Oxaferrocene cryptands as efficient molecular switches for alkali and alkaline earth metal ions

Article abstract of DOI:10.1021/om970681f

The reactions of 3,4-diphenylcyclopent-2-en-1-one, cyclopent-2-en-1-one, 3,4-dimethylcyclopent-2-en-1-one, and 2,3,4,5-tetramethylcyclopent-2-en-1-one with various trialkylsilyl triflates (SiR₃ = SiMe₃, SiEt₃, SiMe₂^tBu, SiⁱPr₃) result in the formation of the respective cyclopentadienyl silyl ethers in virtually quantitative yields. The reactions of these cyclopentadienes with n-BuLi and FeCl₂ give the respective substituted ferrocenyl trialkylsilyl ethers in 48-57% yields, which are synthetic equivalents of hydroxyferrocenes as generated by a suitable cleavage protocol: SiMe₃-protected ferrocene alcohols (CH₃CN + 10% H₂O + Na₂CO₃); SiEt₃ (CH₃CN + 10% H₂O + Na₂CO₃ + NaF); and SiMe₂^tBu protection (NBu₄F·3H₂O in CH₃CN). Carrying out the deprotection in the presence of carbon electrophiles leads to the respective ferrocenyl alkyl ethers in excellent yields. Using 1-chloro-2-tosylethane results in 1,1′-bis(2-chloroethoxy)-3,3′,4,4′-tetraphenylferrocene which was converted into 1,1′-bis-(2-iodoethoxy)-3,3′,4,4′-tetraphenylferrocene in 89% yield, to be finally reacted with diazacrown ethers (diaza-12-C-4, diaza-15-C-5, diaza-18-C-6). In this manner the respective ferrocene cryptands 23, 24, and 25 were generated as the 1 + 1 products in yields of 67-80%. The coordination of alkali and alkaline earth metal ions within the cavities of these ferrocene cryptands results in large anodic shifts of the redox potentials of the ferrocene units (23, E = +0.290 V; 23 + Ca²⁺, E = +0.670 V; 23 + Na⁺, E = +0.505 V; 24, E = +0.285 V; 24 + K⁺, E = +0.455 V) as determined by cyclic voltammetry. These anodic shifts are correlated with a decrease in the complex stability constant K of the respective metal ion complex (redox-switching effect) and in the case of 23 and Ca²⁺ amounts to a reduction of K by 3.4 × 10⁶.

Full text of DOI:10.1021/om970681f

5950
Organometallics 1997, 16, 5950-5957
Oxa fer r ocen e Cr yp ta n d s a s Efficien t Molecu la r Sw itch es
for Alk a li a n d Alk a lin e Ea r th Meta l Ion s
Herbert Plenio* and Clemens Aberle
Institut fu¨r Anorganische und Analytische Chemie, Albertstr. 21, 79104 Freiburg, Germany
Received August 4, 1997X
The reactions of 3,4-diphenylcyclopent-2-en-1-one, cyclopent-2-en-1-one, 3,4-dimethylcy-
clopent-2-en-1-one, and 2,3,4,5-tetramethylcyclopent-2-en-1-one with various trialkylsilyl
triflates (SiR3 ) SiMe3, SiEt3, SiMe2 Bu, SiiPr3) result in the formation of the respective
t
cyclopentadienyl silyl ethers in virtually quantitative yields. The reactions of these
cyclopentadienes with n-BuLi and FeCl2 give the respective substituted ferrocenyl trialkylsilyl
ethers in 48-57% yields, which are synthetic equivalents of hydroxyferrocenes as generated
by a suitable cleavage protocol: SiMe3-protected ferrocene alcohols (CH3CN + 10% H2O +
t
Na2CO3); SiEt3 (CH3CN + 10% H2O + Na2CO3 + NaF); and SiMe2 Bu protection (NBu4F‚3H2O
in CH3CN). Carrying out the deprotection in the presence of carbon electrophiles leads to
the respective ferrocenyl alkyl ethers in excellent yields. Using 1-chloro-2-tosylethane results
in 1,1′-bis(2-chloroethoxy)-3,3′,4,4′-tetraphenylferrocene which was converted into 1,1′-bis-
(2-iodoethoxy)-3,3′,4,4′-tetraphenylferrocene in 89% yield, to be finally reacted with diaza-
crown ethers (diaza-12-C-4, diaza-15-C-5, diaza-18-C-6). In this manner the respective
ferrocene cryptands 23, 24, and 25 were generated as the 1 + 1 products in yields of 67-
80%. The coordination of alkali and alkaline earth metal ions within the cavities of these
ferrocene cryptands results in large anodic shifts of the redox potentials of the ferrocene
units (23, E ) +0.290 V; 23 + Ca2+, E ) +0.670 V; 23 + Na+, E ) +0.505 V; 24, E ) +0.285
V; 24 + K+, E ) +0.455 V) as determined by cyclic voltammetry. These anodic shifts are
correlated with a decrease in the complex stability constant K of the respective metal ion
complex (redox-switching effect) and in the case of 23 and Ca2+ amounts to a reduction of K
by 3.4 × 106.
In tr od u ction
cal stimulus, which may be a light- or chemically-driven
conformational change,11 a variation of proton concen-
tration,12 or an electron transfer (redox reaction).13,14
For our work we have chosen to investigate the redox-
switched bonding of alkali or alkaline earth metal ions
within the cavities of ferrocene-crown ethers. Conse-
quently the guest is a metal ion, whose binding proper-
ties within a ferrocene cryptand are switched on or off,
depending on the redox state of the ferrocene unit.
Group I and II metal ions are ideally suited for such
switching devices, since their coordination chemistry is
characterized by fast complexation and decomplexation
reactions.
The making of molecules capable of performing cer-
tain tasks has been termed property-directed synthesis.1
Consequently the investigation of specifically designed
molecular devices constitutes a subdivision of supramo-
lecular chemistry.2,3 Examples for such devices are mo-
lecular switches, in which a characteristic property such
as conductivity along a molecular wire,4 fluorescence,5
refractive index,6 magnetic behavior,7 photochromism,8
catalytic behavior,9 or affinity toward a guest molecule
can be switched on or off.10 The minimum requirements
for the switched complexation of guests are the ability
to incorporate molecular or ionic units into a suitable
receptor framework as well as the reversible change of
the binding properties triggered by a chemical or physi-
(10) The term switch may be somewhat misleading, in that it
suggests an all or nothing situation, which is, quite in contrast to a
mechanical switch, not possible on a molecular level. According to the
laws of thermodynamics there can only be a certain increase or
decrease in the binding affinity as expressed by a change of the stability
constant K for the equilibrium under observation.
(11) Shinkai, S.; Ogawa, T.; J usano, Y.; Manabe, O.; Kikugawa, K.;
Goto, T.; Matsuda, T. J . Am. Chem. Soc. 1982, 104, 1960.
(12) Izatt, R. M.; Lamb, J . D.; Hawkins, R. T.; Brown, P. R.; Izatt,
R. S.; Christensen, J . J . J . Am Chem. Soc. 1983, 105, 1782.
(13) (a) vanVeggel, F. C. J . M.; Verboom, W.; Reinhoudt, D. N.
Chem. Rev. 1994, 94, 279. (b) Beer, P. D. Adv. Inorg. Chem. 1992, 39,
79.
(14) (a) Hall, C. D.; Leineweber, A.; Tucker, J . H. R.; Williams, D.
J . J . Organomet. Chem. 1996, 523, 13. (b) Beer, P. D.; Stokes, S. E.
Polyhedron 1995, 14, 2631. (c) Beer, P. D.; Crane, C. G.; Danks, J . P.;
Gale, P. A.; McAleer, J . F. J . Organomet. Chem. 1995, 490, 143. (c)
J ulie, L. F.; Schatz, E.; Ward, M. D.; Weber, F.; Yellowlees, L. J . J .
Chem. Soc., Chem. Commun. 1994, 799. (d) Nabeshima, T.; Furusawa,
H.; Yano, Y. Angew. Chem. 1994, 106, 1849; Angew. Chem., Int. Ed.
Engl. 1994, 33, 1750. (e) Rahman, S. M. F.; Fukuishi, K. J . Chem.
Soc., Chem. Commun. 1994, 917. (f) Plenio, H.; Diodone, R. J .
Organomet. Chem. 1995, 492, 73. (g) Blanc, S.; Yakirevitch, P.; Leize,
E.; Meyer, M.; Libman, J .; VanDorsselaer, A.; Albrecht-Gary, A.-M.;
Shanzer, A. J . Am. Chem. Soc. 1997, 119, 4934. (h) Fabrizzi, L.; Poggi,
A. Chem. Soc. Rev. 1995, 197.
X Abstract published in Advance ACS Abstracts, December 15, 1997.
(1) Gokel, G. W.; Medina, J . C.; Li, C. Synthesis 1991, 677.
(2) Lehn, J .-M. Supramolecular Chemistry; VCH; Weinheim, 1995.
(3) Comprehensive Supramolecular Chemistry; Lehn, J .-M., Atwood,
J . L., Davies, J . E. D., McNicol, D. D., Vo¨gtle, F., Eds.; Pergamon
Press: Oxford, 1996.
(4) Kawai, S. H.; Gilat, S. L.; Lehn, J .-M. J . Chem. Soc., Chem.
Commun. 1994, 1011.
(5) (a) J ames, T. D.; Linnane, P.; Shinkai, S. J . Chem. Soc., Chem.
Commun. 1996, 281. (b) DeSilva, A. P.; Gunaratne, H. Q. N.;
Gunnlaugson, T.; Nieuwenhuizen, M. J . Chem. Soc., Chem. Commun.
1996, 1967. (c) DeSilva A. P.; Gunaratne, H. Q. N.; McCoy, C. P. J .
Chem. Soc., Chem. Commun. 1996, 2399.
(6) (a) Meerholz, K. Angew. Chem. 1997, 109, 981; Angew. Chem.,
Int. Ed. Engl. 1997, 36, 945. (b) Peng, Z.; Gharavi, A. R.; Yu, L. J .
Am. Chem. Soc. 1997, 119, 4622.
(7) Gu¨tlich, P.; Hauser, A.; Spiering, H. Angew. Chem. 1994, 106,
2109; Angew. Chem., Int. Ed. Engl. 1994, 33, 2024.
(8) Pina, F.; Melo, M. J .; Maestri, M.; Ballardini, R.; Balzani, V. J .
Am. Chem. Soc. 1997, 119, 5556.
(9) Hembre, R. T.; McQueen, J . S.; Day, V. W. J . Am. Chem. Soc.
1996, 118, 798.
S0276-7333(97)00681-X CCC: $14.00 © 1997 American Chemical Society

Oxaferrocene Cryptands as Molecular Switches
Organometallics, Vol. 16, No. 26, 1997 5951
Ch a r t 1
Sch em e 1. Gen er a l Syn th esis of th e
Cyclop en ta d ien yl Silyl Eth er s a n d F er r ocen yl Silyl
Eth er sa
a
Reagents: (a) CF3SO3SiR3′′, Et3N, petroleum ether; (b)
n-BuLi, FeCl2, THF.
Sch em e 2. Cyclop en ta d ien yl Silyl Eth er s
For an optimum performance of such molecular
devices it is very important to allow an efficient elec-
tronic communication between the metal ion coordi-
nated within the macrocyclic cavity and the redox active
unit. Recently it was shown by us that the switching
process in most ferrocene cryptands relies predomi-
nantly on electrostatic interactions.15 One of the best
molecular devices based on this principle is (1,1′-
ferrocenylenedimethylene) bis(1,4,10,13-tetraoxa-7,16-
diazacyclooctadecane) (A in Chart 1),16 but it is difficult
to see how such switches primarily relying on Coulomb
type interactions could be further optimized. Conse-
quently, to develop such molecular machines to the point
at which they can be used to emulate naturally occur-
ring switching molecules such as for example calmodu-
lin,17 the previous concept based only on electrostatic
interactions has to be modified and extended. In this
vein we have recently demonstrated in an investigation
of transition metal complexes of bis(1,1′-di(2-picolyl)-
amino)-3,3′,4,4′-tetraphenylferrocene (B in Chart 1) that
the electronic communication between the two metal
centers can be improved drastically when at least one
of the donor atoms coordinated to the metal ion is
directly coupled electronically to the metallocene unit.18
Due to the often unfavorable thermodynamic and
kinetic properties (with respect to switching action!)
transition metal complexes are not well suited for redox-
switching, while unfavorable soft-hard interactions in
the complexes of the hard alkali metal ions with the soft
nitrogen donors of aminoferrocenes did not result in
significantly improved devices (C in Chart 1).19 The
obvious solution therefore was to synthesize ferrocene
cryptands in which one or two hard oxygen donor
centers are directly attached to the redox-active moiety.
In fact the first ferrocene crown ethers prepared used
this concept (D in Chart 1),20 which was abandoned later
on, probably due to the difficult availability of the
hydroxyferrocenes needed to synthesize such com-
pounds21 and because the coronand type ligands syn-
thesized then did not allow the stable bonding of alkali
metal ions required for an efficient switching device.
To overcome these problems we present here a new
and facile synthesis of a number of different silylated
hydroxyferrocenes,22 their reactions to yield oxafer-
rocene cryptands, and an investigation into the electro-
chemistry, i.e., switching behavior, of such redox-active
macrocycles.
Resu lts a n d Discu ssion
Syn th eses of Oxygen -Su bstitu ted Cyclop en ta -
d ien es. Cyclopentenones are very versatile starting
materials for the synthesis of various cyclopenta-
dienes,23,24 and we have recently communicated the
reactions of 3,4-diphenylcyclopent-2-en-1-one, cyclopent-
2-en-1-one, 3,4-dimethylcyclopent-2-en-1-one, and 2,3,4,5-
tetramethylcyclopent-2-en-1-one with a variety of tri-
alkylsilyl triflates (SiMe3, SiEt3, SitBuMe2, SiiPr3) yielding
the respective cyclopentadienyl silyl ethers 1-14 in
almost quantitative yields (Scheme 1, Scheme 2) when
applying procedures related to those by Simchen et al.25
and Reetz et al.26 for silyl enol ether formation.
(20) (a) Biernat, J . F.; Wilczewski, T. Tetrahedron 1980, 275. (b)
Akabori, S.; Habata, Y.; Sakamoto, Y.; Sato, M.; Ebine, S. Bull. Chem.
Soc. J pn. 1983, 56, 537. (c) Saji, T. Chem. Lett. 1986, 275.
(21) Herberhold, M. In Ferrocenes; Togni, A., Hayashi, T., Eds.;
VCH: Weinheim, 1995.
(22) A preliminary report of our work has been published: Plenio,
H.; Aberle, C. J . Chem. Soc., Chem. Commun. 1996, 2123.
(23) (a) Plenio, H. Organometallics 1992, 11, 1856. (b) J u¨ngling, S.;
Mu¨lhaupt, R.; Plenio, H. J . Organomet. Chem. 1993, 460, 191. (c)
Plenio, H.; Diodone, R. J . Org. Chem. 1993, 58, 6650. (d) Bunel, E. E.;
Campos, P.; Ruz, J .; Valle, L.; Chadwick, I.; Ana, M. S.; Gonzalez, G.;
Manriquez, J . M. Organometallics 1988, 7, 474. (e) Plenio, H.; Burth,
D. Organometallics 1996, 15, 1151. (f) Plenio, H.; Burth, D. J .
Organomet. Chem. 1996, 519, 269.
(15) (a) Plenio, H.; Yang, J .; Diodone, R.; Heinze, J . Inorg. Chem.
1994, 33, 4098. (b) Plenio, H.; Diodone, R. Inorg. Chem. 1995, 35, 3964.
(c) Allgeier, A. M.; Slone, C. S.; Mirkin, C. A.; Liable-Sands, L. M.;
Yap, G. P. A.; Rheingold, A. L. J . Am. Chem. Soc. 1997, 119, 550.
(16) (a) Medina, J . C.; Goodnow, T. T.; Rojas, M. T.; Atwood, J . L.;
Lynn, B. C.; Kaifer, A. E.; Gokel, G. W. J . Am. Chem. Soc. 1992, 114,
10583. (b) Plenio, H.; Burth, D. Chem. Ber. 1993, 126, 2403.
(17) Bioinorganic Chemistry; Bertini, I., Gray, H. B., Lippard, S.
J ., Valentine, J . S., Eds.; University Science Books: Sausalito, CA,
1994.
(18) (a) Plenio, H.; Burth, D. Organometallics 1996, 15, 4054. (b)
Plenio, H.; Burth, D. Angew. Chem. 1995, 107, 881; Angew. Chem.,
Int. Ed. Engl. 1995, 34, 800.
(24) (a) Plenio, H.; Warnecke, A. Organometallics 1996, 15, 5066.
(b) Plenio, H.; Warnecke, A. J . Organomet. Chem. 1997, 544, 133.
(25) Reetz, M. T.; Neumeier, G. Chem. Ber. 1979, 112, 2209.
(26) Simchen, G.; Kober, W. Synthesis 1976, 259.
(19) Plenio, H.; Burth, D. Organometallics 1996, 15, 1151.

5952 Organometallics, Vol. 16, No. 26, 1997
Plenio and Aberle
Sch em e 3. Differ en t Isom er s of th e Silyl
Cyclop en ta d ien yl Eth er s
Sch em e 5. Syn th esis of th e
1,1′-Dia lk oxyfer r ocen esa
a
Conditions: (a) cleavage reagents for OSiMe3, CH3CN/10%
Sch em e 4. 1,1′-Disiloxyfer r ocen es
H2O/Na2CO3; and OSiEt3, CH3CN/10% H2O/Na2CO3/NaF; and
SitBuMe2, Bu4NF‚3H2O/Na2CO3.
t
for more reliable results SiMe2 Bu protection is to be
preferred. Quite in contrast the synthesis of the 1,1′-
bis(trialkylsiloxy)octamethylferrocenes is unreliable. We
were able to isolate a dark red oil, which most likely
appears to contain the desired ferrocene. The reason
for these problems appears to be that small protective
groups are unstable with respect to cleavage reactions,
while the anions of the extremely bulky tetramethyl-
(trialkylsiloxy)cyclopentadienes are rather oxidatively
coupled upon attempted ferrocene synthesis. The redox
potentials of the ferrocenes reflect the electronic effects
of the different substituents, and it is obvious that siloxy
groups are strongly electron-donating substituents, with
each being responsible for a cathodic shift of about 100
The cyclopentadienes 1-3 and 8-14 derived from 3,4-
diphenylcyclopent-2-en-1-one, 3,4-dimethylcyclopent-2-
en-1-one, and 2,3,4,5-tetramethylcyclopent-2-en-1-one,
respectively, are stable compounds. However, the cy-
clopentadienes synthesized from the unsubstituted cy-
clopent-2-en-1-one are prone to Diels-Alder dimeriza-
tion and must be stored at temperatures below -20 °C
but can be handled without special precautions during
workup at room temperature. All yields were excellent,
as long as an excess of acid (trialkylsilyl triflate) with
respect to the added base (Et3N) was avoided, since this
invariably leads to the decomposition of the products.
The characterization of the cyclopentadienyl silyl
ethers by NMR spectroscopy is not straightforward since
they exist as a mixture of double-bond isomers (Scheme
3).27 The product formed under kinetic control is the
1,4-diene, which rearranges within a time interval
ranging from a few minutes to several days to give the
thermodynamically favored 1,3-diene. Both isomers
have the siloxy groups in the vinyl position, while
another possible isomer with the siloxy substituents in
the allyl position was never observed in detectable
concentrations.
Syn th esis of F er r ocen yl Silyl Eth er s a n d F er r o-
cen yl Alk yl Eth er s. We have synthesized a number
of cyclopentadienylsilyl ethers with different protective
groups to evaluate which of them is best suited for the
synthesis of the respective ferrocenes (Scheme 1, Scheme
4). Ideally the protective group is just stable enough
to survive the reaction conditions during the formation
of the ferrocenyl silyl ethers and can be easily cleaved
off afterward under basic conditions, to allow the in situ
functionalization to ferrocenyl alkyl ethers. The stabil-
ity of the protective group primarily depends on the
steric bulk of the trialkylsilyl unit28 and to a minor
extent also on the additional substituents attached to
the cyclopentadiene. Accordingly we found that SiMe3-
protected 1 is ideally suited for the synthesis of the
corresponding tetraphenylferrocene 15, when using a
standard procedure in which the cyclopentadienyl silyl
ethers are deprotonated with n-BuLi and the respective
anions reacted with FeCl2. For the more electron-rich
metallocenes at least SiEt3 protection is required and
the synthesis of the corresponding ferrocenes is possible
under carefully controlled reaction conditions. However,
mV: 15 E1/2 ) +0.28 V; 17 E1/2 ) +0.18 V; 19 E1/2
)
-0.03 V.29
The ferrocenyl silyl ethers described here are masked
alcohols forming stable compounds, quite unlike the
closely related hydroxyferrocenes.21 For the synthesis
of the desired ferrocenyl alkyl ethers, it is most conve-
nient to cleave the silyl ether and in situ react the
formed alcohol with carbon electrophiles to generate the
ferrocenyl alkyl ethers (Scheme 5). The cleavage of the
alcohol as well as formation of the alkyl ether ideally
proceeds in one step under basic conditions, and we have
performed a number of experiments with SiMe3-, SiEt3-,
t
and SiMe2 Bu-protected alcohols in the presence of
different cleavage reagents and benzyl bromide to
determine the optimum conditions. The SiMe3-pro-
tected ferrocene 15 is best reacted in acetonitrile
containing 10% of water and excess Na2CO3. Perform-
ing this cleavage in the presence of benzyl bromide leads
directly to the respective benzyl ether in more than 85%
yield. This procedure does not work very well in the
t
case of the SiEt3- and SiMe2 Bu-protected ferrocenes and
it is much more convenient to cleave the latter two silyl
ethers in a different manner: SiEt3 protection (excess
t
NaF + Na2CO3 in CH3CN with 10% water) and SiMe2 Bu
protection (n-Bu4N+F-‚3H2O in CH3CN). The fluoride
reagent is highly reactive and quantitatively cleaves the
silyl ethers within a few minutes at room temperature.
Carrying out both cleavage reactions in the presence of
benzyl bromide leads after several hours under reflux
to the respective ferrocenyl benzyl ethers.
Syn th esis of th e F er r ocen e Cr yp ta n d s. After
synthesizing a number of ferrocenyl silyl ethers as well
as working out conditions for their cleavage and trans-
formation into the respective alkyl ethers, we had to
decide which ferrocene was best suited for the synthesis
of ferrocene cryptands. While an obvious choice might
have been the unsubstituted ferrocene 17, we prefer the
tetraphenylferrocene 15 instead, because it is much
(27) J utzi, P. Chem. Rev. 1986, 86, 983.
(28) Greene, T. W.; Wuts, P. G. M. Protective Groups in Organic
Chemistry; Wiley: New York, 1991.
(29) Lu, S.; Strelets, V. V.; Ryan, M. F.; Pietro, M. F.; Lever, A. P.
B. Inorg. Chem. 1996, 35, 1013.

Oxaferrocene Cryptands as Molecular Switches
Organometallics, Vol. 16, No. 26, 1997 5953
Sch em e 6. Syn th esis of 1,1′-Bis(2-iod oeth oxy)-
3,3′,4,4′-tetr a p h en ylfer r ocen ea
Sch em e 8. Histogr a m of th e Sh ifts of th e Red ox
P oten tia ls (∆E1/2) As In d u ced by th e Coor d in a tion
of th e Alk a li a n d Alk a lin e Ea r th Meta l Ion s w ith in
th e Ca vities of th e Oxa fer r ocen e Cr yp ta n d s 23, 24,
a n d 25
a
Reagents and conditions: (a) CH3CN + 10% H2O, Na2CO3;
ClC2H4(tosylate), reflux; (b) NaI, acetone reflux.
Sch em e 7. Syn th esis of th e Oxa fer r ocen e
Cr yp ta n d s
Cyclic Volta m m etr ic In vestiga tion s of th e F er -
r ocen e Cr yp ta n d s a n d th eir Meta l Com p lexes. In
the metal complexes of the oxaferrocene cryptands 23,
24, and 25 two oxygen donors are shared between the
redox-active metallocene and the group I or II metal ion
bonded within the macrocyclic cavity. These two oxygen
atoms are vital for the electronic communication be-
tween the two metal centers for two reasons: (I) A
chelating coordination of a metal ion by the two oxygen
centers bonded directly to the cyclopentadienyl rings
will result in a short iron to metal ion distance and
consequently in an increase of the electrostatic interac-
tion between the two metal centers. (II) The coordina-
tion of a metal ion by the same two oxygen atoms will
allow a through bond interaction between the two metal
centers via the cyclopentadienyl units. More evidence
for the participation of the two oxygen atoms attached
to the five-membered ring in the complexation of the
metal cations could have been obtained from crystal
structure analysis; however, all crystals investigated
(23‚NaClO4, 23‚NaCF3SO3, 24‚KCF3SO3) either proved
to be high-order twins or poorly diffracting species.
We expected that the aforementioned two effects
would give rise to a highly efficient switching device, to
be quantified by measuring the redox potentials of the
oxaferrocene cryptands (E1) as well as the anodically
shifted redox potentials (E2) of the respective complexes
with group I or II metal ions. The difference E2 - E1 )
∆E1/2 is correlated with the difference in the stability
constants for the metal ion complexes of the neutral and
the oxidized oxaferrocene cryptand.32
more convenient for a number of reasons: (I) 3,4-
diphenylcyclopentenone can be synthesized in two steps
from easily available starting materials in large amounts
(>100 g).18a,30 (II) The steric bulk of the two phenyl
groups hinders the Diels-Alder dimerization of the
respective cyclopentadienylsilyl ethers. (III) The electron-
withdrawing effect of the phenyl groups renders the
resulting ferrocenes less sensitive toward oxidation,
which facilitates chromatographic workup. (IV) The
SiMe3 protective group can be cleaved off very easily.
The tetraphenylferrocene 15 was cleaved according
to Scheme 6 and the resulting dihydroxyferrocene in situ
reacted with a 5-fold excess of 1-tosyl-2-chloroethane.
This reaction yields after chromatographic purification
1,1′-bis(2-chloroethoxy)-3,3′,4,4′-tetraphenylferrocene (21)
in 80% yield together with a small amount (<5%) of the
undesired ferrocenophane, resulting from an intramo-
lecular ring closure of the second hydroxy group with
the OCH2CH2Cl unit. Iodide can be replaced more
easily than chloride, and a Finkelstein reaction was
employed to generate the iodide 22 from the chloride
21 in 89% yield.
Upon analyzing the CV data obtained for the oxafer-
rocene cryptands 23, 24, and 25 and their group I or II
metal ion complexes, two overlapping trends are visible
(Scheme 8, Table 1): (I) The larger the cavity, i.e., the
more donor atoms present (23 f 24 f 25), the smaller
the anodic shift ∆E1/2 of the redox potentials. (II) Those
metal ions which are complementary to the size of the
cavity of the ferrocene cryptand produce the largest
anodic shift ∆E1/2 of the redox potentials upon complex-
ation.
This iodide was reacted with three different macro-
cycles (diaza-12-crown-4, diaza-15-crown-5, diaza-18-
crown-6) in the presence of Na2CO3 or K2CO3 to yield
the desired ferrocene cryptands in excellent yields: 23
(yield 80%), 24 (yield 67%), and 25 (yield 68%) (Scheme
7). Such extremely high yields are quite unusual in the
formation of macrocycles, especially since no high dilu-
tion conditions were applied.31
Both trends can be identified, when the ∆E1/2 values
of the different alkaline earth metal ion complexes of
(30) (a) J app, F. R.; Knox, J . J . Chem. Soc. 1905, 87, 679. (b)
Geissman, T. A.; Koelsch, C. F. J . Org. Chem. 1939, 3, 498.
(31) Aza-Crown Macrocycles (published in the series: The Chemistry
of Macrocyclic Compounds); Bradshaw, J . S., Krakowiak, K. E., Izatt,
R. M., Eds.; Wiley: New York, 1993.
(32) Miller, S. R.; Gustowski, D. A.; Chen, Z.; Gokel, G. W.;
Echegoyen, L.; Kaifer, A. E. Anal. Chem. 1988, 60, 2021.

5954 Organometallics, Vol. 16, No. 26, 1997
Plenio and Aberle
Ta ble 1. Su m m a r y of th e Red ox P oten tia ls of th e F er r ocen e Cr yp ta n d s 23, 24, a n d 25 a n d Th eir Differ en t
Alk a lin e a n d Alk a lin e Ea r th Meta l Ion Com p lexes. All CV’s Wer e Recor d ed in Aceton itr ile a t Am bien t
Tem p er a tu r es Usin g Gr ou p I a n d II Tr ifla tes or P er ch lor a tes
metal ion
Li+
Na+
K+
Rb+
Cs+
Ca2+
Sr2+
Ba2+
Ferrocene 23
E1/2 [V]
∆E1/2 [mV]
+0.290
+0.285
+0.285
+0.450
+160
+0.505
+215
+0.390
+0.390
+0.385
+95
+0.670
+380
+0.650
+360
+0.595
+305
+100
+100
Ferrocene 24
+0.455
+170
E1/2 [V]
∆E1/2 [mV]
+0.385
+100
+0.465
+180
+0.430
+145
+0.375
+90
+0.605
+320
+0.565
+280
+0.585
+300
Ferrocene 25
+0.410
+125
E1/2 [V]
∆E1/2 [mV]
+0.365
+80
+0.365
+80
+0.435
+150
+0.400
+115
+0.445
+160
+0.480
+195
+0.550
+265
23, 24, and 25 are compared. With 23 for Ca2+, Sr2+
,
ferrocenyl silyl ethers which are used as stable synthetic
equivalents of the otherwise unstable hydroxyferrocenes
and on the fact that the new oxaferrocene cryptands are
available in excellent yields from fairly simple starting
materials. Consequently, we have described the syn-
thesis of highly efficient molecular switches based on
ferrocenyl alkyl ether cryptands, in which two oxygen
atoms act as an electronic relais between the redox-
active ferrocene unit and a group I or II metal ion
coordinated by the macrocyclic unit. The performance
of these devices by far exceeds that of other known redox
switches and is especially high for the biologically
relevant metal ions calcium, sodium, and potassium
with switching effects of up to 3.5 × 106 for calcium
complexation. It can be finally concluded that the
development of molecular switching devices based on
ferrocenes has now reached a point at which one can
seriously envisage applications in terms of imitating
naturally occurring switching and regulatory device.33
The systems presented here appear well suited for this
purpose. In this context it is quite significant that
calcium is the element responsible for controls and
triggers in organisms.34
and Ba2+ the largest ∆E1/2 of +380, +360, and +305 mV
within the whole series was observed. But it is also
obvious that the cavity of 23 is best suited for Ca2+, since
in the complexes of this metal with 23, 24, and 25 ∆E1/2
drops very significantly from +380 f +320 f +160 mV.
On the other hand the drop of ∆E1/2 in the same series
of ligands is much less pronounced for the Ba2+ com-
plexes, i.e., +305 f +300 f +265 mV.
The second trend is more obvious upon analyzing the
∆E1/2 values of the alkali metal complexes with 23, 24,
and 25. For 23 the Na+ complex displays the largest
∆E1/2 value by a wide margin; while 24/Na+ still shows
the strongest anodic shift of all alkali metal ions, but
now by a much smaller margin. When 23 and 24 are
compared, the Na+ value decreases from +215 f +180
mV, whereas the K+ ∆E1/2 value has increased from
+100 f +170 mV; this trend continues with 25 and here
the +80 mV of Na+ is substantially smaller than the
+125 mV determined for K+, while the largest shift is
observed with the larger Rb+ +150 mV. It is quite
remarkable that in 24 even the comparatively soft Cs+
shows a larger ∆E1/2 than that of Na+. The ∆E1/2 values
were calculated to correspond to the following reductions
in the stability constants: 23/Na+ ) 5 × 103, 23/Ca2+
) 3.5 × 106, and 24/K+ ) 8.4 × 102.
Exp er im en ta l Section
An explanation for the trends described above can be
easily given. The general drop of ∆E1/2 upon increasing
the size of the ligand cavity is caused by the decreased
relative importance of the two oxygen atoms directly
bonded to the cyclopentadienyl ring for metal coordina-
tion, as evidenced by the increasing number of donor
atoms in the diaza crown ether substructure. This also
causes an increase in the iron-metal ion distance and
hence a reduced electrostatic interaction. The two
oxygen atoms directly bonded to the ferrocene will be
most important when the metal ion is complementary
to the cavity of the respective ferrocene cryptand.
It should finally be noted that the anodic shifts of the
redox potentials observed in the ferrocene cryptands 23,
24, and 25 are the largest ever observed in ferrocene
crown ethers and this is also the first time that sub-
stantial effects have been observed for the larger alkali
metal ions K+, Rb+, and Cs+. The switching effects are
extremely strong for the alkaline earth metal ions, and
the Ca2+ value of ∆E1/2 ) +380 mV is 110 mV more
positive than the previous best of ∆E1/2 ) +270 mV
observed by Gokel et al.16 in a ferrocene cryptand, which
appears to rely mainly on electrostatic interactions.
Con clu sion s. The work described here is based on
the development of a new and facile synthesis of
Commercially available solvents and reagents were purified
according to literature procedures. Chromatography was
carried out on silica MN60. NMR spectra were recorded at
300 K on a Bruker Avance (1H NMR 200 MHz, 13C NMR 50.3
MHz) instrument. 1H NMR were referenced to residual 1H
impurities in the solvent and 13C NMR to the solvent signals:
CDCl3 (7.26 ppm, 77.0 ppm), C6D6 (7.16, 128.0 ppm), CD3CN
(1.93 ppm, 1.30 ppm). For the purpose of 1H NMR signal
assignment CpH denotes a proton attached to the sp2 carbon
of cyclopentadiene or to the carbon of a ferrocene η5-cyclopen-
tadienyl ring. Mass spectra: Finnigan MAT 3800. IR spec-
tra: Bruker IFS-25: solid materials as KBr tablets. Elemental
analyses: Mikroanalytisches Laboratorium der Chemischen
Laboratorien, Universita¨t Freiburg. Cyclic Voltammetry: The
standard electrochemical instrumentation consisted of an Amel
System 5000 potentiostat/galvanostat. All cyclic voltammo-
gramms were recorded in dry CH3CN under an argon atmo-
sphere at ambient temperature and processed using Amel
software on a PC. A three-electrode configuration was em-
ployed. The working electrode was a Pt disk (diameter 1 mm)
sealed in soft glass with a Pt wire counter electrode. The
pseudoreference electrode was an Ag wire. Potentials were
calibrated internally against the formal potential of cobalti-
(33) Plenio, H.; Aberle, C. Submitted for publication.
(34) Frausto da Silva, J . J . R.; Williams, R. J . P. The Biological
Chemistry of Elements; Oxford University Press: Oxford, 1991.

Oxaferrocene Cryptands as Molecular Switches
Organometallics, Vol. 16, No. 26, 1997 5955
cinium perchlorate (-0.94 V vs Ag/AgCl) or octamethylferro-
cene (-0.025 V vs Ag/AgCl). Solutions were ca. 2 × 10-4 mol
dm-3 in compound. NBu4PF6 (0.1 M) was used as a supporting
electrolyte. In all complexes the separation of anodic and
cathodic peak potentials is smaller than 100 mV (scan speed
100 mV/s). The procedures given for 3,4-diphenyl-3-hydroxy-
cyclopent-4-enone and 3,4-diphenylcyclopent-2-enone are modi-
fied from the publications by J app/Knox and Geissman/
Koelsch.18a,30 Starting materials were available commercially
or prepared according to literature procedures: diaza-12-
crown-4,35 3,4-dimethylcyclopent-2-en-1-one.36
1-(Tr iisop r op ylsiloxy)cyclop en ta d ien e (7). Yield: cy-
clopent-2-en-1-one (430 mg, 5.00 mmol), Et3N (0.84 mL, 6.00
mmol), CF3SO3SiiPr3 (1.30 mL, 5.00 mmol). Yield: 1.13 g
(95%) yellow oil. 1H NMR (CDCl3): δ 1.11 (m, 21H, SiiPr),
2.91 (m, 2H, CH2), 5.25 (m, 1H, CpH), 6.27-6.38 (m, 2H, CpH).
Syn th eses of 1-(Tr ia lk ylsiloxy)-3,4-d im eth ylcyclop en -
ta d ien es 8-11 a n d 1-(Tr ia lk ylsiloxy)-2,3,4,5-tetr a m eth -
ylcyclop en ta d ien es 12-14. A solution of 3,4-dimethylcy-
clopent-2-en-1-one or 2,3,4,5-tetramethylcyclopent-2-en-1-one
and NEt3 in 30/50 petroleum ether (50 mL) was cooled to 0 °C
and CF3SO3SiR3 (R3 ) Me3, Et3, tBuMe2, or iPr3) added
dropwise. The reaction mixture was allowed to warm to room
Syn th esis of 1-(Tr ia lk ylsiloxy)-3,4-d ip h en ylcyclop en -
ta d ien es 1, 2, a n d 3. 3,4-Diphenylcyclopent-2-enone was
added to 30/50 petroleum ether and treated with Et3N.
-
temperature and the Et3NH+CF3SO3 precipitate filtered off
after 15 min. The solvents were removed from the filtrate and
the residue dried under vacuo for 30 min resulting in a yellow
oil.
i
CF3SO3SiR3 (R ) Me, Et, Pr) was added dropwise to the ice-
cooled, stirred suspension. The reaction mixture was allowed
to warm to room temperature and stirring continued for 12 h.
Precipitated Et3NH+CF3SO3- was filtered off and the solvent
removed from the filtrate in vacuo, whereupon a red-brown
oil remains behind.
1-(Tr im eth ylsiloxy)-3,4-d im eth ylcyclop en ta d ien e (8).
Scale: 3,4-dimethylcyclopent-2-en-1-one (570 mg, 5.00 mmol),
Et3N (0.84 mL, 6.00 mmol), CF3SO3SiMe3 (0.93 mL, 5.00
mmol). Yield: 856 mg (94%) yellow oil. 1H NMR (CDCl3): δ
0.05-0.26 (m, 9H, SiCH3), 1.03-1.37 (m, CpCH3), 1.75-2.30
(m, CpCH3), 2.57-2.80 (m, CH2), 4.41-4.57 (m, CpH), 5.04-
5.23 (m, CpH), 5.74-5.96 (m, CpH).
1-(Tr iet h ylsiloxy)-3,4-d im et h ylcyclop en t a d ien e (9).
Scale: 3,4-dimethylcyclopent-2-en-1-one (570 mg, 5.00 mmol),
Et3N (0.84 mL, 6.00 mmol), CF3SO3SiEt3 (1.05 mL, 5.00 mmol).
Yield: 1.08 g (96%) yellow oil. 1H NMR (CDCl3): δ 0.69 (q, J
) 15.4 Hz, 6H, SiCH2), 0.84-1.21 (m, CH3), 1.89 (m, CpCH3),
2.17 (m, CpCH3), 4.41-4.76 (m, CpH), 5.05-5.32 (m, CpH),
5.76-5.96 (m, CpH).
1-(Tr im eth ylsiloxy)-3,4-d ip h en ylcyclop en ta d ien e (1).
Scale: 3,4-diphenylcyclopent-2-en-1-one (1.00 g, 4.3 mmol),
Et3N (1.20 mL, 5.20 mmol), CF3SO3SiMe3 (0.80 mL, 4.30
mmol). Yield: 1.3 g (97%) red-brown oil. 1H NMR (CDCl3):
δ 0.27 (s, 9H, SiCH3), 3.43 (s, 2H, CH2), 5.56 (s, 1H, CpH),
7.14-7.30 (m, 10H, ArH). 13C NMR (CDCl3): δ -0.24, 54.69,
113.21, 125.83, 126.31, 126.87, 127.70, 128.19, 128.30, 128.61,
139.30, 150.52, 154.78.
1-(Tr iet h ylsiloxy)-3,4-d ip h en ylcyclop en t a d ien e (2).
Scale: 3,4-diphenylcyclopent-2-en-1-one (1.00 g, 4.3 mmol),
Et3N (1.20 mL, 5.20 mmol), CF3SO3SiEt3 (0.90 mL, 4.30 mmol).
Yield: 1.43 g (96%) red-brown oil. 1H NMR (CDCl3): δ 0.78
(q, J ) 15.2 Hz, 6H, CH2CH3), 1.05 (t, J ) 15.1 Hz, 9H,
CH2CH3), 3.45 (s, 2H, CH2), 5.58 (s, 1H, CpH), 7.11-7.39 (m,
10H, ArH).
1-(ter t-Bu tyld im eth ylsiloxy)-3,4-d im eth ylcyclop en ta -
d ien e (10). Scale: 3,4-dimethylcyclopent-2-en-1-one (570 mg,
t
5.00 mmol), Et3N (0.84 mL, 6.00 mmol), CF3SO3SiMe2 Bu (1.15
mL, 5.00 mmol). Yield: 1.09 g (97%) yellow oil. 1H NMR
(CDCl3): δ 0.15 (s, 6H, SiCH3), 0.92 (s, 9H, SitBu), 1.02-1.42
(m, CpCH3), 1.88-2.07 (CpCH3), 4.39-4.57 (m, CpH), 5.15-
5.25 (m, CpH), 5.72-5.92 (m, CpH).
1-(Tr iisopr opylsiloxy)-3,4-diph en ylcyclopen tadien e (3).
Scale: 3,4-diphenylcyclopent-2-en-1-one (1.00 g, 4.30 mmol),
Et3N (1.20 mL, 5.20 mmol), CF3SO3SiiPr3 (1.10 mL, 4.30
mmol). Yield: 1.63 g (97%) red-brown oil. 1H NMR (CDCl3):
1-(Tr iisop r op ylsiloxy)-3,4-d im et h ylcyclop en t a d ien e
(11). Scale: 3,4-dimethylcyclopent-2-en-1-one (570 mg, 5.00
mmol), Et3N (0.84 mL, 6.00 mmol), CF3SO3SiiPr3 (1.30 mL,
5.00 mmol). Yield: 1.30 g (98%) yellow oil. 1H NMR (CDCl3):
δ 1.03-1.28 (m, CH3), 1.89 (br, CpCH3), 2.08 (m, CpCH3),
4.40-4.56 (m, CpH), 5.05-5.25 (m, CpH), 5.70-5.82 (m, CpH).
1-(Tr im et h ylsiloxy)-2,3,4,5-t et r a m et h ylcyclop en t a d i-
en e (12). Scale: 2,3,4,5-tetramethylcyclopent-2-en-1-one (690
mg, 5.00 mmol), Et3N (0.84 mL, 6.00 mmol), CF3SO3SiMe3
(0.93 mL, 5.00 mmol). Yield: 1.00 g (94%) yellow oil. 1H NMR
(CDCl3): δ 0.05 (s, 9H, SiCH3), 0.93-0.99 (m, CpCH3), 1.48-
1.79 (m, CpCH3), 4.45-4.50 (m, 1H, CpH).
1-(Tr ie t h ylsiloxy)-2,3,4,5-t e t r a m e t h ylcyclop e n t a d i-
en e (13). Scale: 2,3,4,5-tetramethylcyclopent-2-en-1-one (690
mg, 5.00 mmol), Et3N (0.84 mL, 6.00 mmol), CF3SO3SiEt3 (1.05
mL, 5.00 mmol). Yield: 1.20 g (95%) yellow oil. 1H NMR
(CDCl3): δ 0.68 (q, J ) 8.0 Hz, 6H, SiCH2), 0.86-1.16 (m, CH3),
1.57-1.94 (m, CpCH3), 4.44-4.51 (m, 1H, CpH).
1-(Tr iisop r op ylsiloxy)-2,3,4,5-tetr a m eth ylcyclop en ta -
d ien e (14). Scale: 2,3,4,5-tetramethylcyclopent-2-en-1-one
(690 mg, 5.00 mmol), Et3N (0.84 mL, 6.00 mmol), CF3SO3SiiPr3
(1.30 mL, 5.00 mmol). Yield: 1.44 g (96%) yellow oil. 1H NMR
(CDCl3): δ 1.00-1.18 (m, CH3), 1.51-1.96 (m, CpCH3), 4.44-
4.53 (m, CpH).
i
δ 1.05-1.13 (m, 21H, Pr), 3.43 (s, 2H, CH2), 4.55 (d, J ) 2.6
Hz, 1H, CpH), 5.34 (s, 1H, CpH), 6.70 (d, J ) 2.7 Hz, 1H,
ArCH), 7.08-7.34 (m, 10H, ArH). 13C NMR (CDCl3): δ 12.48
(SiCH(CH3)2), 17.90 (CH3), 54.62, 113.09, 125.47, 125.82,
126.25, 126.79, 127.43, 127.75, 128.27, 128.56, 128.62, 134.87,
139.48, 150.21, 155.31.
Syn th eses of 1-(Tr ia lk ylsiloxy)cyclop en ta d ien es 4, 5,
6, a n d 7. A solution of cyclopent-2-en-1-one and NEt3 in 30/
50 petroleum ether (50 mL) was cooled to 0 °C and CF3SO3SiR3
(R3 ) Me3, Et3 or iPr3) added dropwise. The reaction mix-
ture was allowed to warm to room temperature and the
-
Et3NH+CF3SO3 precipitate filtered off after 15 min. The
solvents were carefully evaporated from the filtrate in vacuo
without warming, and the products were stored at -20 °C to
retard Diels-Alder dimerizations. It is very important that
there is no excess of acid present.
1-(Tr im eth ylsiloxy)cyclop en ta d ien e (4). Scale: cyclo-
pent-2-en-1-one (430 mg, 5.00 mmol), Et3N (0.84 mL, 6.00
mmol), CF3SO3SiMe3 (0.93 mL, 5.00 mmol). Yield: 733 mg
(95%) yellow oil. 1H NMR (CDCl3): δ 0.24 (s, 9H, Si(CH3)3),
2.87-2.91 (m, 2H, CH2), 5.23 (m, 1H, CpH), 6.19-6.34 (m, 2H,
CpH).
1-(Tr ieth ylsiloxy)cyclop en ta d ien e (5). Scale: cyclopent-
2-en-1-one (430 mg, 5.00 mmol), Et3N (0.84 mL, 6.00 mmol),
CF3SO3SiEt3 (1.05 mL, 5.00 mmol). Yield: 953 mg (97%)
yellow oil. 1H NMR (CDCl3): δ 0.72 (q, J ) 7.8 Hz, 6H,
SiCH2CH3), 0.99 (t, J ) 7.7 Hz, 9H, CH2CH3), 2.91 (d, J ) 1.5
Hz, 2H, CH2), 5.24 (t, J ) 1.8 Hz, 1H, CpH), 6.24-6.35 (m,
2H, CpH).
Syn th eses of 1,1′-Bis(tr ia lk ylsiloxy)-3,4-d ip h en ylfer -
r ocen es 15 a n d 16. A solution of the respective 1-(trialkyl-
siloxy)cyclopentadiene 1 or 3 in THF (50 mL) was cooled to
-40 °C and treated with n-BuLi (2.5 M in hexane), resulting
in the green fluorescent lithium cyclopentadienide. After
stirring for 15 min, the solution was warmed to 0 °C, FeCl2
was added, and the reaction mixture was slowly allowed to
warm to room temperature and stirred until the green
fluorescence had disappeared. The solvents were evaporated,
the oily residue was extracted with cyclohexane, and the
(35) Anelli, P. L.; Montanari, F.; Quici, S.; Ciani, G.; Sironi, A. J .
Org. Chem. 1988, 53, 5292.
(36) (a) Conia, J . M.; Leriverend, M. L. Bull. Chim. Soc. Fr. 1978,
2981. (b) Plenio, H.; Diodone, R. J . Org. Chem. 1993, 58, 6650.

5956 Organometallics, Vol. 16, No. 26, 1997
Plenio and Aberle
solution was filtered over a silica plug. Finally the cyclohexane
was evaporated and the residue dried in vacuo.
1,1′-Bis(t r im e t h ylsiloxy)-3,3′,4,4′-t e t r a p h e n ylfe r r o-
cen e (15). Scale: 1-(trimethylsiloxy)-3,4-diphenylcyclopenta-
diene (2.00 g, 6.51 mmol), n-BuLi (2.60 mL, 6.51 mmol, 2.5 M
in hexane), FeCl2 (0.40 g, 3.26 mmol). Yield: 1.23 g (57%)
4.37 (s, 4H, CpH), 7.02-7.24 (m, 20H, ArH). 13C NMR
(CDCl3): δ 41.80, 60.95, 70.08, 81.20, 124.77, 126.16, 127.60,
129.74, 136.26. 3,3′,4,4′-Tetraphenyldioxa[4]ferrocenophane.
MS, m/z (%): 548 (36) [M+]. 1H NMR (CDCl3): δ 4.35 (s, 4H,
CpH), 4.84 (s, 4H, CpOCH2), 6.90-7.08 (m, 20H, ArH).
1,1′-Bis(2-iodoeth oxy)-3,3′,4,4′-tetr aph en ylfer r ocen e (22).
1,1′-Bis(2-chloroethoxy)-3,3′,4,4′-tetraphenylferrocene (2.70 g,
4.20 mmol) and NaI (6.22 g, 42.00 mmol) were dissolved in
acetone (250 mL) and heated under reflux for 6 d. The cold
solution was filtered, and the volatiles were evaporated. The
residue was taken up in CH2Cl2 and washed with NaS2O3 (10%
aqueous). The organic layer was separated, dried over MgSO4,
and filtered and the solvent evaporated. The crude product
was purified by chromatography (cyclohexane-ethyl acetate
) 10/1). Yield: 3.10 g (89%), mp 158 °C, orange-colored solid.
Anal. Calcd for C38H32FeI2O2 (830.3): C, 55.14; H, 4.03.
Found: C, 54.97; H, 3.88. 1H NMR (CDCl3): δ 2.81 (t, J )
6.6 Hz, 4H, CH2I), 3.54 (t, J ) 6.8 Hz, 4H, OCH2), 4.30 (s, 4H,
CpH), 7.05-7.40 (m, 20H, ArH). 13C NMR (CDCl3): δ 1.00
(CH2I), 61.07 (CpH), 70.74 (OCH2), 81.15, 124.51, 126.18,
127.62, 129.74, 136.30.
1,1′-[(1,7-Dioxa -4,10-d ia za cyclod od e ca n e -4,10-d iyl)-
3,3′,4,4′-tetr a p h en ylfer r ocen e (23). A mixture of 1,1′-bis(2-
iodoethoxy)-3,3′,4,4′-tetraphenylferrocene (250 mg, 0.30 mmol),
diaza-12-crown-4 (52 mg, 0.30 mmol), and Na2CO3 (160 mg,
1.50 mmol) in CH3CN (50 mL) was heated under reflux for 5
d. The cold reaction mixture was filtered, the volatiles were
evaporated in vacuo, and the residue was dissolved in a
mixture of CH2Cl2 (20 mL) and water (10 mL). The organic
layer was separated and washed with water twice (10 mL),
dried over MgSO4, filtered, and evaporated. The residue was
purified by chromatography (ethyl acetate-diethylamine: 20/
1). Yield: 180 mg (80%), mp 215 °C; orange-colored powder.
Anal. Calcd for C46H48FeN2O4 (748.7): C, 73.79; H, 6.46; N,
3.74. Found: C, 73.88; H, 6.42; N, 3.70. MS, m/z (%): 748
(55) [M+]. 1H NMR (CD3CN): δ 2.64-2.88 (m, 8H, NCH2),
2.94 (t, J ) 6.8 Hz, 4H, CpOCH2CH2N), 3.53-3.61 (m, 8H,
CH2O), 4.28 (t, J ) 6.8 Hz, 4H, CpOCH2), 4.43 (s, 4H, CpH),
6.92-7.08 (m, 20H, ArH). 13C NMR (CD3CN): δ 55.74, 57.00,
62.38, 70.69, 72.55, 82.71, 126.86, 127.00, 128.12, 128.42,
130.81, 137.65.
orange-colored powder. Anal. Calcd for
C40H42O2Si2Fe
(666.8): C, 72.73; H; 6.48. Found: C, 72.05; H, 6.35. 1H NMR
(CDCl3): δ 0.20 (s, 18H, SiCH3), 4.16 (s, 4H, CpH), 7.00-7.26
(m, 20H, ArH). 13C NMR (CDCl3): δ 0.30, 65.76, 81.21, 120.42,
125.56, 127.23, 130.01, 136.79.
1,1′-Bis(tr iisop r op ylsiloxy)-3,3′,4,4′-tetr a p h en ylfer r o-
cen e (16). Scale: 1-(triisopropylsiloxy)-3,4-diphenylcyclopen-
tadiene (2.00 g, 5.12 mmol), n-BuLi (2.60 mL, 5.12 mmol, 2.5
M in hexane), FeCl2 (0.40 g, 2.56 mmol). Yield: 1.17 g (55%)
orange-colored powder. 1H NMR (CDCl3): δ 1.12-1.26 (m,
i
42H, Pr), 4.23 (s, 4H, CpH), 6.88-7.26 (m, 20H, ArH). 13C
NMR (CDCl3): δ 12.17, 17.92, 66.70, 81.56, 121.80, 125.46,
127.17, 130.04, 136.67.
Syn th eses of 1,1′-Bis(tr ia lk ylsiloxy)fer r ocen es 17-20.
A solution of the respective 1-(trialkylsiloxy)cyclopentadiene
in THF (50 mL) was cooled to -40 °C and treated with n-BuLi
(2.5 M in hexane). After stirring for 15 min, the solution was
warmed to 0 °C, FeCl2 was added, and the reaction mixture
was slowly allowed to warm to room temperature and stirred.
The solvents were evaporated, the oily residue was extracted
with cyclohexane, and the solution was filtered over a silica
plug. Finally the cyclohexane was evaporated and the residue
dried in vacuo.
1,1′-Bis(ter t-b u t yld im et h ylsiloxy)fer r ocen e
(17).
Scale: 1-(tert-butyldimethylsiloxy)cyclopentadiene (1.00 g, 5.09
mmol), n-BuLi (2.04 mL, 5.09 mmol, 2.5 M in hexane), FeCl2
(0.31 g, 2.50 mmol). Yield: 0.58 g (52%); orange-colored oil.
t
1H NMR (C6D6): δ 0.15 (s, 12H, SiCH3), 0.97 (s, 18H, Bu),
3.78 (br, 4H, CpH), 4.05 (br, 4H, CpH). 13C NMR (C6D6): δ
-4.44, 18.31, 25.88, 60.75, 63.21, 122.45.
1,1′-Bis(tr iisop r op ylsiloxy)fer r ocen e (18). Scale: 1-(tri-
isopropylsiloxy)cyclopentadiene (1.00 g, 5.09 mmol), n-BuLi
(2.04 mL, 5.09 mmol, 2.5 M in hexane), FeCl2 (0.31 g, 2.50
mmol). Yield: 1.11 g (56%); mp. 172 °C; orange-colored pow-
i
der. 1H NMR (CDCl3): δ 1.02-1.14 (m, 42H, Pr), 3.75 (“t”, J
) 1.9 Hz, 4H, CpH), 3.98 (“t”, J ) 1.9 Hz, 4H, CpH). 13C NMR
(CDCl3): δ 12.28, 17.82, 60.88, 62.52, 122.29.
1,1′-[(1,4,10-Tr ioxa -7,13-d ia za cyclop en t a d eca n e-7,13-
d iyl)d ieth oxy]-3,3′,4,4′-tetr a p h en ylfer r ocen e (24). A mix-
ture of 1,1′-bis(2-iodoethoxy)-3,3′,4,4′-tetraphenylferrocene (500
mg, 0.60 mmol), diaza-15-crown-5 (127 mg, 0.60 mmol), and
Na2CO3 (320 mg, 3.0 mmol) in CH3CN (50 mL) was heated
under reflux for 7 d. The cold reaction mixture was filtered,
the volatiles were evaporated in vacuo, and the residue
was dissolved in a mixture of CH2Cl2 (20 mL) and water
(10 mL). The organic layer was separated and washed with
water twice (10 mL), dried over MgSO4, filtered, and evapo-
rated. The residue was purified by chromatography (ethyl
acetate-diethylamine: 20/1). Yield: 320 mg (67%), mp 210
°C; orange-colored powder. Anal. Calcd for C48H52FeN2O5
(792.8): C, 72.72; H, 6.61; N, 3.53. Found: C, 72.63; H, 6.77;
N, 3.66. MS, m/z (%): 792 (39) [M+]. 1H NMR (CD3CN): δ
1,1′-Bis(t r iet h ylsiloxy)-3,3′,4,4′-t et r a m et h ylfer r ocen e
(19). Scale: 1-(triethylsiloxy)-3,4-dimethylcyclopentadiene
(1.50 g, 6.67 mmol), n-BuLi (2.67 mL, 6.67 mmol, 2.5 M in
hexane), FeCl2 (0.42 g, 3.34 mmol). Yield: 0.81 g (48%) dark
yellow oil. 1H NMR (CDCl3): δ 0.45-0.71 (m, 12H, SiCH2CH3),
0.88-0.99 (m, 4H, CH2CH3), 1.78 (s, 12H, CpCH3), 3.55 (s, 4H,
CpH). 13C NMR (CDCl3): δ 11.97, 12.19, 17.90, 63.44, 76.52,
118.13.
1,1′-Bis(tr iisop r op ylsiloxy)-3,3′,4,4′-tetr a m eth ylfer r o-
cen e (20). Scale: 1-(triisopropylsiloxy)-3,4-dimethylcyclopen-
tadiene (2.00 g, 7.50 mmol), n-BuLi (3.00 mL, 7.50 mmol, 2.5
M in hexane), FeCl2 (0.47 g, 3.75 mmol). Yield: 1.19 g (54%)
i
dark yellow oil. 1H NMR (CDCl3): δ 1.07 (s, 42H, Pr), 1.81
(s, 12H, CpCH3), 3.58 (s, 4H, CpH). 13C NMR (CDCl3): δ 11.98,
12.19, 17.89, 63.42, 76.36, 117.95.
2.65-2.69 (m, 8H, NCH2), 2.87 (t,
J ) 6.8 Hz, 4H,
CpOCH2CH2N), 3.47-3.52 (m, 12H, CH2O), 4.21 (t, J ) 6.9
Hz, 4H, CpOCH2), 4.33 (s, 4H, CpH), 6.79-7.00 (m, 20H, ArH).
13C NMR (CD3CN): δ 55.89, 57.63, 58.41, 61.94, 70.75, 70.97,
71.36, 71.68, 82.87, 126.85, 127.01, 127.64, 128.41, 130.84,
137.45.
1,1′-Bis(2-ch lor et h oxy)-3,3′,4,4′-t et r a p h en ylfer r ocen e
(21). A mixture of 1,1′-bis(trimethylsiloxy)-3,3′,4,4′-tetraphen-
ylferrocene (5.00 g, 7.50 mmol), 1-tosyl-2-chloroethane (17.5
g, 75.0 mmol), Na2CO3 (10.0 g , 95.5 mmol), CH3CN (500 mL)
and H2O (50 mL) were heated under reflux for 3 d. After the
reaction mixture was cooled to room temperature, the Na2CO3
was filtered off and the solvents were evaporated in vacuo.
The residue was purified by chromatography (silica, cyclo-
hexane-ethyl acetate: 4/1). Yield: 3.88 g (80%), mp 154 °C;
orange-colored powder. In addition to the desired product a
small amount of 3,3′,4,4′-tetraphenyldioxa[4]ferrocenophane
was isolated. Anal. Calcd for C38H32Cl2FeO2 (647.4): C, 70.50;
H, 4.98. Found: C, 70.20; H, 5.23. 1H NMR (CDCl3): δ 3.52
(t, J ) 5.7 Hz, 4H, CH2Cl), 3.82 (t, J ) 5.9 Hz, 4H, OCH2),
1,1′-[(1,4,10,13-Tet r a oxa -7,16-d ia za cyclooct a d eca n e-
7,16-d iyl)d ieth oxy]-3,3′,4,4′-tetr a p h en ylfer r ocen e (25). A
mixture of 1,1′-bis(2-iodoethoxy)-3,3′,4,4′-tetraphenylferrocene
(500 mg, 0.60 mmol), diaza-18-crown-6 (156 mg, 0.60 mmol),
and K2CO3 (840 mg, 6.0 mmol) in CH3CN (75 mL) was heated
under reflux for 7 d. The cold reaction mixture was filtered,
the volatiles were evaporated in vacuo, and the residue was
dissolved in CH2Cl2 (20 mL) and water (10 mL). The organic
layer was separated and washed with water twice (10 mL),
dried over MgSO4, filtered, and evaporated. The residue was

Oxaferrocene Cryptands as Molecular Switches
Organometallics, Vol. 16, No. 26, 1997 5957
purified by chromatography (ethyl acetate-diethylamine: 20/
1). Yield: 340 mg (68%), mp 208 °C, orange-colored powder.
Anal. Calcd for C50H56FeN2O6 (836.8): C, 71.76; H, 6.74; N,
3.35. Found: C, 71.55; H, 6.87; N, 3.27. MS, m/z (%): 836
(53) [M+]. 1H NMR (CD3CN): δ 2.71-2.77 (m, 8H, NCH2),
2.94 (t, J ) 7.0 Hz, 4H, CpOCH2CH2N), 3.55-3.62 (m, 16H,
CH2O), 4.21 (t, J ) 6.9 Hz, 4H, CpOCH2), 4.40 (s, 4H, CpH),
6.91-7.09 (m, 20H, ArH). 13C NMR (CDCl3): δ 54.90, 56.72,
60.45, 69.75, 70.08, 71.02, 81.95, 125.61, 125.86, 127.28,
129.83, 135.91.
Ack n ow led gm en t. This work was supported by the
Deutsche Forschungsgemeinschaft through the Gra-
duiertenkolleg Ungepaarte Elektronen in Chemie und
Biologie, the Institut fu¨r Anorganische und Analytische
Chemie, the Fonds der Chemischen Industrie, the
Freiburger Wissenschaftliche Gesellschaft, and the DFG
through a Heisenberg fellowship to one of us (H.P.).
OM970681F