ 43.9 (s). Lit.20a in THF/THF-
d8: δ 43.3 (s). 31P NMR of solid: δ 64.1 (s, br), 44.3 (s, br),
27.2 (s, br). 13C NMR of solid: δ 129.3 (s), 23.3 (s).
IR spectra were recorded in several solvents, and data are
listed in Table 3. (Subtraction of benzene from the IR spectra
is often a problem because of a band in the IR spectrum of
benzene at ∼1960 cm-1; this coincides with one of the bands
of Ta(CO)4dppe and sometimes results in abnormal intensities
for the 1960 cm-1 band of Ta(CO)4(dppe).) In toluene and
benzene, Ta(CO)4(dppe) was observed to decompose slowly
(over a few hours) at/below 285 K. Above 285 K, the complex
decomposed in about 1 h to a material which did not exhibit
any carbonyl bands in the IR spectrum. In CH2Cl2, the
complex decomposed faster, but it could be observed for at least
15 mins below 285 K; carbonyl bands attributable to TaH(CO)4-
(dppe) and TaCl(CO)4(dppe) were observed to grow in. IR
spectra of Ta(CO)4(dppe) could not be obtained in more polar
solvents, although samples were taken immediately upon
dissolving the solid. In diethyl ether, no carbonyl stretching
bands were observed. The sample was pumped dry and
redissolved in toluene. No carbonyl bands were observed, as
the complex had completely decomposed to a non-carbonyl-
containing species. In THF and acetonitrile, carbonyl bands
for TaH(CO)4(dppe) were observed in solution. Ta(CO)4(dppe)
apparently reacts with these solvents or with adventitious
water.21
An IR spectrum was obtained of a toluene solution of
Ta(CO)4(dppe) at ∼243 K, and carbonyl bands at 1962 (m),
1891 (s), 1878 (vs), and 1810 (w, br) cm-1 were observed. NMR
spectra of Ta(CO)4(dppe) (saturated solution in C6D6) were
recorded (∼2000 scans) over the range δ 1100 to -1100. The
spectrum was completely featureless except for two very broad
resonances in the regions δ 0.5-3.5 and δ 6.7-8.2; lowering
the temperature to 213 K had little effect on band widths.
A sample of Ta(CO)4(dppe) was prepared in toluene-d8, and
31P NMR spectra were collected over the temperature range
298-190 K. As the temperature decreased, a very broad (∼75
ppm) resonance centered at ∼δ 48 appeared at ∼233 K. The
solid-state MAS 31P NMR spectrum exhibited resonances at δ
85.0 (s, br), 60.2 (s, br), 33.6 (s, br), and 15.7 (s, br).
Ma gn etic Su scep tibility. Evans’ NMR method16 for the
determination of the paramagnetism of a compound in solution
(20) (a) Rehder, D.; Fornalczy, M.; Oltmanns, P. J . Organomet.
Chem. 1987, 331, 207. (b) Puttfarcken, U.; Rehder, D. J . Organomet.
Chem. 1978, 157, 321.
(21) Blaine, C. A.; Ellis, J . E.; Mann, K. R. Inorg. Chem. 1995, 34,
1552.
+
+

3294 Organometallics, Vol. 15, No. 15, 1996
Koeslag et al.
obtained at room temperature,13b,23 and using toluene
as solvent for the TaCl5, yields were improved to ∼26%.
Additional changes made to optimize the yield are
described in the Experimental Section. In all cases, the
reactions were carried out using glass-covered stir bars
and stainless steel cannulas because alkali metal naph-
thalenides reportedly react with Teflon.19
Although Calderazzo et al.11c reported a one-pot
synthesis of TaH(CO)4(dppe), involving the slow addi-
tion of HCl to a toluene solution of [Et4N][Ta(CO)6] and
dppe at 195 K, we were unable to duplicate their
method. Using this approach, the only products we
observed by IR spectroscopy were TaCl(CO)4(dppe) and
TaCl(CO)2(dppe)2, although much of the starting mate-
rial decomposed to non-carbonyl-containing complexes.
In order to attempt to avoid formation of the chloro
species, other proton acids were also tried, but the
hydride was not obtained; decomposition was usually
observed.
A previously reported two-step approach was there-
fore chosen. In this method, the dppe-substituted anion
was synthesized first, followed by protonation to form
the desired hydride. Photolysis of [Et4N][Ta(CO)6] in
the presence of dppe was found to proceed in ∼50%
yield, as described by Davison and Ellis.10d Rehder et
al.20a reported obtaining yields of 84% for this complex
(over a shorter period of time) but apparently with
different equipment.10d
Although Davison and Ellis10d report that no mono-
substituted anion was observed during photolysis of the
niobium or tantalum hexacarbonylates in the presence
of dppe, we did observe an apparent monosubstituted
intermediate by IR spectroscopy. With carbonyl bands
at 1968 (m) and 1822 (vs) cm-1 in THF, the spectrum
of the species compared favorably with that of
[Et4N][Ta(CO)5PPh3] (νCO (THF) 1973 (s), 1863 (m) and
1828 (vs) cm-1).10d The third band (∼1860 cm-1) would
be obscured by the major carbonyl band for the hexa-
carbonylate (1859 cm-1 in THF).
In several reports, TaCl5 has been reduced in the
presence of phosphine ligands, giving substituted com-
plexes which can then be reductively carbonylated.13b
In addition, Rehder et al. reported the synthesis of NbCl-
(CO)3(bdpm) (bdpm ) MeP(CH2CH2CH2PMe2)2) from
the reaction of NbCl4(THF)2 with bdpm and CO under
reducing conditions.24a Ellis et al.24b report that PMe3
was required for the reductive carbonylation of MCl4‚THF
(M ) Zr, Hf) with sodium naphthalenide. Apparently
the phosphine stabilized an intermediate species, al-
though the phosphine was not present in the isolated
complex.24b However, the substituted complexes M(CO)4-
(trmpe) (M ) Ti, Zr, Hf; trmpe ) MeC(CH2PMe2)3) were
isolated from a similar reductive carbonylation,24c and
this procedure was therefore attempted with dppe and
TaCl5.
was used to characterize Ta(CO)4(dppe). In all experiments,
Ta(CO)4(dppe) was prepared by reaction of TaH(CO)4(dppe)
with a deficiency of (t-BuC6H4)3C•. After 15 min, the solvent
was removed under reduced pressure and the solid residue
was dissolved in C6D6 (or toluene-d8) containing mesitylene
and cyclohexane (1-5% v/v each) as susceptibility shift
resonances. Portions of this solution were placed in two 5 mm
NMR tubes, one of which was used to collect a reference
spectrum. A sample of the solvent mixture was placed in a 3
mm tube to a depth of 3 cm, and this tube was then lowered
into the second 5 mm NMR tube containing TaH(CO)4(dppe)
sufficiently carefully that no mixing of the two solutions
occurred. Spectra were collected at room temperature and at
temperatures in the range 213-298 K (toluene-d8), concentra-
tions being calculated by making solvent density corrections.16d
No precipitates were observed in the samples studied at low
temperature.
Using a J ohnson Matthey magnetic susceptibility balance,
the solid Ta(CO)4(dppe) was found to be diamagnetic (øg ) -3.1
× 10-6 cm3 g-1).
EP R Sp ectr oscop y. No EPR signal was observed for
either a solid sample or a toluene solution of Ta(CO)4(dppe)
at 11 or 223 K. To ensure that the solution sample had not
decomposed, an NMR sample of the material was subsequently
prepared and shown to be paramagnetic using the Evans
method (µ ≈ 1.6 µB).
Rea ction s of Ta n ta lu m Com p lexes w ith Or ga n ic Ha -
lid es. The compounds [Et4N][Ta(CO)4(dppe)] (at 295 K),
Ta(CO)4(dppe) (below 285 K), and TaH(CO)4(dppe) (at 295 K)
(10-15 mg in 3-5 mL of toluene or 1-1.5 mL of toluene-d8)
were reacted with allyl bromide, benzyl bromide, methyl
iodide, ethyl iodide, isopropyl iodide, tert-butyl iodide, CCl4,
and (t-BuC6H4)3CBr (2-10-fold molar excess), the reactions
1
being monitored by IR and/or H NMR spectroscopy. Spectra
of reaction mixtures were generally collected within 5 min,
and in some cases, reactions were monitored for up to 1 h.
Resu lts a n d Discu ssion
Syn th eses of [Ta (CO)6]-, [Ta (CO)4(d p p e)]-, a n d
Ta H(CO)4(d p p e). Two general procedures have been
reported for the synthesis of [Ta(CO)6]-. In the first,
using magnesium and zinc in pyridine (eq 2), the
reaction must be heated to 85 °C under a pressure of
110 atm of carbon monoxide in order to achieve a
reasonable rate of reaction, and a yield of ∼35% of
[Ta(CO)6]- as a solvated sodium salt is obtained.10a In
the second, using alkali metal naphthalenides as reduc-
ing agents (eq 4), high pressures of carbon monoxide
are not required, and yields as high as 53% have been
reported.10b,e
The lithium naphthalenide reduction procedure was
therefore chosen for this work, and yields of ∼15% were
obtained. Although these yields do not correspond to
the reported yield, our results are in agreement with
those of Rehder et al.,22 who obtained yields of 15%
using lithium naphthalenide. Following a later report
by Ellis et al.19 for the synthesis of [M(CO)6]2- (M ) Ti,
Zr, Hf), we used sodium naphthalenide. Yields were
improved, albeit only to ∼20%, by maintaining the
reactions under a blanket of CO rather than by bubbling
CO through the reaction mixture.
Substitution and carbonylation of TaCl5 did, in fact,
occur in the presence of dppe, and the yields of the
substituted anion by this method were 30%, a definite
(23) Cotton, F. A.; Diebold, M. P.; Roth, W. J . Polyhedron 1985, 4,
1103. (b) Morancais, J .-L.; Hubert-Pfalzgraph, L. G. Transition Met.
Chem. 1984, 9, 130.
(24) (a) Felten, C.; Richter, E.; Priebsch, W.; Rehder, D. Chem. Ber.
1989, 122, 1617. (b) Ellis, J . E.; Chi, K. M.; DiMaio, A.-J .; Frerichs, S.
R.; Stenzel, J . R.; Rheingold, A. L.; Haggerty, B. S. Angew. Chem., Int.
Ed. Engl. 1991, 30, 194. (c) Blackburn, D. W.; Chi, K. M.; Frerichs, S.
R.; Tinkham, M. L.; Ellis, J . E. Angew. Chem., Int. Ed. Engl. 1988,
27, 437.
One of the difficulties with this preparation is that
TaCl5 forms an unidentified byproduct if it is added too
quickly to DME (even at -78 °C).10b However, solutions
of TaCl5 in toluene (or benzene) have apparently been
(22) Fornalczyk, M.; Su¨ssmilch, F.; Priebsch, W.; Rehder, D. J .
Organomet. Chem. 1992, 426, 159.
+
+

Ta(CO)4(Ph2PCH2CH2PPh2)
Organometallics, Vol. 15, No. 15, 1996 3295
Ta ble 1. IR Da ta for [Et4N][Ta (CO)4(d p p e)]
Ta ble 2. IR Da ta for Ta H(CO)4(d p p e)
solvent
νCO (cm-1
)
solvent/medium
Nujol
νCO (cm-1
)
acetonitrile 1904 (s), 1786 (vs), 1755 (m, sh)
CH2Cl2
THF
1987 (m), 1896 (vs), 1879 (vs), 1844 (vs); lit.
1990 (m), 1865 (vs, br)20a
2008 (m), 2001 (m), 1929 (m), 1906 (m),
1870 (vs)
1903 (m), 1790 (vs), 1772 (s, sh), 1742 (m, br)
1906 (m), 1799 (vs), 1779 (m), 1750 (m); lit. (THF)
hexanes
1908 (s), 1800 (vs), 1779 (s), 1752 (m)10d
toluene
1893 (m), 1789 (m), 1756 (m), 1735 (vs); lit. (Nujol)
1893 (s), 1778 (s), 1755 (s), 1736 (vs)10d
benzene, toluene 2004 (m), 1997 (m), 1914 (s), 1894 (s),
1866 (vs); lit. (toluene) 1994 (m), 1911 (sh),
1892 (s), 1863 (vs)11c
CH2Cl2
2002 (m), 1998 (m), 1907 (m), 1885 (s),
1864 (vs)
2004 (m), 1999 (m), 1917 (s), 1898 (s),
1870 (vs)
improvement over the previous results. It should be
noted, however, that all manipulations had to be carried
out at or below room temperature during the workup
of this reaction because [Et4N][Ta(CO)4(dppe)] is ex-
tremely thermally labile. As well, [Et4N]Br cannot be
added as an aqueous solution when the substituted
anionic complex is the desired product because the latter
reacts with water to form the hydride. However, while
mixtures of the anions [Ta(CO)6]-, [Ta(CO)5(dppe)]-,
and [Ta(CO)4(dppe)]- were occasionally obtained, pho-
tolysis of the recovered solids (after sublimation of
naphthalene) always resulted in overall conversion to
the desired [Ta(CO)4(dppe)]-. In most cases [Ta-
(CO)4(dppe)]- was essentially the only product. IR
spectral data for [Et4N][Ta(CO)4(dppe)] in various media
are listed in Table 1. As can be seen, the carbonyl
stretching bands are generally very broad and the
positions of the bands vary significantly with the
solvent.
ethyl ether
THF, acetonitrile 2001 (m), 1996 (m), 1906 (m, sh), 1889 (s, sh),
1869 (vs, br); lit. (THF) 1998 (m), 1860
(vs, br)11c
Ta ble 3. IR Da ta for Ta (CO)4(d p p e)
solvent
νCO (cm-1
)
Nujol
1958 (m), 1870 (vs, br), 1830 (m, sh), 1793 (m)
1960 (w), 1892 (s), 1877 (vs)
1962 (m), 1891 (s), 1878 (vs)
benzene
toluene
CH2Cl2
1962 (w), 1870 (vs, br)
strongly solvent dependent. Moreover, the spectra
obtained are different from some reported in the litera-
ture (Table 2). However, once we had fully assessed the
chemical behavior of the hydride in the various solvents
and had learned how to purify and work with it,
completely consistent spectra of this compound were
obtained over several years. Moreover, the spectra of
all possible carbonyl-containing impurities are also now
recognizable, and we have confidence in the accuracy
of the data in Table 2.
Syn th esis of Ta (CO)4(d p p e) via Hyd r id e Atom
Abstr a ction Usin g (t-Bu C6H4)3C•. Eisenberg et al.15c
reported the generation of (t-BuC6H4)3C• from (t-
BuC6H4)3CBr in a variety of solvents, using several
metal reducing reagents and at several temperatures.
In order to obtain the best possible conversion of
(t-BuC6H4)3CBr to Ta(CO)4(dppe)], a variety of reaction
conditions were assessed, the processes being monitored
by UV-vis spectroscopy (for the radical, λmax ) 524 nm,
ꢀ ) 825 M-1 cm-1).15c Note that sensitivity of the radical
to room light necessitated all manipulations being
carefully carried out in the absence of light. Although
Eisenberg et al.15c reported the successful manipulation
of the radical in THF, we found that solutions in THF
decompose rapidly. In contrast, solutions of the radical
benzene or toluene are reasonably stable for a few
minutes, during manipulations and on exposure to
adventitious light.
An instantaneous color change from red to dark
brown was noted when a solution of TaH(CO)4(dppe) in
toluene or benzene at room temperature was treated
with a solution of (t-BuC6H4)3C•; the 17-electron com-
pound Ta(CO)4(dppe) was formed. Monitoring by IR
spectroscopy showed that the reaction was complete by
the time a spectrum could be collected, although the
reaction solution was stirred for a further 10 min in the
light in order to decompose any unreacted (t-BuC6H4)3C•.
The product was isolated by removing the solvent under
reduced pressure followed by extraction of organic
(trityl-containing) byproducts with hexanes.
The method of Rehder et al.20 was chosen for the
synthesis of the hydride from [Et4N][Ta(CO)4(dppe)]. By
eluting the substituted anion through silica gel, the
authors reported yields of the hydride of 84%;20a we have
obtained 68%. Other methods of protonation of
[Ta(CO)4(dppe)]- have been reported,6d but use of a
column of silica gel generally results in the synthesis
of the hydride free of all reagents used for its synthesis
except dppe; some dppe is always found in the eluted
solution. Using the original three-step synthesis, the
overall yield of the hydride was 5-9%. By synthesis of
[Et4N][Ta(CO)4(dppe)] directly from TaCl5, overall yields
were three to four times higher (∼20%).
In an attempt to further improve yields of the hydride,
a reaction mixture produced by carbonylation of TaCl5
in the presence of dppe was eluted directly through
silica gel; the yield of hydride was now 20-25%. For
the synthesis of substituted hydride complexes, this
method represents a significant decrease in the number
of manipulations and in the time required. However,
there was little improvement in yield when Na[Ta(CO)4-
(dppe)] was chromatographed directly as compared to
purifying [Et4N][Ta(CO)4(dppe)] first. However, it is
easier to purify the hydride than the anionic complex
because the hydride is, perhaps surprisingly, less sensi-
tive to temperature.
Infrared spectra were collected in a variety of sol-
vents, and the carbonyl stretching frequencies are
reported in Table 2. The hydride presumably assumes
a distorted pentagonal bipyramid structure, as observed
for the analogous vanadium hydride,25 but is probably
stereochemically nonrigid. Indeed, observation of five
carbonyl stretching bands in most solvents suggest that
more than one geometric isomer exists in solution. As
with [Et4N][Ta(CO)4(dppe)], the carbonyl stretching
bands are generally broad and the frequencies are
Ta(CO)4(dppe) was found to be thermally very un-
stable in a variety of solvents. Thus decomposition
occurred during all attempts to recrystallize Ta(CO)4-
(dppe), and uncoordinated dppe was found to contami-
(25) Greiser, T.; Puttfarcken, U.; Rehder, D. Transition Met. Chem.
1979, 4, 168.
+
+

3296 Organometallics, Vol. 15, No. 15, 1996
Koeslag et al.
Ta ble 4. Com p a r a tive IR a n d Ra m a n Da ta for Ta (CO)4(d p p e) a n d Rela ted Com p lexes
conditions
IR (toluene)
complex
νCO (cm-1
)
[Et4N][Ta(CO)4(dppe)]
TaH(CO)4(dppe)
Ta(CO)4(dppe)
1893 (m), 1789 (m), 1756 (m), 1735 (vs)
2004 (m), 1997 (m), 1914 (s), 1894 (s), 1866 (vs)
1962 (m), 1891 (s), 1878 (vs)
TaBr(CO)4(dppe)11c
W(CO)4(dppe)
2024 (s), 1950 (m), 1903 (vs), 1881 (s)11c,32
2017 (m), 1917 (s), 1904 (vs), 1887 (s)
1962 (m), 1891 (s), 1878 (vs), 1810 (w, br)
1987 (m), 1896 (vs), 1879 (vs), 1844 (vs)
1988 (m), 1890 (vs), 1867 (vs, sh), 1856 (vs)
1958 (m), 1870 (vs, br), 1830 (m, sh), 1793 (m)
1985 (s), 1889 (s), 1870 (vs), 1849 (vs)
1893 (vs), 1790 (w), 1779 (w), 1742 (m)
1986 (m), 1890 (m), 1862 (vs), 1694 (w)
2013 (s), 1901 (vs), 1854 (vs)
IR (toluene, 243 K)
IR (Nujol mull)
Ta(CO)4(dppe)
TaH(CO)4(dppe)
HV(CO)4(dppe)6d
Ta(CO)4(dppe)
V(CO)4(dppe)6d
Raman (powder)
Raman (Nujol)
[Et4N][Ta(CO)4(dppe)]
TaH(CO)4(dppe)
W(CO)4(dppe)
Ta(CO)4(dppe)
1960 (vs), 1881 (s), 1823 (m), 1812 (m, sh), 1799 (w, sh)
nate all solid materials obtained. Solid Ta(CO)4(dppe)
was even found to decompose at 243 K over a few weeks
under a nitrogen atmosphere, and thus, it was not
possible to obtain pure material for elemental analyses.
The characterization of Ta(CO)4(dppe), while very in-
volved and complicated, was nevertheless unambigu-
ously achieved through a combination of spectroscopic
and chemical studies.
The monomeric compound Ta(CO)4(dppe) should as-
sume essentially a cis-octahedral structure (C2v), giving
rise to four carbonyl stretching bands (2a1 + b1 + b2),
all of which are IR and Raman active.26 Moreover, the
IR spectrum should exhibit a characteristic pattern of
intensities, i.e. a sharp, medium-intensity, high-fre-
quency a1 mode and, to lower frequencies, medium to
strong a1, b1, and b2 modes, possibly overlapping.26
Solution IR spectra of Ta(CO)4(dppe) (Table 3) in toluene
and benzene exhibited three carbonyl bands, a relatively
narrow band at ∼1960 cm-1 and two stronger, overlap-
ping bands at lower frequencies. Since the two lower
frequency bands are sufficiently broad that either could
obscure a fourth band, the spectra are indeed compatible
with the suggested structure; unfortunately, low solu-
bilities precluded the acquisition of a solution Raman
spectrum.
F igu r e 1. IR spectra of TaH(CO)4(dppe) (A), Ta(CO)4-
(dppe) (B), and W(CO)4(dppe) (C).
observed in addition to differences in the terminal
carbonyl region. It is thus apparent that the new
tantalum(0) system assumes different structures in the
solid state and in solution. A low-temperature (243 K)
solution IR spectrum of Ta(CO)4(dppe) exhibits a weak
band at 1810 cm-1 in addition to the bands observed in
the room-temperature spectrum, and we conclude that
the monomeric Ta(CO)4(dppe) is the dominant species
in solution over a wide range of temperature but that
it exists in solution in equilibrium with the dimer,
[Ta(CO)4(dppe)]2, which contains at least two bridging
carbonyl groups. The compound assumes only the
carbonyl-bridged dimeric structure in the solid state,
and thus the tantalum compound stands in strong
contrast to the vanadium system, where dimerization
does not occur.1,2,6a
Consistent with these results, the tantalum(0) system
is EPR silent both in the solid state, where it probably
exists as a metal-metal-bonded diamagnetic dimer, and
in solution, where it apparently undergoes severe
broadening because of facile monomer-dimer exchange.
Similar results have been found previously for the
chromium-centered radicals, CpCr(CO)3 and Cp*Cr(CO)3,
the EPR spectra of which could only be observed on
samples doped into diamagnetic matrices.27 Unfortu-
nately, all attempts to dope Ta(CO)4(dppe) into crystals
Comparisons of the IR solution spectra of Ta(CO)4-
(dppe) with the carbonyl stretching data of other
compounds in Table 4 are very useful. The carbonyl
stretching bands of Ta(CO)4(dppe) lie to lower frequen-
cies than those of the parent hydride, the differences
being greater than is observed between HV(CO)4(dppe)
and V(CO)4(dppe).6d The carbonyl stretching bands of
Ta(CO)4(dppe) also lie to lower frequencies than those
of the tungsten analogue, W(CO)4(dppe), surprisingly
implying a greater degree of back-donation in the d5
system, but are comparable in frequency to those of the
vanadium analogue, V(CO)4(dppe).6d We note that the
IR spectrum of the putative Ta(CO)6, formed by co-
condensing tantalum atoms and CO molecules in an
argon matrix at ∼4 K, exhibits a carbonyl stretching
band (1967 cm-1 7a
which lies to significantly lower
)
frequency than does the corresponding band of solid
W(CO)6 (∼1982 cm-1).7b The spectra of TaH(CO)4-
(dppe), Ta(CO)4(dppe), and W(CO)4(dppe) are illustrated
for purposes of comparison in Figure 1.
The Nujol mull IR spectrum of Ta(CO)4(dppe) is very
different from the solution spectrum, at least two bands
in the bridging carbonyl region, 1795-1830 cm-1, being
(27) (a) Morton, J . R.; Preston, K. F.; Cooley, N. A.; Baird, M. C.;
Krusic, P. J .; McLain, S. J . J . Chem. Soc., Faraday Trans. 1 1987, 83,
3535. (b) Cooley, N. A.; Baird, M. C.; Morton, J . R.; Preston, K. F.;
LePage, Y. J . Magn. Reson. 1988, 76, 325. (c) J aeger, T. J .; Watkins,
W. C.; MacNeil, J . H.; Fortier, S.; Watson, K. A.; Hensel, K.; Baird, M.
C.; Preston, K. F.; Morton, J . R.; LePage, Y.; Charland, J .-P.; Williams,
A. J .; Ziegler, T. J . Am. Chem. Soc. 1991, 113, 542.
(26) Adams, D. M. Metal-Ligand and Related Vibrations; Edward
Arnold Publishers: London, 1967; p 100.
+
+

Ta(CO)4(Ph2PCH2CH2PPh2)
Organometallics, Vol. 15, No. 15, 1996 3297
resulted in decomposition to non-carbonyl-containing
species and the reactions were not investigated further.
Ma gn et ic Su scep t ib ilit y Mea su r em en t s on
Ta (CO)4(d p p e). The Evans NMR method16a-c was
used to determine the magnetic susceptibility of
Ta(CO)4(dppe) in solution. For NMR samples prepared
by utilizing two concentric tubes as described in the
Experimental Section, the mass susceptibility of the
solute, øg, is given by eq 10,16a-c where øo is the mass
øg ) øo + 3000∆ν/4πνocM + øo(do + ds)/c (10)
susceptibility of the pure solvent, ∆ν is the change in
frequency (or susceptibility shift, in Hz) of the reso-
nances of inert species in solution (cyclohexane, mesit-
ylene), νo is the spectrometer frequency (400 MHz), c is
the concentration of the paramagnetic substance in mol
L-1, and M is its molecular weight in g mol-1. The third
term of eq 10 is a correction term to account for the
difference in density of the pure solvent, do, and the
solution, ds; this term is generally very small for highly
paramagnetic samples and was neglected, as
usual.16a-c,29a
F igu r e 2. IR (A) and Raman (B) spectra of Ta(CO)4(dppe)
(both in Nujol).
The Evans NMR method has proven to be very useful
for the characterization of Ta(CO)4(dppe) even though
the compound cannot be purified. As outlined in the
Experimental Section, hydrogen abstraction reactions
were carried out using slight excesses of TaH(CO)4-
(dppe) so that residual resonances for the hydride could
be observed in the reference spectrum. In particular,
the presence of the high-field hydride resonances was
used for identification since it was well-separated from
any other resonances in the 1H NMR spectra. With
excess hydride compound present in solution, there
would not be any (t-BuC6H4)3C• remaining and thus any
paramagnetism must be due to tantalum species present.
As well, (t-BuC6H4)3CH was observed in the NMR
spectrum, the tertiary CH resonance being character-
istic and well shifted away from all other resonances.
Concentrations of Ta(CO)4(dppe) in the reaction mix-
tures were calculated as follows. From the reference
spectrum prepared as described in the Experimental
Section, the methyl resonances of (t-BuC6H4)3CH were
integrated relative to the total methyl resonances for
these species present (including species formed from
decomposition). This allowed the calculation of the
concentration of (t-BuC6H4)3CH from the total number
of moles of (t-BuC6H4)3CBr added. On the basis of the
stoichiometry of the reaction of eq 11, the concentration
of Ta(CO)4(dppe) was then calculated.
F igu r e 3. Suggested structure for [Ta(CO)4(dppe)]2.
of the 18-electron tungsten analogue, W(CO)4(dppe),
failed because of the low thermal stability of the
tantalum compound.
Also consistent with facile equilibration between a
diamagnetic dimer and a paramagnetic monomer, a
room-temperature 1H NMR spectrum exhibited only
extremely broadened resonances at δ 0.5-3.5 and δ
6.7-8.2, presumably corresponding to the methylene
and phenyl protons, respectively. The 31P NMR spec-
trum at 298 K exhibited no resonances except for those
of free dppe and some minor impurities, but as the
temperature was lowered, a very broad (∼75 ppm)
resonance centered at ∼δ 48 appeared below ∼233 K.
The solid-state 31P NMR spectrum exhibited resonances
of comparable chemical shifts at δ 85.0 (s, br), 60.2 (s,
br), 33.6 (s, br), and 15.7 (s, br), and these are undoubt-
edly to be attributed to the dimer, [Ta(CO)4(dppe)]2. The
multiplet pattern in the 31P spectrum in the solid state
presumably arises from dipolar interactions with the
7
181Ta nucleus (I ) /2),28 but such effects are normally
impossible to observe in solution because of the ef-
ficiency of the quadrupolar relaxation mechanism.
The Raman spectrum (Nujol) of the solid material,
illustrated in Figure 2, also exhibits two terminal and
two bridging carbonyl stretching bands, with frequen-
cies very similar to those in the IR spectrum. Thus it
seems likely that the dimer is of low symmetry and is
certainly not centrosymmetric. A reasonable structure
for [Ta(CO)4(dppe)]2, compatible with the spectroscopic
data, is shown in Figure 3.
(t-BuC6H4)3C• + TaH(CO)4(dppe) f
(t-BuC6H4)3CH + Ta(CO)4(dppe) (11)
Measurements of the susceptibility shifts in the
temperature range 213-298 K were all made in dupli-
cate, and the mass susceptibilities, corrected for changes
to the solvent density at the temperatures studied, were
calculated as in eq 10. The changes in øg as a function
Attempts to obtain solution IR spectra of Ta(CO)4-
(dppe) in other than aromatic solvents provided some
interesting observations on the reactivity of this com-
pound. The hydride, TaH(CO)4(dppe), was formed in
THF and acetonitrile, but dissolving in ethyl ether
(29) (a) Earnshaw, A. Introduction to Magnetochemistry; Academic
Press: New York, 1968. (b) Szafran, Z.; Pike, R. M.; Singh, M. M.
Microscale Inorganic Chemistry; J ohn Wiley & Sons: Toronto, 1991.
(c) Abeles, T. P.; Bos, W. G. J . Chem. Educ. 1967, 44, 438. (d) Mulay,
L. N. In Theory and Applications of Molecular Paramagnetism;
Boudreaux, E. A., Mulay, L. N., Eds.; J ohn Wiley & Sons: Toronto,
1976; Chapter 9.
(28) Harris, R. K.; Olivieri, A. C. Prog. NMR Spectrosc. 1992, 24,
435.
+
+

3298 Organometallics, Vol. 15, No. 15, 1996
Koeslag et al.
of a 63 mg sample of solid Ta(CO)4(dppe) was also
determined using a J ohnson Matthey magnetic suscep-
tibility balance; the material was found to be diamag-
netic (øg ) -3.1 × 10-8 cm3 g-1).
The solution magnetic susceptibility results are com-
pletely compatible with dimerization of Ta(CO)4(dppe)
to [Ta(CO)4(dppe)]2 at low temperatures, through metal-
metal bond formation. The results are quite also
consistent with the observation of diamagnetism for the
solid compound and with our observations that the
solution and solid state IR and Raman spectra are
different. Assuming monomer and dimer magnetic
moments of 1.73 and 0 µB, respectively, an equilibrium
constant for the monomer-dimer equilibrium was cal-
culated from the solution magnetic susceptibility data.
The theoretical molar susceptibility of a paramagnetic
species in solution at temperature T can be calculated
from eq 13, where µ0 is 1.73.29a The concentration
N(µ0)2â2
F igu r e 4. Plot of 1/øM vs T.
ørad
)
(13)
3kT
obtained from eq 11 represents the total tantalum
concentration (monomer plus dimer), and the molar
susceptibility calculated from this value represents the
total susceptibility of the sample. Since the mass
susceptibility (eq 10) and hence the molar susceptibility
varies as the inverse of the concentration, the concen-
tration of the radical can be calculated as
øtotal
[Ta(CO)4(dppe)] ) [total Ta]
(14)
ørad
and the equilibrium constant is given by
[Ta(CO)4(dppe)]2
K )
(15)
(16)
[dimer]
where
[total Ta] - [Ta(CO)4(dppe)]
[dimer] )
F igu r e 5. Plot of µ vs T.
2
of temperature were found to be completely reversible,
and the results of the two experiments are in very close
agreement. The molar susceptibilities, øM, were calcu-
lated and corrected for the diamagnetic effects of the
ligands about the metal center in the usual way.16,27 The
magnetic moments µ (µB) were determined29a,b and, for
samples measured at room temperature, were found to
be very close to the value for one unpaired electron (1.73
µB).16,29
The magnetic susceptibility of a magnetically dilute,
paramagnetic compound generally decreases as the
temperature increases, thus exhibiting Curie law
behavior29a,d as in eq 12. Here C is the Curie constant
From a plot of ln K vs 1/T (Figure 6; deviation from
linearity at the lowest temperature, where ∆ν is at a
minimum, probably reflects experimental error), ∆H,
213
∆S, and ∆G
for the dissociation of the dimer were
calculated to be 4.0 kcal mol-1, 12 cal mol-1 K-1, and
1.4 kcal mol-1, respectively. While the enthalpy of
dissociation is significantly less than those of the
dimeric compounds [CpCr(CO)3]2 and [Cp*Cr(CO)3]2,
which also dissociate significantly in solution,30 our
observation that dimerization occurs at all with
Ta(CO)4(dppe) is in strong contrast to the findings for
vanadium(0) 17-electron complexes, which exist solely
as paramagnetic monomers.1,2 However, metal-metal
bonds of the 5d metals are generally much stronger than
those of the 3d analogues.31
øM ) C/T
(12)
Rea ction s of Ta (CO)4(d p p e) w ith Or ga n ic Ha -
lid es. A reaction which is characteristic of 17-electron,
and is a characteristic of the compound.29a,d A plot of
1/øM vs T should result in a straight line, and the
magnetic moment should be independent of
temperature.29a However, a plot of 1/øM vs T (Figure
4) is a curve and a plot of µ vs T (Figure 5) is a straight
line, with the magnetic moment decreasing significantly
as the temperature decreases. The mass susceptibility
(30) Watkins, W. C.; J aeger, T.; Kidd, C. E.; Fortier, S.; Baird, M.
C.; Kiss, G.; Roper, G. C.; Hoff, C. D. J . Am. Chem. Soc. 1992, 114,
907.
(31) 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; p 111.
+
+

Ta(CO)4(Ph2PCH2CH2PPh2)
Organometallics, Vol. 15, No. 15, 1996 3299
surprising for the hydride, as transition metal hydrides
are often chlorinated by CCl4.
Reactions with the homologous series of four alkyl
iodides in toluene were rather more informative, as
Ta(CO)4(dppe) reacted very slowly with MeI (weak
carbonyl bands for TaI(CO)4(dppe)11c,32 at 2023 (m), 1978
(m), 1904 (vs), and 1880 (m) cm-1 after 45 min),
somewhat less slowly with EtI (weak carbonyl bands
for TaI(CO)4(dppe) after 15 min, medium-intensity
bands after 1 h), much faster with i-PrI (medium-
intensity carbonyl bands for TaI(CO)4(dppe) after 5
min), and very rapidly for t-BuI (conversion to TaI(CO)4-
(dppe) complete within 1 min). In some spectra, weak
bands attributable to TaI(CO)2(dppe)2 were observed at
1837 (vs) and 1764 (m, br) cm-1
.
The reactions of MeI, EtI, and i-PrI were too slow for
1H NMR spectroscopy to be useful. In all cases,
decomposition to non-carbonyl-containing materials was
significant and the carbonyl bands for the probable iodo
products are sufficiently broad that their identification
F igu r e 6. Plot of ln K vs 1/T.
1
in low concentrations is not possible. In addition, H
NMR resonances of the iodo products are all sufficiently
broad that identification of low concentrations is not
feasible. Reactions of t-BuI with TaH(CO)4(dppe) and
[Et4N][Ta(CO)4(dppe)] were very slow.
metal-centered radicals is the abstraction of halogen
atoms from alkyl halides (eq 17).1 When LnM• is present
LnM• + RX f LnMX + R•
(17)
Reaction of Ta(CO)4(dppe) with allyl bromide in
toluene was completed within 1 min, the strong carbonyl
bands of TaBr(CO)4(dppe) (2024 (s), 1950 (m), 1903 (vs),
in sufficiently high concentrations that bimolecular
coupling reactions of LnM• and R• become competitive,
the corresponding alkylmetal compounds LnMR may
also be formed. Then the net overall reaction is
and 1881 (s) cm-1 11c,32
)
being observed. In addition,
weaker carbonyl bands of TaBr(CO)2(dppe)2 (1838 (vs)
and 1761 (m) cm-1 32
were also present. Interestingly,
)
monitoring of the reaction by 1H NMR spectroscopy
revealed both the resonances of TaBr(CO)4(dppe)32 and
a set of resonances which seem attributable to the new
compound η1-C3H5Ta(CO)4(dppe) (1H NMR in toluene-
d8: δ 3.43 (d, J ) 6.6 Hz), 4.87 (d, J ) 10 Hz), 5.02 (d,
J ) 17 Hz), 5.83 (m)). No apparent reaction of allyl
bromide with TaH(CO)4(dppe) or [Et4N][Ta(CO)4(dppe)]
was observed after 5 min, although very weak bands
corresponding to TaBr(CO)4(dppe) could be observed
after 35 min.
2LnM• + RX f LnMX + LnMR
(18)
In such cases, the rates of reaction increase as the
strength of the R-X bond decreases, i.e. I > Br > Cl,
t-Bu > i-Pr > Et > Me; benzylic and allylic halides also
react relatively quickly.
In the case of Ta(CO)4(dppe), reactions as in eq 17
would yield as primary products the halo species
Ta(CO)4(dppe)X, which have been reported;11c in the
presence of free dppe, which is always present in
solutions of Ta(CO)4(dppe), conversion to Ta(CO)2-
(dppe)2X occurs slowly. Well-characterized organo-
tantalum compounds of the type Ta(CO)4(dppe)R have
not been reported, although compounds of the type
Ta(CO)2(dppe)2R are known.13b
The reaction of Ta(CO)4(dppe) with benzyl bromide
in toluene was complete within ∼1 min, and again TaBr-
(CO)4(dppe) and TaBr(CO)2(dppe)2 were observed
(IR).11c,32 In an NMR experiment, weak resonances of
the bromo complex were observed in the phenyl region;
the resonance at δ 6.80 is well separated from the rest
of the phenyl resonances in the spectrum32 and was used
to identify the bromo complex in the spectrum (because
of the breadth of the CH2 resonances, they are difficult
to observe when only a small quantity of the complex
is present).32 As well, a new singlet was observed at δ
4.05 (in toluene-d8); this singlet is downfield of the
singlet for benzyl bromide at δ 3.95 and is tentatively
attributed to the benzyltantalum complex PhCH2Ta-
(CO)4(dppe).
Accordingly solutions of Ta(CO)4(dppe) (toluene for IR,
1
C6D6 or toluene-d8 for NMR studies) were reacted with
carbon tetrachloride, allyl, benzyl, and tris(tert-butyl-
phenyl) bromide, and methyl, ethyl, isopropyl, and tert-
butyl iodide. For purposes of comparison and to ensure
that any reactions observed were not a result of adven-
titious TaH(CO)4(dppe) or [Et4N][Ta(CO)4(dppe)], these
complexes were also treated with the same organic
halides. It should be noted that the latter is sufficiently
1
soluble in benzene and toluene that useful IR and H
The compounds η1-C3H5Ta(CO)4(dppe) and PhCH2Ta-
(CO)4(dppe) were apparently being observed in NMR
tube reactions, where the compounds were formed and
kept in the dark, but not in IR reactions, where the
samples were exposed to fluorescent lighting. To check
the possibility that light-sensitive organotantalum com-
pounds were being prepared, as in eq 18, we reacted
[Et4N][Ta(CO)4(dppe)] with benzyl bromide in THF.
Under fluorescent lighting, only very weak carbonyl
NMR spectra could be obtained in these solvents.
Reactions of CCl4 with TaH(CO)4(dppe) and Ta(CO)4-
(dppe) in toluene both proceeded very rapidly to give
TaCl(CO)4(dppe) (νCO 2024 (s), 1956 (m), 1902 (vs), and
1885 (m) cm-1) and TaCl(CO)2(dppe)2 (νCO 1837 (m) and
1762 (w) cm-1), identified by IR11c,32 and 1H NMR32
spectroscopy. This perhaps inconclusive result is not
(32) Koeslag, M. D.; Baird, M. C. Unpublished results.
+
+

3300 Organometallics, Vol. 15, No. 15, 1996
Koeslag et al.
bands were observed after 5 min and these were difficult
to identify; the sample had largely decomposed. In
contrast, a reaction run in the dark was completed
within 5 min, and an IR spectrum of the reaction
mixture exhibited bands corresponding to TaBr(CO)4-
(dppe) and a second compound. The latter, with car-
bonyl bands at 2000 (m), 1996 (m), and 1868 (vs) cm-1
,
very similar to those of TaH(CO)4(dppe), is probably the
benzyl complex, PhCH2Ta(CO)4(dppe). No effort was
made to isolate the benzyl compound in view of its
apparent sensitivty to visible light. However, perhaps
consistent with the low stabilities of the two presumed
alkyltantalum compounds, it has been reported that the
methyl analogue TaMe(CO)2(dmpe)2 contains an un-
ually long Ta-Me bond33 and thus a presumption of low
stability for σ-allyl and benzyl compounds seems war-
ranted.
Reactions of TaH(CO)4(dppe) and [Et4N][Ta(CO)4-
(dppe)] with benzyl bromide in toluene were very slow,
very weak bands corresponding to TaBr(CO)4(dppe)
being observed only after 15 min. Reaction of Ta(CO)4-
(dppe) with (t-BuC6H4)3CBr in toluene, monitored by IR
spectroscopy, occurred instantly to form TaBr(CO)4-
(dppe).
F igu r e 7. CV scan of 1 mM [Et4N][Ta(CO)4(dppe)] in
CH2Cl2 at Pt electrode, with T ) 245 K and v ) 0.5 V/s.
Arrows show progress of the scan.
The results of the reactions of Ta(CO)4(dppe) with
organic halides are clearly distinguishable from the
much slower reactions of TaH(CO)4(dppe) and, thus, do
not somehow involve the hydride as an intermediate.
However, the reaction products are as anticipated from
a metal-centered radical and support our conclusions,
based on physical properties, that Ta(CO)4(dppe) exists
as such in solution. The relative rates of the homolo-
gous series of alkyl iodides and the relatively high rates
of reaction of C3H5Br, PhCH2Br, and (t-BuC6H4)3CBr,
although the data are qualitative, are exactly as ex-
pected for processes in which a metal-centered radical
abstracts halogen atoms (eq 17). And while most of the
reactions were too slow for alkyltantalum products to
be observed, the light-sensitive products of the rapid
reactions of Ta(CO)4(dppe) with C3H5Br and PhCH2Br
are almost certainly as formulated.
Electr och em ica l Resu lts. During the course of our
investigation, a paper appeared detailing through vol-
tammetry and IR spectroelectrochemistry the reduction
of the Ta(I) complexes TaX(CO)4(dppe), X ) Br, I.21
These results, by Blaine and co-workers, are relevant
to the present work owing to the fact that the 18-
electron anion [Ta(CO)4(dppe)]- is formed upon reduc-
tion of the halide complexes (eq 19). Re-oxidation of
F igu r e 8. CV scan of 0.85 mM [Et4][Ta(CO)4(dppe)] to
more positive potentials than those of Figure 7. Condi-
tions: solvent ) CH2Cl2, T ) 249 K, Pt electrode, v ) 1
V/s.
cause our results with the last of these are in basic
agreement with the conclusions of Blaine et al., indicat-
ing formation of TaH(CO)4(dppe) as well as other
products during thin-layer electrolysis, the spectro-
electrochemical results will not be presented. Rather,
we concentrate on voltammetric data that give ad-
ditional insight into the anodic mechanism of
[Ta(CO)4(dppe)]-.
The anionic complex [Ta(CO)4(dppe)]- was studied by
cyclic voltammetry, and oxidation was found to be
chemically irreversible in THF (Epa ) -1.10 V), CH2Cl2
(Epa ) -1.10 V), and o-difluorobenzene (Epa ) -1.07 V)
(potentials quoted for v ) 0.1 V/s). We show below that
the anodic current function (Ipa/v1/2, independent of v)
(wave A, Figure 7) corresponds to passage of ∼1
electron. This feature is followed by a reversible oxida-
tion (wave B, Figure 7) of about half the Faradaic
intensity with E1/2 ) -0.48 V in CH2Cl2 and -0.49 V in
difluorobenzene. If the scan was continued to more
positive potentials, another apparently reversible oxida-
tion was observed (E1/2 ) -0.1 V, wave C, Figure 8).
Yet another (quasi-reversible) anodic wave (E, not
shown) appears at E1/2 ) +0.41 V and is of about the
same height as that of wave B. Scanning to highly
negative potentials after first scanning through wave
A gave a broad and irreversible cathodic feature (wave
D, Figure 7). The processes responsible for product
waves B-E were identified by matching their potentials
TaX(CO)4(dppe) + 2e- h [Ta(CO)4(dppe)]- + X-
(19)
[Ta(CO)4(dppe)]- in the presence of X- re-forms the
original halide complex in addition to some TaH(CO)4-
(dppe).21 In the context of our interest in the 17-electron
radical Ta(CO)4(dppe), we were meanwhile pursuing
studies of the oxidation of the 18-electron anion in order
to probe the lifetime of the radical and establish a formal
potential for the couple [Ta(CO)4(dppe)]0/-
.
We have studied the oxidation of [Et4N][Ta(CO)4(dppe)]
in CH2Cl2/0.1 M [Bu4N][PF6] by cyclic voltammetry,
bulk coulometry, and IR spectroelectrochemistry. Be-
(33) Protasiewicz, J . D.; Bianconi, P. A.; Williams, I. D.; Liu, S.; Rao,
C. P.; Lippard, S. J . Inorg. Chem. 1992, 31, 4134.
+
+

Ta(CO)4(Ph2PCH2CH2PPh2)
Organometallics, Vol. 15, No. 15, 1996 3301
2[Ta(CO)4(dppe)]- f 2e- + TaH(CO)4(dppe) +
1/2TaCl(CO)2(dppe)2 + 1/2[Ta(CO)6]- (24)
Ta ble 5. Electr och em ica l Da ta of Ta n ta lu m
Com p lexes w ith P oten tia ls in V vs F er r ocen e a n d
Su p p or tin g Electr olyte 0.1 M [Bu 4N][P F 6]
complex
solvent process variable potential
refs
[Ta(CO)6]-
CH2Cl2 oxdn
Epa
Epa
Epa
E1/2
Epa
Epa
E1/2
E1/2
E1/2
E1/2
∼-0.5
this work
-1.10 this work
We must also consider the possibility that Ta(CO)4-
(dppe) reacts by simple ionic diproportionation (eq 25)
[Ta(CO)4(dppe)]- CH2Cl2 oxdn
[Ta(CO)4(dppe)]- THF
TaH(CO)4(dppe) CH2Cl2 0/+
TaBr(CO)4(dppe) CH2Cl2 oxdn
oxdn
-1.16 21
-0.09 this work
+0.41 21
2Ta(CO)4(dppe) h [Ta(CO)4(dppe)]+ +
TaI(CO)4(dppe)
CH2Cl2 oxdn
+0.39 21
[Ta(CO)4(dppe)]- (25)
TaCl(CO)2(dppe)2 CH2Cl2 0/+
-0.44 this work
+0.42 this work
-0.49 this work
+0.32 this work
TaCl(CO)2(dppe)2 CH2Cl2 +/2+
TaCl(CO)2(dppe)2 DFB
TaCl(CO)2(dppe)2 DFB
0/+
+/2+
giving the 16-electron species [Ta(CO)4(dppe)]+, which
could form the hydride TaH(CO)4(dppe) and, perhaps,
TaCl(CO)4(dppe) through hydride and halide abstrac-
tion, respectively, from the solvent. We cannot rule out
TaCl(CO)4(dppe) as a possible product since it is ex-
pected21 to have an oxidation wave in the vicinity of +0.4
V (near wave E); ambiguity arises because TaCl(CO)2-
(dppe)2 has a wave at the same potential arising from
its oxidation to a dication.
The strongest evidence against eq 25 being an im-
portant part of the mechanism comes from the observed
electron stoichiometry. Note that the simple dispro-
portionation of eq 25 returns 1 equiv of the starting
anion to the reaction sequence, so that a net two
electrons/equiv of [Ta(CO)4(dppe)]- is required when eqs
20 and 25 are combined, in contrast to the one-electron
stoichiometry predicted from eq 24.
with those of independently prepared compounds sus-
pected to be reaction products. Relevant potentials are
collected in Table 5.
Figure 7 suggests that the oxidation of [Ta(CO)4-
(dppe)]- gives rise to a mixture of electroactive products
with oxidations at E1/2 ) -0.48 and -0.1 V (waves B
and C, respectively). These potentials are matched by
those of TaCl(CO)2(dppe)2 (E1/2 ) -0.44 V) for wave B
and the expected hydride product TaH(CO)4(dppe) (E1/2
) -0.09 V)34 for wave C. The presence of TaCl(CO)2-
(dppe)2 is confirmed by the anodic wave E (+0.42 V
measured for TaCl(CO)2(dppe)2+/2+, +0.41 V in Figure
7), which corresponds to the second one-electron oxida-
tion of the chloro complex.
The actual electron stoichiometry was measured by
two methods, (a) bulk coulometry and (b) comparison
of voltammetric peak heights of [Ta(CO)4(dppe)]- with
a suitable standard. Exhaustive electrolysis was per-
formed a number of times at various temperatures, and
The mechanism in eqs 20-23 accounts for the ob-
served product waves. In this scenario, once the 17-
electron radical is formed (eq 20) it is subject to two
[Ta(CO)4(dppe)]- h e- + Ta(CO)4(dppe) (20)
Ta(CO)4(dppe) + SH f TaH(CO)4(dppe) (21)
a
net stoichiometry of 0.7-0.9 electrons/mol of
[Ta(CO)4(dppe)]- was obtained on each occasion. We
take this to be consistent with an overall napp ) 1
electron process (eq 24).
For approach (b), it is important to employ a standard
Ta(CO)4(dppe) h 1/2[Ta(CO)2(dppe)2]+ +
with
a
diffusion coefficient similar to that of
[Ta(CO)4(dppe)]-. With this in mind we used
[CpMo(CO)3]- (Epa ∼ -0.55 V vs Fc) owing to its similar
size and identical charge. This anion is reported to give
a one-electron oxidation in CH2Cl2.35,36 The current
functions of [CpMo(CO)3]- and [Ta(CO)4(dppe)]- in
CH2Cl2 were virtually identical at concentrations of
either 0.3 or 0.7 mM, leading us to again conclude that
the primary oxidation wave of [Ta(CO)4(dppe)]- (wave
A) is a one-electron process.
1/2[Ta(CO)6]- (22)
1/2[Ta(CO)2(dppe)2]+ + CH2Cl2 f
1/2TaCl(CO)2(dppe)2 + CH2Cl‚ (?) (23)
major pathways for subsequent reactions, namely hy-
drogen atom abstraction from the solvent (SH) or the
electrolyte cation [eq 21, giving TaH(CO)4(dppe), wave
C] and ionic disproportionation with ligand exchange
(eq 22), producing the 16-electron species [Ta(CO)2-
(dppe)2]+, which then abstracts halide from the solvent
(eq 23) or, possibly, chloride-containing, adventitious
impurities. The chloride abstraction reaction gives
TaCl(CO)2(dppe)2, the oxidation of which is responsible
for waves B and E. A net transfer of one electron/equiv
of [Ta(CO)4(dppe)]- is predicted by the overall reaction
(eq 24). No direct voltammetric evidence for [Ta(CO)6]-
was obtained, not surprising owing to its poor electro-
chemical behavior (vide infra). The net overall reaction
for tantalum-containing species is then
The species giving rise to the small and broad reduc-
tion wave D in Figure 7 is uncertain. Both TaH(CO)4-
(dppe) and TaX(CO)4(dppe) have irreversible waves in
this region, and other possibilities exist as well.
One fact deserving mention is that wave B comes at
virtually the same potential in both CH2Cl2 and o-
difluorobenzene. At first sight this appears to argue
against assignment of wave B as TaCl(CO)2(dppe)2 in
the former [and TaF(CO)2(dppe)2 in the latter]. The
waves for the two possible halide products may, how-
ever, be coincidental, and it is known that the identity
of X has very little effect on the potentials of compounds
of the types TaX(CO)4(dppe) or TaX(CO)2(dppe)2.21,37
(35) Kadish, K.; Lacombe, D. A.; Anderson, J . E. Inorg. Chem. 1986,
25, 2246.
(36) Tilset, M. Inorg. Chem. 1994, 33, 3121.
(37) Bond, A. M.; Bixler, J . W.; Mocellin, E.; Datta, S.; J ames, E. J .;
Wreford, S. S. Inorg. Chem. 1980, 19, 1760.
(34) In ref 21, the potential of the oxidation of TaH(CO)4(dppe) is
given as -0.48 V vs Fc. For this system we obtain a value of -0.09 V
vs Fc ((0.01 V) in several experiments in CH2Cl2 and -0.03 V in
o-difluorobenzene.
+
+

3302 Organometallics, Vol. 15, No. 15, 1996
Koeslag et al.
The possibility also exists that wave B arises from
TaF(CO)2(dppe)2 in both solvents, with the supporting
electrolyte anion, [PF6]-, acting as the source of fluoride
ion in both cases.
Voltammetry experiments conducted at high sweep
rates (up to v ) 50 V/s at 273 K in CH2Cl2) failed to
reveal any reversibility for the oxidation of
[Ta(CO)4(dppe)]-. This observation is in contrast to the
much greater stability of Ta(CO)4(dppe) in nonionic
media, reflecting the additional reaction pathways
available to the radical in ionic solutions. Although the
chemical irreversibility of the couple [Ta(CO)4(dppe)]0/-
precludes a precise measure of its formal potential, the
E1/2 value must lie close to -1.2 V vs Fc. Reasonable
limits for this quantity are -1.2 V ( 0.1 V.
Ack n ow led gm en t. We gratefully acknowledge fi-
nancial support from the Natural Sciences and Engi-
neering Research Council (Research Grant to M.C.B.),
Alcan (scholarship to M.K.), and the the National
Science Foundation for work at the University of
Vermont (Grant NSF CHE 94-16611). We also thank
Drs. K. Preston and P. Kaiser for running EPR spectra,
and Dr. Kent Mann for helpful conversations concerning
the results in ref 21.
OM960133M
+
+