One-pot synthesis of 1,1′-bis(2-amino)ethyl-substituted ferrocenes from the reaction of spiro[2.4]hepta-4,6-diene with lithium amides: An expedient route into functionalized ferrocenophanes

Article abstract of DOI:10.1021/om049222m

The reaction of lithium amides with spiro[2.4]hepta-4,6-diene 2 affords the corresponding (2-aminoethyl)CpLi cyclopropane-ring-opened product that can be trapped in situ with FeCl₂ to provide a direct route into 1,1′-bis-aminoethyl ferrocenes 3. In addition to the desired bis-amine products, a number of interesting ferrocene byproducts have been identified. Their formation can be attributed to the associated Bronsted basicity of the lithium amide nucleophiles and their ability to participate in SET processes, opening-up alternative reaction pathways. The desired aminoethyl-substituted ferrocene products are derived primarily from direct nucleophilic cyclopropane ring-opening rather than via amidolithiation of vinylCp intermediates.

Full text of DOI:10.1021/om049222m

358
Organometallics 2005, 24, 358-366
One-Pot Synthesis of 1,1′-Bis(2-amino)ethyl-Substituted
Ferrocenes from the Reaction of
Spiro[2.4]hepta-4,6-diene with Lithium Amides: An
Expedient Route into Functionalized Ferrocenophanes
Ke´vin M. Joly,† Benson M. Kariuki,†,§ Diane M. Coe,‡ and Liam R. Cox*,†
School of Chemistry, The University of Birmingham, Edgbaston, Birmingham, B15 2TT, U.K.,
and GlaxoSmithKline, Medicinal Chemistry 1, Medicines Research Centre,
Gunnels Wood Road, Stevenage, Herts. SG1 2NY, U.K.
Received October 8, 2004
The reaction of lithium amides with spiro[2.4]hepta-4,6-diene 2 affords the corresponding
(2-aminoethyl)CpLi cyclopropane-ring-opened product that can be trapped in situ with FeCl2
to provide a direct route into 1,1′-bis-aminoethyl ferrocenes 3. In addition to the desired
bis-amine products, a number of interesting ferrocene byproducts have been identified. Their
formation can be attributed to the associated Brønsted basicity of the lithium amide
nucleophiles and their ability to participate in SET processes, opening-up alternative reaction
pathways. The desired aminoethyl-substituted ferrocene products are derived primarily from
direct nucleophilic cyclopropane ring-opening rather than via amidolithiation of vinylCp
intermediates.
Introduction
investigate their mode of coordination with a range of
metals in different oxidation states8 and their dynamic
behavior in solution. Such ligands will form complexes
in which the metal also provides a stereogenic center.
Moreover, since the ligand will be enantiomerically
pure, metal complexation will proceed through diaste-
reoisomeric transition states that should differ in en-
ergy, allowing the potential for diastereocontrol and the
isolation of products containing a single configuration
at the metal center. Complexation of an amino acid to
a Cp metal species usually results in the formation of a
mixture of diastereoisomers, although in certain cases,
excellent levels of stereocontrol can be obtained.9 Even
if diastereocontrol can be achieved, the configurational
Cyclopentadienyl (Cp) metal complexes possessing
side-chains that incorporate additional groups capable
of behaving as donors to the metal are attracting
considerable interest.1-4 If a ligand system is designed
such that the pendant donor group binds only weakly
to the metal center, it may be regarded as a temporary
or hemilabile ligand,5 capable of stabilizing reactive,
coordinatively unsaturated intermediates by reversible
coordination to the metal. This property is particularly
important in metal-catalyzed processes, and not sur-
prisingly, the application of metal complexes containing
this class of ligand as catalysts in a range of reactions,6
in particular olefin polymerization,7 has received ap-
preciable attention. Although a wide range of donor
groups can be envisaged, the synthesis of metal com-
plexes derived from Cp ligands containing appended
nitrogen donors has been the most widely studied.4b
As a long-term goal of this project, we are interested
in developing methods for preparing chiral, substituted
Cp ligands that contain multiple donor groups (hard and
soft groups based on N, O, S) in the side-chain substitu-
ent that can be derived from an amino acid. We wish to
(6) For recent examples: (a) Britovsek, G. J. P.; Cavell, K. J.; Keim,
W. J. Mol. Catal. A 1996, 110, 77-87. (b) Lindner, E.; Schmid, M.;
Wegner, P.; Nachtigal, C.; Steimann, M.; Fawzi, R. Inorg. Chim. Acta
1999, 296, 103-113. (c) Qadir, M.; Mo¨chel, T.; Hii, K. K. Tetrahedron
2000, 56, 7975-7979. (d) Standfest-Hauser, C.; Slugovc, C.; Mereiter,
K.; Schmid, R.; Kirchner, K.; Xiao, L.; Weissensteiner, W. J. Chem.
Soc., Dalton Trans. 2001, 2989-2995. (e) Clarke, M. L.; Cole-Hamilton,
D. J.; Woollins, J. D. J. Chem. Soc., Dalton Trans. 2001, 2721-2723.
(f) Faller, J. W.; Stokes-Huby, H. L.; Albrizzio, M. A. Helv. Chim. Acta
2001, 84, 3031-3042. (g) Bo¨rner, A. Eur. J. Inorg. Chem. 2001, 327-
337. (h) Kuriyama, M.; Nagai, K.; Yamada, K.-I.; Miwa, Y.; Taga, T.;
Tomioka, K. J. Am. Chem. Soc. 2002, 124, 8932-8939. (i) RajanBabu,
T. V.; Nomura, N.; Jin, J.; Nandi, M.; Park, H.; Sun, X. J. Org. Chem.
2003, 68, 8431-8446.
(7) (a) Flores, J. C.; Chien, J. C. W.; Rausch, M. D. Organometallics
1994, 13, 4140-4142. (b) Deckers, P. J. W.; Hessen, B.; Teuben, J. H.
Angew. Chem., Int. Ed. 2001, 40, 2516-2519. (c) Deckers, P. J. W.;
Hessen, B.; Teuben, J. H. Organometallics 2002, 21, 5122-5135. (d)
Groux, L. F.; Zargarian, D. Organometallics 2003, 22, 3124-3133. (e)
de Bruin, T. J. M.; Magna, L.; Raybaud, P.; Toulhoat, H. Organome-
tallics 2003, 22, 3404-3413. (f) Tobisch, S.; Ziegler, T. Organometallics
2003, 22, 5392-5405.
* Corresponding author. Fax: +44 (0)121 414 4403. Tel: +44 (0)-
121 414 3524. E-mail: l.r.cox@bham.ac.uk.
† The University of Birmingham.
§ To whom details regarding the X-ray structure should be ad-
dressed. E-mail: B.M.Kariuki@bham.ac.uk.
‡ GlaxoSmithKline.
(1) Coville, N. J.; du Plooy, K. E.; Pickl, W. Coord. Chem. Rev. 1992,
116, 1-267.
(2) Siemeling, U. Chem. Rev. 2000, 100, 1495-1526.
(3) Butenscho¨n, H. Chem. Rev. 2000, 100, 1527-1564.
(4) (a) Mu¨ller, C.; Vos, D.; Jutzi, P. J. Organomet. Chem. 2000, 600,
127-143. (b) Jutzi, P.; Redeker, T. Eur. J. Inorg. Chem. 1998, 663-
674.
(5) (a) Jeffrey, J. C.; Rauchfuss, T. B. Inorg. Chem. 1979, 18, 2658-
2666. (b) Slone, C. S.; Weinberger, D. A.; Mirkin, C. A. Prog. Inorg.
Chem. 1999, 48, 233-350.
(8) (a) Fryzuk, M. D.; Mao, S. S. H.; Zaworotko, M. J.; MacGillivray,
L. R. J. Am. Chem. Soc. 1993, 115, 5336-5337. (b) du Plooy, K. E.;
Moll, U.; Wocadlo. S.; Massa, W.; Okuda, J. Organometallics 1995, 14,
3129-3131. (c) Mu, Y.; Piers, W. E.; MacQuarrie, D. C.; Zaworotko,
M. J.; Young, V. G., Jr. Organometallics 1996, 15, 2720-2726.
(9) Grotjahn, D. B.; Joubran, C.; Hubbard. J. L. Organometallics
1996, 15, 1230-1235.
10.1021/om049222m CCC: $30.25 © 2005 American Chemical Society
Publication on Web 12/22/2004

Functionalized Ferrocenophanes
Organometallics, Vol. 24, No. 3, 2005 359
Scheme 1. Lithium-Amide-Mediated Opening of
the Cyclopropane Ring in
stability of the metal center will also be important for
any application in asymmetric synthesis. In the case of
Cp-amino acid complexes, epimerization at the metal
center is sometimes rapid at room temperature; in
others, it occurs only at high temperatures, at which
point application in asymmetric synthesis becomes a
realistic goal.10 We hope that by linking the Cp ligand
to the amino acid derivative not only will complexation
be more diastereoselective, but the resulting metal
complexes should also be more configurationally stable
owing to the chelating nature of the ligand system.
Spiro[2.4]hepta-4,6-diene
As a first target, we chose to develop a synthesis of a
more simple Cp-amino hybrid ligand system in which
an ethylene spacer separates the Cp ligand from the
nitrogen donor site.11 The vast majority of Cp-amino
ligand systems contain a methylene or SiMe2 spacer
between the Cp ligand and appended nitrogen group.12
Systems containing an additional linking atom are
comparatively unusual, which probably reflects the
relative difficulty associated with their preparation.
CpCH2CH2NR2 ligands are normally accessed by a
nucleophilic displacement reaction between NaCp and
the corresponding 2-chloroethylamino compound.13 Al-
though this multistep approach provides a very useful
route to some (2-amino)ethyl-substituted Cp ligands,
especially if the desired 2-chloroethylamino compound
is commercially available,14 we found elimination of HCl
to be a major competing process in the nucleophilic
displacement step in a number of systems that we
examined.15 An alternative and potentially more con-
vergent approach that would not be reliant on the
synthesis of 2-chloroethylamino ligand precursors was
therefore sought.
Scheme 2. Two-Step Synthesis of
1,1′-Bis(2-diethylamino)ethylferrocene 3aa
a Reagents and conditions: (a) CpH (1.2 equiv), NaH (2.6
equiv), THF, RT, 12 h, 62% (X ) Cl), 46% (X ) Br); (b) LiNEt2
(1.0 equiv), THF, reflux, 15 h, then anhydrous FeCl2, (0.55
equiv), reflux, 8 h (42%).
direct synthesis of our desired ligand system. A search
in the literature revealed that a large number of soft
nucleophiles (based on S, As, P) have already been used
successfully to open the cyclopropane ring in 2;2,16 in
sharp contrast, the use of harder nitrogen and oxygen
nucleophiles has received scant attention, with only
isolated examples being reported to date.17,18 The only
example that we could find in which an amine nucleo-
phile had been reacted with 2 involved the rather
unusual lithium amide 1, which provided the desired
Cp-amino hybrid ligand in good yield (Scheme 1).17 We
therefore set about investigating whether more conven-
tional nitrogen nucleophiles could be used to prepare
substituted (2-amino)ethylCp ligands from 2 and pro-
vide a potentially facile and general entry into this class
of ligands.
Employing spiro[2.4]hepta-4,6-diene 2 as a precursor
to both the Cp ligand and the two carbons required for
our ethylene spacer seemed to be a particularly attrac-
tive option worth exploring. Release of ring-strain,
accompanied by the formation of a relatively stable Cp
anion, would hopefully render opening of the cyclopro-
pane by a nitrogen nucleophile facile and present a
Results and Discussion
(10) (a) Carmona, D.; Lahoz, F. J.; Atencio, R.; Oro, L. A.; Lamata
M. P.; San Jose´, E. Tetrahedron: Asymmetry 1993, 4, 1425-1428. (b)
Carmona, D.; Mendoza, A.; Lahoz, F. J.; Oro, L. A.; Lamata M. P.;
San Jose´, E. J. Organomet. Chem. 1990, 396, C17-C21.
(11) Such a C2 linker between the two ligand donor sites should
hopefully maximize the likelihood that the side-chain amino group
would interact with the metal center. See for example: (a) Ti
complex: Herrmann, W. A.; Morawietz, M. J. A.; Priermeier, T.;
Mashima, K. J. Organomet. Chem. 1995, 486, 291-295. (b) Cr
complex: Do¨hring, A.; Go¨hre, J.; Jolly, P. W.; Kryger B.; Rust, J.;
Verhovnik, G. P. J. Organometallics 2000, 19, 388-402. (c) Co
complex: Jutzi, P.; Kristen, M. O.; Dahlhaus, J.; Neumann, B.;
Stammler, H.-G. Organometallics 1993, 12, 2980-2985.
(12) For some recent examples of CpCH2NR2 and CpSiMe2NR2
complexes: (a) Strauch, J. W.; Petersen, J. L. Organometallics 2001,
20, 2623-2630. (b) Hultzsch, K. C.; Voth, P.; Beckerle, K.; Spaniol, T.
P.; Okuda, J. Organometallics 2000, 19, 228-243. (c) Amor, F.; Butt,
A.; du Plooy, K. E.; Spaniol, T. P. Okuda, J. Organometallics 1998,
17, 5836-5849. (d) Seo, H.; Park, H.; Kim, B. Y.; Lee, J. H.; Son, S.
U.; Chung, Y. K. Organometallics 2003, 22, 618-620. (e) Knoesen, O.;
Wessels, P. L.; Gorls, H.; Lotz, S. Inorg. Chem. 2001, 40, 1199-1205.
(f) Gambs, C.; Consiglio, G.; Togni, A. Helv. Chim. Acta 2001, 84, 3105-
3126.
(13) (a) Jutzi, P.; Dahlhaus, J. Coord. Chem. Rev. 1994, 137, 179-
199. (b) Bradley, S.; Camm, K. D.; Liu X.; McGowan, P. C.; Mumtaz,
R.; Oughton, K. A.; Podesta, T. J.; Thornton-Pett, M. Inorg. Chem.
2002, 41, 715-726.
Synthesis of 1,1′-Bis(2-amino)ethyl-Substituted
Ferrocenes. Although spiro[2.4]hepta-4,6-diene 2 is
commercially available, we found it more convenient to
prepare this starting material ourselves. Using a modi-
fication of the published procedure,19 treatment of a
THF solution of freshly cracked cyclopentadiene and 1,2-
dichloroethane20 with 2.6 equiv of NaH provided spiro-
[2.4]hepta-4,6-diene 2 in good yield (Scheme 2).21 Using
these conditions, diene 222 was obtained in high purity
(>95%)23 by distillation.
(16) Kauffmann, T.; Ennen, J.; Lhotak, H.; Rensing, A.; Steinseifer,
F.; Woltermann, A. Angew. Chem., Int. Ed. Engl. 1980, 19, 328-329.
(17) For the use of a nitrogen nucleophile: Qian, B.; Henling, L.
M.; Peters, J. C. Organometallics 2000, 19, 2805-2808.
(18) For the use of an oxygen nucleophile: ref 11b.
(19) Wilcox, C. F., Jr.; Craig, R. R. J. Am. Chem. Soc. 1961, 83,
3866-3871.
(20) The use of 1,2-dibromoethane as an alternative bis-electrophile
offered no significant advantage and provided diene 2 in reduced yield
(46%).
(21) We found it particularly important to use an excess of cyclo-
pentadiene in this reaction to ensure complete consumption of the 1,2-
dichloroethane, since any residual starting material proved impossible
to separate from the product.
(22) 2 can be stored under a nitrogen atmosphere at 4 °C for several
months without significant decomposition.
(14) For example, ClCH2CH2NH2‚HCl, ClCH2CH2NEt2‚HCl, ClCH2-
CH2NMe2‚HCl, and ClCH2CH2NiPr2‚HCl are commercially available.
(15) For example, reaction of ClCH2CH2NTsCH(CH2Ph)CH2OTHP
with NaCp resulted in exclusive formation of the enamine elimination
product.

360 Organometallics, Vol. 24, No. 3, 2005
Joly et al.
Table 1. One-Pot Synthesis of
1,1′-Bis(2-amino)ethyl-Substituted Ferrocenes 3
lithium amide
LiNEt2
product
isolated yield (%)
3a
3b
42
52
LiN(CH2)4CH2
3c
3d
3e
52
65
39
LiN(CH2)2N(CH3)CH2CH2
LiN(CH2)2OCH2CH2
LiN(CH2)3CH2
LiNHBn
LiNHBu
Figure 1. Reaction of spiro[2.4]hepta-4,6-diene with
LiNEt2 and then FeCl2 provided a range of ferrocene
products. aIsolated yield after purification by column
chromatography. bYield calculated from GC analysis of the
crude reaction mixture.
3f
3g
12
7
Since we expected the (2-amino)ethylCp ligands to be
unstable and difficult to isolate, we elected to test the
efficiency of our cyclopropane-opening reaction by trap-
ping the intermediate Cp anion directly with Fe(II). This
would generate a ferrocene derivative, which we ex-
pected would be much more readily characterized.
Ferrocenes containing pendant amino groups are them-
selves of interest and find use in a wide range of
applications.24 In a first experiment, heating a solution
of 2 with diethylamine in THF under reflux for 12 h,
followed by the addition of anhydrous FeCl2, failed to
provide the desired ferrocene product. No polar, ring-
opened products were obtained, and the starting diene
2 slowly decomposed on prolonged heating. Moving to
the more nucleophilic lithium amide improved mat-
ters: heating a 1:1 mixture of LiNEt2 and 2 in THF for
15 h, followed by the addition of FeCl2 and further
heating for 8 h, provided the desired 1,1′-bis(2-dieth-
ylamino)ethyl-substituted ferrocene product 3a in mod-
erate yield (Scheme 2).
Carrying out the reaction at room temperature in an
effort to reduce the loss of starting material, presumably
through thermally induced polymerization pathways,
failed to improve matters; in this case the desired ring-
opening reaction simply became extremely sluggish.
Neither the use of other solvents (Et2O, toluene) nor the
inclusion of 1 equiv of TMEDA (per lithium amide) in
the reaction mixture provided any improvement in the
yield of the desired product. Since the basicity of the
lithium amide nucleophile was another source of side-
reactions (see below), we also attempted the ring-
opening reaction using the corresponding magnesium
amide Mg(NEt2)2, prepared from Bu2Mg and HNEt2,25
and the aluminum amide prepared from Me3Al and
HNEt2.26 While both of these amide nucleophiles are
significantly less basic than their lithium congeners,
they unfortunately also lacked the necessary nucleo-
philicity to effect ring-opening.
A range of primary and secondary lithium amides was
then investigated using our optimized conditions, and
the results are summarized in Table 1. The desired 1,1′-
bis(2-amino)ethyl-substituted ferrocenes 3a-g were
isolated from this one-pot operation, after aqueous
workup and purification by flash column chromatogra-
phy or recrystallization, in low to good yields, depending
on the starting amine. The best results were generally
obtained with cyclic secondary lithium amides; acyclic
lithium amides, in particular those derived from pri-
mary amines, provided lower yields of the desired
products, which is reflective of their reduced nucleophi-
licity. Attempted ring-opening with lithium amides
derived from oxazolidinone and imidazole failed. Crys-
tals of the bis-morpholine complex 3d were grown from
EtOAc/petroleum ether (bp 40-60 °C), and the structure
was solved by single-crystal X-ray diffraction, thus
confirming our structural assignment made from spec-
troscopic analyses (see Supporting Information).
Characterization of Reaction Byproducts. In
addition to the desired 1,1′-bis(2-amino)ethyl ferrocenes
3, we observed a range of other interesting ferrocene
products, which helped to account for some of the mass
balance in the reaction. For example, when spiro[2.4]-
hepta-4,6-diene (2) was treated with LiNEt2, followed
by FeCl2, five other ferrocene products, 4-8, were
identified, in addition to the desired product 3a (Figure
1).
(23) Even after distillation, spiro[2.4]hepta-4,6-diene always con-
tained a residual amount of CpH (<5%).
(24) Water-soluble ferrocene and ferrocenium species (amino-
substituted ferrocenes are readily rendered water-soluble by quater-
nization or simple protonation) exhibit a range of useful biological
properties: (a) DNA binding: Georgopoulou, A. S.; Mingos, D. M. P.;
White, A. J. P.; Williams, D. J.; Horrocks, B. R.; Houlton, A. J. Chem.
Soc., Dalton Trans. 2000, 2969-2974. (b) Antitumor activity: Kopf-
Maier, P.; Ko¨pf, H.; Neuse, E. W. Angew. Chem., Int. Ed. Engl. 1984,
23, 456-457. This property possibly arises through the ability of
ferrocenium species to generate hydroxyl radicals, which are commonly
implicated in DNA damage: Osella, D.; Ferrali, M.; Zanello, P.; Laschi,
F.; Fontani, M.; Nervi, C.; Cavigiolo, G. Inorg. Chim. Acta 2000, 306,
42-48. (c) Antimalarial activity: (i) Biot, C.; Delhaes, L.; Maciejewski,
L. A.; Mortuaire, M.; Camus, D.; Dive, D.; Brocard, J. S. Eur. J. Med.
Chem. 2000, 35, 707-714. (ii) Biot, C.; Delhaes, L.; Abessolo, H.;
Domarle, O.; Maciejewski, L. A.; Mortuaire, M.; Delcourt, P.; Deloron,
P.; Camus, D.; Dive, D.; Brocard, J. S. J. Organomet. Chem. 1999, 589,
59-65. (d) Redox sensors for the commercially important flavoenzyme
glucose oxidase: Forrow, N. J.; Sanghera, G. S.; Walters, S. J. J. Chem.
Soc., Dalton Trans. 2002, 3187-3194. (e) Anion-recognition; Beer, P.
D. Acc. Chem. Res. 1998, 31, 71-80. (f) Ligands in Pd-catalyzed cross-
coupling processes: Weng, Z.; Teo, S.; Koh, L. L.; Hor, T. S. A.
Organometallics 2004, 23, 3603-3609.
The formation of the first byproduct, 1,1′-divinyl
ferrocene 4,27 can be understood by assuming that the
(25) (a) Lesse`ne, G.; Tripoli, R.; Cazeau, P.; Biran, C.; Bordeau, M.
Tetrahedron Lett. 1999, 40, 4037-4040. (b) Henderson, K. W.; Kerr,
W. J. Chem. Eur. J. 2001, 7, 3430-3437.
(26) Compe`re, D.; Marazano, C.; Das, B. C. J. Org. Chem. 1999, 64,
4528-4532.
(27) Rausch, M. D.; Siegel, A. J. Organomet. Chem. 1968, 11, 317-
324.

Functionalized Ferrocenophanes
Organometallics, Vol. 24, No. 3, 2005 361
Scheme 3. Reaction of Lithium Amides with
Spiro[2.4]hepta-4,6-diene
ferrocene was recovered intact. In contrast, when a THF
solution of bis(2-pyrrolidino)ethyl ferrocene 3e at reflux
was treated with LiNEt2, analysis of the reaction
mixture after prolonged heating (24 h) revealed the
presence of some monoelimination product 5e (31%
isolated yield), although the major component of the
reaction mixture remained the starting material (60%
isolated yield). From these reactions, it would appear
that vinyl-substituted ferrocenes can be formed from the
desired bis-amino ferrocene product although only when
the more strongly basic acyclic lithium amides are
employed; the major source of these compounds remains
the direct pathway outlined in Scheme 3.
The occurrence of the third byproduct 6 containing
an unsubstituted Cp ligand, in trace (<3%) amounts,
was readily rationalized by the fact that the spiro[2.4]-
hepta-4,6-diene 2 used in the complexation studies
invariably contained a residual amount of cyclopenta-
diene (<5%).23 The origins of the two remaining side-
products, however, are more interesting. GC-MS and
NMR analysis of 7 indicated the presence of an ethylCp
ligand. We propose that the formation of such a ligand,
and therefore of ethyl-substituted ferrocene 7, proceeds
via a single-electron transfer (SET) process (Scheme
3): one-electron reduction of 2 by the lithium amide,
followed by cyclopropane ring-opening and abstraction
of a hydrogen atom,28 provides a route to EtCpLi. The
formation of such products in larger quantities has been
observed when 2 is treated with butyllithium reagents,29
which are well known to react through SET processes.
The structure of the fourth byproduct 8 eluded us for
some time. While mass spectral analysis indicated the
presence of a Cp ligand with a molecular formula C7H7,
i.e., equivalent to a vinylCp group, we had already
attributed this structure to ferrocene 5. We therefore
had to assume that 8 contained a Cp ligand and a
double-bond equivalent, namely, an appended ring.
Ultimately, chemical derivatization studies allowed us
to confirm this and the identity of the compound as 8
(Scheme 4). Thus, reaction of a mixture of ferrocenes
5a, 7a, and 8a (in this case, the mixture contained
negligible amounts of ferrocene 6a) with methyl iodide
generated the corresponding tetraalkylammonium salts.
Without isolation, treatment of this mixture with KOtBu
lithium amide behaves as a base, rather than as a
nucleophile, and abstracts a proton from one of the
cyclopropane methylene groups (Scheme 3). The result-
ing cyclopropanyl anion will then be in equilibrium with
a much more stable vinylCp anion in which the negative
charge has been delocalized into the Cp ring. The
additional aromatic stabilization in the vinylCp species
presumably serves to help “siphon off” any of the
conjugate base that is formed: it is therefore effectively
acting as a thermodynamic sink and readily accounts
for the formation of both this byproduct and 5 (see
below). Ferrocenes 4 and 5 were produced in varying
amounts depending on the relative basicity of the
lithium amide employed; thus amides derived from
cyclic amines generally provided smaller quantities of
these byproducts, which was clearly associated with a
concomitant increase in the yield of the desired product
3.
The identification of the remaining side-products (5-
8) proved to be much more difficult, owing to our
inability to separate them by either silica gel flash
column chromatography or HPLC, irrespective of the
amine employed; as a consequence, they were charac-
terized as a mixture. Analysis of this mixture by GC-
MS revealed the presence of four compounds of molec-
ular mass 285, 311, 311, and 313 Da (using HNEt2 as
the amine). Fragmentation characteristic of the (2-
amino)ethylCp ligand was present in the EI mass
spectra in all cases and suggested a series of ferrocenes
each containing a single (2-amino)ethyl-substituted Cp
ligand.
Since the bis(2-amino)ethyl ferrocene 3 and bis-vinyl
ferrocene 4 were observed, the isolation of ferrocene 5
containing one of each ligand was predictable. We were
keen to check whether the ferrocenes 4 and 5 containing
a vinylCp ligand were derived from the pathway de-
scribed in Scheme 3 and were not the products from a
secondary elimination reaction between our target bis-
(2-amino)ethyl ferrocene 3 and unreacted lithium amide
since this would obviously be compromising the yield
of our target compounds. To this end, bis(2-pyrrolidino)-
ethyl ferrocene 3e was treated with lithium morpho-
linide at reflux for 24 h. Analysis of the reaction mixture
by GC showed that this particular lithium amide did
not effect eliminationsnor amine substitution for that
mattersunder the reaction conditions, and the starting
t
in BuOH effected a Hoffmann elimination to provide
the corresponding olefin products.30 Subsequent hydro-
genation of the olefin functionality present in these
molecules reduced the number of compounds from three
to two, providing 9 and 10. At this stage we were luckily
able to separate the compounds by reversed-phase
HPLC and obtained a pure sample of the derivatized
final mystery product 10. Extensive NMR studies, and
comparison of these data with related compounds
reported in the literature,31,32 allowed us to assign the
structure of 10 to a ferrocenocyclobutene.
It is interesting to speculate on the mechanism of
formation of 8. Some time ago, Oda and Breslow showed
(28) We have not identified the source of hydrogen: carrying out
the reaction in d8-THF did not lead to any deuterium incorporation in
the product.
(29) Kauffmann, T.; Bisling, M.; Ko¨nig, R.; Rensing, A.; Steinseifer,
F. Chem. Ber. 1985, 118, 4517-4530.
(30) Rodrigo, R.; Manske, R. H. F.; MacLean, D. B.; Baczynskyj, L.;
Saunders, J. K. Can. J. Chem. 1972, 50, 853-861.
(31) Trahanovsky, W. S.; Ferguson, J. M. Organometallics 1992, 11,
2006-2007.
(32) Oda, M.; Breslow, R. Tetrahedron Lett. 1973, 2537-2540.

362 Organometallics, Vol. 24, No. 3, 2005
Joly et al.
Table 2. Amidolithiation of 1,1′-Divinyl Ferrocene
Scheme 4. Chemical Derivatization Enabled the
Structure of Ferrocene 8 To Be Determineda
isolated isolated isolated
reaction
conditions
yield
yield
yield
entry
amine
3 (%)a
5 (%)a
14 (%)a
1
morpholine THF, reflux
morpholine THF, RT
69
30
5
5
41
28
0
15
0
42
0
2b
3
Et2NH
THF, reflux
THF, reflux
4c
iPr2NH
0
a Isolated yields following purification by flash column chroma-
tography. b Starting material 4 was also recovered (10%). c Start-
ing material was recovered.
An Alternative Route to the Desired Products
and Adventitious Entry into Some Novel [3]Fer-
rocenophanes. In an effort to improve the yield of our
desired product, we entertained the possibility that the
lithium amide nucleophile might add to vinylCp ligands
in analogy with a similar reaction that has been
reported with styrene.34 If this were to be the case, and
assuming elimination pathways were not important, as
seems to be the case, then using an excess of the lithium
amide nucleophile might allow an increased yield of the
desired bis-amino ferrocenes; vinyl ferrocenes 4 and 5
would effectively be intermediates and, providing the
reaction was given sufficient time, would then converge
on the desired bis-amino ferrocene compound 3. To test
this possibility, 1,1′-divinyl ferrocene 4 was treated
separately with a selection of lithium amides under a
range of reaction conditions (Table 2). The progress of
the reactions was monitored by GC.
a Reagents and conditions: (a) MeI (5 equiv), THF, RT, 15
h; (b) tBuOK (8 equiv), tBuOH, reflux, 15 h; (c) H2 (1 atm), 5%
Pd/C, EtOH, 5 h; (d) reversed-phase HPLC.
Scheme 5. Isomerization of
Spiro[2.4]hepta-4,6-diene and
Bicyclo[3.2.0]hepta-1,3-diene
Scheme 6. Possible Routes to
Ferrocenocyclobutene 8
We were gratified to observe that the more nucleo-
philic lithium morpholinide did indeed add fairly readily
to the vinyl substituents in 4 to provide the bis-amino
ferrocene 3d in very good yield, along with a small
amount of the monoaddition product 5d, after heating
at reflux for 24 h. In addition to these two expected
addition products we also isolated the novel [3]ferro-
cenophane 14d, whose formation can be rationalized by
the cascade process outlined in Scheme 7. Thus inter-
molecular amidolithiation of one vinyl group affords the
organolithium intermediate 15d, which can then react
in one of two ways: protonation of this intermediate,
presumably with the THF solvent acting as the proton
source, provides the monoaddition product 5d, which
can then react in a similar fashion to provide the desired
1,1′-bis(2-amino)ethyl ferrocene adduct 3d. Alterna-
tively, 15d may react with the remaining olefin through
an intramolecular carbolithiation pathway to provide
[3]ferrocenophane 14d after workup.35
that protonation of anion 11 provided bicyclo[3.2.0]-
hepta-1,3-diene 12, which underwent an energetically
favorable and rapid [1,5]-alkyl shift at -50 °C to provide
spiro[2.4]hepta-4,6-diene 2 (Scheme 5).32 Kloosterziel
and co-workers have shown that the reverse reaction
can also be effected by thermolysis at 345-400 °C
(Scheme 5).33
The fact that such high temperatures are required to
isomerize 2 to 12 suggests a [1,5]-alkyl shift pathway
directly on this neutral species is not the source of the
anion 11 required for the formation of ferrocenocy-
clobutene 8, although we do not discount the possibility
that this rearrangement may be more facile on the
carbanion 13 (Scheme 6), which we have already
proposed to be an intermediate in the synthesis of vinyl
ferrocenes 4 and 5 (Scheme 3). Other mechanisms
proceeding via carbanion 13 can be envisaged. For
example, a tandem ring-opening/carbene 1,4-insertion
mechanism would also provide a route to anion 11
(Scheme 6).
(34) (a) Beller, M.; Breindl, C. Tetrahedron 1998, 54, 6359-6368.
(b) Seijas, J. A.; Va´zquez-Tato, M. P.; Entenza, C.; Mart´ınez, M. M.;
OÄ nega, M. G.; Veiga, S. Tetrahedron Lett. 1998, 39, 5073-5076. (c)
Koradin, C.; Dohle, W.; Rodriguez, A. L.; Schmid, B.; Knochel, P.
Tetrahedron 2003, 69, 1571-1587. (d) Hamana, H.; Iwasaki, F.;
Nagashima, H.; Hattori, K.; Hagiwara, T.; Narita, T. Bull. Chem. Soc.
Jpn. 1992, 65, 1109-1113.
(35) For recent examples of intramolecular carbolithiation reac-
tions: (a) Deng, K.; Bensari, A.; Cohen, T. J. Am. Chem. Soc. 2002,
124, 12106-12107. (b) Coldham, I.; Price, K. N.; Rathmell, R. E. Org.
Biomol. Chem. 2003, 2111-2119. (c) Bailey, W. F.; Luderer, M. R.;
Mealy, M. J. Tetrahedron Lett. 2003, 44, 5303-5305. (d) Wei, X.;
Taylor, R. J. K. Tetrahedron Lett. 2003, 44, 7143-7146. (e) Yus, M.;
Ortiz, R.; Huerta, F. F. Tetrahedron 2003, 59, 8525-8542.
(33) Krekels, J. M. E.; de Haan, J. W.; Kloosterziel, H. Tetrahedron
Lett. 1970, 2751-2754.

Functionalized Ferrocenophanes
Organometallics, Vol. 24, No. 3, 2005 363
Scheme 7. Amidolithiation of 1,1′-Divinyl
Ferrocene 4 Provided an Alternative Entry into
the Desired Products and a Route to a Novel
Amine-Functionalized [3]Ferrocenophane
lecular carbolithiation becomes competitive. Not sur-
prisingly, the use of the very poor amide nucleophile
LDA led to no reaction, with starting material being
recovered.
Conclusions
In summary, we have shown that the action of lithium
amides on spiro[2.4]hepta-4,6-diene 2 provides the cor-
responding (2-aminoethyl)CpLi species that can be
trapped in situ with FeCl2 to provide 1,1′-bis(amino)-
ethyl ferrocenes. A number of interesting side-products,
which we propose result from SET processes and the
associated Brønsted basicity of the lithium amide nu-
cleophiles, have also been identified and led us to
investigate the ring-opening reaction in more detail.
Control experiments have shown that the major path-
way to the target bis-amines is through a nucleophilic
opening of the cyclopropane moiety in 2 rather than a
possible competing amidolithiation of vinylCp interme-
diates. The simplicity of this operation and rapidity of
complex assembly provides a useful entry into this class
of amino-substituted ferrocenes, although the length of
some of the reaction steps still detracts somewhat from
the method. We are actively seeking alternative syn-
thetic approaches that address this specific issue.
Experimental Section
General Comments. Infrared spectra were recorded either
neat as thin films between sodium chloride disks or as a Nujol
mull between sodium chloride disks. The intensity of each band
is described as s (strong), m (medium), or w (weak) and with
the prefix v (very) and suffix br (broad) where appropriate.
1H NMR and 13C NMR spectra were recorded in CDCl3 at 500
and 125 MHz, 400 and 100 MHz, or 300 and 75 MHz,
respectively. Chemical shifts are reported as δ values (ppm)
referenced to the following solvent signals: CHCl3, δΗ 7.26;
CDCl3, δC 77.0. The term, “stack” is used to describe a region
where resonances arising from nonequivalent nuclei are
coincident, and multiplet, m, to describe a region where
resonances arising from a single nucleus (or equivalent nuclei)
are coincident but coupling constants cannot be readily as-
signed. Connectivities were deduced from COSY90, HSQC, and
HMBC experiments. Mass spectra were recorded utilizing
electrospray ionization (and a methanol mobile phase), or
m-nitrobenzyl alcohol (3-NOBA) as the matrix, and are
reported as m/z (%). HRMS were recorded using a lock mass
incorporated into the mobile phase. Melting points were
determined using open capillaries and are uncorrected.
Reversed-phase preparative HPLC was performed using a
C18(2) 100 × 21.2l D 5 µM column at 20 mL min-1 and
monitored using a photodiode array detector (used at 200 and
330 nm) using a helium-degassed HPLC grade water/aceto-
nitrile gradient, with formic acid additive [t ) 0-1 min: 20%
of A (0.1% formic acid in H2O):80% of B (95% MeCN, 5% of
0.05% formic acid in H2O) then gradient to t ) 12 min: 5%
A:95% B, then gradient to t ) 14 min: 1% A:99% B].
Analytical GC data were collected with FID detection:
carrier gas, He; 5% phenyl 95% dimethylpolysiloxane column,
15 m × 0.53 mm i.d.
The fact that the bis-amino ferrocene 3d was the
major product suggests the second carbolithiation step
is either slow or possibly reversible. These experiments
seemed to confirm our predictions that vinyl ferrocenes
could be treated as intermediates. However, using an
excess of lithium morpholinide in the reaction with 2,
or adding excess after the addition of FeCl2, and heating
the mixture at reflux for a prolonged period (15 h), failed
to lead to a significant improvement in the yield of 3d.
Carrying out the amidolithiation reaction at room
temperature for 24 h served to reduce the rate of
addition, although now we were able to completely
suppress ferrocenophane formation. Although styrene
derivatives have been shown to react with lithium
amides at 0 °C,34b carrying out our reaction at this lower
temperature simply resulted in the recovery of starting
material after 24 h. The reduced susceptibility of our
vinyl group toward amidolithiation compared with that
in styrene was further emphasized when we observed
no reaction at all when morpholine in the presence of
substoichiometric quantities (10 mol %) of nBuLi at
reflux was employed.34a
Performing the same reaction between 1,1′-divinyl
ferrocene 4 and LiNEt2 proved less successful. The
reduced nucleophilicity of this lithium amide provided
the double-addition product 3a in greatly reduced yield
along with the monoaddition product 5a and the cor-
responding [3]ferrocenophane 14a in 28% and 42%
yield, respectively, after 24 h at reflux. The increased
amount of [3]ferrocenophane product probably reflects
the reduced rate of amidolithiation such that intramo-
Reactions were monitored by thin-layer chromatography
using precoated glass-backed silica-rapid plates (60A F254) and
visualized by UV detection (at 254 nm) and with ammonium
molybdate(IV)-cerium(IV) sulfate staining dip. Column chro-
matography was performed on silica gel (particle size 40-63
µm mesh).
All reactions were conducted in oven-dried (140 °C) or flame-
dried glassware under a nitrogen atmosphere, and at ambient

364 Organometallics, Vol. 24, No. 3, 2005
Joly et al.
temperature (20 to 25 °C) unless otherwise stated, with
magnetic stirring. Volumes of 1 mL or less were measured and
dispensed with gastight syringes. Evaporation and concentra-
tion under reduced pressure were performed at 50-500 mbar.
Residual solvent was removed under high vacuum (<1 mbar).
All reagents were obtained from commercial sources and
used without further purification unless stated otherwise.
Cyclopentadiene was obtained by cracking dicyclopentadiene
(160-170 °C) under N2 and collected by fractional distillation.36
All amines were distilled under a N2 atmosphere from KOH
and stored under N2 at room temperature over activated 4 Å
molecular sieves (activated by heating under a vacuum for 15
min with a Bunsen flame immediately before use). Dichlo-
romethane was freshly distilled under nitrogen from CaH2.
Tetrahydrofuran and diethyl ether were freshly distilled under
nitrogen from sodium benzophenone ketyl. All solutions are
aqueous and saturated unless stated otherwise.
General Procedure for Formation of 1,1′-Bis(2-amino)-
ethyl-Substituted Ferrocenes (3). nBuLi (7.19 mL of a 1.6
M solution in hexane, 12.0 mmol) was added to a solution of
the amine (13 mmol) in THF (21 mL) at 0 °C. The solution
was then stirred at RT for 15 min and then transferred via
syringe to a solution of spiro[2.4]hepta-4,6-diene 2 (10 mmol)
in THF (5 mL) at RT. The resulting orange solution was heated
under reflux overnight (∼12-14 h) and then cooled to RT,
whereupon anhydrous FeCl2 (5 mmol) was added. The dark
brown solution was heated under reflux for a further 18 h.
After cooling to RT, Et2O (200 mL) and NH4Cl solution (200
mL) were added. The mixture was basified to pH 9 with NaOH
solution (2 M), and the two phases were separated. The
aqueous phase was extracted with Et2O (2 × 100 mL) and then
with CH2Cl2 (2 × 100 mL). The combined organic fractions
were washed with brine (300 mL) and then dried (MgSO4).
The drying agent was removed by filtration and the filtrate
concentrated under reduced pressure to provide a red oil,
which was purified by flash column chromatography.
mmol), spiro[2.4]hepta-4,6-diene 2 (0.5 mL, 4.85 mmol), and
FeCl2 (0.31 g, 2.4 mmol). After workup, purification by flash
column chromatography (94% EtOAc, 5% EtOH, 1% Et3N)
afforded bis-amine 3d (650 mg, 65%) as a red powder: mp 70-
72 °C; Rf ) 0.30 (94% EtOAc, 5% EtOH, 1% Et3N); λmax
(CH2Cl2)/nm 435 (ꢀ/dm3 mol-1 cm-1 160), 322 (166); IR (Nujol
mull) 3070w, 1377m, 1131m, 1116s, 1038m cm-1; 1H NMR (300
MHz) δ 2.42-2.55 (stack, 16H), 3.70-3.74 (stack, 8H), 3.99
(s, 8H); 13C NMR (75 MHz) δ 25.9 (CH2), 52.9 (CH2), 59.6 (CH2),
66.1 (CH2), 67.1 (CH), 67.9 (CH), 85.6 (quat. C); MS (EI) m/z
413 ([M + 1]+, 47%), 312 (5, [M - CH2N(CH2CH2)O]+), 100
(100, [O(CH2CH2)2NCH2)]+). Anal. Calcd for C22H32FeN2O2: C,
64.08; H, 7.82; N, 6.79. Found: C, 64.09; H, 7.91; N, 6.70.
1,1′-Bis[2-pyrrolidin-1-yl]ethylferrocene (3e). Bis-amino
ferrocene 3e was prepared from pyrrolidine (0.52 mL, 6.3
mmol), spiro[2.4]hepta-4,6-diene 2 (0.5 mL, 4.85 mmol), and
FeCl2 (0.31 g, 2.4 mmol). After workup, purification by flash
column chromatography (90% acetone, 9% EtOH, 1% Et3N)
afforded bis-amine 3e (360 mg, 39%) as a red powder: mp 46-
48 °C; Rf ) 0.07 (94% EtOAc, 5% EtOH, 1% Et3N); IR (Nujol
mull) 3084s, 1460s, 1378m, 1350m, 1329m, 1280m, 1144m,
1114m, 1038s cm-1; 1H NMR (300 MHz) δ 1.73-1.85 (m, 8H),
2.43-2.66 (stack, 16H), 3.99 (s, 8H); 13C NMR (75 MHz) δ 23.4
(CH2), 29.1 (CH2), 54.2 (CH2), 57.8 (CH2), 67.9 (CH), 68.7 (CH),
86.8 (quat. C); MS (EI) m/z 380 ([M]+, 22%), 84 (100, [(CH2CH2)2-
NCH2)]+); HRMS calcd for C22H33FeN2 [M + H]+ 381.1993,
found 381.1988.
1,1′-Bis[2-benzylamin-1-yl]ethylferrocene (3f). Bis-
amino ferrocene 3f was prepared from BnNH2 (0.66 mL, 6.3
mmol), spiro[2.4]hepta-4,6-diene 2 (0.5 mL, 4.85 mmol), and
FeCl2 (0.31 g, 2.4 mmol). After workup, purification by flash
column chromatography (94% EtOAc, 5% EtOH, 1% Et3N)
afforded bis-amine 3f (72 mg, 12%) as an orange oil: Rf ) 0.29
(94% EtOAc, 5% EtOH, 1% Et3N); IR (film) 3312w, 3085m,
3026m, 1454s, 1118m, 1040m, 1027m cm-1 1H NMR (300
;
MHz) δ 2.30 (br s, 2H), 2.54 (t, J 7.0, 4H), 2.75 (t, J 7.0, 4H),
3.79 (s, 4H), 3.98 (s, 8H), 7.19-7.39 (stack, 10H); 13C NMR
(75 MHz) δ 29.5 (CH2), 50.2 (CH2), 53.6 (CH2), 68.0 (CH), 68.9
(CH), 86.4 (quat. C), 127.2 (CH), 128.4 (CH), 128.7 (CH), 140.5
(quat. C); MS (TOF ES+) m/z 453.2 ([M + H]+, 100%); HRMS
calcd for C28H33FeN2 [M + H]+ 453.1993, found 453.1998.
1,1′-Bis[2-butylamin-1-yl]ethylferrocene (3g). Bis-amino
ferrocene 3g was prepared from BuNH2 (1.00 mL, 10.1 mmol),
spiro[2.4]hepta-4,6-diene 2 (0.8 mL, 7.75 mmol), and FeCl2
(0.31 g, 2.4 mmol). After workup, purification by flash column
chromatography (90% acetone, 8% EtOH, 2% Et3N) afforded
bis-amine 3g (100 mg, 8%) as an orange oil: Rf ) 0.28 (90%
acetone, 8% EtOH, 2% Et3N); IR (film) 3420w, 3086w, 1470s,
1,1′-Bis[2-diethylamino]ethylferrocene (3a). Bis-amino
ferrocene 3a was prepared from Et2NH (0.65 mL, 6.3 mmol),
spiro[2.4]hepta-4,6-diene 2 (0.5 mL, 4.85 mmol), and FeCl2
(0.31 g, 2.4 mmol). After workup, purification by flash column
chromatography (94% EtOAc, 4% EtOH, 1% Et3N) afforded
bis-amine 3a (394 mg, 42%) as a red-orange oil: Rf ) 0.26 (94%
EtOAc, 5% EtOH, 1% Et3N); IR (film) 3089m, 1471m, 1383m,
1293m, 1206m, 1069m, 1040m cm-1 1H NMR (300 MHz) δ
;
1.03 (t, J 7.2, 12H), 2.41-2.61 (stack, 16H), 3.98 (br s, 8H);
13C NMR (75 MHz) δ 12.3 (CH3), 27.4 (CH2), 47.3 (CH2), 54.8
(CH2), 68.3 (CH), 69.1 (CH), 87.4 (quat. C); MS (EI) m/z 384
([M]+, 40%), 86 (100, [(CH3CH2)2NCH2)]+); HRMS calcd for
C22H37FeN2 [M + H]+ 385.2306, found 385.2294.
1411m, 1378m, 1304m, 1265m cm-1 1H NMR (300 MHz) δ
;
0.88 (t, J 7.0, 6H), 1.20-1.52 (stack, 8H), 1.73 (br s, 2H), 2.44-
2.76 (stack, 12H), 3.98 (s, 8H); 13C NMR (75 MHz) δ 13.9 (CH3),
20.4 (CH2), 29.9 (CH2), 32.1 (CH2), 49.5 (CH2), 51.1 (CH2), 68.0
(CH), 68.9 (CH), 86.4 (quat. C); MS (TOF ES+) m/z 385.2 ([M
+ H]+, 100%); HRMS calcd for C22H37FeN2 [M + H]+ 385.2306,
found 385.2318.
1,1′-Bis[2-methylpiperazin-1-yl]ethylferrocene (3c). Bis-
amino ferrocene 3c was prepared from N-methylpiperazine
(0.28 mL, 2.5 mmol), spiro[2.4]hepta-4,6-diene 2 (0.2 mL, 1.94
mmol), and FeCl2 (0.135 g, 1.1 mmol). After workup, purifica-
tion by flash column chromatography (94% EtOAc, 5% EtOH,
1% Et3N) afforded bis-amine 3c (0.221 mg, 52%) as a red-
orange powder: mp 78-80 °C; Rf ) 0.13 (94% EtOAc, 5%
EtOH, 1% Et3N); IR (film) 3088w, 3045w, 1450s, 1372m,
1356m, 1284s, 1266s, 1163s, 1148s, 1013s cm-1; 1H NMR (300
MHz) δ 2.31 (s, 6H), 2.35-2.65 (stack, 24H), 3.99 (s, 8H); 13C
NMR (75 MHz) δ 26.9 (CH2), 45.8 (CH3), 52.9 (CH2), 54.9 (CH2),
59.8 (CH2), 67.8 (CH), 68.6 (CH), 86.4 (quat. C); MS (EI) m/z
438 ([M]+, 42%), 247 (7, [M - CpCH2CH2N(CH2CH2)2NCH3]+),
176 (7, [M - FeCpCH2CH2N(CH2CH2)2NCH3]+), 113 (100,
[CH2N(CH2CH2)2NCH3]+), 70 (47, [N(CH2CH2)2]+). Anal. Calcd
for C24H38FeN4: C, 65.75; H, 8.74; N, 12.78. Found: C, 65.88;
H, 8.86; N, 12.54.
1, 1′-Divinylferrocene (4) and 1-[2-Diisopropylamino]-
ethyl-1′-ethenylferrocene (5h). nBuLi (3.48 mL of a 1.68 M
solution in hexane, 5.81 mmol) was added to a solution of
iPr2NH (0.89 mL, 6.30 mmol) in THF (10 mL) at 0 °C under
N2. The solution was stirred at RT for 15 min and then added
via syringe to a solution of spiro[2.4]hepta-4,6-diene 2 (0.5 mL,
4.85 mmol) in THF (2.5 mL) at RT. The resulting orange
solution was heated under reflux overnight and then cooled
to RT. Anhydrous FeCl2 (0.307 g, 2.42 mmol) was added. The
dark brown solution was heated under reflux for a further 22
h. After cooling to RT, Et2O (100 mL) and ammonium chloride
solution (100 mL) were added. The mixture was basified to
pH 9 with NaOH solution (2 M), and the layers were separated.
The aqueous layer was extracted with Et2O (2 × 50 mL) and
then with CH2Cl2 (2 × 50 mL). The combined organic layers
were washed with brine (150 mL) and then dried (MgSO4).
1,1′-Bis[2-morpholin-4-yl]ethylferrocene (3d). Bis-ami-
no ferrocene 3d was prepared from morpholine (0.55 mL, 6.3
(36) Kiselev, V. D.; Kashaeva, E. A.; Konovalov, A. I. Tetrahedron
1999, 55, 1153-1162.

Functionalized Ferrocenophanes
Organometallics, Vol. 24, No. 3, 2005 365
CpFeCpCH2]+), 121 (15, [CpFe]+), 56 (16, [Fe]+); HPLC tR
Concentration under reduced pressure afforded a red-orange
oil, which was purified by flash column chromatography
(hexane, then 94% EtOAc, 5% EtOH, 1% Et3N) to afford, in
order of elution, first the bis-olefin 4 (0.202 g, 35%) as a red
powder: mp 38-40 °C; Rf ) 0.50 (hexane); IR (Nujol mull)
)
7.89 min; data were consistent with those reported in the
literature.37
1-[2-Pyrrolidin-1-yl]ethyl-1′-ethenylferrocene (5e). nBuLi
(0.177 mL of a 1.78 M solution in hexane, 0.32 mmol) was
added to a solution of Et2NH (35 µL, 0.34 mmol) in THF (3
mL) at 0 °C. The pink solution was stirred at RT for 15 min
and then added via syringe to a solution of 1,1′-bis[2-pyrrolidin-
1-yl]ethylferrocene 3e (100 mg, 0.26 mmol) in THF (2.5 mL)
at RT. The resulting orange solution was heated under reflux
for 48 h. After cooling to RT, workup as described for the
preparation of bis-amines 3 provided a red-orange oil, which
was purified by flash column chromatography (94% EtOAc,
5% EtOH, 1% Et3N) to afford, in order of elution, first 5e (25
mg, 31%) as a red-orange oil: Rf ) 0.28 (94% EtOAc, 5% EtOH,
1% Et3N); IR (film) 3086m, 1629s, 1461m, 1382m, 1350m,
1330m, 1292m, 1241m, 1202m, 1145m, 1118m, 1041m, 1020m
3088m, 1631s, 1464m, 1378m, 1239m, 1047m, 1028m cm-1
;
1H NMR (300 MHz) δ 4.09-4.25 (stack, 8H), 5.05 (dd, J 10.7,
1.5, 2H), 5.28 (dd, J 17.3, 1.5, 2H), 6.36 (dd, J 17.3, 10.7, 2H);
13C NMR (75 MHz) δ 67.9 (CH), 69.9 (CH), 84.2 (quat. C), 111.3
(CH2), 134.1 (CH); MS (TOF ES+) m/z 238.0 ([M + H]+, 100%);
HRMS calcd for C14H14Fe [M]+ 238.0445, found 238.0441. Anal.
Calcd for C14H14Fe: C, 70.62; H, 5.93. Found: C, 70.54; H,
6.01. Then amine 5h (55 mg, 8%) was afforded as a red oil: Rf
) 0.44 (94% EtOAc, 5% EtOH, 1% Et3N); IR (film) 3406m,
3005m, 1629s, 1463m, 1381m, 1267m, 1241m, 1205m, 1046m,
1029m cm-1; 1H NMR (300 MHz) δ 1.01 (d, J 6.6, 12H), 2.30-
2.56 (stack, 4H), 2.94-3.11 (m, 2H), 3.86-4.32 (stack, 8H), 5.03
(d, J 10.7, 1H), 5.31 (d, J 17.3, 1H), 6.41 (dd, J 17.3, 10.7, 1H);
13C NMR (100 MHz) δ 20.7 (CH3), 31.9 (CH2), 47.5 (CH2), 48.9
(CH), 67.2 (CH), 68.4 (CH), 69.2 (CH), 69.5 (CH), 83.6 (quat.
C), 87.6 (quat. C), 110.9 (CH2), 134.5 (CH); MS (EI) m/z 339
([M]+, 15%), 114 (100, [CH2N(CH(CH3)2)2]+); HRMS calcd for
C20H30FeN [M + H]+ 340.1728, found 340.1725.
cm-1 1H NMR (300 MHz) δ 1.77-1.83 (m, 4H), 2.47-2.63
;
(stack, 8H), 3.98-4.27 (stack, 8H), 5.03 (dd, J 10.7, 1.5, 1H),
5.30 (dd, J 17.6, 1.5, 1H), 6.40 (dd, J 17.6, 10.7, 1H); 13C NMR
(75 MHz) δ 26.1 (CH2), 31.3 (CH2), 56.9 (CH2), 60.3 (CH2), 69.9
(CH), 71.2 (CH), 71.9 (CH), 72.0 (CH), 86.4 (quat. C), 89.7
(quat. C), 113.8 (CH2), 137.0 (CH); MS (EI) m/z 309 ([M]+,
25%), 84 (100, [(CH2CH2)2NCH2)]+). Anal. Calcd for C18H23-
FeN: C, 69.91; H, 7.50; N, 4.53. Found: C, 69.69; H, 7.25; N,
4.32. Then starting material 3e (60 mg, 60%) eluted.
1,1′-Diethylferrocene (9) and 1-Ethyl-1′,2′-cyclobutyl-
ferrocene (10). Methyl iodide (0.49 mL, 3.21 mmol) was
added to a mixture of ferrocenes 5a, 7a, and 8a (0.200 g, ∼0.64
mmol) in THF (10 mL) and the resulting brown solution heated
under reflux for 16 h. After cooling to RT the volatiles were
removed under reduced pressure to afford a mixture of
tetraalkylammonium salts as a brown oil (0.303 g), which was
used directly in the next step without further purification. KOt-
Bu (0.58 g, 5.15 mmol) was added to a solution of the
tetraalkylammonium salts in tBuOH (5 mL), and the resulting
orange mixture was heated under reflux for 15 h. The
brownish-orange cloudy reaction mixture was then cooled to
RT, and CH2Cl2 (20 mL) and water (20 mL) were then added.
The two phases were separated, and the aqueous layer was
extracted with CH2Cl2 (2 × 15 mL). The combined organic
fractions were washed with brine (20 mL) and then dried
(MgSO4). Concentration under reduced pressure afforded three
olefin products as a dark red oil (0.162 g). These were used
directly in the next step without further purification. A
solution of the three olefin products (0.162 g) in ethanol (5 mL)
was added to 10% palladium on activated charcoal (0.036 g,
0.0034 mmol), and the resulting black slurry was stirred
vigorously under a H2 atmosphere for 7 h. The resulting black
mixture was filtered through a plug of Celite, washing
thoroughly with ethanol (40 mL). Evaporation of the solvent
under reduced pressure provided two products as an orange
oil (0.123 g). Purification by reversed-phase HPLC afforded,
in order of elution, first ferrocene 10 (0.013 g) as an orange
oil: Rf ) 0.59 (1% Et3N in cyclohexane); IR (film) 3088m,
1260s, 1217m, 1094s, 1040s, 1019s cm-1; 1H NMR (400 MHz)
δ 1.19 (t, J 7.6, 3H, CpCH2CH3), 2.37 (q, J 7.6, 2H, CpCH2-
CH3), 2.82-2.91 (m, 2H, CCH2CH2C), 3.03-3.12 (m, 2H,
CCH2CH2C), 3.74-3.78 (m, 1H, CpH), 3.93-4.03 (stack, 6H,
CpH); 13C NMR (100 MHz) δ 14.8 (CH3, CpCH2CH3), 21.5 (CH2,
CpCH2CH3), 28.5 (CH2, CCH2CH2C), 62.2 (CH, Cp), 65.9 (CH,
Cp), 69.0 (CH, Cp), 69.2 (CH, Cp), 91.5 (quat. C, CpCH2CH3),
92.3 (quat. C, CpCH2CH2); MS (EI) m/z 240 ([M]+, 100%), 91
(24, [CpCH2CH2]+), 56, (32, [Fe]+); HRMS calcd for C14H16Fe
[M]+ 240.0601, found 240.0602; HPLC tR ) 6.86 min. Then
ferrocene 9 (0.009 g) was afforded as an orange oil: Rf ) 0.59
(1% Et3N in cyclohexane); IR (film) 3089w, 1260s, 1094s, 1020s
1,1′-(1-Diethylaminomethylpropanylene)ferrocene (14a)
and 1-[2-Diethylamino]ethyl-1′-ethenylferrocene (5a). A
solution of LiNEt2 [prepared as described above from nBuLi
(0.308 mL of a 2.40 M solution in hexane, 0.74 mmol) and
HNEt2 (80 µL, 0.78 mmol)] in THF (4 mL) was added via
cannula to a solution of 1,1′-divinyl ferrocene 4 (88 mg, 0.37
mmol) in THF (5 mL) at RT. The resulting orange solution
was heated under reflux for 45 h. After cooling, workup as
described for the preparation of bis-amine 3 afforded an orange
solid, which was purified by flash column chromatography
(94% EtOAc, 5% EtOH, 1% Et3N) to afford, in order of elution,
first the ferrocenophane 14a (0.048 g, 42%) as an orange
solid: Rf ) 0.36 (94% EtOAc, 5% EtOH, 1% Et3N); λmax(CH2-
Cl2)/nm 442 (ꢀ/dm3 mol-1 cm-1 141), 320 (52); IR (film) 3088m,
3047w, 1465m, 1383m, 1266s, 1204m, 1150w, 1071m, 1041m,
1024m cm-1 1H NMR (300 MHz) δ 0.99 (t, J 7.0, 6H,
;
N(CH2CH3)2), 1.63-1.83 (m, 2H, CpCH2CH2CH), 1.97-2.11
(m, 1H, CpCH), 2.23-2.40 (m, 2H, CpCH2CH2CH), 2.41-2.65
(stack, 6H, CH2N(CH2CH3)2), 3.85-4.13 (stack, 8H, CpH); 13
C
NMR (75 MHz) δ 11.7 (CH3, N(CH2CH3)2), 24.5 (CH2,
CpCH2CH2CH), 35.5 (CH, CpCH), 40.3 (CH2, CpCH2CH2CH),
47.5 (CH2, N(CH2CH3)2), 58.4 (CH2, CpCHCH2N), 65.8 (CH,
CpH), 67.2 (CH, CpH), 67.5 (CH, CpH), 67.7 (CH, CpH), 68.6
(CH, CpH), 68.7 (CH, CpH), 70.6 (CH, CpH), 71.0 (CH, CpH),
87.3 (quat. C, Cp), 88.3 (quat. C, Cp); MS (EI) m/z 311 ([M]+,
38%), 225 (24, [M - CH2N(CH2CH3)2]+), 86 (100, [CH2N(CH2-
CH3)2]+). Anal. Calcd for C18H25FeN: C, 69.46; H, 8.10; N, 4.50.
Found: C, 69.44; H, 8.12; N, 4.21. Then amine 5a (0.032 g,
28%) was afforded as an orange oil: Rf ) 21 (94% EtOAc, 5%
EtOH, 1% Et3N); IR (film) 3086m, 1630m, 1470m, 1383m cm-1
;
1H NMR (300 MHz) δ 1.06 (t, J 7.2, 6H), 2.37-2.67 (stack,
8H), 3.92-4.06 (stack, 4H), 4.12-4.21 (stack, 2H), 4.23-4.32
(stack, 2H), 5.04 (dd, J 10.8, 1.4, 1H), 5.32 (dd, J 17.4, 1.4,
1H), 6.41 (dd, J 17.4, 10.8, 1H); 13C NMR (75 MHz) δ 11.8
(CH3), 26.5 (CH2), 46.9 (CH2), 54.3 (CH2), 67.2 (CH), 68.5 (CH),
69.2 (CH), 69.4 (CH), 83.7 (quat. C), 87.3 (quat. C), 111.1 (CH2),
134.4 (CH); MS (TOF ES+) m/z 312.1 ([M + H]+, 100%); HRMS
calcd for C18H26FeN2 [M + H]+ 312.1415, found 312.1424. Then
bis-amine 3a (0.007 g, 5%) was afforded as an orange oil.
1,1′-(1-Morpholinomethylpropanylene)ferrocene (14d)
and 1-[2-Morpholino]ethyl-1′-ethenylferrocene (5d). A
solution of lithium morpholinide [prepared as described above
1
cm-1; H NMR (400 MHz) δ 1.17 (t, J 7.6, 6H), 2.35 (q, J 7.6,
4H), 4.00 (s, 8H); 13C NMR (100 MHz) δ 14.8 (CH3), 22.1 (CH2),
67.9 (CH), 69.0 (CH), 91.0 (quat. C); MS (EI) m/z 242 ([M]+,
100%), 227 (24, [CH3CH2CpFeCpCH2]+), 212 (17, [CH2-
(37) Carroll, M. A.; White, A. J. P.; Widdowson, D. A.; Williams, D.
J. J. Chem. Soc., Perkin Trans. 1 2000, 1551-1557.

366 Organometallics, Vol. 24, No. 3, 2005
Joly et al.
from nBuLi (0.350 mL of a 1.78 M solution in hexane, 0.84
mmol) and morpholine (77 µL, 0.88 mmol)] in THF (2 mL) was
added via cannula to a solution of 1,1′-divinyl ferrocene 4 (100
mg, 0.42 mmol) in THF (2.5 mL) at RT. The resulting orange
solution was heated under reflux overnight. After cooling,
workup as described for the preparation of bis-amine 3
afforded an orange solid, which was purified by flash column
chromatography (94% EtOAc, 5% EtOH, 1% Et3N) to afford,
in order of elution, first the ferrocenophane 14d (0.017 g, 12%)
as a pale orange solid: Rf ) 0.64 (94% EtOAc, 5% EtOH, 1%
(CH2), 60.2 (CH2), 66.9 (CH2), 67.2 (CH), 68.5 (CH), 69.2 (CH),
69.3 (CH), 83.7 (quat. C), 86.9 (quat. C), 111.3 (CH2), 134.7
(CH); MS (TOF ES+) m/z 326.1 ([M + H]+, 100%); HRMS calcd
for C18H24FeNO [M + H]+ 326.1207, found 326.1212. Then bis-
amine 3d (0.119 g, 69%) was afforded as an orange solid.
Acknowledgment. We gratefully acknowledge fi-
nancial support from the Engineering and Physical
Sciences Research Council, The University of Birming-
ham, and GlaxoSmithKline (CASE award) (to K.M.J.),
and Dr. Steve J. Vickers for assistance with the elec-
trochemical characterization.
Et3N); IR (film) 3053w, 1266s, 1117m cm-1 1H NMR (300
;
MHz) δ 1.58-1.85 (stack, 2H), 2.04-2.20 (m, 1H), 2.22-2.59
(stack, 8H), 3.57-3.78 (m, 4H), 3.89-4.14 (stack, 8H); 13C
NMR (75 MHz) δ 23.9 (CH2), 33.8 (CH), 39.8 (CH2), 53.9 (CH2),
64.1 (CH2), 65.9 (CH), 66.9 (CH2), 67.3 (CH), 67.6 (CH), 67.8
(CH), 68.6 (CH), 68.8 (CH), 70.4 (CH), 70.9 (CH), 87.1 (quat.
C), 87.9 (quat. C); MS (EI) m/z 325 ([M]+, 20%), 225 (50, [M -
CH2N(CH2CH2)2O]+), 100 (100, [CH2N(CH2CH2)2O]+), 56 (37,
[Fe]+); HRMS calcd for C18H24FeNO [M + H]+ 326.1207; found
326.1205. Then amine 5d (0.014 g, 10%) was afforded as an
orange oil: Rf ) 0.50 (94% EtOAc, 5% EtOH, 1% Et3N); IR
Supporting Information Available: General experimen-
tal details, experimental procedure, and full characterization
1
data for 2 and 3b, H and 13C NMR data for all compounds,
X-ray structure for 3b, and selected UV spectra for 3d and
14a. GC-MS data for the formation of 3a and associated
byproducts, GC-MS analysis of chemical derivatization study
(converting mixture of 5a, 7a, 8a to 9 and 10), GC-MS data
for reaction of lithium amides with 4, and selected electro-
chemical data for 3e. This material is available free of charge
via the Internet at http://pubs.acs.org.
1
(film) 3085m, 1629m, 1457m, 1118s, 1037m, 1006m cm-1; H
NMR (300 MHz) δ 2.43-2.55 (stack, 8H), 3.74 (t, J 4.6, 4H),
3.94-4.05 (stack, 4H), 4.11-4.19 (stack, 2H), 4.22-4.30 (stack,
2H), 5.04 (dd, J 10.7, 1.5, 1H), 5.31 (dd, J 17.5, 1.5, 1H), 6.40
(dd, J 17.5, 10.7, 1H); 13C NMR (75 MHz) δ 26.0 (CH2), 53.6
OM049222M