Conformational control of intramolecular arene stacking in ferrocene complexes bearing tert-butyl and pentafluorophenyl substituents

Article abstract of DOI:10.1016/S0022-328X(01)00878-6

The reaction of sodium tert-butylcyclopentadienide with C₆F₆ and NaH in THF affords either 1-pentafluorophenyl-3-tert-butylcyclopentadiene (1, as a mixture of double-bond regioisomers) or 1,2-bis(pentafluorophenyl)-4-tert-butylcyclopentadiene (2), depending on the reaction conditions. Treatment of 1 with NaH in THF affords sodium 1-pentafluorophenyl-3-tert-butylcyclopentadienide (3). Treatment of 2 with NaH in THF affords sodium 1,2-bis(pentafluorophenyl)-4-tert-butyl-cyclopentadienide (4). The reaction of the monoarylated ligand (3, two equivalents) with FeBr₂ in THF affords 1,1′-bis(pentafluorophenyl)-3,3′-di-tert-butylferrocene as two diastereomers (meso-5 and rac-5), which were separated by fractional crystallization from hexane. The X-ray crystal structure of rac-5 reveals a conformation in which the two opposing tert-butyl substituents are splayed, enabling transannular stacking of the C₆F₅ groups. In meso-5, steric repulsion of the two tert-butyl substituents prevents intramolecular arene stacking. The reaction of the diarylated ligand (4, two equivalents) with FeBr₂ in THF affords 1,1′,2,2′-tetrakis(pentafluorophenyl)-4,4′-di-tert-butylferrocene (6). The crystal structure of 6 reveals no arene stacking. Variable-temperature NMR studies showed that the Cp-Fe-Cp torsional barrier (ΔH^?) of 6 is 17(1) kcalmol^-1 in dichloromethane-d₂ solution.

Full text of DOI:10.1016/S0022-328X(01)00878-6

Journal of Organometallic Chemistry 637–639 (2001) 107–115
www.elsevier.com/locate/jorganchem
Conformational control of intramolecular arene stacking in
ferrocene complexes bearing tert-butyl and pentafluorophenyl
substituents
Paul A. Deck a,*, Caleb E. Kroll b,1, W. Gary Hollis Jr. b, Frank R. Fronczek c
a Department of Chemistry, Virginia Tech, Blacksburg, VA 24061-0212, USA
b Department of Chemistry, Roanoke College, Salem, VA 24153, USA
c Department of Chemistry, Louisiana State Uni6ersity, Baton Rouge, LA 70803-1804, USA
Received 13 November 2000; received in revised form 20 December 2000; accepted 27 December 2000
Abstract
The reaction of sodium tert-butylcyclopentadienide with C6F6 and NaH in THF affords either 1-pentafluorophenyl-3-tert-
butylcyclopentadiene (1, as a mixture of double-bond regioisomers) or 1,2-bis(pentafluorophenyl)-4-tert-butylcyclopentadiene (2),
depending on the reaction conditions. Treatment of 1 with NaH in THF affords sodium 1-pentafluorophenyl-3-tert-butylcy-
clopentadienide (3). Treatment of 2 with NaH in THF affords sodium 1,2-bis(pentafluorophenyl)-4-tert-butyl-cyclopentadienide
(4). The reaction of the monoarylated ligand (3, two equivalents) with FeBr2 in THF affords 1,1%-bis(pentafluorophenyl)-3,3%-di-
tert-butylferrocene as two diastereomers (meso-5 and rac-5), which were separated by fractional crystallization from hexane. The
X-ray crystal structure of rac-5 reveals a conformation in which the two opposing tert-butyl substituents are splayed, enabling
transannular stacking of the C6F5 groups. In meso-5, steric repulsion of the two tert-butyl substituents prevents intramolecular
arene stacking. The reaction of the diarylated ligand (4, two equivalents) with FeBr2 in THF affords 1,1%,2,2%-tetrakis(pen-
tafluorophenyl)-4,4%-di-tert-butylferrocene (6). The crystal structure of 6 reveals no arene stacking. Variable-temperature NMR
studies showed that the CpꢀFeꢀCp torsional barrier (DH‡) of 6 is 17(1) kcal mol−1 in dichloromethane-d2 solution. © 2001
Elsevier Science B.V. All rights reserved.
Keywords: Crystal engineering; Arene stacking; Ferrocene; Pentafluorophenyl; Conformational analysis
1. Introduction
controlling arene stacking through rational structural
modification of organometallic complexes.
Arene stacking is a persistent theme in the solid-state
structures of metallocenes and other cyclopentadienyl
(Cp) complexes bearing pentafluorophenyl (C6F5) sub-
stituents [1,2]. Hughes and co-workers have rationalized
intramolecular arene stacking in the crystal structures
of 1,1%-diarylferrocenes (aryl=phenyl, pentafluoro-
phenyl, or one of each) in terms of fundamental crystal
engineering concepts [3]. As a part of our continuing
study of the synthesis and reactivity of C6F5-substituted
Cp complexes [4], we aim to explore the possibility of
In this report, we show that a bulky tert-butyl (t-Bu)
group on cyclopentadiene effects complete regiocontrol
over the subsequent attachment of one or two C6F5
groups. We then describe the synthesis of ferrocenes in
which each Cp ligand bears one t-Bu group and one or
two C6F5 groups. We then show, using crystallographic
data, how transannular steric interactions of the t-Bu
groups either enable or prohibit CpꢀFeꢀCp conforma-
tions in which arene stacking can occur, depending on
their stereochemical arrangement. We also compare the
enthalpy of activation for CpꢀFeꢀCp torsion of
1,1%,2,2%-tetrakis(pentafluorophenyl)-4,4%-di-tert-butyl-
ferrocene (6) with two related complexes, 1,1%,2,2%,4,4%-
hexakis(pentafluorophenyl)ferrocene or 1,1%,2,2%,4,4%-
hexa(tert-butyl)ferrocene.
* Corresponding author. Tel.: +1-540-231-3493; fax: +1-540-231-
3255.
E-mail address: pdeck@vt.edu (P.A. Deck).
1 Undergraduate research participant.
0022-328X/01/$ - see front matter © 2001 Elsevier Science B.V. All rights reserved.
PII: S0022-328X(01)00878-6

108
P.A. Deck et al. / Journal of Organometallic Chemistry 637–639 (2001) 107–115
2. Experimental
2.1. General procedures
1H), 3.17 (s, 2H), 1.23 (s, 9H). 19F-NMR (CDCl3): l
1a, −141.26 (m, 2F), −159.08 (t, 1F, 3J=21 Hz),
−163.86 (m, 2F); 1b, −141.75 (m, 2F), −159.77 (t,
3
1F, J=21 Hz), −164.02 (m, 2F); 1c, −141.38 (m,
3
Standard inert-atmosphere techniques were used for
all reactions [5]. Hexafluorobenzene was used as re-
ceived from Oakwood. Sodium tert-butylcyclopentadi-
enide and potassium tert-butylcyclopentadienide were
prepared from tert-butylcyclopentadiene (Aldrich,
freshly distilled prior to use) and NaH or KH, respec-
tively, in THF. NMR experiments were performed
using JEOL ECP500 and Varian U400 instruments.
Dynamic NMR measurements were carried out on the
Varian U400, and the temperature scale was calibrated
using a MeOH standard. THF-d8 was vapor-transferred
into resealable NMR tubes from Na–K alloy. 19F-
NMR spectra were referenced to external C6F6 in
CDCl3 at −163.00 ppm. {19F}13C-NMR spectra were
particularly useful for resolving signals in the aromatic
CꢀF regions where extensive 19F–13C couplings in the
1H-decoupled spectra prevented us from distinguishing
closely spaced signals. Furthermore, 19F decoupling
provides substantial NOE enhancement of signals that
are ordinarily weak under the usual conditions of
broadband 1H decoupling. CꢀF coupling constants
were approximated from observed splittings in the
{1H}13C-NMR spectra. Elemental analyses were per-
formed by Desert Analytics (Tucson, AZ).
2F), −157.65 (t, 1F, J=21 Hz), −163.52 (m, 2F).
Anal. Found: C, 62.24; H, 4.37. Calc. for C15H13F5: C,
62.50; H, 4.55%.
2.3. 1,2-Bis(pentafluorophenyl)-4-tert-butylcyclo-
pentadiene (2)
A mixture of potassium tert-butylcyclopentadienide
(3.20 g, 20.0 mmol), hexafluorobenzene (17.0 g, 91.4
mmol), sodium hydride (2.4 g, 0.10 mol) and THF (100
ml) was stirred at 65 °C for 60 h. After removing the
solvent, hexane (100 ml) and water (50 ml, dropwise at
first) were added to hydrolyze the mixture. The biphasic
mixture was filtered through Celite and then separated.
The aqueous layer was extracted with 25 ml of hexanes.
The combined organic layers were washed with water,
dried over anhydrous magnesium sulfate, filtered, and
evaporated to afford 6.07 g of a yellow oil. The crude
oil was subjected to flash chromatography on silica gel
(40×7 cm2), eluting with hexanes to afford first 230 mg
(0.80 mmol, 4%) of 1, and then 4.44 g (9.8 mmol, 49%)
of pure 2. Because 2 is initially formed as a viscous oil
that retains solvent, we waited until the oil sponta-
neously crystallized (several hours) and then sublimed a
small portion (0.05 Torr, 80 °C) to obtain an analytical
1
2.2. 1-(Pentafluorophenyl)-3-tert-butylcyclopentadiene
(1)
sample. H-NMR (CDCl3): l 6.29 (s, 1H), 3.54 (s, 2H),
1.26 (s, 9H). 19F-NMR (CDCl3): l −140.42 (m, 2F),
−141.10 (m, 2F), −154.68 (t, 1F, 3J=21 Hz), −
3
A mixture of sodium tert-butylcyclopentadienide
(1.44 g, 10.0 mmol), hexafluorobenzene (3.72 g, 20.0
mmol), sodium hydride (0.48 g, 20 mmol) and THF
(100 ml) was stirred at 25 °C for 20 h. After removing
the solvent, hexane (100 ml) and water (50 ml, dropwise
at first) were added to hydrolyze the mixture. The
biphasic mixture was filtered through Celite and then
separated. The aqueous layer was extracted with 25 ml
of hexanes. The combined organic layers were washed
with water, dried over anhydrous magnesium sulfate,
filtered, and evaporated to afford 1.85 g of a brown oil.
TLC analysis (silica gel, 250 mm, hexanes) of the oil
showed two major products: compound 1 at Rf=0.57
and compound 2 at Rf=0.47. The crude oil was sub-
jected to flash chromatography on silica gel (40×4 cm2
column), eluting with hexanes to afford 1.02 g (3.54
mmol, 35%) of pure 1 as a colorless oil. A subsequent
fraction afforded 160 mg (0.4 mmol, 4%) of pure 2.
Spectroscopic analysis showed that 1 is obtained as a
mixture of three double-bond regioisomers (1a, 53%;
155.36 (t, 1F, J=21 Hz), −162.01 (m, 2F), −162.16
(m, 2F). {1H}13C-NMR (CDCl3): l 163.0 (s, C-t-Bu),
3
133.3 (s, C-C6F5), 129.0 (t, C-C6F5, JCF=2 Hz), 126.0
(s, CH), 44.1 (t, CH2, 4JCF=2 Hz), 33.6 (s, CMe3), 30.6
(s, CH3). {19F}13C-NMR (CDCl3): l 143.9 (s, CF),
143.8 (s, CF), 140.8 (s, CF), 140.3 (s, CF), 137.76 (s,
CF), 137.72 (s, CF), 111.6 (s, C6F5 ipso), 111.2 (s, C6F5
1
4
ipso), 44.1 (td, CH2, JCH=127 Hz, JCH=9 Hz), 30.6
1
(q, CH3, JCH=126 Hz). Anal. Found: C, 55.47; H,
2.33. Calc. for C21H12F10: C, 55.52; H, 2.66%.
2.4. Sodium 1-pentafluorophenyl-3-tert-butylcyclo-
pentadienide (3)
A mixture of 1 (1.02 g, 3.54 mmol), sodium hydride
(127 mg, 5.31 mmol), and THF (50 ml) was stirred at
25 °C for 15 h. The dark mixture was filtered, the
solvent was evaporated, and then most of the residual
THF was removed by heating the crude product at
60 °C under a high vacuum (6×10−6 mm) for 6 h to
afford a dark, sticky solid as the sole product in
approximately quantitative yield. 1H-NMR (THF-d8): l
6.35 (m, 2H), 5.89 (m, 1H), 1.27 (s, 9H). 19F-NMR
(THF-d8): l −146.12 (m, 2F), −167.79 (m, 2F), −
1
1b, 33%; and 1c, 14%). Assignments are tentative. H-
NMR (CDCl3): l 1a, 7.24 (m, 1H), 6.20 (m, 1H), 3.48
(s, 2H), 1.22 (s, 9H); 1b, 7.08 (m, 1H), 6.30 (m, 1H),
3.47 (s, 2H), 1.24 (s, 9H); 1c, 6.66 (m, 1H), 6.39 (m,

P.A. Deck et al. / Journal of Organometallic Chemistry 637–639 (2001) 107–115
109
3
4
174.92 (tt, 1F, J=2 Hz, J=6 Hz). {1H}13C-NMR
afforded 132 mg (0.21 mmol, 13%) of pure rac-5. In an
alternating fashion, further crops enabled the recovery
of additional small amounts of each isomer. Purity for
4
(THF-d8): l 108.2 (t, CH, JCF=8 Hz), 105.3 (t, CH,
5J=3 Hz), 104.7 (t, CH, JCF=8 Hz), 33.4 (s, CH3),
4
1
32.6 (s, CMe3). Satisfactory microanalytical data were
each compound was confirmed by H- and 19F-NMR,
not obtained. H- and 19F-NMR spectra (Figs. S1 and
while the structural assignments were confirmed by
1
S2) are available from the corresponding author as
evidence of substantial purity.
single-crystal X-ray diffraction.
1
Data for meso-5: H-NMR (CDCl3): l 4.43 (m, 2H),
4.40 (m, 2H), 4.38 (m, 2H), 1.19 (s, 18H). 19F-NMR
3
2.5. Sodium 1,2-bis(pentafluorophenyl)-4-tert-butyl-
cyclopentadiene (4)
(CDCl3): l −140.35 (m, 4F), −158.41 (t, 2F, J=21
Hz), −163.44 (m, 4F). {1H}13C-NMR (CDCl3): l
4
104.4 (s, C-C6F5), 71.8 (s, C-t-Bu), 71.5 (t, CH, JCF
=
4
A mixture of 2 (3.37 g, 7.42 mmol), sodium hydride
(0.27 g, 11 mmol), and THF (100 ml) was stirred at
25 °C for 15 h. The resulting dark mixture was filtered,
the filtrate was evaporated, and the residual THF was
removed by heating the crude product at 80 °C under a
high vacuum (6×10−6 mm) for 12 h to afford a gray
5 Hz), 68.7 (s, CH), 68.3 (t, CH, JCF=6 Hz), 31.4 (s,
CH3), 30.6 (s, CMe3). {19F}13C-NMR (CDCl3): l 144.4
(s, CF), 138.7 (s, CF), 138.0 (s, CF), 114.5 (C6F5 ipso).
Anal. Found: C, 57.35; H, 3.82. Calc. for C30H24F10Fe:
C, 57.16; H, 3.84%.
1
Data for rac-5: H-NMR (CDCl3): l 4.90 (m, 2H),
1
solid (quantitative). H-NMR (THF-d8): l 6.08 (s, 2H),
4.68 (m, 2H), 4.24 (m, 2H). 19F-NMR (CDCl3): l
−139.68 (m, 4F), −159.2 (t, 2F, 3J=21 Hz), −
163.88 (m, 4F). {1H}13C-NMR (CDCl3): l 104.8 (s,
1.29 (s, 9H). 19F-NMR (THF-d8): l −146.81 (m, 4H),
3
4
−168.78 (m, 4H), −170.81 (tt, 2H, J=21 Hz, J=2
Hz). {1H}13C-NMR (THF-d8): l 144.2 (d, CF, JCF
=
C-C6F5), 71.7 (s, C-t-Bu), 70.4 (t, CH, JCF=6 Hz),
1
4
1
4
252 Hz), 138.4 (d, CF, JCF=248 Hz), 136.2 (d, CF,
1J=242 Hz), 136.2 (s, C-C6F5), 120.4 (t, C6F5 ipso,
2JCF=19 Hz), 109.3 (s, CH), 101.2 (s, C-t-Bu), 33.3 (s,
CH3), 32.4 (CMe3). {19F}13C-NMR (THF-d8): d 144.2
(s, CF), 138.4 (s, CF), 136.2 (s, CF), 120.4 (s, C6F5
ipso), 109.3 (dd, CH, 1JCH=156 Hz, 4JCH=8 Hz),
68.6 (s, CH), 67.6 (t, CH, JCF=7 Hz), 31.5 (s, CH3),
30.7 (s, CMe3). {19F}13C-NMR (CDCl3): l 144.1 (s,
CF), 138.4 (s, CF), 137.6 (s, CF), 112.7 (C6F5 ipso).
Anal. Found: C, 57.09; H, 3.68. Calc. for C30H24F10Fe:
C, 57.16; H, 3.84%.
2
1
101.2 (t, C-t-Bu, JCH=7 Hz), 33.3 (q, CH3, JCH
=
2.7. 1,1%,2,2%-Tetrakis(pentafluorophenyl)-4,4%-
di-tert-butylferrocene (6)
126 Hz), 32.4 (s, CMe3). Satisfactory microanalytical
data were not obtained. 1H- and 19F-NMR spectra
(Figs. S3 and S4) are available from the corresponding
author as evidence of substantial purity.
A solution of 4 (3.40 g, 7.14 mmol) and FeBr2 (721
mg, 3.34 mmol) in THF (100 ml) was stirred at 25 °C
for 24 h. The solvent was evaporated to afford a red
residue. The residue was dissolved in 100 ml of benzene
and filtered through a bed of neutral alumina (30×10
mm2), using 100 ml of hexanes to elute all of the red
product. The dark red filtrate was evaporated to afford
a red crystalline solid, which was rinsed with pentane
(2×5 ml) and dried under vacuum to afford 3.20 g
(3.32 mmol, 98%) of a dark red solid. An analytical
sample was obtained by recrystallization from hexanes.
1H-NMR (CD2Cl2, 295 K): l 4.86 (br s, 2H), 4.78 (br s,
2H), 1.11 (s, 18H). 19F-NMR (CD2Cl2, 295 K): l
−139.89 (br, 4F), −140.64 (m, 4F), −159.30 (t, 2F,
2.6. 1,1%-Bis(pentafluorophenyl)-3,3%-di-tert-
butylferrocene (5)
A solution of 3 (3.54 mmol) and iron(II) bromide
(350 mg, 1.60 mmol) in THF (100 ml) was stirred at
25 °C for 6 h. The solvent was evaporated to afford a
dark residue. Benzene (75 ml) and water (25 ml) were
added, and the biphasic mixture was filtered through
Celite, which was rinsed with 100 ml of hexanes. The
biphasic mixture was then separated, and the aqueous
layer was extracted with 20 ml of hexanes. The com-
bined organic layers were washed with brine (3×20
ml), dried over anhydrous magnesium sulfate, filtered,
and evaporated to afford 0.98 g (1.56 mmol, 97%) of a
red–orange crystalline solid. TLC analysis (silica, 250
mm, hexanes) showed a single red spot (Rf=0.35) and
traces of the hydrolyzed ligand (1, Rf=0.57), suggest-
ing that meso-5 and rac-5 could not be separated by
flash chromatography. Instead, the crude product (5)
was fractionally crystallized from about 10 ml of hex-
anes, reducing the mother liquors by half for each
successive crop. The first crop afforded 190 mg (0.30
mmol, 19%) of pure meso-5, while the second crop
3JFF=21 Hz), −160.28 (t, 2F, J=21 Hz), −166.70
(m, 4F), −167.97 (m, 4F). Anal. Found: C, 52.04; H,
2.23. Calc. for C42H22F20Fe: C, 52.41; H, 2.30%.
3
2.8. Crystallography
Crystals of meso-5 and rac-5 were obtained by cool-
ing concentrated hexanes solutions to −5 °C for 24 h.
Crystals of 6 obtained from hexanes revealed disor-
dered solvent molecules that were difficult to model.
Superior crystals were obtained by cooling a concen-
trated CDCl3 solution of 6 to −5 °C for 2 days.

110
P.A. Deck et al. / Journal of Organometallic Chemistry 637–639 (2001) 107–115
Relevant crystal data are assembled in Table 1. Details
of the crystallographic experiments as well as complete
tables of metric parameters are available from the
corresponding author.
which is apparently prohibited by steric congestion in
the nucleophilic aromatic substitution transition state,
or (b) placing the aryl group, the t-Bu group, or both
on the sp3-hybridized carbon atom of the cyclopentadi-
ene. Moving the aryl group to the sp3-hybridized posi-
tion would disrupt conjugation, while moving the t-Bu
group to the sp3-hybridized position would violate
Saytzeff’s rule. Regarding the precise spectral assign-
ments, we were unable to further distinguish unambigu-
ously which 1H- and 19F-NMR spectra signals
corresponded to each of the three isomers. Selectivity
for monoarylation or diarylation is achieved by varying
the reaction conditions. At 25 °C (20–40 h), the
monoarylated cyclopentadiene (1) is the major product,
whereas a longer reaction (60 h) at 65 °C affords nearly
pure diarylated product (2), both in moderate yields.
The two products (1 and 2) are readily separated and
purified by silica gel chromatography. For the reasons
outlined above, the diarylated cyclopentadiene (2) is
obtained as a single regioisomer.
3. Results and discussion
3.1. Ligand synthesis
We showed earlier that Cp anions react with pe-
rfluoroarenes in nucleophilic aromatic substitution re-
actions to afford fluoroarylated cyclopentadienes [4]. In
the present study, we started with tert-butylcyclopenta-
diene in order to examine the effect of the bulky t-Bu
group on the regioselectivity of this process. Scheme 1
shows that this reaction procedes with complete re-
gioselectivity for arylation distal to the tert-butyl
group. The 1H- and 19F-NMR spectra of the crude
monoarylation product (1) revealed the presence of
three distal double-bond regioisomers in a ratio of
about 5:3:1. Additional isomers would require either (a)
attaching the aryl group proximal to the t-Bu group,
As shown in Scheme 1, both of the dienes (1 and 2)
are readily converted to the corresponding sodium salts
in essentially quantitative yields. Formation of a single
Table 1
Crystallographic data
Compound
meso-5
30H24F10Fe
rac-5
6
Empirical formula
Formula weight
Diffractometer
Crystal size (mm)
Crystal system
C
C30H24F10Fe
630.34
Nonius Kappa CCD
0.35×0.25×0.22
Monoclinic
C42H22F20Fe·CDCl3
1081.81
Nonius Kappa CCD
0.35×0.25×0.08
Triclinic
630.34
Nonius Kappa CCD
0.28×0.27×0.12
Monoclinic
Unit cell dimensions
,
a (A)
10.0450(3)
8.9440(2)
29.5860(8)
14.5540(3)
12.6950(4)
14.4570(4)
11.1680(2)
20.0280(6)
20.2450(6)
70.8390(12)
80.4250(17)
81.9660(18)
4200.1(2)
,
b (A)
,
c (A)
h (°)
i (°)
k (°)
V (A )
91.0590(14)
95.6590(17)
3
,
2657.63(12)
P21/c (no. 14)
4
1.575
0.7
2658.10(12)
P21/c (no. 14)
4
1.575
0.7
(
Space group
Z
P1 (no. 2)
4
Dcalc (Mg m−3
)
1.711
0.7
2152
Absorption coefficient (mm−1
F(000)
)
1280
1280
,
u (Mo–Ka) (A)
Temperature (K)
0.71073
120
0.71073
120
0.71073
120
Theta range for data collection (°)
No. of reflections collected
No. of independent reflections
Absorption correction method
Max/min transmission
H treatment
2.7–32.0
21967
9125
Multi-scan
0.925, 0.806
Constrained
9125/0/376
0.034
2.6–30.0
12876
7739
Multi-scan
0.869, 0.832
Constrained
7739/0/376
0.035
2.5–30.0
43374
24072
Multi-scan
0.948, 0.828
Constrained
24072/0/1219
0.053
Data/restraints/parameters
R [I\2 and (I)]
Rw [I\2 and (I)]
0.080
0.89
0.43 and −0.40
0.096
1.02
0.32 and −0.38
0.142
0.92
1.03 and −1.14
Goodness-of-fit on F2
−3
,
Largest difference peak and hole (e A
)

P.A. Deck et al. / Journal of Organometallic Chemistry 637–639 (2001) 107–115
111
stant and one larger coupling constant would be ex-
pected. This regiochemical assignment was ultimately
confirmed by crystallographic analysis of the corre-
sponding ferrocene derivatives.
Chart (1)
3.2. Metallocene synthesis
The arylated ligands (3 and 4) reacted smoothly with
FeBr2 in THF (Scheme 2) to afford high yields of the
corresponding ferrocene complexes (5 and 6). The
monoarylated ligand (3) afforded a roughly 1:1 mixture
of meso-5 and rac-5. By careful fractional crystalliza-
tion from hexanes, we isolated pure meso-5 from the
first, third, and fifth crops of crystals. The second and
fourth crops contained mostly rac-5. Purities of individ-
Scheme 1.
1
ual crops were established by H- and 19F-NMR spec-
troscopy, and individual well-formed crystals were
selected for crystallographic analysis (see below). Both
complexes are crystalline solids having the dark orange
color that is typical of ferrocene complexes. The te-
traarylated ferrocene (6) was isolated as a deep red
crystalline solid. All three complexes (meso-5, rac-5,
and 6) are air-stable and were fully characterized by
1H-, 19F-, and 13C-NMR spectroscopy as well as ele-
mental microanalysis and single-crystal X-ray diffrac-
tion (see below).
3.3. Structural studies
Scheme 2.
The molecular structure of meso-5 is shown in Fig.
1a. Selected metric data are provided in Table 2. Note-
worthy is the absence of intramolecular arene stacking
(Fig. 1b). A survey of prior studies of 1,1%-bis(pen-
tafluorophenyl)-substituted ferrocene complexes (e.g.
7–9) reveal that the structure in which the two C6F5
groups are engaged in p-stacking is the solid-state
conformational minimum. However, based on Okuda’s
earlier investigations of CpꢀFeꢀCp conformational dy-
namics in t-Bu-substituted ferrocenes, we also know
that the transannular steric interaction of two t-Bu
groups in either eclipsing [hꢁÚ(t-BuꢀCpꢀCp%ꢀt-Bu%)=
0°] or even gauche (h=36°) conformations is strongly
avoided [9]. These interactions prevent meso-5 from
adopting a conformation (h:0°) that would enable
stacking of the two C6F5 groups. The actual conforma-
tion is defined by h=82.5°.
Cp anion from the monoarylated diene (1) supports the
assignment of a single regiochemical relationship be-
tween the t-Bu and C6F5 groups. Both sodium salts (3
and 4) are thermally stable (to at least 100 °C) but
air-sensitive. 13C-NMR analysis of the monoarylated
anion (3) provides evidence of the distal regiochemistry
(Chart 1). The three signals assigned to the methine
(CH) carbons at 108.2, 105.3, and 104.7 ppm exhibit
long-range CꢀF couplings of 8, 3, and 8 Hz, respec-
tively. In the observed distal regioisomer, two CH
carbons would exhibit the larger four-bond coupling
constant (8 Hz), and one CH carbon would exhibit a
smaller five-bond coupling constant (3 Hz), whereas in
the proximal regioisomer, two smaller coupling con-

112
P.A. Deck et al. / Journal of Organometallic Chemistry 637–639 (2001) 107–115
The molecular structure of rac-5 (Fig. 2, Table 2)
supports this analysis. In rac-5, stacking of the two
C6F5 groups (Fig. 2b) can be achieved without ap-
proaching conformational angles in which the two t-Bu
engage in strong mutual repulsions. The observed con-
formation is defined by h=82.6°, nearly the same value
as observed for meso-5. Interestingly, the C6F5ꢀC6F5
stacking motif in rac-5 is ‘tighter’ than in other related
complexes that we have analyzed. For example, the
,
C6F5(centroid)ꢀC6F5(centroid) distance of 3.39 A is
,
nearly 0.2 A shorter than the non-butylated parent
compound (7), which exhibits a corresponding distance
,
of 3.58 A [3]. The angle of 4.4° between the two C6F5
planes of rac-5 also suggests a well-organized stacking
arrangement. Corresponding interplanar angles in other
C6F5-substituted ferrocenes that we have studied range
from 5° to 17° [1,2]. The C6F5(centroid)ꢀC6F5(least-
,
squares plane) distances (3.225 and 3.271 A) of rac-5
are less than C6F5(centroid)ꢀC6F5(centroid) distance be-
cause of the tilting of aryl groups relative to their
respective Cp ligands by 28 and 26°, respectively. Cor-
responding CpꢀC6F5 tilting angles in 7 are both 25° [3].
Fig. 1. Crystal structure of meso-5. (a) Thermal ellipsoid plot shown
at 50% probability. H atoms were removed for clarity. F atoms are
numbered as the attached C atoms. (b) Ball-and-stick diagram show-
ing CpꢀFeꢀCp conformation.
Table 2
a
,
Selected bond lengths (A) and bond angles (°) for meso-5, rac-5, and 6
meso-5
rac-5
6a
6b
meso-5
rac-5
6a
6b
FeꢀC1
FeꢀC2
FeꢀC3
FeꢀC4
2.041(1)
2.049(1)
2.083(1)
2.052(1)
2.037(2)
1.655(1)
1.479(2)
2.056(1)
2.049(1)
2.079(1)
2.061(1)
2.037(1)
1.660(1)
1.475(2)
2.091(3)
2.111(3)
2.076(3)
2.107(3)
2.054(3)
1.698(3)
1.472(4)
1.481(4)
2.071(3)
2.116(3)
2.077(3)
2.104(3)
2.054(3)
1.695(3)
1.490(4)
1.478(4)
FeꢀC6
FeꢀC7
FeꢀC8
FeꢀC9
FeꢀC10
FeꢀCp2 b
C6ꢀC61
C7ꢀC71
C8ꢀC81
C9ꢀC91
2.043(1)
2.047(1)
2.076(1)
2.054(1)
2.034(1)
1.653(1)
1.476(2)
2.049(1)
2.050(1)
2.080(1)
2.056(1)
2.038(1)
1.659(1)
1.479(2)
2.107(3)
2.076(3)
2.055(3)
2.102(3)
2.074(3)
1.694(3)
1.476(4)
1.489(4)
2.111(3)
2.087(3)
2.050(3)
2.103(3)
2.081(3)
1.695(3)
1.477(4)
1.478(4)
FeꢀC5
b
FeꢀCp1
C1ꢀC11
C2ꢀC21
C3ꢀC31
C4ꢀC41
1.522(2)
1.520(2)
1.524(2)
1.519(2)
1.522(4)
1.518(4)
1.536(4)
1.525(4)
c
Cp1ꢀFeꢀCp2n
178
22
180
29
174
39
44
174
35
47
h
83
28
83
26
100
45
29
102
41
32
b1 d
b2 d
b3 d
b4 d
a Crystalline 6 shows two nearly identical molecules per asymmetric unit, designated 6a and 6b.
b Cp1 and Cp2 are the centroid of C1ꢀC5 and C6ꢀC10, respectively.
c h is the torsion angle C(t-Bu)ꢀCp1ꢀCp2ꢀC(t-Bu) (absolute value).
d b1ꢀb4 are Cp-aryl interplanar torsion angles (absolute values). b1: (C1ꢀC5)ꢀ(C11ꢀC16); b2: (C1ꢀC5)ꢀ(C21ꢀC26); b3: (C6ꢀC10)ꢀ(C61ꢀC66);
and b4: (C6ꢀC10)ꢀ(C71ꢀC76).

P.A. Deck et al. / Journal of Organometallic Chemistry 637–639 (2001) 107–115
113
angles (b1ꢀb4, Table 2) lie in the range 20–50, typical
of other C6F5-substituted ferrocenes that we have stud-
ied [2]. These angles tend to the lower end of this range
when additional steric constraints such as those im-
posed by bulky vicinal substituents are absent, whereas
6 exhibits b angles that are toward the upper end of the
typical range. The longer FeꢀCp distances in 6 also
suggest that 6 shows transannular steric repulsion, be-
cause one would expect the additional electron-with-
drawing C6F5 groups to decrease FeꢀC distances [7].
3.4. Dynamic NMR studies
The NMR spectra of 6 reveal an exchange process
described in Fig. 4. The ligand substitution pattern
alone suggests time-averaged C2h (6a) or C26 (6f) sym-
metry with all four Cp protons and all four C6F5
groups chemically equivalent. In fact, Hanusa and co-
workers found that [1,2,4-Cy3C5H2]2Fe (Cy=cyclo-
hexyl) adopts approximate C2h symmetry in the solid
state [8]. However, the 1H-NMR spectrum of 6 at
23 °C shows a pair of exchange-broadened signals. At
room temperature, the two diastereotopic C6F5 groups
Fig. 2. Crystal structure of rac-5. (a) Thermal ellipsoid plot shown at
50% probability. H atoms were removed for clarity. F atoms are
numbered as the attached C atoms. (b) Ball-and-stick diagram viewed
along the CpꢀFeꢀCp axis showing transannular arene stacking.
The molecular structure of 6 is shown in Fig. 3a.
Rationalizing the observation that 6·CDCl3 does not
engage in intramolecular arene stacking requires closer
inspection of transannular interactions (Fig. 3b). The
F26ꢀF72 and F16ꢀF66 distances are 2.56 and 2.59,
respectively. Turning the ‘bottom’ ligand to increase the
stacking interaction between the C21ꢀC26 and
C61ꢀC66 aryl substituents would decrease both of these
FꢀF distances and introduce additional strain. The
C16ꢀC66 and C71ꢀC76 aryl groups cannot twist about
their CpꢀC6F5 axes to relieve this strain because of
transannular interactions with the tert-butyl groups via
C44 and C94.
Neither meso-5, rac-5, nor 6·CDCl3 exhibits inter-
molecular arene stacking. The packing diagram of
meso-5 shows a pairwise CpꢀC6F5 ‘stacking’ motif, but
the Cp(centroid)-to-C6F5(centroid) distance of 3.828
and the Cp-to-C6F5 interplanar angle of 29.4° suggest
that this feature of the packing arragement is poorly
organized and insignificant.
Additional metric data in Table 2 shows that while
meso-5 and rac-5 are relatively unstrained about their
metallocene cores, the tetraarylated complex shows a
slight ‘bend’ in the CpꢀFeꢀCp angle (visible in Fig. 3a).
The FeꢀC distances show that none of these complexes
exhibits significantly ‘slipped’ or ‘hinged’ Cp ligands [6],
although the carbon atoms (C3, C8) bearing the tert-
butyl substituents tend to be more distant from the iron
atom in meso-5 and rac-5. The CpꢀC6F5 interplanar
Fig. 3. Crystal structure of 6·CDCl3. (a) Thermal ellipsoid plot shown
at 50% probability. H and F atoms were removed for clarity. (b)
Ball-and-stick diagram showing CpꢀFeꢀCp conformations. Darkened
F and dotted lines atoms indicate transannular F···F repulsions that
prevent stacking of the C21ꢀC26 and C61ꢀC66 aryl substituents.

114
P.A. Deck et al. / Journal of Organometallic Chemistry 637–639 (2001) 107–115
4. Conclusions
A single tert-butyl substituent on the Cp anion di-
rects arylation by C6F6 to distal positions. Isomeric
ferrocene complexes bearing one t-Bu and one C6F5
group on each Cp ligand exhibit intramolecular arene
stacking only if the stacking conformation is consistent
with the conformational preferences of opposing t-Bu
groups.
5. Supplementary material
The following data are available from the corre-
Fig. 4. Conformational analysis of 6. Boxed diagrams correspond to
the crystal structure. Filled circle=tert-butyl; open circle=C6F5
(Fig. 3).
sponding author: NMR data for 3 and 4, variable-tem-
1
perature H-NMR spectra of 6 and an associated table
of rate data.
Crystallographic data for the structural analysis have
been deposited with the Cambridge Crystallographic
Data Centre, CCDC nos. 154483, 154484, and 154485,
for compounds rac-5, meso-5, 6·CDCl3, respectively.
Copies of this information may be obtained free of
charge from The Director, CCDC, 12 Union Road,
Cambridge CB2 1EZ, UK (Fax: +44-1223-336033;
e-mail: deposit@ccdc.cam.ac.uk or www: http://
www.ccdc.cam.ac.uk).
Acknowledgements
This work was supported in part by a grant from the
Jeffress Memorial Trust to W.G.H., a Roanoke College
Summer Scholarship to C.E.K., and the Virginia Tech
Chemistry Department.
Fig. 5. Eyring plot of rate constants for CpꢀFeꢀCp rotation in 6.
are also in the slow-exchange spectral regime in the
1
19F-NMR spectrum. Lineshape analyses of the H spec-
trum at several temperatures in the slow-exchange
regime (−20 to 22 °C, spectra are available from the
corresponding author) and application of Eq. (1) lead
to an Eyring plot (Fig. 5), from which we obtain the
References
[1] (a) M.P. Thornberry, C. Slebodnick, P.A. Deck, F.R. Fronczek,
Organometallics 19 (2000) 5352;
(b) M.P. Thornberry, C. Slebodnick, P.A. Deck, F.R. Fronczek,
Organometallics 20 (2001) 920.
[2] P.A. Deck, M.J. Lane, J.L. Montgomery, C. Slebodnick, F.R.
Fronczek, Organometallics 19 (2000) 1013.
activation parameters (DH‡=1792 kcal mol−1
,
DS‡=696 e.u., and DG2‡98=1592 kcal mol−1) for
CpꢀFeꢀCp rotation in 6. Notably, this enthalpy of
activation is comparable to corresponding values ob-
[3] M.D. Blanchard, R.P. Hughes, T.E. Concolino, A.L. Rheingold,
Chem. Mater. 12 (2000) 1604.
tained for hexakis(tert-butyl)ferrocene (15 kcal mol−1
)
[9] but slightly higher than that obtained for hexak-
is(pentafluorophenyl)ferrocene (11 kcal mol−1) [1]. Ex-
amination of the crystal structure of 6 (Fig. 3) suggests
that this exchange could be represented by the racem-
ization 6c ? 6c% [5], probably proceeding via 6a to
avoid the t-Buꢀt-Bu interaction in 6f. Estimating the
CpꢀFeꢀCp and CpꢀC6F5 rotational barriers was not
achieved for meso-5 and rac-5, because we were unable
to observe the slow-exchange regimes of the 19F-NMR
spectra even after cooling the samples to −90 °C.
[4] P.A. Deck, W.F. Jackson, F.R. Fronczek, Organometallics 15
(1996) 5287.
[5] (a) A.B. Pangborn, M.A. Giardello, R.H. Grubbs, R.K. Rosen,
F.J. Timmers, Organometallics 15 (1996) 1518;
(b) D.F. Shriver, M.A. Drezdzon, The Manipulation of Air-Sensi-
tive Compounds, Wiley, New York, 1986.
[6] (a) T.B. Marder, J.C. Calabrese, C.D. Roe, T.H. Tulip,
Organometallics 6 (1987) 2012;
(b) S.A. Wescott, A.K. Kakkar, G. Stringer, N.J. Taylor, T.B.
Marder, J. Organomet. Chem. 394 (1990) 777;
(c) J.W. Faller, R.H. Crabtree, A. Habib, Organometallics 4
(1985) 929;

P.A. Deck et al. / Journal of Organometallic Chemistry 637–639 (2001) 107–115
115
(d) R.T. Carl, R.P. Hughes, A.L. Rheingold, T.B. Marder, N.J.
[9] (a) J. Okuda, Top. Curr. Chem. 160 (1991) 97;
Taylor, Organometallics 7 (1988) 1613;
(e) R.T. Baker, T.H. Tulip, Organometallics 5 (1986) 839.
[7] P.G. Gassman, C.H. Winter, J. Am. Chem. Soc. 110 (1988)
6130.
(b) J. Okuda, E. Herdtweck, Chem. Ber. 121 (1988) 1899;
(c) C. Janiak, H. Schumann, Adv. Organomet. Chem. 33 (1991)
291;
(d) J. Okuda, J. Organomet. Chem. 367 (1989) C1;
&
[8] J.A. Burman, M.L. Hays, D.J. Burkey, P.S. Tanner, T.P. Hanusa,
J. Organomet. Chem. 479 (1994) 135.
(e) E.W. Abel, N.J. Long, K.G. Orrell, A.G. Osborne, V. Sik, J.
Organomet. Chem. 403 (1991) 195.