Synthesis and characterization of zirconium and iron complexes containing substituted indenyl ligands: Evaluation of steric and electronic parameters

Article abstract of DOI:10.1021/om0495675

Evaluation of the steric and electronic influence of a family of silyl-and alkyl-substituted indenyl liganda on zirconium and iron centers has been accomplished by a combination of X-ray diffraction, IR spectroscopy, solution NMR dynamics, and electrochemical measurements. Three tetrasubstituted, bis-indenyl zirconocene dichloride complexes have been characterized by X-ray diffraction and adopt a gauche ligand conformation such that the interactions between tertiary substituents on adjacent rings are minimized. Similar solid state conformations were also observed in two of the corresponding iron compounds. Evaluation of the electronic environment about each zirconium center was achieved by measurement of the CO stretching frequencies of the dicarbonyl derivatives. Simple inductive effects govern the electronic properties of each zirconocene where silyl groups are relatively electron withdrawing and alkyl groups electron donating. For the most hindered zirconocene dicarbonyl derivatives, population of three vibrationally distinct rotamers has been detected by IR spectroscopy. Independent assessment of these stereoelectronic parameters by variable-temperature NMR spectroscopy and electrochemistry with the analogous series of iron complexes provided the same relative ordering of the indenyl ligands.

Full text of DOI:10.1021/om0495675

5332
Organometallics 2004, 23, 5332-5346
Syn th esis a n d Ch a r a cter iza tion of Zir con iu m a n d Ir on
Com p lexes Con ta in in g Su bstitu ted In d en yl Liga n d s:
Eva lu a tion of Ster ic a n d Electr on ic P a r a m eter s
Christopher A. Bradley, Samuel Flores-Torres, Emil Lobkovsky,
Hector D. Abrun˜a, and Paul J . Chirik*
Department of Chemistry and Chemical Biology, Baker Laboratory, Cornell University,
Ithaca, New York 14853
Received J une 14, 2004
Evaluation of the steric and electronic influence of a family of silyl- and alkyl-substituted
indenyl ligands on zirconium and iron centers has been accomplished by a combination of
X-ray diffraction, IR spectroscopy, solution NMR dynamics, and electrochemical measure-
ments. Three tetrasubstituted, bis-indenyl zirconocene dichloride complexes have been
characterized by X-ray diffraction and adopt a gauche ligand conformation such that the
interactions between tertiary substituents on adjacent rings are minimized. Similar solid
state conformations were also observed in two of the corresponding iron compounds.
Evaluation of the electronic environment about each zirconium center was achieved by
measurement of the CO stretching frequencies of the dicarbonyl derivatives. Simple inductive
effects govern the electronic properties of each zirconocene where silyl groups are relatively
electron withdrawing and alkyl groups electron donating. For the most hindered zirconocene
dicarbonyl derivatives, population of three vibrationally distinct rotamers has been detected
by IR spectroscopy. Independent assessment of these stereoelectronic parameters by variable-
temperature NMR spectroscopy and electrochemistry with the analogous series of iron
complexes provided the same relative ordering of the indenyl ligands.
In tr od u ction
cyclometalated hydrides are obtained, while in another,
N2 coordination to form an activated, side-on bound N2
complex results.6 More recently we have demonstrated
that the number of cyclopentadienyl methyl groups on
a low-valent zirconocene dictates both the hapticity and
reactivity of coordinated N2. Whereas [(η5-C5Me5)2Zr-
(η1-N2)]2(µ2,η1,η1-N2) contains weakly activated, end-on
bound dinitrogen ligands that are easily displaced by
H2,7 removal of one methyl group from each Cp* ring
yields a strongly activated zirconocene dinitrogen com-
plex that promotes the hydrogenation of N2 to ammonia
with mild thermolysis.8 We have also discovered pro-
found reactivity differences on replacing cyclopenta-
dienyl with their indenyl counterparts. Instead of cyclo-
metalation or N2 activation, reductive elimination of
isobutane from (η5-C9H5-1,3-(SiMe3)2)2Zr(CH2CHMe2)H
yields a zirconium sandwich complex that provides
convenient access to an isolable Zr(II) equivalent. This
unusual “zirconocene” promotes C-H activation under
mild conditions, readily coordinates alkynes, and under-
goes haptotropic rearrangement to its η6, η5 isomer upon
addition of THF.9
Substituted cyclopentadienyl and related indenyl
anions occupy a prominent role in organometallic chem-
istry, serving as versatile ligands for main group ele-
ments, transition metals, and actinides.1 Seemingly
subtle changes in either cyclopentadienyl or indenyl
ligand substitution can have profound consequences on
chemical reactivity. For example, manipulation of alkyl
and silyl groups on the ligand scaffold in group 4
metallocene-promoted olefin polymerization determines
catalyst symmetry and therefore stereochemistry and
physical properties of the resulting polyolefin.2 While
these effects are reasonably well understood for rigid
ansa-metallocenes,3 unbridged cyclopentadienyl and
indenyl ligated zirconium precatalysts often display
variable activities and levels of stereocontrol that are
dependent on both activator4 and ring substituents.5
Investigations in our laboratory focusing on the
chemistry of low-valent group 4 metallocenes have
demonstrated the importance of cyclopentadienyl and
indenyl substituent effects. Both the steric and elec-
tronic influences imparted by cyclopentadienyl substit-
uents govern the rate and outcome of alkane reductive
elimination reactions. In some cases, stable zirconocene
Prompted by these striking effects on chemical reac-
tivity, we sought to methodically characterize the stereo-
electronic properties of a class of 1,2-disubstituted
(1) For leading reviews on metallocene chemistry see: Metallo-
cenes: Synthesis, Reactivity and Applications; Togni, A., Halterman,
R. L., Eds.; Wiley-VCH: Weinheim, 1998.
(2) Coates, G. W. Chem. Rev. 2000, 100, 1223.
(3) Ewen, J . A. Sci. Am. 1997, 276, 86.
(6) Pool, J . A.; Lobkovsky, E.; Chirik, P. J . J . Am. Chem. Soc. 2003,
125, 2241.
(7) Manriquez, J . M.; Bercaw, J . E. J . Am. Chem. Soc. 1974, 96,
6229.
(4) Chen, M.-C.; Roberts, J . A. S.; Marks, T. J . J . Am. Chem. Soc.
2004, 126, 4605.
(5) Mo¨hring, P. C.; Coville, N. J . J . Organomet. Chem. 1994, 479, 1.
(8) Pool, J . A.; Lobkovsky, E.; Chirik, P. J . Nature 2004, 427, 527.
(9) Bradley, C. A.; Lobkovsky, E.; Chirik, P. J . J . Am. Chem. Soc.
2003, 125, 8110.
10.1021/om0495675 CCC: $27.50 © 2004 American Chemical Society
Publication on Web 09/29/2004

Zr and Fe Complexes with Indenyl Ligands
Organometallics, Vol. 23, No. 22, 2004 5333
F igu r e 1. Synthesis of 1,3-diisopropylindene.
indenyl ligands with sterically demanding silyl and
alkyl substituents. A similar, comprehensive study on
the electronic properties of cyclopentadienyl ligands in
the context of elucidating the “ansa-effect” in group 4
metallocene chemistry has recently appeared.10 The goal
of this study is to gain a fundamental understanding of
the features of assorted indenyl ligands such that
“zirconocenes” with specifically tailored properties and
reactivity patterns can be synthesized. X-ray diffraction,
infrared spectroscopy, and dynamic NMR studies pro-
vided insight into the steric influence of the family of
ligands, while cyclic voltammetry and infrared spec-
troscopy were used as a measure of ligand electronics.
condensation with acetone in the presence of KOH
furnished 3-isopropyl-8,8-dimethylbenzofulvene. This
result contrasts a previous report where an isopropyl-
idene-linked bis-indene was claimed from a similar
procedure.15 The synthesis of the desired 1,3-diisopro-
pylindene was completed by reduction of 3-isopropyl-
8,8-dimethylbenzofulvene with NaBEt3H. Attempts to
reduce the fulvene with LiAlH4 yielded a mixture of
products.
With a range of indenyl ligands in hand, homoleptic
bis-indenyl zirconocene dichlorides were prepared in a
straightforward manner by reacting 2 equiv of the
appropriate indenide anion with ZrCl4 in diethyl ether
(eq 1). Each of the compounds prepared is shown in
Figure 2 along with a shorthand designation and
isolated yield. It should be noted that the synthesis of
4-Cl2 yielded a near equimolar mixture of rac and meso
isomers that were not separated.
Resu lts a n d Discu ssion
Syn th esis of Bis-in d en yl Zir con ocen e Dich lor id e
Com p lexes. Our studies began with the synthesis of a
variety of tetrasubstituted bis-indenyl zirconium dichlo-
ride complexes where the substituents on the indenyl
ligands were systematically varied. The requisite 1,3-
disilyl-substituted indenide anions were prepared in a
straightforward manner by treatment of the appropriate
lithium indenide with various silyl chlorides.11 Complete
experimental details for each synthesis can be found in
the Supporting Information. We also targeted prepara-
tion of a 1,3-dialkyl-substituted indene. Reports of the
synthesis of the 1,3-diisopropylindenide anion and its
coordination chemistry have appeared previously,12,13
although to our knowledge a detailed synthesis of the
ligand has not been described. Our preparation com-
menced with the condensation of acetone with com-
mercially available indene in the presence of stoichio-
metric pyrollidine base to cleanly afford 8,8-dimethyl-
benzofulvene in 68% yield (Figure 1).14 Subsequent
reduction with LiAlH4 followed by careful hydrolysis
provided 3-isopropylindene. In our hands, a second
The synthesis of “mixed ring” bis-indenyl zirconocene
dichlorides was also achieved. Preparation of the req-
uisite mono-indenyl, piano stool complex, [(η5-C9H5-1,3-
(SiMe3)2)ZrCl3‚(Et2O)2LiClx]n, was accomplished by ad-
dition of 1 equiv of Li[C9H5-1,3-(SiMe3)2] to ZrCl4 in
1
diethyl ether (eq 2). Both H and 13C NMR spectroscopy
confirmed the presence of 2 equiv of diethyl ether.
Satisfactory elemental analyses were not obtained due
to coordination of variable amounts of LiCl from sample
to sample. While isolated material was effective for
subsequent transformations, we discovered that in situ
generation of the piano stool compound was the most
convenient and reproducible method for the synthesis
of the desired mixed ring bis-indenyl zirconocene dichlo-
ride complexes. In this manner, the tetrasubstituted
compounds 6-Cl2 and 7-Cl2 were isolated in modest
yields (Figure 2). This procedure was also extended to
include the synthesis of trisubstituted complexes such
(10) Zachmanoglou, C. E.; Bridgewater, B. M.; Parkin, G. E.;
Brandow, C. G.; Bercaw, J . E.; J ardine, C. N.; Lyall, M.; Green, J . C.;
Kiester, J . B. J . Am. Chem. Soc. 2002, 124, 9525.
(11) Brady, E. D.; Overby, J . S.; Meredith, M. B.; Mussman, A. B.;
Cohn, M. A.; Hanusa, T. P.; Yee, G. T.; Pink, M. J . Am. Chem. Soc.
2002, 124, 9556.
(12) Overby, J . S.; Hanusa, T. P.; Sellers, S. P.; Yee, G. T. Organo-
metallics 1999, 18, 3561.
(13) Overby, J . S.; Hanusa, T. P. Organometallics 1996, 16, 2212.
(14) Stone, K. J .; Little, R. D. J . Org. Chem. 1984, 49, 1849.
(15) Nifant’ev, I. E.; Ivchenko, P. V.; Kuz’ina, L.; Luzikov, Y. N.;
Sitnikov, A. A.; Sizan, O. E. Synthesis 1997, 469.

5334 Organometallics, Vol. 23, No. 22, 2004
Bradley et al.
F igu r e 2. Labeling scheme and isolated yields of bis-indenyl zirconocene dichlorides.
as 8-Cl2 and 9-Cl2 (Figure 2). All of the bis-indenyl
zirconocene dichloride complexes can be easily isolated
as bright yellow, slightly air-sensitive crystalline solids
in multigram quantities.
structures,10 where the five-membered rings of the
indenyl ligands are essentially eclipsed and four carbon
atoms from each ring lie over the ZrCl2 fragment. The
most hindered example crystallographically character-
ized, 2-Cl2, has one silyl phenyl substituent oriented
above an indenyl ring with the other directed away from
the metallocene wedge. In contrast to tetrasubstituted
zirconocenes, the indenyl ligands in the trisubstituted
zirconocene dichloride, 9-Cl2, become more eclipsed due
to the removal of one of the large tertiary silyl or alkyl
substituents. In this case the structure can be catego-
rized as a class III zirconocene,10 where the five-
membered rings stagger and there are three carbons
over the ZrCl2 portion of the molecule. The observed
bond lengths in 1-Cl2, 2-Cl2, 7-Cl2, and 9-Cl2 are typical
for group 4 bent metallocene derivatives1 and can found
in the Supporting Information.
Each of the dichloride complexes was characterized
1
by H and 13C NMR spectroscopy, elemental analysis,
and in many cases, single-crystal X-ray diffraction. For
1-Cl2, 2-Cl2, 3-Cl2, and 5-Cl2, 1H NMR spectra in
benzene-d6 display spectral features diagnostic of C2v
symmetric bent zirconocene derivatives. For example,
the 1H NMR spectrum of 1-Cl2 exhibits one [SiMe3]
resonance, one cyclopentadienyl resonance, and two
indenyl resonances, consistent with rapidly rotating
indenyl ligands. This dynamic behavior was observed
even at temperatures as low as -80 °C in toluene-d8.
The lower symmetry complexes such as 6-Cl2, 7-Cl2,
8-Cl2, and 9-Cl2 exhibit the expected number of 1H and
13C resonances, the details of which can be found in the
Experimental Section and Supporting Information.
The X-ray crystal structures of 1-Cl2, 2-Cl2, 7-Cl2, and
9-Cl2 are shown in Figures 3-6. Crystals of 1-Cl2
contain two independent, enantiomeric molecules in the
asymmetric unit that differ slightly by the orientation
of one of the indenyl rings. The two bulky rings are
rotated in a perpendicular gauche-type arrangement to
avoid steric interactions between the [SiMe3] substitu-
ents. Similar conformations are observed for 2-Cl2 and
7-Cl2. These metallocenes can be classified as class IV
A more quantitative comparison of the structural
parameters for each of the crystallographically charac-
terized zirconocene dichlorides is presented in Tables 1
and 2. Using the methods originally described by Faller
and Crabtree16 and later expanded by Veiros,17 the
hapticity and relative orientation of the indenyl ligands
can be defined (Table 1). The parameters for (η5-C9H7)2-
ZrCl2 and (η5-C9Me7)2ZrCl2 are also included for
comparison. The slip distortion (∆M-C) and the fold
angle (Ω) are used as a measure of the hapticity of the
indenyl ligands. The first quantity (∆M-C) is defined
as the difference between the average distance from the
metal center to the carbon atoms shared by the five-
and six-membered rings (C(4) and C(9), Table 1) and
18
19
(16) Faller, J . W.; Crabtree, R. H.; Habib, A. Organometallics 1985,
4, 929.
(17) (a) Calhorda, M. J .; Felix, V.; Veiros, L. Coord. Chem. Rev. 2002,
230, 49. (b) Parkin has also updated these parameters: Trnka, T. M.;
Bonanno, J . B.; Bridgewater, B. M.; Parkin, G. Organometallics 2001,
20, 3255.
(18) Repo, T.; Klinga, M.; Mutikainen, L.; Su, Y.; Leskela, M.;
Polamo, M. Acta Chem. Scand. 1996, 50, 1116.
(19) O’Hare, D.; Murphy, V.; Diamond, G.; Arnold, P.; Mountford,
P. Organometallics 1994, 13, 4689.

Zr and Fe Complexes with Indenyl Ligands
Organometallics, Vol. 23, No. 22, 2004 5335
F igu r e 3. (a) Molecular structure of 1-Cl2 with 30% probability ellipsoids. Hydrogen atoms are omitted for clarity. (b)
Top view of the molecule with hydrogen atoms and silyl methyl groups omitted for clarity.
F igu r e 5. Molecular structure of 7-Cl2 with 30% prob-
ability ellipsoids. Hydrogen atoms are omitted for clarity.
F igu r e 4. Molecular structure of 2-Cl2 with 30% prob-
ability ellipsoids. Hydrogen atoms are omitted for clarity.
the adjacent carbon atoms unique to the five-membered
ring (C(1), C(2), and C(3), Table 1). Typically metallo-
cenes with slip distortions less than 0.30 Å are consid-
ered to have η5 indenyl ligands.16 The fold angle (Ω) is
defined as the angle between the planes formed by the
carbons unique to the five-membered ring (C(1)-C(2)-
C(3)) and C(1)-C(3)-C(4)-C(9) (Table 1). Values of Ω
less than 8° are usually associated with η5 coordination.

5336 Organometallics, Vol. 23, No. 22, 2004
Bradley et al.
Ta ble 2. Ca lcu la ted Geom etr ica l Da ta for
Bis-in d en yl Zir con ocen e Dich lor id e, Dica r bon yl
Com p lexes, a n d Selected Cyclop en ta d ien yl
Der iva tives
compound
1-Cl2
R (deg)
â (deg)
γ (deg)
τ (deg)
47.7
47.9
48.5
44.7
50.6
33.9
27.2
53.5
54
59
51
41
43.7
59.3
41
132.3
132.1
131.5
135.3
129.4
146.1
152.8
126
126
121
129
121
135.9
135.9
135.3
137.6
133.3
151.7
150.1
129
128
129
130
128
1.8
1.9
1.9
1.2
2.0
2.8
-1.5
1.4
1.0
4.0
1.0
4.3
2.7
3.8
-0.4
2-Cl2
7-Cl2
9-Cl2
1-(CO)2
3-(CO)2
Cp2ZrCl2
a
a
a
Cp′2ZrCl2
Cpt2ZrCl2
a
a
a
Cp′′2ZrCl2
Cptt2ZrCl2
Cp*2ZrCl2
136
121
139
130
128
138
(η5-C9H7)2ZrCl2
(η5-C9Me7)2ZrCl2
a
b
Data taken from ref 10. Abbreviations used: Cp ) η5-C5H5,
Cp′ ) η5-C5H4-SiMe3, Cpt ) η5-C5H4-CMe3, Cp′′ ) η5-C5H3-1,3-
(SiMe3)2, Cptt ) η5-C5H3-1,3-(CMe3)2, Cp* ) η5-C5Me5.
The conformational preference of the indenyl ligands
in the dichloride complexes can be described by the
rotational angle (RA).17 This quantity is the dihedral
angle between the two planes of the indenyl defined by
the metal-C(2) and the midpoint of the C(4)-C(9)
carbons. A rotational angle of 0° corresponds to an
eclipsed conformation, whereas a value of 180° indicates
staggered indenyl ligands. It is important to note that
for bent metallocenes such as those reported in this
study the value of RA is also a measure of the angle
between the rings. The data contained in Table 1
demonstrate the structural similarity of the tetrasub-
stituted bis-indenyl zirconocene dichloride complexes.
For 1-Cl2, 2-Cl2, and 7-Cl2, rotational angles near 90°
are observed, indicative of an essentially gauche ar-
rangement of the indenyl ligands. Similar observations
have been made for analogous chromium complexes11
as well as in sterically hindered bis-indenyl iron and
cobalt compounds.20 The gauche conformation maxi-
mizes the distance between the sterically demanding
substituents on adjacent indenyl rings. In contrast,
9-Cl2 has a rotational angle of approximately 120°,
demonstrating that rotational preference of each com-
plex is dictated by the large, isotropic silyl or alkyl
substituents. In all examples, crystal packing effects
may also influence some of the observed rotational
angles.
F igu r e 6. (a) Molecular structure of 9-Cl2 with 30%
probability ellipsoids. Hydrogen atoms are omitted for
clarity. (b) Top view of the molecule with hydrogen atoms
and silyl and tert-butyl substituents omitted for clarity.
Ta ble 1. Str u ctu r a l Da ta for Bis-in d en yl
Zir con ocen e Dich lor id e a n d Dica r bon yl Com p lexes
slip distortion
fold angle
rotational
compound
1-Cl2
∆M-C (Å)
Ω (deg)
angle (deg)
0.121(6), 0.126(4) 3.3(2), 5.1(2)
0.106(5), 0.119(4) 2.994), 3.7(2)
0.110(2), 0.098(2) 4.4(2), 4.5(2)
0.149(3), 0.104(3) 5.2(3), 3.7(2)
0.071(6), 0.098(6) 3.5(5), 4.2(4)
0.130(4), 0.114(3) 4.2(2), 3.6(2)
0.057(8), 0.022(8) 0.6(6), 0.5(6)
0.115(6), 0.112(5) 5.2(4), 6.2(5)
89.7(5)
95.3(5)
85.3(2)
79.8(2)
120.9(4)
86.6(3)
12.0(5)
36.3(5)
88.7(4)
2-Cl2
7-Cl2
9-Cl2
1-(CO)2
3-(CO)2
(η5-C9H7)2ZrCl2
The angular parameters, R, â, γ, and τ, typically
defined for bent bis-cyclopentadienyl complexes10 have
also been computed (Table 2). Values for several Cp
compounds are also presented in Table 2 for comparison.
The angular parameters for the tetrasubstituted com-
pounds 1-Cl2, 2-Cl2, and 7-Cl2 are similar, indicating
(η5-C9Me7)2ZrCl2 0.039(4), 0.041(4) 5.9(4), 6.4(3)
The range of the slip distortions, 0.071-0.149 Å, and
fold angles, 2.9-5.2°, are consistent with η5 hapticity
for all of the indenyl ligands in the dichloride complexes
crystallographically characterized in this study.
(20) Calhorda, M. J .; Veiros, L. F. J . Organomet. Chem. 2001, 635,
197.

Zr and Fe Complexes with Indenyl Ligands
Organometallics, Vol. 23, No. 22, 2004 5337
little variation in geometry as a function of tertiary
substituent. As is known in cyclopentadienyl chemis-
try,10 the large substituents on the indenyl rings orient
to avoid steric interactions in the narrow portion of the
bent metallocene, producing a more linear arrange-
ment of the rings. The values of R and â for these
molecules are similar to those observed for Cp*2ZrCl2
and CpMe4Et2ZrCl2,10 suggesting that disubstitution on
an indenyl ligand with tertiary silyl groups produces a
ligand of similar steric disposition to a peralkylated
cyclopentadiene. Deviation from this trend is observed
for 9-Cl2, where replacement of one large substituent
with a hydrogen atom results in decreased values of â
and γ.
Eva lu a tion of th e Electr on ic In flu en ce of th e
In d en yl Liga n d s. Having examined the structural
features of several zirconocene dichlorides, a quantita-
tive comparison of the electronic properties imparted
by each substituted indenyl ligand was desired. Previous
work with bis-cyclopentadienyl zirconium complexes has
established that carbonyl stretching frequencies of
(CpRn)2Zr(CO)2 derivatives are a reliable measure of
relative ligand electronics.10 Preparation of the requisite
bis-indenyl zirconium dicarbonyl complexes was ac-
complished by reduction of the corresponding dichloride
with magnesium under an atmosphere of CO (eq 3).
Each zirconocene dicarbonyl complex was isolated as a
green, microcrystalline solid in moderate yield.
9
F igu r e 7. Top view of 1-(CO)2 with 30% probability
ellipsoids. Hydrogen atoms and silyl methyl groups are
omitted for clarity.
The homoleptic zirconocene dicarbonyl complexes
1-(CO)2, 2-(CO), 3-(CO)2, and 5-(CO)2 display the
expected number of 1H and 13C NMR resonances for C2v
symmetric molecules with rapidly rotating indenyl
rings. This behavior is preserved even upon cooling to
-80 °C. As observed for 4-Cl2, a near equimolar mixture
of rac and meso isomers of 4-(CO)2 was also isolated.
This synthetic procedure was extended to include the
“mixed ring” derivatives 7-(CO)2, 8-(CO)2, and 9-(CO)2.
The solid state structure of 1-(CO)2 has been com-
municated,9 and the indenyl ligands adopt a conforma-
tion similar to that observed for 1-Cl2. As Figure 7 and
the parameters in Tables 1 and 2 illustrate, the molecule
adopts a gauche arrangement with a rotational angle
of 86.6(3)°. Comparing the structures of 1-Cl2 and
1-(CO)2 demonstrates a contraction of R with con-
comitant expansion of â upon substitution of chloride
for carbon monoxide. Such an effect is consistent with
the metallocene becoming more linear as the steric
congestion in the wedge is reduced. This behavior has
also been noted upon replacement of chloride with
hydride ligands.21
F igu r e 8. (a) Molecular structure of 3-(CO)2 with 30%
probability ellipsoids. Hydrogen atoms are omitted for
clarity. (b) Top view with hydrogen atoms and silyl sub-
stituents omitted for clarity.
The zirconocene dicarbonyl complex containing large
tert-butyldimethylsilyl substitutents, 3-(CO)2, was also
characterized by X-ray diffraction. The solid state
structure is shown in Figure 8 and exhibits a somewhat
unexpected ligand conformation. The silyl groups are
rotated such that the tert-butyl substituents are oriented
toward the most open regions of the zirconocene, above
and below the planes of their respective indenyl ligands.
This conformation places the silyl methyl groups toward
the interior of the molecule. The indenyl ligands are
essentially eclipsed, with a rotational angle of only
(21) Chirik, P. J .; Day, M. W.; Bercaw, J . E. Organometallics 1999,
18, 1873.

5338 Organometallics, Vol. 23, No. 22, 2004
Bradley et al.
Ta ble 3. Ca r bon yl Str etch in g F r equ en cies of
Bis-in d en yl Zir con iu m Dica r bon yl Com p lexes
Recor d ed in P en ta n e
com-
pound ν(CO)sym ν(CO)sym ν(CO)sym ν(CO)asym ν(CO)asym ν(CO)asym
1-(CO)2 1980a
2-(CO)2 1982
3-(CO)2 1976
4-(CO)2 1974
5-(CO)2 1952
7-(CO)2 1976
8-(CO)2 1980
9-(CO)2 1977
1964
1965
1956
1951
1956
1957
1901
1902
1895
1893
1859
1895
1902
1896
1882
1883
1874
1863
1874
1877
1956
1968
1965
1874
1885
1880
a
Frequencies denoted in italics indicate the major isomer.
cm-1 to the value of 1915.5 cm-1 reported for (η5-C5H3-
1,3-(CHMe2)2)2Zr(CO)2 suggests that, in this case, the
indenyl ligand is more electron donating than the
corresponding cyclopentadiene.10 In fact, the electronic
environment of 5-(CO)2 most closely resembles that of
a tetrasubstituted cyclopentadiene such as (η5-C5Me4H)2-
Zr(CO)2, which has an average carbonyl stretching
frequency of 1904.5 cm-1 in pentane solution.10
The infrared spectra of 7-(CO)2, 8-(CO)2, and 9-(CO)2
contained four carbonyl bands. For 7-(CO)2 the bands
are approximately of equal intensity, whereas for 8-(CO)2
and 9-(CO)2, a set of major and minor bands are
observed (Table 3, Figure 9b). The zirconocene dicar-
bonyl complexes with larger indenyl ligands, 1-(CO)2
and 2-(CO)2, exhibit a total of six carbonyl stretches.
Similar observations were made in the solid state (KBr);
however the resolution of the peaks is significantly
reduced (Supporting Information). Despite the differ-
ence in the time scale of the IR experiment and the
rotation of cyclopententadienyl ligands, to our knowl-
edge this is the first observation of different rotamers
of group 4 metallocene dicarbonyl complexes by IR
spectroscopy.23
The spectrum of 1-(CO)2 in pentane solution exhibits
two major bands centered at 1980 and 1901 cm-1, as
well as four smaller peaks of approximately equal
intensity (Figure 9c). To rule out the possibility that the
additional bands were due to impurities, the single
crystals used for both elemental analysis and X-ray
diffraction9 were examined by infrared spectroscopy and
produced identical spectra. Furthermore, the relative
intensities of the peaks were consistent from sample to
sample, also arguing against impurities. The additional
carbonyl bands arise from population of three different
rotamers that have measurably different vibrational
spectra (Figure 10).23 Observation of different rotamers
in solution complements Waymouth’s recent report on
the application of on-resonance spin-lattice relaxation
in the rotating frame to measure the rate of rotation in
several aryl-substituted bis-indenyl zirconocenes.24
Starting with the gauche arrangement observed in the
solid state structure of 1-(CO)2, rotamers resulting from
90° rotation of one indenyl ligand are considered for
simplicity, although other intermediate rotations are
certainly plausible. Regardless of rotamer structure, it
F igu r e 9. Infrared spectra of (a) 5-(CO)2, (b) 8-(CO)2, and
(c) 1-(CO)2.
12.0(5)° (Table 1). In this orientation, all four silyl
substituents avoid narrow portions of the metallocene
and unfavorable transannular interactions with the
benzo fragments of the indenes. This structure also
produces a more linear zirconocene with values of R and
â equal to 27.2 and 152.8°, respectively, as compared
to 33.9 and 146.1° in 1-(CO)2, where the indenyl ligands
adopt the more familiar gauche arrangement (Table 2).
A similar eclipsed indenyl conformation is observed for
(η5-C9H7)2Zr(CO)2,22 although in this case the two rings
are oriented approximately 90° from the metallocene
wedge whereas in 3-(CO)2 the benzo rings are anti,
approximately 180°, from the Zr(CO)2 fragment.
The electronic environment imparted by each indenyl
ligand was assessed by infrared spectroscopy. Repre-
sentative spectra are shown in Figure 9, and the
carbonyl stretching frequencies are compiled in Table
3. As typically observed for idealized C2v symmetric bent
zirconocene dicarbonyls, 5-(CO)2 exhibits two intense
CO bands centered at 1952 and 1859 cm-1, correspond-
ing to a symmetric and asymmetric stretch, respectively
(Figure 9a). Comparing the average value of 1905.5
(23) For examples in iron and chromium chemistry see: (a) Mackie,
S. C.; Baird, M. C. Organometallics 1992, 11, 3712. (b) DeGrace, S.
A.; Berry, S. W.; Mackie, S. C. J . Organomet. Chem. 1998, 560, 63. (c)
Polowin, J .; Mackie, S. C.; Baird, M. C. Organometallics 1992, 11, 3724.
(24) Wilmes, G. M.; France, M. B.; Lynch, S. R.; Waymouth, R. M.
Organometallics 2004, 23, 2405.
(22) Rausch, M. D.; Moriarity, K. J .; Atwood, J . L.; Hunter, W. E.;
Samuel, E. J . Organomet. Chem. 1987, 327, 39.

Zr and Fe Complexes with Indenyl Ligands
Organometallics, Vol. 23, No. 22, 2004 5339
F igu r e 11. Pentane solution infrared spectrum of 3-(CO)2.
minor bands appear at 1976 and 1895 cm-1. We at-
tribute this difference to population of only two rotamers
for this complex instead of the three observed for
1-(CO)2 and 2-(CO)2. Recall that the solid state struc-
ture of 3-(CO)2 exhibits an anomalous, nearly eclipsed
rotamer, which contrasts the gauche arrangements
observed for other crystallographically characterized
tetrasubstituted zirconocene derivatives. It is also pos-
sible that other rotamers are indeed populated but have
CO stretching frequencies that are not resolved by the
IR experiment.
Observation of two sets of bands for the trisubstituted
zirconocene dicarbonyls, 8-(CO)2 and 9-(CO)2, is con-
sistent with population of two vibrationally distinct
rotamers. Only two bands are observed for 5-(CO)2,
indicative of either population of one rotamer or over-
lapping bands. Unlike isotropic tertiary silyl and tert-
butyl groups, the isopropyl substituents have a confor-
mational flexibility that allows the methine hydrogens
to point toward the interior of the molecule, reducing
transannular interactions. As a result, the barrier for
interconversion of rotamers may be reduced or popula-
tion of a single rotamer becomes more significant. A
similar trend was noted in the conformational energy
profiles of iron piano stool complexes, where more
hindered ligands resulted in higher barriers for rota-
tion.23
From these data, the electronic influence of each
indenyl ligand can be estimated, although absolute
ordering of the ligands is difficult due to the significant
differences in CO stretching frequencies observed for
each rotamer. For the trisubstituted zirconocene dicar-
bonyls, 8-(CO)2 and 9-(CO)2, comparing like bands
establishes that substitution of a [CMe3] group for a
[SiMe3] substituent produces a systematic red shift,
demonstrating the increased electron-donating charac-
ter of the [CMe3] group. Using a similar analysis, an
analogous trend can be discerned for the tetrasubsti-
tuted derivatives, 1-(CO)2, 7-(CO)2, and 4-(CO)2. Suc-
cessive replacement of [SiMe3] groups with [CMe3]
substituents results in progressive red shifting of the
stretching frequencies. These observations may be
rationalized by simple inductive effects and are sup-
ported by the Hammett σm values of -0.04 for [SiMe3]
and -0.10 for [CMe3].25
F igu r e 10. Gauche, eclipsed, and staggered rotamers of
1-(CO)2 and 2-(CO)2. Silyl substituents are omitted for
clarity.
is remarkable that slight changes in the orientational
preferences of the indene ligands produce such large
(∼15 cm-1) changes in the frequencies of the carbonyl
bands. On the basis of the crystallographic data (Figure
7), we tentatively attribute the major bands to rotamer
A, although a detailed computational study will be
required for definitive assignment. Averaging the fre-
quencies of the major bands observed experimentally
produces a value of 1939.5 cm-1, which is blue shifted
15 cm-1 from the corresponding cyclopentadienyl com-
plex, (η5-C5H3-1,3-(SiMe3)2)2Zr(CO)2, suggesting the in-
denyl derivative is electron withdrawing relative to the
Cp in this case. Thus incorporation of two silyl substit-
uents on an indenyl anion produces a ligand with steric
properties that resemble permethylcyclopentadiene but
with electron-withdrawing properties that are virtually
unparalleled in alkyl- or silyl-substituted Cp chemistry.
The minor rotamers produce a more electron rich
zirconium environment.
Similar infrared spectroscopic features are observed
for 2-(CO)2 and can also be rationalized by population
of three vibrationally distinct rotamers in analogy to
1-(CO)2. The similarity of the frequencies for 1-(CO)2
and 2-(CO)2 suggests that similar rotamers are ob-
served. Although the major isomer in 2-(CO)2 displays
bands at 1982 and 1902 cm-1, the intensity of the minor
peaks approaches those of the major. Attempts to
measure perturbations to the equilibrium population as
a function of temperature (22-75 °C) for 1-(CO)2 by
React-IR spectroscopy have been unsuccessful.
The solution IR spectrum of the most hindered
member of the series, 3-(CO)2, exhibits predominantly
two CO stretches, although a very minor set of ad-
ditional peaks is also detectable (Figure 11). The major
bands are centered at 1956 and 1874 cm-1, while the
The data collected in Table 3 also allow comparison
of the electronic properties of the silyl substituents.
(25) Hansch, C.; Leo, A.; Taft, R. W. Chem. Rev. 1991, 91, 165.

5340 Organometallics, Vol. 23, No. 22, 2004
Bradley et al.
F igu r e 12. Labeling scheme and isolated yields of bis-indenyl iron(II) compounds.
Ch a r t 1
Replacing [SiMe3] groups with [SiMe2CMe3] substitu-
ents results in a slight reduction of the CO stretching
frequencies, indicating a small but measurable increase
in electron donation in moving from 1-(CO)2 to 3-(CO)2.
The opposite effect is observed with [SiMe2Ph], where
a slight increase in carbonyl stretching frequency is
observed when comparing 1-(CO)2 with 2-(CO)2, dem-
onstrating that C9H5-1,3-(SiMe2Ph)2 engenders the most
electrophilic zirconium center. Unfortunately, IR spec-
troscopy is unable to differentiate between 3-(CO)2 and
4-(CO)2 (Chart 1).
P r ep a r a tion of Bis-in d en yl Ir on Com p lexes. Due
to the complications arising from the observation of
different numbers of carbonyl bands as a function of
indenyl ligand, we sought an independent measure of
ligand electronics. Because the preparation of bis-
cyclopentadienyl iron(II) complexes is trivial and their
reversible electrochemistry is well established, we tar-
geted the synthesis of the corresponding series of bis-
indenyl iron(II) compounds. Following Wilkinson’s origi-
nal procedure for Ind2Fe (Ind ) η5-C9H7), which calls
for addition of 2 equiv of lithium indenide to an ethereal
slurry of FeCl2,26 dibenzoferrocene complexes with both
mono- and disubstituted indenyl ligands were prepared
(Figure 12).
stereomers by successive recrystallizations have been
unsuccessful. Incorporation of a larger [SiMe2CMe3]
substituent produced only one diastereomer of 12 upon
metalation with FeCl2. On the basis of our current data,
we are unable to definitively assign the stereochemistry,
although steric arguments would suggest that the rac
isomer should be favored.
Dibenzoferrocene derivatives containing 1,2-disubsti-
tuted indenide anions were also prepared (Figure 12).
As with 10-12, these molecules were isolated in good
yield as blue-green, crystalline solids. For 17, an equi-
molar mixture of rac and meso diastereomers were
obtained and not separated. The solid state structures
of both 13 and 17 were determined by X-ray diffraction
and are shown in Figures 13 and 14, respectively. For
17, only the rac diastereomer was observed in the
1
lattice. Because H NMR spectra of the crystals exhib-
ited resonances for both the rac and meso isomers of
17, we believe that serendipitous selection of the crystal
resulted in crystallographic characterization of pure rac-
17. No evidence for disorder in the structure was
detected.
The angular parameters R, â, γ, and τ as well as the
rotational angles have been computed for both 13 and
17 and are presented in Table 4. In both complexes, the
slip distortions (∆M-C) and fold angles (Ω) are indica-
tive of η5-coordinated indenyl ligands. While the fold
angles for 13 and 17 are quite similar to those observed
with the zirconocene dichlorides, the slip distortions are
considerably smaller. The angular parameters R and â
indicate a slight bending of 13 and 17 as compared to
Both (η5-C9H6-SiMe3)2Fe (10) and (η5-C9H6-CMe3)2Fe
(11) were synthesized and isolated as blue-green solids
in good yields as near equimolar mixtures of dia-
stereomers (Figure 12). Attempts to separate the dia-
(26) Wilkinson, G.; Pauson, P. L. J . Am. Chem. Soc. 1954, 70, 2024.

Zr and Fe Complexes with Indenyl Ligands
Organometallics, Vol. 23, No. 22, 2004 5341
F igu r e 14. Molecular structure of rac-14 with 30% prob-
ability ellipsoids. Hydrogen atoms are omitted for clarity.
indenyl doublets (Figure 15), suggesting a time-aver-
aged C2v symmetric metallocene arising from fast ring
rotation on the time scale of the NMR experiment.
Cooling the sample (T ) 210 K, Figure 15) resulted in
broadening of the [SiMe3] and indenyl resonances,
ultimately resolving into separate peaks. At 180 K, the
lowest temperature accessible experimentally, near
baseline resolution of the [SiMe3] groups and one set of
indenyl resonances is observed, suggesting restricted
indenyl rotation. A similar temperature profile is ob-
served in dichloromethane-d2. Line shape analysis28 was
used to determine rate constants of the dynamic process,
and these values were used to construct an Eyring plot
(Supporting Information). As a reference value, a rate
constant of 23 s-1 was measured at -83 °C. From these
data, activation parameters of ∆Hq ) 11.3(5) kcal/mol
and ∆Sq ) 4(2) eu were obtained for ring rotation. These
values are in good agreement with activation energies
of 11.0 and 15.3 kcal/mol reported for (η5-C5H2-1,2,4-
(SiMe3)3)2Fe29 and (η5-C5H2-1,2,4-(CMe3)3)2Fe,30 respec-
tively.
F igu r e 13. (a) Molecular structure of 13 with 30%
probability ellipsoids. Hydrogen atoms are omitted for
clarity. (b) Top view of the molecule with hydrogen atoms
and silyl methyl groups omitted for clarity.
Ind2Fe26 or Ind*2Fe27 (Table 4). In addition, the [SiMe3]
substituents of 13 are bent 13.5° and 16.6° out of the
plane of the indenyl ligand. A similar observation is
made for rac-17, where the [CMe3] groups are displaced
15.4° and the [SiMe3] groups 17.9° from the plane of
the ligand. The rotational angles of 85.2(2)° and 86.7(2)°,
respectively, highlight the gauche conformation in the
solid state and are comparable to the zirconocene
dichloride structures. All of these observations are a
result of the increasing steric demand of the ligand due
to the bulky [SiMe3] and [CMe3] substituents. Similar
geometric preferences have been observed by Hanusa
in related bis-indenyl chromium complexes11 and by
Treichel for [(η5-C9H5-1,3-(CH3)2)2Fe]PF6.27
The solution behavior of the substituted dibenzo-
ferrocenes was examined by variable-temperature 1H
NMR spectroscopy with the goal of understanding
indenyl ligand conformational dynamics. At 23 °C, the
1H NMR spectrum of 13 exhibits equivalent [SiMe3]
groups, one cyclopentadienyl hydrogen, and two distinct
The other tetrasubstituted dibenzoferrocenes display
1
similar H NMR dynamics. Unfortunately, peak overlap
and spectral complexity for 14 and 15 prevent a quanti-
tative line shape analysis and extraction of activation
parameters. For 14, four resonances can be observed at
180 K for the two sets of diastereotopic methyl groups
on the [SiMe2Ph] substituents. On the basis of the gross
features of the data, the barrier to ring rotation is on
the same order of magnitude as 13, suggesting similar
steric environments imparted by each ligand.
The cyclopentadienyl region of the toluene-d8 1H NMR
spectrum of the mixture of rac and meso isomers of 17
is shown in Figure 16. At ambient temperature, the
spectrum exhibits two cyclopentadienyl resonances of
approximately equal area, indicative of an equimolar
Ta ble 4. Str u ctu r a l P a r a m eter s for Bis-in d en yl Ir on (II) Com p lexes
compound
∆(M-C) (Å)
Ω (deg)
RA (deg)
R (deg)
â (deg)
γ (deg)
τ (deg)
13
0.058(5), 0.066(4)
0.065(3), 0.065(3)
0.054(6), 0.051(5)
0.029(4)
3.2(3), 3.9(4)
3.7(2), 3.2(2)
2.6(4), 2.7(5)
2.4(4)
85.2(5)
86.7(2)
13.0(4)
151.3(3)
5.8
7.2
3.2
0.3
174.2
172.8
176.8
179.7
178
1.9
2.25
1.2
rac-17
Ind2Fe26
Ind*2Fe34
177.3
179.2
179.2
-0.25

5342 Organometallics, Vol. 23, No. 22, 2004
Bradley et al.
1
F igu r e 15. Variable-temperature H NMR spectra of 13 in toluene-d8.
1
F igu r e 16. Cyclopentadienyl region of the H NMR spectrum of rac/meso-17 in toluene-d8 as a function of temperature.
mixture of the two diastereomers in a fast exchange
regime. At the lowest temperature examined, 180 K,
four distinct cyclopentadienyl peaks are observed, two
of which are of equal area (Figure 16). Warming the
sample to 195 K resulted in coalescence of these two
resonances with concomitant broadening of the remain-
ing two peaks. The resonances coalesce at 223 K,
suggesting interconversion of two rotamers on the NMR
time scale. We interpret these data in terms of the
model presented in Figure 17. On the basis of crystal-
lographic data and simple steric arguments, gauche
rotamers should be observed in preference to eclipsed
or staggered isomers due to unfavorable interactions
between the bulky [SiMe3] and [CMe3] substituents in
these orientations. If only gauche rotamers are present
at the static limit, rac-17 would be expected to exhibit
two cyclopentadienyl resonances due to the distinct
rotamers A and C (Figure 17). Because of different
transannular interactions, the populations of A and C
could be measurably different. A similar analysis on
meso-17 predicts two cyclopentadienyl hydrogens in the
static limit. In this case, rotamers A and C are equiva-
lent, and therefore the two cyclopentadienyl peaks must
be of equal intensity. The spectra presented in Figure
16 are consistent with this model and clearly establish
restricted ring rotation at low temperatures for the two
diastereomers of 17.
(27) Treichel, P. M.; J ohnson, J . W.; Calabrese, J . C. J . Organomet.
Chem. 1975, 88, 215.
(28) Marat, K. Spinworks; University of Manitoba, 2004.
(29) Okuda, J .; Herdtweck, E. Chem. Ber. 1988, 121, 1899.
(30) Abel, E. W.; Long, N. J .; Orrell, K. G.; Osborne, A. G.; Sik, V.
J . Organomet. Chem. 1991, 403, 195.

Zr and Fe Complexes with Indenyl Ligands
Organometallics, Vol. 23, No. 22, 2004 5343
Ta ble 5. F or m a l P oten tia ls for a Ser ies of
Su bstitu ted Diben zofer r ocen esa
compound
E1/2 (mV)
Cp2Fe
10
12
Ind2Fe
11
14
13
15
17
16
0
-54
-71
-85
-115
-20
-118
-122
-153
-250
a
Reported versus ferrocene/ferrocenium.
All dibenzoferrocene derivatives examined exhibited
well-behaved cyclic voltammetric responses. The voltam-
metric profiles are those anticipated for a freely diffus-
ing redox couple that is chemically reversible. The peak
potential difference (∆Ep) is somewhat larger than the
anticipated value of 59 mV, but this is due to ohmic
losses in solution, not sluggish kinetics. This is evi-
denced by the fact that ferrocene itself, which is known
to show rapid interfacial charge transfer kinetics,
exhibited a similar response likely arising from the high
resistance of the solvent.
All of the formal potentials were negatively shifted,
relative to ferrocene, suggesting that every indenyl
ligand is more electron donating than Cp (Table 5),32,33
contrasting some of the results from the infrared study
where disilylated indenes appeared less electron donat-
ing. In agreement with the carbonyl stretching frequen-
cies, the relative electronic properties determined from
electrochemistry follow the trends anticipated from
inductive effects. In the series of disubstituted dibenzo-
ferrocenes 10-12, the relative electron-donating ability
of the indenyl substituents is [SiMe3] < [SiMe2CMe3]
< [CMe3].
A similar trend has been observed for the correspond-
ing ferrocenes, where silyl groups are electron with-
drawing and alkyl groups such as tert-butyl are electron
donating.34 Importantly, the formal potentials for silyl-
ated compounds 10 and 12 are shifted positive of Ind2Fe,
suggesting that the silyl groups are indeed electron
withdrawing relative to hydrogen. However, as is evi-
dent from the data in Table 5, the presence of additional
silyl substituents produces a shift in the formal poten-
tials to values that are negative of dibenzoferrocene.
This apparent contradiction, which suggests that the
second silyl group gives rise to a net electron-donating
effect, likely indicates that steric effects play a larger
role than anticipated. Similar behavior has been re-
ported by Okuda in silylated ferrocenes and was ra-
tionalized by the alleviation of unfavorable steric inter-
actions upon oxidation to the less congested ferrocenium
ion.34 While a similar steric argument may be used here,
the iron-carbon bond distances determined for 13 are
within experimental error of those reported for Ind2Fe26
and Ind*2Fe.35 However, the increased steric congestion
associated with 13, 14, 15, and 17 is manifested in the
F igu r e 17. Rotamers of rac- and meso-17.
Applying the model originally developed by Okuda for
[SiMe3]- and [CMe3]-substituted ferrocenes,29,31 the con-
formational dynamics of the tetrasubstituted dibenzo-
ferrocenes may be understood. Because the indenyl
substituents are isotropic, we believe that the barrier
of rotation arises from geared, synchronous rotations
of the bulky substituents. Thus to avoid unfavorable
transannular interactions, the [SiMe3] groups on one
indenyl ring of 13 must rotate simulataneously with the
other. Similar behavior has also been observed for 14
and the mixture of rac and meso isomers of 17.
Each of the dibenzoferrocenes reported in this study
was characterized by cyclic voltammetry in 0.1 M
TBAH/THF (TBAH ) tetrabutylammonium hexafluoro-
phosphate) solutions using platinum as counter and
working electrodes. A silver wire was used as a quasi-
reference electrode, and its potential was calibrated
against ferrocene.
(32) Treichel, P. M.; J ohnson, J . W.; Wagner, K. P. J . Organomet.
Chem. 1975, 88, 227.
(33) Robbins, J . L.; Edelstein, N.; Spencer, B.; Smart, J . C. J . Am.
Chem. Soc. 1982, 104, 1882.
(31) Thornberry, M. P.; Slebodnick, C.; Deck, P. A.; Fronczek, F. R.
Organometallics 2000, 19, 5352.
(34) Okuda, J .; Albach, R. W.; Herdtweck, E. Polyhedron 1991, 10,
1741.

5344 Organometallics, Vol. 23, No. 22, 2004
Bradley et al.
Ch a r t 2
solution NMR dynamics of these molecules, where
restricted ring rotation is observed at low temperature.
Only time-averaged spectra consistent with rapid ring
rotation are obtained for 10-12. Interestingly, the
formal potential for 14 is shifted to a more positive value
than Ind2Fe but not past the value for ferrocene,
suggesting that the electron-donating effects overcome
some of the steric effects.
tablished and should provide insight into understanding
the chemistry of their low-valent zirconium derivatives.
Exp er im en ta l Section
Gen er a l Con sid er a tion s. All air- and moisture-sensitive
manipulations were carried out using standard vacuum line,
Schlenk, or cannula techniques or in an M. Braun inert
atmosphere drybox containing an atmosphere of purified
nitrogen. Solvents for air- and moisture-sensitive manipula-
tions were initially dried and deoxygenated using literature
procedures.36 Benzene-d6 and toluene-d8 for NMR spectroscopy
were distilled from sodium metal under an atmosphere of
argon and stored over 4 Å molecular sieves or sodium metal.
Dichloromethane-d2 was distilled from CaH2 under vacuum
prior to use. Chlorotrimethylsilane was purchased from Acros
Organics and dried over CaH2 before use. Indene, Adogen-464,
tert-butyl bromide, SiMe2PhCl, SiMe2(CMe3)Cl, 1.6 M nBuLi
in hexanes, and 1.0 M NaBEt3H in toluene were purchased
from Acros and used as received. Iron(II) chloride and carbon
monoxide were purchased from Aldrich. Carbon monoxide was
The steric arguments discussed above may be used
to reconcile the differences between the electrochemical
and infrared spectroscopic data. From this study and
that previously reported by Okuda,34 it appears that the
electrochemical experiments are more sensitive to steric
effects and caution must be exercised when dealing with
highly substituted cyclopentadienyl or indenyl ligands
or conclusions about the absolute electronic properties
of a given ligand may be tenuous. Still, we believe that
the data presented in Table 5 are a reliable measure of
the relative electronic properties of a given ligand set,
as long as similar steric environments are maintained.
Thus, whereas general comparisons among disubstitut-
ed and tetrasubstituted dibenzoferrocenes might not be
as compelling, comparisons within each respective series
are. Because each of the tetrasubstituted dibenzoferro-
cenes display similar variable-temperature NMR spec-
troscopic profiles, we feel confident in establishing the
trend, shown in Chart 2, of the relative electron-
donating properties of each disubstituted indenyl ligand.
Importantly, these data are consistent with the trend
established by infrared spectroscopy for the zirconocene
dicarbonyl compounds.
passed through
a liquid nitrogen-cooled trap before use.
Preparation of the substituted indenyl ligands was carried out
using standard procedures, and more specific details can be
found in the Supporting Information. 1-Cl29 and 1-(CO)29 were
prepared according to previously reported procedures.
1H and 13C NMR spectra were recorded on a Varian Inova
400 spectrometer operating at 399.779 MHz (1H) and 110.524
MHz (13C). Variable-temperature 1H experiments were con-
ducted on a Varian Inova 500 spectrometer operating at 500.62
MHz. All chemical shifts are reported relative to SiMe4 using
1H (residual) or 13C NMR chemical shifts of the solvent as a
secondary standard. Infrared spectroscopy was conducted on
a Mattson RS-10500 Research Series FT-IR spectrometer
calibrated with a polystyrene standard. Cyclic voltammograms
(CVs) were collected using three-compartment electrochemical
cells, with Pt working and counter electrodes and Ag wire as
a reference in a drybox equipped with electrochemical outlets.
CVs were recorded using a Bioanalytical Systems CV-27
voltammograph and plotted with a Soltec VP-6423S X-Y
recorder. All CVs were run at a scan rate of 100 mV/s.
Solutions of the individual compounds were prepared by
charging 2-7 mg of the appropriate ferrocene and 0.500 g (1.3
mmol) of [n-Bu4N][PF6] in a vial and dissolving the solids in
13 mL of THF. This produced solutions of approximately 1 mM
in ferrocene and 0.1 M in electrolyte. After recording the
baseline of a standard 0.1 M solution of electrolyte, CVs were
collected for each of the different ferrocenes, including Cp2Fe.
Oxidation potentials were then referenced to the formal
potential of ferrocene. All compounds displayed reversible
electrochemical behavior.
Con clu sion s
The electronic properties of a family of tetrasubsti-
tuted zirconocene dicarbonyl and iron(II) sandwich
complexes containing silylated and alkylated indenyl
ligands have been studied by infrared spectroscopy and
electrochemical experiments, respectively. Both tech-
niques established that simple inductive effects are
useful in predicting the relative electron-donating or
-withdrawing capability of a ligand within a series.
Pentane solution infrared spectra of highly substituted
bis-indenyl zirconocenes allow observation of multiple
rotamers in solution with significantly different CO
stretching frequencies, the number of which depends on
the size of the indenyl substituents. Special care must
be taken in interpreting electrochemical data on these
types of molecules, as steric effects influence the mea-
sured formal potentials. Despite these complications, the
stereoelectronic properties of a family of synthetically
accessible disubstituted indene ligands have been es-
Single crystals suitable for X-ray diffraction were coated
with polyisobutylene oil in a drybox and were quickly trans-
ferred to the goniometer head of a Siemens SMART CCD area
detector system equipped with a molybdenum X-ray tube (λ
) 0.71073 Å). Preliminary data revealed the crystal system.
A hemisphere routine was used for data collection and deter-
mination of lattice constants. The space group was identified,
(35) Westcott, S. A.; Kakkar, A. K.; Stringer, G.; Taylor, N. J .;
Marder, T. B. J . Organomet. Chem. 1990, 394, 777.
(36) Pangborn, A.; Giardello, M.; Grubbs, R. H.; Rosen, R.; Timmers,
F. Organometallics 1996, 15, 1518.

Zr and Fe Complexes with Indenyl Ligands
Organometallics, Vol. 23, No. 22, 2004 5345
and the data were processed using the Bruker SAINT program
and corrected for absorption using SADABS. The structures
were solved using direct methods (SHELXS) completed by
subsequent Fourier synthesis and refined by full-matrix least-
squares procedures. Elemental analyses were performed at
Robertson Microlit Laboratories, Inc., Madison, NJ .
and the reaction mixture was stirred for 3 h. Once complete,
the flask was transferred outside of the drybox, and ap-
proximately 100 mL of H2O was added to quench any excess
NaBEt3H. A workup procedure similar to that described for
8,8-dimethylbenzofulvene was followed. Vacuum distillation
afforded 3.1 g (65%) of 1,3-diisopropylindene suitable for
deprotonation. 1H NMR (benzene-d6): δ 0.60 (d, 10 Hz, 3H,
CH(CH3)2), 0.99 (d, 10 Hz, 3H, CH(CH3)2), 1.21 (m, 6H,
CH(CH3)2), 2.17 (m, 1H, CH(CH3)2), 2.78 (m, 1H, CH(CH3)2),
3.24 (br s, 1H, Cp), 6.05 (s, 1H, Cp), 7.26 (m, 4H, Ind). 13C
NMR (benzene-d6): δ 17.54, 21.62, 21.98, 22.19, 27.21, 30.36
(CHMe2), 55.32, 119.644, 123.33 (Cp), 124.90, 126.54, 145.33,
148.22, 151.35 (Ind). One indene resonance not located.
P r ep a r a tion of (η5-C9H5-1,3-(SiMe2P h )2)2Zr Cl2 (2-Cl2).
A 500 mL round-bottomed flask was charged with 3.24 g (8.30
mmol) of Li[C9H5-1,3-(SiMe2Ph)2] and approximately 150 mL
of diethyl ether. The resulting yellow solution was chilled in
a cold well cooled to liquid nitrogen temperature, and 0.969 g
(4.16 mmol) of ZrCl4 was added. The reaction mixture was
stirred for 1 day, and the solvent was removed in vacuo. The
resulting solid was dissolved in dichloromethane and filtered
through a pad of Celite. Removal of dichloromethane in vacuo
yielded a brown oil that was recrystallized from diethyl ether
at -35 °C to afford 2.58 g (67%) of 2-Cl2 as a yellow solid.
Anal. Calcd for C50H54Si4ZrCl2: C, 64.61; H, 5.86. Found: C,
64.37; H, 5.45. 1H NMR (benzene-d6): δ 0.69 (s, 12H, SiMe2Ph),
0.75 (s, 12H, SiMe2Ph), 6.84 (m, 4H, Ind), 7.23 (s, 2H, Cp),
7.25 (m, 20H, Ph), 7.60 (m, 4H, Ind). 13C NMR (benzene-d6):
δ -0.37 (SiMe2Ph), -0.02 (SiMe2Ph), 119.89, 126.43 (Cp),
128.05, 128.39, 129.27, 134.09, 136.45, 138.34, 139.42 (Ind/
Ph).
P r ep a r a t ion of [(η5-C9H 5-1,3-(SiMe3)2)Zr Cl3‚(E t 2O)2‚
LiClx]n . A 100 mL round-bottomed flask was charged with
0.931 g (4.00 mmol) of ZrCl4, and approximately 50 mL of
diethyl ether was added. The slurry was chilled in a cold well
cooled to liquid nitrogen temperature for an additional 15 min.
The flask was removed from the cold well, and 1.07 g (3.98
mmol) of Li[C9H5-1,3-(SiMe3)2] was then added in small
portions over the course of several minutes. The reaction
mixture was allowed to warm to ambient temperature and
stirred overnight. The resulting orange-yellow solution was
filtered through Celite, and the solvent was removed in vacuo
to yield 2.02 g (91%) of an oily orange-yellow solid identified
as [(η5-C9H5-1,3-(SiMe3)2)ZrCl3‚(Et2O)2‚LiClx]. 1H NMR (benzene-
d6): δ 0.44 (s, 18H, SiMe3), 1.04 (br s, 12H, CH3CH2O), 3.24
(br s, 8H, CH3CH2O), 7.05 (br s, 2H, Ind), 7.38 (s, 1H, Cp),
7.79 (m, 2H, Ind). 13C NMR (benzene-d6): δ 0.41 (SiMe3), 15.06
(CH3CH2O), 65.98 (CH3CH2O), 121.42, 126.15 (Cp), 126.57,
136.64, 138.22 (Ind).
P r ep a r a tion of 8,8-Dim eth ylben zofu lven e. A 500 mL
round-bottomed flask was charged with 20.0 g (0.17 mol) of
indene and 19.0 mL (0.26 mol) of acetone. After the liquids
were diluted in approximately 150 mL of MeOH, 21.0 mL (0.25
mol) of pyrollidine was added. The reaction mixture was stirred
for 24 h at room temperature. The excess pyrollidine was then
quenched with 21 mL (0.25 mol) of concentrated acetic acid.
Water (300 mL) was added, and the product was extracted
four times with 250 mL portions of diethyl ether. The resulting
diethyl ether layer was collected and washed twice with 300
mL portions of water and then dried over MgSO4. Filtration
followed by solvent removal in vacuo afforded 18.5 g (68%) of
8,8-dimethylbenzofulvene, which was used without additional
purification. 1H NMR (benzene-d6): δ 1.74 (s, 3H, Me), 1.88
(s, 3H, Me), 6.63 (s, 2H, Cp), 7.08 (m, 2H, Ind), 7.20 (d, 9 Hz,
1H, Ind), 7.56 (d, 9 Hz, 1H, Ind). 13C NMR (benzene-d6): δ )
50.79 (Me), 54.61 (Me), 121.38, 123.89, 125.01 (Cp), 126.35,
128.82, 136.40, 137.42, 142.33, 144.60 (Ind). One resonance
not located.
P r ep a r a tion of 3-Isop r op ylin d en e. In a drybox, a 500
mL three-necked flask was charged with 5.33 g (0.14 mol) of
LiAlH4. Against an Ar counterflow, 150 mL of diethyl ether
was added by cannula. Through an addition funnel, a 150 mL
of a diethyl ether solution containing 18.5 g (0.12 mol) of 8,8-
dimethylbenzofulvene was added dropwise over the course of
1 h. The reaction was stirred for an additional 8 h, after which
time approximately 20 mL of a 15% aqueous NaOH solution
was then carefully added over the course of 1 h or until the
evolution of dihydrogen was complete. Ca u tion : Th e r ea c-
tion is h igh ly exoth er m ic a n d p r od u ces la r ge a m ou n ts
of H2(g)! Approximately 200 mL of H2O was then added to
the reaction mixture, and the product was extracted with three
300 mL portions of diethyl ether. The organic layer was
collected and washed with two 300 mL portions of water and
then dried over MgSO4. Filtration followed by solvent removal
in vacuo afforded 16.0 g (85%) of a red oil identified as
3-isopropylindene, which was used without further purifica-
tion. 1H NMR (benzene-d6): δ 1.12 (d, 9 Hz, 6H, CH(CH3)2),
2.70 (s, 9 Hz, 1H, CH(CH3)2), 2.96 (s, 2H, CH2), 5.88 (s, 1H,
Cp), 7.07 (m, 2H, Ind), 7.14 (d, 9 Hz, 1H, Ind), 7.24 (d, 9 Hz,
1H, Ind).
P r ep a r a tion of 3-Isop r op yl-8,8-d im eth ylben zofu lven e.
A 1 L round-bottomed flask was charged with 16.0 g (0.11 mol)
of 3-isopropylindene and 10.0 g (0.18 mol) of crushed KOH
pellets. A reflux condenser was attached, and approximately
150 mL of DME was added. The resulting reaction mixture
was heated to reflux, and 16 mL (0.22 mmol) of acetone was
added. The reaction mixture was then heated for an additional
12 h, forming a purple solution. After cooling to room temper-
ature, approximately 300 mL of H2O was added, and a workup
similar to that described for 8,8-dimethylbenzofulvene was
performed. This procedure yielded 19.0 g (86%) of a red oil
identified as 3-isopropyl-8,8-dimethylbenzofulvene, which was
used without further purification. 1H NMR (benzene-d6): δ
1.25 (d, 8 Hz, 6H, CH(CH3)2), 1.90 (s, 3H, Me), 2.03 (s, 3H,
Me), 2.88 (s, 9 Hz, 1H, CH(CH3)2), 6.57 (s, 1H, Cp), 7.21 (m,
2H, Ind), 7.31 (d, 8 Hz, 1H, Ind), 7.69 (d, 8 Hz, 1H, Ind). 13C
NMR (benzene-d6): δ 22.22, 22.46, 24.44, 26.99 (CH3 and
CHMe2), 28.07 (CH(CH3)2), 119.41, 120.86, 123.90 (Cp), 124.99,
126.03, 136.48, 137.64, 139.11, 144.21, 148.59 (Ind).
P r ep a r a tion
of
(η5-C9H5-1,3-(SiMe3)2)(η5-C9H5-1,3-
(SiMe2P h )2)Zr Cl2 (6-Cl2). In a 100 mL round-bottomed flask,
6.62 mmol of [(η5-C9H5-1,3-(SiMe3)2)ZrCl3‚(Et2O)2‚LiClx] was
generated in situ, and the resulting diethyl ether slurry was
chilled in a liquid nitrogen cooled cold well for 30 min. The
flask was removed, 2.59 g (6.63 mmol) of Li[C9H5-1,3-
(SiMe2Ph)2] was added to the mixture, and the reaction was
stirred overnight. The solvent was removed in vacuo, the
resulting yellow solid was extracted into dichloromethane, and
the solution was filtered through Celite. Removal of the solvent
in vacuo and recrystallization of the resulting yellow solid from
diethyl ether at -35 °C afforded 2.24 g (42%) of 6-Cl2. Anal.
Calcd for C40H50Si4ZrCl2: C, 59.66; H, 6.26. Found: C, 59.94;
H, 5.87. 1H NMR (benzene-d6): δ 0.29 (s, 18H, SiMe3), 0.73 (s,
6H, SiMe2Ph), 0.77 (s, 6H, SiMe2Ph), 6.80 (m, 2H, Ind), 6.98
(m, 2H, Ind), 7.00 (s, 1H, Cp), 7.10 (m, 3H, Cp/Ph), 7.15 (m,
4H, Ph), 7.22 (m, 4H, Ph), 7.62 (m, 2H, Ind), 7.67 (m, 2H, Ind).
13C NMR (benzene-d6): δ -0.31 (SiMe2Ph), 0.24 (SiMe2Ph),
1.14 (SiMe3), 119.66, 121.84, 126.24, 126.37 (Cp), 129.28,
134.13, 135.57, 138.07, 138.48, 139.47 (Ind/Ph). Four Ind/Ph
resonances not located.
P r ep a r a tion of 1,3-Diisop r op ylin d en e. A 250 mL round-
bottomed flask was charged with 5.0 g (25 mmol) of 3-iso-
propyl-8,8-dimethylbenzofulvene and approximately 100 mL
of diethyl ether. To the resulting red solution, 27.5 mL (28
mmol) of a 1.0 M solution of NaBEt3H in toluene was added,

5346 Organometallics, Vol. 23, No. 22, 2004
Bradley et al.
P r ep a r a tion of (η5-C9H5-1,3-(SiMe3)2)(η5-C9H6-SiMe3)Zr -
(CO)2 (8-(CO)2). A thick walled glass bomb was charged with
0.275 g (0.45 mmol) of 8-Cl2 and 0.115 g (4.7 mmol) of activated
Mg turnings. On a vacuum line, approximately 15 mL of THF
was vacuum transferred onto the solids and the resulting
slurry frozen in liquid nitrogen. One atmosphere of CO was
added to the vessel at -196 °C. The vessel was thawed by
warming to room temperature and then placed in a 45 °C oil
bath for 18 h. The CO atmosphere and the solvent were then
removed in vacuo, leaving a green residue. The vessel was
taken into the drybox and the residue extracted into pentane.
Filtration through Celite followed by solvent removal and
recrystallization from pentane at -35 °C afforded 0.142 g
(53%) of 8-(CO)2 as a forest green solid. Anal. Calcd for
1H NMR (benzene-d6): δ 0.40 (s, 18H, SiMe3), 0.45 (s, 18H,
SiMe3), 3.28 (br s, 1H, Cp), 3.93 (br s, 1H, Cp), 4.63 (br s, 1H,
Cp), 4.88 (br s, 1H, Cp), 6.50 (t, 4 Hz, 2H, Ind), 6.74 (m, 4H,
Ind), 6.8-7.0 (m, 6H, Ind), 7.51 (d, 8 Hz, 2H, Ind), 7.59 (d, 8
Hz 2H, Ind). 13C NMR (benzene-d6): δ 0.56 (SiMe3), 0.64
(SiMe3), 61.83, 65.39, 65.80, 74.27, 78.25, 90.13 (Cp), 90.38,
90.60, 122.39, 122.57, 122.72, 124.15, 124.49, 124.57, 126.47,
127.22, 129.95, 130.77 (Ind).
Ack n ow led gm en t. We would like to thank the
Cornell University Department of Chemistry and Chemi-
cal Biology, the donors of the Petroleum Research Fund
administered by the American Chemical Society, and
the National Science Foundation (predoctoral fellow-
ship to C.A.B. and CAREER Award to P.J .C.) for
generous financial support. P.J .C. is also a Cottrell
Scholar and would like to acknowledge financial support
from the Research Corporation. We also thank Mark
Wall and David Bray of the MIT Department of Chem-
istry Instrument Facility for access to spectrometer
time.
C
29H38Si3ZrO2: C, 58.63; H, 6.45. Found: C, 57.95; H, 6.52.
1H NMR (benzene-d6): δ 0.20 (s, 9H, SiMe3), 0.28 (s, 9H,
SiMe3), 0.31 (s, 9H, SiMe3), 4.29 (d, 5 Hz, 1H, Cp), 5.19 (s, 1H,
Cp), 5.92 (d, 5 Hz, 1H, Cp), 6.78 (m, 2H, Ind), 6.82 (m, 1H,
Ind), 6.87 (m, 2H, Ind), 7.21 (m, 1H, Ind), 7.29 (m, 1H, Ind),
7.33 (d, 10 Hz, 1H, Ind). 13C NMR (benzene-d6): δ 0.01 (SiMe3),
0.36 (SiMe3), 0.48 (SiMe3), 89.98, 95.19, 95.97, 96.41, 101.23,
106.07 (Cp), 121.31, 122.56, 123.27, 123.49, 123.75, 124.14,
124.54, 125.85, 126.17, 126.50, 127.19, 127.90 (Ind), 260.42
(Zr-CO), 264.04 (Zr-CO).
Su p p or tin g In for m a tion Ava ila ble: Additional experi-
mental details for the preparation of indenyl ligands,37 zir-
conocene dichloride, and dicarbonyl complexes as well as
dibenzoferrocene derivatives. Crystallographic data for 1-Cl2,
2-Cl2, 7-Cl2, 9-Cl2, 3-(CO)2, 13, and rac-17 including full atom
labeling schemes and complete bond distances and angles.
Representative infrared spectra recorded in KBr, variable-
temperature NMR data, and Eyring plot for 13. This material
is available free of charge via the Internet at http://pubs.acs.org.
P r ep a r a tion of r a c/m eso-(η5-C9H6-(SiMe3))2F e (10). A 20
mL scintillation vial was charged with 0.076 g (0.60 mmol) of
FeCl2 and approximately 5 mL of diethyl ether. The resulting
slurry was chilled to -35 °C in the drybox freezer, after which
time 0.233 g (1.20 mmol) of Li[C9H6-(SiMe3)] was added as an
ethereal solution over the course of 5 min. The reaction
mixture was stirred for 3 h and then filtered through a pad of
Celite. The diethyl ether was removed in vacuo and the
resulting solid recrystallized from diethyl ether at -35 °C,
affording 0.234 g (91%) of a blue-green solid identified as 10
as an equimolar mixture of rac and meso isomers. Anal. Calcd
for C24H30Si2Fe: C, 66.95; H, 7.02. Found C, 67.22; H, 6.77.
OM0495675
(37) (a) Resconi, L.; Balboni, C.; Baruzzi, G.; Flori, C.; Guidotti, S.;
Mercandelli, P.; Sirioni, A. Organometallics 2000, 19, 420. (b) Davison,
A.; Rakita, P. E. J . Organomet. Chem. 1970, 23, 407.