Spectroscopy and photophysics of self-organized zinc porphyrin nanolayers. 1. Optical spectroscopy of excitonic interactions involving the soret band

Article abstract of DOI:10.1021/jp058054o

The photophysical properties of excited singlet states of zinc tetra-(p-octylphenyl)-porphyrin in 5-25-nm-thick films spin-coated onto quartz slides have been investigated by optical spectroscopy. Analysis of the polarized absorption spectra using a dipole-dipole exciton model with two mutually perpendicular transition dipole moments per molecule shows that the films are built from linear aggregates, i.e., stacks with a slipped-deck-of-cards configuration. The molecular planes of the porphyrins in the stacks are found to be perpendicularly oriented with respect to the substrate plane. Assuming a value of 2-3 for the dielectric constant of the film, from the excitonic shift, an angle of 44° ± 3° and an interplanar distance of 0.35-0.36 nm between adjacent porphyrins are calculated, close to the ground-state geometry in solution. The ordering in these films was further investigated by the effects of various solvents and temperature annealing. Spin-coating from toluene as a solvent results in highly ordered films, and annealing of these films has little effect on their absorption spectra. However, spin-coating from chloroform or pyridine or exposure of the films to these solvents in their vapor phases changes their ordering presumably due to incorporation of residual solvent molecules. Annealing yields absorption spectra identical to those of films spin-coated from toluene. The absorption spectra are insensitive to atmospheric moisture, in contrast to those of zinc tetraphenylporphyrin films lacking octyl substituents.

Full text of DOI:10.1021/jp058054o

J. Phys. Chem. B 2005, 109, 17031-17037
17031
ARTICLES
Spectroscopy and Photophysics of Self-Organized Zinc Porphyrin Nanolayers. 1. Optical
Spectroscopy of Excitonic Interactions Involving the Soret Band
Harry Donker,* Rob B. M. Koehorst, and Tjeerd J. Schaafsma
Laboratory for Biophysics, Department of Agrotechnology and Food Sciences, Wageningen UniVersity,
Dreijenlaan 3, 6703 HA Wageningen, The Netherlands
ReceiVed: February 16, 2005; In Final Form: May 19, 2005
The photophysical properties of excited singlet states of zinc tetra-(p-octylphenyl)-porphyrin in 5-25-nm-
thick films spin-coated onto quartz slides have been investigated by optical spectroscopy. Analysis of the
polarized absorption spectra using a dipole-dipole exciton model with two mutually perpendicular transition
dipole moments per molecule shows that the films are built from linear aggregates, i.e., stacks with a slipped-
deck-of-cards configuration. The molecular planes of the porphyrins in the stacks are found to be perpendicularly
oriented with respect to the substrate plane. Assuming a value of 2-3 for the dielectric constant of the film,
from the excitonic shift, an angle of 44° ( 3° and an interplanar distance of 0.35-0.36 nm between adjacent
porphyrins are calculated, close to the ground-state geometry in solution. The ordering in these films was
further investigated by the effects of various solvents and temperature annealing. Spin-coating from toluene
as a solvent results in highly ordered films, and annealing of these films has little effect on their absorption
spectra. However, spin-coating from chloroform or pyridine or exposure of the films to these solvents in
their vapor phases changes their ordering presumably due to incorporation of residual solvent molecules.
Annealing yields absorption spectra identical to those of films spin-coated from toluene. The absorption spectra
are insensitive to atmospheric moisture, in contrast to those of zinc tetraphenylporphyrin films lacking octyl
substituents.
Introduction
This paper on excitonic interactions in the second excited
singlet state of zinc tetra-(p-octylphenyl)-porphyrin (ZnTOPP)
(Figure 1) is the first of a series of three closely related papers
reporting the spectroscopic and photophysical properties of the
first and second excited singlet states as well as the lowest
excited triplet state in 5-25-nm-thick, self-organized films of
this porphyrin on quartz slides as a substrate and the effects of
various solvents and temperature annealing on their degree of
order. The following paper treats excitation transport in these
films studied by doping with varying concentrations of fluo-
rescence quenchers. Finally, the results of magnetic resonance
spectroscopy of triplet states in the porphyrin films are reported
and discussed in the third paper.
Figure 1. Molecular structure of zinc tetra-(p-octylphenyl)-porphyrin;
R ) -C
8 17
H .
There is a rapidly growing interest in the properties of organic
films with nanometer thicknesses in optoelectronic devices as
a result of the relatively simple deposition methods, low
processing costs, and a large variety of molecular structures as
compared to those of inorganic layers. During the past decade,
the number of applications of inorganic films has rapidly
from photo detectors, light-emitting diodes, sensors, solar cells,
data storage devices, and displays to coatings for (bio)molecular
recognition or nonlinear optics. For almost all of these applica-
tions, the creation, recombination, and transport of excitons and/
or charge carriers play an essential role. A drawback of organic
films is their usually low conductivity and dielectric constant,
1
,2
increased, and a breakthrough of nanoscale organic film
5
so that charge separation is energetically less favorable. This
3,4
technology appears to be just around the corner. These kinds
of films have a broad spectrum of potential applications, ranging
drawback may be partially compensated by the ordering of these
materials at the molecular scale, increasing the efficiency of
these transport processes. Several recent reviews cover the
*
Author to whom correspondence should be addressed. Present ad-
6-8
progress of research on organic films.
dress: Laboratory for Inorganic Chemistry, Delft University of Technology,
Delft, The Netherlands. Phone: +31-15-2783891. Fax: +31-15-2788047.
E-mail: H.Donker@tnw.tudelft.nl.
The optoelectronic properties of the films are largely deter-
mined by the degree of molecular ordering, the domain size,
1
0.1021/jp058054o CCC: $30.25 © 2005 American Chemical Society
Published on Web 08/18/2005

17032 J. Phys. Chem. B, Vol. 109, No. 36, 2005
Donker et al.
the distance and orientation of the molecules with respect to
each other and to the substrate, and the optical, electric, and
electrochemical properties of the film as well as those of the
9
substrate. In particular, tetrapyrrole compounds, such as por-
phyrins and phthalocyanines, and their supramolecular as-
10
semblies offer fascinating options due to their self-organizing
and steric properties,11-17 stability, optical absorption in the
visible range, and variability. The electrochemical properties
can be controlled by varying the metal center, whereas the
molecular architecture of the films is determined by substituting
18
suitable functionalizing side groups.
The orientation of the porphyrins with respect to the substrate
in nanolayers is essential for most applications, which require
anisotropic transport of excitons or charge carriers to the
substrate. Much progress has been made in controlling and
Figure 2. Absorption spectra of ZnTOPP in toluene (dashed line),
hexane (dotted line), and a film spin-coated from 5 mM toluene solution
onto quartz (solid line). The absorption spectrum of the spin-coated
film was obtained with unpolarized light at an incidence angle of 10°
and is uncorrected for reflection effects. For clarity, the spectra have
been scaled to the same height.
1
9-22
changing the orientation of porphyrins in such systems.
Experimental Section
Thin ZnTOPP films of up to several tenths of nanometers in
thickness on quartz plates (Suprasil, L 15 mm, 1-mm-thickness)
were prepared by spin-coating at 2500 rpm from 1-5 mM
toluene, pyridine, or chloroform solutions. Before spin-coating,
the quartz plates were subsequently rinsed with aqua regia,
water, methanol, and toluene and blown dry with nitrogen.
Through the use of the optical density of the films at 550 nm,
their estimated thicknesses were 5-25 ( 5 nm, respectively.
For the thickest film (25 nm), the amount of deposited porphyrin
was also determined by dissolving the film in a known volume
2
3
of toluene. Through the use of the extinction coefficient, ꢀ550
-
1
)
22 000 L M cm , the calculated thickness was in good
agreement with the above-mentioned value and corresponded
to ∼100 porphyrin layers for the investigated film. All solvents
were p.a. quality, unless stated otherwise. Films of the analogous
compounds without alkyl chains, i.e., tetraphenyl porphyrins
Figure 3. Absorption spectra of (A) a ZnTOPP film spin-coated from
5 mM toluene solution, (B) a ZnTOPP film spin-coated from 5 mM
chloroform solution, and (C) a ZnTPP film spin-coated from 5 mM
chloroform solution. Solid lines are before, dashed lines after 5 min of
annealing at 175 °C. The spectra have been scaled to the highest
intensity in the unannealed films.
(TPP), were used as references.
ZnTOPP and ZnTPP were prepared by metallization of H2-
TOPP and H2TPP, respectively, by refluxing in dimethylfor-
mamide (DMF) with ZnCl2 (Merck, p.a.).24 H2TOPP and H2TPP
were synthesized by condensation of 4-(n-octyl)benzaldehyde
and benzaldehyde (Kodak, 99%), respectively, with pyrrole
xenon lamp as an excitation source. Spectra were measured in
the front-face mode at an angle close to 0° or in the right-angle
mode at ∼45° between the normal to the substrate and the
incident light beam. Excitation spectra were corrected for lamp
output, and emission spectra were corrected for photomultiplier
sensitivity. Both types of spectra were also corrected for the
monochromator transmission. Low-temperature (1.4 K) fluo-
rescence measurements were carried out using a homemade
(
Janssen Chimica, 99%) in refluxing propionic acid (Merck,
25,26
z.s.).
The porphyrins were purified by chromatography on
silica (Merck, silica gel 60) with toluene or chloroform as the
eluent. For duplicate samples, ZnTOPP purchased from Por-
phyrin Products was used. All porphyrins are estimated to be
>
99% pure as shown by thin-layer chromatography, absorption,
and fluorescence spectroscopy.
Throughout this paper, films spin-coated from toluene are
named films 1, those spin-coated from chloroform films 2, those
spin-coated from toluene and subsequently exposed to chloro-
form vapor films 3, and those spin-coated from pyridine films
+
liquid helium bath cryostat and a CW Ar ion laser (Coherent
CR-5) as excitation source. Emission light was selected with
an OG 540 cutoff filter and a Jobin-Yvon HR1000 mono-
chromator and detected with an S20 (EMI) photomultiplier.
4
.
Absorption spectra of spin-coated films were recorded on a
Results and Discussion
Cary 5E spectrophotometer, equipped with an integrating diffuse
reflectance sphere (DRA-CA-50, Labsphere), varying the angle
between the normal of the substrate and the incident light beam
between 10° and 50° The absorption spectra are uncorrected
for reflection since the films are highly anisotropic, as is evident
from the absorption spectra. Polarized absorption spectra were
recorded by placing sheet polarizers (Polaroid type HNP′B) in
the sample and reference beams, with their polarization direction
either perpendicular (As) or parallel (Ap) to the plane of incidence
of the incoming light.
Absorption Spectra. Absorption spectra of spin-coated
ZnTOPP and ZnTPP films and solutions in toluene and hexane
are shown in Figures 2 and 3. The spectra of the films are
uncorrected for reflection effects. Results for purchased and
home-synthesized ZnTOPP were identical.
The absorption spectra of ZnTOPP and ZnTPP are similar
2
3
to those of a number of other porphyrins. The weak bands at
∼585 (film), 590 (toluene), and 584 nm (hexane) are assigned
to the Q(0,0) transition, the bands at ∼550 (film, toluene) and
546 nm (hexane) to the Q(1,0) transition, and the weak band at
507 (film), 515 (toluene), and 509 nm (hexane) to the Q(2,0)
Fluorescence spectra were recorded at room temperature using
a Spex Fluorolog 3-22 fluorometer equipped with a 450-W

Optical Spectroscopy of Zinc Porphyrin Nanolayers
J. Phys. Chem. B, Vol. 109, No. 36, 2005 17033
Figure 5. Polarized absorption spectra of ZnTOPP films spin-coated
from 5 mM toluene solution onto quartz for two angles of incidence.
Figure 4. Absorption spectra of a ZnTOPP film spin-coated from
pyridine vs time.
The B-bands of ZnTOPP films are similar to those of spin-
transition. The strong absorption bands at 425 (toluene) and 416
nm (hexane) with weak sidebands in the 350-490-nm region
are assigned to the strongly allowed B transitions. By contrast,
the absorption spectra of the ZnTOPP films show three or four
different peaks, depending on the sample treatment. ZnTOPP
spin-coated from toluene has absorption bands at 375, 413, and
27
coated films of H2TOPP and aggregates of ZnTPP derivatives
1
1-13
in solution.
The B-band of these aggregates is split into a
strong red-shifted component and a weak blue-shifted compo-
nent. Similar splittings have been reported for porphyrin LB
2
8,29
films.
For the aggregated derivative reported in refs 11-13
with a molecular structure closely resembling that of ZnTOPP,
the B-band has two main components, i.e., at 446 and 375 nm,
very similar to the 450 and 375 nm bands for the ZnTOPP films
in this work. The 413 nm absorption band of ZnTOPP films
was found to be absent for the aggregates in solution. This band
is very close to the B-band of ZnTOPP in hexane (cf. Figure 2)
and can be related either to ZnTOPP monomers embedded in
the alkyl side chains of the ZnTOPP aggregates or to an
excitonic shift of the B-band resulting from a slightly different
geometry of the aggregates in the film as compared to that in
solution.
Polarized Absorption Spectra. The strong polarization
anisotropy of the absorption spectra at different incidence angles
R, as shown in Figure 5, indicates that the ZnTOPP molecules
in the film spin-coated from toluene are highly ordered. Defining
the absorbance at perpendicular and parallel polarization as As
and Ap, respectively, from the 375-nm absorption at different
incidence angles R (Figure 5) the reduced linear dichroism
4
50 nm, whereas ZnTOPP spin-coated from chloroform shows
an additional band at 390 nm. This band disappears upon
annealing of the film, however, as shown in Figure 3B. For
∼
25-nm-thick films spin-coated from 5 mM toluene solution,
the band at 450 nm seemed to consist of two components;
however, spin-coated 5-nm-thick films from 1 mM solutions
produced only a single band without shoulders, suggesting that
the shoulder observed for the thicker film is due to reflection
effects, which are expected to increase with the film thickness.
ZnTPP films show only a single, broad, structureless absorption
band centered at 440 nm. The 375 nm absorption band is also
found in the fluorescence excitation spectra. Its intensity
increases with increasing layer thickness, demonstrating that the
observed polarization is a bulk property of the film and not
that of a surface layer. This is further supported by the
observation that in films spin-coated from pyridine this band is
absent at first but fully reappears with time during slow
evaporation of the solvent from the film (Figure 4).
The shifts and splittings of the B-bands in the films with
respect to their position in dilute solutions may in principle result
from (i) porphyrin-substrate interaction, (ii) the presence of
different aggregates, (iii) excitonic interactions between neigh-
boring porphyrin molecules in the film, or from combinations
of i, ii, and iii. The interaction of type i should give rise to
spectra that depend on the thickness of the film, which is not
observed. Note also that the Q-band for the films is at almost
the same wavelength as for solutions, supporting the conclusion
that there are no strong porphyrin-substrate interactions. The
two bands at 375 and 450 nm in the B-band region cannot be
ascribed to the presence of two different types of aggregates
either, since the results of fluorescence quenching measurements
Ap - As
Ap + 2As
r
LD )
(1)
is calculated to be 0.9 ( 0.05, demonstrating that this band
results from a highly ordered molecular array with a polarized
transition parallel to the plane of incidence, i.e., the transition
dipole moment M associated with this band is perpendicular to
the substrate. Since M is in the molecular plane, this implies
that all porphyrin planes in the stacks are perpendicular to the
substrate plane. In agreement with this, the intensity of the 375-
nm band decreases with the decreasing incidence angle as shown
in Figure 6 and completely disappears in spectra recorded at a
normal incidence of the exciting light. The polarization of the
450-nm red-shifted band appears to be perpendicular to the
substrate plane, but reflection effects at this wavelength are too
strong to justify a conclusion about its polarization. At 375 and
450 nm, As and Ap at R ) 10° are almost equal, indicating that
the absorbance is averaged in the substrate plane over the angle
æ (Figure 7A), i.e., there is no preferential orientation of the
various domains around the normal to the substrate plane,
resulting in averaging of the orientational distribution of the
domains and thus of the stacks within the illuminated area. This
implies that optical polarization data only yield information on
the orientation of the porphyrin molecular plane with respect
(paper 2) do not depend on the excitation wavelength. Efficient
energy transfer between two different aggregates in the films
as a possible cause of the occurrence of two bands can also be
excluded, since the distance between aggregates is at least 5
times that between two porphyrin molecules within one stack,
considering the presence of alkyl and phenyl substituents at each
porphyrin macrocycle.
Therefore, the shift of the prominent ∼420-nm B-band in
solution to 450 nm for the film and the splitting into three or
four components must be ascribed to excitonic interactions, and
possible effects of the change of the molecular environment
can be neglected.

17034 J. Phys. Chem. B, Vol. 109, No. 36, 2005
Donker et al.
Figure 8. Axis system for the calculation of the dipole-dipole
excitonic interaction.
Figure 6. Absorbance at 375 nm of a spin-coated ZnTOPP film vs
the angle of incidence.
corner shifted arrays (or any other shift direction as long as the
dimer is parallel) yield identical results. The blue-shifted
transition (or one of the two if there are two blue-shifted
transitions) is always polarized perpendicular to the shift
direction. For tetrasulfophenyl porphyrin aggregates in solution,
absorption bands red- and blue-shifted with respect to the
monomer B-band have indeed been found. Flow-induced linear
dichroism measurements have shown that the transition dipole
moments of these bands are mutually perpendicular, where the
red-shifted band was found to be polarized along the long axis
of the aggregate and the blue-shifted band along the short axis,
in agreement with the dipole-dipole exciton model for J-type
linearly aggregated porphyrins.
31
For arrays containing mutually tilted molecular planes, in
principle, transitions to each of the four exciton states are
allowed.
A tilted configuration is highly unlikely, however, in view
of the symmetry of the porphyrin monomers constituting the
Figure 7. Structural arrangement of a slipped-deck-of-cards array. (A)
In the three-dimensional view, the angle R is the incidence angle;
rotation around æ takes the random orientation of the stacks in the
substrate plane into account. (B) In the top view, θ is the angle between
the line connecting the porphyrin centers and the porphyrin molecular
planes. Note that the laboratory frame of reference {x,y,z} of part B
does not coincide with the molecular frame of reference {X,Y,Z}.
32-34
array. In fact, crystal structures of tetraphenyl porphyrins
3
5
and NMR spectra of porphyrin dimers and oligomers in
solution invariably show that adjacent porphyrins in dimers, at
least in their ground state, preferably assume a slipped-deck-
of-cards configuration. Such an arrangement is also supported
by fluorescence anisotropy decay measurements (paper 2).
Spectral shifts and splittings for porphyrin aggregates have
to the substrate and not on the mutual orientation of the
porphyrin planes in the stack with respect to each other. To
determine the mutual orientation and position of the porphyrin
monomers in a stack, additional information is required, e.g.,
the number of exciton components, the most stable ground-
state conformation of oligomers in solution, and/or the crystal
structure.
The spectral effects of the excitonic interactions between the
two degenerate transition moments of metalloporphyrins for
various dimer geometries are well-known.30 For a parallel
porphyrin array and transitions from the ground state to the
excited state, two exciton states are forbidden for one pair and
allowed for another pair, with mutually perpendicular polariza-
tion. For a cofacial dimer, there are two allowed blue-shifted
transitions at the same energy.
frequently been explained by the Kasha dipole-dipole exciton
36,37
model
using a single transition dipole moment for each of
the monomers in the aggregate. Similar to what occurs for
aggregates with a single transition dipole moment per monomer,
for large aggregates of ZnTOPP monomers with two degenerate
transition dipole moments, both the number of exciton states
and the exciton bandwidth increase. The number of allowed
transitions and the direction of their spectral shifts compared
to the monomer band remain the same, but the magnitude of
the shift approaches twice the interaction potential V by
increasing the aggregate size.
The dipolar interaction potential V is given by36-38
V ) V ) V [sin θ sin θ cos ψ - 2 cos θ cos θ ]
ij
ji
0
1
2
1
2
i,j ) X,Y (2)
If the monomers are alternately shifted in two perpendicular
directions, resulting in helical or zigzag-type aggregates, there
is only one allowed transition to a twofold degenerate exciton
state, giving rise to a red or blue spectral shift, depending on
the angle θ between the molecular planes and the center-to-
center vector. For translationally shifted parallel arrays denoted
as “slipped-decks-of-cards”, this degeneracy is lifted, one of
the allowed transitions always being blue-shifted, whereas the
other transition can be blue- or red-shifted from the monomer
position, depending on the angle between the center-to-center
vector and the molecular planes. Edge-to-edge or corner-to-
where the angles θ1, θ2, and ψ are defined in Figure 8 and
2
2
-
4
M
-4
M
3
V0(eV) ) 6.24 × 10
) 6.24 × 10
sin θ (3)
3
3
ꢀrR
ꢀrd
with M expressed in Debye units, R and d in nanometers, ꢀr the
high-frequency dielectric constant, and d the distance between
the molecular planes.
Since we assume that the molecular planes in the aggregate
are parallel, ψ ) 0 and d ) R sin θ (Figure 7B). For parallel

Optical Spectroscopy of Zinc Porphyrin Nanolayers
J. Phys. Chem. B, Vol. 109, No. 36, 2005 17035
and cofacial arrays, the off-diagonal elements VXY ) VYX ) 0,
since in a common laboratory axis system the molecular X- and
Y-transition moments of adjacent porphyrins in this array are
perpendicular.
A linear array of a slipped-deck-of-cards is predicted to have
two allowed exciton states with orthogonal transition dipole
moments, resulting in blue- and red-shifted B-bands with mutual
perpendicular polarization.
between the value of θ for the film and in solution, we may
draw a cautious conclusion that the internal relative dielectric
constant of the film is rather low and equals ꢀr ≈ 2-3.
Substituting ꢀr ) 1.89 for hexane, M ) 9.5 D for the B-band,
calculated from the solution spectra, the previously determined
angle θ ) 41°, and the blue shift of 0.326 eV into eq 3, the
distance d between two adjacent porphyrins is calculated to be
0.37 nm. Replacing these data by those for toluene (ꢀr ≈ 2.24,
λ0 ) 425 nm) as a reference results in θ ) 47° and d ) 0.36
nm and thus does not lead to significant changes in the
interplanar distance. For both media, the calculated value of d
is again rather close to the ground-state value of 0.31 nm,
supporting our previous conclusion that most likely the internal
dielectric constant of the film is ∼2-3.
Recently, it has been shown39 that intensity from the red-
shifted exciton component can be transferred to the Q-band by
nonzero values of VXY ) VYX or higher-order Coulombic
perturbation terms in the Hamiltonian, resulting in enhanced
intensity of this band, as is indeed experimentally found (Figure
2
).
These observations are in agreement with the structure of
For the S2 exciton to be observable in the absorption
spectrum, the excitonic interaction energy in frequency units
must exceed the lifetime broadening of the S2 excited state,
the films containing linear porphyrin stacks with a slipped-deck-
of-cards geometry, with their long axis parallel to the substrate.
Then, the 375-nm band of the ZnTOPP stacks corresponds to a
transition with a polarization along the short axis of the stacks.
Since the 375-nm band has been found to be a bulk property of
the film, we may conclude that the film has a lamellar structure.
The 450-nm band results from two contributions:
-1
which is (2πτ) with τ its lifetime. For a substituted zinc
porphyrin monomer similar to ZnTOPP, τ ≈ 1-2 ps, resulting
-1
12 -1
in τ (S2) ≈ (0.5-1) × 10 s . For its linked porphyrin arrays,
40
τ has been found to be <1 ps , however, and the same
shortening of τ may be expected for the S2 lifetime in the
porphyrin arrays in this work. The fact that both excitonic
components, i.e., the blue- as well as the red-shifted one, are
observed implies that the lifetime broadening of the S2 excited
(
i) interaction of the incident parallel and perpendicularly
polarized light with the X-component of M averaged over the
angle æ (Figure 7A), giving rise to absorbances Ap(av) and As(av)
(ii) reflection effects, resulting in additional polarization with
state must be considerably less than the excitonic interaction
unknown magnitude and sign. Unfortunately, this contribution
prevents an accurate prediction of the net polarization of the
red-shifted exciton component.
14 -1
energy of ∼1.3 × 10
s
derived from the splitting between
the 375 and 450 nm excitonic bands.
As predicted by the dipole-dipole exciton model, the
Q-bands do not show any splitting or shift, due to their much
smaller transition moments M ≈ 0.7 D for Q(0,0), and remain
twofold degenerate. Even though the porphyrins are oriented
perpendicularly to the substrate, the Q-bands do not exhibit any
polarization. As stated before, the Q-bands will predominantly
mix with the red-shifted exciton component of the B-band as
According to the Kasha model, the transition energy of an
aggregate (∆Eagg) is given by36,37
∆
Eagg ) ∆Emon + ∆D ( 2V
(4)
where ∆Emon is the transition energy of the monomer, ∆D is
the difference between the van der Waals terms for the ground
and the excited state, assumed to be equal for the ground and
excited state, and 2V is the exciton interaction energy for an
aggregate of N monomers, with N a large number, as is assumed
to apply to the porphyrin stacks.
3
9
compared with the blue-shifted component, resulting in an
increase of the transition moment of the Q-band parallel to the
substrate. This effect, together with the random orientation of
this transition moment, explains the vanishing of the Q-band
polarization. The absorbances resulting from the interaction of
the electric component E of the incident light with each of the
two perpendicular transition moments add up. Averaging over
all orientations of the stacks in the substrate plane and assuming
that the components Es and Ep are equal results in the
disappearance of the Q-band polarization.
For θ1 ) θ2 ) θ and ψ ) 0, eq 2 transforms to
2
V ) V ) V (1 - 3 cos θ) i,j ) X,Y
(5)
ij
ji
0
with θ the angle between the center-to-center vector R and both
transition dipole moments3 (Figure 7B). Since M lies in the
porphyrin plane, θ is also the angle between the center-to-center
6,37
At the calculated interplanar distance of less than 0.4 nm
between adjacent porphyrins, a simple point dipole model is at
the limit of its validity and may even break down. The point
dipole approximation is only valid when the distance between
vector and that plane. The energy of the blue-shifted transition
blue
is given by substituting θ1 ) θ2 ) 90° in eq 5, yielding ꢀ
)
red
2
2
VYY ) 2V0. However, ꢀ ) 2VXX ) 2V0 (1 - 3 cos θ). Thus,
4
1
the molecules is roughly larger than their size, which does
not apply to the above-mentioned type of aggregates. Using an
θ can be calculated from
4
2,43
ꢀred/ꢀ ) (1 - 3 cos θ)
blue
2
extended dipole model
does not change the essence of the
(6)
discussion above, however, since it has been found to lead to
only minor corrections for d and θ.
A film structure with a linear slipped-deck-of-cards arrange-
Through the use of the B-band of the porphyrin in hexane at
λ0 ) 416 nm as a reference, the shifts of the 375 and 450 nm
exciton bands are calculated as 0.326 and 0.225 eV, respectively,
resulting in θ ≈ 41°. If we use the absorption band of the
porphyrin in toluene at λ0 ) 425 nm, then θ changes to ∼47°.
It is interesting to note that these angles are close to that in the
ground state of dimers of zinc tetraphenylporphyrin derivatives
in solution. NMR ring current shifts for these solution dimers
yield a parallel conformation with a 0.42-nm translation along
the phenyl-phenyl axis and a plane-to-plane distance of 0.31
28,29
ment is similar to that proposed for porphyrin LB films
and
1
2
aggregates in solution. Atomic force microscopy of supra-
molecular porphyrin monolayers deposited from solution on
graphite has shown that the monomers in these layers are
cofacial and perpendicularly oriented on the substrate10 as in
the present work.
Recently, Monte Carlo simulations have been applied to the
same ZnTOPP films using preliminary results from the present
work on the film structure and comparing the simulated data
3
5
nm, resulting in θ ≈ 40°. Considering the striking agreement

17036 J. Phys. Chem. B, Vol. 109, No. 36, 2005
Donker et al.
film to chloroform in liquid or gaseous form is that it disrupts
the interaction between the porphyrin octylchains and thus the
molecular ordering of the stacks. The equivalence of the
porphyrins within a stack is thereby destroyed, and as a result
the excitonic interaction is reduced or even disappears, depend-
ing on the remaining degree of order. Similar to what happens
in ZnTPP films, the ZnTOPP molecules are now exposed to
traces of water, which are always present in the solvent. Water
may act as a weak aquo ligand to the porphyrin zinc center,
resulting in the formation of excitation traps within each stack.
We cannot completely exclude that chloroform itself could be
the ligand instead of water, but this is less likely, since
chloroform is not known to have ligating properties. Whatever
the ligand is, upon annealing of the film, chloroform evaporates
from the film, and its order as well as the excitonic interactions
between the ZnTOPP molecules are restored; i.e., the 390-nm
band shifts back to 375 nm, and the 450-nm band shifts slightly
to the red as shown in Figure 3. To investigate this phenomenon
further, we also prepared films by spin-coating from the strongly
ligating pyridine as a solvent. Figure 4 shows the room-
temperature absorption spectra of ZnTOPP films spin-coated
from pyridine (film 4). Immediately after spin-coating (Figure
Figure 9. Emission and excitation spectra: solid line, excitation
spectrum at normal incidence, λem ) 640 nm; dashed line, excitation
spectrum at 45° angle of incidence, λem ) 640 nm; dotted line, emission
spectrum, λexc ) 450 nm.
with the experimental energy-transfer kinetics.44 The results of
these simulations have shown that the above-mentioned struc-
tural model of the nanolayers is essentially correct.
Annealing Effects. Figure 3 illustrates the changes in the
room-temperature absorption spectra of ZnTOPP films following
annealing to 175 °C during ∼5 min, resulting in the formation
of ordered domains.1
1-13
For ZnTOPP films 1, annealing leaves
the shape and width of the B-band unaffected. By contrast, the
90-nm band for films 2 disappears upon annealing, and the
full width at half-maximum (fwhm) line width of the B-band
3
8), the (0,0) and (0,1) Q-bands are red-shifted compared to those
of the films spin-coated from chloroform (2) or toluene (1), with
a concomitant increase of the intensity ratio IQ(0,0)/IQ(0,1). These
-
1
at 450 nm decreases from 27 ( 1 nm (1400 cm ) to 24 ( 1
-
1
nm (1200 cm ). After annealing, the spectrum of film 2 is
identical to that of 1. By contrast, Figure 3C shows that
annealing of ZnTPP films spin-coated from toluene or chloro-
form results in considerable broadening of the B-band, i.e., from
48-50
features identify the species as pyridine-ligated ZnTOPP
The B-band is split into several components, with the main
component at 437 nm broadened with respect to the same band
in films 1. Note that in this spectrum the 375-nm absorption
band is absent. Slow evaporation of pyridine from the film
results in a gradual blue shift of the Q-bands and a return of
the IQ(0,0)/IQ(0,1) intensity ratio to that of film 1. Furthermore,
we note that the B-bands initially have shoulders at 390 and
-
1
-1
3
5 ( 3 nm (1800 cm ) to 59 ( 3 nm (3000 cm ). Also,
whereas the 450 nm B-band of the ZnTOPP film is red-shifted
by ∼25 nm from that in solution spectra, the B-band of ZnTPP
is on average only ∼10 nm red-shifted. For the ZnTOPP films,
the Q-bands around 550 nm are unaffected by annealing, but
the Q-bands of ZnTPP films 3 show a small blue shift and a
small change in the intensity ratio IQ(0,0)/IQ(1,0) (Figure 3C). These
changes are similar to those observed upon drying a solution
of ZnTPP in toluene on sodium wire, and we therefore ascribe
these changes to the removal of an aquo ligand. We suggest
that the broadening upon annealing is due to the decrease in
the intermolecular distance between the porphyrin molecules
in the film by removal of the aquo ligands, thereby increasing
the intermolecular excitonic interaction. However, from the
intensity ratio and positions of the Q-bands of ZnTOPP films
spin-coated from nondried toluene, we conclude that ZnTOPP
in these films is not bound to an aquo ligand. This is also
supported by the finding that sodium-drying of toluene before
use and storing the film for several weeks in the dark and
ambient air leaves the spectra unchanged. These results strongly
indicate that the octyl chains in ZnTOPP screen the porphyrin
core of the aggregates from atmospheric moisture.
400 nm. The 390-nm band gradually disappears during evapora-
tion of the solvent in favor of the 375-nm polarized absorption
band, and the entire B-band narrows and shifts to 450 nm. Only
after annealing at 150 °C, the film becomes completely ordered
(Figure 4). Exposing film 1 to pyridine vapor results in
essentially the same spectral changes. The origin of the 390-
nm band is somewhat unclear. Note that it appears in both
spectra of films spin-coated from chloroform as well as from
pyridine, even though these solvents have quite different ligand
strengths. This could be an indication that it arises from a
modified excitonic interaction between less ordered porphyrins,
as a result of their interaction with the solvent.
Since the above-mentioned experiments lead to complete
conversion of the nonligated to the ligated ZnTOPP species,
we may again conclude that the 375-nm absorption band is a
bulk property of the film.
There is an important difference between the spectral effects
of exposure of the film to chloroform and pyridine vapor.
Treatment of ZnTOPP films with chloroform vapor does not
result in a net shift of the B-band as occurs with pyridine.
Obviously, the effect of chloroform represents a more subtle
change of the porphyrin environment in the ordered film, in
particular of the excitonic interaction with neighboring porphy-
rins, than with pyridine. These changes only occur at low
temperature since room-temperature absorption spectra of
ZnTOPP in neat chloroform and toluene are identical. Evidently,
chloroform interacts only weakly with ZnTOPP molecules in
the film, reducing the excitonic interaction, but without com-
pletely destroying its order, as occurs with pyridine as a ligand.
The molecular ordering inside the films is reflected by their
absorption and fluorescence spectra (Figures 3 and 9). Whereas
in films 1 the 450-nm B-band has about the same width before
and after annealing (Figure 3A), the narrowing of this band in
films 2 (Figure 3B) upon annealing indicates a transition to a
more ordered film structure4
5-47
The 390-nm band in the absorption spectra of films 2 spin-
coated from chloroform is absent in spectra of ZnTOPP in the
same solvent, implying that this band is a property of the film.
From the low-temperature fluorescence spectra, we may con-
clude that after exposure to chloroform a large fraction of the
ZnTOPP molecules has been changed to the ligated form, acting
as fluorescent traps.48 (For details, see paper 2.)
The changes in the film structure as revealed by the absorption
spectra can be summarized as follows:
The most likely explanation for the effect of exposure of the

Optical Spectroscopy of Zinc Porphyrin Nanolayers
J. Phys. Chem. B, Vol. 109, No. 36, 2005 17037
•
•
Films 1 show spontaneous ordering.
Chloroform has a disordering effect on the films, as
(21) Yokoyama, T.; Yokoyama, S.; Kamikado, T.; Mashiko, S. J. Chem.
Phys. 2001, 115, 3814.
(22) Wienke, J.; Kleima, F. J.; Koehorst, R. B. M.; Schaafsma, T. J.
demonstrated for films 2 and 3, most likely by exposing the
porphyrins to traces of water, acting as aquo ligands to the
porphyrin zinc centers.
Thin Solid Films 1996, 279, 87.
(23) Gouterman, M. In The Porphyrins; Dolphin, D., Ed.; Academic
Press: New York, 1978; Vol. III, Chapter 1.
(
24) Adler, A. D.; Longo, F. R.; Kampas, F.; Kim, J. J. Inorg. Nucl.
Chem. 1970, 32, 2443.
25) Adler, A. D.; Longo, F. R.; Shergalis, W. J. Am. Chem. Soc. 1964,
86, 3145.
•
Interaction with strong ligands, e.g., pyridine, as in film 4,
disrupts the molecular ordering of the films, which is restored
upon evaporation of the solvent.
(
(26) Little, R. G.; Anton, J. A.; Loach, P. A.; Ibers, J. A. J. Heterocycl.
Chem. 1975, 12, 343.
Acknowledgment. T.J.S. thanks Dr. Herbert van Amerongen
for useful comments. This work was supported by The
Netherlands Organization for Energy and Environment under
Contract No. 146.100-024.4.
(27) Kroon, J. M.; Sudh o¨ lter, E. J. R.; Wienke, J.; Koehorst, R. B. M.;
Savenije, T. J.; Schaafsma, T. J. In Proceedings of the 13th European
PhotoVoltaic Solar Energy Conference, Nice, France, Oct 1995; H. S.
Stephens & Associates: Bedford, U.K., 1995; pp 1295-1298.
(
28) Schick, G. A.; Schreiman, I. C.; Wagner, R. W.; Lindsey, J. S.;
Bocian, D. F. J. Am. Chem. Soc. 1989, 111, 1344.
29) Azumi, R.; Matsumoto, M.; Kawabata, Y.; Kuroda, S.; Sugi, M.;
King, L. G.; Crossley, M. J. J. Phys. Chem. 1993, 97, 12862.
30) Tran-Thi, T. H.; Lipskier, J. F.; Maillard, P.; Momenteau, M.;
Lopez-Castillo, J. L.; Jay-Gerin, J.-P. J. Phys. Chem. 1992, 96, 1073.
31) Ohno, O.; Kaizu, Y.; Kobayashi, H. J. Chem. Phys. 1993, 99,
128.
References and Notes
(
(
1) Springer Handbook of Nanotechnology; Bushan, B., Ed.; Springer-
Verlag: New York, 2004.
2) Ohring, M. The Materials Science of Thin Films; Elsevier Science
and Technology Books: Amsterdam, The Netherlands, 2001.
3) Tredgold, R. H. Order in Organic Films; Cambridge University
Press: New York, 1994.
4) Frank, C. W. Organic Thin Films: Structure and Applications; ACS
Symposium Series 695; American Chemical Society: Washington, DC,
(
(
(
4
8
1
(
(32) Hamor, M. J.; Hamor, T. A.; Hoard, J. L. J. Am. Chem. Soc. 1964,
6, 1938.
(
(33) Fleischer, E. B.; Miller, C. K.; Webb, L. E. J. Am. Chem. Soc.
964, 86, 2342.
1
998.
(
(
34) Silvers, S. J.; Tulinsky, A. J. Am. Chem. Soc. 1967, 89, 3331.
35) Hofstra, U.; Koehorst, R. B. M.; Schaafsma, T. J. Magn. Reson.
(
(
(
5) W o¨ hrle, D.; Meissner, D. AdV. Mater. 1991, 3, 129.
6) Liu, C.-Y.; Bard, A. J. Acc. Chem. Res. 1999, 32, 235.
Chem. 1987, 25, 1069.
7) Peumans, P.; Yakimov, A.; Forrest, S. R. J. Appl. Phys. 2003, 93,
(
36) McRae, E. G.; Kasha, M. J. Chem. Phys. 1958, 28, 721.
3
693.
8) Drain, C. M.; Hupp, J. T.; Suslick, K. S.; Wasielewski, M. R. J.
Porphyrins Phthalocyanines 2002, 6, 243.
9) Friedrich, M.; Gavrila, G.; Himcinschi, C.; Kampen, T. U.; Kobitski,
(37) Kasha, M.; Rawls, H. R.; El-Bayoumi, M. A. Pure Appl. Chem.
(
1
965, 11, 371.
38) Stomphorst, R. G.; Schaafsma, T. J.; van der Zwan, G. J. Phys.
Chem. A 2001, 105, 4226.
39) Zimmermann, J.; Siggel, U.; Fuhrhop, J.-H.; R o¨ der, B. J. Phys.
Chem. B 2003, 107, 6019.
40) Cho, H. S.; Song, N. W.; Kim, Y. H.; Jeoung, S. C.; Hahn, S.;
Kim, D. J. Phys. Chem. A 2000, 104, 3287.
(
(
A. Yu.; M e´ ndez, H.; Salvan, G.; Cerrill o´ , I.; M e´ ndez, J.; Nicoara, N.; Bar o´ ,
A. M.; Zahn, D. R. T. J. Phys.: Condens. Matter 2003, 15, 2699.
(
(
10) Guo, Q.; Yin, J.; Yin, F.; Palmer, R. E.; Bampos, N.; Sanders, J.
K. M. J. Phys.: Condens. Matter 2003, 15, 3127.
11) Kroon, J. M.; Sudh o¨ lter, E. J. R.; Schenning, A. P. H. J.; Nolte, R.
J. M. Langmuir 1995, 11, 214.
12) Kroon, J. M.; Schenkels, P. S.; van Dijk, M.; Sudh o¨ lter, E. J. R. J.
Mater. Chem. 1995, 5, 1309.
13) Kroon, J. M.; Koehorst, R. B. M.; van Dijk, M.; Sanders, G. M.;
Sudh o¨ lter, E. J. R. J. Mater. Chem. 1997, 7, 615.
14) (a) Koti, A. S. R.; Periasamy, N. J. Mater. Chem. 2002, 12, 2313.
b) Koti, A. S. R.; Periasamy, N. Chem. Mater. 2003, 15, 369.
15) Shimizu, Y.; Miya, M.; Nagata, A.; Ohta, K.; Yamamoto, I.;
Kusabayashi, S. Liq. Cryst. 1993, 14, 795.
16) Shimizu, Y.; Higashiyama, T.; Fuchita, T. Thin Solid Films 1998,
31, 279.
17) Chandrasekhar, S. In Handbook of Liquid Crystals, Low Molecular
(
(
(
41) F o¨ rster, Th. Ann. Phys. 1948, 2, 55.
(42) Czikklely, V.; Forsterling, H. D.; Kuhn, H. Chem. Phys. Lett. 1970,
(
6
, 207.
(43) Von Maltzan, B. Z. Naturforsch., A: Phys. Sci. 1985, 40, 389.
(
(44) (a) Yatskou, M. M., Ph.D. Thesis, Wageningen University, 2001.
(
(b) Yatskou, M. M.; Donker, H.; Novikov, E. G.; Koehorst, R. B. M.; van
Hoek, A.; Apanasovich, V. V.; Schaafsma, T. J. J. Phys. Chem. A 2001,
105, 9498. (c) Yatskou, M. M.; Donker, H.; Koehorst, R. B. M.; van Hoek,
A.; Schaafsma, T. J. Chem. Phys. Lett. 2001, 345, 141. (d) Apanasovich,
V. V.; Novikov, E. G.; Yatskou, M. M.; Koehorst, R. B. M.; Schaafsma,
T. J.; van Hoek, A. J. Appl. Spectrosc. 1999, 66, 613.
(45) Gregg, B. A.; Fox, M. A.; Bard, A. J. J. Phys. Chem. 1989, 93,
4227.
(46) Markovitsi, D.; L e´ cuyer, I.; Simon, J. J. Phys. Chem. 1991, 95,
3620.
(47) Sigal, H.; Markovitsi, D.; Gallos, L. K.; Argyrakis, P. J. Phys. Chem.
1996, 100, 10999.
(48) Humphrey-Baker, R.; Kalyanasundaram, K. J. Photochem. 1985,
(
(
(
3
(
Weight Liquid Crystals II; Demus, D., Goodby, J., Gray, G. W., Spiess,
H.-W., Vill, V., Eds.; Wiley-VCH: Weinheim, Germany, 1998; Vol. 2B,
Chapter 8.
(
18) The Porphyrin Handbook; Kadish, K. M., Smith, K. M., Guilard,
R., Eds.; Academic Press: San Diego, Tokyo, 1999; Vols. 6 and 8.
19) Ogi, T.; Ohkita, H.; Shinzaburo, I.; Masahide, Y. Thin Solid Films
002, 415, 228.
20) Hayashi, K.; Kawato, S.; Fujii, Y.; Horiuchi, T.; Matsushige, K.
Appl. Phys. Lett. 1997, 70, 1384.
(
2
31, 105.
(
(49) Nardo, J. V.; Dawson, J. H. Inorg. Chim. Acta 1986, 123, 9.
(50) Miller, J. R.; Dorough, G. D. J. Am. Chem. Soc. 1952, 74, 3977.