Improved synthesis of (η5-CpR)M(CO)4 compounds and the Nujol matrix photochemistry of (η5-C5H5)M(CO)4 and (η5-C9H7)M(CO)4, where M=Nb and Ta

Article abstract of DOI:10.1016/S0022-328X(97)00735-3

Improved preparative routes for cyclopentadienyl (Cp) and substituted Cp(CO)₄ derivatives of niobium and tantalum and their precursors are described. New compounds are fully characterized by IR, NMR, and elemental analysis. The molecular struct

Full text of DOI:10.1016/S0022-328X(97)00735-3

Journal of Organometallic Chemistry 557 (1998) 77–92
Improved synthesis of (p5-CpR)M(CO)4 compounds and the Nujol
matrix photochemistry of (p5-C5H5)M(CO)4 and (p5-C9H7)M(CO)4,
where M=Nb and Ta
Thomas E. Bitterwolf a,*, Skip Gallagher a, J. Timothy Bays a, Bruce Scallorn a,
Arnold L. Rheingold b, Ilia A. Guzei b, Louise Liable-Sands b, John C. Linehan c
a Department of Chemistry, Uni6ersity of Idaho, Moscow, ID 83844-2343, USA
b Department of Chemistry, Uni6ersity of Delaware, Newark, DE 19716, USA
c Pacific Northwest National Laboratory, PO Box 999, Richland, WA 99352, USA
Received 28 July 1997
Abstract
Improved preparative routes for cyclopentadienyl (Cp) and substituted Cp(CO)4 derivatives of niobium and tantalum and their
precursors are described. New compounds are fully characterized by IR, NMR, and elemental analysis. The molecular structures
˚
˚
˚
of five compounds were determined: (vinylCp)Nb(CO)4: triclinic, P¯ı, a=6.770(1) A, b=6.937(1) A, c=12.706(2) A, h=
3
˚
90.42(1)°, i=91.07(1)°, k=105.560(8)°, V=574.71(12) A , Z=2, R(F)=2.82; (acetylCp)Nb(CO)4: monoclinic, P21/c, a=
10.316(2) A, b=14.490(3) A, c=7.957(3) A, i=94.68(2)°, V=1185.4(4) A , Z=4, R(F)=2.58; (Cp)Ta(CO)4: orthorhombic,
Pnma, a=7.756(1) A, b=12.159(2) A, c=10.580(1) A, V=997.7(2) A , Z=4, R(F)=2.60; (indenyl)Ta(CO)4: monoclinic,
P21/n, a=6.602(2) A, b=12.399(1) A, c=7.4775(9) A, i=92.05(2)°, V=611.2(2) A , Z=2, R(F)=3.05; and (ben-
zoylCp)Ta(CO)4: monoclinic, P21/c, a=12.552(10) A, b=6.5247(2) A, c=18.4682(2) A, i=98.9152(4)°, V=1494.24(5) A ,
Z=4, RF)=2.83 are reported. Photolysis of CpMCO)4 and indenyl)MCO)4, where M=Nb and Ta, in Nujol glass matrices were
found to result in loss of one to three CO ligands as a function of the wavelength of the incident light. Annealing or reverse
photolysis resulted in reversal of these CO-loss reactions. © 1998 Elsevier Science S.A. All rights reserved.
3
˚
˚
˚
˚
3
˚
˚
˚
˚
3
˚
˚
˚
˚
3
˚
˚
˚
˚
Keywords: Photochemistry; Niobium; Tantalum
1. Introduction
be reported since then employs unusually high pressures
(4700 psi) and requires 135 h of reaction time [2].
Building upon elegant work by Calderazzo, Pampaloni
and their coworkers [3], we have recently reported the
synthesis of several previously inaccessible ring-substi-
tuted derivatives via Scheme I [4].
In this paper we present the synthesis of additional
ring-substituted CpNb and Ta(CO)4 derivatives by this
route, and describe the use of Na[M2(CO)8(v-I)3] as a
preferred precursor to these compounds (Scheme Ib). A
high yield synthesis of (p5-C5H5)M(CO)4 from (p5-
C5H5)MCl4 (Scheme II), employing analogous, al-
though somewhat less demanding, conditions to those
described by Herrmann is also reported. The photolysis
In contrast to the vast literature of the cyclopentadi-
enyl (Cp) metal carbonyl compounds of most of the
transition metals, the paucity of chemistry of the nio-
bium and tantalum compounds is notable. The chief
reason for the limited exploration of this class of com-
pounds has been the difficulties associated with their
preparation and in the preparation of their precursors.
Although King reported the first synthesis of (p5-
C5H5)Nb(CO)4 in 1963 [1], the only high yield route to
* Corresponding author. Tel.: +1 208 8856552; fax: +1 208
8856173; e-mail: bitterte@uidaho.edu
0022-328X/98/$19.00 © 1998 Elsevier Science S.A. All rights reserved.
PII S0022-328X(97)00735-3

78
T.E. Bitterwolf et al. / Journal of Organometallic Chemistry 557 (1998) 77–92
Scheme I. Synthesis of ring-substituted derivatives using [M(CO)6]−
.
of several compounds in Nujol glass matrices is also
presented.
low pressure reductive carbonylation of (p5-
C5H5)NbCl4 to yield (p5-C5H5)Nb(CO)3Cl2 [5]. In con-
trast, under very high pressure conditions, Herrmann
has prepared (p5-C5H5)Nb(CO)4 and its CH3Cp and
(CH3)5Cp derivatives. We speculated that intermediate
conditions with an appropriate reducing agent and
solvent might also yield tetracarbonylated derivatives.
Indeed, reaction of (p5-C5H5)NbCl4, (p5-C5H5)TaCl4,
(p5-C5H4C4H9)NbCl4, [(p5-C5H4SiCH3)3]NbCl4 and
[(p5-C5H4SiCH3)3]TaCl4 in pyridine in the presence of
Zn/Mg under CO (1600 psi) gave, after work-up and
sublimation, good yields of the desired tetracarbonyl
derivatives. Scale-up of the reactions to batches of 10 g
was readily achieved and further scale-up was only
limited by the capacity of the Parr bomb. Since pres-
sures in the range of 1600 psi may be easily achieved in
a standard high pressure reaction bomb at typical tank
pressures of CO, this preparative route should greatly
improve the availability of these compounds. The
mono-CpM(CO)4 derivatives were prepared by reaction
of MCl5 with RCpTMS and used without purification.
We have been unable to prepare the indenyl, acetyl-
or formylCp Group V compounds by this route al-
though preliminary results suggest that the ester deriva-
tives may be prepared in this way.
2. Synthesis of ring-substituted CpM(CO)4 derivatives
As
previously
reported,
reaction
of
Na[M(CO)6]·THF, where M=Nb or Ta, with cop-
per(I) chloride or MCl5 in THF under an atmosphere
of CO yields Na[M2(CO)8(v-Cl)3] which is allowed to
react without isolation with NaC5H4R or LiC5H4R.
After flash filtration of the reaction mixture and re-
moval of solvent, the resulting oily residues are sub-
limed to yield (p5-C5H4R)M(CO)4. In many cases the
sublimed products were contaminated with an organic
oil believed to be bis(Cp) compounds formed by copper
coupling of the Cp anions. Repeated sublimations
yielded analytically pure products, albeit at the expense
of yield.
Removal of the organic oils from the Group V
compounds is complicated by the instability of these
organometallic compounds to chromatographic condi-
tions. Even brief residence on either neutral alumina or
silica gel results in decomposition. In an effort to avoid
the side reactions leading to bis(Cp) compounds, we
examined the use of alternate reagents for conversion of
Na[MCO)6]·THF to the bimetallic, Na[M2(CO)8(v-X)3]
reagents. We have found that the use of a stoichiomet-
ric quantity of iodine to Na[M(CO)6]·THF results in
formation of Na[M2(CO)8(v-I)3] as shown by IR. This
reagent reacts with NaCpR or LiCpR to yield the
corresponding CpM(CO)4 derivatives. Work-up and
subsequent sublimation is not complicated by organic
impurities resulting in somewhat higher yields with
considerably less effort required in purification.
While the use of Na[M2(CO)8(v-I)3] simplifies the
purification of these compounds, the reaction yields
remain low to good. This, and the difficulties associated
with the synthesis of Na[M(CO)6]·THF, prompted us
to examine more direct routes to the desired com-
pounds. Cardoso, Clark and Moorhouse have reported
Scheme II. High yield synthesis of (p5-C5H5)M(CO)4 from (p5-
C5H5)MCl4.

T.E. Bitterwolf et al. / Journal of Organometallic Chemistry 557 (1998) 77–92
79
Table 1
Herrman and Biersack [8] have reported the prepara-
tion of 13CO and 13C18O enriched (p5-C5H5)Nb(CO)4
via (p5-C5H5)Nb(CO)3THF. Mass spectrometric and
infrared spectral studies were conducted on these mate-
rials, but there is no mention in this or subsequent
papers of 13C-NMR measurements on this enriched
compound. For natural abundance 13CO, coupling be-
tween the 9/2 spin 93Nb nucleus and the carbonyl
carbon results in a broad, ill-resolved resonance in the
metal carbonyl region, Fig. 1a. A 13CO enriched sample
was prepared by exchange between 13CO and (p5-
C5H5)Nb(CO)4 in the dark at room temperature (r.t.) in
benzene over a period of several months. IR spectra
suggest that on the average slightly more than one CO
group had exchanged. The 13C-NMR spectrum of this
sample revealed the expected ten-line carbonyl reso-
nance illustrated in Fig. 1b. The 13C–93Nb coupling
constant was found to be 222 Hz.
IR and 93Nb-NMR data for niobium and tantalum compounds
(C5H4R)M(CO)4 IR (CH2Cl2) w(CO), cm−1
NMR (C6D6)
93Nb
R
Nb
Ta
H
2034, 1918
2041, 1929
2038, 1928
2040, 1930
2040, 1930
2040, 1931
2033, 1919
2033, 1919
2033, 1921
2031, 1915
2031, 1917
2031, 1910
2037, 1921
2036, 1921
2037, 1922
2036, 1921
2036, 1921
2030, 1911
−2016
−1918
−1919
−1921
−1904
−1917
−1953
−1946
−1912
−1995
CO2Me
CO2Et
C(ꢀO)Me
C(ꢀO)C6H5
C(ꢀO)CH2C6H5
C(CH3)ꢀCH2
CHꢀCH2
Indene
2030, 1912
2028, 1908
2029, 1917
CH2(CH2)2CH3
TMS
Attempts to carry out Friedel–Crafts acylation on
both (p5-C5H5)Nb(CO)4, [p5-C5H4Si(CH3)3]Nb(CO)4,
and (p5-C5H5)Ta(CO)4 resulted in complete decomposi-
tion of the organometallic species. Lokshin et al. [6]
reported that AlCl3 appeared to complex exclusively at
the metal in (p5-C5H5)Nb(CO)4, and the metal may also
be the preferred point of reaction in acylation reactions
as well. The failure of these reactions leaves the reac-
tions described above as the only viable routes to many
ring substituted compounds in these series.
2.2. Molecular structures
The molecular structures of five compounds have
been determined. Crystallographic parameters are sum-
2.1. Spectroscopic studies
Carbonyl stretching frequencies of CpM(CO) com-
pounds are known to be sensitive to ring-substituents.
The carbonyl stretching frequencies of the Nb and Ta
series of compounds, Table 1, are found to increase
(relative to R=H) for electron withdrawing groups
and decrease for electron donating groups, although the
range of values exhibited by each series is too small to
carry out meaningful correlations with Hammett sub-
stituent constants. Comparison of the carbonyl stretch-
ing frequencies between the Nb and Ta series indicates
a high degree of correlation between the two, recogniz-
ing the limitations imposed by the small range of
values.
93Nb-NMR spectra were recorded on the compounds
of the series with C6D6 as solvent and these values are
also reported in Table 1. Comparison of the 93Nb
chemical shifts with the carbonyl stretching frequencies
indicates that there is no correlation between the chem-
ical shifts and the electron donating or withdrawing
effects of the ring substituents. In contrast, Rausch and
Graham found good correlations between Hammett
substituent parameters or carbonyl stretching frequen-
cies and the 103Rh resonances in a series of (p5-
C5H4R)Rh(CO)2 compounds [7]. The reason for the
poor correlations in the Nb compounds is not
apparent.
Fig. 1. 13C-NMR of the carbonyl region of (a) (p5-
C5H4CO2C2H5)Nb(CO)4 at natural abundance (230000 scans) and b)
(p5-C5H5)Nb(CO)3
(
13CO) showing 13C–93Nb coupling (264000
scans).

80
T.E. Bitterwolf et al. / Journal of Organometallic Chemistry 557 (1998) 77–92
Table 2
Crystallographic data for (C5H4CHꢀCH2)Nb(CO)4 (1), (indenyl)Ta(CO)4 (2), CpTa(CO)4 (3), (C5H4C(O)Ph)Ta(CO)4 (4), (C5H4C(O)CH3)
Nb(CO)4 (5)
1
2
3
4
5
Formula
C11H7NbO4
296.08
P¯ı
C
408.14
P21/m
6.602(2)
12.399(1)
7.4775(9)
13H7TaO4
C9H5TaO4
358.08
Pnma
7.756(1)
12.159(2)
10.580(1)
C16H9TaO5
462.18
P21/c
12.55200(10)
6.5347(2)
18.4682(2)
C11H7NbO5
312.08
P21/c
10.316(2)
14.490(3)
7.957(3)
Formula weight
Space group
˚
a (A)
6.770(1)
6.937(1)
12.706(2)
90.42(1)
91.07(1)
105.560(8)
574.71(12)
2
˚
b (A)
˚
c (A)
h (°)
i (°)
k (°)
93.05(2)
98.9152(4)
94.68(2)
3
˚
V (A )
611.2(2)
2
997.7(2)
4
1494.24(5)
4
1185.4(4)
4
Z
Crystal color
Orange
1.711
10.41
248(2)
Mo–Kh (u=0.71073 A)
Orange
2.219
90.00
243(2)
Orange
2.384
109.99
298(2)
Orange
2.054
73.76
Red
Calculated density (g cm−3
)
1.749
10.20
243(2)
v(Mo–Kh) (cm−1
Temperature (K)
Radiation
)
223(2)
˚
R(F)a (%)
2.82
8.36
3.05
8.50
2.60
5.94
2.83
6.74
2.58
6.37
R(wF2)a %)
a Quantity minimized=R(wF2)=S[w(Fo2−Fc2)2]/S[(wF2o)2]1/2; R=SD/S(Fo), D=ꢀ(Fo−Fc)ꢀ.
marized in Table 2, while selected bond length and
angle data are presented in Tables 3–7. The molecular
structures and numbering schemes of these compounds
are presented in Figs. 2–6. In all cases, these molecules
and those reported previously are found to possess the
expected four-legged piano stool geometry. Compari-
son of metalꢁcarbonyl and carbonyl CꢁO bond lengths
for these compounds reveals no statistically significant
differences attributable to the electronic effects of the
ring substituents.
The molecular structure of (vinylCp)Nb(CO)4 is, to
the best of our knowledge, the first half-sandwich
molecular structure of a vinylCp derivative to be re-
ported. As expected, the vinyl group lies in the plane of
the Cp ring. The bond length of the CꢀC is within
normal ranges.
also been reported in frozen gas matrices at 12 K [13].
Poliakoff and coworkers [11] have reported the photol-
ysis of (p5-C5H5)Ta(CO)4 in an argon matrix at 12 K,
as well as observation of the (p5-C5H5)M(CO)3 pho-
tointermediates by time resolved IR.
Nujol glass matrix photochemical studies in this re-
search were carried out in an apparatus designed by Dr
Antony Rest of the University of Southampton. The
apparatus and techniques for these photochemical stud-
ies have been described previously [14]. Nujol solutions
for these studies were prepared in a glove box with
minimal background light. Samples were frozen to
about 77 K and photolyzed using optical filters to
precisely define the wavelengths of light being transmit-
Table 3
˚
Selected bond lengths (A) and angles (°) for (C5H4CHꢀCH2)Nb(CO)4
2.3. Nujol glass matrix photochemical studies
Atoms
Distance
Atoms
Angle
Rehder and Bechthold [9] have prepared a series of
(p5-C5H5)Nb(CO)3(PR3) complexes by photolysis of the
parent tetracarbonyl in the presence of phosphines and
phosphites. Nesmayanov et al. [10] reported the photol-
ysis of the group V compounds with diphenylacetylene
to give (p5-C5H5)M(CO)2Ph2C2) and other products.
Poliakoff and coworkers [11] have examined the pho-
tolysis of the Group V series with H2, N2 and silanes in
both n-heptane (r.t.) and liquid xenon (−80°C) using
time resolved IR. Hitam and Rest [12] have carried out
a detailed photochemical study of (p5-C5H5)V(CO)4
and its 13CO isotopomers in frozen gas matrices at 12
K. A similar photochemical study of ring substituted
vanadium compounds including (p5-C9H7)V(CO)4 has
NbꢁC(1)
NbꢁC(2)
NbꢁC(3)
NbꢁC(4)
C(1)ꢁO(1)
C(2)ꢁO(2)
C(3)ꢁO(3)
C(4)ꢁO(4)
NbꢁC(5)
NbꢁC(6)
NbꢁC(7)
NbꢁC(8)
NbꢁC(9)
C(5)ꢁC(10)
C(10)ꢁC(11)
2.094(5)
2.082(4)
2.073(4)
2.093(4)
1.127(5)
1.137(5)
1.144(5)
1.134(5)
2.411(4)
2.378(4)
2.395(4)
2.417(4)
2.420(4)
1.486(7)
1.244(7)
C(1)ꢁNbꢁC(2)
C(1)ꢁNbꢁC(4)
C(2)ꢁNbꢁC(3)
C(3)ꢁNbꢁC(4)
C(1)ꢁNbꢁC(3)
C(2)ꢁNbꢁC(4)
C(5)ꢁC(10)ꢁC(11)
73.7(2)
74.3(2)
73.9(2)
74.01(14)
117.3(2)
115.9(2)
127.5(6)

T.E. Bitterwolf et al. / Journal of Organometallic Chemistry 557 (1998) 77–92
81
Table 4
Selected bond lengths (A) and angles (°) for (indenyl)Ta(CO)4
Table 6
˚
˚
Selected bond lengths (A) and angles °) for (C5H4C(O)Ph)Ta(CO)4
Atoms
Distance
Atoms
Angle
Atoms
Distance
Atoms
Angle
TaꢁC(1)
TaꢁC(2)
C(1)ꢁO(1)
C(2)ꢁO(2)
TaꢁC(3)
TaꢁC(4)
TaꢁC(5)
C(5)ꢁC(6)
C(6)ꢁC(7)
C(7)ꢁC(7a)
2.064(8)
2.075(7)
1.153(10)
1.142(8)
2.349(10)
2.380(7)
2.436(6)
1.419(11)
1.32(2)
C(1)ꢁTaꢁC(1a)
C(1)ꢁTaꢁC(2)
C(1a)ꢁTaꢁC(2a)
C(2)ꢁTaꢁC(2a)
C(1)ꢁTaꢁC(2a)
C(1a)ꢁTaꢁC(2)
75.8(5)
73.3(3)
73.3(3)
74.2(4)
117.0(3)
117.0(3)
TaꢁC(1)
TaꢁC(2)
TaꢁC(3)
TaꢁC(4)
TaꢁC(5)
TaꢁC(6)
TaꢁC(7)
TaꢁC(8)
2.397(5)
2.398(6)
2.427(5)
2.440(5)
2.431(5)
2.073(6)
2.093(6)
2.070(6)
2.075(5)
1.507(7)
1.221(6)
1.514(7)
C(6)ꢁTaꢁC(7)
C(6)ꢁTaꢁC(9)
C(7)ꢁTaꢁC(8)
C(8)ꢁTaꢁC(9)
C(6)ꢁTaꢁC(8)
C(7)ꢁTaꢁC(9)
C(1)ꢁC(10)ꢁC(11)
C(1)ꢁC(10)ꢁO(10)
73.5(2)
73.1(2)
73.4(2)
73.7(2)
115.4(2)
115.6(2)
120.3(5)
119.1(5)
TaꢁC(9)
1.43(3)
C(1)ꢁC(10)
C(10)ꢁO(10)
C(10)ꢁC(11)
ted to the sample. A summary of IR data for com-
pounds and photoproducts is presented in Table 8.
Preliminary observations suggested that compounds
in this series are exceptionally light sensitive, an obser-
vation that has also been made for the analogous
vanadium compounds [15]. Samples of the indenyl
compounds had to be prepared in near darkness to
avoid photolysis by incident light. In fact, IR spectra
recorded on samples of the Ta compound before and
after routine UV–vis spectral measurements indicated
detectable quantities of photoproducts were formed.
As shown in Fig. 8a and Fig. 11a, both (p5-
C5H5)Nb(CO)4 and (p5-C5H5)Ta(CO)4 have electronic
absorption bands at about 430 and 370 nm and a small
band at about 310 nm. In dilute samples, both com-
pounds have a strong band at about 250 nm. Rest et al.
assign the corresponding long wavelength bands in
(p5-C5H5)V(CO)4 to d“d and M“y*(CO) transitions,
respectively [16]. The short wavelength bands are typi-
cally assigned to metal to ligand charge transfer bands
although no detailed analysis of the bonding in these
compounds is presently available.
associated with the carbonyl groups in the starting
materials decrease, while a band attributable to ‘free’
CO (2132 cm−1) and three additional bands consistent
with a species of the formula CpM(CO)3 (Cs symmetry)
grow in. The band positions for (p5-C5H5)M(CO)3,
where M=Nb and Ta, are within a few cm−1 of those
reported by Poliakoff et al. [11] in n-heptane (r.t.) and
for (p5-C5H5)Ta(CO)3 in frozen argon (12 K). The
observation of carbonyl loss at 600 nm is somewhat
puzzling as there are no apparent electronic absorption
bands in this region.
Electronic spectra of (p5-C5H5)M(CO)4, M=Nb or
Ta, after photolysis, Fig. 8b and Fig. 11b, and the
difference spectra, Figs. 9 and 12, show the bands at
430, 390 and 250 nm (the 250 nm band for Nb is not
observed in Fig. 8b due to the high concentration of
this sample) to have been bleached while new bands at
about 660, 390 and 280 nm grow in. These new bands
are associated with the carbonyl-loss photoproducts,
(p5-C5H5)M(CO)3.
Upon photolysis of these Nb and Ta compounds at
higher energy (340BuB420 nm), additional bands
were observed that may be assigned to the cis and trans
dicarbonyl-loss products by analogy to bands assigned
Figs.
7
and 10 present IR spectra of (p5-
C5H5)Nb(CO)4 and (p5-C9H7)Ta(CO)4 in frozen Nujol
matrices upon photolysis at various wavelengths and
upon annealing. Photolysis of (p5-C5H5)M(CO)4 and
(p5-C9H7)M(CO)4, M=Nb or Ta, at wavelengths as
long as 600 nm (480BuB620 nm) were found to
stimulate CO loss as evidenced by difference IR spectra,
Fig. 7b and Fig. 10b. These figures show that bands
Table 7
˚
Selected bond lengths (A) and angles (°) for (C5H4C(O)CH3)Nb(CO)4
Atom
Distance
Atoms
Angle
NbꢁC(1)
NbꢁC(2)
NbꢁC(3)
NbꢁC(4)
NbꢁC(5)
NbꢁC(8)
NbꢁC(9)
NbꢁC(10)
NbꢁC(11)
C1)ꢁC(6)
C6)ꢁO(6)
C6)ꢁC(7)
2.370(3)
2.420(3)
2.454(3)
2.425(3)
2.382(3)
2.091(3)
2.096(3)
2.078(3)
2.069(3)
1.488(4)
1.211(4)
1.495(4)
C(8)ꢁNbꢁC(9)
C(8)ꢁNbꢁC(11)
C(9)ꢁNbꢁC(10)
C(10)ꢁNbꢁC(11)
C(8)ꢁNbꢁC(10)
C(9)ꢁNbꢁC(11)
C(1)ꢁC(6)ꢁC(7)
C(1)ꢁC(6)ꢁO(6)
74.88(14)
73.75(13)
74.46(13)
73.67(12)
117.04(12)
117.09(12)
118.3(3)
Table 5
˚
Selected bond lengths (A) and angles (°) for CpTa(CO)4
Atoms
Distance
Atoms
Angle
TaꢁC(1)
TaꢁC(2)
TaꢁC(3)
TaꢁC(11)
TaꢁC(12)
2.347(11)
C(11)ꢁTaꢁC1(1a)
C(11)ꢁTaꢁC1(2)
C(11a)ꢁTaꢁC1(2a)
C(12)ꢁTaꢁC1(2a)
C(11)ꢁTaꢁC1(2a)
74.9(5)
74.5(3)
74.5(3)
73.4(5)
117.4(3)
2.362(8)
2.357(9)
2.081(9)
2.082(10)
120.0(3)
C(11a)ꢁTaꢁC(12) 117.4(3)

82
T.E. Bitterwolf et al. / Journal of Organometallic Chemistry 557 (1998) 77–92
Fig. 4. Molecular structure of (p5-C5H5)Ta(CO)4.
Fig. 2. Molecular structure of (p5-C5H4CHꢀCH2)Nb(CO)4.
species. While photoreversal has been observed by Rest
[12], there is no current explanation for how stimula-
tion of an excited band in an isolated metal species
prompts ligand capture.
by Rest to (p5-C5H5)V(CO)2. In the case of (p5-
C5H5)Nb(CO)4, the trans species is not observed upon
photolysis, but appears upon annealing the sample. It is
possible that the cis species is initially formed in all
cases and undergoes either thermal or photochemical
conversion to trans. Small, but distinct, bands were also
observed that correspond to bands assigned by Rest to
tricarbonyl-loss fragments [12]. It is likely that Nujol
solvates all of the carbonyl-loss products although it is
difficult to understand how hydrocarbon solvation
might stabilize a formally 12-electron species such as
CpM(CO).
As noted above, bands found at about 660 nm in the
difference UV–vis spectra are identified with the single-
carbonyl loss products, (p5-C5H5)M(CO)3. Photolysis
into this band for both the Nb and Ta compounds
resulted in recapture of CO by the electron deficient
Upon annealing freshly photolyzed samples to 133
K, the bands attributable to ‘free’ CO, and carbonyl-
loss species decayed, while bands of the parent com-
pound grew in. For both (p5-C5H5)Ta(CO)4, and
(p5-C9H7)Ta(CO)4 Figs. 10–12 several new bands in the
terminal carbonyl region were also observed to grow in
upon annealing. We tentatively attribute these bands to
a CO capture species in which the p5-indenyl ring has
slipped to its p3-form. We are initiating solution and
supercritical studies of this compound in the presence
of a pressure of CO to establish whether such a species
may be formed under these conditions. Processes in-
volving what appear to be similar p5 to p3 interconver-
sions have been observed by Casey and coworkers for
rhenium compounds [17].
The observed photochemical reactions are summa-
rized in Scheme III. Of particular interest in this series
of compounds is the ease with which two carbonyl
groups can be lost. As the resulting species is formally
a 14-electron species, and since highly electron deficient
species are known to be important in polymerization
catalysis, we are continuing to investigate the photo-
chemistry of these Group V compounds in solution and
supercritical fluids such as ethylene.
3. Experimental
All operations were carried out under a nitrogen or
argon atmosphere using standard Schlenk or vacuum
line techniques. Both nitrogen and argon were dried
with molecular sieves and trace oxygen was removed
with BASF catalyst. CO (99.5%) was purchased from
Fig. 3. Molecular structure of (p5-C9H7)Ta(CO)4.

T.E. Bitterwolf et al. / Journal of Organometallic Chemistry 557 (1998) 77–92
83
Fig. 5. Molecular structure of (p5-C5H4CO)C6H5)Ta(CO)4.
Liquid Carbonic Specialty Gases and used without
further purification. 13C-enriched CO (99%) was pur-
chased from Cambridge Isotope Laboratories and used
as received. Solvents were dried using appropriate dry-
ing agents and distilled under nitrogen.
Column chromatography was carried out using
Fisher brand Florisil, silica gel, or Aldrich ꢀ150 mesh)
neutral alumina. The deactivated silica gel and alumina
were placed under vacuum for 24 h at r.t. then stored
under nitrogen. Celite was obtained from Fisher Scien-
tific and oven dried prior to use.
NbCl5 and TaCl5 were purchased from Strem Chem-
icals and were purified by sublimation prior to use.
Particularly impure NbCl5 was refluxed in thionyl chlo-
ride for 24 h. The solvent was removed under reduced
pressure and the yellow solid was sublimed twice to
produce pure NbCl5.
Proton and carbon NMR spectra were performed on
either an IBM NR 300 or a Varian Gemini-300. IR
spectra were recorded using either a Digilab Qualimatic
FT-IR or a Perkin Elmer FT-IR Spectrum 1000. Mass
spectra were obtained using a Vacuum Generators
7070HS GC/MS.
High pressure reactions were carried out in a 300
cm−3 Parr series 4560 reactor, equipped with a mag-
netic stirrer. When necessitated by low tank pressures,
the CO pressure in the Parr reactor was boosted using
a 10000 psi, 30 cc, syringe pump, model 62-6-10, from
High Pressure Equipment of Eire, PA. Pressure was
monitored using a 10000 psi, temperature compensated,
Heise gauge, from Newton Connecticut.
Melting points were determined on a Mel-Temp II
apparatus (Laboratory Devices) and are uncorrected.
Elemental analysis was performed by Desert Analytics
of Tucson, AZ. Li(C5H4CHꢀCH2) [18], (CH3)3SiC5H5
[19] and [(CH3)3Si]2C5H4 [20] were prepared by litera-
ture methods.
3.1. Preparation of (p5-C9H7)Nb(CO)4 6ia
Na[Nb2(CO)8(v-I)3]
A 100 cm3 Schlenk flask was loaded with Na[Nb-
(CO)6]·THF (1.2 g, 3.37 mmol) and THF (30 cm3). The
flask was placed under 8 psi of CO and maintained at
r.t.. A solution containing iodine (0.85 g, 3.35 mmol) in
THF (15 cm3) was slowly injected into the rapidly
stirred mixture. Prompt CO evolution was accompanied
by a color change to brown/purple. After the addition
was complete the reaction was stirred for 15 min prior
to the addition of a THF (25 cm3) solution of NaC9H7
(0.95 g, 6.88 mmol). Although there was no visible
change in the appearance of the mixture, IR established
the reaction to be complete within 5 min. The solvent
was removed under reduced pressure. The residue was
extracted with benzene (3×25 cm3) and the benzene
Fig. 6. Molecular structure of (p5-C5H4C((O)CH3)Nb(CO)4.

84
T.E. Bitterwolf et al. / Journal of Organometallic Chemistry 557 (1998) 77–92
Table 8
IR absorption spectra for niobium and tantalum intermediatesa
RM(CO)4
RM(CO)3
RM(CO)2
RM(CO)
1831
(3-C9H7)M(CO)4X)
cis
trans
CpV(CO)b4
InV(CO)d4
2030
1951
1930
2029
1957
1934
2034
1942
1927
2034
1947
1930
2033
1934
1918
2032
1937
1921
1953
1896
1857
1994
1917
1903
1982
1881
1871
1980
1883
1873
1977
1872
1859
1975
1873
1864
1914
1813
1992
1895
2055c
1982
1946
2052e
1978
CpNb(CO)4
InNb(CO)4
CpTa(CO)4
InTa(CO)4
1898
1804
2006
1844
f
1899
1809
1881
1790
1991
1905
1844
1846
2050
1985
1951
2049
1955
1897
1887
1800
1996
1906
a
Cp, p5-C5H5; In, p5-C9H7; R=Cp or In. b In ref. [12], CH4 matrix. c Data reported for (3-C9H7)MCO)5 in a CO matrix. d In ref. [13], CH4
matrix. e Carbonyl stretching frequencies reported for (3-C9H7)M(CO)4(N2) in a nitrogen matrix. The N2 stretching frequency is 2220 cm−1
f Obscured by an overlapping band of another photoproduct, or of low intensity due to low abundance.
.
extracts filtered through a 3×3 cm plug of Celite. The
solvent was removed and the residue sublimed to yield
a maroon solid. There was no evidence of 1,1%-biindenyl
in either the crude product or the sublimed product.
Spectral data were identical to those previously re-
ported. Yield: 0.50 g, 46%.
3.3. Preparation of (p5-C5H4CH2)3(CH3)Ta(CO)4
A 100 cm3 Schlenk flask was loaded with Na[T-
a(CO)6]·THF 1.1 g, 2.48 mmol). After addition of
dimethoxyethane (15 cm3) the flask was placed under
an atmosphere of CO (8 psi). In a Schlenk addition
funnel a slurry of finely ground TaCl5 (0.68 g, 1.89
mmol) and dimethoxyethane (16 cm3) under a CO
atmosphere was prepared. The TaCl5 slurry was slowly
added to the flask containing the hexacarbonyl while
the temperature was maintained between −50 and
−40°C. The flask was allowed to slowly warm to r.t.
producing a dark red solution. The solution was stirred
at r.t. for 30 min then cooled to 0°C and added to
3.2. Preparation of (p5-C5H4CHꢀCH2)Nb(CO)4
(p5-C5H4CHꢀCH2)Nb(CO)4 was prepared as a deep
maroon solid by reaction of LiC5H4CHCH2 with
Na[Nb2(CO)8v-Cl)3] by the procedure reported previ-
ously [4]. The compound appears to be very sensitive to
light (even as a solid) and during the first sublimation
noticeable discoloration was observed. The material
was protected from light during the second sublimation
but previous decomposition may account for the partic-
ularly poor yield of this compound. Crystals suitable
for X-ray analysis were obtained on the second subli-
mation of the product. Limited sample prevented deter-
mination of melting point. IR spectrum (CH2Cl2): 2033
LiC5H4(CH2)3CH3
(0.56
g,
4.41
mmol)
in
dimethoxyethane (18 cm3). The resulting solution was
heated to 85°C for 14 h. After removal of solvent under
reduced pressure, dichloromethane (15 cm3) was added
to give a brown solution which was flash chro-
matographed through a 12 cm plug of degassed alu-
mina producing a broad orange band. Decomposition
and CO evolution was observed while the compound
was in contact with the column. The solvent was re-
moved under reduced pressure yielding a small quantity
1
(s), 1919 (vs) cm−1. H-NMR (CDCl3): l 6.30 m, 1 H,
(CHꢀC), 5.80 br, 2 H, (Cp), 5.45 br, 3 H, Cp and
(CꢀCH2), 5.00 d, 1 H, (CꢀCH2). 13C-NMR (CDCl3): l
129.08 (CHꢀCH2), 113.56 (CHꢀCH2), 93.26 br, (Cp).
93Nb-NMR (C6D6): l −1945.9.0. Anal. Calc. for
C11H7O4Nb: C, 44.62; H, 2.38. Found: C, 43.86; H,
2.49. Yield: B5%.
of red oil. IR (CH2Cl2): 2028 (s), 1908 (vs) cm−1
.
1H-NMR (CDCl3): l 5.7 s, 2 H, (Cp), 5.4 s, 2 H, (Cp),
2.3 m, 2 H, (CH2), 1.4 br m, 4 H, (CH2), 0.9 t, 3 H,
(CH3). 13C-NMR (CDCl3): l 245.4 (CO), 119.6 ipso

T.E. Bitterwolf et al. / Journal of Organometallic Chemistry 557 (1998) 77–92
85
(Cp), 94.0 (Cp), 91.4 (Cp), 38.7 (CH2), 28.8 (CH2), 22.3
(CH2), 13.9 (CH3). Yield: B5%.
allowed to settle and the clear pale yellow layer was
filtered into a 500 cm3 Schlenk flask using a Schlenk filter.
The filtrate was cooled in an ice bath and methyl lithium
(95 cm3 of 1.5 M in Et2O), was slowly added by syringe
producing a clear solution. The mixture was stirred for
5 h at r.t.. The flask was again cooled with an ice bath
and trimethylsilylchloride (14.5 g, 0.133 mol), was in-
jected into the solution. A pale yellow solution quickly
formed and the mixture was allowed to warm to r.t. while
stirring overnight. The precipitate was allowed to settle
and the solution filtered and concentrated to a bright
3.4. Preparation of (CH3)3SiC5H4(CH2)3CH3
A 500 cm3 Schlenk flask was loaded with NaC5H5 (12.5
g, 0.142 mol) and ethyl ether (350 cm3). The flask was
placed in an ice bath and 1-bromobutane (19.0 g, 0.139
mol) was slowly injected into the rapidly stirred solution.
After 30 min the ice bath was removed and the flask
allowed to warm to r.t.. After 20 h the mixture was
Fig. 7. IR spectra with (p5-C5H5)Nb(CO)4 isolated in a Nujol matrix: (a) starting material, (b) subtraction spectrum, 60 min irradiation at
480BuB620 nm minus starting material, (c) subtraction spectrum, 30 min irradiation at 340BuB420 nm minus starting material, (d)
subtraction spectrum, difference between two 15 min periods of irradiation at 340BuB420 nm, (e) subtraction spectrum, anneal to 133 K minus
final photolysis. * Arise from (p5-C5H5)Nb(CO)3 13CO) present in natural abundance.
(

86
T.E. Bitterwolf et al. / Journal of Organometallic Chemistry 557 (1998) 77–92
Fig. 8. Electronic absorption spectrum of (p5-C5H5)Nb(CO)4 isolated in a Nujol matrix: (a) starting material, (b) after 60 min irradiation at
480BuB620 nm and 25 min irradiation at 340BuB420 nm. * Irregularity arising from an instrument artifact.
yellow oil. The oil was transferred to a 14/20 short path
3.5. Preparation of (p5-C5H5)Nb(CO)4
distillation apparatus and a pale yellow distillate col-
lected (28–34°C at 0.5 torr). H-NMR (CDCl3): l 6.50
A 500 cm3 Schlenk flask was loaded with finely
ground NbCl5 (10.6 g, 39.2 mmol) and CH2Cl2 (400
cm3). This mixture was rapidly stirred for several hours
to produce a fine suspension. To the r.t. mixture
(CH3)3SiC5H5 (6.6 g, 47.8 mmol) was added by syringe
over 40 min producing an immediate color change to
brown/red. After an additional hour stirring was
stopped and the fine solid allowed to settle. The brown
solution was removed using a filter stick. The residual
1
(s, 1 H, vinylic), 6.47 (s, 1 H, vinylic), 6.11 (s, 1 H,
vinylic), 3.28 (s, 1 H, HCTMS), 2.45 (t, 2 H, J=7.3
Hz, CH2), 1.57 (m, 2 H, CH2), 1.42 (m, 2 H, CH2), 0.96
(t, 3 H, J=7.1 Hz, CH3), 0.00 (s, 9 H, TMS). 13C-
NMR (CDCl3): l 146.0 (ipso), 133.6 (vinylic, CH),
132.3 (vinylic, CH), 126.8 (vinylic CH), 50.8 (HCTMS),
32.2 (CH2), 29.5 (CH2), 22.6 (CH2), 14.0 (CH3), −1.7
(SiCH3)3). Yield: 7.7g, 30%.

T.E. Bitterwolf et al. / Journal of Organometallic Chemistry 557 (1998) 77–92
87
solvent was removed under reduced pressure and 11.0 g
of fine copper-colored solid was recovered. No further
attempt was made to purify the CpNbCl4. The crude
tetrachloride was loaded into a 300 cm3 Parr reactor with
magnesium (2.3 g, 94.7 mmol) and zinc (2.3 g, 35.1
mmol). Freshly distilled pyridine (220 cm3) was trans-
ferred into the bomb. Once sealed the bomb was
promptly pressurized to 1800 psi and heated to 60°C.
Pressure was maintained by frequently recharging with
CO. After 86 h the brown solution was poured into a 250
cm3 Schlenk flask under a flow of nitrogen. The solvent
was removed under reduced pressure using a hot water
bath to yield a dark brown solid. The solid was broken
up with a spatula and 40 cm3 of benzene added. After
stirring for 30 min, the solution was filtered through a
3×3 cm plug of Celite and the residual solid was washed
with fresh benzene (4×25 cm3). The benzene was re-
moved under reduced pressure to produce an orange/red
solid. The solid was sublimed twice to produce 5.1 g of
dark maroon solid. IR and NMR data are identical to
those previously published. Yield based on the crude
(p5-C5H5)NbCl4 yield: 58%.
Fig. 9. Difference electronic absorption spectrum of (p5-C5H5)Nb(CO)4 isolated in a Nujol matrix. The difference was taken between the final
photolysis 60 min irradiation at 480BuB620 nm and 25 min irradiation at 340BuB420 nm) and the starting material.

88
T.E. Bitterwolf et al. / Journal of Organometallic Chemistry 557 (1998) 77–92
Fig. 10. IR spectra with (p5-C9H7)Ta(CO)4 isolated in a Nujol matrix: (a) starting material, (b) subtraction spectrum, 60 min irradiation at
340BuB620 nm (100 min total at u\340 nm) minus starting material, (c) subtraction spectrum, 20 min irradiation at uB285 nm minus 100
min irradiation at u\340, (d) subtraction spectrum, anneal to 173 K minus 125 min total photolysis at u\225 nm.
This same preparation was attempted with toluene
and acetonitrile under identical conditions as those
given above but neither produced the desired product.
Changing the solvent to THF did produced the (p5-
C5H5)Nb(CO)4 which was identified by IR. However,
difficulties in purification prohibited the isolation of the
compound and no yield could be reported. The above
reaction was also conducted under 1.4 atmospheres of
CO in pyridine but an IR of the benzene extract
showed no metal carbonyls. Increasing the quantity of
zinc (3.2 g) while reducing that of magnesium (1.2 g)
resulted in a slightly reduced yield (52%).
3.6. Preparation of (p5-C5H5)Nb(CO)4−n(13CO)n
(p5-C5H5)Nb(CO)4 (0.1 g, 0.37 mmol) was loaded
into a 60 cm3 Griffin–Worden tube and benzene (10
cm3) was added. The solution was freeze/pump/thawed
three times to remove nitrogen. The tube was charged
to a pressure of 30 psi with 13C-labeled CO. The
solution was protected from light and stirred. After 3
months the solution was withdrawn via syringe, placed
in a 50 cm3 Schlenk flask and the solvent removed,
producing a orange/red solid. IR Spectrum (CH2Cl2):
2033 (s), 2024 (s), 2016 (s), 1918 (vs), 1897 (vs sh)

T.E. Bitterwolf et al. / Journal of Organometallic Chemistry 557 (1998) 77–92
89
Fig. 11. Electronic absorption spectrum of (p5-C5H5)Ta(CO)4 isolated in a Nujol matrix: (a) starting material, (b) after 240 min irradiation at
u\285 nm . * Irregularity arising from an instrument artifact or an artifact from the cryostat.
cm−1.1H-NMR (CDCl3): l 5.62–5.55 (br, C5H5). 13C-
NMR (CDCl3): l 251.23 (CO, J=222), 95.24–93.98
(br, C5H5).
45 min. There was an immediate color change to or-
ange/brown. The mixture was stirred for several hours
at r.t. producing a fine orange precipitate in a brown
solution. The flask was placed in the deep freeze (−
80°C) for several hours and the (p5-C5H5)TaCl4 was
allowed to settle. The brown solution was removed with
a filter stick and discarded. The residual solvent was
removed from the yellow/orange solid under vacuum.
After transferring the flask to the glove box, the crude
(p5-C5H5)TaCl4 (11.2 g) was loaded into the Parr bomb
3.7. Preparation of (p5-C5H5)Ta(CO)4
A 500 cm3 Schlenk flask was loaded with finely
ground TaCl5 (12.0 g, 33.5 mol) and of CH2Cl2 (300
cm3). The mixture was rapidly stirred at r.t. for several
hours to produce a fine suspension. CH3)3SiC5H5 (5.4 g,
39.1 mmol) was injected to the TaCl5 suspension over

90
T.E. Bitterwolf et al. / Journal of Organometallic Chemistry 557 (1998) 77–92
Fig. 12. Difference electronic absorption spectrum of (p5-C5H5)Ta(CO)4 isolated in a Nujol matrix. The difference was taken between the final
photolysis (240 min irradiation at u\285 nm) and the starting material. * Irregularity arising from an instrument artifact.
with magnesium (2.1 g, 86.4 mmol) and zinc (2.1 g, 32.1
mmol). Freshly distilled pyridine (220 cm3) was trans-
ferred into the bomb. Upon sealing, the bomb was
promptly pressurized with CO to 1600 psi and heated
to 60°C. Pressure was maintained by frequently
recharging with CO. After 96 h the brown solution was
poured into a 250 cm3 Schlenk flask under a flow of
nitrogen. The remaining work up followed that previ-
ously outlined in the preparation of (p5-C5H5)Nb(CO)4.
(p5-C5H5)Ta(CO)4 was doubly sublimed (85°C, 0.5
Torr) to produce 5.4 g of orange/red crystals. Crystals
suitable for X-ray analysis were obtained from the
second sublimation. IR and NMR data were identical
to those previously published. Yield (based on the
crude CpTaCl4 yield): 52%.
3.8. Preparation of (p5-C5H4CH2)3(CH3)Nb(CO)4
With some noted exceptions, this compound was
prepared in a similar manner as that described for
(p5-C5H5)Nb(CO)4, producing a deep maroon oil. Ow-
ing to the high solubility of (p5-C5H4CH2)3(CH3)NbCl4
the compound could not be separated from the reaction
solution by filtration as was the case with (p5-

T.E. Bitterwolf et al. / Journal of Organometallic Chemistry 557 (1998) 77–92
91
C5H5)NbCl4. Solvent was removed from the reaction
mixture under vacuum and the impure (p5-
C5H4CH2)3(CH3)NbCl4 was loaded directly into the
bomb. Once sealed the bomb was promptly pressurized
to 1980 psi and heated to 55°C for 19 h. After solvent
removal the compound was extracted with several por-
tions of benzene. The resulting red solution was then
filtered through a 3×3 cm plug of Celite followed by a
plug of silica. Removal of the solvent left a brown/red
oil. The oil was purified using a very short path distilla-
tion apparatus and a hot oil bath (120–136°C at 0.5
torr). The resulting maroon oil is extremely light sensi-
tive and was protected from light. IR (CH2Cl2): 2031
208 (210–H2, 11%). Exact Mass: Calc. 325.9877;
Found, 325.9887. Yield (based on NbCl5): 30%.
3.9. Preparation of [(p5-C5H4Si(CH3)3]Nb(CO)4
With some noted exceptions, this compound was
prepared in a similar manner as that described for
(p5-C5H5)Nb(CO)4, producing a deep maroon solid. As
in the case of the butyl derivatives, [p5-
C5H4Si(CH3)3]NbCl4 could not be purified by filtration
and the crude material was used directly after solvent
removal. The Parr bomb was pressurized to 2150 psi
and
heated
to
50°C
for
22
h.
[p5-
1
(s), 1915 (vs), cm−1. H-NMR (CDCl3): l 5.57 br s, 2
C5H4Si(CH3)3]Nb(CO)4 and its Ta analog are extremely
light sensitive and must be stored in the dark to prevent
decomposition. M.p. 47°C; dec. 208°C. IR (CH2Cl2):
2031 (s), 1917 (vs), cm−1. 1H-NMR (CDCl3): l 5.70 (br
s, 2 H, Cp), 5.55 (br s, 2 H, Cp), 0.18 s, (SiCH3)3).
13C-NMR (CDCl3): l the ipso Cp was not observed,
98.78 (br, Cp), −0.12 (Si(CH3)3). Anal. Calc. for
C12H13O4SiNb: C, 42.12; H, 3.83. Found: C, 41.87; H,
3.68. Yield (based on NbCl5): 52%.
H, (Cp), 5.43 br s, 2 H, (Cp), 2.29 br m, 2 H, (CH2),
1.44 m, 2 H, (CH2), 1.33 m, 2 H, (CH2), 0.89 t, 3 H,
J=7.2 Hz, (CH3). 13C-NMR (CDCl3): l 120.90 br, ipso
(Cp), 94.83 br, (Cp), 92.45 br, (Cp), 37.09 (CH2), 28.38
(CH2), 22.27 (CH2), 13.72 (CH3). 93Nb-NMR (C6D6): l
−1995.0. MS (m/z): 326 (M+, 6%), 298 (M+ –1CO,
2%), 268 (M+ –2CO–H2, 9%), 240 (M+ –3CO–H2,
16%), 212 (M+ –4CO–H2, 36%), 210 (212–H2, 32%),
Scheme III. (i) UV–vis irradiation 222BuB620 nm; (ii) UV irradiation 340BuB420 nm, M=Nb, or (p5-C9H7)Ta(CO)4; (iii) UV irradiation
340BuB420 nm, M=Nb, or (p5-C5H5)Nb(CO)4 or (p5-C9H7)Ta(CO)4; (iv) UV irradiation 270BuB290 nm for (p5-C5H5)Ta(CO)4 or u\285
nm for (p5-C9H7)Ta(CO)4; (v) M=Ta.

92
T.E. Bitterwolf et al. / Journal of Organometallic Chemistry 557 (1998) 77–92
3.10. Preparation of [p5-C5H4 Si(CH3)3]Ta( CO)4
library [21]. Program DIFABS is described by Walker
and Stuart [22].
With some noted exceptions, this compound was
prepared in a similar manner as that described for
(p5-C5H5)Ta(CO)4, producing a highly light-sensitive
bright orange solid. [p5-C5H4Si(CH3)3]TaCl4 could not
be separated from its reaction mixture so the crude
material was used after removal of solvent. The Parr
bomb was pressurized to 1970 psi and heated to 55°C
for 63 h. M.p. 59°C; dec. 225°C. IR (CH2Cl2): 2029 (s),
Acknowledgements
Thomas E. Bitterwolf thanks the Research Corpora-
tion for their generous support for the purchase of the
Perkin Elmer FT-IR Spectrum 1000 Spectrometer and
the University of Idaho Research Council for funding
this research. The University of Delaware acknowledges
the National Science Foundation for their support of
the purchase of the CCD-based diffractometer (Grant
CHE-9628768).
1909 (vs), cm−1 1H-NMR (CDCl3): l 5.73 s, 2 H,
.
(Cp), 5.55 s, 2 H, (Cp), 0.20 s, (Si(CH3)3). 13C-NMR
(CDCl3): l 244.62 (CO), 105.05 ipso (Cp), 98.44 (Cp),
97.79 (Cp), −0.12 (Si(CH3)3). Anal. Calc. for
C12H13O4SiTa: C, 33.50 H, 3.05. Found: C, 33.20; H,
2.90. MS (m/z): 430 (M+, 18%), 402 (M+ –1CO, 9%),
387 (402–Me, 6%), 374 (M+ –2CO, 7%), 359 (374–Me,
5%), 346 (M+ –3CO, 42%), 318 (M+ –4CO, 100%), 246
(M+ –4CO–TMS, 9%). Exact Mass: Calc. 430.0036;
Found, 430.0047. Yield (based on TaCl5): 42%.
References
[1] R.B. King, Z. Naturforsch. 18B (1963) 157.
[2] W.A. Herrmann, W. Kalcher, H. Biersack, I. Bernal, M.
Creswick, Chem. Ber. 114 (1981) 3558.
[3] F. Calderazzo, M. Castellani, G. Pampaloni, P.F. Zanazzi, J.
Chem. Soc. Dalton Trans. (1985) 1989.
3.11. Crystallographic structure determination
[4] T.E. Bitterwolf, S. Gallagher, A.L. Rheingold, G.P.A. Yap, J.
Organometal. Chem. (1997), in press.
[5] A.M. Cardoso, R.J.H. Clark, S. Moorhouse, J. Organometal.
Chem. 186 (1990) 237.
[6] B.V. Lokshin, E.B. Rusach, Z.P. Valueva, A.G. Ginzburg, N.E.
Kolobova, J. Organomet. Chem. 102 (1975) 535.
[7] P.B. Graham, M.D. Rausch, K. Taeschler, W. Von Philipsborn,
Organometallics 10 (1991) 3049.
The single crystal X-ray diffraction experiments were
performed on a Siemens P4 diffractometer for 1–3 and
5, and on a Siemens P4/CCD diffractometer for 4.
The systematic absences in the diffraction data were
consistent for the reported space group. For 2, either of
the monoclinic space groups P21 and P21/m was indi-
cated; for 3, either of the orthorhombic space groups
Pna21 and Pnma was indicated; in both cases the latter
centrosymmetric space groups were preferred based on
the chemically reasonable and computationally stable
results of refinement. The structures were solved using
direct methods, completed by subsequent difference
Fourier synthesis and refined by full-matrix least-
squares procedures. The data were corrected for ab-
sorption for 2, 4 and 5. The absorption corrections
were not required for 1 and 3 because the variation in
the integrated C-scan intensities was B10%; the empir-
ical absorption corrections for 2 and 4 were applied by
using the program DIFABS; the semi-empirical absorp-
tion correction data for 5 were collected by the C-scan
technique. All non-hydrogen atoms were refined with
anisotropic displacement coefficients. All hydrogen
atoms were treated as idealized contributions. In 2,
atoms Ta, C(3), and H(3A) reside in a mirror plane. In
3, atoms (Ta, C1), and H(1A) reside in a mirror plane.
In the case of 2, the final difference map contained a
[8] W.A. Herrmann, H. Biersack, J. Organomet. Chem. 191 (1980)
397.
[9] J.D. Rehder, H.-Ch. Bechthold, Z. Naturforsch. 39b (1984) 323.
[10] A.N. Nesmeyanov, K.N. Anisimov, N.E. Kolobova, A.A.
Pasynskii, Izv. Akad. Nauk B. SSR Ser. Khim. Nauk (1968)
2814 (Engl. trans. p. 2664). Izv. Akad. Nauk B. SSR Ser. Khim.
Nauk (1969) 100 (Engl. trans. p. 87). Izv. Akad. Nauk B. SSR
Ser. Khim. Nauk (1969) 183 (Engl. trans. p. 179).
[11] (a) M.T. Haward, M.W. George, S.M. Howdel, M. Poliakoff, J.
Chem. Soc. Chem. Comm. (1990) 913. (b) M.W. George, M.T.
Haward, P.A. Hamley, C. Hughes, F.P.A. Johnson, V.K. Popov,
M. Poliakoff, J. Am. Chem. Soc. 115 (1993) 2286.
[12] R.B. Hitam, A.J. Rest, Organometallics 8 (1989) 1598.
[13] A.J. Rest, M. Herberhold, M. Schrepfermann, Organometallics
11 (1992) 3646.
[14] T.E. Bitterwolf, K.A. Lott, A.J. Rest, J. Mascetti, J.
Organometal. Chem. 419 (1991) 113.
[15] M. Hoch, A. Duch, D. Rheder, Inorg. Chem. 25 (1986) 2907.
[16] A.J. Rest, M. Herberhold, M. Schrepfermann, Organometallics
11 (1992) 3646.
[17] (a) C.P. Casey, J.T. Brady, Am. Chem. Soc. 210th National
Meeting, Chicago, IL, 20–24 August, 1995, INORG182. (b) C.P.
Casey, J.M. O’Connor, Organometallics 4 (1984) 384.
[18] O. Massao, D.T. Mallin, D.W. Macomber, M.D. Rausch, R.D.
Rogers, A.N. Rollins, J. Organometal. Chem. 405 (1991) 41.
[19] (a) E.W. Abel, M.O. Dunster, J. Organomet. Chem. 33 (1971)
161. (b) K.C. Frisch, J. Am. Chem. Soc. 75 (1953) 6050.
[20] Yu.A. Ustynyuk, A.V. Kisin, I.M. Pribykova, A.A. Zenkin,
N.D. Antonova, J. Organometal. Chem. 42 (1972) 47.
[21] G. Sheldrick, Siemens XRD, Madison, WI.
3
˚
single large peak (ca. 1.02 eA ) in a chemically unrea-
sonable position as its only prominent feature. The final
difference maps in the cases of 1 and 3–5 were
featureless.
All software and sources of the scattering factors are
contained in the SHELXTL (version 5.03) program
.
[22] N. Walker, D. Stuart, Acta Crystallogr. A39 (1983) 158.
.