Ferrocenyl-polyphenylenes: Toward metallo-organic polyaromatics

Article abstract of DOI:10.1021/om100862a

This work continues our investigations into substituted polyaromatic hydrocarbons (PAHs). Presented is a new series of ferrocenyl-substituted polyphenylenes and the successful Lewis acid-catalyzed cyclodehydrogenation of one of these to form an eight-ring

Full text of DOI:10.1021/om100862a

Organometallics 2010, 29, 6541–6547 6541
DOI: 10.1021/om100862a
Ferrocenyl-polyphenylenes: Toward Metallo-organic Polyaromatics
Dilwyn J. Roberts, Daniel J. Gregg, Chris M. Fitchett, and Sylvia M. Draper*
School of Chemistry, University of Dublin, Trinity College, D2, Ireland
Received September 6, 2010
This work continues our investigations into substituted polyaromatic hydrocarbons (PAHs).
Presented is a new series of ferrocenyl-substituted polyphenylenes and the successful Lewis acid-
catalyzed cyclodehydrogenation of one of these to form an eight-ring fused PAH supporting a
ferrocene moiety (16-ferrocenyl(tribenzo[e;g,h,i;k])perylene (6)), the largest of its kind. The failure to
generate further metallo-organic superaromatics is discussed in relation to the structure of the
precursors and the oxidation of the ferrocene metal center. The synthesis of the polyphenylenes
involves the [2þ4] Diels-Alder reaction of phenyl, ferrocenyl, and hydrogen-terminated ferrocenyl
acetylenes and tetraphenylcyclopentadienone. The full spectroscopic characterization and the
electronic and redox properties of all the systems are described and compared. The single-crystal
X-ray structure of tetraphenyl ferrocenyl benzene (1), the precursor to 6, is also discussed.
Introduction
transfer in glucose oxidase and anion, cation, and DNA
biosensors.9
Carbon-rich organometallic compounds represent a rapidly
growing subfield at the interface of organic and organometallic
chemistry. Carbon-rich ligands offer a range of interesting
properties including facile electron transfer to (and from) a
metal surface. As a result, organometallic fragments assembled
on substituted benzenes, polyaromatics, and heteroaromatics,
such as pyridine, have attracted considerable interest as poten-
tial models for organometallic polymers,1 as targets for the
study of electron-transfer processes, and in the generation of
molecular wires.2 Indeed in the past decade, the interest in
organometallic polymers has increased owing to their applica-
tion in light-emitting diodes,3 molecular magnets,4 gas sensors,5
and photovoltaic devices.6
Among these systems, ferrocene has found many applica-
tions in synthetic organic chemistry and material sciences.7
The stable coordination of the cyclopentadienyl ligands, the
synthetic ease of functionalization, the possibility of intro-
ducing chirality, and the reversible redox behavior of the
metal center are attractive features of ferrocene-based
molecules.8 Examples include ferrocene-mediated electron
Ferrocenylacetylenes in particular offer a wide range of
potential applications in NLO,10 in catalysis, and as models
of charge transfer, and they are key synthetic building blocks
for the generation of more complex molecules with well-
defined magnetic11 and electronic properties.12 Most ferro-
cene complexes that are investigated for their optical non-
linearities comprise ferrocene as donor and -CN- or
-CHO-substituted benzene, stilbene, or azobenzene deriva-
tives, as acceptor.13
The spacing between the cyclopentadienyl rings in a
˚
ferrocenyl moiety is about 3.3 A, geometrically similar to
the interplanar spacing in π-π stacking arrangements.11 In
this regard, ferrocene is particularly suited for incorporation
into all-carbon polycyclic and N-doped polycyclic mole-
cules. Polyaromatic hydrocarbons (PAHs) have a number
of attributes that could make them useful as linking groups in
donor-acceptor organometallic systems. They are planar
and relatively rigid with a fixed geometry. Their highly
delocalized π-systems with small HOMO-LUMO energy
separations and low-energy vacant π*-orbitals are likely to
facilitate electronic communication between donors and
acceptors. Surprisingly, although large polycyclic materials
incorporating up to 222 carbon atoms have been character-
ized, very few include an organometallic moiety. Known
examples include Bunz’s ethynylated cyclobutadienes,14-17
*To whom correspondence should be addressed. E-mail: smdraper@
tcd.ie.
(1) Manners, I. Angew. Chem., Int. Ed. Engl. 1996, 35, 1603.
(2) Ribou, A. C.; Launay, J. P.; Sachtleben, M. L.; Li, H.; Spangler,
C. W. Inorg. Chem. 1996, 35, 3735.
(3) Koga, Y.; Ueno, K.; Matsubara, K. J. Polym. Sci., Part A: Polym.
Chem. 2006, 44, 4204.
(4) Clendenning, S. B.; Fournier-Bidoz, S.; Pietrangelo, A.; Yang,
G. C.; Han, S. J.; Brodersen, P. M.; Yip, C. M.; Lu, Z. H.; Ozin, G. A.;
Manners, I. J. Mat. Chem. 2004, 14, 1686.
(5) Penza, A.; Cassano, G.; Sergi, A.; Lo Sterzo, C.; Russo, M. Sens.
Actuators, B 2001, 81, 88.
(6) Cyr, P. W.; Klem, E. J. D.; Sargent, E. H.; Manners, I. Chem.
Mater. 2005, 17, 5770.
(7) Wilkinson, G.; Rosenblum, M.; Whiting, M. C.; Woodward, R. B.
(9) Sehnert, J.; Hess, A.; Metzler-Nolte, N. J. Organomet. Chem.
2001, 637, 349.
(10) Long, N. J. Angew. Chem., Int. Ed. Engl. 1995, 34, 21.
(11) McKinney, J. D.; Anderson, P. A.; Hamor, T. A.; Jones, C. J.;
Paxton, K.; Porch, A. J. Organomet. Chem. 1998, 558, 147.
(12) Stepnicka, P.; Trojan, L.; Kubista, J.; Ludvik, J. J. Organomet.
Chem. 2001, 637, 291.
(13) Calabrese, J. C.; Cheng, L. T.; Green, J. C.; Marder, S. R.; Tam,
J. Am. Chem. Soc. 1952, 74, 2125.
(8) Molina, P.; Tarraga, A; Curiel, D.; Velasco, M. D. J. Organomet.
Chem. 2001, 637, 258.
W. J. Am. Chem. Soc. 1991, 113, 7227.
(14) Bunz, U. H. F.; Wiegelmann Kreiter, J. E. C. Chem. Ber. 1996,
129, 785.
r
2010 American Chemical Society
Published on Web 11/04/2010
pubs.acs.org/Organometallics

6542 Organometallics, Vol. 29, No. 23, 2010
Roberts et al.
Scheme 1. Scheme for the Synthesis of Polyphenylenes 1, 2, 3, 4, and 5 (All Synthesized by the Diels-Alder Reaction between
Tetraphenylcyclopentadienone and the Appropriate Acetylene) and the Ferrocenyl-Supported PAH 6a
a (i)-(iv) Tetraphenylcyclopentadienone, benzophenone, 200 °C, 48 h; (i) 1 70%; (ii) 2 36%; (iii) 3 44%, 4 18%; (iv) 5 75%; (v) AlCl3/CuCl2, CS2, RT,
48 h, 6 15%.
the incorporation of tricarbonylchromium into an HBC
derivative,18 and an L-shaped 4-(2-ferrocenylvinyl)pyrene
compound.11
in common chlorinated and aromatic solvents, but mostly
insoluble in hexane. The orange to red color is a result of an
absorption band at around 440 nm and is characteristic of
ferrocene compounds. All the complexes were studied spec-
troscopically and characterized by 1H and 13C NMR and IR
spectroscopy and accurate mass and ESI-mass spectrometry.
Compounds 1, 2, and 5 were synthesized by the [2þ4]-
cycloaddition reaction of ethynylferrocene,21 phenylethynyl-
ferrocene,22 and 1,4-bis(ferrocenyl)butadiyne,24 respectively,
with commercially available tetraphenylcyclopentadienone
(tetracyclone) in a benzophenone melt at 200 °C. After work-
up, column chromatography isolated polyphenylene 1 as an
air-stable, orange-red, crystalline solid in 70% yield. 2 was
isolated in a similar manner as an orange-yellow, air-stable
solid in 36% yield. The NMR and IR spectroscopic data as
well as mass spectrometry of 2 concurred with those already
reported.19
Polyphenylenes 3 and 4 were synthesized by the Diels-
Alder reaction of 1,10-bis(phenylethynyl)ferrocene with 1.2
and 2.2 equiv of tetracyclone, respectively. 1,10-Bis(phenyl-
ethynyl)ferrocene was synthesized by the Stille coupling of
1,10-iodoferrocene with 2 equiv of phenylethynyltrimethyl-
stannane.23 Compound 3 was isolated as the mono-Diels-
Alder adduct, whereas 4 was isolated as the diadduct. Com-
pounds 3 and 4 were isolated as orange, crystalline solids in
44% and 18% yield, respectively. Compound 5 was isolated
as a red, air-stable solid in 75% yield.
To broaden the number of molecules of this type in the
literature, we devised the synthetic strategy presented in this
work. This provides a route to incorporating ferrocenyl moi-
eties into polyphenylenes, by introducing the metal complex at
an early synthetic stage. Despite the need for it to survive the
reaction conditions of all subsequent steps, we felt confident
that the good chemical stability and attractive materials proper-
ties of ferrocene offered a reasonable chance of success.
Interest in the fluxional processes of sterically demanding
chiral propeller compounds of the type CnArn has given rise
to the limited examples of monoferrocenyl polyphenylene
compounds in the literature to date, e.g., sterically crowded
C6Ph5Fc (2), shown in Scheme 1.19 It is, however, important
to expand the nature and type of such systems and to include
electrochemical and spectroscopic evaluations of the result-
ing molecules, as illustrated by the work of Rathcore et al. in
their investigations of hexakis(4-ferrocenylphenyl)benzene.20
Results and Discussion
In general, the new ferrocenyl compounds synthesized in
this work are yellow/red, crystalline, air-stable solids, soluble
(15) Laskoski, M.; Steffen, W.; Smith, M. D.; Bunz, U. H. F. Chem.
Commun. 2001, 691.
(16) Laskoski, M.; Roidl, G.; Smith, M. D.; Bunz, U. H. F. Angew.
The novel polyphenylenes 3, 4, and 5 were fully characterized
by NMR and IR spectroscopy as well as mass spectrometry.
Chem., Int. Ed. 2001, 40, 1460.
(17) Laskoski, M.; Morton, J. G. M.; Smith, M. D.; Bunz, U. H. F.
J. Organomet. Chem. 2002, 652, 21.
(18) Herwig, P. T.; Enkelmann, V.; Schmelz, O.; Mullen, K. Chem.;
(21) Rosenblum, M; Brawn, N.; Papenmeier, J.; Applebaum, M.
J. Organomet. Chem. 1966, 6, 173.
(22) Fu, N.; Zhang, Y. R.; Yang, D.; Chen, B. H.; Wu, X. L. Catal.
Eur. J. 2000, 6, 1834.
(19) Gupta, H. K.; Brydges, S.; McGlinchey, M. J. Organomettallics
1999, 18, 115.
(20) Chebney, V. J.; Dhar, D.; Lindeman, S. V.; Rathore, R. Org.
Commun. 2008, 9, 976.
(23) Ingham, S. L.; Khan, M. S.; Lewis, J.; Long, N. J.; Raithby, P. R.
J. Organomet. Chem. 1994, 470, 153.
Lett. 2006, 8, 5041.
(24) Hay, A. S. J. Org. Chem. 1962, 27, 3320.

Article
Organometallics, Vol. 29, No. 23, 2010 6543
Full details of the characterization of all the compounds can
be seen in the Experimental Section.
Cyclodehydrogenation. Given our interest in synthesizing
ferrocenyl-supported PAHs, the cyclodehydrogenation of
polyphenylenes 1, 2, 3, 4, and 5 was attempted; however only
1 underwent successful cyclodehydrogenation using AlCl3
and CuCl2. The cyclodehydrogenation of 1 yielded 6 (Scheme 1)
in 15% yield as a pale yellow solid. As a step toward an organo-
metallic model of an intercalated graphitic sheet, 6 was expected
to possess interesting geometric and electronic properties.
Under the conditions for cyclodehydrogenation 1 also
produced insoluble and intractable material. This accounts
for the low yield of 6 and is believed to be a consequence of
the partial oxidative decomposition of the ferrocene nuclei
by CuII salts.15 Evidence for the successful cyclodehydro-
genation came from the large aromatic shifts in the 1H NMR
spectrum of 6 on comparison with those of the ferrocenyl
polyphenylene 1. The aromatic proton signals in the δ 6.6-
7.2 range, for 1, moved downfield to δ 7.2-9.8 in 6, as a
direct result of the increase in ring-current-induced deshield-
ing by the PAH.
Figure 1. The X-ray crystal structures of ferrocenyl polypheny-
lene 1.
Attempts to cyclodehydrogenate ferrocenyl polypheny-
lenes 2, 3, 4, and 5 to form the planar, ring-closed system were
unsuccessful, yielding only black, insoluble, and uncharac-
terizable solids, indicating the complete decomposition of
the polyphenylenes.
Table 1. UV-Vis Data for Ferrocene and 1, 2, 3, 4, 5, and 6 in
Toluene
Single-Crystal X-ray Analysis. Room-temperature, single-
crystal X-ray diffraction of 1 was carried out to determine the
solid-state molecular arrangement. Orange crystals of 1 were
obtained, in air, from the slow diffusion of hexane into a
dichloromethane solution of 1.
All investigated crystals of 1 proved to be heavily twinned.
The sample finally chosen consisted of at least five domains,
four of which could be investigated. The structure was solved
and refined for two domains, with an R1 of 4.79%. It
crystallized in space group P1, and its molecular structure is
shown in Figure 1. The propeller-like molecular structure
shows twisting of the phenyl and ferrocene subunits with
respect to the central phenyl ring, moving clockwise around
the central ring, and starting at C1 the twist of the ferrocene
moiety is 36° out of the plane formed by the central ring.
The other phenyls rings are tilted by 56°, 58°, 70°, and 72°,
respectively.
λ
max, nm
(ε, L mol-1 cm-1 ꢀ 103)
327.0 (0.5)
344.0 (2.0)
350.5 (2.5)
358.0 (3.1)
low-energy
d-d transition
compound
Fc
1
2
3
4
438.5 (0.5)
447.5 (0.5)
464.5 (0.5)
460.0 (0.7)
464.0 (1.0)
453.0 (1.0)
ca. 440
300.0 (19.6), 359 (sh) (3.7)
324.0 (11.0), 359.0 (sh) (3.5)
302.5 (10.0), 312.0 (14.0),
355.0 (4.0), 371.5 (4.0), 405.0 (0.5)
5
6
thus assigned as the 1E1g r 1A1g ligand field (LF) transition
(“d-d” transition) and the higher energy band around 325 nm
as the E2g r A1g LF transition. Organometallics are highly
covalent, so these metal-centered d orbitals actually have con-
siderable ligand character and are sensitive to ligand binding.27
Substitution onto a cyclopentadienyl ring particularly by con-
jugated acceptor groups is generally followed by drastic changes
in the UV-vis absorption spectra.
The electronic absorption spectra of 1, 2, 3, 4, 5, and 6 were
recorded in toluene, and the data are presented in Table 1.
Excluding 6, each of the complexes displays two or three
prominent absorption bands in the 300-600 nm region.
Similar to other ferrocenyl species, they exhibit a weak
low-energy band attributable to the d-d transition arising
from the ferrocene moiety.27,28 These broad, low-energy
features account for the orange to red colors of these
compounds. The two weak absorption bands of the ferrocene
unit (λ = 327.0 and 438.5 nm) are magnified, in some instances,
indicating perturbation of the ferrocene frontier orbitals upon
substitution.
1
1
This is unusual but similar to the structure reported19 for
2, where each peripheral ring is displaced slightly more than
its preceding neighbor (from 51° for ferrocenyl ring to 120°).
More common is the regular propeller configuration of
C6Ph6-type molecules with uniform twists on the order of
67° between the outer and central phenyl rings.25
Molecules of 1 are offset stacked and head-to-tail to avoid
steric interactions between the ferrocenyl units of neighbor-
ing molecules. The ferrocene moiety remains unperturbed by
the presence of the polyphenylene, and the cyclopentadienyl
rings of the ferrocene are 3° off planarity.
UV-Vis Absorption Spectroscopy. The electronic absorp-
tion spectra of ferrocene and its substituted derivatives have
revealed that the metal-centered dz2 orbital is generally the
HOMO, and the combination of the dxz and dyz is the
LUMO.13,26 The absorption around 440 nm in ferrocene is
In ferrocene there is a very weak feature at around
438.5 nm. The ferrocene derivatives in this study also display
this typical ferrocenyl transition in the visible region; how-
ever they are shifted to the 440 to 465 nm region, again
(25) Bart, J. C. J. Acta Crystallogr., Sect. B 1968, 24, 1277.
(26) Jayaprakash, K. N.; Ray, P. C.; Matsuoka, I.; Bhadbhade,
M. M.; Puranik, V. G.; Das, P. K.; Nishihara, H.; Sarkar, A. Organo-
metallics 1999, 18, 3851.
(27) Thomas, K. R. J.; Lin, J. T. J. Organomet. Chem. 2001, 637, 139.
(28) Sohn, Y. S.; Hendrick., D.; Gray, H. B. J. Am. Chem. Soc. 1971,
93, 3603.

6544 Organometallics, Vol. 29, No. 23, 2010
Roberts et al.
Table 2. UV-Vis Data of 6 and Dodecyl-Substituted
Tribenzoperylene29
compound
λmax (nm)
6a
312
315
355
365
372
384
405
418
tribenzoperyleneb
a Toluene. b Chloroform.
Figure 2. UV-vis absorption spectrum of 6 in toluene.
suggesting frontier orbital interaction between donor and
acceptor.
In the UV-vis spectrum of 1 there is a peak at 344 nm and
an additional broad feature at 447.5 nm. There are similar
features in the spectrum of 2 at 350.5 and 464.5 nm and in the
spectrum of 3 at 358 and 460 nm. The spectrum of 4 has a
shoulder at 359 nm, a broad absorption at 464.0 nm, and an
additional shoulder at 300 nm. This is mirrored in the spec-
trum of 5, which also has a shoulder at 359.0 nm, a broad
absorption at 453 nm, and an additional band at 324 nm. 1, 2,
3, 4, and 5 show similar electronic properties but smaller
HOMO-LUMO energy gaps than ferrocene, as reflected in
the lower energy of their transitions.
An interesting question concerns how the electronic prop-
erties of the ferrocencyl moiety in 1 are effected by ring
closure of the polyphenylene. The UV-vis spectrum of ferro-
cenyl PAH derivative 6 is very different from polyphenylene 1,
with four intense features in the 300-425 nm region
(Figure 2). In the UV-vis spectrum of 6, the broad signature
of the ferrocene d-d transition almost disappeared, and
instead only a small shoulder at around 440 nm occurs. It
is likely that the weaker ferrocenyl band is hidden beneath
the broad 405 nm band, which is assigned to a πfπ*
transition. In addition, 6 is highly absorbing between 300
and 450 nm. In general, the spectrum of 6 is dominated by
πfπ* absorption bands with high absorption coefficients
(Table 1) and is reminiscent of spectra obtained for triben-
zoperylene.
Due to insolubility, the UV-vis spectra for PAHs are
often restricted to measurement in thin films (solid-state
characterization), unless they are substituted by large solu-
bilizing chains.29 In Table 2 the ground-state absorption
properties of 6 are compared to those of dodecyl-substituted
tribenzoperylene.
Compound 6 can be considered to be an L-shaped mole-
cule in which a ferrocene donor group is linked to a planar
polyaromatic moiety. It is well known that the solubility of
the simple PAH (tribenzoperylene, C30H16) fragment at-
tached to the ferrocene unit in 6 is extremely low. It is
necessary, in order to obtain soluble analogues of this
PAH, to attach long solubilizing chains on the periphery.29
The fact that its ferrocenyl analogue 6 is soluble in organic
solvents suggests there is an underlying factor preventing
Figure 3. Structures of ferrocenyl PAH 6 and tribenzoperylene.
aggregation. Aggregation of the PAH unit is caused by a
π-π stacking, which, in 6, could be prohibited by the bulky
ferrocenyl group. There appears to be little perturbation of
the PAH fluorophore in 6. The πfπ* transitions in the
UV-vis spectra of the alkyl-substituted PAH are at the same
energy as for the soluble ferrocene analogue. It is suggested,
therefore, that steric interactions between individual ferro-
cenyl molecules impact on the ability of the PAH subunit to
stack or form dimers. This has been seen in single-crystal
X-ray studies of ferrocenylnaphthalimide molecules,30 which
place the bulky ferrocenyl moieties on the periphery of a
columnar network.
Comparison of the spectral properties of 6 to tribenzoper-
ylene serves as further evidence for the success of the
cyclodehydrogenation reaction in forming the closed ring
system of 6. It shows that the πfπ* transitions of the PAH
unit dominate the absorption spectrum of 6. The UV-vis
spectra of 6 and long chain substituted tribenzoperylene
show extremely similar shape and λmax values. Both spectra
show three types (groups) of bands (R, β, p), which are typical
of aromatic hydrocarbons. The β-bands appear around
315 nm, and the p-bands around 370 nm. The R-band at
410 nm is an extremely low intensity absorption band.
Investigations into the oxidation of substituted ferrocenes
with air in acidic solutions have been carried out by numer-
ous workers.31 It has been reported that the rate of the
oxidation is influenced by the ability of a proton to coordi-
nate to the iron atom. A greater reactivity can be accounted
for by a specific, tilted conformation of the two cyclopenta-
dienyl rings.31 The same authors also report a correlation
between the wavelength of the low-energy band at around
450 nm (see Table 1), which is found to be sensitive to the
inclination of the cyclopentadienyl rings of ferrocene,31 and
the rate of oxidation in acidic solution. As an example,
[2]ferrocenophane, which possesses tilted cyclopentadienyl
rings, exhibits a longer wavelength absorption and facile
protonation of its iron center.31
(30) McAdam, C. J.; Robinson, B. H.; Simpson, J. Organomettallics
2000, 19, 3644.
(29) Dotz, F.; Brand, J. D; Ito, S.; Gherghel, L.; Mullen, K. J. Am.
Chem. Soc. 2000, 122, 7707.
(31) Okada, Y.; Nakagawa, T.; Hayashi, T. Inorg. Chim. Acta 2001,
312, 197.

Article
Organometallics, Vol. 29, No. 23, 2010 6545
Table 3. Electrochemical Data for Ferrocenyl Complexes 1, 2, 3,
4, 5, and 6a
of the resulting [Fc/Fcþ] couple is dependent on the substitu-
tion of the ligand.10,32
Compared to the ferrocene standard, the oxidation po-
tentials of the ferrocenyl units in 1, 2, and 4 exhibit a slight
cathodic (less positive) shift (50-160 mV). The relatively low
value for the ferrocene oxidation potential indicates that
these ferrocenyl species are electron rich. The ease of oxida-
tion in the three related compounds could be attributed to
the electron-donating ability of the polyphenylene core,
stabilizing the oxidized form with respect to ferrocene itself.
The two redox potentials in diferrocenyl 5 could be as-
signed to the oxidation of the Fe(II)-Fe(II) complex to the
mixed-valence Fe(II)-Fe(III) complex and, at a more posi-
tive potential, oxidation to the Fe(III)-Fe(III) complex. The
first (E°=-0.06 V), which is essentially the same as 1 (E°=
-0.05), is assigned to the Fc(polyphenylene) group and the
second (E° = 0.15 V) to oxidation of the ethynylferrocene
substituent (ethynylferrocene: E°=0.13 V).
The redox potential of 3 is anodically shifted (30 mV)
compared to ferrocene and is in between that of ferrocenyl
polyphenylenes 1, 2, 4, and 6. Clearly the slight electron-
withdrawing ability of the ethynylbenzene is being seen.
The cyclic voltammogram of 6 exhibits a redox potential
of E° = 0.04 V. The significant anodic shift (90 mV) as
compared to the oxidation potential for 1 (E°=-0.05 V) is
due to the electron-withdrawing effect of the fully cyclized
PAH substituent, which is bound directly to the ferrocenyl
moiety. This makes oxidation of the ferrocene unit in 6 elec-
trochemically more difficult than that of uncyclized 1 and
ferrocene. A similar though lesser effect is seen for 3, which
shows a slight anodic shift (by 30 mV) compared to that of
ferrocene. In this case it is due to the electron-withdrawing
effect of the acetylene group.
complex
E° (V)
ΔEp (mV)
1
2
3
4
5
6
-0.05
117
81
90
79
76, 76
70
-0.08
þ0.03
-0.16
-0.06, 0.15
þ0.04
a Redox potentials are given in dichloromethane and referenced vs the
ferrocenium/ferrocene couple using platinum disk working electrode, Pt
wire auxiliary electrode, Ag/Agþ reference electrode, 100 mV s-1 scan
rate, with 0.1 M (Bu4N)PF6 as supporting electrolyte. ΔEp is given by
Epa - Epc
.
The low-energy band (Table 1) of 1 appears at the shortest
wavelength of the polyphenylenes synthesized and is the
closest to that of ferrocene. In polyphenylenes 2, 3, 4, and
5 the low-energy bands are shifted to longer wavelengths
than ferrocene by between 14.5 and 25.5 nm. Therefore the
basicity of the iron center(s) in 2, 3, 4, and 5 would be
expected to be high because of the increased nonplanarity
of the cyclopentadienyl rings. The resulting accelerated
protonation of the metal centers may have serious implica-
tions in the highly acidic conditions of oxidative cyclodehy-
drogenation and perhaps explains the failure of 2, 3, 4, and
5 to generate superaromatics.
Electrochemistry. Ferrocene/ferrocenium itself represents
a strictly reversible one-electron redox couple, although in
some derivatives reversibility may be significantly lowered by
decomposition of the electrogenerated ferrocenium species.12
Ferrocene substituents may also be regarded as a redox-
active probe on a molecular level, since following the ferro-
cene/ferrocenium redox potentials offers a unique opportu-
nity to detect alterations in the electronic structure of the
ferrocene-derived molecules, as well as the metal-π-ligand
bonding. It also provides information about the changes of
electron density distribution and HOMO energies in a series
of related compounds.12
Table 3 compares the redox potentials of, 1, 2, 3, 4, and 5,
in dichloromethane (0.1 M (Bu4N)PF6 as electrolyte). Under
the experimental conditions, all derivatives studied, except
diferrocenyl 5, exhibit a fully reversible one-electron oxida-
tion attributable to the ferrocene/ferrocenium couple. Deri-
vative 5 displays two such reversible one-electron oxidations.
All potentials are quoted versus the [Fc/Fcþ] couple,
which had a potential of 0.51 V using the Ag/AgCl nonaque-
ous reference electrode used in this work. Peak separations
are in the range ΔE = 70-120 mV, close to the reversible
limit. These values are similar to those of the ferrocene
standard (160 mV under the same voltammetric conditions),
indicating that the one-electron [Fc/Fcþ] oxidation processes
are electrochemically reversible in dichloromethane.
For 6 a reversible one-electron redox process was observed
at E°= 0.04 V with ΔE = 70 mV, while the diferrocenyl,
complex 5, exhibits two redox processes at E°=-0.06 and
0.15 V. The peak separations (ΔE = 76 mV) and almost
identical peak currents indicate two reversible one-electron
oxidations occurring independently at each ferrocene center.
Although all the ferrocenyl compounds show reversible
one-electron oxidation, it is interesting to note that the redox
potentials of the ferrocene units show a broad range of values
from -0.08 to 0.15 V, caused by the electronic effect of the
substituents. The net result of the two synergic interactions
between the metal and the ligand frontier orbitals and the E°
Conclusions
Four novel ferrocenyl polyphenylenes, 1, 3, 4, and 5, have
been synthesized and spectroscopically characterized. The
redox chemistry of these ferrocenyl derivatives has been
investigated by cyclic voltammetry and compared to ferro-
cene and the previously reported 2. The molecular structure
of 1 has been elucidated by single-crystal X-ray analysis.
The work describes the successful cyclodehydrogenation
of 1 to the novel ferrocenyl PAH derivative 6 and demon-
strates for the first time the conditions for the generation of
metallo-organic superaromatics.
The electronic spectra of the complexes were measured in
toluene. Polyphenylenes 1, 2, 3, 4, and 5 all display similar
features in their UV-vis absorption spectra; however the
spectral data support the proposition that the two cyclopen-
tadienyl rings are significantly tilted in 2, 3, 4, and 5.
Consequently, protonation of the iron atom of the ferrocene
units in these compounds is expected to be accelerated, and
their oxidation reactivity increased in acidic media during
cyclodehydrogenation. Conversely the cyclopentadienyl
rings in 1 are less tilted, and as a result, the ferrocenyl moiety
was able to withstand cyclodehydrogenation to yield 6. The
absorption spectrum of ferrocenyl complex 6 was dominated
by the πfπ* absorption bands of the polycyclic aromatic
substituent, and the absorption due to the d-d transitions at
lower energy was effectively swamped. The electrochemistry
(32) Wang, X. Y.; Deng, Z. X.; Jin, B. K.; Tian, Y. P.; Lin, X. Q.
Electrochim. Acta 2002, 47, 1537.

6546 Organometallics, Vol. 29, No. 23, 2010
Roberts et al.
of all the compounds was investigated by cyclic voltammetry
in dichloromethane, and all derivatives exhibit the one-
electron reversible oxidation wave associated with the ferro-
cene/ferrocenium couple. The corresponding E° values are
influenced by the electronic effects of the substituent and
were found to support the aromatic shifts observed in the
NMR data. The value of the oxidation potential of ferro-
cenyl 6 is considerably more positive than that of 1, due to the
electron-withdrawing effect of the polyaromatic function
directly attached to the ferrocene unit. The diferrocenyl
polyphenylene derivative 5 displayed separated, reversible,
one-electron-oxidation steps (E° = -0.06 and 0.15 V), oc-
curring independently at each ferrocenyl center. Oxidation
of 5 thus involves two consecutive one-electron steps to give
the corresponding mono- and then diferrocenium species in
the potential region examined.
performed using SMART and SAINT-NT (v. 6.2). Intensities
were corrected for Lorentz and polarization effects and for
absorption using SADABS (v. 2).33 Space groups were deter-
mined from systematic absences and checked for higher sym-
metry. The structures were solved by direct methods using
SHELXS (v. 5.1) and refined on F2 using all data by full-matrix
least-squares procedures with SHELXL (v. 5.1).34 The functions
minimized were w(Fo2 - Fc ), with w = [s2(Fo2)þ(aP)2þbP]-1
,
2
where P = [max(Fo)2þ2Fc ]/3. In all cases, final Fourier syn-
theses showed no significant residual electron density in chemi-
cally sensible positions.
2
General Procedure for the Synthesis of 1, 2, 3, 4, and 5 via
[2þ4]-Cycloaddition Diels-Alder Reaction. A mixture of appro-
priate acetylene (1 equiv), tetraphenylcyclopentadienone (1 equiv,
or 1.2 or 2.2 equiv in the case of 3 and 4), and benzophenone was
mixed in a round-bottomed flask and attached to an air con-
denser. The mixture was heated at 200 °C for 48 h. The reaction
mixture was subsequently cooled to RT, and the residue was
extracted into dichloromethane and filtered. After removal of
the solvent the recovered material was subjected to flash chro-
matography (SiO2, 1:1 hexane/dichloromethane).
This work demonstrates the formation of an organo-
metallic molecular graphene, which could have wide-reaching
applications for the future study and production of molec-
ular materials.
1
Compound 1. Yield: 70%, mp 213 °C. H NMR (400 MHz,
CDCl3): δ 7.95 (s, 1H, Ph H), 7.17-6.66 (m, 20H, Ph H), 4.06
(s, 5H, Cp H), 3.96 (t, 2H, J=2 Hz, Cp H), 3.89 (t, 2H, J=2 Hz,
Cp H). 13C NMR (100 MHz, CDCl3): δ 141.2, 140.3, 139.7,
139.4, 139.2, 139.1, 137.0, 136.9, 136.0, 130.7, 130.5, 130.4,
130.2, 128.9, 126.7, 126.1, 125.8, 125.5, 125.2, 124.8, 124.4,
124.2, 85.7, 69.4, 68.5, 66.7. Anal. Calcd for C40H30Fe: C,
84.80; H, 5.33. Found: C, 84.68; H, 5.38. ESI-MS (CH3CN):
m/z (%) [M]þ 566.3 (100) (calcd 566.4). IR (KBr disk, cm-1):
ν(C-H st) 3080, 3055, 3022, ν(CdC st) 1599, 1494, 1441, 696.
Compound 2. Yield: 36%, mp >300 °C. 1H NMR (400 MHz,
CDCl3): δ 7.08 (m 10H, Ph H), 6.81 (m, 15H, Ph H), 3.88 (s, 2H,
Cp H), 3.80 (s, 2H, Cp H), 3.73 (s, 5H, Cp H). 13CNMR(100MHz,
CDCl3): δ 131.7, 130.8, 130.6, 126.2, 125.8, 125.3, 124.6, 124.5,
72.8, 68.9, 66.5. ESI-MS (CH3CN): m/z (%) [M]þ 642.6 (100)
(calcd 642.6). ESI-MS (CH3CN): calculated for C46H34Fe [M]þ
m/z 642.2010, found 642.2036. IR (KBr disk, cm-1): ν(C-H st)
3080, 3056, 3022, ν(CdC st) 1599, 1497, 1441, 699, ν(CH rock,
Cp) 1108, 1070, 1029, 1006.
Experimental Section
Materials and Methods. Unless otherwise stated, all reactions
were carried out under an argon atmosphere using standard
Schlenk techniques. All solvents were distilled under nitrogen
from appropriate drying agents and degassed prior to use. Flash
chromatography was performed using silica gel (Fluka 60)
or activated alumina (Brockman I, Aldrich Chemical) as the
stationary phase, and all separations were carried out in air.
Tetraphenylcyclopentadienone (Aldrich) and ferrocene (Aldrich)
were used as received. Ethynylferrocene,21 phenylethynylferrocene,22
1,10-bis(phenylethynyl)ferrocene,23 and 1,4-bis(ferrocenyl)but-
adiyne24 were synthesized according to literature procedures.
Physical and Analytical Measurements. NMR spectra were
recorded in CDCl3 with (i) a Bruker Avance DPX-400 MHz
spectrometer at the frequencies 400.13 MHz for 1H and 100.6 MHz
for 13C or (iii) an AV-600 MHz spectrometer at 600.13 MHz for
1H and 150.6 MHz for 13C. The signals for 1H and 13C spectra
were referenced to TMS at δ 0.00, and coupling constants were
recorded in hertz (Hz). Electrospray mass spectra were re-
corded on a micromass LCT tof electrospray mass spectro-
meter. Despite multiple attempts, it was not possible to obtain
consistent, accurate elemental analysis results except for 1.
This is not uncommon where large polyphenylene or thermally
robust polyaromatic frameworks are involved. In the alter-
Compound 3. Yield: 44%, mp 239-240 °C. 1H NMR (400 MHz,
CDCl3): δ 7.49-7.43 (m, 2H, Ph H), 7.33-7.32 (m, 3H, Ph H),
7.16-7.06 (m, 10H, Ph H), 6.85-6.77 (m, 15H, Ph H), 4.08 (t,
2H, J=2 Hz, Cp H), 3.94 (t, 2H, J=2 Hz, Cp H), 3.81 (t, 2H, J=
2 Hz, Cp H), 3.63 (t, 2H, J=2 Hz, Cp H). 13C NMR (150 MHz,
CDCl3): δ 141.67, 141.47, 141.39, 140.64, 140.46, 139.62, 136.73,
133.08, 131.94, 131.21, 131.07, 130.84, 128.11, 126.61, 126.35,
126.12, 126.09, 125.75, 124.81, 88.37, 87.79, 85.56, 84.20, 74.11,
71.98, 70.60, 69.92. ESI-MS (methanol) for C54H38Fe: [M]þ
calculated m/z 742.2323, found m/z 742.2297. IR (KBr disk,
cm-1): ν(C-H st, aromatic) 3079, 3053, 3023, ν(CtC st) 2207,
ν(CdC st): 1599, 1494, 1441, ν(C-H rock Fc) 1070, 1026, 1010.
Compound 4. Yield: 18%, mp >200 °C (dec). 1H NMR
(400 MHz, CDCl3): δ 7.00 (m, 20H, Ph H), 6.78 (m, 30H, Ph H),
3.43 (s, 4H, Cp H), 3.19 (s, 4H, Cp H). ESI-MS (toluene) for
C82H58Fe: [M]þ calculated m/z 1098.3888, found m/z 1098.3873.
IR (KBr disk, cm-1): ν(C-H st, aromatic) 33078, 3053, 3022,
ν(CdC st) 1599, 1464, ν(C-H rock Fc) 1073, 1027.
1
native the H NMR spectra are provided for all other novel,
numbered compounds (3-6) in the Supporting Information,
and accurate mass spectra recorded against lucine enkephalin
(555.6 g mol-1) are reported with accuracy within 5 ppm in all
cases.
Cyclic voltammetry was performed using a CH Instruments
electrochemical analyzer model 600B and a conventional three-
electrode cell with a Ag/AgCl (saturated KCl) reference elec-
trode. All solutions were degassed with N2 before measurements
were taken, and a blanket of N2 was maintained over the solu-
tion for the duration of the experiment. Tetrabutylammonium
hexafluorophosphate was used as the supporting electrolyte for
all the experiments.
Compound 5. Yield: 75%, mp >300 °C. 1H NMR (400 MHz,
CDCl3): δ 7.29 (m, 5H, Ph H), 7.00 (m, 3H, Ph H), 6.84 (m, 10H,
Ph H), 6.72 (m, 2H, Ph H), 4.40 (s, 2H, Cp H), 4.22 (s, 5H, Cp H),
4.20 (s, 4H, Cp H), 4.09 (s, 7H, Cp H). 13C NMR (100 MHz,
CDCl3): δ 144.5, 141.4, 141.3, 140.4, 140.2, 140.1, 139.8, 138.6,
136.9, 131.3, 130.8, 130.7, 130.4, 126.8, 126.6, 126.1, 126.0,
125.6, 125.1, 124.8, 124.7, 122.6, 100.8, 87.2, 84.6, 71.3, 70.2,
All photophysical studies were carried out with solutions
contained within 1 ꢀ 1 cm2 quartz cells in HPLC grade solvents
and were degassed using N2 bubbling. UV-visible absorption
spectra were recorded on a Shimadzu UV-2450 spectropho-
tometer and the data analyzed using UVProbe software.
X-ray analysis was performed with a Bruker SMART APEX
CCD diffractometer using graphite-monochromised Mo KR
(33) SAINTþ; Bruker-AXS, 1997-1999.
(34) (a) Sheldrick, G. M. SADABS; University of Gottingen, 1998.
(b) Sheldrick, G. M. Acta Crystallogr. 1990, A46, 467. (c) Sheldrick, G. M.
SHELX-97, University of Gottingen, 1997.
˚
(λ = 0.71073 A) radiation at 153 K. A full sphere of data was
obtained using the omega scan method. The data reduction was

Article
Organometallics, Vol. 29, No. 23, 2010 6547
69.4, 68.3, 66.5, 65.9. ESI-MS (CH3CN): m/z (%) [M]þ 774.2
(100) (calcd 774.5). ESI-MS (CH3CN): calculated for C52H38Fe2
[M]þ m/z 774.1672, found 774.1686. IR (KBr disk, cm-1):
ν(C-H st) 3082, 3056, 3022, ν(CtC st) 2211, ν(CdC st):
1600, 1496, 1441, 697, ν(CH rock, Cp) 1106, 1069, 1027, 1001.
Synthesis of 16-Ferrocenyl(tribenzo[e;g,h,i;k])perylene (6). Com-
pound 1 (180 mg; 0.318 mmol), AlCl3 (700 mg; 5.25 mmol), and
CuCl2 (700 mg; 5.21 mmol) were stirred in CS2 (60 mL) at 25 °C for
48 h under an argon atmosphere. The black solid was allowed to
settle and the CS2 removed by cannula. The solid was dried in vacuo
and stirred in dilute ammonia solution. The product was extracted
into chloroform and dried over MgSO4. Chromatography (SiO2,
7:3 hexane/diethyl ether).
(100 MHz, CDCl3): δ 133.8, 130.7, 130.1, 130.0, 129.8, 129.7,
129.6, 129.5, 129.4, 127.5, 125.8, 124.7, 124.4, 124.3, 124.0,
130.2, 127.4, 127.1, 126.7, 126.4, 126.2, 126.0, 124.1, 123.4,
123.1, 122.9, 121.6, 121.4, 121.2, 120.8, 70.7, 69.5, 67.9. ESI-
MS (toluene): m/z (%) [M]þ 559.8 (100) (calcd 560.4). ESI-MS
(toluene): calculated for C40H24Fe [M]þ m/z 560.1227, found
560.1226. IR (KBr disk, cm-1): ν(C-H st) 3083, 3071, ν(CdC
st) 1459, 1383, 1261, ν(CH rock, Cp) 1096, 1075, 1021.
Acknowledgment. We thank Dr. John O’Brien for tech-
nical assistance. D.J.G. thanks Enterprise Ireland award IF-
2001\369, and D.R. thanks IRCSET and Science Founda-
tion Ireland (05 PICA-I819) for financial support.
1
Yield: 15%, mp >300 °C. H NMR (400 MHz, CDCl3): δ
9.83 (s, 1H, Ph H), 9.07 (d, 2H, 3JHH=8 Hz, Ph H), 9.01 (t, 2H,
3JHH=7 Hz, Ph H), 8.92 (d, 2H, 3JHH=8 Hz, Ph H), 8.71 (d, 1H,
3JHH=8 Hz, Ph H), 8.11 (m, 3H, Ph H), 7.89 (t, 1H, 3JHH=7 Hz,
Ph H), 7.83 (t, 1H, 3JHH=7 Hz, Ph H), 7.59 (t, 1H, 3JHH=7.5 Hz,
Ph H), 7.24 (t, 1H, 3JHH=7.5 Hz, Ph H), 4.77 (t, 2H, J=2 Hz, Cp H),
4.50 (t, 2H, J = 2 Hz, Cp H), 4.12 (s, 5H, Cp H). 13C NMR
Supporting Information Available: The crystallographic data
for compound 1 as well the complete 1H NMR spectra for novel
compounds 3, 4, 5, and 6. Tables showing H-H TOCSY and
HMQC information for 6. This material is available free of
charge via the Internet at http://pubs.acs.org.