Synthesis and group electronegativity implications on the electrochemical and spectroscopic properties of diferrocenyl meso-substituted porphyrins

Article abstract of DOI:10.1021/om060373u

Two different condensations of appropriate dipyrromethanes and aldehydes resulted in two structural isomers of metal-free, meso-substituted diferrocenyldipentafluorophenyl porphyrins, one with the ferrocenyl groups in the 5,15 positions, 6, and the other with the ferrocenyl groups in the 5,10 positions, 7. UV/vis spectroscopic and cyclic voltammetric (CV) techniques could not unambiguously distinguish between the isomers, but ¹H NMR clearly distinguished between them. Use of the CH₂Cl₂/0.1M [N(ⁿBu)₄][B(C₆F₅)₄] solvent/supporting electrolyte system allowed good resolution between the two ferrocenyl CV waves with ΔE′ = 111 and 115 mV for 6 and 7, 102 mV for the Zn derivative of 6, namely 8, and 109 mV for 6's Ni derivative, 9. ΔE′ values for the Zn (10) and Ni (11) derivatives of 7 were 103 and 95 mV, respectively. The formal reduction potentials, E′, at which the two observed ring-based electrochemical reductions and one oxidation process were detected, varied in a manner that depended on the cationic electronegativities, χ_Zn2+ or χ_Ni2+, of the coordinated central cations. Differences in E′ of 6-11 with respect to that of meso tetraphenyl porphyrin, 2Htpp, were found to be related to the group electronegativities, χ_C6H5, χ_C6F5, χ_Fc, and χ_Fc+, of each meso substituent.

Full text of DOI:10.1021/om060373u

102
Organometallics 2007, 26, 102-109
Synthesis and Group Electronegativity Implications on the
Electrochemical and Spectroscopic Properties of Diferrocenyl
meso-Substituted Porphyrins
Aure´lien Auger and Jannie C. Swarts*
Department of Chemistry, UniVersity of the Free State, P.O. Box 339, Bloemfontein, 9300,
Republic of South Africa
ReceiVed May 1, 2006
Two different condensations of appropriate dipyrromethanes and aldehydes resulted in two structural
isomers of metal-free, meso-substituted diferrocenyldipentafluorophenyl porphyrins, one with the ferrocenyl
groups in the 5,15 positions, 6, and the other with the ferrocenyl groups in the 5,10 positions, 7. UV/vis
spectroscopic and cyclic voltammetric (CV) techniques could not unambiguously distinguish between
the isomers, but 1H NMR clearly distinguished between them. Use of the CH2Cl2/0.1M [N(nBu)4][B(C6F5)4]
solvent/supporting electrolyte system allowed good resolution between the two ferrocenyl CV waves
with ∆E°′ ) 111 and 115 mV for 6 and 7, 102 mV for the Zn derivative of 6, namely 8, and 109 mV
for 6’s Ni derivative, 9. ∆E°′ values for the Zn (10) and Ni (11) derivatives of 7 were 103 and 95 mV,
respectively. The formal reduction potentials, E°′, at which the two observed ring-based electrochemical
reductions and one oxidation process were detected, varied in a manner that depended on the cationic
electronegativities, øZn2+ or øNi2+, of the coordinated central cations. Differences in E°′ of 6-11 with
respect to that of meso tetraphenyl porphyrin, 2Htpp, were found to be related to the group
electronegativities, øC6H5, øC6F5, øFc, and øFc+, of each meso substituent.
fluorinated macrocycles, van Lier and co-workers found that a
series of asymmetrically substituted fluorinated phthalocyanines
are effective photosensitizers for the production of singlet
oxygen.9 The photochemical and electron transport activity of
fluorinated porphyrins have also been investigated.10
Ferrocene-based materials, on the other hand, have found
increasing use in transition metal catalyzed processes.11 The
powerful electron-donating properties of the ferrocenyl group
(group electronegativity ) 1.87)12 cause ferrocene-based deriva-
tives to represent an important class of ligands for use in
coordination chemistry.13 The biological activity of ferrocene
derivatives in the treatment of cancer has also been the focus
of many literature reports.14
Introduction
Porphyrins and their derivatives are well-known tetrameric
macrocycles that, owing to their expanded aromatic π-electron
system, display unique physical and chemical properties.1 Their
electrochemical2 and spectrophysical3 properties in particular
make these macrocyclic chromophores and their derivatives ideal
candidates to mimic biological processes found in nature.4
Porphyrin derivatives show rich electrochemical behavior in that
six2 ring-centered redox processes may be observed, the metals
coordinated in the macrocyclic cavity may be redox active,5
and electroactive meso or â-pyrrole substituents6 may further
expand the broad variety of electrochemical processes that this
class of compounds may exibit. The electrochemistry of a
compound can often be strongly related to its activity as a
prospective drug.7 In addition, the aromatic porphyrin and
related phthalocyanine macrocycles are well-known photo-
sensitisers for use in photodynamic therapy.8 In terms of
Porphyrins with strong electron donor and acceptor groups
have found applications as second-order nonlinear optical
systems.15 This generates the promise that asymmetrically
(8) (a) Detty, M. R.; Gibson, S. L.; Wagner, S. J. J. Med. Chem. 2004,
47, 3897. (b) Ackroyd, R.; Kelty, C.; Brown, N.; Reed, M. Photochem.
Photobiol. 2001, 74, 656.
(9) Sharman, W. M.; Van Lier, J. E. Bioconjugate Chem. 2005, 16, 1166.
(10) Nango, M.; Iida, K.; Kawakita, T.; Matsura, M.; Harada, Y.;
Yamashita, K.; Tsuda, K.; Kimura, Y. J. Chem. Soc., Chem. Commun. 1992,
545.
(11) (a) Hayashie, T. In Ferrocenes: Homogeneous Catalysis, Organic
Synthesis, Material Science; Togni, A., Hayashi, T., Eds.; VCH: Weinheim,
1995; p 105. (b) Bishop, J. J.; Davison, A.; Katcher, M. L.; Lichtenberg,
D. W.; Merril, R. E.; Smart, J. C. J. Organomet. Chem. 1971, 22, 241. (c)
Cullen, W. R.; Woolins, J. D. Coord. Chem. ReV. 1981, 39, 1. (d)
Marqading, D.; Klusacek, A.; Gokel, G.; Hoffman, P.; Ugi, I. J. Am. Chem.
Soc. 1970, 92, 5389.
* To whom correspondence should be addressed. E-mail:
swartsjc@mail.uovs.ac.za.
(1) (a) Walker, F. A. In The Porphyrin Handbook; Kadish, K. M., Smith,
K. M., Guilard, R., Eds.; Academic Press: New York, 2000; Vol. 5, p 81
(b) Pandey, R. K.; Zheng, G. In The Porphyrin Handbook; Kadish, K. M.,
Smith, K. M., Guilard, R., Eds.; Academic Press: New York, 2000; Vol.
6, p 157.
(2) Kadish, K. M.; Royal, G.; Van Caemelbecke, E.; Gueletti, L. In The
Porphyrin Handbook; Kadish, K. M., Smith, K. M., Guilard, R., Eds.;
Academic Press: New York, 2000; Vol. 9, p 1.
(3) Ali, H.; van Lier, J. E. Chem. ReV. 1999, 99, 2379.
(4) Kadish, K. M.; Guo, N.; Van Caemelbecke, E.; Froiio, A.; Paolesse,
R.; Monti, D.; Tagliatesta, P.; Boschi, T.; Prodi, L.; Boletta, F.; Zaccheroni,
N. Inorg. Chem. 1998, 37, 2358.
(12) Du Plessis, W. C.; Vosloo, T. G.; Swarts, J. C. J.Chem. Soc., Dalton
Trans. 1998, 2507.
(5) (a) Walker, F. A.; Beroiz, D.; Kadish, K. M. J. Am. Chem. Soc. 1976,
98, 3484. (b) Kadish, K. M.; Lin, X. Q.; Han, B. C. Inorg. Chem. 1987,
26, 1987.
(6) Beer, P. D.; Drew, M. G. B.; Jagessar, R. J. Chem. Soc., Dalton
Trans. 1997, 881.
(7) Swarts, J. C.; Swarts, D. M.; Maree, D. M.; Neuse, E. W.; La
Madeleine, C.; Van Lier, J. E. Anticancer Res. 2001, 21, 2033.
(13) Togni, A. Angew. Chem., Int. Ed. 1996, 108, 1581.
(14) (a) Swarts, J. C. In Macromolecular Symposia; Guiseppi-Elie, A.,
Levon, K., Eds.; Wiley-VCH: Weinheim, 2002; Vol. 186, p 123. (b) Swarts,
J. C.; Van Rensburg C. E. J. European patent 1345951, 2004.
(15) (a) Samyn, C.; Verbiest, T.; Persoons, A. Macromol. Rapid Commun.
2001, 21, 1. (b) Di Bella, S. Chem. Soc. ReV. 2001, 30, 355. (c) Kelley, A.
M.; Leng, W.; Blanchard-Desce, M. J. Am. Chem. Soc. 2003, 125, 10520.
10.1021/om060373u CCC: $37.00 © 2007 American Chemical Society
Publication on Web 12/07/2006

Electrochemistry of Diferrocenyl Porphyrins
Organometallics, Vol. 26, No. 1, 2007 103
Scheme 1. Syntheses of Porphyrins 6-11
substituted porphyrins bearing simultaneously electron-donating
ferrocenyl and electron-withdrawing fluorine substituents may
have interesting new chemical, electrochemical, and other
physical properties. A key aspect in studies involving porphyrins
with different substituents is the selective synthesis, isolation,
and characterization of single isomers from mixtures of different
substituted macrocycles.16 A need clearly exists to develop the
chemistry surrounding porphyrin complexes bearing multiple
substituent types, including those having simultaneously electron-
withdrawing and electron-donating substituents. Consequently,
in this publication, we report the synthesis and electronic and
electrochemical properties of symmetrically and unsymmetri-
cally substituted porphyrin isomers bearing simultaneously two-
electron-withdrawing pentafluorophenyl and two-electron-
donating ferrocenyl groups.17
positions 5 and 15. We label this structural isomer the trans
isomer. In the case of isomer 7, the ferrocenyl units are situated
in the 5 and 10 meso positions, and this conformation is referred
to as the cis structural isomer.
Porphyrins bearing two different substituents in the meso
positions (trans-A2B2 or cis-A2B2) have been available via a
statistical condensation of pyrrole 1 with two different alde-
hydes, e.g., 2 and 3.18,19 However, the synthesis of the present
ferrocene-containing porphyrins 6 and 7 involved higher yielding
condensations of dipyrromethanes with suitable aldehydes. The
trans isomer 6 was obtained by reacting ferrocenecarbox-
aldehyde, 3, with the dipyrromethane 4, under Schlenck
conditions, following Lindsey protocols.20 To fully aromatize
the porphyrinoid species that formed, the reaction was quenched
with an excess of DDQ (2,3-dichloro-5,6-dicyanobenzoquinone).
A different route was considered for the preparation of the
cis isomer, 7. It involved the simultaneous mixed condensation
of equimolar amounts of dipyrromethanes 4 and 5 with
aldehydes 2 and 3. Chromatographic separation rid 7 of different
porphyrin compounds and other byproducts. Porphyrin 7 was
isolated in 4% yield and was characterized by elemental analysis
as well as with various spectroscopic methods. The poor yield
of 7 (4%) as compared to the yield for 6 (22%) is explained by
noting that, statistically, the cyclocondensation to obtain 7 can
lead to more porphyrin products with different meso substituent
patterns (e.g., AAAA, AAAB, ABAB, or ABBB) than in the
case of 6.
Results and Discussion
Synthesis. The overall synthetic route to the target porphyrins
of this study is illustrated in Scheme 1. The reaction of pyrrole,
1, with 2,3,4,5,6-pentafluorobenzaldehyde, 2, or ferrocenecar-
boxaldehyde, 3, gave respectively, after addition of a catalytic
amount of trifluoroacetic acid and chromatographic purification,
dipyrromethanes 4 (29% yield) and 5 (77% yield). The high-
yield synthesis of 5, which decomposed faster in solution when
exposed to a normal air atmosphere, was also described
elsewhere18 and is considered the result of the presence of the
strong electron-donating ferrocenyl group.
Metalation of 6 and 7 with zinc acetate or nickel acetate was
done in DMF (N,N-dimethylformamide). High-yield formations
of the zinc derivatives 8 (76%) and 10 (93%) and nickel
The porphyrins 6 and 7 are structural isomers of each other.
In porphyrin 6, the meso-substituted ferrocenyl groups are in
(16) Bonnett, R. Chem. ReV. 1995, 24, 19.
(19) Littler, B. J.; Miller, M. A.; Hung, C.; Wagner, R. W.; O’Shea, D.
F.; Boyle, P. D.; Lindsey, J. S. J. Org. Chem. 1999, 64, 1391
(20) (a) Lindsey, J. S. In The Porphyrin Handbook; Kadish, K. M., Smith,
K. M., Guilard, R., Eds.; Academic Press: New York, 2000; Vol. 1, p. 45.
(b) Rao, P. D.; Dhanalekshmi, S.; Littler, B. J.; Lindsey, J. S. J. Org. Chem.
2000, 65, 7323.
(17) After this study was completed we also became aware of novel work
performed by Gryko in this area; Koszarna, B.; Butenscho¨n, H.; Gryko, D.
T. Org. Biol. Chem. 2005, 3, 2640.
(18) Gallagher, J. F.; Moriarty, E. Acta Crystallogr., Sect. C: Cryst.
Struct. Commun. 1999, 55, 1079.

104 Organometallics, Vol. 26, No. 1, 2007
Auger and Swarts
â-pyrrole proton peak positions to appear at much higher field
than does the zinc derivative. The zinc cation is so weakly
electronegative that it enriches electron density on the porphyrin
macrocycle compared to both the Ni porphyrin and the 2H
porphyrin. On the assumption that there is a linear relationship
1
between pyrrole H NMR peak positions and cation electro-
negativity, it follows that the electronegativity of the proton pair
should be ca. ø2(H+) ) 1.5-1.6. This value is very much
substantiated by the Soret band λmax electronic spectra results
(see below) and differs by 0.5-0.6 units from the atomic Pauling
øH ) 2.1 value.
1
The pyrrole chemical shifts and H NMR signal patterns
recorded for the cis isomer 7 were more complex. Rather than
1
two clean doublets, multiple H NMR signals for the eight
â-pyrrolic protons were observed (Figure 1). Two nearly
overlapping doublets that integrated for four protons were
observed at 8.65 ppm. Further downfield, a multiplet at 9.83
ppm (for 1H) and two doublets at 9.97 (for 2H) and 10.05 ppm
(for 1H) were observed. This rather unusual pattern may be
associated with atropisomerism resulting from the presence of
two ferrocenyl units in the 5 and 10-meso positions. Also,
restricted rotation of the ferrocenyl groups by virtue of their
close proximity could influence the electronic environments of
the number 7 and 8 â-pyrrolic protons located between these
organometallic substituents. It is important to notice that, so
1
far, H NMR spectroscopy was the only method that could
distinguish 6 from 7. Since their molecular weight as well as
molecular formulas are identical due to their isomeric relation-
ship (Scheme 1), their mass spectra and elemental analyses data
were the same. Further upfield in the spectra, the typical signals
of monosubstituted ferrocenyl units were in agreement with the
trans and cis substitution patterns of 6 and 7. Chemical shifts
and multiplicity of signals at 5.54 (t), 4.89 (t), and 4.21 (s) ppm
for 6 are in agreement with freely rotating ferrocenyl groups in
the 5 and 15 meso positions. The 1H NMR signals of the
ferrocenyl group of 7 were recorded at 5.54 (m), 4.88 (m), and
4.16 (two singlets). The multiplicity of these signals, Figure 1,
is again consistent with a cis 5,10-ferrocenyl substitution pattern
Figure 1. 1H NMR spectra of porphyrins 6 (bottom) and 7 (middle
and top).
derivatives 9 and 11 (62% and 57%, respectively) were
observed. The new porphyrins 6-11 are all stable in air and
are soluble in common organic solvents.
1
1H NMR Spectroscopy. H NMR characterization distin-
guished conclusively between the trans conformational isomer
6 and the cis conformational isomer 7. The room-temperature
1H NMR spectrum obtained for 6 in CDCl3 (Figure 1) showed
a simple pattern in the aromatic region. Two doublets at 9.97
and 8.66 ppm integrated for four protons each and were assigned
to two sets of â-pyrrolic protons. This reflects perfectly the
trans-A2B2 symmetry of the isomer 6. The electron-withdrawing
effect of the pentafluorophenyl groups through the π-conjugated
system of the porphyrin core is believed to deshield the
â-pyrrolic protons adjacent to the pentafluorophenyl groups and
results in the low-field chemical shift peak position of 9.97 ppm.
Inversely, the electron-donating effect of the ferrocenyl groups
shields the â-pyrrolic protons adjacent to the ferrocenyl groups,
causing the resonance peak to be at the high-field position of
8.66 ppm. The zinc derivative of 6, namely, porphyrin 8,
exhibited a similar 1H NMR spectrum in this region. However,
the two pyrrolic doublet signals were observed at the relatively
more downfield positions of 10.21 and 8.86 ppm, respectively.
In contrast, the signals associated with the â-pyrrolic protons
of the nickel porphyrin derivative 9 appeared at 9.64 and 8.52
ppm. These observations can be related to the cationic electro-
negativities, ø(cat)n+ of the coordinating metallic cations. Zinc
cations (øZn2+ ) 1.1) are less electronegative than nickel cations
(øNi2+ ) 2.00).21 This implies that the Ni2+ cation withdraws
more electron density from the porphyrin ring, causing its
1
that invokes restricted rotation. H NMR data for 10 and 11
also clearly differentiated the cis zinc and nickel derivatives,
different from the corresponding trans analogues; data are
presented in the Experimental Section.
The ferrocenyl-based 1H NMR signals of the zinc derivatives
8 and 10 and nickel derivatives 9 and 11 were observed at almost
the same chemical shifts as were recorded for 6 and 7, indicating
the Zn2+ and Ni2+ metal cation centers did not transmit their
relative electronic effects to meso substituents as effectively as
they do to the â-pyrrole positions.
Furthermore, the chemical shifts for the imino (NH) protons
of 6 and 7 were observed as two overlapping singlets at -1.66
ppm (for 6) and as two separate singlets at -1.63 and -1.73
ppm for 7, respectively. The presence of two ferrocenyl units
adjacent to each other at the 5 and 10-meso positions of 7
implies that the two inner imino protons of isomer 7 are
magnetically nonequivalent and therefore that the two inner
imino protons are situated in different chemical environments.
No signal assignable to any NH protons was observed in the
case of 8-11, providing further evidence for the elemental
analyses of the successful insertion of metal cations in the
macrocyclic cavities of 8 and 10 (Zn2+) and 9 and 11 (Ni2+).
Electronic Absorption Spectroscopy. Porphyrins 6-11 are
dark purple in the solid state but have a greenish tint in solution.
As demonstrated in Figure 2, the electronic spectra of porphyrins
6-11 in THF all exhibited the characteristic strong Soret band
(21) Duffy, A. J. Chem. Soc., Dalton Trans. 1983, 1475.

Electrochemistry of Diferrocenyl Porphyrins
Organometallics, Vol. 26, No. 1, 2007 105
Table 1. Cited potentials are corrected to be referenced against
Fc/Fc+, as required by IUPAC.26 Experimentally, to prevent
signal overlapping, decamethylferrocene (Fc*) was used as
internal reference. Under our conditions, the (Fc*)/(Fc*)+ couple
was found to be at -610 mV versus Fc/Fc+. Electrochemical
reversibility for one-electron-transfer processes at 25 °C is
characterized by peak potential differences of Epa - Epc ) ∆Ep
) 59 mV (pa ) peak anodic, pc ) peak cathodic) and is
independent of the scan rate.27 Because under our experimental
conditions decamethylferrocene exhibited an experimentally
determined ∆Ep value of 78 mV and ferrocene itself gave ∆Ep
) 81 mV, within the context of this paper, experimentally
determined ∆Ep values smaller than 90 mV are considered to
imply electrochemically reversible behavior. Larger ∆Ep values
are considered to represent electrochemically quasi-reversible
behavior (Table 1).
Compounds 6-11 showed two ring-centered one-electron-
transfer reduction waves labeled 1 and 2 in Figures 3 and 4,
two ferrocenyl substituent one-electron-transfer oxidation waves
labeled 3 and 4, and one ring-centered one-electron-transfer
oxidation wave labeled 5 within the potential window that
CH2Cl2 as solvent allows. 2Htpp exhibited a second ring-
centered one-electron oxidation process in the CH2Cl2 potential
window, and this process is labeled 6 in the CV shown in Figure
4. Except for wave 1 for the nickel complexes 9 and 11, waves
1 and 6 for 2Htpp, and wave 5 for the cis complexes 10 and
11, every electron-transfer process was electrochemically revers-
ible, with ∆Ep < 90 mV at scan rate 100 mV s-1 (Table 1).
Chemical reversibility was good with peak current ratios
(calculated as the current of the reverse scan divided by the
current of the forward scan) approaching 1 for all processes
except those associated with wave 6 of 2Htpp, wave 1 for the
nickel complex 9, and wave 5 for complexes 6-11. The general
current ratio deviation from unity of wave 5 is consistent with
a weak onset of electrode pollution, probably via electrode
deposition after the formation of the poorly soluble, doubly
charged [M(Fc+)2Por] species during wave 4 (Figures 3 and
4). It will also explain the prewave sometimes observed at wave
5. Re-reduction at wave 4 must redissolve any precipitated
material from the electrode surface, as wave 4 has very
reversible characteristics.
Figure 2. Electronic spectra of porphyrins 6 (thin black line, 2.17
× 10-6 M), 8 (bold gray line, 2.23 × 10-6 M), and 9 (bold black
line, 7.49 × 10-7 M) in THF. The inset highlights the Q-band
region.
in the 400-450 nm region. The spectra of the free base
porphyrins 6 and 7 (not shown) overlapped almost exactly. They
both showed the intense Soret maximum at 424 nm and a
shoulder at 496 nm. The Q-bands featured two peaks very close
to each other at 631 (Qx) and 697 nm (Qy). Two-fold degeneracy
of the lowest energy singlet state of the metal-free derivatives
6 and 7 explains the splitting of the Q-band. The spectra of 8
and 10 (not shown) and 9 and 11 (not shown) overlapped also
almost exactly. The zinc-metalated derivatives 8 and 10
exhibited a red-shifted Soret band with peak maxima at 428
nm compared to the free base 6. A blue-shifted Soret band with
peak maxima at 421 nm was observed for the nickel complexes
9 and 11. It is also apparent that by introducing a metal in the
cavity of 6 (or 7) the Soret band extinction coefficient was
reduced significantly (Figure 2). Lower extinction coefficients
are consistent with stronger aggregation of metalated porphyrins
in the absence of axial ligation. Upon comparing Soret band
maxima of 6 with that of 5,10,15,20-tetraphenylporphyrin
(2Htpp), porphyrin 6 exhibits a red-shift in the Soret band
maxima from 416 to 424 nm. This may be rationalized in terms
of the sum of group electronegativities, øtot, of the combined
meso substituents. Since øFc ) 1.8712 and øC6F5 ) 2.46,22 øtot,6
) 8.66 on the Gordy scale. For 2Htpp, having four meso-
substituted phenyl groups (øC6H5 ) 2.2112]), øtot ) 8.84. It
follows that the macrocyclic core of 6 and 7 should be more
electron rich than that of 2Htpp, because the combined group
electronegativity of the meso substituents of 6 and 7 is 0.22
Gordy scale units less than for 2Htpp. This translates to a longer
wavelength Soret band λmax value for 6 than for 2Htpp. That
longer λmax values are linearly associated with smaller group
electronegativity values, i.e., more electron-rich species, was
also observed for the ν(CO) stretching frequency wave numbers
of [Rh(â-diketonato)(CO)(PPh3)] complexes23 (ν ) 107/λ with
λ in nm, ν in cm-1 units) and for the ν(CdO) stretching
frequencies of methyl esters of the type RCOOMe.24 Following
the [Rh(â-diketonato)(CO)(PPh3)] and RCOOMe queue, a linear
relationship between λmax and øcations of 8 and 9 was assumed,
and it led upon extrapolation to ø2(H+) ) 1.5 for the free base
porphyrin 6. This value is mutually consistent with that obtained
Linear sweep voltammetry (LSV) confirmed waves 1-6 are
all involved in processes where the same amount of electrons
(one electron) is transferred. Figure 3 demonstrates this for 6.
The partially resolved peaks 3 and 4 assigned to the two
ferrocenyl substituents were much better resolved by Osteryoung
square wave voltammetry. This is also demonstrated in Figure
3 for 6. The use of the extreme noninteracting CH2Cl2/0.1 mol
dm-3 [N(nBu4][B(C6F5)4] solvent electrolyte system28 minimized
ion pairing to the point that the two ferrocene waves did not
coalesce but became resolved. Repetition of the CV experiment
of 6, in the coordinating solvent acetonitrile using the traditional
[N(nBu4)][PF6] electrolyte, resulted in two strongly overlapping
ferrocene-based electrochemical waves, which could not be
resolved at all. Different formal reduction potentials for side
groups on symmetrical complexes in which mixed-valent
1
from the H NMR â-pyrrole peak position/øcations relationship
described above.
Electrochemistry. Cyclic voltammetry (CV) experiments
were conducted in dry CH2Cl2 utilizing 0.1 mol dm-3 [NBu4]-
[B(C6F5)4] as supporting electrolyte in order to restrict any
solvent/porphyrin or electrolyte/porphyrin ion-pairing interac-
tions to a minimum.25 Electrochemical data are summarized in
(25) Barriere, F.; Kirss, R. U.; Geiger, W. E. Organometallics 2005, 24,
48, and references therein.
(26) Gritzner, G.; Kuta, J. Pure Appl. Chem. 1984, 56, 461.
(27) (a) Evans, D. H.; O’Connell, K, M.; Peterson, R. A.; Kelly, M. J.
J. Chem. Educ. 1983, 60, 291. (b) Kissinger, P. T.; Heineman, W. R. J.
Chem. Educ. 1983, 60, 702. (c) Van Benshoten, J. J.; Lewis, J. Y.;
Heineman, W. R. J. Chem. Educ. 1983, 60, 772. (d) Sawyer, T.; Roberts,
J. L. In Experimental Electrochemistry for Chemists; John Wiley and
Sons: New York, 1974; p 118.
(22) Swarts, J. C.; Erasmus, E. Unpublished results.
(23) Conradie, J.; Lamprecht, G. J.; Otto, S.; Swarts, J. C. Inorg. Chim.
Acta 2002, 328, 191.
(24) Kagarisa, R. E. J. Am. Chem. Soc. 1955, 77, 1377.
(28) Barriere, F.; Geiger, W. E. J. Am. Chem. Soc. 2006, 128, 3980.

106 Organometallics, Vol. 26, No. 1, 2007
Auger and Swarts
Table 1. Cyclic Voltammetry Data (potentials vs Fc/Fc+, scan rate ) 100 mV s-1) of 1.0 mmol dm-3 Solutions of Analytes in
CH2Cl2 Solutions Containing 0.1 mol dm-3 [N(nBu)4][B(C6H5)4] at 25 °C
a ipa/ipc b When the switching potential is chosen to exclude wave 6 from the scan, ipc/ipa ) 0.96 c Approximate value, the weak shoulder directly before
wave 5 causes baseline uncertainty d Slow electron-transfer kinetics and/or electrode deposition distorts this value.
Figure 4. Cyclic voltamogramms of 1 mmol dm-3 solutions of
(in order from bottom to top) ferrocene (Fc) and decamethyl-
ferrocene (Fc*), 11 (NiPcis), 9 (NiPtrans), 8 (ZnPtrans), 7 (2HPcis),
6 (2HPtrans), 2Htpp, and 2Htpp scanned over the limited potential
range of -0.80 to 0.70 V in dichloromethane containing 0.1 mol
dm-3 [N(nBu)4][B(C6F5)4] at 25 °C at a scan rate of 100 mV s-1
on a glassy-carbon working electrode. The prewaves before wave
5 for the 2H and Ni derivatives are not due to an impurity. They
probably arise from electrode pollution (deposition) after the
formation of the poorly soluble doubly charged [M(Fc+)2Por]
species during wave 4. Inset: Below the 2Htpp scan is a blowup
of this scan showing clearly the peak at -331 mV, which is linked
to wave 6. As the top CV shows, this peak is absent when the
turn-around potential excludes wave 6.
Figure 3. Top: Osteryoung square wave voltammogram of 1 mmol
dm-3 solutions of 6 in dichloromethane containing 0.1 mol dm-3
[N(nBu)4][B(C6F5)4] at 10 Hz and 25 °C. Middle: Linear sweep
voltammetry at 2 mV s-1. Bottom: Cyclic voltammograms of 6 in
dichloromethane at a scan rate of 100, 200, 300, 400, and 500 mV
s-1 on a glassy-carbon working electrode. The peak labeled Fc* is
that of decamethylferrocene, which has been added as an internal
standard to the solution. The prewave before wave 5 is commented
on in Figure 4.
intermediates are generated are well known in systems that allow
electron delocalization, either through intramolecular through-
bond paths or from a direct metal-metal interaction.29,30 Both
the aromatic porphyrinoid core and the ferrocenyl side groups
allow delocalization of electrons, and the ease at which
multiferrocenyl systems generate mixed-valent intermediates has
been described prior to this work by our research group.31 The
inequivalence of the ferrocenyl and ferrocenium groups of the
mixed-valent intermediates of 6-11 are highlighted in terms
of the group electronegativities, øFc ) 1.8712,30 and øFc+
)
2.82.31,32 The electron-withdrawing power of the ferrocenium
(29) (a) Creutz, C.; Taube, H. J. Am. Chem. Soc. 1969, 91, 3988. (b)
Geiger, W. E.; Van Order, N.; Pierce, D. T.; Bitterwolf, T. E.; Reingold,
A. L.; Chasteen, N. D. Organometallics 1991, 10, 2403. (c) Van Order,
N.; Geiger, W. E.; Bitterwolf, T. E.; Reingold, A. L. J. Am. Chem. Soc.
1987, 109, 5680. (d) Pierce, D. T.; Geiger, W. E. Inorg. Chem. 1994, 33,
373.
group is almost as high as that of the CF3 group (øCF3
)
3.01).12,31
(31) Du Plessis, W. C.; Erasmus, J. J. C; Lamprecht, G. J.; Conradie, J.;
Cameron, T. S.; Aquino, M. A. S.; Swarts, J. C. Can. J. Chem. 1999, 77,
378.
(30) Conradie, J.; Cameron, T. S.; Aquino, M. A. S.; Lamprecht, G. J.;
Swarts, J. C. Inorg. Chim. Acta 2005, 358, 2530.

Electrochemistry of Diferrocenyl Porphyrins
Organometallics, Vol. 26, No. 1, 2007 107
Scheme 2. Ring-Based and Ferrocenyl Group Electron-Transfer Reactions Associated with Redox Potentials for Compounds
6-11 and 2Htpp
The individual electron-transfer reactions associated with
all five redox steps, E°′ values for waves 1, 2, and 5 of the
nickel complexes were lower rather than higher compared to
the corresponding waves of 6 and 7. The deviation from the
expected pattern relating to waves 1 and 5 of the Ni complexes
is in part attributed to the fact that some of these couples are
not electrochemically reversible (∆Ep g 100 mV), implying that
the indicated E°′ values are not true thermodynamic quantities.
Electrode deposition of oxidized products associated especially
with wave 4 may also play a role. Different degrees of ruffling
for the Ni complexes compared to that of the other complexes
upon redox reactions may also contribute to lower than expected
reduction potentials. The fact that the macrocycles are only for
waves 1-6 in Figures 3 and 4 are assigned in Scheme 2.
Inspection of the formal reduction potentials, E°′ ) 1/2(Epa
+
Epc), associated with these potentials reveal two key factors
emanating from this study.
The first is associated with the observation that E°′ of peak
1 of the zinc and nickel complexes is smaller (i.e., more
negative) than that of 2Htpp (Figure 4). This is interpreted as a
consequence of øtot ) 8.84 for the four phenyl substituents in
2Htpp being 0.22 Gordy eletronegativity units larger than øtot
) 8.66 for the two C6F5 and two ferrocenyl meso substituents
of 8-11. These øtot values show 8-11 should be more electron
rich than 2Htpp and therefore slightly easier to reduce. Peak 2
does not follow this trend unambiguously, as 8-11 and 2Htpp
have comparable E°′ values. In contrast, oxidation wave 5 for
the zinc and nickel complexes 8-11 as well as for the free base
complexes 6 and 7 are found at E°′ potentials 0.5-0.7 V larger
than for 2Htpp. This striking difference is easily explained upon
recognizing that the ferrocenyl groups have been oxidized to
the ferrocenium species with øFc+ ) 2.82 at waves 3 and 4.
This changes the combined substituent group electronegativities
for 6-9 at wave 5 to øtot ) 10.56. This large øtot value implies
that 6-11 are much more electron poor and therefore more
difficult to oxidize than 2Htpp at potentials associated with the
redox processes of wave 5. The introduction of redox-tuneable
side chains to porphyrins 6-11 therefore led to the unusual
result that ring-centered reductions occur in electron-rich
macrocycles, but ring-centered oxidations occur in electron-
deficient species.
waves 3 and 4 in the same redox state, namely, (porphyrin)2-
,
1
than for the H NMR and the electronic spectroscopy study,
must also contribute to the observed deviations.
With respect to 2Htpp, oxidation waves 5 and 6 introduced
interesting features in the present solvent electrolyte system.
Oxidation of the [MPor] species to [MPor]•+ at wave 5 happened
smoothly, but the reduction half-wave associated with the
reaction [MPor]•+ f [MPor] was less intense than expected
(ipc current values are small, ipc/ipa ) 0.86). During the cathodic
cycle, a new reduction wave appeared at -331 mV (Figure 4,
blowup under CV second from the top). However, when the
switching potential during the positive scan was lowered from
1.3 V to 0.6 V, i.e., when wave 6 was not engaged during the
CV cycle, wave 5 had the characteristic CV appearance with
∆Ep ) 87 mV and ipc/ipa ) 0.96, and the new peak at -331
mV was absent (Figure 4, top). It follows that the species that
leads to the new peak at -331 mV must be the oxidized species
The second key point that can be quantified from the observed
formal reduction potentials relates to the relative electronic
influence of the Zn2+, Ni2+, and 2(H+) centers. The relatively
higher cation electronegativity of Ni2+ over Zn2+ (øZn2+ ) 1.1,
generated at wave 6, namely, a follow-up product of [MPor]2+
.
Wave 6 is associated with the [MPor]•+/[MPor]2+ redox system
and represents a chemically and electrochemically irreversible
process because ∆Ep ≈ 148 mV and ipc/ipa ) 0.59. Electrode
deposition and enhanced aggregation of the doubly charged
[MPor]2+ species in CH2Cl2 are the most probable causes of
this behavior. The process leading to the new wave at -331
mV partially removes material from the bilayer next to the
electrode, which would account for the lower than expected ipc/
ipa ) 0.86 ratio for wave 5. At slow scan rates, the peak at
-331 mV was very weak, but it became stronger at fast scan
rates.
ø
Ni2+ ) 2.0) implies that 8 (and 10) has a higher ring-centered
electron density than 9 and 11. This implies that E°′ for all
couples of 8 and 10 should be observed at lower potentials than
for 9 and 11. Experimentally this was verified for couples 2-5
(Scheme 2). The failure of wave 1 to follow this trend may be
related to the electrochemical irreversibility of wave 1 of
complexes 9 and 11 (∆E ) 109 and 114 mV, respectively), or
alternatively, the degree of ruffling for the doubly reduced Ni
complexes may be different from that of the other complexes.
Additionally, the 1H NMR and electronic spectra-derived 2(H+)
electronegativity of ø2(H+) ) 1.5-1.6 would imply that com-
plexes 6 and 7 should have E°′ values intermittently between
that of 8-11 in the resting state, i.e., at redox processes
associated with waves 2 and 3. In practice this was only
marginally observed for wave 3, where electron-transfer did not
change the redox state of the macrocycle itself, but that of a
substituent. In addition, while the zinc complexes 8 and 10 did
show the expected lower E°′ values compared to 6 and 7 for
Conclusion
In summary, new porphyrin isomers 6-11 substituted by two
ferrocenyl units and two pentafluorophenyl units were synthe-
sized and fully characterized. The rich electrochemistry of these
new complexes was found to be strongly dependent on the group
electronegativity of the pentafluorophenyl and ferrocenyl sub-
stituents and also of the cationic electronegativity of the central
metallic cation Mn+ coordinated in the porphyrin cavity. To our
knowledge, porphyrin formal reduction potentials were for the
first time quantitatively placed in perspective utilizing group
and cationic electronegativities. In addition, the introduction of
(32) Davis, W. L.; Shago, R. F.; Langner, E. H. G.; Swarts, J. C.
Polyhedron 2005, 24, 1611.

108 Organometallics, Vol. 26, No. 1, 2007
Auger and Swarts
aldehyde (196 mg, 1.00 mmol) were dissolved in DCM (400 mL)
at room temperature before trifluoroacetic acid (1.14 g, 10.00 mmol)
was added. After 15 h of stirring at room temperature, DDQ was
added (0.34 g, 1.50 mmol). Stirring continued for an additional
hour before purification by column chromatography over silica gave
three fractions. The first fraction (eluent hexane/DCM, 2:1) con-
tained the 5-ferrocenyl-10,15,20-tris(pentafluorophenyl)porphyrin
(10 mg, 1%); δH (300 MHz, CDCl3) 10.13 (2H, d, J 5), 8.79 (4H,
m), 8.74 (2H, d, J 5), 5.62 (2H, s), 4.95 (2H, s), 4.26 (5H, s), -2.28
(2H, s). The second fraction (eluent hexane/DCM, 2:1) gave 5,10-
bisferrocenyl-15,20-bis(pentafluorophenyl)porphyrin, 7 (36 mg,
4%), as a purple solid, mp >200 °C. Anal. Found: C, 61.74; H,
2.96; N, 5.58. C52H28N4F10Fe2 requires C, 61.81; H, 2.79; N, 5.54.
δH (300 MHz, CDCl3) 10.05 (1H, d, J 5), 9.97 (2H, d, J 5), 9.83
(1H, m), 8.65 (4H, m), 5.53 (4H, m), 4.88 (4H, m), 4.16 (10H, m),
-1.63 (1H, s), -1.73 (1H, s); λmax (2.77 × 10-6 M in THF) 424
(ꢀ ) 2.20 × 105) nm.
redox-tuneable side chains led to the unusual result that ring-
centered porphyrin reductions occur in electron-rich macro-
cycles, but ring-centered oxidations occur in electron-deficient
macrocycles. The potential use of these complexes as second-
generation photosensitisers is now under investigation.
Experimental Section
General Procedures. Solid reagents and liquid reactants except
pyrrole (Merck or Aldrich) were used without further purification.
Solvents were distilled prior to use, and water was double distilled.
Pyrrole was distilled from CaH2 under argon. All reactions required
anhydrous conditions and were therefore conducted under a nitrogen
atmosphere. [N(nBu)4][B(C6F5)4] was synthesized according to
published procedures.33
Column chromatography was performed on silica gel 60 (particle
size 0.040-0.063 mm). 1H NMR spectra were recorded on a Bruker
Advance DPX 300 NMR spectrometer (300 MHz) in CDCl3
solutions. Chemical shifts are reported relative to SiMe4 at 0 ppm.
UV/vis spectra were recorded on a Cary 50 probe UV/visible
spectrophotometer. Elemental analyses were performed by the
Canadian Microanalytical Service Ltd., Canada.
5-(Pentafluorophenyl)dipyrromethane (4). The procedure is
a modification of that described in ref 19. A solution of 2,3,4,5,6-
pentafluorobenzaldehyde (0.70 g, 3.60 mmol) and pyrrole (10 mL,
0.144 mol, 40 equiv) was degassed by bubbling nitrogen for 10
min before trifluoroacetic acid (0.026 mL, 0.36 mmol) was added.
The solution was stirred at room temperature for 15 min in the
dark, diluted with DCM (2 × 100 mL), washed with 0.1 M NaOH
(100 mL), and dried over magnesium sulfate. Solvents and the
excess of pyrrole were removed under reduced pressure. Purification
of the resulting slurry by column chromatography over silica (eluent
cyclohexane/EtOAc/triethylamine, 80:20:1) gave 5-(pentafluoro-
phenyl)dipyrromethane, 4 (0.32 g, 29%), as a colorless solid; mp
110 °C (lit. mp 110-112 °C);19 δH (300 MHz, CDCl3) 8.06 (2H,
br s, NH), 6.73 (2H, m), 6.22 (2H, m), 6.11 (2H, m), 5.93 (1H, s).
5-Ferrocenyldipyrromethane (5). A sample of ferrocenecar-
boxaldehyde (0.75 g, 3.50 mmol) was treated as described in the
procedure for 4, affording 5-ferrocenyldipyrromethane (0.88 g, 77%)
as a yellow powder; mp 130-131 °C (lit. 131-132 °C);18 δH (300
MHz, CDCl3) 8.00 (2H, br s, NH), 6.79 (2H, m), 6.18 (2H, m),
6.02 (2H, m), 5.22 (1H, s), 4.19 (2H, m), 4.09 (7H, m).
[5,15-Bisferrocenyl-10,20-bis(pentafluorophenyl)porphyrinato]-
zinc(II) (8) and [5,15-Bisferrocenyl-10,20-bis(pentafluorophenyl)-
porphyrinato]nickel(II) (9). 5,15-Bisferrocenyl-10,20-bis(penta-
fluorophenyl)porphyrin (20 mg, 0.02 mmol) and zinc acetate
dihydrate (3 equiv, 13 mg, 0.06 mmol) were added to N,N-
dimethylformamide (10 mL), and the resulting solution was heated
to reflux for 4 h. Then water (100 mL) was added to the mixture,
and the precipitate was filtered. The resultant material was purified
by column chromatography over silica (eluent hexane/DCM, 1:1)
to give [5,15-bisferrocenyl-10,20-bis(pentafluorophenyl)porphyri-
nato] zinc(II) (8) (16 mg, 76%) as a dark brown solid, mp >200
°C. Anal. Found: C, 58.32; H, 2.63; N, 5.28. C52H26N4F10Fe2Zn
requires C, 58.16; H, 2.44; N, 5.22. δH (300 MHz, CDCl3) 10.21
(4H, d, J 5), 8.86 (4H, d, J 5), 5.53 (4H, t, J 2), 4.87 (4H, t, J 2),
4.24 (10H, s); λmax (2.23 × 10-6 M in THF) 428 (ꢀ ) 2.33 × 105),
650 (ꢀ ) 1.16 × 104) nm.
[5,15-Bisferrocenyl-10,20-bis(pentafluorophenyl)porphyrinato]-
nickel(II) (9) was synthesized in exactly the same manner utilizing
nickel acetate dihydrate (3 equiv, 15 mg, 0.06 mmol) rather than
zinc acetate, yielding 11 mg (62%) as a dark brown solid, mp >200
°C. Anal. Found: C, 58.29; H, 2.49; N, 5.31. C52H26N4F10Fe2Ni
requires C, 58.52; H, 2.46; N, 5.25. δH (300 MHz, CDCl3) 9.64
(4H, d, J 5), 8.52 (4H, d, J 5), 5.16 (4H, br s), 4.75 (4H, br s), 4.02
(10H, s); λmax (7.49 × 10-7 M in THF) 421 (ꢀ ) 1.56 × 105) nm.
[5,10-Bisferrocenyl-15,20-bis(pentafluorophenyl)porphyrinato]-
zinc(II) (10) and [5,10-Bisferrocenyl-15,20-bis(pentafluoro-
phenyl)porphyrinato]nickel(II) (11). These cis derivatives were
synthesized and isolated as described above for the corresponding
trans derivatives by replacing 6 with 7 and using the chromato-
graphic eluent hexane/DCM, 1:1.
Characterization data for dark purple 10: Yield 93%. Anal.
Found: C, 58.35; H, 2.68; N, 5.57. C52H26N4F10Fe2Zn requires C,
58.16; H, 2.44; N, 5.22%); δH (300 MHz, CDCl3) 10.11 (3H, br
m),10.01 (1H, s), 8.73 (4H, m), 5.64 (4H, br s), 4.95 (4H, br s),
4.24 (10H, m); λmax (1.49 × 10-6 M in THF) 429 (ꢀ ) 2.61 ×
105) nm.
Characterization data for dark purple 11: Yield, 57%; (Found:
C, 58.34; H, 2.51; N, 5.30. C52H26N4F10Fe2Ni requires C, 58.52;
H, 2.46; N, 5.25. δH (300 MHz, CDCl3) 9.57 (3H, br m), 9.47 (1H,
s), 8.50 (4H, m), 5.58 (4H, br s), 4.85 (4H, br s), 4.11 (10H, s);
λmax (1.125 × 10-5 M in THF) 421 (ꢀ ) 2.49 × 105) nm.
5,15-Bisferrocenyl-10,20-bis(pentafluorophenyl)porphyrin (6).
Ferrocenecarboxaldehyde (137 mg, 0.64 mmol) and 5-(pentafluoro-
phenyl)dipyrromethane (200 mg, 0.64 mmol) were dissolved in
DCM (100 mL) at room temperature in the dark before trifluoro-
acetic acid (3 mg, 0.025 mmol) was added to initiate the condensa-
tion. After the mixture was stirred for 15 h at room temperature,
the reaction was quenched with DDQ (0.22 g, 0.96 mmol). Stirring
continued for 1 h before triethylamine (2.5 mg, 0.025 mmol) was
added to neutralize the acid. The solvents were removed under
reduced pressure, and the resultant material was purified by column
chromatography over silica (eluent: hexane/DCM, 1:1) to give a
sticky purple solid. The crude product was triturated with methanol
and filtered off to afford 5,15-bisferrocenyl-10,20-bis(pentafluoro-
phenyl)porphyrin, 6 (70 mg, 22%), as a purple powder, mp >200
°C. Anal. Found: C, 61.75; H, 2.84; N, 5.70. C52H28N4F10Fe2
requires C, 61.80; H, 2.79; N, 5.54. δH (300 MHz, CDCl3) 9.97
(4H, d, J 5), 8.66 (4H, d, J 5), 5.54 (4H, m), 4.89 (4H, m), 4.21
(10H, s), -1.66 (2H, s); λmax (2.17 × 10-6 M in THF) 423 (ꢀ )
2.76 × 105) nm.
Cyclic Voltammetry. CV experiments were performed on ca.
1 mM solutions of H2tpp and 6-11 in dry CH2Cl2/0.100 M
[N(nBu)4][B(C6F5)4], utilizing a standard three-electrode cell with
glassy carbon electrode of surface area 7.07 mm2 pretreated by
polishing on a Buehler microcloth first with 1 µm and then 1/4 µm
diamond paste, a Pt-wire counter electrode, and a Ag/Ag+ reference
electrode. The reference electrode was constructed by immersing
a Ag wire in an acetonitrile solution containing 0.010 M AgNO3
and 0.100 M [N(nBu)4][B(C6F5)4] in a thin inner Luggin capillary
5,10-Bisferrocenyl-15,20-bis(pentafluorophenyl)porphyrin (7).
Ferrocenecarboxaldehyde (214 mg, 1.00 mmol), 5-(pentafluoro-
phenyl)dipyrromethane (312 mg, 1.00 mmol), 5-ferrocenyldipyrro-
methane (330 mg, 1.00 mmol), and 2,3,4,5,6-pentafluorobenz-
(33) LeSuer, R. J.; Buttolph, C.; Geiger, W. E. Anal. Chem. 2004, 76,
6395.

Electrochemistry of Diferrocenyl Porphyrins
Organometallics, Vol. 26, No. 1, 2007 109
with Vicor tip. This Luggin was immersed into a second Luggin
capillary that contained a 0.100 M [N(nBu)4][B(C6F5)4] dichloro-
methane solution. All measurements were conducted under a blanket
of argon at 25 °C in a Faraday cage connected to a BAS 100 B/W
electrochemical workstation interfaced with a personal computer.
Data, uncorrected for junction potentials, were collected with
standard BAS 100 software and exported to Excel for manipulation
and analyses. Successive experiments under the same experimental
conditions showed that all formal reduction and oxidation potentials
were reproducible within 5 mV. All potentials in this study were
experimentally referenced against the Ag/Ag+ couple but were then
manipulated on Excel to be referenced against Fc/Fc+ as recom-
mended by IUPAC.26 Under our conditions the Fc/Fc+ couple
exhibited ipc/ipa ) 0.94, ∆Ep ) 81 mV, and E°′ ) 0.087 V versus
Ag/Ag+. All temperatures were kept constant to within 0.2 °C.
Acknowledgment. The authors acknowledge the NRF under
grant 2054243 and the UFS for financial support.
OM060373U