Meso-aryl phenanthroporphyrins: Synthesis and spectroscopic properties

Article abstract of DOI:10.1002/chem.201002596

The successful synthesis of tetraphenyltetraphenanthroporphyrin (TPTPhenP; 5a) in 2006 under modified Rothemund-Lindsey conditions yielded a tetraphenyl porphyrinoid with a B band redshifted to an unprecedented 576 nm. Radially symmetric fused-ring expans

Full text of DOI:10.1002/chem.201002596

FULL PAPER
DOI: 10.1002/chem.201002596
meso-Aryl Phenanthroporphyrins: Synthesis and Spectroscopic Properties
Hai-Jun Xu,[a] John Mack,[b] Ana B. Descalzo,[c, d] Zhen Shen,*[a] Nagao Kobayashi,*[b]
Xiao-Zeng You,[a] and Knut Rurack*[c]
ꢀ
ꢀ
Abstract: The successful synthesis of
tetraphenyltetraphenanthroporphyrin
(TPTPhenP; 5a) in 2006 under modi-
fied Rothemund–Lindsey conditions
based on a perimeter model approach
that makes use of time-dependent DFT
calculations and magnetic circular di-
chroism spectroscopy and also based
on a detailed analysis of the fluores-
cence emission. Attempts to introduce
substituents at the ortho and meta posi-
tions of the meso-phenyl groups and to
insert a central metal proved unsuc-
cessful. The synthesis of a series of
TPTPhenPs with strong electron-with-
drawing ( CN, NO2) and -donating
ꢀ
ꢀ
G
NACHTNGUTERNU(NG CH3)2) substituents at the
para positions of the meso-phenyl rings
is reported. Marked redshifts of the
main spectral bands were consistently
observed. The most pronounced spec-
tral changes were observed with N-
(CH3)2 groups (5i) due to a marked de-
stabilization of the HOMO, which has
large MO coefficients on the meso-
carbon atoms. Protonation of 5i at
both the ligand core and at the N-
(CH3)2 groups resulted in unprecedent-
yielded
a tetraphenyl porphyrinoid
with a B band redshifted to an unpre-
cedented 576 nm. Radially symmetric
fused-ring expansion of tetraphenylpor-
phyrin with phenanthrene moieties re-
sults in very deep saddling due to steric
crowding and very marked redshifts of
the Q and B (or Soret) porphyrinoid
absorption bands. The extent to which
the TPTPhenP structure can be further
modified is explored, and the optical
properties of TPTPhenPs are analyzed
ꢀ
A
ꢀ
A
Keywords: charge transfer · circu-
ed Q00 band absorption at wavelengths
greater than 1200 nm.
lar dichroism
· dyes/pigments ·
porphyrinoids · protonation
Introduction
cationic or anionic analytes,[2] and as dyes for applications in
nonlinear optics (NLO),[3] optical limiting,[4] opto-electronic
devices,[5] and solar-energy conversion.[6] A redshift of the
absorption and emission bands of porphyrins can usually be
achieved through an extension of the p-conjugation system.
A number of different approaches have been adopted over
the last decade to achieve this, including the introduction of
meso-alkynyl substituents;[7] an increase in the number of
core pyrrole rings to form sapphyrins, hexaphyrins, octa-
phyrins, and so on;[8] coplanar polymerization of porphy-
The study of redshifted porphyrinoid chromophores has at-
tracted considerable attention in recent years because of
their potential use as photosensitizers in photodynamic ther-
apy (PDT),[1] as fluorescent probes for the recognition of
[a] Dr. H.-J. Xu, Prof. Z. Shen, Prof. X.-Z. You
State Key Laboratory of Coordination Chemistry
Nanjing National Laboratory of Microstructures
School of Chemistry and Chemical Engineering
Nanjing University, Nanjing 210093 (P.R. China)
Fax : (+86)25-8331-4502
rins;[9] the expansion of the
AHCTUNGRETNNUNG p system with exocyclic
rings;[10,11] and conformational distortions due to steric
crowding of peripheral substituents.[12,13] Although the
fusion of exocyclic rings at the b-pyrrole positions is proba-
bly the most obvious strategy for modifying the electronic
and optical properties of the p system on this basis, only
limited redshifts have been observed upon substitution with
fused benzene rings.[10] The radially symmetric addition of a
E-mail: zshen@nju.edu.cn
[b] Dr. J. Mack, Prof. N. Kobayashi
Department of Chemistry, Graduate School of Science
Tohoku University, Sendai 980, 8578 (Japan)
Fax : (+81)22-795-7719
E-mail: nagaok@m.tohoku.ac.jp
[c] Dr. A. B. Descalzo, Dr. K. Rurack
Fachgruppe 1.5, BAM Bundesanstalt
fꢀr Materialforschung und -prꢀfung
Richard-Willstꢁtter-Strasse 11, 12489 Berlin (Germany)
Fax : (+49)30-8104-1157
second
set
of
benzo
groups
to
form
tetra-
AHCTUNGTREG[NNNU 2,3]naphthoporphyrin (TNP) from tetrabenzoporphyrin
(TBP) (Scheme 1) results in only a minor redshift of the B
(or Soret) band from 433 to 439 nm.[14,15] Spence and Lash
have demonstrated that the introduction of peripheral ace-
naphthalene rings has a far larger impact, and that further
redshifts are observed when phenyl substituents are added
at the meso-carbon atoms.[16] They reported that the B band
of the doubly protonated form of tetraacenaphthoporphyrin
(TANPH22+) in CHCl3 shifts significantly to the red from
528 nm[17] to around 565 nm in the case of meso-tetraphenyl-
E-mail: knut.rurack@bam.de
[d] Dr. A. B. Descalzo
Present address: Departamento de Quꢂmica Orgꢃnica
Facultad de Ciencias Quꢂmicas
Universidad Complutense de Madrid
Avda. Complutense s/n, 28040 Madrid (Spain)
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201002596.
Chem. Eur. J. 2011, 17, 8965 – 8983
ꢄ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
8965

A
by
extending this approach to tet-
raphenanthroporphyrin
(TPhenP) since this can be ex-
pected to yield a porphyrinoid
with a still higher degree of
ligand nonplanarity. Although
TPhenP was successfully pre-
pared some years ago,[17,22] the
preparation of meso-tetrasubsti-
tuted tetraphenanthroporphy-
AHCTUNGTRENGNrUN ins initially proved unsuccess-
ful under normal Lindsey por-
phyrin condensation condi-
tions[16] due to the steric hin-
drance between the meso-aryl
substituents and the relatively
bulky
phenanthrene
rings
(Scheme 1). Recently, we were
able to synthesize meso-tetra-
AHCTUNGTRENNUNG
a
modification
to
Lindseyꢅs
Scheme 1. The molecular structures of porphyrinoid compounds discussed in this paper including porphyrin
(P), tetraphenylporphyrin (TPP), meso-tetrakis(4-N,N-dimethylaminophenyl)porphyrin (6i), tetrabenzopor-
phyrin (TBP), tetraphenyltetrabenzoporphyrin (TPTBP), tetranaphthoporphyrin (TNP), tetraphenyltetranaph-
thoporphyrin (TPTNP), tetraacenaphthoporphyrin (TANP), tetraphenyltetraacenaphthoporphyrin (TPTANP),
tetraphenanthroporphyrin (TPhenP), tetraphenyltetraphenanthroporphyrin (TPTPhenP, 5a), and dodecaphe-
nylporphyrin (DPP). For title compounds 5a–i, see Scheme 2.
method.[22] In this paper, we ex-
plore the effect of modifying
the
aryl
substituents
of
TPhenPs by preparing a range
of TPTPhenP compounds in-
cluding
meso-(4-halogeno-
tetraacenaphthoporphyrin (TPTANPH22+).[16] The role of
meso substitution clearly extends beyond an enhancement
of the solubility properties[18] in this context.
ACHTUNGTRENtNUNG etraphenyl) derivatives (5c–f) and newly synthesized com-
pounds with strongly electron-donating and -withdrawing
groups at the para position of the meso-phenyl rings (5b,
5g–i; Scheme 2). A detailed analysis of the electronic struc-
tures and optical spectroscopy of these compounds is carried
out based on an application of Michlꢅs perimeter model.[20]
Mack et al.[19] recently demonstrated—based on an appli-
cation of Michlꢅs perimeter model[20] to the magnetic circu-
lar dichroism (MCD) spectroscopic analysis and theoretical
calculations of a wide range of radially symmetric zinc por-
phyrinoids—that key trends observed in the spectral proper-
ties of porphyrinoids are largely determined by the relative
energies of the four frontier p MOs associated with Gouter-
manꢅs 4-orbital model.[21] These MOs retain the nodal prop-
erties and hence the orbital angular momentum properties
of a parent hydrocarbon polyene associated with the inner
perimeter of the ligand. Saddled porphyrinoids such as
ZnTPTANP exhibit large redshifts in the B-band region due
to limited configurational interaction between the B and
higher-energy p–p* states. A key issue faced by the use of
TPTANP in near-infrared (NIR) dye applications is that the
forbidden and allowed nature of the Q and B bands[21] re-
sults in relatively weak Q00-band absorption. Building on
this earlier research, our aim has been to introduce structur-
al perturbations to the porphyrin p system, which modify
the energies of the frontier p MOs in a predictable manner,
so that rationally designed porphyrinoids can be synthesized
with markedly redshifted and relatively intense Q bands.
The next logical step was to build on recent research by
the Lash and Lukꢅyanets groups on the meso-tetraaryl sub-
stitution of tetraacenaphthoporphyrins[16,18] and tetra-
Electronic structure of porphyrinoids: Many of the key
breakthroughs in understanding the electronic structures of
porphyrinoids have been based on the application of MCD
spectroscopy.[23] It currently is not possible to calculate
MCD spectra by using commercially available DFT-based
software packages. Older semiempirical approaches contin-
ue to play a major role in the analysis of MCD spectra,
since they provide a readily accessible conceptual frame-
work for the experimentalist that can enable the prediction
of key trends in the spectral properties and thus facilitate
the rational design of porphyrinoids for practical applica-
tions. Moffitt[24] and Michl[20] demonstrated that the relative
intensities of the major electronic absorption bands of aro-
matic p systems can be successfully described in terms of
perturbations to the structure of a high-symmetry parent hy-
2ꢀ
drocarbon (C16H16 in the case of metal tetrapyrrole por-
phyrinoid complexes, or C18H18 for free-base compounds).
The nodal patterns of the p-system MOs are retained even
when the symmetry of the cyclic perimeter is modified
(Figure 1 and Figure S10 in the Supporting Information). As
a consequence, there is an ML =0, ꢁ1, ꢁ2, ꢁ3, ꢁ4, ꢁ5, ꢁ6,
8966
www.chemeurj.org
ꢄ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 8965 – 8983

Phenanthroporphyrins
FULL PAPER
structural perturbation intro-
AHCTUNGTRENNUNG
a
mixing of the allowed and for-
bidden properties of the Q and
B bands and a marked intensifi-
cation of the Q band, which as
a consequence can become the
dominant spectral feature for
some porphyrinoids such as the
phthalocyanines.[25]
MCD spectroscopy[23] can be
used to identify the main elec-
tronic bands and to derive
structural
information
that
cannot be obtained from UV/
Vis absorption and NMR spec-
tra alone, since the signs and in-
tensities of the spectral bands
are determined by the OAM
properties of the ground and
excited states. Analysis of the
MCD spectrum is based on in-
tensity mechanisms described
Scheme 2. Scheme for the preparation of 5a (Ar=C6H5), 5b (Ar=4-Me-C6H4), 5c (Ar=4-F-C6H4), 5d (Ar=
4-Cl-C6H4), 5e (Ar=4-Br-C6H4), 5 f (Ar=4-I-C6H4), 5g (Ar=4-NC-C6H4), 5h (Ar=4-O2N-C6H4) and 5i
(Ar=4-Me2N-C6H4). Compounds 5j (Ar=3-O2N-C6H4) and 5k (Ar=2,5-Cl2-C6H3) could not be obtained on
this basis.
ꢁ7, 8 sequence in ascending energy terms in the p MOs of
metal porphyrinoids in an analogous manner to the ML =0,
ꢁ1, ꢁ2, 3 p MO sequence that forms the basis of the aro-
matic properties of benzene. Within the band nomenclature
of Goutermanꢅs 4-orbital model,[21] there is an electronically
allowed B transition and a forbidden Q transition that links
the frontier ML =ꢁ4, ꢁ5 p MOs based on DML =ꢁ1 and
ꢁ9 b transitions, respectively. Electromagnetic radiation can
rotate a maximum of once per wavelength, so only DML =0,
ꢁ1 transitions are electric-dipole-allowed. The incorpora-
tion of four pyrrole rings into the parent perimeter to form
planar metal porphyrin (MP) complexes reduces the symme-
try from D16h to D4h, which is further lowered to D2h in the
context of free-base compounds. The LUMO and LUMO+1
of MP complexes retain ML =ꢁ5 nodal patterns (Figure 1)
and are orbitally degenerate in the context of metal com-
plexes (1eg*) with only a slight splitting typically predicted
in the context of free-base compounds (1b2g* and 1b3g*).[25]
The MOs associated with the HOMO of the parent perime-
ter can be readily identified based on their ML =ꢁ4 nodal
patterns and are symmetry split under both D4h (1a1u and
1a2u) or D2h (1au and 1b1u) symmetry. In the context of por-
phyrins, these two MOs remain accidentally near degener-
ate, and the orbital angular momentum (OAM) properties
of the lowest-lying p–p* excited states therefore remain rel-
atively close to those of the parent hydrocarbon perimeter.
Since there are only minor splittings of the doubly degener-
ate HOMO and LUMO of the parent perimeter (referred to
as the DHOMO and DLUMO values by Michl[20]; Figure 2),
the Q bands that lay at lower energy remain almost fully
forbidden and there is typically very weak electronic absorp-
tion intensity in the Q-band region based primarily on vibra-
tional borrowing from the allowed B00 bands.[26] When a
by the Faraday a1, b0, and c0 terms.[23] Since the main p!
p* bands of D4h-symmetry MPs arise from transitions out of
orbitally nondegenerate 1a1u and 1a2u MOs into an orbitally
degenerate 1eg* LUMO, there is a nondegenerate ground
state (1A1g) and orbitally degenerate excited states (1Eu).
The MCD intensity mechanism is based on the differential
absorbance of left and right circularly polarized light (lcp
and rcp) (DAlꢀr), with lcp being absorbed during DML =+1
transitions and rcp being absorbed during DML =ꢀ1 transi-
tions. When a magnetic field is applied parallel to the axis
of light propagation, there is a splitting of the orbitally de-
1
generate Eu p–p* excited states into MJ +1 microstates and
a first derivative-shaped Faraday a1 term is observed due to
the Zeeman splitting of states. If a structural modification
removes the main four-fold axis of symmetry so that there is
a complete lifting of ground- and excited-state degeneracies,
coupled oppositely signed and Gaussian-shaped Faraday b0
terms replace the derivative-shaped Faraday a1 terms,
based on field-induced mixing of close-lying excited states
linked by a magnetic dipole transition moment. When the
DLUMO value is small relative to the bandwidths of the
Gaussian-shaped b0 terms, a pseudo-a1 term is formed,
which can be difficult to distinguish from an a1 term. Al-
though free-base TPTPhenPs 5a–i lack a four-fold axis of
symmetry and are predicted to have deeply saddled struc-
tures (Figure 3), a1 terms are anticipated in the spectra of
protonated dication species, since an S4 axis of symmetry is
retained.
Key information about the electronic structure can be de-
rived from the sign sequences observed in the a1 terms or
coupled oppositely signed b0 terms associated with the Q
and B bands, since the ML =ꢁ1 property of the incident lcp
and rcp photons must be conserved based on Newtonꢅs third
Chem. Eur. J. 2011, 17, 8965 – 8983
ꢄ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
8967

Z. Shen, N. Kobayashi, K. Rurack et al.
Figure 2. Left: Michlꢅs perimeter model[20] for C16H162ꢀ. The circle repre-
sents a diagrammatic representation of the clockwise and counterclock-
wise motion of p-system electrons on the inner ligand perimeter that gen-
erate the ML value for each complex p MO. Center: The alignment and
magnitude of the magnetic moments induced by the electron motion
within each p MO can be predicted based on the right-hand rule and
linear combination of atomic orbitals (LCAO) calculations. The moments
aligned along the z axis with or against the applied field are shown dia-
grammatically.[20] Right- and left-handedness is defined within classical
optics looking towards the light source of the CD spectrometer (lcp and
rcp=left and right circularly polarized light, respectively).
Figure 1. Frontier p MOs of a C18H18 parent hydrocarbon perimeter, por-
phyrin (P), tetraphenylporphyrin (TPP), tetrabenzoporphyrin (TBP), and
tetraphenyltetraphenanthroporphyrin (TPTPhenP, 5a) at an isosurface
value of 0.03 a.u. The same nodal patterns are consistently observed re-
flecting the ML =ꢁ4, ꢁ5 properties of the four frontier p MOs of fused-
ring-expanded porphyrinoids. Under the D2h symmetry of P, the MOs
from left to right are 1b3g*, 1b2g*, 1b1u, and 1au, respectively.
Figure 3. The B3LYP-optimized geometries of (top) Zn5a and (bottom)
5i exhibit a deep saddled conformation when viewed along the (left) z
and (right) y axes.
law of motion.[20,23] When an electron is electronically excit-
ed from the HOMO to the LUMO, the circular motion of
the electron on the perimeter is no longer balanced by that
of an electron in an MO of the opposing-handedness proper-
ties. Absorption of a photon by a cyclic p system, therefore,
results in a circular redistribution of charge and hence to a
magnetic moment aligned along the z axis perpendicular to
the p system (Figure 2). The alignment of the magnetic mo-
ments with or against the applied field is determined by the
OAM properties of the frontier p MOs derived from the
HOMO and LUMO of the parent perimeter. When the
OAM is greater in the LUMO level than in the HOMO
level, as is normally the case for most porphyrinoids, a nega-
tive particle is effectively formed on the perimeter, which
circulates with the left- or right-handed motion of the inci-
dent photon. In contrast, when there is a marked quenching
of the OAM of the LUMO level due to the introduction of
a large DLUMO value by a structural perturbation, a posi-
tive particle circulates instead based on the hole left within
the HOMO level. Application of Ampereꢅs rule for sole-
noids can be used to demonstrate that when the LUMO
level has greater OAM, a positive Faraday a1 term (i.e., a
8968
www.chemeurj.org
ꢄ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 8965 – 8983

Phenanthroporphyrins
FULL PAPER
negative/positive (+ve/ꢀve) sign sequence in ascending
energy terms) is formed within the MCD spectrum, since
the microstate associated with the absorption of rcp (DML =
ꢀ1) is stabilized as the induced magnetic moment is aligned
with the applied field, whereas the microstate associated
with the absorption of lcp (DML =+1) is destabilized as the
induced magnetic moment is aligned against the field.
5b–i were obtained in 10–19% yields. In contrast, porphyr-
ins 5j and 5k could not be formed from 3-nitrobenzalde-
hyde and 2,5-dichlorobenzaldehyde under these conditions
due to the increased steric hindrance associated with sub-
stituents at the ortho and meta positions. In contrast with
the analogous TPTANP compounds, repeated attempts to
synthesize the metal complexes of 5b–i based on metal in-
sertion reactions with metal salts also consistently failed.
Michl[23] referred to cyclic polyenes in which DHOMO,
DLUMOꢂ0 (such as TPTANPs) as soft MCD chromo-
phores,[19] since small structural modifications can potentially
reverse the alignment of the induced magnetic moments of
the p–p* excited states and hence the sign sequence ob-
served for the Faraday a1 terms or coupled oppositely
signed Faraday b0 terms associated with the main Q00 and
B00 bands. In the context of TPTANPs, metal complexes
were found to exhibit anomalous negative Faraday a1 terms
(i.e., a +ve/ꢀve sign sequence in ascending energy terms),
whereas the normal ꢀve/+ve sign sequence was observed
for free-base compounds and dication species.[19] The unusu-
al Faraday a1 terms in the MCD spectra of the metal com-
plexes are believed to arise, because ligand saddling quench-
es the OAM properties of the LUMO level to a greater
extent than those of the HOMO level. The availability of
the more deeply saddled TPTPhenPs (Figure 3) provides ad-
ditional insight into the effect of ligand saddling on the elec-
tronic structure and optical properties of porphyrinoids.
Recent studies on core-modified tetrabenzoporphyrins[27]
and a-phenylated phthalocyanines[28] have demonstrated
that the results of DFT geometry optimizations are very
similar to those obtained by X-ray crystallography when
steric hindrance between peripheral fused rings results in
severe ligand folding due to the lack of conformational flexi-
bility. Trends observed in the experimental data for
TPTPhenPs can, therefore, be readily compared to those
predicted in theoretical calculations for B3LYP-optimized
geometries, even in the absence of X-ray structures.
Characterization of 5a–i: The compounds are all moderately
to weakly soluble in most organic solvents with reddish-pink
(e.g. 5a, 5c), greenish-blue (5 f, 5h), and blue (5g, 5i) colors.
The structures were characterized by MALDI-TOF MS,
1H NMR spectroscopy, and elemental analysis (see the Ex-
perimental Section and Figures S1–S9 in the Supporting In-
formation). For example, the MALDI-TOF mass spectrum
of 5g contains the anticipated molecular ion peak
1
N
CDCl3 in the presence of 1% deuterated trifluoroacetic acid
(Figure 4) contains six peaks between d=6.8–8.9 ppm,
Figure 4. Partial 500 MHz 1H NMR spectrum of 5g in CDCl3 containing
1% deuterated trifluoroacetic acid. The assignment of the proton signal
is based on the integrated intensities and coupling patterns.
Results and Discussion
Synthesis of meso-tetraaryl tetraphenanthroporphyrins: A
modified version of the Lindsey method[29] was used to pre-
pare a series of TPTPhenPs (5a–i) (Scheme 2). Phenanthro-
ACHTUNGTRENNUNG[9,10]pyrrole (2) is readily available based on a Barton–
Zard reaction of 9-nitrophenanthrene (1) with ethyl isocya-
which is consistent with the anticipated symmetric aromatic
structure. The phenanthrene protons closest to the meso-
phenyl rings are shielded relative to those in TPhenP and
can, therefore, be assigned to the upfield triplet at d=
6.86 ppm.[31] The doublets at d=8.89 and 8.48 ppm, respec-
tively, can be assigned to the ortho and meta protons of the
meso-phenyl rings, since they are deshielded by both the
porphyrin macrocyclic and phenanthrene ring currents. The
remaining protons lie in the typical aromatic range between
d=7.0–7.5 ppm. These results for 5g, and the broadly simi-
noacetate in the presence of the non-nucleophilic base 1,8-
diazabicycloACHTUNGTRENNUNG
[5.4.0]undec-7-ene (DBU).[30] The ethyl ester
group can be removed by heating 2 with potassium hydrox-
ide in ethylene glycol at 1708C to afford the tetracyclic com-
pound 3 in excellent yield. A BF3·OEt2-catalyzed condensa-
tion of 3 with benzaldehyde in dry CH2Cl2 at ꢀ508C for 3 h
and subsequently at room temperature for 24 h generated a
porphyrinogen (4), which was oxidized with 2,3-dichloro-5,6-
dicyano-1,4-benzoquinone (DDQ) to afford the target por-
phyrin compound 5a in 15% yield.[22] A series of 4-substitut-
ed benzaldehydes (R=F, Cl, Br, I, Me, CN, NO2, NMe2)
were condensed with 3 as described above and porphyrins
1
lar H NMR spectra of 5a–f, 5h, and 5i (Figures S1–S6, S8,
and S9 in the Supporting Information), demonstrate that the
structures of TPTPhenPs have reasonably high symmetry
similar to those reported during recent studies on
1
TPTANPs.[16,18] The H NMR spectra are consistent with the
B3LYP geometry optimizations for 5a and 5i,[32] which pre-
Chem. Eur. J. 2011, 17, 8965 – 8983
ꢄ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
8969

Z. Shen, N. Kobayashi, K. Rurack et al.
dict a deep saddlelike distorted conformation of the p
system that results in a C2v symmetry (Figure 3). Despite re-
peated efforts, single crystals suitable for X-ray crystallogra-
phy could not be obtained. It should be noted that no X-ray
structures have been reported for the similarly deeply sad-
dled TPTANPs[10,16,18,19] probably due to the impact of the
severe ligand saddling on the p–p stacking properties in the
solid phase.
Absorption, MCD, and fluorescence spectroscopy of free-
base 5a–i: The optical spectra of 5a–i in CHCl3 are provid-
ed in Figures 5 and 6. Key data related to the major spectral
bands are summarized in Table 1. The B bands are located
at longer wavelengths than is observed for comparable
Figure 6. Absorption and fluorescence spectra of 5a and 5c–f in CHCl3
at 298 K. The fluorescence bands are normalized on the lowest-energy
absorption maxima.
Table 1. Absorption and fluorescence properties of 5a–i in CHCl3 at
298 K.
B band
Q band
Fluorescence
Ff
labs
loge
labs
loge
lem
tf
[nm]
[nm]
[nm]
A
]
[ns]
ꢀ
5a
5b
5c
5d
5e
5 f
5g
5h
5i
H
576
584
576
582
583
586
596
603
595
4.92
4.97
4.96
4.95
4.93
4.82
4.98
4.98
4.87
794
803
794
805
801
805
798
804
931
4.14
4.31
4.14
4.17
4.14
4.06
4.14
4.18
4.33
831
851
833
870
888
892
907
924
1004
5.7
5.2
6.0
3.9
3.1
2.0
2.3
1.7
1.2
0.38
0.32
0.41
0.25
0.21
0.14
0.19
0.17
0.14
ꢀ
CH3
ꢀ
F
ꢀ
Cl
ꢀ
Br
ꢀ
I
ꢀ
CN
ꢀ
NO2
ꢀ
NACTHNURGTENUNG(CH3)2
TPTANP compounds (575–605 nm) and are redshifted by
more than 150 nm relative to the parent meso-tetraphenyl-
porphyrin (TPP) compound,[10] although they retain high
molar absorption coefficients of loge>4.9 despite the deep
saddling of the ligand. The MO coefficients of the frontier p
MOs are largely located on the inner ligand perimeter
(Figure 1), so there is still considerable overlap between the
MOs. Weaker Q00 and Q01 absorption bands are observed in
the NIR region. The Q00 bands have comparable MCD in-
tensity to the B bands due to the larger magnetic moments
associated with the forbidden (DML =ꢁ9) transition. In con-
trast with most planar porphyrins[26] and with TPTANPs,[19]
the absorption intensity of the Q00 bands is greater than that
of the Q01 vibrational bands. This suggests that the Q00
bands become partially allowed due to a significant splitting
of the MOs derived from the HOMO of the parent perime-
ter, and that vibrational borrowing from the allowed B band
is therefore not the main intensity mechanism in the Q-band
Figure 5. Absorption and MCD spectra of 5a, 5b, and 5 f–i recorded in
99:1 (v/v) CHCl3/ethanol at 298 K. The coupled pairs of overlapping
Faraday b0 terms associated with the Q and B bands can be viewed as
pseudo-a1 terms that arise from transitions into close-lying excited states
that would be orbitally degenerate if a central metal were present.
8970
www.chemeurj.org
ꢄ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 8965 – 8983

Phenanthroporphyrins
FULL PAPER
region. The MCD intensity to the blue of the B band is very
weak, since the MOs involved tend to be associated with the
peripheral phenanthrene moieties rather than the inner pe-
rimeter of the p system. The OAM properties of the higher-
energy excited states are, therefore, largely quenched by the
ligand saddling. Across the entire 5a–i series, a single in-
tense set of oppositely signed b0 terms is observed for the
Q00 and B bands of each spectrum with a ꢀve/+ve/ꢀve/+
ve sign sequence in ascending energy terms as has been re-
ported previously for free-base TPTANP. This is the pattern
anticipated when the DHOMO value is greater than the
DLUMO value and the quenching of the OAM properties
of the HOMO level of the parent perimeter is greater than
that of the LUMO level.[19,20]
with an internal heavy-atom effect that has been reported
previously for porphyrins.[39–42] Phosphorescence could not
be detected for any of the dyes probably due either to the
limited detection range (ꢃ1100 nm) of the instrument em-
ployed or to the weakness of such luminescence due to the
marked redshift of the lowest-energy absorption band.
2+
Absorption and fluorescence spectroscopy of 5aH2
–
5hH22+: Compound 5i possesses four peripheral sites at the
ꢀ
N
N
can, therefore, be readily transformed from being strongly
electron-donating to strongly electron-withdrawing. Before
discussing 5i in detail, it is important to first understand the
spectral changes that occur when the inner perimeters of the
p systems of 5a–5h are protonated to form conventional di-
cation species. Data were obtained in the presence of 1%
trifluoroacetic acid (TFA) so that only doubly protonated
species are present in solution (Table 2 and Figure 7). The B
In the series 5a, 5c–f, slight redshifts of 5–10 nm are ob-
served for the B and Q00 bands, in an order consistent with
the electron-withdrawing ability of the 4-halogeno substitu-
ent. In the near-infrared region (NIR), there are weak Q01
vibrational bands (ꢂ720 nm) and more intense Q00 bands
(ꢂ800 nm) with around 5–10% and 15% of the intensity of
Table 2. Absorption and fluorescence properties of protonated species of
5a–i in CHCl3 at 298 K.
the corresponding B bands, respectively. Substitution with
strongly electron-withdrawing and -donating groups results
in more pronounced changes to the TPTPhenP spectrum. In
B band
Q band
Fluorescence
labs
loge
labs
loge
lem
Ff
[ꢆ10ꢀ3
]
ꢀ
the presence of strongly electron-withdrawing CN (5g) and
[nm]
[nm]
[nm]
U
2+
2+
ꢀ
NO2 (5h) para substituents, there is a greater splitting of
ꢀ
5aH2
5bH2
H
579
587
579
584
586
589
596
602
636
580
5.04
5.01
5.06
5.04
5.02
4.99
5.18
5.11
4.70
n.d.[b]
803
816
804
807
808
811
803
806
1015
784
4.21
4.34
4.25
4.32
4.33
4.29
4.28
4.38
4.55
n.d.[b]
838
858
837
842
844
851
834
839
n.d.[a]
810
8.0
ꢀ
CH3
10.0
10.1
6.8
6.5
3.9
4.1
4.4
–
the x- and y-polarized Q00 bands, and the lower-energy Q00
band is observed as a broad shoulder between 850–970 nm.
The B bands also shift markedly to 596 and 603 nm, respec-
tively. The largest spectral changes are observed with strong-
ꢀ
2+
ꢀ
5cH2
5dH2
F
2+
2+
ꢀ
Cl
ꢀ
5eH2
5 fH2
Br
2+
ꢀ
I
2+
ꢀ
ly electron-donating
NACHTNGUTERNU(NG CH3)2 groups (5i). There is a
5gH2
5hH2
CN
2+
ꢀ
NO2
marked redshift and intensification of the Q00 bands, where-
as in contrast the B band lies at the same wavelength as that
of 5b and 5g and is considerably broadened and lower in in-
tensity.
3+
ꢀ
5iH3
5iH6
NACHTNGUTRENNU(G CH3)2
6+
ACHTUNGTRENNUNG
(4.0)[b]
ꢀ
N
A
[a] No fluorescence could be detected since the detection system could
only be employed up to 1100 nm. [b] Could not be reliably determined.
All of the TPTPhenPs reported exhibit NIR fluorescence
with quantum yields between 0.1 and 1.0% (Table 1). The
fluorescence spectra generally form an approximate mirror
image of the Q00-band absorption spectra with maxima be-
tween 830 nm (5a, 5c) and 1004 nm (5i), which represents
an unprecedented wavelength for tetrapyrrole porphy-
2+
bands of 5aH22+–5 fH2 shift slightly to the red by around
3 nm while gaining markedly in intensity. No redshifts are
observed for 5gH2 and 5hH22+. Similar trends are ob-
2+
served in the Q bands of TPTPhenPs with weakly electron-
donating or -accepting para substituents. Redshifts of 7–
A
served.[33] The fluorescence lifetimes are generally monoex-
ponential and range between 0.14 and 0.41 ns for 5i and 5c,
respectively. This suggests that the emissive state is rapidly
quenched by intramolecular processes. The fluorescence
quantum yield and lifetime data are of the same order of
magnitude as those found for other highly distorted por-
phyrins such as dodecaphenylporphyrin[34–36] (DPP): Ff =2ꢆ
10ꢀ3, tf =0.095 ns in CHCl3.[36] As is the case with DPP, the
relatively low emissivity of these compounds can be ascribed
to the high degree of nonplanarity,[13] which facilitates inter-
system crossing and internal conversion, and the growing in-
fluence of internal conversion as a function of the shift of
the emission band into the NIR region due to the energy-
gap rule.[37] The increase in the rate constant of nonradiative
deactivation[38] observed for the 4-halogeno compounds, 2.4
(5c)<3.9 (5d)<4.7 (5e)<7.0ꢆ109 sꢀ1 (5 f), is consistent
15 nm and moderate intensity increases are observed in the
2+
spectra of 5aH22+–5 fH22+. The Q bands of 5gH2
and
2+
5hH2 narrow significantly and, in contrast with the free-
base spectra, display features very similar to those of
2+
5aH22+–5 fH2 (Figures 5 and 7). Despite the marked dif-
ferences in the electron-withdrawing and -donating proper-
ties of the para substituents, the absorption and MCD spec-
2+
6+
tra of 5aH22+, 5 fH22+, 5gH2 and 5iH6 are almost identi-
cal, in marked contrast with the corresponding free-base
spectra. Positive Faraday a1 terms are consistently observed
for the Q00, Q01, and B00 bands. The changes in the fluores-
cence spectra closely reflect those observed in the absorp-
tion spectra. Virtually identical emission bands are observed
2+
for 5aH22+–5hH2 at (846ꢁ12) nm with Stokes shifts of
(530ꢁ70) cmꢀ1. The fluorescence quantum yields are be-
tween 150–250% higher than those of the corresponding
Chem. Eur. J. 2011, 17, 8965 – 8983
ꢄ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
8971

Z. Shen, N. Kobayashi, K. Rurack et al.
LUMO level that is believed to cause the anomalous nega-
tive a1 terms observed in the MCD spectra of TPTANP
metal complexes.[19]
Absorption, MCD, and fluorescence spectroscopy of the
protonated species of 5i and theoretical calculations: Titra-
tions were carried out for the protonation of 5i in the pres-
ence of TFA (Figure 8). Up to six protons can be added,
ꢀ
since the four
NACHTNGUTERUNN(G CH3)2 groups are susceptible to protona-
6+
tion. Complete protonation to form the 5iH6 was achieved
in a 99:1 CHCl3/ethanol solvent mixture. DFT geometry op-
timizations and TD-DFT calculations of TPTPhenPs are ex-
pensive in terms of machine time, so they have only been
carried out for 5a, 5i, and 5iH22+. Additional calculations
were carried out on the corresponding TPP model complex
with no peripheral phenanthrene rings, meso-tetrakis(4-N,N-
dimethylaminophenyl)porphyrin (6i) (Figure 9; for chemical
structure, see Scheme 1). The spectra of free-base 5i and of
the protonated species are expected to be broadly similar to
those of 6i since the absorption spectrum of 5a is very simi-
lar to that of TPP with the exception of the marked redshift
of the Q and B bands. The additional information provided
by MCD spectroscopy can be used to identify the spectra of
each protonated species during the titration of 5i. Since
there is no three-fold or higher axis of symmetry and hence
no orbitally degenerate excited states, coupled oppositely
2+
3+
Figure 7. Absorption and MCD spectra of 5aH22+, 5 fH22+, and 5gH2
recorded in 99:1 (v/v) CHCl3/ethanol that contained 1% of TFA; 5iH3
signed and Gaussian-shaped x- and y-polarized b0 terms are
4+
anticipated in the MCD spectra of 5i, 5iH33+, 5iH4
,
adj
obtained during a titration of a solution of 5i in 99:1 (v/v) CHCl3/ethanol
after the addition of 50 equiv of TFA; and 5iH6 recorded in pure TFA.
5iH4 opp, and 5iH55+, whereas x/y-polarized and first-deriv-
4+
6+
The close correspondence between the absorption band maxima and the
zero point of the first-derivative-shaped MCD band is consistent with the
presence of three Faraday a1 terms that arise from x/y-polarized Q00,
Q01, and B00 bands.
ative shaped a1 terms are anticipated in the spectra of
2+
6+
5iH2 and 5iH6 due to the presence of an S4 axis of sym-
metry (Figure 8).[46] The visible and near-IR regions of the
6+
calculated spectra of 5i and 5iH6 are dominated by the Q
and B bands. TD-DFT and ZINDO/s calculations predict
free bases, as has been reported previously for other highly
sterically hindered porphyrins such as DPP (DPPH22+: Ff =
7ꢆ10ꢀ3, tf =0.33 ns in CHCl3).[36]
the presence of additional coupled and oppositely signed b0
4+
terms or a1 terms in the spectra of 5iH22+, 5iH33+, 5iH4
4+
5+
adj, 5iH4 opp, and 5iH5 based on configurational interac-
tion between the Q and B excited states and states associat-
ed with two meso-aryl substituent MOs, which have the
same symmetry properties as the MOs that arise from the
HOMO of the parent perimeter (Figures 9 and 10; and Fig-
ures S10–S12 in the Supporting Information). These MOs
are markedly destabilized upon partial protonation of the
ꢀ
2+
The B3LYP geometry optimization for 5iH2 (Table 3)
predicts that the electrostatic repulsion introduced by the
two additional protons on the inner ligand perimeter leads
to a further deepening of the saddled structure. Similar ef-
fects have previously been reported for tetraphenylporphyri-
noids with the meso-phenyl rings tilting into the plane of the
inner perimeter of the p system.[43,44] A marked intensifica-
peripheral
NACHTNGUTERNU(NG CH3)2 groups (Figure 10).
tion of the Q-band region has been reported on this basis
The most striking features of the initial stages of the titra-
tion spectra are the intense band centered at around
1000 nm and a weaker, extremely redshifted band between
1150–1300 nm region (Figure 8). These bands can be as-
signed as Q00 bands, since a marked destabilization of the
HOMO of 6i is predicted upon protonation (Figure 10 and
Figure S12 in the Supporting Information). The energy of
the other MO that arises from the doubly degenerate
HOMO of the parent perimeter (i.e., the 1au MO of P) re-
mains relatively constant due to the lack of significant MO
coefficients on the pyrrole nitrogen atoms (Figure 10). Since
there is a close correspondence between band centers in the
absorption and MCD spectra after the addition of 20 equiv
2+ [45]
for DPPH2
.
A destabilization of the HOMO, which has
large MO coefficients on the four pyrrole nitrogen atoms
(Figure 1), leads to an increase in the DHOMO value from
2+
0.85 to 1.12 eV on going from 5i to 5iH2 and hence to the
increase in the molar absorption coefficients that is observed
2+
for the Q00 bands of 5aH2 and 5cH22+–5 fH22+, since there
is a mixing of the allowed and forbidden properties of the Q
and B bands. An increased DHOMO value also accounts for
the positive a1 terms that are observed for the Q00 and B
2+
bands of TPTPhenPH2
species. The quenching of the
OAM properties of the HOMO level on this basis can be
expected to outweigh the saddling-related quenching of the
8972
www.chemeurj.org
ꢄ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 8965 – 8983

Phenanthroporphyrins
FULL PAPER
Table 3. Distances [ꢇ] between the carbon atoms of the bonds that define the b positions of opposite pyrroles
(see structure for atom numbering) of Zn5a and a series of free-base porphyrinoid compounds and ZnII com-
plexes based on B3LYP geometry optimizations with 6-31G(d) basis sets, the average HOMO–LUMO band-
spectra could not be measured
for the various protonated spe-
cies due to their strongly red-
shifted absorption bands.
A
Influence of annelation on con-
formation and band positions:
The optical spectroscopy of ra-
dially symmetric porphyrinoids
can be readily accounted for
based on the influence of differ-
ent structural perturbations on
the energies of the four frontier
p MOs with ML =ꢁ4, ꢁ5 nodal
properties that arise from the
HOMO and LUMO of the
parent perimeter (Figure 11,
Tables 3 and 4; and Figure S13
in the Supporting Information).
With the exception of a slight
lifting of the orbital degeneracy
of the 1eg* LUMO due to the
loss of the fourfold axis of sym-
metry, the energies of the MOs
of the free-base compounds are
broadly similar to those of the
corresponding ZnII complexes.
The closed-shell d10 configura-
tion of ZnII makes ZnII–por-
phyrinoids ideal model com-
plexes for studying how trends
in the energies of the four fron-
tier p MOs affect the wave-
lengths of the Q and B bands.
Fused-ring expansion with ben-
zene rings destabilizes the 1au
MO, which has large MO co-
Carbon atoms
HꢀL
Qobs
Bobs
Solvent[a]
Ref.
1–3
1’–3’
2–4
2’–4’
2+
5iH2
5i
5a
6.64
6.79
6.90
7.27
8.40
7.84
8.07
8.62
8.18
8.55
8.30
8.27
8.69
8.65
8.07
8.52
8.64
8.60
8.62
8.60
6.64
6.79
6.90
7.26
8.41
7.84
8.07
8.62
8.18
8.55
8.30
8.27
8.69
8.65
8.07
8.52
8.64
8.60
8.62
8.60
6.64
6.74
6.87
7.27
8.40
8.07
8.07
8.62
8.18
8.55
8.32
8.27
8.70
8.85
8.05
8.54
8.61
8.60
8.63
8.60
6.64
6.74
6.87
7.27
8.41
7.99
8.07
8.62
8.18
8.55
8.32
8.27
8.70
8.65
8.05
8.54
8.61
8.60
8.63
8.60
–
–
–
–
–
–
–
4.31
4.36
4.43
4.77
4.94
4.42
4.63
4.65
4.83
4.68
4.78
4.89
4.97
4.78
4.89
4.97
4.97
5.12
5.20
993, 902[c]
809, 776[c]
–
601, 573[c]
586, 568[c]
–
CHCl3
CHCl3
–
CHCl3
THF
Zn5a
ZnTPhenP
TPTANP
ZnTPTANP
ZnTANP[b]
ZnTPNP
ZnNP[b]
TPTBP
ZnTPTBP
TBP[b]
668[d]
482[d]
[29]
[19]
[19]
[17]
[49a]
[49b]
[49d]
[49d]
[27]
[14]
[49c]
[45]
[19]
[19]
[19]
[21]
776, 750[c]
717[c]
575, 558[c]
563[c]
CH2Cl2
CHCl3
CHCl3
DMF
CHCl3
CHCl3
CHCl3
DMF
o-DCB
CH2Cl2
DMA
CH2Cl2
DMA
C6H6
702[d]
528[d]
710
701
470
439
670, 640[c]
651[c]
472, 456[c]
459[c]
663, 614[c]
628
435, 418[c]
433
ZnTBP[b]
ZnDPP
6i
639
463
634[d]
447[d]
TPP[b]
641, 544[c]
586[c]
422, 413[c]
419[c]
ZnTPP[b]
P[b]
618, 525[e]
561
412, 397[e]
401
ZnP[b]
[a] THF, DMF, o-DCB, and DMA refer to tetrahydrofuran, diACTHUNTGRNEUNGmethylformamide, o-diclorobenzene, and dime-
thylacetamide, respectively. [b] Planar ZnII complexes with D4h symmetry and free-base compounds with D2h
symmetry have values between 8.50 and 8.70 ꢇ. [c] Data derived from MCD spectroscopy. [d] Absorption-
band values are for dication species generated in the presence of 1% TFA rather than for ZnII complexes.
[e] Data derived from an MCD spectrum of free-base octaethylporphyrin, since comparatively little research
has been carried out on the parent unsubstituted porphyrin.
AHCTUNGTRENNUNG
of TFA rather than with the zero points of first-derivative-
shaped signals as is typically observed with a1 terms
(Figure 7), it is safe to conclude that the spectrum is domi-
nated by well-resolved Faraday b0 terms, and can, therefore,
This leads to a significant decrease in the first oxidation po-
tential reported for TBP and ZnTBP relative to P and ZnP,
Table 5), since it is easier to remove an electron from the
HOMO. A marked redshift and intensification of the Q
band is observed upon going from ZnP to ZnTBP to
ZnTNP due to the lower average HOMO–LUMO bandgap
and larger DHOMO value (Figure 11 and Table 4).[19] An
average bandgap is calculated because when real rather
than complex MOs are considered, the Q and B bands of
porphyrins arise from close to equal contributions of one-
electron transitions from the two occupied MOs derived
from the parent perimeter HOMO to the two unoccupied
MOs derived from the parent perimeter LUMO
(Table 4).[19,27]
3+
2+
be assigned to the 5iH3 species rather than to the 5iH2
dication. Only slight changes are observed in the absorption
spectrum at low acid concentrations, thus making it difficult
to trace the initial protonation of the two core pyrrolic ni-
trogen atoms. Similar results were reported by Walter et al.
during the titration of the 6i model compound.[17,48] Full pro-
6+
tonation to form 5iH6 can be readily achieved when the
compound is dissolved in concentrated TFA and results in a
2+
“normal” 5H2 spectrum (Figure 7) with an intense and
narrow B band at 580 nm and well-resolved Q01 and Q00
bands at 710 and 784 nm, since the electron-donating prop-
ꢀ
erties of all of the
N
(CH3)2 groups are eliminated. The
The introduction of phenyl groups to form ZnTPP,
ZnTPTBP, and ZnTPTNP destabilizes the 1b1u MO due to
the large MO coefficients at the meso-carbon atoms, thereby
resulting in a further decrease in the average HOMO–
6+
Faraday a1 terms associated with these 5iH6 bands can
be readily identified in the latter stages of TFA titrations of
5i (Figure S11). With the exception of 5iH66+, fluorescence
Chem. Eur. J. 2011, 17, 8965 – 8983
ꢄ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
8973

Z. Shen, N. Kobayashi, K. Rurack et al.
tier p MOs derived from the
HOMO and LUMO of the
parent perimeter, whereas the
redox properties are deter-
mined by the energies of the
HOMO and LUMO.
In contrast with what is ob-
served upon fused-ring expan-
sion with benzene rings, there is
only a minor destabilization of
the 1au MO when acenaphtha-
lene and phenanthrene moieties
are added to the porphyrin p
system. A stabilization of the
LUMO results in a narrowing
of the average HOMO–LUMO
bandgap, however, and
a
marked redshift of the Q and B
bands of ZnTPhenP and
ZnTPTANP relative to ZnTBP.
Although the average HOMO–
LUMO bandgap of ZnTPhenP
is predicted to be larger than
that of ZnTANP, slightly great-
er redshifts are predicted for
Figure 8. Absorption and MCD titration spectra of 5i (14 mm) with TFA in a 99:1 CHCl3/ethanol solvent mix-
ture. The molar equivalents of TFA added and the assignment of spectra to species 5iH22+, 5iH33+, and
the
Q
and
B
bands of
4+
5iH4
are indicated. Asterisks denote narrow regions of the absorption spectra in which the data sets are
opp
TPTPhenP and ZnTPTPhenP
relative to those of TPTANP
potentially slightly problematic due to the subtraction of strong solvent bands. Later stages of the titration are
included as Supporting Information. A consistent set of isosbestic points is not observed between 0–20 equiv
TFA (A), so an intermediate species is probably present, which is speculatively assigned as 5iH22+. The MCD
spectrum at 20 equiv is dominated by Faraday b0 terms, since the MCD and absorption band maxima corre-
spond with each other, and can therefore be assigned to 5iH33+, since first-derivative-shaped a1 terms would
be anticipated for 5iH22+. The spectrum of the next protonated species is observed at 500 equiv (B), which is
and ZnTPTPANP due to
a
marked destabilization of the
1b1u MO caused by the greater
saddling of the ligand (Table 3),
which rotates the meso-phenyl
rings into the plane of the p
system (Figure 3). The decreas-
es anticipated in the first oxida-
tion and reduction potentials of
4+
4+
assigned tentatively to 5iH4
gies for 6iH3 and 6iH4
instead of 5iH4
on the basis of the similarity of predicted NIR-band ener-
opp
adj
3+
4+
(Figure 9).[46] Isosbestic points are not observed on subsequent additions of TFA
opp
5+
(Figure S11 in the Supporting Information) due to partial ring decomposition, but the spectra of 5iH5 and
5iH6 can be readily identified.
6+
LUMO bandgap and redshifts of the Q and B bands. Addi-
tional marked redshifts are observed in the spectrum of
ZnDPP relative to that of ZnTPP due to the effect of steric
hindrance at the ligand periphery on the four frontier p
MOs (Table 3), which is predicted by B3LYP geometry opti-
mizations to result in a marked saddling distortion of the p
system. The decrease in the average HOMO–LUMO band-
5i and 6i relative to 5a and TPP, respectively, upon substitu-
ꢀ
N
tion with
N
bands and the narrowing of the calculated HOMO–LUMO
bandgaps (Figures 5 and 9–11), are observed (Table 5), since
it is easier to add an electron when the LUMO is stabilized
and to remove an electron when the HOMO is destabilized.
Although there is a relatively small redshift of the B
bands of TPTPhenP (568 and 586 nm) relative to TPTANP
(558 and 575 nm),[19] the B and Q bands of ZnTPhenP (482
and 668 nm) are predicted to lie well to the blue of those of
ZnTANP (528 and 702 nm) (Figure 11). Similarly, the elec-
ACHTUNGTRENNUNGgap predicted for ZnDPP is associated primarily, but not ex-
clusively, with a destabilization of the 1b1u MO (Figure 11
and Table 3). This results in a marked decrease in the first
oxidation potential reported for ZnDPP relative to ZnTPP
(Table 5). As was demonstrated in a recent study of nonpla-
nar phthalocyanines with peripheral phenyl substituents,[28]
trends in the gap between the first oxidation and reduction
potentials (DEoꢀr) mirror those observed in the HOMO–
LUMO bandgap and in the observed and calculated Q-band
energies, since the first oxidation and reduction potentials
vary with the energies of the HOMO and LUMO. The
agreement is not expected to be exact, however, since the
Q-band energies are related to the energies of all four fron-
2+
tronic absorption spectrum of TPTANPH2 contains B and
Q00 bands at 565 and 760 nm, respectively,[16] whereas the
2+
corresponding bands for 5aH2 are centered only slightly
to the red at 579 and 803 nm, and the Q and B bands of
2+
TANPH2 (702 and 528 nm)[17] lie well to the red of those
2+
of TPhenPH2
(668 and 482 nm).[17] The significantly
deeper saddling (Table 3), therefore, appears to be the main
cause of the redshift of the Q and B bands of the TPTPhenP
ligand relative to those of TPTANP, rather than the incorpo-
8974
www.chemeurj.org
ꢄ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 8965 – 8983

Phenanthroporphyrins
FULL PAPER
level as is normally the case
with radially symmetric por-
phyrinoids. Although attempts
to form metal complexes of
TPTPhenPs have proven unsuc-
cessful, it is worth noting that
the calculated Q- and B-band
wavelengths for Zn5a also lie
slightly to the red of those pre-
dicted for ZnTPTANP, despite
the fact that the corresponding
bands for ZnTPhenP lie well to
the blue of those of ZnTANP
(Figure 11). A slightly shallower
saddling is predicted for Zn5a
relative to the free-base com-
pound and dication species
(Table 3). In contrast, although
a saddling of the ligand is pre-
dicted for ZnTPTANP, the
B3LYP-optimized structure of
TPTANP contains markedly
different levels of ligand folding
along the x and y axes with the
x-axis acenaphthalene moieties
tilted slightly out of a parallel
alignment (Table 3). The appar-
ent loss of structural flexibility
when phenanthrenes replace
the acenaphthalene moieties
suggests that the steric crowd-
ing at the periphery of the
TPTPhenP ligand results in a
degree of saddling that is too
deep and too rigid for a metal
to be readily incorporated into
the central cavity. It should be
noted that the ZINDO/s calcu-
lation for free-base TPTANP
Figure 9. Left: The TD-DFT- and (right) ZINDO/s-calculated electronic absorption spectra of 5a, 5i, TPP, 6i,
4+
4+
6iH22+, 6iH33+, 6iH4 adj, 6iH4 opp, 6iH55+, and 6iH66+. Transitions associated primarily with the four frontier
p MOs from Goutermanꢅs 4-orbital model,[21] and two additional MOs with the same symmetry properties as
the HOMO and HOMOꢀ1 of P located primarily on the phenyl substituents (Figure 11), which are responsi-
ble for the additional charge-transfer bands of protonated species, are indicated with large black and light-gray does not provide an accurate
diamonds, respectively. Details of the most intense bands of the ZINDO/s calculations are provided as Sup-
porting Information. The TD-DFT calculations predict multiple sets of intense x- and y-polarized bands in the
B-band regions of the TPP and 6a spectra, whereas only one intense set of bands is observed experimentally
as predicted in the ZINDO/s calculations (Figure 7). Similar issues have recently been reported with other por-
phyrinoid TD-DFT calculations.[19,27]
prediction of the wavelengths
of the
Q
and
B
bands
(Figure 11), whereas the wave-
lengths predicted by TD-DFT
for the same B3LYP-optimized
structure are consistent with the
ration of the phenanthrene moieties. The marked increase
in the DHOMO value due to a destabilization of the 1b1u
MO (Figure 1 and Figure 12) leads to an intensification of
the Q-band region of 5i relative to that of 5a.[20] The larger
DHOMO values predicted in TPTPhenP calculations rela-
tive to those of the corresponding TPTANPs coupled with
the relatively small DLUMO values (Figure 12) accounts for
the absence of the anomalous MCD band sign sequences
observed in the MCD spectra of metal TPTANP com-
plexes,[19] since the OAM of the incident photon is con-
served by the circulation of negative charge in the LUMO
trends observed in the experimental and calculated data
(see the Supporting Information).
Influence of the electronic nature of the meso substituents
on band positions: TD-DFT and ZINDO/s calculations pre-
ꢀ
dict that the introduction of electron donating
N
G
groups at the para positions of the phenyl groups to form 5i
and 6i destabilizes the four frontier p MOs (Figure 11 and
Figure S13 in the Supporting Information). The introduction
of substituents on the phenyl group provides scope for fine-
tuning the properties of TPTPhenPs based on the Hammett
Chem. Eur. J. 2011, 17, 8965 – 8983
ꢄ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
8975

Z. Shen, N. Kobayashi, K. Rurack et al.
Figure 11. Top: ZINDO/s MO energies of a series of the free-base com-
pounds and zinc complexes of porphyrinoids based on a set of B3LYP ge-
ometry optimizations with 6-31G(d) basis sets. Black circles, triangles, di-
amonds, and squares denote the 1au, 1b1u, 1b2g*, and 1b3g* frontier p
MOs of P, respectively, and the MOs of the other compounds that exhibit
the same ML =ꢁ4, ꢁ5 nodal patterns. Gray diamonds denote the aver-
age HOMO–LUMO bandgap taking into account all four MOs associat-
ed with the HOMO and LUMO of the parent perimeter. Bottom: Exper-
imental and calculated Q- and B-band energies are denoted with gray di-
amonds and triangles and black lines, respectively.[10b,17,19,21,27,31,48,49] Black
diamonds and triangles denote values derived from dication species
rather than the ZnII complexes of TANP, TPhenP, and 5a.[17,31] The corre-
sponding plot for the TD-DFT calculations is provided in Figure S13 in
the Supporting Information.
Figure 10. Top: The nodal patterns of the four key frontier p HOMOs of
6i and 6iH2 at an isosurface of 0.02 a.u. based on a B3LYP geometry-
2+
optimized structure derived using 6-31G basis sets. The axes are aligned
so that the x and y axes lie horizontal and vertical in the plane of the
page, respectively. H and Hꢀn denote the HOMO and the nth MO in de-
scending energy from the HOMO, respectively. Bottom: Trends in the
relative energies [eV] of the key frontier p MOs of P, TPP, 5a, 5i,
4+
4+
6+
5iH22+, 6i, 6iH22+, 6iH33+, 6iH4 adj, 6iH4 opp, 6iH55+, and 6iH6 rela-
tive to the average energy of the two MOs associated with the LUMO of
the parent perimeter (i.e., 1b2g* and 1b3g* in the case of P; Figure 1).
Black circles, triangles, squares, and diamonds denote MOs derived from
the 1au, 1b1u, 1b2g*, and 1b3g* MOs of P, respectively (Figure 1 and Fig-
ure S10 in the Supporting Information), which are the four frontier p
MOs that form the basis of Goutermanꢅs 4-orbital model.[21] Light-gray
triangles and circles denote additional MOs introduced by the peripheral
ꢀ
facilitating conjugation between the meso groups and the
inner perimeter of the p system. As a result, even subtle
electronic effects like those that arise from the halogen sub-
stituent series, 5c–f, alter the optical spectroscopy to a readi-
ly observable extent (Figure 6 and Table 6). For example, in
the case of electron-withdrawing substituents, a plot of the
B-band maxima versus 4sp yields a strong correlation for
5c–e, 5g, and 5h (Figure 13). A change in the aryl substitu-
ent can be expected to modify the energies of the Q and B
bands based on changes in the DEoꢀr value and the magni-
tude of the HOMO–LUMO bandgap. This is more difficult
to follow in the Q-band region in which much of the absorp-
tion band intensity is vibrational rather than electronic in
origin and there is often a marked splitting of the Qx and
Qy bands.
When there is limited configurational interaction between
the S1 state and higher p–p* energy excited states, trends in
the energy of the lowest-energy p!p* band will mirror
those observed in the HOMO–LUMO bandgap and DEoꢀr
values. In the context of porphyrinoids, the S2 state can also
be involved in this regard because, as is described in Gouter-
manꢅs 4-orbital model,[21] the Q- and B-band energies are
NACHTUNGTRENNUNG(CH3)2 groups with b1u and au symmetry (under the D2h symmetry of
P), respectively (Figure 10, top), which play a key role in additional
charge-transfer bands calculated to lie at similar energies to the Q and B
4+
4+
5+
bands in the spectra of 6iH22+, 6iH33+, 6iH4 adj, 6iH4 opp, and 6iH5
.
The corresponding plot of the ZINDO/s MO energies is provided in Fig-
ure S12 in the Supporting Information.
parameter (sp). It has long been known that linear plots can
1
be obtained for E = versus 4sp for the first and second reduc-
2
tion and oxidation potentials of structurally related tetra-
phenylporphyrins.[51] For example, in the mid-1970s, Kadish
and Morrison[52] reported linear trends in the redox values
ꢀ
ꢀ
ꢀ
of free-base (p-X)TPP compounds with OCH3, CH3, H,
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
F, COOCH3, Cl, CN, and NO2 substituents. The deep
saddling distortion of the p system of (p-X)TPTPhenPs pre-
dicted in the B3LYP geometry optimizations of 5a and 5i
(Figure 3 and Table 3) tilts the meso-phenyl rings almost
into the plane of the four pyrrole nitrogen atoms, thereby
8976
www.chemeurj.org
ꢄ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 8965 – 8983

Phenanthroporphyrins
FULL PAPER
Table 4. The DHOMO (DH) and DLUMO (DL) values, the percentage contribution to the Q (Q4ob%) and B
(B4ob%) bands of the one-electron transitions that link the lower-energy MO derived from the parent perimeter
HOMO to the MOs derived from the LUMO, the calculated Q- and B-band intensities (Qint, Bint), and wave-
lengths (Qcalcd, Bcalcd) derived from ZINDO/s calculations.
determined by configurational
interaction between four spin-
allowed p–p* excited states,
which arise from four one-elec-
tron transitions between the
two MOs derived from the
DH
DL
Q4ob%
B4ob%
Qint
Bint
Qcalcd
Bcalcd
C18H18
P
TPP
6i
TBP
TPTBP
TPTANP
5a
5i
C16H16
ZnP
ZnTPP
0.00
0.36
0.10
0.07
1.19
0.72
0.09
0.14
0.38
0.00
0.62
0.11
0.03
1.44
0.96
1.82
1.35
0.52
0.03
0.78
0.06
0.00
0.02
0.04
0.02
0.26
0.33
0.06
0.05
0.07
0.00
0.00
0.00
0.01
0.00
0.00
0.00
0.00
0.00
0.00
0.01
0.00
96, 97
96, 97
96, 97
92, 93
96, 96
95, 96
88, 84
92, 94
92, 94
96, 96
96, 96
96, 96
96, 96
94, 94
95, 95
95, 95
93, 93
89, 89
91, 91
93, 93
94, 93
97, 97
77, 95
88, 94
86, 91
86, 92
91, 93
55, 44
83, 91
86, 91
97, 97
94, 94
94, 94
91, 91
92, 92
93, 93
78, 78
67, 67
75, 75
90, 90
88, 87
90, 90
0.00, 0.00
0.03, 0.05
0.00, 0.00
0.00, 0.00
0.19, 0.42
0.05, 0.26
0.01, 0.06
0.01, 0.01
0.05, 0.05
0.00, 0.00
0.07, 0.06
0.00, 0.00
0.00, 0.00
0.35, 0.35
0.18, 0.18
0.57, 0.57
0.37, 0.37
0.16, 0.16
0.00, 0.00
0.20, 0.19
0.00, 0.00
4.51, 4.51
1.75, 2.30
2.56, 2.89
2.81, 3.07
2.61, 2.28
2.72, 2.52
0.80, 1.07
1.40, 1.85
1.56, 1.87
3.70, 3.70
2.31, 2.31
2.82, 2.82
2.86, 2.87
2.47, 2,47
2.55, 2.55
2.93, 2.93
1.53, 1.54
2.11, 2.11
2.33, 2.34
2.67, 2.67
1.78, 1.78
898, 677
743, 657
784, 691
796, 702
807, 775
880, 828
685, 546
1016, 885
1040, 902
724, 724
674, 674
704, 704
752, 752
770, 770
812, 812
850, 850
886, 886
762, 762
840, 840
772, 771
906, 906
359, 359
383, 369
405, 400
416, 413
405, 378
451, 428
515, 496
550, 537
555, 552
2ꢀ
HOMO of the C16H16 parent
perimeter and the two MOs de-
rived from the LUMO. When
DHOMOꢂDLUMOꢂ0, how-
ever, as is predicted to be the
case
for
TPTPhenPs
2<M->
350, 350 (Figure 12), there are close to
367, 367
400, 400
428, 428
fully forbidden Q bands and
fully allowed B bands based on
ZnDPP
ZnTBP
the DML =ꢁ1 and ꢁ9 proper-
382, 382
ties of the Q and B transitions.
When there is only limited con-
figurational interaction between
the B and higher p–p* states, it
is reasonable to expect that the
ZnTPTBP
ZnTNP
ZnTPTNP
ZnTANP
ZnTPTANP
ZnTPhenP
Zn5a
430, 430
392, 392
449, 449
465, 465
514, 514
430, 430
534, 533 B-band energy will be modified
in a readily predictable manner
when the para substituent of
Table 5. The first oxidation (ox.) and first reduction (red.) potentials [V] of porphyrinoid compounds relative
to the saturated calomel electrode and a comparison of trends in the DEoꢀr [eV] values, the HOMO (H) and
LUMO (L) energies [eV] in B3LYP geometry optimizations, the HOMO–LUMO bandgap (jHꢀLj), and the
average observed and TD-DFT-calculated Q00-band energies [eV].
the phenyl groups on the meso-
carbon atoms is changed based
on a consideration of the nodal
Solvent
1st ox.
1st red.
DEoꢀr
Ref.
H
L
jHꢀLj
Q(calcd)
Q(obs) patterns of the four key frontier
+0.94[a]
+1.00
+0.44
+0.55
+0.90
+0.48
+0.72[a]
+0.80
+0.48
+0.38
ꢀ1.36[a]
ꢀ1.23
ꢀ1.38
ꢀ1.13
ꢀ0.93
ꢀ0.89
ꢀ1.33[a]
ꢀ1.35
ꢀ1.34
ꢀ1.46
2.30[a]
2.23
1.82
1.68
1.83
1.37
2.05[a]
2.15
1.82
1.84
[50a]
[50b]
[50c]
ꢀ5.30
ꢀ4.91
ꢀ4.22
ꢀ4.69
ꢀ4.47
ꢀ3.74
ꢀ5.21
ꢀ4.98
ꢀ4.72
ꢀ4.62
ꢀ2.24
ꢀ2.19
ꢀ1.75
ꢀ2.26
ꢀ2.58
ꢀ2.08
ꢀ2.14
ꢀ2.12
ꢀ2.23
ꢀ2.10
3.06
2.72
2.47
2.43
1.89
1.66
3.07
2.86
2.49
2.52
2.37
2.24
2.13
1.95
1.59
1.43
2.46
2.33
2.05
2.05
2.19[a]
p MOs using a perimeter model
P
TPP
6i
TBP
5a
5i
ZnP
ZnTPP
ZnDPP
ZnTBP
CH2Cl2
CH2Cl2
CH2Cl2
DMSO
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
DMSO
2.12
2.02
2.15
1.57
1.32
2.22
2.12
1.95
1.98
approach. Kadish and Morri-
son[52] demonstrated in the con-
text of (p-X)TPP compounds
that an increase in the electron-
[50d]
[b]
[b]
donating properties causes
a
[50a]
[50b]
[50e]
[50d]
negative shift in the reduction
and oxidation potentials. Simi-
lar negative shifts are observed
for (p-X)TPTPhenPs (Figure 14
and Table 6) and are consistent
with a destabilization of both
[a] Redox data for P and ZnP are derived from free-base and zinc octaethylporphyrin, since comparatively
little research has been carried out on the parent unsubstituted porphyrin. [b] This work.
the HOMO and LUMO. A consideration of the
nodal properties makes it clear that the HOMO,
LUMO, and LUMO+1 are more likely to be desta-
bilized than the HOMOꢀ1 (Figure 14), since there
are large MO coefficients on the meso-carbon
atoms. The negative shift is more marked in the
case of the oxidation potential (Figure 13), since all
four meso-carbon atoms lie on nodal planes or an-
tinodes of the HOMO-1 and HOMO, respectively,
due to the alignment of mirror planes along the x
and y axes and the DML =ꢁ4 nodal properties of
the MOs derived from the HOMO of the 16-atom
parent hydrocarbon perimeter (Figures 10 and 14).
In contrast, the meso-carbon atoms lie between the
nodal planes of the LUMO and LUMO+1 due to
the DML =ꢁ5 nodal properties, so the MO coeffi-
cients are smaller than is the case with the HOMO.
Table 6. The first and second reduction (1st red., 2nd red.) and oxidation (1st ox., 2nd
ox.) potentials [V] relative to the saturated calomel electrode, the gap between the
first oxidation and reduction potentials (DEoꢀr) and the Hammett parameters (sp)[53]
of TPThenP compounds in CH2Cl2.
[d]
4sp
2nd
1st
1st
2nd
DEoꢀr
Q[e] B[e]
red.[a]
red.[a]
ox.[a]
ox.[a]
ꢀ0.89[b]
0.48[c]
0.92[c]
1.37
1.34 2.02
ꢀ
N-
ACHTUNGTRENNUNG(CH3)2
5i
ꢀ3.32
–
ꢀ1.13[b]
ꢀ0.93[b]
ꢀ0.92[b]
ꢀ0.96[b]
ꢀ0.99[b]
0.82[c]
0.90[c]
0.90[c]
0.84[c]
0.78[c]
–
–
–
1.95
1.83
1.82
1.80
1.77
1.55 2.13
1.57 2.16
1.55 2.12
1.55 2.14
1.54 2.06
ꢀ
5b
5a
5 f
5e
5h
CH3
ꢀ
H
ꢀ0.68
0.00
0.72
–
–
–
–
ꢀ
I
1.16[c]
1.32[c]
ꢀ
Br
0.92
3.12 ꢀ1.50[c]
ꢀ
NO2
[a] Peak potentials were determined based on differential pulse voltammetry experi-
ments. [b] Reversible peak. [c] Irreversible peak. [d] The gap between the first oxida-
tion and reduction potentials (DEoꢀr) [eV]. [e] The observed energies of the Q and B
bands [eV].
Chem. Eur. J. 2011, 17, 8965 – 8983
ꢄ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
8977

Z. Shen, N. Kobayashi, K. Rurack et al.
fects that are responsible for
the trend observed with 5a and
5c–f. Michl has demonstrated
that electron-donating substitu-
ents tend to have a greater
effect on the DHOMO value
due to the presence of addition-
al high-lying occupied MOs
with symmetry similar to those
of the MOs derived from the
HOMO of the parent hydrocar-
bon perimeter (Figure 10).[20b]
The mixing of these MOs due
to the mesomeric effect, there-
fore, significantly modifies the
energies of the frontier p MOs
relative to 5a and 5c–f in which
an inductive substituent effect
is the dominant factor. The
redox properties and the ener-
gies of the Q and B bands are,
therefore, modified significant-
ly.
In the context of the meso-
meric interactions of the elec-
ꢀ
ꢀ
tron-accepting CN and NO2
substituents in 5g and 5h, there
is less scope for a significant
DLUMO value to be intro-
duced because of the introduc-
tion of additional low-lying un-
occupied MOs, due to the
manner in which the alignment
of the nodal planes of the four
key frontier p MOs is deter-
mined by the planes of symme-
try, which are introduced along
the x and y axes by the saddling
of the ligand and the protona-
Figure 12. The effect of structural perturbations on the DHOMO and DLUMO values predicted by (top left)
B3LYP and by (top right) INDO/s calculations for the same set of B3LYP-optimized geometries. In the light-
gray shaded areas in which DHOMO>DLUMO, a ꢀve/+ve MCD sign sequence is anticipated[20] for the Q
band in ascending energy terms, whereas a +ve/ꢀve sign sequence is anticipated in the unshaded areas in
which DLUMO>DHOMO. Bottom right: The calculated Q-band oscillator strengths from TD-DFT and
(bottom left) INDO/s and calculations based on 6-31G(d) basis sets plotted against the DHOMO value. As
the DHOMO value increases, the calculated oscillator strength increases. A greater destabilization of the 1b1u
MO (Figure 1) tends to be predicted in the TD-DFT calculations, thereby resulting in significantly greater cal-
culated oscillator strengths for TPTPhenPs.
The B-band energies of 5b and 5i do not lie on the trend
line formed by the energies of the electron-withdrawing sub-
stituents (Figure 13). This is not surprising, since Kadish and
tion of two pyrrole nitrogen atoms. In the context of the
LUMO and LUMO+1, there are significant MO coefficients
on the meso-carbon atoms, so there is less scope for the in-
troduction of a significant DLUMO value based on the elec-
tron-withdrawing or -donating properties of the four aryl
groups (Figure 14). The protonated and nonprotonated pyr-
role nitrogen atoms are the positions on the inner ligand pe-
rimeter in which there is most scope for substantially modi-
fying the DLUMO value based on the alignment of nodes
and antinodes along the x and y axes in the LUMO and
Morrison[52] reported that the closest agreement with the E =
1
2
versus 4sp trend was observed with electron-accepting sub-
stituents. Marked differences are observed in the spectra of
these compounds relative to those of 5a and 5c–f (Figures 5
and 6). There is a marked intensification of the Q-band
region of the 5b and 5i spectra, which may be related in
part to increased vibrational band intensity caused by the
ꢀ
ꢀ
ꢀ
CH3 and
N
(CH3)2 groups. The presence of vibrational
LUMO+1. Electron-accepting aryl substituents, such as
ꢀ
bands in the B-band region could significantly alter the posi-
tion of the observed B-band centers in the absorption spec-
trum, which results from the overlap of the symmetry-split
x- and y-polarized B00 bands. In the context of 5i, there is a
strong mesomeric interaction between the nitrogen lone
ꢀ
CN and NO2, draw electron density away from these
atoms, however, thereby minimizing the effect of the elec-
tronegativity difference on these atoms due to protonation.
The large DHOMO value that is introduced (Figure 12) in
the electronic structure of 5i causes increased mixing of the
allowed and forbidden properties of the Q and B bands, so
the Q bands gain significant intensity. Previous studies on a
pairs of
NACHTUNGTRENNUNG(CH3)2 with the p system of the core porphy-
ACHTUNGTRENNUNGrinoid macrocycle (Figure 10) in addition to the inductive ef-
8978
www.chemeurj.org
ꢄ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 8965 – 8983

Phenanthroporphyrins
FULL PAPER
higher-energy p–p* states, thereby resulting in a blueshift of
the B band and an increase in the energy gap between the
Q and B bands.[19] The increased energy separation of the Q
and B bands, and the broadening and loss of intensity of the
B band of 5i (Figures 5 and 12) is consistent with this. From
the standpoint of NIR dye applications, the properties of
the Q bands of TPTPhenPs are considerably more impor-
tant than those of the B band, however, since they lie at the
red end of the visible region. The ability to fine-tune the
band centers and intensities of the Q00 bands by varying the
Hammett parameter of the para substituent and by intro-
ducing electron-donating substituents with a strong meso-
meric interaction with the TPTPhenP chromophore should
greatly facilitate the design of TPTPhenPs suitable for spe-
cific practical applications.
Conclusion
Substituted TPTPhenPs exhibit the most highly redshifted B
and Q bands to have been reported to date for radially sym-
metric tetrapyrrole porphyrinoids. The optical properties of
TPTPhenP compounds can be fine-tuned by carrying out
the synthesis with different benzaldehydes so that meso-aryl
groups with strongly electron-withdrawing or -donating
properties are introduced into the p system. An unprece-
dented redshift of the Q band is observed for 5i based on
ꢀ
Figure 13. Plot of the B absorption-band energy in CHCl3 versus Ham-
mett parameter sp (4sp because four substituents are attached to the por-
phyrin) for dyes 5c–h with electron-withdrawing substituents (black
squares) and 5a, 5b, and 5i (top; white squares). The solid line repre-
sents a fit of the data (r=0.996) based on a linear regression analysis.
The trends observed in the Q-band region broadly mirror those observed
for the B band, but the correlation factor is considerably lower since
there are substantially greater changes in the relative intensities and sep-
arations of the x- and y-polarized Q00 bands (Figures 5 and 6). Bottom:
The observed Q- and B-band wavelengths and DEoꢀr values [eV] denoted
by black triangles, diamonds and circles, respectively, and plotted versus
4sp. Redox potentials for the oxidation and reduction steps relative to
the saturated calomel electrode are plotted as light gray squares against a
secondary axis.
the strong electron-donating properties of the
NACHTUNGTRENNUNG(CH3)2
groups, due to a marked destabilization of the HOMO and
hence a decrease in the HOMO–LUMO bandgap. An in-
crease in the DHOMO value (Table 4 and Figure 12) results
in a further marked intensification of the Q00 bands and a
significantly greater separation of the Q and B bands (Fig-
ures 11 and 12), since there is a greater level of configura-
tional interaction between the B and higher-energy p–p*
states. The protonated species of TPTPhenPs exhibit batho-
chromic shifts of the B and Q00 bands of up to 150 nm com-
2+
pared to the meso-unsubstituted TPhenPH2 parent com-
3+
wide range of different radially symmetric fused-ring-ex-
panded porphyrinoids have demonstrated that when
DHOMO>DLUMO there tends to be an increase in the
level of configurational interaction between the B and
pound. The redshift of the lowest-energy of 5iH3
(Figure 7) is an unprecedented 550 nm relative to the spec-
trum of TPP. These are by far the largest bathochromic
shifts observed for a radially symmetric tetrapyrrole por-
2ꢀ
Figure 14. The alignment of the nodal planes in the four frontier p MOs of 5a based on the ML =ꢁ4 and ꢁ5 properties of the parent C16H16 parent pe-
rimeter. MO labels are shown for both D2h and C2v symmetry to aid comparison with higher-symmetry porphyrinoids (Figure 1). The protonated pyrrole
nitrogen atoms are aligned with the vertical y axis in the plane of the page, whereas the z axis points out of the page towards the reader.
Chem. Eur. J. 2011, 17, 8965 – 8983
ꢄ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
8979

Z. Shen, N. Kobayashi, K. Rurack et al.
from CHCl3/methanol, afforded 5b as dark red crystals in 11% yield
(35 mg). M.p. >3008C; H NMR (500 MHz, [D7]DMF): d=1.01 (b, 2H),
phyrinoid. Acid–base chemistry in conjunction with potent
donor/acceptor groups can enable unprecedented NIR dye
properties, which respond in a predesigned way to a specific
chemical input. These properties could lead to applications
as precursors or building blocks for the construction of
NIR-active supramolecular ensembles suitable for use in
nonlinear optics, opto-electronics, solar-energy conversion,
or as photosensitizers for photodynamic therapy. Still larger
bathochromic shifts may still be achievable. The close corre-
spondence observed in the trends predicted by TD-DFT and
ZINDO/s calculations and those observed experimentally by
optical spectroscopy and electrochemistry (Figure 11) dem-
onstrate that this question can be approached based on ra-
tional design rather than by trial and error.
1
2.44 (s, 12H), 6.83 (m, 8H), 7.26 (m, 8H), 7.48–7.50 (m, 8H), 8.27 (m,
8H), 8.47 (m, 8H), 8.73 ppm (m, 8H); IR (KBr): n˜max =3425, 3086, 1604,
1445, 1352, 1079, 1047, 758, 728 cmꢀ1; MS (MALDI-TOF): m/z: 1271.41
[M+]; elemental analysis calcd (%) for C96H62N4: C 90.68, H 4.91, N 4.41;
found: C 90.65, H 4.89, N 4.86.
Porphyrin 5c: Compound 5c was prepared as described above from p-
fluorobenzaldehyde
(124 mg,
1 mmol),
phenanthroACTHNGUTER[NNUG 9,10-c]pyrrole
(217 mg, 1 mmol), BF3·Et2O (40 mL, 0.32 mmol), and DDQ (227 mg,
1 mmol) in CH2Cl2 (350 mL). Column chromatography on silica gel, elut-
ing gradually from 40% CHCl3/petroleum ether to chloroform, followed
by recrystallization from CHCl3/methanol, afforded 5c as a dark red
powder in 10% yield (32 mg). M.p. >3008C; 1H NMR (500 MHz,
[D7]DMF): d=0.89 (b, 2H), 6.91–6.94 (m, 8H), 7.46–7.49 (m, 8H), 7.60–
7.63 (m, 8H), 7.86–7.88 (m, 8H), 8.65–8.67 (m, 8H), 8.86–8.88 ppm (m,
8H); IR (KBr): n˜max =3426, 3086, 1597, 1449, 1352, 1158, 1048, 758,
724 cmꢀ1; MS (MALDI-TOF): m/z: 1288.4 [M+]; elemental analysis
calcd (%) for C92H50F4N4: C 85.83, H 3.91, N 4.35; found: C 85.86, H
3.88, N 4.21.
Porphyrin 5d: Compound 5d was prepared as described above from p-
chlorobenzaldehyde (140 mg, 1 mmol), phenanthroACTHUNRTGNEUNG[9,10-c]pyrrole
Experimental Section
(217 mg, 1 mmol), BF3·Et2O (40 mL, 0.32 mmol), and DDQ (227 mg,
1 mmol) in CH2Cl2 (350 mL). Column chromatography on silica gel, elut-
ing gradually from 40% CHCl3/petroleum ether to 1% triethylamine/
chloroform, followed by recrystallization from CHCl3/methanol, afforded
5d as a dark red powder in 13% yield (44 mg). M.p. >3008C; 1H NMR
(500 MHz, [D7]DMF): d=0.84 (b, 2H), 7.02–7.05 (m, 8H), 7.42–7.44 (m,
8H), 7.57–7.59 (m, 8H), 7.78–7.81 (m, 8H), 8.68–8.71 (m, 8H), 8.92–
8.95 ppm (m, 8H); IR (KBr): n˜max =3421, 3084, 1586, 1447, 1350, 1047,
1092 758, 726 cmꢀ1; MS (MALDI-TOF): m/z: 1353.3 [M+H+]; elemental
analysis calcd (%) for C92H50Cl4N4: C 81.66, H 3.72, N 4.14; found: C
81.45, H 3.80, N 4.21.
General: Unless otherwise noted, all chemicals and solvents were of ana-
lytical reagent grade and used as received. Dry CH2Cl2 was freshly dis-
tilled over CaH2 under nitrogen. Dry tetrahydrofuran (THF) was distilled
from sodium/benzophenone under nitrogen. Column chromatography
and TLC were performed using C-200 and Kieselgel 60 F254 instruments,
respectively. 1H and 13C NMR spectra were acquired in [D7]DMF or
CDCl3 using a Bruker DRX500 spectrometer operating at 500 MHz for
1H and 125 MHz for 13C. Chemical shifts are reported in units of ppm
and are referenced using TMS as the internal standard. MALDI-TOF
mass spectra were recorded using a Bruker Autoflex IITM spectrometer.
Elemental analyses for C, H, and N were performed using a Perkin–
Elmer 240C elemental analyzer. Melting points were determined using a
Reichert thermometer without correction. IR spectra were recorded
using a VECTOR 22 spectrophotometer with KBr discs in the 4000–
400 cmꢀ1 region. The redox data were obtained using a Zahner Electro-
chemical Workstation. Data were measured for 5a, 5b, 5e, 5 f, 5h, and 5i
using a platinum working electrode, a platinum wire counter electrode,
and an Ag/AgCl reference electrode using Fc/Fc+ as an internal stan-
dard. The measurements were carried out in CH2Cl2 solution using 0.1m
Bu4NPF6 supporting electrolyte. All experiments were carried out under
an argon atmosphere.
Porphyrin 5e: Compound 5e was prepared as described above from 4-
bromobenzaldehyde (185 mg, 1 mmol), phenanthroACTHUNRTGNEUNG[9,10-c]pyrrole
(217 mg, 1 mmol), BF3·Et2O (40 mL, 0.32 mmol), and DDQ (227 mg,
1 mmol). Column chromatography on silica gel, eluting gradually from
50% CHCl3/petroleum ether to CHCl3, followed by recrystallization
from CHCl3/methanol, afforded 5e as a dark red powder in 11% yield
1
(42 mg). M.p. >3008C; H NMR (500 MHz, [D7]DMF): d=0.86 (b, 2H),
6.99–7.03 (m, 8H), 7.40–7.43 (m, 8H), 7.55–7.59 (m, 8H), 7.83–7.86 (m,
8H), 8.66–8.69 (m, 8H), 8.85–8.88 ppm (m, 8H); IR (KBr): n˜max =3422,
3082, 1630, 1442, 1350, 1077, 1047, 759, 725 cmꢀ1; MS (MALDI-TOF): m/
z: 1531.1 [M+]; elemental analysis calcd (%) for C92H50Br4N4: C 72.17, H
3.29, N 3.66; found: C 72.17, H 3.36, N 3.57.
Synthesis: PhenanthroACHTUNGTRENNUNG
[9,10-c]pyrrole,[30] ethyl isocyanoacetate,[54] and 4-
iodobenzaldehyde[55] were prepared according to literature methods.
Porphyrin 5 f: Compound 5 f was prepared as described above from 4-
iodoACTHNUTRGENNGbU enzaldehyde (232 mg, 1 mmol), phenanthroAHCTUNGTRENN[NGU 9,10-c]pyrrole (217 mg,
Porphyrin 5a: Benzaldehyde (106 mg, 1 mmol) was added to a stirred so-
lution of phenanthroACHTUNGTRENNUNG[9,10-c]pyrrole (217 mg, 1 mmol) in dry CH2Cl2
1 mmol), BF3·Et2O (40 mL, 0.32 mmol), and DDQ (227 mg, 1 mmol).
Column chromatography on silica gel, eluting gradually from 40%
CHCl3/petroleum ether to 80% CHCl3/petroleum ether, followed by re-
crystallization from CHCl3/methanol, afforded 5 f as a dark red powder
in 15% yield (65 mg). M.p. >3008C; 1H NMR (500 MHz, [D7]DMF): d=
0.82 (b, 2H), 6.98–7.05 (m, 8H), 7.19–7.25 (m, 8H), 7.38–7.46 (m, 8H),
7.57–7.63 (m, 8H), 8.02–8.10 (m, 8H), 8.67–8.72 ppm (m, 8H); IR (KBr):
(350 mL), and the resulting solution was purged with argon for 10 min.
The solution was cooled to ꢀ508C and kept in the dark. BF3·Et2O
(40 mL, 0.32 mmol) was added. The solution was stirred at ꢀ508C for 2 h,
warmed to room temperature, and stirred for another 48 h. Three drops
of triethylamine were added followed by the addition of DDQ (227 mg,
1 mmol). After stirring for an additional 1 h, the solvents were removed
under reduced pressure. The residue was purified by column chromatog-
raphy on silica gel, eluting gradually from 50% chloroform/petroleum
ether to CHCl3. The product was collected as a deep-red fraction. Re-
crystallization from CHCl3/methanol afforded 5a as dark red solids in
17% yield (52 mg). M.p. >3008C; 1H NMR (500 MHZ, [D7]DMF): d=
1.11 (b, 2H), 6.93 (m, 8H), 7.36–7.38 (m, 8H), 7.71–7.74 (m, 12H), 7.95–
n˜max =3424, 3080, 1576, 1446, 1348, 1060, 758, 727 617 cmꢀ1
; MS
(MALDI-TOF): m/z: 1718.9 [M+]; elemental analysis calcd (%) for
C92H50I4N4: C 64.28, H 2.93, N 3.26; found: C 64.13, H 3.01, N 3.15.
Porphyrin 5g: Compound 5g was prepared as described above from 4-cy-
anobenzaldehyde (131 mg, 1 mmol), phenanthroACTHNUTRGENUG[N 9,10-c]pyrrole (217 mg,
7.97 (m, 8H), 8.59–8.63 (m, 8H), 8.98 ppm (m, 8H); IR (KBr): n˜max
=
1 mmol), BF3·Et2O (40 mL, 0.32 mmol), and DDQ (227 mg, 1 mmol).
Chromatography on silica, eluting gradually from 50% CHCl3-petroleum
ether to CHCl3, followed by recrystallization from CHCl3/methanol, af-
forded 5g as dark red crystals in 14% yield (49 mg). M.p. >3008C;
1H NMR (500 MHz, [D]TFA/CDCl3): d=0.89 (b, 2H), 6.86 (t, J1 =
7.5 Hz, J2 =8.0 Hz, 8H), 7.19 (d, J=8.0 Hz, 8H), 7.42 (t, J1 =7.5 Hz, J2 =
8.0 Hz, 8H), 7.94 (d, J=7.5 Hz, 8H), 8.48 (d, J=8.5 Hz, 8H), 8.89 ppm
(d, J=8.5 Hz, 8H); IR (KBr): n˜max =3386, 3084, 2225, 1599, 1446, 1343,
1166, 759, 725 cmꢀ1; MS (MALDI-TOF): m/z: 1315.4 [M+]; elemental
3424, 3081, 1629, 1443, 1352, 1048, 759, 726 cmꢀ1; MS (MALDI-TOF): m/
z: 1216.6 [M+H+]; elemental analysis calcd (%) for C92H54N4: C 90.91,
H 4.48, N 4.61; found: C 90.80, H 4.60, N 4.42.
Porphyrin 5b: Compound 5b was prepared as described above from p-
tolyl aldehyde (120 mg, 1 mmol), phenanthroACTHNUTRGNE[UNG 9,10-c]pyrrole (217 mg,
1 mmol), BF3·Et2O (40 mL, 0.32 mmol), and DDQ (227 mg, 1 mmol) in
dry CH2Cl2 (350 mL). Chromatography on silica, eluting gradually from
50% CHCl3/petroleum ether to CHCl3, followed by recrystallization
8980
www.chemeurj.org
ꢄ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 8965 – 8983

Phenanthroporphyrins
FULL PAPER
analysis calcd (%) for C96H50N8: C 87.65, H 3.83, N 8.52; found: C 87.51,
H 3.87, N 8.44.
(Scheme 1), 5i, 5iH22+, and the free-base (6i) and protonated species
(6iH22+, 6iH33+, 6iH4 adj, 6iH4 opp, 6iH55+, 6iH66+) of a TPP model
complex for 5i (Scheme 2) through the use of the B3LYP functional of
the Gaussian 03W software package[32] by using 6-31G or 6-31G(d) basis
sets. ZINDO/s and TD-DFT calculations were carried out based on
B3LYP-optimized geometries.
4+
4+
Porphyrin 5h: Compound 5h was prepared as described above from 4-ni-
trobenzaldehyde (151 mg, 1 mmol), phenanthroACTHNUTRGENUG[N 9,10-c]pyrrole (217 mg,
1 mmol), BF3·Et2O (40 mL, 0.32 mmol), and DDQ (227 mg, 1 mmol).
Chromatography on silica gel, eluting gradually from 50% CHCl3/petro-
leum ether to CHCl3, followed by recrystallization from CHCl3/methanol,
afforded 5h as dark green crystals in 6% yield (21 mg). M.p. >3008C;
1H NMR (500 MHz, [D7]DMF): d=0.83 (b, 2H), 7.01 (m, 8H), 7.39 (m,
8H), 7.51 (m, 8H), 8.53 (m, 8H), 8.69 (m, 8H), 9.21–9.28 ppm (m, 8H);
IR (KBr): n˜max =3417, 3084, 1588, 1515, 1354, 1337, 1046, 852, 758,
727 cmꢀ1; MS (MALDI-TOF): m/z: 1396.4 [M+H+]; elemental analysis
calcd (%) for C92H50N8O8: C 79.19, H 3.61, N 8.03; found: C 79.13, H
3.49, N 8.03.
Acknowledgements
Financial support from the National Natural Science Foundation of
China (nos. 20971066 and 21021062), the Chinese Ministry of Educationꢅs
Program for New Century Excellent Talents in Universities (no. NCET-
08-0272), the Major State Basic Research Development Program of
China (grant nos. 2011CB808704 and 2007CB925103), a Grant-in-Aid for
Scientific Research on Innovative Areas (no. 20108007, “pi-Space”) pro-
vided by the Japanese Ministry of Education, Culture, Sports, Science,
and Technology (MEXT), the Alexander von Humboldt Foundation and
the European Commissionꢅs Human Resources & Mobility Programme
for a Marie Curie Intra-European Fellowship is highly appreciated. We
thank D. Pfeifer (BAM Div. I.3) for help with protonation experiments
by NMR spectroscopy, which were fruitless because of solubility prob-
lems.
Porphyrin 5i: Compound 5i was prepared as described above from 4-di-
methylamino-benzaldehyde (149 mg, 1 mmol), phenanthroACTHNUTRGNE[NUG 9,10-c]pyrrole
(217 mg, 1 mmol), BF3·Et2O (45 mL, 0.37 mmol), and DDQ (31 mg,
0.14 mmol). Chromatography on silica gel, eluting gradually from CHCl3
to ethyl acetate, followed by recrystallization from CHCl3/methanol af-
forded 5i as deep blue crystals in 19% yield (68 mg). M.p. >3008C;
1H NMR (500 MHz, [D7]DMF): d=1.86 (b, 2H), 3.79 (s, 24H), 7.18 (m,
16H), 7.55 (d, 8H), 7.92 (d, 8H), 8.69 ppm (m, 16H); MS (MALDI-
TOF): m/z: 1387.6 [M+]; elemental analysis calcd (%) for C100H74N8: C
86.55, H 5.37, N 8.07; found: C 86.42, H 5.31, N 8.05.
Optical spectroscopy: CHCl3 and trifluoroacetic acid employed for the
fluorescence measurements were of UV spectroscopic grade. CHCl3 sta-
bilized by 0.5% ethanol supplied by Nacalai Tesque was used for MCD
measurements. Steady-state absorption and fluorescence measurements
were carried out using a Cary 5000 UV/Vis-NIR spectrophotometer and
a Spectronics Instrument 8100 spectrofluorometer. For all measurements,
the temperature was kept constant at (298ꢁ1) K. All absorption meas-
urements were carried out in 50 mm cuvettes. Molar absorption coeffi-
cients were determined from N=4 independent solutions and only very
dilute stock solutions (1–10 mm) were used in these studies to avoid ag-
gregation or solubility problems. The uncertainties of measurement
amount to around ꢁ10% for 5b and 5d and to ꢁꢃ5% for the other
dyes. Fluorescence studies were performed in 10 mm cuvettes (908 stan-
dard geometry, polarizers set at 54.78 for emission and 08 for excitation)
with optically dilute solutions with an absorbance between 0.05 and 0.1
at the maximum of the B band. The fluorescence quantum yields (Ff)
were determined relative to cresyl violet in methanol (Ff =0.61ꢁ0.01).[56]
The fluorescence spectra presented here were spectrally corrected.[57] The
uncertainty of the Ff measurements were determined to be ꢁ20%. Fluo-
rescence lifetimes (tf) were determined using a unique customized laser-
impulse fluorometer with picosecond time resolution.[56] The fluorescence
was collected at right angles (polarizer set at 54.78; monochromator with
spectral bandwidths of 16 nm). While realizing typical instrumental re-
sponse functions of full-width half-maximum (fwhm) of around (25–
30) ps, the time division was 4.8 pschannelꢀ1 and the experimental accu-
racy amounted to ꢁ3 ps, respectively. The laser beam was attenuated
using a double prism attenuator from LTB, and typical excitation ener-
gies were in the nanowatt to microwatt range (average laser power). The
fluorescence lifetime profiles were analyzed using a PC with the Global
Unlimited V2.2 software package (Laboratory for Fluorescence Dynam-
ics, University of Illinois). The goodness-of-fit of the single decays as
judged by reduced chi-squared (cR2) and the autocorrelation function
C(j) of the residuals was always below cR2 <1.2. The parallel spectropho-
tometric and MCD titrations were performed in CHCl3 using a Hitachi
U-3410 spectrophotometer and a Jasco J-725 spectrodichrometer coupled
to a Jasco electromagnet that produces a magnetic field of up to 1.09 T
using a 10 mm cuvette.
[1] a) R. Bonnett, Chem. Soc. Rev. 1995, 24, 19–33; b) Z. Zhang, G. M.
Ferrence, T. D. Lash, Org. Lett. 2009, 11, 1249–1252; c) T. Lu, P.
Shao, I. Mathew, A. Sand, W. Sun, J. Am. Chem. Soc. 2008, 130,
15782–15783; d) C. Comuzzi, S. Cogoi, M. Overhand, G. A. Van der
Marel, H. S. Overkleeft, L. E. Xodo, J. Med. Chem. 2006, 49, 196–
204; e) M. Drobizhev, Y. Stepanenko, Y. Dzenis, A. Karotki, A.
Rebane, P. N. Taylor, H. L. Anderson, J. Phys. Chem. B 2005, 109,
7223–7236.
[2] a) C. Ikeda, N. Sakamoto, T. Nabeshima, Org. Lett. 2008, 10, 4601–
4604; b) D. Wu, A. B. Descalzo, F. Weik, F. Emmerling, Z. Shen, X.-
Z. You, K. Rurack, Angew. Chem. 2008, 120, 199–203; Angew.
Chem. Int. Ed. 2008, 47, 193–197; c) X.-J. Zhu, S.-T. Fu, W.-K.
Wong, J.-P. Guo, W.-Y. Wong, Angew. Chem. 2006, 118, 3222–3226;
Angew. Chem. Int. Ed. 2006, 45, 3150–3154.
[3] a) J. M. Lim, Z. S. Yoon, J.-Y. Shin, K. S. Kim, M.-C. Yoon, D. Kim,
Chem. Commun. 2009, 261–273; b) K. S. Kim, Z. S. Yoon, A. B.
Ricks, J.-Y. Shin, S. Mori, J. Sankar, S. Saito, Y. M. Jung, M. R. Wa-
sielewski, A. Osuka, D. Kim, J. Phys. Chem. A 2009, 113, 4498–
4506; c) Z. S. Yoon, D.-G. Cho, K. S. Kim, J. L. Sessler, D. Kim, J.
Am. Chem. Soc. 2008, 130, 6930–6931.
[4] M. Calvete, G. Y. Yang, M. Hanack, Synth. Met. 2004, 141, 231–243.
[5] a) D. Holten, D. F. Bocian, J. S. Lindsey, Acc. Chem. Res. 2002, 35,
57–69; b) X. He, H. Liu, Y. Li, Y. Liu, F. Lu, Y. Li, D. Zhu, Macro-
mol. Chem. Phys. 2005, 206, 2199–2205; c) R. Iqbal, S. C. Moratti,
A. B. Holmes, G. Yahioglu, L. R. Milgrom, F. Cacialli, J. Morgado,
R. H. Friend, J. Mater. Sci.: Mater. Electron. 2000, 11, 97–103;
d) R. W. Wagner, J. S. Lindsey, J. Seth, V. Palaniappan, D. F. Bocian,
J. Am. Chem. Soc. 1996, 118, 3996–3997.
[6] a) D. Wu, Z. Shen, Z.-L. Xue, X.-Z. You, Chin. J. Inorg. Chem.
2007, 23, 1–14; b) J. A. A. W. Elemans, R. Van Hameren, R. J. M.
Nolte, A. E. Rowan, Adv. Mater. 2006, 18, 1251–1266; c) C.-P.
Hsieh, H.-P. Lu, C.-L. Chiu, C.-W. Lee, S.-H. Chuang, C.-L. Mai, W.-
N. Yen, S.-J. Hsu, E. W.-G. Diau, C.-Y. Yeh, J. Mater. Chem. 2010,
20, 1127–1134; d) S. Hayashi, M. Tanaka, H. Hayashi, S. Eu, T.
Umeyama, Y. Matano, Y. Araki, H. Imahori, J. Phys. Chem. C 2008,
112, 15576–15585.
Calculations: Molecular structures were optimized for free-base and/or
ZnII complexes porphyrin (P), tetraphenylporphyrin (TPP), tetrabenzo-
porphyrin (TBP), tetraphenyltetrabenzoporphyrin (TPTBP), tetranaph-
thoporphyrin (TNP), tetraphenyltetranaphthoporphyrin (TPTNP), tetra-
ACHTUNGTRENNUNGacenaphthoporphyrin (TANP), tetraphenyltetraacenaphthoporphyrin
(TPTANP), tetraphenanthroporphyrin (TPhenP), tetraphenyltetraphe-
[7] a) I. A. Maretina, Russ. J. Gen. Chem. 2009, 79, 1544–1581; b) H. L.
Anderson, Chem. Commun. 1999, 2323–2330. H. Zou, M. J. Ther-
ien, J. K. Blasie, J. Phys. Chem. B 2008, 112, 1350–1357; c) M. U.
Winters, J. Karnbratt, M. Eng, C. J. Wilson, H. L. Anderson, B. Al-
binsson, J. Phys. Chem. C 2007, 111, 7192–7199; d) Z. Shen, H.
Uno, Y. Shimizu, N. Ono, Org. Biomol. Chem. 2004, 2, 3442–3447;
e) S. M. Kuebler, R. G. Denning, H. L. Anderson, J. Am. Chem. Soc.
nanthroporphyrin (TPTPhenP, 5a), and dodecaphenylporphyrin (DPP)
Chem. Eur. J. 2011, 17, 8965 – 8983
ꢄ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
8981

Z. Shen, N. Kobayashi, K. Rurack et al.
2000, 122, 339–347; f) P. N. Taylor, A. P. Wylie, J. Huuskonen, H. L.
Anderson, Angew. Chem. 1998, 110, 1033–1037; Angew. Chem. Int.
Ed. 1998, 37, 986–989.
[32] Gaussian 03, Revision C02, M. J. Frisch, G. W. Trucks, H. B. Schle-
gel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, J. A. Montgom-
ery, Jr., T. Vreven, K. N. Kudin, J. C. Burant, J. M. Millam, S. S.
Iyengar, J. Tomasi, V. Barone, B. Mennucci, M. Cossi, G. Scalmani,
N. Rega, G. A. Petersson, H. Nakatsuji, M. Hada, M. Ehara, K.
Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda,
O. Kitao, H. Nakai, M. Klene, X. Li, J. E. Knox, H. P. Hratchian,
J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E.
Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W.
Ochterski, P. Y. Ayala, K. Morokuma, G. A. Voth, P. Salvador, J. J.
Dannenberg, V. G. Zakrzewski, S. Dapprich, A. D. Daniels, M. C.
Strain, O. Farkas, D. K. Malick, A. D. Rabuck, K. Raghavachari,
J. B. Foresman, J. V. Ortiz, Q. Cui, A. G. Baboul, S. Clifford, J. Cio-
slowski, B. B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I. Komaro-
mi, R. L. Martin, D. J. Fox, T. Keith, M. A. Al-Laham, C. Y. Peng,
A. Nanayakkara, M. Challacombe, P. M. W. Gill, B. Johnson, W.
Chen, M. W. Wong, C. Gonzalez, J. A. Pople, Gaussian, Inc., Wall-
ingford CT, 2004.
[8] a) R. Misra, T. K. Chandrashekar, Acc. Chem. Res. 2008, 41, 265–
279; b) J. L. Sessler, D. Seidel, Angew. Chem. 2003, 115, 5292–5333;
Angew. Chem. Int. Ed. 2003, 42, 5134–5175; c) J. S. Reddy, V. G.
Anand, J. Am. Chem. Soc. 2009, 131, 15433–15439; d) Y. Kami-
mura, S. Shimizu, A. Osuka, Chem. Eur. J. 2007, 13, 1620–1628;
e) Y. Inokuma, T. Matsunari, N. Ono, H. Uno, A. Osuka, Angew.
Chem. 2005, 117, 1890–1894; Angew. Chem. Int. Ed. 2005, 44, 1856–
1860.
[9] a) S. Tokuji, T. Yurino, N. Aratani, H. Shinokubo, A. Osuka, Chem.
Eur. J. 2009, 15, 12208–12211; b) M. Siczek, P. J. Chmielewski,
Angew. Chem. 2007, 119, 7576–7580; Angew. Chem. Int. Ed. 2007,
46, 7432–7436; c) H. Uno, Y. Kitawaki, N. Ono, Chem. Commun.
2002, 116–117; d) T. Ogawa, Y. Nishimoto, N. Yoshida, N. Ono, A.
Osuka, Angew. Chem. 1999, 111, 140–142; Angew. Chem. Int. Ed.
1999, 38, 176–179.
[10] a) T. D. Lash, J. Porphyrins Phthalocyanines 2001, 5, 267–288;
b) T. D. Lash in The Porphyrin Handbook, Vol. 2 (Eds.: K. M.
Kadish, K. M. Smith, R. Guilard), Academic, New York, 1999,
Chapter 10, pp. 125–199, and references therein.
[11] a) N. K. S. Davis, M. Pawlicki, H. L. Anderson, Org. Lett. 2008, 10,
3945–3947; b) J. M. Manley, T. J. Roper, T. D. Lash, J. Org. Chem.
2005, 70, 874–891; c) N. Ono, T. Yamamoto, N. Shimada, K. Kuroki,
M. Wada, R. Utsunomiya, T. Yano, H. Uno, T. Murashima, Hetero-
cycles 2003, 61, 433–447.
[12] R. E. Haddad, S. Gazeau, J. Pecaut, J. Marchon, C. J. Medforth,
J. A. Shelnutt, J. Am. Chem. Soc. 2003, 125, 1253–1268.
[13] J. A. Shelnutt, X.-Z. Song, J.-G. Ma, S.-L. Jia, W. Jentzen, C. J. Med-
forth, Chem. Soc. Rev. 1998, 27, 31–41.
[14] S. A. Vinogradov, D. F. Wilson, J. Chem. Soc. Perkin Trans. 2 1995,
103–111.
[15] V. N. Kopranenkov, A. M. Vorotnikov, S. N. Dashkevich, E. A. Lu-
kꢅyanets, Russ. J. Gen. Chem. 1985, 55, 803–809.
[16] J. D. Spence, T. D. Lash, J. Org. Chem. 2000, 65, 1530–1539.
[17] N. Ono, H. Hironaga, K. Ono, S. Kaneko, T. Murashima, T. Ueda,
C. Tsukamura, T. Ogawa, J. Chem. Soc. Perkin Trans. 1 1996, 417–
423.
[18] T. D. Lash, P. Chandrasekar, J. Am. Chem. Soc. 1996, 118, 8767–
8768.
[19] J. Mack, Y. Asano, N. Kobayashi, M. J. Stillman, J. Am. Chem. Soc.
2005, 127, 17697–17711.
[20] a) J. Michl, J. Am. Chem. Soc. 1978, 100, 6801–6811; b) J. Michl, Tet-
rahedron 1984, 40, 3845–3934.
[21] M. Gouterman in The Porphyrins, Vol. 3 (Ed.: D. Dolphin), Aca-
demic, New York, 1978, pp. 1–165.
[22] H.-J. Xu, Z. Shen, T. Okujima, N. Ono, X.-Z. You, Tetrahedron Lett.
2006, 47, 931–934.
[23] J. Mack, M. J. Stillman, N. Kobayashi, Coord. Chem. Rev. 2007, 251,
429–453.
[24] a) W. J. Moffitt, J. Chem. Phys. 1954, 22, 320–333; b) W. J. Moffitt, J.
Chem. Phys. 1954, 22, 1820–1829.
[25] a) J. Mack, M. J. Stillman in The Porphyrin Handbook, Vol. 16
(Eds.: K. Kadish, K. Smith, R. Guilard), Academic, New York,
2003, Chapter 103, pp. 43–116; b) J. Mack, M. J. Stillman, Coord.
Chem. Rev. 2001, 219–221, 993–1032.
[26] M. H. Perrin, J. Chem. Phys. 1973, 59, 2090–2104.
[27] J. Mack, M. Bunya, Y. Shimizu, H. Uoyama, N. Komobuchi, T. Oku-
jima, H. Uno, S. Ito, M. J. Stillman, N. Ono, N. Kobayashi, Chem.
Eur. J. 2008, 14, 5001–5020.
[33] When shoulders appear on the low-energy side of the Q bands like
in 5g and 5h (more prominently) and in 5d–f (less prominently),
the maxima of the respective 0,0 bands after spectral deconvolution
on the energy scale have been used to calculate the Stokes shifts.
[34] A. Harriman, R. J. Hosie, J. Chem. Soc. Faraday Trans. 2 1981, 77,
1695–1702.
[35] S. Gentemann, C. J. Medforth, T. P. Forsyth, D. J. Nurco, K. M.
Smith, J. Fajer, D. Holten, J. Am. Chem. Soc. 1994, 116, 7363–7368.
[36] N. S. Savenkova, R. T. Kuznetsova, I. N. Lapin, V. A. Svetlichnyi,
G. V. Mayer, P. A. Shatunov, Opt. Spectrosc. 2005, 99, 751–758.
[37] W. Siebrand, J. Chem. Phys. 1967, 46, 440–447.
[38] knr =(1ꢀFf)ꢆtfꢀ1
.
[39] R. Bonnett, A. Harriman, A. N. Kozyrev, J. Chem. Soc. Faraday
Trans. 1992, 88, 763–769.
[40] a) A. T. Gradyushko, D. T. Kozhich, K. N. Solovꢅev, M. P. Tsvirko, J.
Appl. Spectrosc. 1970, 12, 838–840; b) K. N. Solovꢅev, M. P. Tsvirko,
A. T. Gradyushko, D. T. Kozhich, Opt. Spectrosc. 1972, 33, 480–483.
[41] E. A. Borisevich, G. D. Egorova, V. N. Knyukshto, K. N. Solovꢅev,
Opt. Spectrosc. 1987, 63, 34–37.
[42] The formalism described in Ref. [37] yields a correlation of r<0.896
for a quantitative analysis of the heavy-atom effect based on a plot
ꢀ
ACTHNGUTERNNUG
of logk versus logx2, in which k=(t5x)ꢀ1 (t5a)ꢀ1 (5x=5d, 5e, 5 f)
and x is the spin–orbit coupling constant of the halogen atoms, be-
cause multiple processes are apparently involved in the quenching.
The fact that the fluorescence-band positions of 5a and 5c–f are not
identical hints at the influence of more than one fluorescence
quenching process, since the emission band positions of the corre-
sponding meso-tetrakis(4-halogenophenyl)porphyrin derivatives are
identical within experimental error.[41] Further evidence is provided
by the longer lifetime of fluoro derivative 5c relative to the refer-
ence compound, 5a, which lacks a heavy-atom substituent.
[43] A. Stone, E. B. Fleischer, J. Am. Chem. Soc. 1968, 90, 2735–2748.
[44] a) R. A. Freitag, D. G. Whitten, J. Phys. Chem. 1983, 87, 3918–3925;
b) A. Rosa, G. Ricciardi, E. J. Baerends, A. Romeo, L. M. Scolaro,
J. Phys. Chem. A 2003, 107, 11468–11482; c) C. J. Medforth, R. E.
Haddad, C. M. Muzzi, N. R. Dooley, L. Jaquinod, D. C. Shyr, D. J.
Nurco, M. M. Olmstead, K. M. Smith, J.-G. Ma, J. A. Shelnutt,
Inorg. Chem. 2003, 42, 2227–2241.
[45] T. V. Karmanova, T. V. Gromova, B. D. Berezin, A. S. Semeikin,
S. A. Syrbu, Russ. J. Gen. Chem. 2001, 71, 803–808.
[46] “Adj” and “opp” indicate the two possible isomers formed when
either two adjacently or two oppositely located (dimethylamino)-
phenyl groups are protonated.
[28] T. Fukuda, S. Homma, N. Kobayashi, Chem. Eur. J. 2005, 11, 5205–
5216.
[47] R. I. Walter, E. C. A. Ojadi, H. Linschitz, J. Phys. Chem. 1993, 97,
13308–13312.
[29] a) J. S. Lindsey, I. C. Schreiman, H. C. Hsu, P. C. Kearney, A. M.
Marguerettaz, J. Org. Chem. 1987, 52, 827–836; b) J. S. Lindsey,
R. W. Wagner, J. Org. Chem. 1989, 54, 828–836.
[48] E. C. A. Ojadi, H. Linschitz, M. Gouterman, R. I. Walter, J. S. Lind-
sey, R. W. Wagner, P. R. Droupadi, W. Wang, J. Phys. Chem. 1993,
97, 13192–13197.
[30] B. H. Novak, T. D. Lash, J. Org. Chem. 1998, 63, 3998–4010.
[31] T. D. Lash, B. H. Novak, Angew. Chem. 1995, 107, 723–725; Angew.
Chem. Int. Ed. Engl. 1995, 34, 683–685.
[49] a) V. N. Kopranenkov, A. M. Vorotnikov, S. N. Dashkevich, E. A.
Lukꢅyanets, Russ. J. Gen. Chem. 1995, 55, 900–907; b) V. N. Kopra-
nenkov, A. M. Vorotnikov, E. A. Lukꢅyanets, Russ. J. Gen. Chem.
8982
www.chemeurj.org
ꢄ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 8965 – 8983

Phenanthroporphyrins
FULL PAPER
1979, 49, 2758–2759; c) J. L. Retsek, C. J. Medforth, D. J. Nurco, S.
Gentemann, V. S. Chirvony, K. M. Smith, D. Holten, J. Phys. Chem.
B 2001, 105, 6396–6411; d) E. A. Lukꢅyanets, S. N. Dashkevich, N.
Kobayashi, Russ. J. Gen. Chem. 1993, 63, 985–988.
[53] C. Hansch, A. Leo, R. W. Taft, Chem. Rev. 1991, 91, 165–195.
[54] a) G. D. Hartman, L. M. Weinstock, Org. Synth. 1979, 59, 183; b) U.
Schçllkopf, D. Hoppe, R. Jentsch, Chem. Ber. 1975, 108, 1580–1592.
[55] M. J. Plater, S. Aiken, G. Bourhill, Tetrahedron 2002, 58, 2405–2413.
[56] Y.-H. Yu, A. B. Descalzo, Z. Shen, H. Rçhr, Q. Liu, Y.-W. Wang, M.
Spieles, Y.-Z. Li, K. Rurack, X.-Z. You, Chem. Asian J. 2006, 1,
176–187.
[57] U. Resch-Genger, D. Pfeifer, C. Monte, W. Pilz, A. Hoffmann, M.
Spieles, K. Rurack, J. Hollandt, D. Taubert, B. Schçnenberger, P.
Nording, J. Fluoresc. 2005, 15, 315–336.
[50] a) G.-Z. Wu, H.-K. Leung, W.-X. Gan, Tetrahedron 1990, 46, 3233–
3244; b) G. Hariprasad, S. J. Dahal, G. G. Maiya, J. Chem. Soc.
Dalton Trans. 1996, 3429–3436; c) A. Bettelheim, B. A. White, S. A.
Raybuck, R. W. Murray, Inorg. Chem. 1987, 26, 1009–1017; d) C. K.
Chang, J. Fajer, J. Am. Chem. Soc. 1980, 102, 848–851; e) J. Takeda,
M. Sato, Chem. Lett. 1995, 939–940.
[51] K. M. Kadish, E. Van Caemelbecke, G. Royal, in The Porphyrin
Handbook, Vol. 8 (Eds.: K. M. Kadish, K. M. Smith, R. Guilard),
Academic, New York, 1999, Chapter 55, pp. 1–114.
[52] K. M. Kadish, M. M. Morrison, J. Am. Chem. Soc. 1976, 98, 3326–
3328.
Received: September 8, 2010
Revised: April 14, 2011
Published online: June 16, 2011
Chem. Eur. J. 2011, 17, 8965 – 8983
ꢄ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
8983