Communication between Co2(CO)4dppm units via polyferrocenylalkyne linkages

Article abstract of DOI:10.1021/om000493g

The synthesis and characterization of a family of ethynylbi- and ethynylterferrocenyls, I-Fc₂C=CR, Fc₂C=CR, and Fc₂(C=CR)₂ (R = Fc, SiMe₃), I-Fc₃C=CSiMe₃, Fc₃C=CSiMe₃, Fc₃(C=CSiMe₃)₂, and their Co₂(CO)₆ and Co₂(CO)₄dppm complexes are described. The structure adopted by Fc₂{[Co₂(CO)₄dppm C₂SiMe₃]}₂ minimizes the steric interactions between the biferrocene and dppm moieties. Electrochemical studies show that identical redox centers-ferrocenyl termini, Co²(CO)₄dppm units, or ferrocenyl groups in the corecommunicate independently with each other.

Full text of DOI:10.1021/om000493g

Organometallics 2000, 19, 5039-5048
5039
Com m u n ica tion betw een Co2(CO)4d p p m Un its via
P olyfer r ocen yla lk yn e Lin k a ges
Leigh-Anne Hore, C. J ohn McAdam, J oy L. Kerr, Noel W. Duffy,
Brian H. Robinson,* and J im Simpson
Department of Chemistry, University of Otago, P.O. Box 56, Dunedin, New Zealand
Received J une 13, 2000
The synthesis and characterization of a family of ethynylbi- and ethynylterferrocenyls,
I-Fc2CtCR, Fc2CtCR, and Fc2(CtCR)2 (R ) Fc, SiMe3), I-Fc3CtCSiMe3, Fc3CtCSiMe3,
Fc3(CtCSiMe3)2, and their Co2(CO)6 and Co2(CO)4dppm complexes are described. The
structure adopted by Fc2{[Co2(CO)4dppm C2SiMe3]}2 minimizes the steric interactions
between the biferrocene and dppm moieties. Electrochemical studies show that identical
redox centerssferrocenyl termini, Co2(CO)4dppm units, or ferrocenyl groups in the cores
communicate independently with each other.
In tr od u ction
the crystal structure of a mixed valence salt revealed
nonequivalence in the ferrocenyl moieties.5b Since both
symmetrical and asymmetrical biferrocene derivatives
are accessible, it has the added advantage of bringing
to bear environmental factors to the magnitude of
communication in the array. Thus biferrocene com-
pounds can be utilized as both links and redox centers.
Polyferrocenes are less well understood, the main reason
being the difficulty in controlling syntheses of function-
alized derivatives, although a series of alkylpolyfer-
rocenes has been reported recently.6 Our strategy was
to start with a simple ferrocene compound and assemble
the polyferrocene in a coherent synthetic sequence. A
useful template is a ferrocenylalkynyl unit. First, the
synthetic methodology for alkyne elaboration is well
established. Second, unbranched ethynyl or polyalkynyl
links appear from both experimental and theoretical
work to be very efficient communicators; this connectiv-
ity is unaffected by the aromaticity of an interpolated
group.7 Third, the alkyne allows the addition of other
electroactive centers. In particular, we have shown that
coordination of a Co2(CO)xLy fragment across the -Ct
C- bond provides a tunable redox center, as the electron
density in the relevant redox orbitals can be varied by
changing L.8,9 The Co2(CO)4dppm fragment is especially
important as, due to “clamping” of the Co-Co bond by
the dppm, it is not affected by the fast ECE reactions
Considerable attention has been given recently to
molecules containing highly conjugated organometallic
systems, especially those with multiple electroactive
centers joined by a conjugated assembly. These com-
pounds are of interest for their potential application in
the construction of molecular scale electronic devices or
the manifestation of advantageous bulk properties such
as catalytic activity, large nonlinear optical responses,
magnetism, and electron transfer.1 Molecules that
contain ligands with switchable redox behavior add
another dimension. Our interest is the integration of
switchable electrophores into arrays containing fluo-
rescent centers. The link, which must be capable of
transmitting the electronic effects while maintaining
stereochemical control, is a crucial component of this
module. In this paper we focus on possible links which
can also function as electrophores: polyferrocenyl-Ct
C- or polyferrocenyl-C2[Co2(CO)4dppm].
Ferrocene, because of its simple redox behavior,2 is a
ubiquitous moiety wherever an electrophore is required,
and it has played a major role in molecular sensors,
electrochemical agents, liquid crystals, and nonlinear
optical material development.3 Biferrocene is another
potential electrophore. Its redox chemistry, which dis-
plays two reversible one-electron couples, has attracted
attention for many years, and electronic communication
between the two Fc+/0 couples is well established.4
Electron transfer rates and charge delocalization of
alkylated biferrocenium salts have been studied,5a and
(5) (a) Hendrickson, D. N.; Oh, S. M.; Dong, T.-Y.; Kambara, T.;
Cohn, M. J .; Moore, M. F. Comments Inorg. Chem. 1985, 4, 329. (b)
Delville, M.-H.; Robert, F.; Gouzerh, P.; Linares, J .; Boukheddaden,
K.; Varret, F.; Astruc, D. J . Organomet. Chem. 1993, 451, C10.
(6) (a) Dong, T.-Y.; Lee, W.-Y.; Su, P.-T.; Chang, L.-S.; Lin, K.-J .
Organometallics 1998, 17, 3323. (b) Dong, T.-Y.; Ho, P.-H.; Lai, X.-Q.;
Lin Z-W.; Lin, K.-J . Organometallics 2000, 19, 1096.
(7) Duffy, N. W.; McAdam, C. J .; Nervi, C.; Osella, D.; Ravera, M.;
Robinson, B. H.; Simpson, J . Inorg. Chim. Acta 1996, 247, 99.
(8) Robinson, B. H.; Simpson, J . In Paramagnetic Organometallic
Species in Activation/ Selectivity Catalysis; Chanon, M. Ed.; Klymer:
Durdrecht, 1989; p 357.
(1) (a) Conjugated Polymer Materials: Opportunities in Electronic,
Optoelectronic and Molecular Electronics; Bredas, J . L., Chance, R. R,
Eds.; NATO ASI Series, Vol. 182; Kluwer: Dordrecht, 1990. (b)
Manners I. R. Soc. Chem. Annu. Rep.-Inorg. Chem. 1994, 91, 131.
(2) Zanello P. In Ferrocenes; Togni, A., Hayashi, T., Eds.; VCH:
Weinheim, 1995; Chapter 7.
(3) (a) Togni, A. In Ferrocenes; VCH: Weinheim, 1995; Chapter 8.
(b) Deschenaux, P.; Goodby, J . W. In Ferrocenes; VCH: Weinheim,
1995; Chapter 9.
(4) (a) Brown, G. M.; Meyer, T. J .; Cowan, D. O.; Le Vanda, C.;
Kaufman, F. J .; Roling, P. V.; Rausch, M. D. Inorg. Chem. 1975, 14,
506. (b) Cowan, D. O.; Le Vanda, C.; Park, J .; Kaufman, F. J . Acc.
Chem. Res. 1973, 6, 1. (c) Muller-Westerhoff, U. T. Angew. Chem., Int.
Ed. Engl. 1986, 25, 702.
(9) McAdam, C. J .; Duffy, N. W.; Robinson, B. H.; Simpson, J .
Organometallics 1996, 16, 3935. McAdam, C. J .; Duffy, N. W.;
Robinson, B. H.; Simpson, J . J . Organomet. Chem. 1997, 527, 179.
Duffy, N. W.; McAdam, C. J .; Robinson, B. H.; Simpson, J . J .
Organomet. Chem. 1998, 565, 19. Duffy, N. W.; McAdam, C. J .;
Robinson, B. H.; Simpson, J . J . Organomet. Chem. 1999, 573, 36.
(10) Arewgoda, M.; Rieger, P. H.; Robinson, B. H.; Simpson, J . J .
Am. Chem. Soc. 1982, 104, 5633.
10.1021/om000493g CCC: $19.00 © 2000 American Chemical Society
Publication on Web 10/27/2000

5040 Organometallics, Vol. 19, No. 24, 2000
Hore et al.
Sch em e 1
seen with many other alkynedicobalt carbonyl redox
centers.10 This clean oxidation response provides a probe
for estimating the efficiency of communication along the
array.
effect of the relative mole ratio of iodine to lithiated
ferrocene on the product composition. It was found that
by varying the ratio of iodine to ferrocene it was possible
to control whether Fc2I2 or terferrocenyliodides “Fc3Ix”
were the major products. Chromatographic separation
of the iodinated Fc3 compounds proved difficult with the
added complication that they degrade on silica or
alumina. The major Fc3 component was 2, which was
the key reagent for further controlled elaboration of the
polyferrocene backbone. In practice, it was convenient
to utilize the crude “Fc3” mixture in reactions with
alkynes, as the products were much easier to separate.
Long and co-workers have reported recently11 both
ethynyl and organometallic derivatives of biferrocene
as precursors for materials having advantageous non-
linear optical behavior. 1′,6′-Bis(trimethylsilylethynyl)-
biferrocene provided the route to the acetylides 1′,6′-
bis{manganese(I)tricarbonyl[1,2-bis(diphenylphosphino)-
methane}biferrocene and 1′,6′-bis{chlororutheniumbis-
(diphenylphosphino)methane}biferrocene. Electrochemi-
cal data showed that the organometallic moiety donated
electron density through the -CtC- connection, which
stabilized the Fe(II) center toward oxidation but de-
creased the relative stability of the Fe(II)Fe(III) mixed
valence species.11 A restriction on the versatility of their
synthetic route is the low yield and unpredictable
syntheses of Fc2I2.12 Our intention was to adapt the
iodination of ferrocene and lithioferrocene so that poly-
ferrocene complexes with -CtC- bonds could be pre-
pared in quantities suitable for further incorporation
into fluorescent arrays. This paper describes biferrocene
and terferrocene linkages with or without cluster redox
centers and with Me3Si or ferrocene end-groups.
Cu(I)/Pd(II)-catalyzed coupling of a terminal alkyne
to 1 gave the desired ethynylbiferrocene (Scheme 1). An
important step forward was the ability to control the
coupling reaction with 1 and HCtCR (R ) SiMe3, Fc),
as this enabled the iodoethynylbiferrocenes 3 and 4 to
be isolated in reasonable yield. Compound 3 is poten-
tially a useful synthon for elaboration of the ferrocenyl
framework. Thus, coupling of 3 with HCtCR gave the
target bisethynylferrocenes 6 (R ) SiMe3) and 7 (R )
Fc) in good yields (Scheme 1). Long and co-workers11
had previously prepared the bis-ethynyl complex 6
directly from 1. Furthermore, the lithiation of 3 and
quenching with water gave the mono-ethynyl derivative
5.
Resu lts a n d Discu ssion
The coupling reaction between 2 and trimethylsilyl-
acetylene gave 8, the terferrocene analogue of 3, but it
was moderately unstable and was only characterized
spectroscopically. The mono-ethynyl 9 and bis-ethynyl
10 complexes were prepared from coupling reactions
similar to those described above for the biferrocene
derivatives (Scheme 2). Similar reactions with ethynyl-
ferrocene appeared to give ferrocenyl analogues of 9 and
10, but they were remarkably unstable; the reason for
this instability is unknown.
Syn th esis. The two strategic reagents were diiodobi-
ferrocene 1 and diiodoterferrocene 2. Rausch,12 Long,11
and their co-workers have reported the preparation of
1 from FcLi. In our experience the yields were quite
variable, and iodinated polyferrocenes were also pro-
duced in this reaction. We therefore investigated the
(11) Colbert, M. C. B.; Hodgson, D.; Lewis, J .; Long, N. J .; Raithby,
P. R. Polyhedron 1995, 14, 2759.
(12) Kovar, R. F.; Rausch, M. D.; Rosenberg. H. Organomet. Chem.
Synth. 1970/1971, 1, 173.
Satisfactory analytical and 1H and 13C NMR data

Communication between Co2(CO)4dppm Units
Organometallics, Vol. 19, No. 24, 2000 5041
Sch em e 2
Sch em e 3
were obtained for 2-10. The complicated NMR in the
ferrocene region of the terferrocene complexes made
assignment difficult, but the ratio of polyferrocene to
trimethylsilyl resonances identified the products; this
ratio was also useful in monitoring the reactions.
Reaction of 5, 6, 9, or 10 with Co2(CO)8 gave the green-
black Co2(CO)6 derivatives 11-14 (Scheme 3). Of note
the deshielding effect of the metal fragment. The ν(CO)
bands were similar to those of other Co2(CO)6 complexes,
the A1 in-phase mode at 2083 cm-1 indicating that there
is no unusual steric congestion around the Co2(CO)6 unit
caused by the polyferrocenyl moiety.9
Substitution reactions of the Co2(CO)6 complexes with
dppm proceeded smoothly to give a series of red-brown
Co2(CO)4dppm complexes 15-20 (Scheme 4).
1
in the H NMR is the downfield shift of δ(SiMe3) due to

5042 Organometallics, Vol. 19, No. 24, 2000
Hore et al.
Sch em e 4
A change from green-black to red-brown color upon
dppm coordination is unusual; in fact the reverse
normally holds. A shift in ν(CO) of ∼60 cm-1 to lower
energy relative to the Co2(CO)6 complexes shows that
the dppm has substituted symmetrically across the Co-
Co bond.9 The increase in electron density on the Co2-
(CO)4(alkyne) unit shifts δH(SiMe3) further downfield
to 0.55-0.56 ppm and clearly delineates the SiMe3
groups in the complexes 16 and 19, where there is one
free CtC group. Both 16 and 19 are produced in the
reactions involving the bis-substituted Co2(CO)6 com-
plexes 12 and 14, respectively, with the metal fragment
being lost during the reaction. This is a common feature
for dppm reactions with polyalkyne-Co2(CO)6 com-
plexes and is presumably because two Co2(CO)4dppm
units cannot be accommodated sterically on two adja-
cent CtC bonds.9 A similar steric explanation could be
advanced to account for the generic instability of 11-
20 and the greater lability of the Co2(CO)4dppm com-
pared to Co2(CO)6 analogues. An X-ray structural
analysis of 17 was therefore undertaken in order to
establish the steric parameters for these compounds.
X-r a y Str u ctu r e Deter m in a tion of 17. The struc-
ture of compound 17 was determined by single-crystal
X-ray structure analysis. A perspective view is shown
in Figure 1, which defines the atom-numbering schemes.
Selected bond length and angle information are given
in Table 1. The unit cell contains discrete, well-
separated molecules of the biferrocene complex and
ethyl acetate solvent, with the closest intermolecular
contact, not involving H atoms, being 3.214(2) Å be-
tween O(111) and C(3A).
Molecules of 17 consist of a biferrocene unit with
approximately tetrahedral (µ-alkyne)dicobalt moieties
bound to the unsubstituted cyclopentadiene rings. They
lie about an inversion center located at the midpoint of
the C(8)-C(8a) bond of the fulvalenide ligand; this
imposes a trans-conformation on the location of the (µ-
alkyne)dicobalt substituents with respect to the inver-
sion center, with a steplike conformation6a of the two
ferrocenyl rings. Both Me3SiC2Co2(CO)4(dppm) substit-
uents point away from the central biferrocenyl unit and
the outer C atoms of the alkyne groups each carry
trimethylsilyl substituents. The cyclopentadiene rings
in the biferrocene adopt an approximately staggered
conformation, with a mean torsion angle C(m)-C(1g)-
C(2g)-C(n) of 26.1(1)° (C(m) ) C(3)-C(7), C(n) ) C(8)-
C(12), and C(1g) and C(2g) are the corresponding ring
centroids), and the dihedral angle between the least-
squares planes of the Cp rings of each ferrocenyl moiety
is 3.59(25)°. The average Fe-C distances (2.049(8) Å)
are not unusual, and the Fe atom is almost equidistant
between the Cp ring (1.6502(9) Å) and fulvalenide
(1.6546(9) Å). These are similar to the distances found
in ferrocene itself.
The alkyne units adopt the normal cis-bent configu-
ration, as expected for perpendicular acetylenes. The
significant variation between the alkyne bend-back
angles to the Me3Si substituent, Si(1)-C(1)-C(2) 148.15-
(15)°, and to the biferrocene, C(3)-C(2)-C(1) 140.86-
(17)°, undoubtedly reflects the requirement to minimize
steric interactions between the (µ-alkyne)dicobalt and
biferrocenyl residues. The (µ-alkyne)dicobalt cores of the
Me3SiC2Co2(CO)4(dppm) moieties are typical of perpen-

Communication between Co2(CO)4dppm Units
Organometallics, Vol. 19, No. 24, 2000 5043
F igu r e 1. Perspective view of 17 showing the atom-numbering scheme, with displacement ellipsoids drawn at the 50%
probablility level. For clarity H atoms have been omitted, and only two C atoms of consecutively numbered cyclopentadienyl
and phenyl rings and the O atoms of the carbonyl ligands have been labeled.
Ta ble 1. Selected Bon d Len gth s (Å) a n d An gles
(d eg) for 17
found where two P donor atoms chelate to a single Co
atom or when a second dppm or dam18a ligand binds in
(µ-η2)-fashion across the Co-Co bond. Co-Co bond
extension in these cases has been attributed to steric
affects, and the lengthening observed here is most likely
a response by the molecule to the significant steric
requirements of the biferrocenyl moiety and the adja-
cent dppm ligands.
The dppm ligands adopt (µ-η2) coordination, bridging
the Co-Co bonds of the C2Co2 fragment. Each cobalt
atom also carries an “axial” CO and “pseudoequatorial”
CO and P ligands. The CO and P atoms deviate only
slightly from the classical “sawhorse” arrangement with
respect to the cobalt atoms (torsion angles P(1)-Co(1)-
Co(2)-P(2) 6.19(2)°; C(111)-Co(1)-Co(2)-C(211) 17.6-
(2)°; (112)-Co(1)-Co(2)-C(212) 7.6(1)°). The Co-P bond
distances are unremarkable and the bridging dppm
ligands bind so as to minimize steric interactions
between the phenyl rings of the dppm ligands and the
adjacent ferrocenyl centers.
Red ox Ch em istr y. The molecules described herein
provided an opportunity to explore the efficiency of
intravalence charge transfer within the polyferrocene
core when alkyne ferrocenyl groups are attached and,
in the Co2(CO)4dppm complexes, where the communica-
tion axis is no longer linear.
Electr och em istr y of 5, 6, 9, a n d 10. These com-
plexes display the typical one-electron, chemically re-
versible, oxidation waves of the ferrocenyl units in the
cyclic and square wave voltammetry. The number of
ferrocenyl units determines the number of i/V responses.
For example, 10 with three units displays three discrete
waves (Figure 2). Ferrocene is a net electron donor, and
Si-C(1)
1.8425(19)
1.458(2)
1.437(2)
1.438(2)
1.421(3)
1.419(3)
1.423(3)
1.429(3)
1.430(3)
1.472(4)
Co(1)-C(1)
Co(1)-C(2)
Co(2)-C(1)
Co(2)-C(2)
Co(1)-P(1)
Co(1)-Co(2)
Co(2)-P(2)
P(1)-C(120)
P(2)-C(120)
C(1)-C(2)
1.9828(18)
1.9658(18)
1.9812(18)
2.0073(17)
2.2139(5)
2.4815(3)
2.2458(5)
1.8411(18)
1.8470(18)
1.353(3)
C(3)-C(2)
C(3)-C(4)
C(3)-C(7)
C(4)-C(5)
C(5)-C(6)
C(6)-C(7)
C(8)-C(9)
C(8)-C(12)
C(8)-C(8)#1
C(2)-C(1)-Si
148.18(15)
140.86(17)
94.189(15)
98.560(15)
C(1)-Co(2)-P(2) 138.29(5)
C(2)-Co(2)-P(2) 100.11(5)
C(1)-C(2)-C(3)
P(1)-Co(1)-Co(2)
P(2)-Co(2)-Co(1)
P(1)-C(120)-P(2) 108.18(9)
dicular alkyne complexes with an interline angle C(1)-
C(2)/Co(1)-Co(2) of 91.5(1)°. The C2Co2 units are linked
via a short, 1.458(2) Å, C-C single bond to the cyclo-
pentadiene rings of the biferrocene. This bond distance
suggests some degree of electron delocalization between
the alkyne and biferrocene units (vide infra). The alkyne
C(1)-C(2) bond length, 1.353(3) Å, is in reasonable
agreement with those found in other C2Co2 systems.
However, distortion of the quasi tetrahedral C2Co2 core
is evidenced by the nonequivalence of the Co-C bond
lengths [Co(1)-C(1) 1.9828(18) Å, Co(1)-C(2) 1.9658-
(18) Å, Co(2)-C(1) 1.9812(18) Å, Co(2)-C(2) 2.0073(17)
Å]; the net effect tilts the C(1)-C(2) vector relative to
the Co-Co bond toward the central biferrocenyl moiety.
The Co(1)-Co(2) bond is long, 2.4815(3) Å, in compari-
son with most (µ-alkyne)dicobalt complexes with (µ-η2)-
bound dppm ligands. In such compounds Co-Co bond
distances are generally in the range 2.46-2.47 Å; indeed
that is also the norm for unsubstituted C2Co2(CO)6
systems.19 Co-Co bonds as long as 2.50-2.52 Å are
as a consequence the potential for the first oxidation
+/0
process E1/2
moves cathodically as the number of
ferrocenyl rings in the polyferrocene Fc(Fc)N (N ) 1-3)
(13) Barlow, S.; O’Hare, D. Chem. Rev. 1997, 96, 637.
(14) Seiler, P.; Dunitz, J . D. Acta Crystallogr. Sect. B 1979, 35, 1068.
(15) Bonnet, J .-J .; Mathieu, R. Inorg. Chem. 1978, 17, 1973.
(16) Hoffman, D. M.; Hoffmann, R.; Fisel, C. R.. J . Am. Chem. Soc.
1982, 104, 3858.
(17) Dickson R. S.; Fraser, P. J . Adv. Organomet. Chem. 1974, 12,
323.
(18) (a) Bird, P. H.; Fraser A. R.; Hall, D. N. Inorg. Chem. 1977, 16,
1023. (b) Lewis, J .; Lin, B.; Khan, M. S.; Al-Mandhary, M. R. A.;
Raithby, P. R. J . Organomet. Chem. 1994, 484, 161. McAdam, C. J .;
Duffy, N. W.; Robinson, B. H.; Simpson, J . Organometallics 1996, 15,
3935.
(19) Bianchini, C.; Dapporto, P.; Meli, A. J . Organomet. Chem. 1979,
174, 205.

5044 Organometallics, Vol. 19, No. 24, 2000
Hore et al.
F igu r e 2. Square wave voltammogram (60 Hz, 2 mV) of
10 in CH2Cl2, Pt, 0.1 nBu4NPF6.
Ta ble 2. Electr och em ica l Da ta
a,b
[E1/2
]
F igu r e 3. Square wave voltammogram (60 Hz, 2 mV) of
7 in CH2Cl2, Pt, 0.1 nBu4NPF6.
[1+/0] [2+/1+] [3+/2+] [4+/3+] ∆E(1)c ∆E(2)c ∆E(3)b,c
Fc2
5
6
0.46
0.50
0.51
0.43
0.54
0.50
0.46
0.48
0.37
0.42
0.41
0.37
0.36
0.31
0.80
0.81
0.81
0.57
0.90
0.95
0.76
0.66
0.59
0.73
0.76
0.55
0.51
0.51
340
310
300
350
360
450
300
180
220
310
350
180
150
200
Sch em e 5
7
0.62
0.97
140
240
21a d
21bd
15
17
1.11
1.02
0.97
0.99
0.99
0.85
0.70
0.76
1.26
360
380
260
230
300
190
250
f
Fc3
9
10
18
20
1.05
1.06e
1.04
200
40e
280
f
∆E1/2(2) ) 230 mV at 200 mV s-1. An explanation for
the variation in the ∆E1/2 values is likely to lie in the
different charge separations in the oxidized species, but
there is no satisfactory theoretical model at present.
Electr och em istr y of 7. Compound 7, which has four
ferrocenyl units, displayed four reversible one-electron
oxidation processes in its cyclic and square wave vol-
tammetry (Figure 3). Although the four Fc units could
be grouped as an “inner” biferrocenyl core and two Fc
termini there, is no reason a priori why all four should
not communicate. Data for the sequence Fc(CtC)nFc22
Fc4
FcC2SiMe3 0.61
FcC2Fc
0.63
0.76
130
a
Volts; all data referenced to decamethylferrocene; Pt, 200 mV
b
s-1 in dichloromethane. Data in italics refer to Co2 redox center.
c In mV. Compounds 21a and 21b, Fc2(CtCPh)2 and Fc2(CtCH)2,
d
respectively.23 e Two-electron process in CV; 40 mV estimated from
SQV. f Ref 4a.
increases (Table 2). The relationship, in dichloromethane
at 200 mV s-1 referenced against decamethylferrocene,
can be expressed as E1/2+/0(V) ) [0.55 - N(0.09)]. A
pendant -CtC-SiMe3 on a ferrocenyl redox unit causes
23
and Fc2(CtCPh)2 show that the interpolation of an
+/0
an anodic shift in E1/2 (Table 2), and for 5 and 9 the
alkyne between Fc termini or attachment of an alkyne
relationship is E1/2+/0(V) ) [0.55 - N(0.09) + 0.05].
Intuitively, the -CtC-SiMe3 functional group should
not withdraw electron density from the ferrocenyl
moiety, and it is probable that the anodic shift is a
reflection of the relative stability of the mixed-valence
Fc-FcN+ cations. In support of a noninductive explana-
tion is the fact that the addition of a second -CtC-
SiMe3 to the polyferrocenyl core has virtually no effect
link to Fc2 causes an anodic shift in E1/2+/0. On the other
hand, a Fc2 group would act as a good donor to the Fc
+/0
termini. For these reasons E1/2
(A) at 0.43 V is
assigned to oxidation of the two communicating Fc
termini. It is unlikely that ∆E1/2 between the “outer”
Fc in 7 would be much greater than that for FcCtCFc
2+/+
(130 mV). E1/2
(B) at 0.57 V [∆E1/2(I) ) 140 mV] is
therefore assigned to the second oxidation process of the
Fc termini (Scheme 5). The remaining two oxidation
processes at E1/2 ) 0.62 V (C) and 0.97V (D) are
assigned to the Fc2 core; as expected, these are anodic
of those for 6 since there is a positive charge on the
ferrocenyl termini. Support for this assignment comes
from the consistency of the “communication parameter”
of ∼300 mV for the Fc2 core in 5-7.
+/0
on E1/2 (cf. compare 5 and 6, 9 and 10).
For the polyferrocenes ∆E1/2(1), the separation be-
tween E1/2+/0 and E1/22+/+, decreases with increasing size
of the ferrocenyl core. It is noteworthy that ∆E1/2(1) <
Fc(Fc)N ∆E1/2(2) for N ) 2 and 3, presumably because
the higher charge is effectively dispersed in the poly-
ferrocene, thereby stabilizing the cations. Given the C2-
SiMe3 substituent effect noted above, it was therefore
surprising to find that, first, ∆E1/2(1) is essentially the
same for the comparable bi- and terferrocenes 5/9 and
6/10 and, second, the reverse order ∆E1/2(1) > ∆E1/2(2)
The important conclusions from these data are the
following: the interpolation of a Fc2 unit does not impede
the communication between the Fc termini, and, to a
first approximation, communication between the “outer”
and “inner” sets of Fc redox centers is mutually exclu-
sive.
for 9 and 10. Thus ∆E1/2(1), the separation between
+/0
E1/2
and E1/22+/+, for 6 is 300 mV, whereas the
sequential values for 10 are ∆E1/2(1) ) 350 mV and
(21) Gregson, D.; Howard, J . A. K. Acta Crystallogr. 1983, C39, 1024
(22) Ribot, A.-C.; Launay, J . P.; Sachtleben, M. L.; Li, H., Spangler,
C. W. Inorg. Chem. 1996, 35, 3735.
(20) Aggarwal, R. P.; Connelly, N. G.; Crespo, M. C.; Dunne, B. J .;
Hopkins, P. M.; Orpen, A. G. J . Chem. Soc., Dalton Trans. 1992, 655.

Communication between Co2(CO)4dppm Units
Organometallics, Vol. 19, No. 24, 2000 5045
Sch em e 7
F igu r e 4. Square wave voltammogram (60 Hz, 2 mV) of
15 in CH2Cl2, Pt, 0.1 nBu4NPF6.
moieties is equivalent whether they are at the terminus
or as a biferrocene unit.
The situation is different for the Fc3 complex 18. Four
reversible one-electron processes are displayed in the
cyclic and square wave voltammetry of 18 (Figure 5);
the first three have E1/2 values similar to those for 10
and are assigned to the Fc3 core; that at E1/2 ) 1.06 V
is assigned to a chemically reversible oxidation of the
Co2(CO)4dppm unit (Scheme 6). Coordination of the
cluster to the alkyne link has reduced ∆E(1) from 310
(cf. 9) to 180 mV but with little change in ∆E(2). This
is in contrast with the Fc2 system, where there was little
perturbation of the intra-metallocene “communication”
by coordinating the cluster unit.
F igu r e 5. Square wave voltammogram (60 Hz, 2 mV) of
18 in CH2Cl2, Pt, 0.1 nBu4NPF6.
Electr och em istr y of Co2(CO)4d p p m Der iva tives.
Because of the instability of the Co2(CO)6 derivatives
and the rapid ECE reactions involving the carbonyl
fragment,8,10 a detailed examination of the oxidation
electrochemistry of only the Co2(CO)4dppm derivatives
was carried out. The latter complexes also provided an
opportunity to examine whether a disparate redox
center also behaved independently of the polyferrocenyl
core.
The effect on the -CtC- bond by coordination and
electron donation from the Co2(CO)4dppm unit caused
a small cathodic shift in the polyferrocenyl core poten-
tials compared with the parent alkynes (Table 2). The
Fc2 complex with one Co2(CO)4dppm unit 15 displayed
cyclic and square wave voltammetric responses which
clearly delineate the redox centers (Figure 4). Two
chemically reversible one-electron processes E1/2 ) 0.46
and 0.76 V are assigned to the Fc2 core and the
chemically reversible one-electron step at more positive
potentials, E1/2 ) 1.11 V, to the Co2(CO)4dppm unit. It
is significant that ∆E1/2[Fc2] for 15 is identical to that
of its parent 5, indicating that perturbation of the -Ct
C- does not affect communication within the Fc2 core,
the mutual exclusivity noted above. Furthermore, E1/2
for the oxidation of the Co2(CO)4dppm unit in 15 is the
same as that in Fc[C2Co2(CO)4dppm]C2Fc,9a illustrating
that the Coulombic effect of two positively charged Fc
The bis-substituted complexes 17 and 20 gave the
opportunity to investigate the interaction between two
Co2(CO)4dppm units modulated by an FcN core. Assign-
ments for these complexes are given in Scheme 7. For
the Fc2 complex 17 the core Fc2 and Co2(CO)4dppm
potentials are little changed from 15, and extensive
communication is retained between the Co2(CO)4dppm
clusters (∆E1/2 ) 240 mV). As a consequence, four one-
electron waves are seen in the voltammetry of 17
(Figure 6). Those associated with the Fc2 core are
chemically reversible, whereas the cluster oxidation
processes are reversible at fast scan rates (>1 V s-1).
In contrast, 20 displays three reversible one-electron
responses due to the Fc3 core and a two-electron
response in the cyclic voltammogram (Figure 7); a small
resolution (∆E ≈ 40mV) of the two-electron response is
just discernible in the square wave voltammogram. The
two-electron Co2(CO)4dppm oxidation process is chemi-
cally reversible only at low temperatures and fast scan
rates. It is interesting that for these bis complexes the
collapse of the linear alkyne link by coordination of two
Sch em e 6

5046 Organometallics, Vol. 19, No. 24, 2000
Hore et al.
of the terminal groups, the Fc3 core did reduce the
interaction between the pendant redox centers. To
design a molecular array with predetermined tunable
electron transfer characteristics, it will be necessary to
have a better description of what is meant by “com-
munication” in these systems. That is, a description of
the orbital energies for the various mixed valence
species in an oxidation sequence. However, at the
present time we have no theoretical model that can
predict the subtle trends that have been noted in this
work. It is doubtful that the ∆E parameters derived for
the Fc2 and Fc3 molecules give other than a crude
qualitative measure of “communication”.
Exp er im en ta l Section
F igu r e 6. Cyclic voltammogram of 17 in CH2Cl2, Pt, 0.1
nBu4NPF6, 100 mV s-1
.
All reactions were carried out under anerobic conditions
using distilled and dried solvents. Transfers involving lithium
reagents were carried out using a cannula. Fc2I2 (1) was
prepared by the literature method.12 Microanalyses were
carried out by the Campbell Microanalytical Laboratory,
University of Otago. The FAB mass spectra were recorded on
a Kratos MS80RFA instrument with an Iontech ZN11NF atom
gun. Solvents were dried and distilled by standard procedures.
Infrared spectra were recorded on a Digilab FX60 spectrom-
1
eter. H NMR were recorded on a Gemini 200 MHz spectrom-
eter and the 31P NMR on a Varian VXR 300 MHz spectrometer.
Voltametric responses were recorded in dichloromethane on
an EG & G PAR 273 electrochemical analyzer interfaced to a
microcomputer using PAR M270 software with the solution
in a three-electrode cell. Potentials were referenced with
decamethylferrocene as an internal reference (ferrocene ) 0.55
V in CH2Cl2).
F igu r e 7. Cyclic voltammogram and square wave volta-
mmogram (60 Hz, 2 mV) of 20 in CH2Cl2, Pt, 0.1 nBu4NPF6,
200 mV s-1
.
Co2(CO)4dppm units reduces communication both within
the FcN core and between the two Co2(CO)4dppm redox
centers.
P r ep a r a tion of 3, 5, a n d 6. To a stirred solution of 1 (293
mg, (0.47 mmol) in diisopropylamine (25 mL) at 20 °C was
added trimethylsilylacetylene (0.13 mL, 0.92 mmol) followed
by 2 mol % Cu(OAc)2/(PPh3)2PdCl2. The reaction mixture was
refluxed for 30 min and solvent stripped in vacuo. The products
were separated on preparative silica gel plates in hexane/
dichloromethane (4:1) to give 3 (band 2) as a dark red solid
(83 mg, 25%). Anal. Calcd for C25H25Fe2ISi: C, 50.71; H, 4.26.
Despite the apparent longevity of the oxidized species
of 6-20 at the electrode surface on the electrochemical
time scale, no mixed valence species could be isolated.
Controlled electrochemical oxidation in an OTTLE cell
and chemical oxidation produced color changes and
shifts in the UV/visible/IR spectra, but the species
produced were very unstable. Bulk oxidation invariably
led to decomposition.
1
Found: C, 50.88; H, 4.22. Mass spectrum: m/e 592 (M+). H
NMR (CDCl3) δ: 0.25 (s, 9H, -SiMe3), 3.9-4.4 (m, 16H, Fc2-
H). 3 (117 mg, 0.20 mmol) was dissolved in dry tetrahydrofu-
ran and cooled to -78 °C. n-Butyllithium (0.2 mL) was then
added and the mixture stirred for 1 h, after which it was
warmed to room temperature. Water was added and the
mixture was left to stir for a further 10 min. The reaction
mixture was extracted into hexane and recrystallized from
hexane/dichloromethane to give 72 mg (75%) of 5 as dark
orange microcrystals. Anal. Calcd for C25H26Fe2Si: C, 64.4; H,
5.62. Found: C, 64.09; H, 5.57. Mass spectrum: m/e 466 (M+).
1H NMR (CDCl3) δ: 0.24 (s, 9H, -SiMe3), 4.0-4.4 (m, 17H,
Fc2-H). Trimethylsilyl acetylene (0.17 mL) was coupled to 3
(372 mg, 0.6 mmol) using 2 mol % Cu(OAc)2/(PPh3)2PdCl2. The
reaction mixture was separated on SiO2 plates in hexane/
dichloromethane (4:1) to yield 76 mg (23%) of 6 (this compound
Con clu sion
These studies show that ethynylpolyferrocenyl units
can function as effective modular links in the construc-
tion of molecular switches and arrays. The controlled
synthesis of the diiodopolyferrocenes provides synthons
which, via standard coupling methods, can be used to
incorporate the polyferrocene core into linear or poly-
hedral arrays. An increase in the lability of the mol-
ecules as the polyferrocene backbone expands does limit
this route. For reasons that are not understood this was
especially evident when ferrocenyl groups occupied the
terminal positions and Fc3-ethynylferrocene compounds
could not be characterized. Nevertheless, it has been
found that a series of molecules in which the biferrocene
is linked to fluorophores have reasonable stability.23 The
observation that each redox center, whether a ferrocenyl
terminus, a Co2(CO)4dppm unit, or the polyferrocenyl
core, can act almost independently of each other while
still being subject to inductive influences is significant.
While the Fc2 core did not modulate the redox chemistry
has been prepared directly from 1).11 Anal. Calcd for C30H34
-
Fe2Si2: C, 64.06; H, 6.09. Found: C, 63.89; H, 6.21. Mass
spectrum (FAB): m/e 562 (M+). 1H NMR (CDCl3) δ: 0.23 (s,
18H, -SiMe3), 4.01 (4H, t, Fc âH), 4.18 (4H, t, Fc RH), 4.21
(4H, t, Fc â′H), 4.37 (4H, t, Fc R′H).
P r ep a r a tion of 4 a n d 7. A stirred solution of 1 (293 mg,
0.23 mmol), ethynyl ferrocene (120 mg, 0.57 mmol), and 2 mol
% Cu(OAc)2/(PPh3)2PdCl2 in diisopropylamine (25 mL) was
refluxed for 15 min. The solvent was stripped in vacuo. The
products were separated on preparative silica gel plates in
hexane/dichloromethane (4:1). Bands 1 and 2 gave unreacted
1 and diferrocenyl butadiyne, respectively. Band 3 gave 4;
FAB-MS: m/e 704 (M+). Band 4 was identified as 7, a dark
(23) Duffy, N. W. Ph.D. Thesis, University of Otago, 1997; work in
progress.

Communication between Co2(CO)4dppm Units
Organometallics, Vol. 19, No. 24, 2000 5047
Ta ble 3. Cr ysta l Da ta a n d Str u ctu r e Refin em en t
for 17
red solid, (23 mg, 13%). Anal. Calcd for C44H34Fe4: C, 67.23;
H, 4.33. Found: C, 66.89; H, 4.21. Mass spectrum (EI): m/e
786 (M+). 1H NMR (CDCl3) δ: 4.06 (4H, t, Fc âH), 4.20 (4H, t,
Fc â′H), 4.22 (4H, t, Fc RH), 4.23 (s, 10H, C5H5), 4.27 (4H, t,
Fc â′′H), 4.43 (4H, t, Fc R′H), 4.44 (4H, t, Fc R′′H).
empirical formula
molecular formula
fw
C96H94Co4Fe2O12P4Si2
C88H78O8Si2P4Fe2Co4, 2(C4H8O2)
1967.19
cryst syst
monoclinic
P r ep a r a tion of 2. To ferrocene (5.6 g) in hexane (170 mL)
was added n-BuLi (40 mL) and TMEDA (10 mL). Iodine (7 g)
in THF (80 mL) was then added with vigorous stirring to the
solution containing the precipitated dilithioferrocene-TMEDA
complex, and the reaction mixture was stirred for 1 h, then
refluxed for 30 min. The mixture was extracted into dichlo-
romethane, and the organic layer washed with water and then
dried over magnesium sulfate. Separation of the iodo polyfer-
rocene derivatives was achieved by column chromatography
on alumina. Hexane eluent was progressively enriched with
CH2Cl2 to remove mono- and biferrocenes. The desired product
was eluted next by ethyl acetate to give 2 as an orange powder
(2.05 g, 24%). Anal. Calcd for C30H24Fe3I2: C, 44.71; H, 3.00.
Found: C, 44.88; H, 2.94. FAB-MS: m/e 806 (M+), 680 (M -
space group
abs coeff
P2(1)/c
1.167 mm-1
final R indices [I>2σ(I)]
R1 ) 0.0295, wR2 ) 0.0796
R1 ) 0.0361, wR2 ) 0.0871
1.131
203(2) K
0.71073 Å
(all data)
goodness-of-fit on F2
temperature
wavelength
unit cell dimens
a ) 11.9329(2) Å
b ) 25.7059(6) Å
c ) 15.7314(2) Å
â ) 107.71(1)°
4596.72(14) Å3
2
volume
Z
density (calcd)
F(000)
1.421 Mg/m3
2028
1
I+). H NMR (CDCl3) δ: 3.91, 4.06, 4.10 {3 × (t, 4H, Fc3-H)},
no. of reflns collected
no. of ind reflns
10 324
10 324 [R(int) ) 0.0000]
4.15 (m, 12H, Fc3-H).
P r ep a r a tion of 8 a n d 10. To a stirred solution of 2 (480
mg, 0.6 mmol) in diisopropylamine (50 mL) was added tri-
methylsilylacetylene (0.34 mL) followed by 2 mol % Cu(OAc)2/
(PPh3)2PdCl2, and the solution was refluxed for 3 h. The solvent
was stripped in vacuo and the residue separated on prepara-
tive silica gel plates, eluting with hexane/dichloromethane (10:
1), to give 8 (50 mg, 11%) as a very unstable brown solid. FAB-
MS: m/e 776 (M+). δH (CDCl3): 0.23 (s, 9H, -SiMe3), 3.9-4.2
(m, 24H, Fc3-H)}; the compound gave a positive Beilstein test.
Recrystallization of the next band from dichloromethane/
hexane gave 10 as needlelike orange crystals (100 mg, 22%).
Anal. Calcd for C40H42Fe3Si2: C, 64.35; H, 5.67. Found: C,
64.15; H, 5.50. FAB-MS: m/e 746 (M+). 1H NMR (CDCl3) δ:
0.22 (s, 18H, -SiMe3); 3.96, 4.05, 4.13 {3 × (t, 4H, Fc3-H)},
4.17 (m, 12H, Fc3-H).
1
3CO...M+ - 6CO). H NMR (CDCl3) δ: 0.43 (s, 9H, -SiMe3),
3.9-4.3 (m, 25H, Fc3-H). IR νCO (hexane): 2083, 2045, 2021,
2013, 2002 cm-1. 14: Compound 10 (100 mg, 0.15 mmol) and
dicobalt octacarbonyl (206 mg, 0.6 mmol) gave 78 mg (44%) of
green-black crystalline 14. Anal. Calcd for C52H42Co4Fe3O12
-
Si2: C, 47.37; H, 3.21. Found: C, 47.85; H, 3.32. FAB-MS: m/e
1318 (M+), sequence 1234-982 (M+ - 3CO...M+ - 12CO). H
1
NMR (CDCl3) δ: 0.43 (s, 18H, -SiMe3), 4.0-4.3 (m, 24H, Fc3-
H). IR νCO (hexane): 2082, 2045, 2021, 2013, 2002 cm-1
.
P r ep a r a tion of 15-20. A solution of 12 (137 mg, 0.12
mmol) and excess bis(diphenylphosphino)methane in benzene
(25 mL) was heated under reflux for 20 min. The solvent was
stripped, and the products were separated on preparative silica
gel plates, eluting with ether/hexane (1:1). Removal of the first
red band gave 16, 83 mg as a red-brown solid. FAB-MS: m/e
1176 (M+). 1H NMR (CDCl3) δ: 0.22 (s, 9H, -CtC-SiMe3),
0.56 (s, 9H, -cluster-SiMe3), 3.2-3.6 (m, 2H, -CH2-), 3.9-
4.4 (m, 16H, Fc2-H), 6.9-7.3 (m, 20H, Ph-H). δP (CDCl3): 34
(s). IR νCO (hexane): 2017, 1991, 1965, 1950 cm-1; νCtC: 2148
cm-1. From the same reaction removal of the second red band
from plate chromatography gave 99 mg of red-brown crystal-
line 17. Anal. Calcd for C88H78Co4Fe2O8P4Si2: C, 59.01; H, 4.39.
Found: C, 58.81; H, 4.72. 1H NMR (CDCl3) δ: 0.56 (s, 18H,
-SiMe3), 3.2-3.6 (m, 2H, -CH2-), 3.9-4.4 (m, 16H, Fc2-H),
6.9-7.3 (m, 40H, Ph-H). δP (CDCl3): 34 (s). IR νCO (hexane):
P r ep a r a tion of 9. The iodide 8 (80 mg) was dissolved in
THF and cooled to -78 °C, n-BuLi (hexane, 0.5 mL) was added,
and the reaction mixture was stirred for 1 h before being
brought to room temperature. Water was added and the
reaction mixture stirred for a further 15 min. The organic layer
was collected and the solvent stripped in vacuo. The residue
was dissolved in hexane, washed with water, dried over
magnesium sulfate, and recrystallized from dichloromethane/
hexane to give dark red 9 (41 mg). Anal. Calcd for C35H34Fe3-
Si: C, 64.64; H, 5.27. Found: C, 64.66; H, 5.40. FAB-MS: m/e
1
650 (M+). H NMR (CDCl3) δ: 0.22 (s, 9H, -SiMe3), 3.9-4.3
2018, 1992, 1965, 1950, 1926 cm-1
By a similar procedure. 15: From 11, with ethyl acetate/
hexane (1:5), gave red-brown 15 (37%). Anal. Calcd For C54H48
.
(m, 25H, -Fc3H).
P r ep a r a tion of 11-14. Dicobaltoctacarbonyl (510 mg, 1.5
mmol) was added to 5 (443 mg, 0.95 mmol) in hexane/
dichloromethane (1:1) and the reaction mixture stirred at
ambient temperature for 2 h. The solvent was removed and
the resulting mixture separated by chromatography on silica
gel preparative plates in hexane/dichloromethane (5:1). The
green band was removed and crystallized from hexane to give
-
Co2Fe2O4P2Si: C, 60.02; H, 4.48. Found: C, 60.29; H, 4.86. 1H
NMR (CDCl3) δ: 0.22 (s, 9H, -CtC-SiMe3), 3.2-3.6 (m, 2H,
-CH2-), 3.9-4.4 (m, 16H, Fc2-H), 6.9-7.3 (m, 20H, Ph-H). δP
(CDCl3): 34 (s). 18: From 13 (104 mg, 0.11 mmol), with ethyl
acetate/hexane (2:5), gave after recrystallization from ethyl
acetate, 53 mg (38%) of red-brown crystals of 18. Anal. Calcd
for C64H56Co2Fe3O4P2Si: C, 60.78; H, 4.46. Found: C, 60.04;
H, 4.55. FAB-MS: m/e 1264 (M+). 1H NMR (CDCl3) δ: 0.55
(s, 9H, -SiMe3), 3.2-3.6 (m, 2H, -CH2-), 3.8-4.3 (m, 25H,
Fc3-H), 6.9-7.3 (m, 20H, Ph-H). δP (CDCl3): 34 (s). IR νCO
(hexane): 2017, 2013, 1991, 1980, 1965, 1950 cm-1. 19,20:
From 14, eluting with ether/hexane (1:1), two major product
bands were obtained. That with the highest Rf gave red-brown
green-black crystals of 11 (102 mg 14%). Anal. Calcd for C31H26
-
Co2Fe2O6Si: C, 49.5; H, 3.48. Found: C,49.56; H, 3.35. 1H
NMR (CDCl3) δ: 0.43 (s, 9H, -SiMe3), 3.9-4.4 (m, 17H, Fc2-
H). IR νCO (hexane): 2083, 2046, 2021, 2013, 2002 cm-1
.
The following complexes were prepared by a similar proce-
dure. 12: Compound 6 (224 mg, 0.4 mmol) and an excess of
dicobalt octacarbonyl gave 181 mg (40%) of green crystalline
12. Anal. Calcd for C42H34Co4Fe2O12Si2: C, 44.47; H, 3.02.
Found: C, 44.48; H, 3.30. FAB-MS: m/e 1134 (M+), sequence
1050-798 (M+ - 3CO...M+ - 12CO). 1H NMR (CDCl3) δ: 0.44
(s, 18H, -SiMe3), 4.0-4.4 (m, 16H, Fc2-H). IR νCO (hexane):
2083, 2046, 2022, 2014, 2003 cm-1. 13: Compound 9 (153 mg,
0.235 mmol) and dicobalt octacarbonyl (200 mg, 0.59 mmol)
gave 72 mg (33%) of green-black crystalline 13. Anal. Calcd
for C41H34Co2Fe3O6Si: C, 52.60; H, 3.66. Found: C, 52.83; H,
19 (21%). FAB-MS: m/e 1360 (M+). H NMR (CDCl3) δ: 0.21
1
(s, 9H, -CtC-SiMe3), 0.55 (s, 9H, -cluster-SiMe3), 3.2-3.6
(m, 2H,-CH2-), 3.8-4.2 (m, 24H, Fc3-H), 6.9-7.3 (m, 40H, Ph-
H). IR νCO (hexane): 2017, 1991, 1965, 1950 cm-1; νCtC: 2149
cm-1. Removal of the second band gave red-brown crystalline
20 (41%). Anal. Calcd for C98H86Co4Fe3O8P4Si2: C, 59.59; H,
4.39. Found: C, 59.33; H, 4.39. FB-MS: m/e 1974 (M+). 1H
NMR (CDCl3) δ: 0.56 (s, 18H, -SiMe3), 3.2-3.6 (m, 4H,
3.63. FAB-MS: m/e 936 (M+), sequence 852-600 (M+
-

5048 Organometallics, Vol. 19, No. 24, 2000
Hore et al.
-CH2-), 3.8-4.2 (m, 24H, Fc3-H), 6.9-7.3 (m, 40H, Ph-H).
moiety; refinement therefore involved only one-half of the
molecular unit. A difference Fourier synthesis, following the
location of all of the non-hydrogen atoms for 17, revealed six
substantial high peaks that could be readily assigned to ethyl
acetate solvate in the crystal lattice. Anisotropic refinement
of the solvate, with isotropic H atoms included as fixed
contributions to Fc, led to a significant reduction in R1 and
the absence of significant residual electron density in subse-
quent difference Fourier maps.
IR νCO (hexane): 2017, 1991, 1965, 1951 cm-1
.
X-r a y Da ta Collection , Red u ction , a n d Str u ctu r e So-
lu tion for 17. Crystal data for 17 are given in Table 3.
Recrystallization of 17 from ethyl acetate yielded red-brown
blocks, one of which was used for data collection. Data were
collected at 203(2) K on a Bruker SMART CCD diffractometer,
processed using SAINT24 with empirical absorption corrections
applied using SADABS25.
The structure was solved using SHELXS26 and refined by
full-matrix least-squares using SHELXL-97.26 Non-hydrogen
atoms were assigned anisotropic temperature factors, and the
H atoms were included in calculated positions. Molecules of
17 lie on a center of symmetry located at the midpoint of the
C-C bond of the fulvalenide ring system of the biferrocene
Ack n ow led gm en t. We thank Prof. W. T. Robinson
(University of Canterbury) for the collection of X-ray
data, the Uniiversity of Otago and the Royal Society
(NZ) Marsden Fund for financial support, and Robinson
College, Cambridge, for a Bye Fellowship (B.H.R.).
(24) SMART (Control) and SAINT (Integration) software; Bruker,
AXS: Madison WI, 1994.
(25) SADABS (correction for area detector data); Bruker AXS:
Madison WI, 1995 and 1997.
(26) Sheldrick, G. M. SHELXS-96; University of Go¨ttingen: Ger-
many, 1996. Sheldrick, G. M. SHELXL-97; University of Go¨ttingen:
Germany, 1997.
Su p p or tin g In for m a tion Ava ila ble: X-ray data for 17
are available free of charge via the Internet at http://
pubs.acs.org.
OM000493G