Tetrachelate porphyrin chromophores for metal oxide semiconductor sensitization: Effect of the spacer length and anchoring group position

Article abstract of DOI:10.1021/ja068218u

Four Zn(II)-tetra(carboxyphenyl)porphyrins in solution and bound to metal oxide (TiO₂, ZnO, and ZrO₂) nanoparticle films were studied to determine the effect of the spacer length and anchoring group position (para or meta) on their b

Full text of DOI:10.1021/ja068218u

Published on Web 03/27/2007
Tetrachelate Porphyrin Chromophores for Metal Oxide
Semiconductor Sensitization: Effect of the Spacer Length
and Anchoring Group Position
Jonathan Rochford,† Dorothy Chu,† Anders Hagfeldt,‡ and Elena Galoppini*,†
Contribution from the Chemistry Department, Rutgers UniVersity, 73 Warren Street, Newark,
New Jersey 07102, and Center of Molecular DeVices, Department of Chemistry, Royal Institute
of Technology, Teknikringen 30, SE 100 44 Stockholm, Sweden
Received November 16, 2006; E-mail: galoppin@rutgers.edu
Abstract: Four Zn(II)-tetra(carboxyphenyl)porphyrins in solution and bound to metal oxide (TiO2, ZnO, and
ZrO2) nanoparticle films were studied to determine the effect of the spacer length and anchoring group
position (para or meta) on their binding geometry and photoelectrochemical and photophysical properties.
The properties of three types of anchoring groups (COOH and COONHEt3) for four Zn(II)-porphyrins (Zn-
(II)-5,10,15,20-tetra(4-carboxyphenyl)porphyrin (p-ZnTCPP), Zn(II)-5,10,15,20-tetra(3-carboxyphenyl)por-
phyrin (m-ZnTCPP), Zn(II)-5,10,15,20-tetra(3-(4-carboxyphenyl)phenyl)porphyrin (m-ZnTCP2P), and Zn(II)-
5,10,15,20-tetra(3-ethynyl(4-carboxyphenyl)phenyl)porphyrin (m-ZnTC(PEP)P)) were compared. In m-ZnTCPP,
m-ZnTCP2P, and m-ZnTC(PEP)P the four anchoring groups are in the meta position on the meso-phenyl
rings of the porphyrin macrocycle, thus favoring a planar binding mode to the metal oxide surfaces. The
three meta-substituted porphyrin salts have rigid spacer units of increasing length (phenyl (P), biphenyl
(P2), and diphenylethynyl (PEP)) between the porphyrin ring and the carboxy anchoring groups, thus raising
the macrocycle from the metal oxide surface. All porphyrins studied here, when bound to TiO2 and ZnO,
exhibited quenching of the fluorescence emission, consistent with electron injection into the conduction
band of the semiconductor. Steady-state UV-vis and fluorescence studies of p-ZnTCPP on insulating
ZrO2 showed evidence of aggregation and exciton coupling. This was not observed in any of the meta-
substituted porphyrins. The photoelectrochemical properties (IPCE, Voc, and Isc) of the porphyrins bound
to TiO2 films in solar cells have been measured and rationalized with respect to the sensitizer binding
geometry and distance from the surface.
the chromophore is attached to the surface.3 For instance, we,
and others, have reported a series of “rigid-rod”4,5 and “tripodal”
linkers6,7 to achieve some level of control over the orientation
and distance of the sensitizer at the metal oxide surface. We
have now designed and synthesized a series of tetra(carbox-
yphenyl)porphyrin derivatives (Figure 1) as models to determine
how the distance and attachment of the sensitizing chromophore
on colloidal TiO2 and ZnO films influence interfacial processes
and solar cell efficiencies. The interest in developing model
1. Introduction
The investigation of semiconductor sensitization by organic
and inorganic dyes for applications in dye-sensitized solar cells
(DSSCs) is a very active area of research due to the need for
efficient, environmentally friendly, and economically viable
renewable energy sources.1 One of the key requirements in the
operation of this kind of solar cell is the sensitization of a wide
band gap semiconductor, typically TiO2 or ZnO, via electron
injection from a photoexcited chromophore.2 The efficiency of
the electron-transfer step at the dye-semiconductor interface
is highly dependent, among numerous other factors, on the way
(4) (a) Hoertz, P. G.; Carlisle, R. A.; Meyer, G. J.; Wang, D.; Piotrowiak, P.;
Galoppini, E. Nano Lett. 2003, 3, 325. (b) Wang, D.; Mendelsohn, R.;
Galoppini, E.; Hoertz, P. G.; Carlisle, R. A.; Meyer, G. J. J. Phys. Chem.
B 2004, 108, 16642. (c) Taratula, O.; Rochford, J.; Piotrowiak, P.;
Galoppini, E.; Carlisle, R. A.; Meyer, G. J. J. Phys. Chem. B 2006, 110,
15734.
(5) Kilså, K.; Mayo, E. I.; Kuciauskas, D.; Villahermosa, R.; Lewis, N. S.;
Winkler, J. R.; Gray, H. B. J. Phys. Chem. A 2003, 107, 3379.
(6) (a) Galoppini, E.; Guo, W.; Qu, P.; Meyer, G. J. J. Am. Chem. Soc. 2001,
123, 4342. (b) Galoppini, E.; Guo, W.; Zhang, W.; Hoertz, P.; Qu, P.;
Meyer, G. J. J. Am. Chem. Soc. 2002, 124, 7801. (c) Piotrowiak, P.;
Galoppini, E.; Wei, Q.; Meyer, G. J.; Wiewio´r, P. J. Am. Chem. Soc. 2003,
125, 5278. (d) Clark, C. C.; Meyer, G. J.; Wei, Q.; Galoppini, E. J. Phys.
Chem. B 2006, 110, 11044.
(7) (a) Loewe, R. S.; Ambroise, A.; Muthukumaran, K.; Padmaja, K.; Lysenko,
A. B.; Mathur, G.; Li, Q.; Bocian, D. F.; Misra, V.; Lindsey, J. S. J. Org.
Chem. 2004, 69, 1453. (b) Wei, L.; Padmaja, K.; Youngblood, W. J.;
Lysenko, A. B.; Lindsey, J. S.; Bocian, D. F. J. Org. Chem. 2004, 69,
1461. (c) Long, B.; Nikitin, K.; Fitzmaurice, D. J. Am. Chem. Soc. 2003,
125, 5152.
† Rutgers University.
‡ Royal Institute of Technology.
(1) (a) O’Regan, B.; Gra¨tzel, M. Nature 1991, 353, 737. (b) Gra¨tzel, M. Nature
2001, 414, 338. (c) Gra¨tzel, M. Inorg. Chem. 2005, 44, 6841. (d)
Kalyanasundaram, K.; Gra¨tzel, M. Coord. Chem. ReV. 1998, 177, 347. (e)
Kamat, P. V. Chem. ReV. 1993, 93, 267-300. (f) Qu, P.; Meyer, G. J. In
Electron Transfer in Chemistry; Balzani, V., Ed.; John Wiley & Sons: New
York, 2001; Vol. IV, Part 2, Chapter 2, pp 355-411.
(2) (a) Anderson, N. A.; Lian, T. Coord. Chem. ReV. 2004, 248, 1231. (b)
Schwarzburg, K.; Ernstofer, R.; Felber, S.; Willig, F. Coord. Chem. ReV.
2004, 248, 1259. (c) Benko¨, G.; Kallioinen, J.; Korppi-Tommola, J. E. I.;
Yartsev, A.; Sundstro¨m, V. J. Am. Chem. Soc. 2002, 124, 489.
(3) (a) Clifford, J. N.; Palomares, E.; Nazeeruddin, Md. K.; Gra¨tzel, M.; Nelson,
J.; Li, X.; Long, N. J.; Durrant, J. R. J. Am. Chem. Soc. 2004, 126, 5225.
(b) Palomares, E.; Mart´ınez-Diaz, M. V.; Haque, S. A.; Torres, T.; Durrant,
J. R. Chem. Commun. 2004, 18, 2112. (c) Durrant, J. R.; Haque, S. A.;
Palomares, E. Coord. Chem. ReV. 2004, 248, 1247.
9
10.1021/ja068218u CCC: $37.00 © 2007 American Chemical Society
J. AM. CHEM. SOC. 2007, 129, 4655-4665
4655

A R T I C L E S
Rochford et al.
Figure 1. Structures of the porphyrins employed in this study. Also shown are the anticipated binding geometries of the COOH and COOEt3NH derivatives
on metal oxide surfaces. This diagram shows ideal behavior of the sensitizers on the metal oxide surface. However, we must assume that other binding
modes are also likely. Due to the steric constraints of these systems it should take at least two of the four carboxy groups bound to the surface to fix the
porphyrin chromophores in the rigid planar fashion depicted.
porphyrins was initiated by our electron transport study in a
solar cell with the working electrode made of a ZnO nanowire
film and sensitized with Zn(II)-5,10,15,20-tetra(3-carboxyphe-
nyl)porphyrin (m-ZnTCPP).8 Eventually, the compounds de-
scribed here could be useful models to study interfacial processes
in ZnO nanowire films.
In this paper we will discuss the synthesis, photophysical,
electrochemical, and photoelectrochemical properties of the
carboxylic acid derivatives (COOH, COOMe, and COOEt3NH)
of four tetrachelate Zn-porphyrins (Figure 1). Two (m-ZnTCP2P
and m-ZnTC(PEP)P) are new compounds. For clarity, for all
four porphyrins, the carboxylic acid derivatives are indicated
with the suffix -[A] in the abbreviated name, the corresponding
methyl esters with the suffix -[E], and the ammonium salts with
the suffix -[S]. The carboxylic acid functionality is the most
commonly used anchoring group for standard Gra¨tzel-type
DSSCs, and numerous DSSC devices employing carboxypor-
phyrins have been reported over the years.1d,3a,c,9
The presence of carboxy anchoring groups attached to the
meso-phenyl ring of the porphyrin has little influence on the
ground-state electronic nature of the dye, because of the
orthogonal orientation of the meso-phenyl ring.10 However, the
position on the carboxylic acid anchoring groups on the meso-
phenyl ring does have a great influence on the binding mode
and, as a result, on other properties. For instance, Campbell et
al. have reported a 5-fold increase in the short circuit photo-
current (Isc) and a significantly higher open circuit photovoltage
(Voc) on changing the position of the COOH groups from the
para (in p-ZnTCPP-[A]) to the meta position (in m-ZnTCPP-
[A]).9a This effect was attributed to a more efficient electron
injection from the meta porphyrin, which binds flat on the TiO2
surface, compared to that of the para, which binds vertically
(Figure 1).9a,11 Our recent study of cells prepared from m-
ZnTCPP-[A] bound to ZnO nanowires and ZnO colloidal
(8) Galoppini, E.; Rochford, J.; Chen, H.; Saraf, G.; Lu, Y.; Hagfeldt, A.;
Boschloo, G. J. Phys. Chem. B 2006, 110, 16159.
(9) (a) Campbell, W. M.; Burrell, A. K.; Officer, D. L.; Jolley, K. W. Coord.
Chem. ReV. 2004, 248, 1363 and references therein. (b) Argazzi, R.;
Murakami Iha, N. Y.; Zabri, H.; Odobel, F.; Bignozzi, C. A. Coord. Chem.
ReV. 2004, 248, 1299. (c) Nazeeruddin, M. K.; Humphry-Baker, R.; Officer,
D. L.; Campbell, W. M.; Burrell, A. K.; Gra¨tzel, M. Langmuir 2004, 20,
6514. (d) Watson, D. F.; Marton, A.; Stux, A. M.; Meyer, G. J. J. Phys.
Chem. B 2004, 108, 11680. (e) Brune, A.; Jeong, G.; Liddell, P. A.;
Sotomura, T.; Moore, T. A.; Moore, A. L.; Gust, D. Langmuir 2004, 20,
8366. (f) Jasieniak, J.; Johnston, M.; Waclawik, E. R. J. Phys. Chem. B
2004, 108, 12962. (g) Michaelis, E.; Nonomura, K.; Schlettwein, D.;
Yoshida, T.; Minoura, H.; Wo¨hrle, D. J. Porphyrins Phthalocyanines 2004,
12, 1366. (h) Schmidt-Mende, L.; Campbell, W. M.; Wang, Q.; Jolley, K.
W.; Officer, D. L.; Nazeeruddin, M. K.; Gra¨tzel, M. Chem. Phys. Chem.
2005, 6, 1253. (i) Hasobe, T.; Hattori, S.; Kamat, P. V.; Urano, Y.;
Umezawa, N.; Nagano, T.; Fukuzumi, S. Chem. Phys. 2005, 319, 243. (j)
Wang, Q.; Campbell, W. M.; Bonfantani, E. E.; Jolley, K. W.; Officer, D.
L.; Walsh, P. J.; Gordon, K.; Humphry-Baker, R.; Nazeeruddin, M. K.;
Gra¨tzel, M. J. Phys. Chem. B 2005, 109, 15397. (k) Hasobe, T.; Hattori,
S.; Kamat, P. V.; Fukuzumi, S. Tetrahedron 2006, 62, 1937. (l) Lo, C.-F.;
Luo, L.; Wei-Guang Diau, E.; Chang, I.-J.; Lin, C.-Y. Chem. Commun.
2006, 1430. (m) Luo, L.; Lo, C.-F.; Lin, C.-Y.; Chang, I.-J.; Wei-Guang
Diau, E. J. Phys. Chem. B 2006, 110, 410.
(10) (a) Schmidt, E. S.; Calderwood, T. S.; Bruice, T. C. Inorg. Chem. 1986,
25, 3718. (b) Seth, J.; Palaniappan, V.; Johnson, T. E.; Prathapan, S.;
Lindsey, J. S.; Bocian, D. F. J. Am. Chem. Soc. 1994, 116, 10578. (c)
Bru¨ckner, C.; Foss, P. C. D.; O’ Sullivan, J.; Pelto, R.; Zeller, M.; Birge,
R. R.; Crundwell, G. Phys. Chem. Chem. Phys. 2006, 8, 2402.
9
4656 J. AM. CHEM. SOC. VOL. 129, NO. 15, 2007

Porphyrins for Semiconductor Sensitization
A R T I C L E S
nanoparticle films motivated us to investigate such orientation
effects further. In particular, we were interested to study the
influence of both distance and binding geometry on the
photophysical properties and solar cell efficiencies of the
sensitizers.
2. Experimental Section
2.1. Metal Oxide Film Preparation. Colloidal TiO2 and ZrO2 films
were prepared by a sol-gel technique that produces mesoporous films
of approximately 10 µm thickness and that consist of nanoparticles
with an average diameter of ∼20 nm (Supporting Information). ZnO
films were prepared by a previously published method that produces
mesoporous films of approximately 2 µm thickness with an average
particle size of 15 nm.12 Repeated depositions for making thicker ZnO
films resulted in cracking and peeling of the film from the glass
substrate. The TiO2 films were prepared by casting the colloidal
solutions by the doctor blade technique onto the substrate over an area
of 1 × 2 cm2, followed by sintering at ∼450 °C for 30 min. Lower
temperatures were used for sintering the ZnO films (∼300-350 °C).
For absorption and fluorescence studies the films were cast on cover
glass slides (VWR). For electrochemical studies the films were cast
on transparent, electronically conductive glass with an indium-tin-
oxide layer (8 Ω/sq ITO, Pilkington). We will refer to films cast on
cover glass or ITO conductive glass as MO/G or MO/ITO, respectively,
where MO is the metal oxide. The preparation of the films for the solar
cells employed in the photoelectrochemical studies followed slightly
different procedures (see the Photoelectrochemistry section).
Figure 2. FT-IR-ATR spectra of m-ZnTC(PEP)P solid and adsorbed on
TiO2 as the acid and the salt derivative. The broad bands observed at ∼3000
cm–1 for the TiO2 samples are assigned to adsorbed moisture on the films.
2.2. Binding. Binding of the dye molecules to the TiO2, ZnO, and
ZrO2 films was accomplished by immersing the films in a 0.4 mM
MeOH solution of the dye at room temperature for 1 h. Short binding
times were chosen to prevent aggregation of the porphyrins at the metal
oxide surfaces. Basic or acidic pretreatment of the metal oxide films
was tested, but binding was found to occur best on untreated films.
All films, which were stored in the air, were dried by heating to 150
°C for 30 min and cooled to 80 °C prior to immersion in the dye
solution. The sensitized films were rinsed, immersed in MeOH to
remove physisorbed dye, and then used for spectroscopic or electro-
chemical measurements. The amount of bound dye was estimated by
calculating the amount of desorbed dye per gram of TiO2 following a
reported method.13 This involved treating a combination of five films
with 5 mL of 0.1 N NaOH(aq) and determining the average concentra-
tion per film of desorbed dye from the UV-vis absorption spectrum,
using the extinction coefficient measured in basic solution. The five
TiO2 films (stripped of the dye) were then immersed in deionized water
for 1 h, dried at 150 °C for 1 h, and the weight of TiO2 on each glass
slide was determined by scraping the films from the surface.
2.3. Spectroscopy. The FT-ATR-IR spectra of all porphyrins were
measured neat (as powders) and bound to TiO2 (Figures 2 and 3 and
Supporting Information). Fourier transform infrared attenuated total
reflectance (FT-IR-ATR) spectra were acquired on a Thermo Electron
Corporation Nicolet 6700 FT-IR. UV-vis absorption spectra were
acquired at ambient temperature in MeOH (Acros, spectroscopic grade)
using a Hewlett-Packard 8453 diode array spectrometer. The sensitized
MO/G films were placed diagonally in a 1 cm square quartz cuvette in
air while recording spectra. Steady-state fluorescence spectra were
acquired on a Spex Fluorolog that had been calibrated with a standard
NIST tungsten-halogen lamp. Sensitized MO/G films were placed
diagonally in a 1 cm square quartz cuvette. The excitation beam was
Figure 3. Main binding modes of the carboxylate group to TiO2.
directed 45° to the film surface, and the emitted light was monitored
from the front face of the sample and from a 90° angle in the case of
fluid solutions. Fluorescence quantum yields for the samples (ΦFl) were
calculated in MeOH solutions deareated by freeze-pump-thaw
technique at λexc ) 565 nm, using the optically dilute technique with
Zn(II)-5,10,15,20-tetraphenylporphyrin (ZnTPP) as the actinometer (Φref
) 0.033),14 according to eq 1.
ΦFl ) (Aref/As)(Is/Iref)(ηs/ηref)2Φref
(1)
The subscript “s” refers to the sample, the subscript “ref” to the
reference sample (in this case ZnTPP), A is the absorbance at the
excitation wavelength, I is the integrated emission area, and η is the
solvent refraction index.
2.4. Electrochemistry. Cyclic voltammetry was performed with a
BAS potentiostat under nitrogen at room temperature with a conven-
tional three-electrode configuration at a scan rate of 100 mV s–1. The
electrochemical properties of the methyl ester derivatives of the dyes
were studied in dichloromethane with 0.1 M Bu4NBF4 supporting
electrolyte. A glassy carbon working electrode (2 mm diameter) was
used along with a platinum wire auxiliary electrode and a Ag/AgCl
reference electrode. When using a porphyrin/MO/ITO film as the
working electrode, acetonitrile was used with a 0.1 M Bu4NClO4
supporting electrolyte. In this case the porphyrins were bound as the
tetra(triethylammonium)carboxyporphyrin salts. The potential of the
Ag/AgCl electrode was calibrated externally versus the ferrocene/
ferrocenium (Fc+/Fc) redox couple. The half-wave redox potentials
(E1/2) were determined as (Epa + Epc)/2, where Epa and Epc are the anodic
and cathodic peak potentials, respectively. It was assumed that E1/2
=
E0′ for all measurements. All potentials were referenced to the normal
hydrogen electrode (NHE) to make a comparison with the conduction
bands of TiO2 and ZnO. Where E1/2 could not be calculated, due to
irreversible oxidation or reduction processes, Epa and Epc were reported,
respectively. All redox processes were also studied by differential pulse
(11) A similar comparison between meta- and para-substituted tetraphosphonic
acid porphyrins bound to TiO2 showed little change between the two
structural isomers, see: Odobel, F.; Blart, E.; Lagre´e, M.; Villieras, M.;
Boujtita, H.; El Murr, N.; Caramori, S.; Bignozzi, C. A. J. Mater. Chem.
2003, 13, 502.
(12) Taratula, O.; Galoppini, E.; Wang, D.; Chu, D.; Zhang, Z.; Chen, H.; Saraf,
G.; Lu, Y. J. Phys. Chem. B 2006, 110, 6506.
(13) Dabestani, R.; Bard, A. J.; Campion, A.; Fox, M. A.; Mallouk, T. E.;
Webber, S. E.; White, J. M. J. Phys. Chem. 1988, 92, 1872.
(14) Strachan, J. P.; Gentemann, S.; Seth, J.; Kalsbeck, W. A.; Lindsey, J. S.;
Holten, D.; Bocian, D. F. J. Am. Chem. Soc. 1997, 119, 11191.
9
J. AM. CHEM. SOC. VOL. 129, NO. 15, 2007 4657

A R T I C L E S
Rochford et al.
voltammetry (DPV), using relationship 2
E1/2 ) Emax + ∆E/2
anoquinone (DDQ), tetrakis(triphenylphosphine)palladium(0), dichlo-
robis(triphenylphosphine)palladium(II), cuprous(I) iodide, zinc(II) ac-
etate, sodium hydroxide, sodium bicarbonate, potassium carbonate, and
tetrabutylammonium fluoride were all used as received. Ethyl acetate
was purchased as HPLC grade and used as received. Pyrrole was freshly
distilled under vacuum and over potassium hydroxide. Anhydrous
tetrahydrofuran (THF) was distilled under nitrogen atmosphere from
sodium/benzophenone immediately prior to use. Dichloromethane and
hexane were distilled over CaCl2 prior to use. Triethylamine and
diisopropylamine were both distilled over CaH2 prior to use. Column
chromatography was performed using 230-600 mesh silica gel (Sorbent
Technologies). Thin-layer chromatography (TLC) was carried out using
silica gel plates with a fluorescent indicator (Sorbent Technologies)
and UV as the detection method. High- and low-resolution mass spectra
were collected at a commercial facility. NMR spectra were recorded
(2)
where Emax is the peak maxima in the DPV scan and ∆E is the pulse
amplitude.15 DPV experiments were carried out at a scan rate of 20
mV s–1, pulse amplitude 50 mV, pulse width 50 mV, and pulse period
0.2 s. In some cases, namely, the third reduction process of some of
the porphyrin dyes, the differential pulse voltammetry data are reported,
due to poor quality cyclic voltammetry scans at such negative potentials.
The excited-state oxidation potentials E1/2(P+/P*) of all tetra-
(triethylammonium)carboxyporphyrins in methanol and bound to TiO2
and ZnO surfaces were calculated from relationship 3
E
1/2(P+/P*) ) E1/2(P+/P) - E0-0
(3)
1
on a Varian INOVA-500 spectrometer at 499.896 MHz for H and
125.711 MHz for 13C under ambient probe temperature using CDCl3,
CD3OD, or THF-d8 as solvent. Chemical shifts (δ) are given in parts
million (ppm) and reported to a precision of (0.01 ppm for proton
and carbon. Proton coupling constants (J) are given in Hertz (Hz) and
reported to a precision of (0.1 Hz. The spectra of all compounds were
referenced to the residual solvent peak, i.e., CHCl3 at 7.27 ppm and
CH3OH at 4.87 ppm, for the 1H spectra and were referenced to CDCl3
at 77.00 ppm for all 13C spectra. Satisfactory 13C spectra for compounds
m-H2TCP2P-[E], m-H2TC(PEP)P-[E], m-ZnTCP2P-[A], m-ZnTC(PEP)P-
[A], p-ZnTCPP-[S], m-ZnTCPP-[S], m-ZnTCP2P-[S], and m-ZnTC-
(PEP)P-[S] could not be obtained because of their poor solubility. Mass
spectrometry analysis (FAB) of all tetra(triethylammonium)carboxy-
porphyrin salts only showed molecular ions of the carboxylic acid
derivatives due to formation of the acids in the matrix. UV-vis
absorption spectra were collected on a Varian Cary-500 spectrometer
using spectrophotometric grade solvents. 3-Ethynylbenzaldehyde was
synthesized by following a published procedure.18 Porphyrins p-H2-
where E1/2(P+/P) is equivalent to the first oxidation potential of the
ground-state porphyrin chromophore and E0-0 is the zero-zero
excitation energy.16 E0-0 was calculated from the intersection of the
porphyrin absorption spectrum with the fluorescence emission spectrum
in methanol at normalized absorption/emission intensity. For the solid-
state measurements E0-0 was similarly calculated from the UV-vis
and fluorescence measurements made on the derivatized MO/G and
ZrO2/G films, respectively. E1/2(P+/P*) values were calculated from
the solid-state cyclic voltammetry, UV-vis, and fluorescence data to
give a closer approximation of E1/2(P+/P*) in a solid-state device as
possible.
2.5. Photoelectrochemistry. The TiO2/FTO working electrodes for
photoelectrochemical measurements were prepared by conventional
methods. Briefly, a sample of EPFL B TiO2 (Solaronics) was spread
by the doctor blade technique over a 0.4 × 0.8 cm2 area on a 1.5 ×
1.5 cm2 fluorine-doped tin-oxide (FTO) conductive glass electrode
(Libby Owens Ford, 8 Ω/sq). The film was dried on a hot plate at 150
°C for 6 min, then sintered at 400 °C for 30 min. The films were cooled
to ∼80 °C and then immediately immersed in the dye solutions (0.4
mM in MeOH). The films thus prepared were ∼10 µm thick (average),
typically after two depositions. The Pt/FTO counterelectrode was
prepared by depositing 15 µL of H2PtCl6 (5 mM in 2-propanol, Sigma-
Aldrich) onto the conductive side of a 1.5 × 1.5 cm2 FTO electrode
followed by sintering for 30 min at 380 °C. An open cell set up was
used, where a porous piece of lens paper tissue, wet with electrolyte
(0.05 M I2, 0.10 M LiI, 0.6 M TBAI in 3-methoxypropionitrile) was
sandwiched between the sensitized TiO2/FTO working electrode and
the Pt/FTO counter electrode. All measurements were carried out under
full sun conditions (1000 W m–2) using an apparatus described
previously.17 The incident monochromatic photon-to-current conversion
efficiency (IPCE) is plotted as a function of excitation wavelength
according to equation (4)
20
TCPP-[E],19 m-H2TCPP-[E],20 p-ZnTCPP-[E],19 m-ZnTCPP-[E], p-
ZnTCPP-[A],19 and m-ZnTCPP-[A]9a have been previously reported
using a similar synthetic methodology. However, 1H NMR spectra were
not always published and are therefore included in the Supporting
Information.
2.6.2. Methyl-4(ethynyl-3-formylphenyl)benzoate (4a). A flask was
charged with methyl-4-iodobenzoate 0.79 g (3 mmol) and 3-ethynyl-
benzaldehyde (0.39 g, 3 mmol) followed by dichlorobis(triphenylphos-
phine)palladium(II) (68 mg, 0.06 mmol, 2% equiv) and CuI (11 mg,
0.06 mmol, 2% equiv) under nitrogen. Triethylamine (5 mL) was then
added to the reaction flask via syringe. The reaction mixture was stirred
for 24 h at room temperature and then poured into 10 mL of water.
The product was extracted with ethyl acetate (10 mL), and the organic
layer was washed repeatedly with water to remove any residual amine.
The organic phase was dried over NaSO4, and the solvent was removed
in vacuo. The crude product was purified by silica gel flash chroma-
tography using dichloromethane/ethyl acetate (1:1) to afford of a pale
yellow solid (547 mg, 2.07 mmol, yield 69%): Rf ) 0.58. 1H NMR δH
(CDCl3): 10.04 (s, 1H), 8.06 (m, 3H), 7.88 (d, 1H, J ) 7.0 Hz), 7.79
(d, 1H, J ) 7.0 Hz)), 7.63 (d, 2H, J ) 9.0 Hz), 7.56 (dd, 1H, J ) 8.0
Hz) ppm. 13C NMR δC (CDCl3): 191.36, 166.44, 162.40,137.05,
136.59, 133.03, 131.62, 129.91, 129.60, 129.40, 129.22, 127.48, 124.04,
90.62, 90.06, 52.27 ppm. FT-IR-ATR: V(CdO) 1716, 1695 cm–1; V(C-
O) 1279 cm–1. LRMS (FAB): m/z 265.3 [MH+].
1240Isc
%IPCE(λ) )
× 100
(4)
λφ
where Isc is the photocurrent density (A cm–2), λ is the monitoring
wavelength, and φ is the incident photon flux (W cm–2).
2.6. Synthesis 2.6.1. General. All reactions involving air- and
moisture-sensitive reagents were performed under nitrogen atmosphere
in oven-dried or flame-dried glassware. A magnetic stirrer was used in
all cases. Reagents were purchased from Fisher-Acros or Sigma-Aldrich
Chemical Co. Benzaldehyde, methyl-4-iodobenzoate, methylbenzoate-
4-carboxaldehyde, methylbenzoate-3-carboxaldehyde, methyl-4(3-
formylphenyl)benzoate, trimethylsilylacetylene, 2,3-dichloro-5,6-dicy-
2.6.3. General Procedure for the Preparation of Tetramethyl-
benzoate Porphyrins. Compounds 3b m-H2TCP2P-[E] and 4b m-H2-
TC(PEP)P-[E] were prepared following the two-step one-flask room-
temperature procedure that is used for the synthesis of meso-substituted
(15) Bard, A. J.; Faulkner, L. R. In Electrochemical Methods: Fundamentals
and Applications, 2nd ed.; Wiley: New York, 2001.
(18) Vicente, M. G. H.; Shetty, S. J.; Wickramashinge, A.; Smith, K. M.
Tetrahedron Lett. 2000, 41, 7623.
(19) Koehorst, R. B. M.; Boschloo, G. K.; Savenije, T. J.; Goosens, A.;
Schaafsma, T. J. J. Phys. Chem. B 2000, 104, 2371.
(20) Bonar-Law, R. P.; Sanders, J. K. M. J. Chem. Soc., Perkin Trans. 1 1995,
3085.
(16) Kuciauskus, D.; Monat, J. E.; Villahermosa, R.; Gray, H. B.; Lewis, N. S.;
McCusker, J. K. J. Phys. Chem. B 2002, 106, 9347.
(17) Lindstro¨m, H.; Magnusson, E.; Holmberg, A.; So¨dergren, S.; Lindquist,
S.-E.; Hagfeldt, A. Sol. Energy Mater. Sol. Cells 2002, 73, 91.
9
4658 J. AM. CHEM. SOC. VOL. 129, NO. 15, 2007

Porphyrins for Semiconductor Sensitization
A R T I C L E S
Scheme 1. Synthetic Strategy Followed for the Preparation of the Porphyrin Sensitizersa
a Reagents and conditions: (i) CH2Cl2/TFA/DDQ, N2/room temperature/12 h, 22-30% yield; (ii) Zn(OAc)2/CH2Cl2/MeOH, room temperature/12 h,
86-92% yield; (iii) 2 N aqueous NaOH/THF, room temperature/3 days, 68-82% yield; (iv) Et3N/THF/EtOAc, room temperature/3 h, 68-76% yield.
1H NMR δH(CDCl3): 8.95 (s, 8H), 8.53 (m, 4H), 8.30 (d, 4H, J )
Scheme 2. Synthesis of 4a
7.5), 8.15 (d, 8H, J ) 8.5), 8.07 (d, 4H, J ) 7.5), 7.93 (m, 12H), 3.94
(s, 12H), -2.70 (s, 2H) ppm. FT-IR-ATR: V(CdO) 1728 cm–1; V(C-
O) 1269 cm–1. LRMS (FAB): m/z 1151.3 [MH+]. UV-vis λmax
(MeOH): 421, 516, 549, 594, 654 nm.
2.6.5. H2-5,10,15,20-tetra(3-ethynyl(4-methylbenzoate)phenyl)-
porphyrin (4b, m-H2TC(PEP)P-[E]). Rf ) 0.65; yield: 175 mg (0.14
1
mmol; 22%). H NMR δH(CDCl3): 8.91 (s, 8H), 8.44 (m, 4H), 8.25
(d, 4H, J ) 7.5), 8.01 (m, 12H), 7.79 (t, 4H, J ) 7.5), 7.64 (d, 4H, J
) 8.0), 3.92 (s, 12H), -2.80 (s, 2H) ppm. FT-IR-ATR: V(CdO) 1727
cm–1; V(C-O) 1267 cm–1. LRMS (FAB): m/z 1247.3 [MH+]. UV-
vis λmax (MeOH): 420, 515, 548, 591, 653 nm.
porphyrins.21 Pyrrole (175 µL, 2.5 mmol) and the aldehydes 3a and
4a, respectively, 2.5 mmol) were dissolved in CH2Cl2 (100 mL) and
the solution was purged with nitrogen for 10 min. The reaction mixture
was then treated with trifluoroacetic acid (186 µL, 2.5 mmol) and stirred
overnight (12 h) at room temperature, resulting in a deep red solution.
DDQ (425 mg, 1.88 mmol) was then added to the solution in one
portion, and stirring was continued for 6 h. NaHCO3 (∼5 g) was then
added in one portion to the reaction mixture, which was stirred until
the acid was neutralized. After gravity filtration, the solvent was
evaporated and the crude mixture was purified by silica gel gravity
chromatography using dichloromethane/ethyl acetate (95:5) to afford
pure porphyrin.
2.6.6. General Procedure for Porphyrin Metalation. The free base
porphyrin (0.10 mmol) was dissolved in CH2Cl2 (20 mL), and the
solution was purged with nitrogen for 10 min. A solution of Zn(OAc)2
(32 mg, 0.15 mmol) in MeOH (5 mL) was added to the porphyrin
solution in one portion, and the reaction mixture was stirred overnight
(12 h). The solvents were removed in vacuo leaving a purple solid,
which was dissolved in CH2Cl2 and washed successively with 5%
aqueous NaHCO3 and water. The organic layer was dried over MgSO4,
and the solvent was removed in vacuo. The metalloporphyrins were
obtained in excellent (close to quantitative) yields following silica gel
chromatography with dichloromethane/ethyl acetate (95:5).
2.6.4. H2-5,10,15,20-tetra(3-(4-methylbenzoate)phenyl)porphyrin
(3b, m-H2TCP2P-[E]). Rf ) 0.59; yield: 198 mg (0.17 mmol; 28%).
2.6.7. Zn(II)-5,10,15,20-tetra(3-(4-methylbenzoate)phenyl)por-
phyrin (3c, m-ZnTCP2P-[E]). Rf ) 0.60; yield: 108 mg (0.089 mmol;
89%). 1H NMR δH(CDCl3): 9.05 (s, 8H), 8.51 (m, 4H), 8.30 (m, 4H),
8.05 (m, 12H), 7.88 (m, 12H), 3.85 (s, 12H) ppm. 13C NMR
(21) Lindsey, J. S.; Schreiman, I. C.; Hsu, H. C.; Kearney, P.; Marguerettaz, A.
M. J. Org. Chem. 1987, 52, 827.
9
J. AM. CHEM. SOC. VOL. 129, NO. 15, 2007 4659

A R T I C L E S
Rochford et al.
precipitated out of solution. The purple solid was filtered and washed
with EtOAc.22
2.6.13. Zn(II)-5,10,15,20-tetra(4-triethylammoniumcarboxyphe-
nyl)porphyrin (1e, p-ZnTCPP-[S]). Yield: 48 mg (0.038 mmol; 76%).
1H NMR δH(CD3OD): 8.84 (s, 8H), 8.34 (m, 8H), 8.22 (m, 8H), 3.02
(m, 24H), 1.20 (m, 36H) ppm. UV-vis λmax (MeOH): 424, 557, 597
nm. FT-IR-ATR: V(CO2 )as 1564 cm–1; V(CO2 )s 1271 cm–1.
–
–
2.6.14. Zn(II)-5,10,15,20-tetra(3-triethylammoniumcarboxyphe-
nyl)porphyrin (2e, m-ZnTCPP-[S]). Yield: 44 mg (0.036 mmol;
1
71%). H NMR δH(CD3OD): 8.80 (m, 12H), 8.39 (m, 8H), 7.81 (t,
4H, J ) 7.5 Hz) ppm. UV-vis λmax (MeOH): 423, 558, 597 nm. FT-
IR-ATR: V(CO2 )as 1564 cm–1; V(CO2 )s 1271 cm–1.
–
–
2.6.15. Zn(II)-5,10,15,20-tetra(3-(4-triethylammoniumcarbox-
yphenyl)phenyl)porphyrin (3e, m-ZnTCP2P-[S]). Yield: 56 mg
(0.036 mmol; 71%). 1H NMR δH(CD3OD): 8.92 (s, 8H), 8.50 (m, 4H),
8.23 (m, 4H), 8.08 (m, 12H), 7.93 (d, 8H, J ) 8.5 Hz), 7.85 (t, 4H, J
) 7.5 Hz), 3.21 (m, 24H), 1.24 (m, 36H) ppm. UV-vis λmax (MeOH):
Figure 4. FT-IR-ATR spectra of p-ZnTCPP-A bound to TiO2. The broad
bands observed at ∼3000 cm–1 for the TiO2 samples are assigned to adsorbed
moisture on the film.
424, 558, 597 nm. FT-IR-ATR: V(CO2 )as 1580 (br) cm–1; V(CO2 )s
–
–
1260 (br) cm–1.
2.6.16. Zn(II)-5,10,15,20-tetra(3-ethynyl(4-triethylammoniumcar-
boxyphenyl)phenyl)porphyrin (4e, m-ZnTC(PEP)P-[S]). Yield: 56
mg (0.034 mmol; 68%). 1H NMR δH(CD3OD): 8.88 (s, 8H), 8.35 (m,
4H), 8.20 (m, 4H), 7.92 (m, 12H), 7.78 (m, 4H), 7.56 (d, 8H, J ) 7.5
Hz), 3.07 (m, 24H), 1.21 (m, 36H) ppm. UV-vis λmax (MeOH): 425,
δC(CDCl3): 166.81, 150.23, 145.25, 143.41, 138.08, 134.12, 133.25,
132.17, 130.12, 128.87, 127.27, 127.19, 126.40, 120.80 ppm. FT-IR-
ATR: V(CdO) 1722 cm–1; V(C-O) 1275 cm–1. HRMS (MALDI): m/z
1212.3140 [M+]. Calcd for C76H52N4O8Zn: 1212.3077. UV-vis λmax
(MeOH): 426, 557, 596 nm.
–
–
558, 598 nm. FT-IR-ATR: V(CO2 )as 1580 (br) cm–1; V(CO2 )s 1260
(br) cm–1.
2.6.8. Zn(II)-5,10,15,20-tetra(3-ethynyl(4-methylbenzoate)phe-
nyl)porphyrin (4c, m-ZnTC(PEP)P-[E]). Rf ) 0.65; yield: 121 mg
(0.092 mmol; 92%). 1H NMR δH(CDCl3): 9.00 (s, 8H), 8.43 (m, 4H),
8.24 (d, 4H, J ) 7.5), 7.97 (d, 4H, J ) 8.5), 7.91 (m, 8H), 7.78 (t, 4H,
J ) 7.5), 7.57 (m, 8H), 3.86 (s, 12H) ppm. 13C NMR δC(CDCl3):
166.22, 150.06, 143.04, 137.26, 134.61, 132.07, 131.29, 131.27, 130.94,
129.25, 129.23, 129.11, 129.09, 127.62, 127.60, 126.77, 121.09, 120.00,
92.38, 88.72 ppm. FT-IR-ATR: V(CdO) 1720 cm–1; V(C-O) 1270
cm–1. HRMS (FAB): m/z 1308.3088 [M+]. Calcd for C84H52N4O8Zn:
1308.3077. UV-vis λmax (MeOH): 425, 557, 597 nm.
2.6.9. General Procedure for Hydrolysis of the Porphyrins Esters
to Acids. The porphyrin tetramethyl ester (0.10 mmol) was dissolved
in THF (10 mL), and 2 N NaOH(aq) (5 mL) was then added in one
portion. The solution was stirred at room temperature for 3 days. The
THF was removed in vacuo, and any unreacted ester was extracted
with chloroform. The acid was precipitated by the slow addition of 1
N HCl(aq) to the aqueous phase. The precipitate was filtered and
thoroughly washed with water and diethyl ether.
3. Results and Discussion
3.1. Synthesis. All porphyrins in this study were prepared
using reported synthetic methodologies. First, the respective
aldehydes were condensed with pyrrole using trifluoroacetic acid
and 2,3-dichloro-5,6-dicyanoquinone (DDQ) as the oxidizing
agent to form the tetra(methylbenzoate) porphyrin (Scheme 1).21
After metalation upon reaction with Zn(OAc)2, the four methyl
ester groups were hydrolyzed in basic conditions to give the
tetracarboxylic acid porphyrins. Porphyrins p-ZnTCPP-[A]19 and
m-ZnTCPP-[A]9a are known compounds and were synthesized
in total yields of ∼19% using methyl-4-formylbenzoate (1a)
and methyl-3-formylbenzoate (2a), respectively. The synthesis
of porphyrins m-ZnTCP2P-[E] and m-ZnTC(PEP)P-[E] was
performed using methyl-4-(3-formylphenyl)benzoate (3a) and
methyl-4-(3-ethynylformylphenyl)benzoate (4a), respectively.
4a was synthesized in 69% yield via the Sonogashira coupling
reaction of 3-ethynylbenzaldehyde with methyl-4-iodobenzoate
(Scheme 2). Both aldehydes undergo efficient porphyrin forma-
tion when condensed with pyrrole under the above conditions
with total yields of 20% and 16% for m-ZnTCP2P-[E] and
m-ZnTC(PEP)P-[E], respectively. Porphyrin m-ZnTCP2P-[A]
was much less soluble in any solvent than p-ZnTCPP-[A],
m-ZnTCPP-[A], and m-ZnTC(PEP)-[A]. To work with com-
pounds that were equally soluble in the same solvents, the tetra-
(triethylammonium)salts were readily prepared from the cor-
responding carboxylic acids using a small excess of triethylamine
with yields of 68-76%.
2.6.10. Zn(II)-5,10,15,20-tetra(3-(4-carboxyphenyl)phenyl)por-
phyrin (3d, m-ZnTCP2P-[A]). Yield: 95 mg (0.082 mmol; 82%). 1H
NMR δH(THF-d8): 8.97 (s, 8H), 8.59 (m, 4H), 8.27 (m, 4H), 8.12 (m,
12H), 8.00 (m, 8H), 7.88 (m, 4H) ppm. FT-IR-ATR: V(CdO) 1686
cm–1; V(C-O) 1283 cm–1. HRMS (FAB): m/z 1157.2516 [MH+]. Calcd
for C72H45N4O8Zn: 1157.2529. UV-vis λmax (MeOH): 424, 557, 596
nm.
2.6.11. Zn(II)-5,10,15,20-tetra(3-ethynyl(4-carboxyphenyl)phe-
nyl)porphyrin (4d, m-ZnTC(PEP)P-[A]). Yield: 99 mg (0.079 mmol;
1
79%). H NMR δH(CD3OD): 8.92 (s, 8H), 8.43 (m, 4H), 8.28 (m,
4H), 7.98 (m, 12H), 7.81 (m, 4H), 7.62 (m, 8H) ppm. FT-IR-ATR:
V(CdO) 1690 cm–1; V(C-O) 1280 cm–1. HRMS (FAB): m/z 1252.2426
[M+]. Calcd for C80H44N4O8Zn: 1252.2451. UV-vis λmax (MeOH):
425, 557, 597 nm.
3.2. FT-IR-ATR. 3.2.1. Neat Samples. All solid samples
(Figure 2 and Supporting Information) showed spectra typical
of the meso-tetraphenylporphyrin macrocycle with bands from
the pyrrole C-H, CdC, and CdN stretches evident over the
2.6.12. General Procedure for the Synthesis of Porphyrin Tetra-
(triethylammonium) Salts. The porphyrin tetracarboxylic acid (0.05
mmol) was dissolved/suspended in THF (10 mL). Et3N (2-3 drops)
was then added, and the solution was stirred at room temperature for
3 h. EtOAc (10 mL) was then added, and the THF was carefully
removed under vacuum, leaving the higher boiling EtOAc behind. As
the THF was removed, the porphyrin tetratriethylammonium salt
(22) During the FAB-MS analysis of all tetrabutylammonium salts it was
observed that all compounds were converted to their tetraacid precursors
in the 3-nitrobenzyl alcohol/glycerol matrix.
9
4660 J. AM. CHEM. SOC. VOL. 129, NO. 15, 2007

Porphyrins for Semiconductor Sensitization
A R T I C L E S
Table 1. Solution UV-Vis Absorption and Fluorescence Emission Data of COOEt3NH Derivatives of All Four Porphyrins in Methanol
UV
−vis absorption
fluorescence
Soret
Q(1,0)
Q(0,0)
λ
max, nm
λ
max, nm
λ
max, nm
1
porphyrin
(
ꢀ × 105, M–1 L–1
)
(
ꢀ × 104, M–1 L–1
)
(
ꢀ × 104, M−1L−
)
λmax, nm (Φ)
(1e) p-ZnTCPP-[S]
(2e) m-ZnTCPP-[S]
(3e) m-ZnTCP2P-[S]
(4e) m-ZnTC(PEP)P-[S]
424 (2.78)
423 (4.44)
424 (5.51)
425 (5.93)
557 (1.39)
558 (2.09)
558 (2.77)
558 (2.63)
597 (0.53)
597 (0.66)
597 (0.96)
598 (0.83)
606, 658 (0.023)
604, 657 (0.016)
605, 659 (0.017)
604, 659 (0.018)
range of 700-1500 cm–1.23 Neat samples of m-ZnTCPP-[A],
m-ZnTCP2P-[A], and m-ZnTC(PEP)P-[A] showed a strong band
in the region of 1680-1690 cm–1 due to the V(CdO) stretch
and a broad band at 1280-1295 cm–1 due to the V(C-O) stretch
of the carboxylic acid groups. Extensive hydrogen bonding of
the carboxylic acid groups in p-ZnTCPP-[A]24 resulted in a shift
to lower frequency and broadening of the V(CdO) stretch at
1638 cm–1 and a shift to higher frequency of the V(C-O) stretch
at 1384 cm–1. The differences observed between the spectra of
the carboxylic acids and the triethylammonium salts correspond
well with those observed between the spectra of benzoic acid
and its triethylammonium salt (see the Supporting Information).
Both p- and m-ZnTCPP-[S] showed new bands at 1564 and
1271 cm–1 which were assigned to the asymmetric and sym-
weak acetylenic V(CtC) stretch was observed from 2215 to
2233 cm–1 for all of the m-ZnTC(PEP)P derivatives. The
characteristic N+-H bending mode of the triethylammonium
counterion is observed in the 2850-2950 cm–1 region for all
salt derivatives.25
3.3. UV-Vis Absorption and Fluorescence Emission
Spectra. 3.3.1. Solution. Table 1 summarizes the solution
absorption and emission data for the COOEt3NH derivatives
of all four porphyrins in methanol. The UV-vis absorption and
fluorescence emission spectra in fluid solutions of all derivatives
(COOH, COOMe, and COOEt3NH) of the zinc porphyrins were
similar. Also, neither the substitution position (para vs meta),
nor the length of the rigid spacer between the porphyrin ring
and the carboxy groups, caused any significant shift in their
–
metric V(CO2 ) stretching modes, respectively. For both
absorption
(Table 1).
or
emission
maxima
–
m-ZnTCP2P-[S] and m-ZnTC(PEP)P-[S] the asymmetric V(CO2 )
stretch was very broad and at slightly higher frequency, ∼1580
cm–1. Their corresponding symmetric stretches were broadened
and shifted to lower frequency, at ∼1260 cm–1. For all
triethylammonium salts there was an additional V(CdO) stretch
at 1702 cm–1 due to hydrogen bonding of the Et3N+H proton
with the carboxylate group.25
All solution UV-vis absorption spectra exhibited the features
typical of the ZnTPP ring, with absorptions at ∼424 nm (the
“Soret” band, corresponding to the S0 f S2 transition) and at
∼558 and ∼597 nm (the “Q bands”, corresponding to the higher
vibrational mode Q(1,0) and the lowest energy vibrational mode
Q(0,0) of the S0 f S1 transition). Figure 5a shows an overlay
of the UV-vis spectra of the COOEt3NH derivatives for all
four porphyrins recorded in MeOH.
The similarities observed in the solution absorption spectra
of all porphyrins here studied are attributed to the lack of
electronic communication between the meso-aryl substituents
and the porphyrin ring. In this respect, p-ZnTCPP, m-ZnTCPP,
m-ZnTCP2P, and m-ZnTC(PEP)P are good models to determine
the effect of binding orientation and sensitizer distance without
significantly changing the electronic properties of the porphyrin
ring. However, the rotational barrier of the o-H-substituted meso-
aryl substituents is low.10c,27 For meso-aryl substituents having
bulkier groups in the ortho position, atropisomers are typically
observed.28 No atropisomers were observed for any porphyrins
synthesized in this study. The solution fluorescence emission
spectra of the four porphyrin ammonium salts p-ZnTCPP-[S],
m-ZnTCPP-[S], m-ZnTCP2P-[S], and m-ZnTC(PEP)P-[S] were
also typical of tetra-aryl Zn(II)porphyrin systems and showed
almost identical λmax in their spectra (Figure 5b).
3.2.2. Bound to TiO2. In the bound spectra of the meta-
substituted tetraacid porphyrins (Figure 2) the V(CdO) and V(C-
O) stretching modes disappear with the appearance of a broad
shoulder on the high-frequency side of the phenyl breathing
mode (1605-1740 cm–1). For all four porphyrin acids and salts,
binding to TiO2 films resulted in the appearance of strong and
broad bands in the 1390-1410 cm–1 region, which are char-
–
acteristic of the symmetric V(CO2 ) stretch. A broad band
observed at ∼1540 cm–1 for the porphyrin acids and salts bound
–
to TiO2 was assigned to the V(CO2 ) asymmetric stretching
mode. Overall, the spectral changes, both for porphyrin acids/
TiO2 and salts/TiO2, are consistent with chelating and/or
bidentate binding modes of the carboxylate groups on the TiO2
surface (Figure 3).26
This is also confirmed by a decrease in the difference between
the symmetric and asymmetric bands of the carboxylate group
of the salt derivatives (∆ ∼ 290 cm–1) on binding to TiO2 (∆
∼ 240 cm–1).26 A number of V(CdO) stretching modes were
still evident for p-ZnTCPP-[A] on TiO2 indicating the presence
of free carboxylic acid groups for this sensitizer (Figure 4). The
3.3.2. Bound to TiO2, ZnO, and ZrO2 Films. Both the
tetracarboxylic acid and tetra(triethylammonium)salt derivatives
(23) (a) Thomas, D. W.; Martell, A. E. J. Am. Chem. Soc. 1959, 81, 5111. (b)
Datta-Gupta, N.; Bardos, T. J. J. Heterocycl. Chem. 1966, 3, 495. (c) Longo,
F. R.; Finarelli, M. G.; Kim, J. B. J. Heterocycl. Chem. 1969, 6, 927.
(24) Diskin-Posner, Y.; Goldberg, I. Chem. Commun. 1999, 1961.
(25) Odinokov, S. E.; Glazunov, V. P.; Nabiullun, J. Chem. Soc., Faraday Trans.
2 1984, 80, 899.
(26) (a) Deacon, G. B.; Phillips, R. J. Coord. Chem. ReV. 1980, 33, 227. (b)
Finnie, K. S.; Bartlett, J. R.; Woolfrey, J. L. Langmuir 1998, 14, 2744. (c)
Vittadini, A.; Selloni, A.; Rotzinger, F. P.; Gra¨tzel, M. J. Phys. Chem. B
2000, 104, 1300. (d) Nazeeruddin, M. K.; Humphrey-Baker, R.; Liska, P.;
Gra¨tzel, M. J. Phys. Chem. B 2003, 107, 8981.
(27) (a) Eaton, S. S.; Eaton, G. R. J. Am. Chem. Soc. 1975, 97, 3660. (b) Eaton,
S. S.; Eaton, G. R. J. Am. Chem. Soc. 1977, 99, 6594. (c) Eaton, S. S.;
Fishwild, D. M.; Eaton, G. R. Inorg. Chem. 1978, 17, 1542. (d) Stolzenberg,
A. M.; Haymond, G. S. Inorg. Chem. 2002, 41, 30. (e) Medforth, C. J.;
Haddad, R. E.; Muzzi, C. M.; Dooley, N. R.; Jaquinod, L.; Shyr, D. C.;
Nurco, D. J.; Olmstead, M. M.; Smith, K. M.; Ma, J.-G.; Shelnutt, J. A.
Inorg. Chem. 2003, 42, 2227. (f) Ayabe, M.; Yamashita, K.; Sada, K.;
Shinkai, S.; Ikeda, A.; Sakamoto, S.; Yamaguchi, K. J. Org. Chem. 2003,
68, 1059. (g) Tong, L. H.; Pascu, S. I.; Jarrosson, T.; Sanders, J. K. M.
Chem. Commun. 2006, 1085
(28) Freitag, R.; Whitten, D. G. J. Phys. Chem. 1983, 87, 3918.
9
J. AM. CHEM. SOC. VOL. 129, NO. 15, 2007 4661

A R T I C L E S
Rochford et al.
Figure 6. (a) UV-vis absorption spectra of tetra(triethylammonium)-
carboxyporphyrin salts on TiO2/G. Comparable UV-vis absorption spectra
were observed for all sensitizers on TiO2 and ZrO2 mesoporous films. (b)
Picture of (i) blank TiO2/ITO, (ii) p-ZnTCPP-[A], (iii) m-ZnTCPP-[A], (iv)
m-ZnTCP2P-[A], and (v) m-ZnTC(PEP)-[A] bound to TiO2/ITO films. Note
the difference in color between the p-ZnTCPP-[S] film and the meta-
substituted systems.
between porphyrin chromophores, and previous studies on TiO2
nanoparticles attributed this effect to the formation of H-
aggregates (face-to-face stacking) between porphyrin mol-
ecules.30 A small red shift (3 nm, λmax ) 601 nm) and
broadening of the lower energy Q(0,0) band was observed for
the p-ZnTCPP acid or salt derivative bound to TiO2, in
comparison to the meta porphyrin acids or salts (λmax ) 598
nm), which are expected to bind flat on TiO2.9a,31
The fluorescence spectra of the meta porphyrin salts bound
to ZrO2, presumably with a planar binding mode, showed little
deviation from their solution-phase spectra. However, the para-
substituted porphyrin p-ZnTCPP (acid or salt) on ZrO2, dis-
played broadening and convergence of the two porphyrin
fluorescence bands (Figure 8). We attributed this effect to the
formation of H-aggregates of the porphyrin ring on the ZrO2
surface.
3.4. Surface Coverage. Coverage of the tetra(triethylammo-
nium)carboxyporphyrin salts, calculated on the TiO2 mesoporous
films, decreased with the increasing length of the spacer:
m-ZnTCPP-[S] (20 µmol g–1) > m-ZnTCP2P-[S] (19 µmol g–1)
> m-ZnTTC(PEP)P-[S] (12 µmol g–1). The highest value was
obtained for p-ZnTCPP-[S] (27 µmol g–1) consistent with a
closer packing in the para porphyrin, for which the “vertical”
binding mode is proposed. The value reported here for p-
Figure 5. Solution (a) UV-vis absorption spectra and (b) fluorescence
emission spectra of tetra(triethylammonium)carboxyporphyrin salts recorded
in methanol.
for all four porphyrins were found to bind efficiently to TiO2,
ZnO,29 and ZrO2 nanoparticle films. Because of the wider band
gap, ZrO2 behaves as an insulator precluding electron injection
from the lowest excited-state of all porphyrins studied here (Ebg
∼ 5 eV for ZrO2 compared to ∼3 eV for TiO2 and ZnO). Since
the fluorescence is not quenched, the ZrO2 films are an excellent
substrate to study the fluorescence emission of the porphyrins
bound to a surface that closely resembles the morphology of
TiO2 films. Significant differences in the absorption and
emission spectra were observed between the para and meta
porphyrin systems (acids or salts). The UV-vis absorption
spectra of the triethylammonium salts bound to TiO2 are shown
in Figure 6.
Comparable UV-vis absorption spectra were observed for
all porphyrins (acids or salts) on TiO2 and ZrO2 mesoporous
films. The thicker (∼10 µm) TiO2 and ZrO2 films had a greater
absorption when compared to that of the thinner ZnO films (∼2
µm). The Soret absorption, which was very intense on TiO2
and ZrO2 films, was less intense on ZnO (Figures 6 and 7).
The most notable feature in the absorption spectra on ZnO
films was the blue shift (10 nm) and splitting (λmax ) 420 and
430 nm) of the Soret band for the para porphyrin p-ZnTCPP
(acid or salt). Such behavior is indicative of exciton interaction
(30) Aratani, N.; Osuka, N.; Cho, H. S.; Kim, D. J. Photochem. Photobiol., C
2002, 3, 25.
(31) There is some precedent in the literature supporting the planar orientation
of meta-substituted tetraphenylporphyrins on various surfaces, see: (a)
Skupin, M.; Li, G.; Fudickar, W.; Zimmerman, J.; Ro¨der, B.; Fuhrhop,
J.-H. J. Am. Chem. Soc. 2001, 123, 3454. (b) Li, G.; Doblhofer, K.; Fuhrhop,
J.-H. Angew. Chem., Int. Ed. 2002, 41, 2730. (c) Bhosale, S.; Li, G.; Li,
F.; Wang, T.; Ludwig, R.; Emmler, T.; Buntkowsky, G.; Fuhrhop, J.-H.
Chem. Commun. 2005, 3559.
(29) The IPCE study was done on TiO2 films. A photoelectrochemical study of
m-ZnTCPP-[A] bound to ZnO nanorods has been previously reported, and
similar studies for all porphyrins here described are forthcoming.
9
4662 J. AM. CHEM. SOC. VOL. 129, NO. 15, 2007

Porphyrins for Semiconductor Sensitization
A R T I C L E S
Figure 9. Calculated molecular dimensions of (a) p-TCPP and (b)
m-ZnTCPP, m-ZnTCP2P, and m-ZnTC(PEP) (ref 33).
Table 2. Solution Redox Potentials of Tetramethylester
Porphyrins in Dichloromethane Reported versus NHE (V)a
oxidation (V)
reduction (V)
E0
(eV)
E1 2(P+/P*)
(eV)
-
0
/
porphyrin
1st
2nd
1st
2nd
3rdb
Figure 7. UV-vis absorption spectra on mesoporous ZnO/G films of the
(a) COOH derivatives and (b) COO–Et3NH+ derivatives for all four
porphyrins.
(1c) p-ZnTCPP-[E]
(2c) m-ZnTCPP-[E]
(3c) m-ZnTCP2P-[E]
1.10 1.47 -1.12 -1.46 -1.64 2.06 -1.04
1.11 1.38 -1.08 -1.45 2.07 -1.04
1.12 1.37 -1.08 -1.31 -1.46 2.07 -1.05
(4c) m-ZnTC(PEP)P-[E] 1.11 1.38 -1.09 -1.33 -1.46 2.07 -1.04
a All cyclic voltammograms were measured at a scan rate of 100 mV
s–1 using a glassy carbon working electrode. b Determined from differential
pulse voltammetry experiments.
expected considering the size (Figure 9) and anticipated binding
geometry (“flat” vs “vertical”) of each sensitizer (Figure 1). A
lower surface coverage would be expected with increasing
footprint size of the sensitizer (Figure 9). Also, larger molecules
may not penetrate the smaller pores in the TiO2 film.
3.5. Electrochemical Properties. 3.5.1. Solution. Table 2
lists the solution redox potentials of tetramethylester porphyrins
p-ZnTCPP-[E], m-ZnTCPP-[E], m-ZnTCP2P-[E], and m-ZnTC-
(PEP)P-[E] reported versus NHE. The solution electrochemical
measurements were performed on the methylester derivatives,
because the acid or ammonium salt derivatives were insoluble
in dichloromethane. All compounds exhibited redox behavior
typical of Zn(II)TPP porphyrins (Table 2).
Figure 8. Fluorescence emission spectra of tetra(triethylammonium)-
carboxyporphyrin salts on ZrO2/G.
ZnTCPP-[S] is lower than that previously reported by Cherian
and Wamser32 for p-ZnTCPP-[A] (47 µmol g–1). However, that
value described the saturation coverage of the mesoporous TiO2
film after binding overnight, whereas the binding times in our
experiments were much shorter (1 h). Variations in film
thickness and nanoparticle size, commonly observed from
laboratory to laboratory, could also lead to such differences.
The trend observed here in the coverage is what would be
Two reversible oxidation processes were observed due to the
formation of the porphyrin radical cation and the diradical cation
(33) For p-TCPP with one bound carboxyl group the approximate height from
the TiO2 surface was taken as the distance from C1 to the midpoint of O1
and O2 and the diameter as the distance from O3 to O4. For p-TCPP with
two bound carboxyl groups, the approximate height from the TiO2 surface
was taken as the distance from the midpoint of C1 and C2 to the midpoint
of O1 and O2 and the diameter as the distance from O3 to O4. For the meta-
substituted porphyrins the height was taken as the distance from the plane
of the porphyrin to the plane defined by the four oxygens O1-O4. The
footprint diameter was calculated as the distance from O5 to O6. Modeling
was done using Spartan ’02 modeling software, Wavefunction, Inc.
(32) Cherian, S.; Wamser, C. C. J. Phys. Chem. B 2000, 104, 3624.
9
J. AM. CHEM. SOC. VOL. 129, NO. 15, 2007 4663

A R T I C L E S
Rochford et al.
Table 3. Redox Potentials of
Tetra(triethylammonium)carboxyporphyrin Salts Bound to TiO2/ITO
and ZnO/ITO Films versus NHE (V)a
TiO2/ITO
(V)
ZnO/ITO
(V)
E0
E1 2(P+/P*)
E0
E1 2(P+/P*)
-0
/
-0
/
porphyrin
1st 2ndb (eV)
(eV)
1st 2ndb (eV)
(eV)
(1e) p-ZnTCPP-[S]
(2e) m-ZnTCPP-[S]
(3e) m-ZnTCP2P-[S]
1.09 1.38 2.03 -1.06 1.07 1.43 2.03 -1.04
1.10 1.36 2.06 -1.04 1.06 1.42 2.06 -1.00
1.09 1.35 2.06 -1.03 1.07 1.42 2.05 -1.02
(4e) m-ZnTC (PEP)P-[S] 1.10 1.36 2.06 -1.04 1.04 1.41 2.05 -0.99
a All samples were run at a scan rate of 100 mV s–1 using the derivatized
MO/ITO films as the working electrode. b Anodic peak maxima (Epa).
Cyclic voltammetry scans for m-ZnTC(PEP)P-[S] bound to
TiO2/ITO and ZnO/ITO working electrodes are shown in Figure
10b. The oxidation of sensitizers bound to wide band gap
semiconductors, such as TiO2, has been explained by an
electron-hole transfer process which permeates through the
adsorbed layer of dye.36 According to this mechanism, the
molecules closest to the conducting glass are oxidized, and the
adjacent dye molecules are then oxidized via electron-hole
transfer. The large peak width observed for m-ZnTC(PEP)P-
[S] bound to TiO2/ITO or ZnO/ITO working electrodes (∼200
mV at half-height) was typical for all compounds and is most
probably due to a slight dispersion in the redox potentials caused
by inhomogeneities of the films. E1/2(P+/P*) is an important
value to consider when designing a chromophore for DSSCs,
as it determines the driving force for electron transfer to the
lower-lying conduction band of the semiconductor. The excited-
state oxidation potential of all porphyrins here studied was about
-1 V versus NHE, which is quite negative with respect to the
conduction bands of both the TiO237 and ZnO38 semiconductors.
In summary, these porphyrins are good photoreductants. A
comparison between the E1/2(P+/P*) values in solution and
bound showed a slight increase in the excited-state oxidation
potential for p-ZnTCPP-[S]. This effect was once again at-
tributed to aggregation at the metal oxide surface for the para
porphyrin and was not observed for the meta porphyrins.
3.6. Photoelectrochemical Measurements in DSSCs. The
main objective of the photoelectrochemical experiments was
to determine the effect of the binding geometries of the sensitizer
on the solar cell efficiencies. The photocurrent action spectra
of all four triethylammonium porphyrin salts bound to TiO2/
FTO are shown in Figure 11.
Figure 10. (a) Solution (CH2Cl2) cyclic voltammogram (i) and differential
pulse voltammogram (ii) scans of m-ZnTC(PEP)P-[E]. (b) Cyclic voltam-
mogram scans of m-ZnTC(PEP)P-[S] bound to (i) TiO2/ITO and (ii) ZnO/
ITO electrodes.
species, respectively. In the negative potential region three quasi-
reversible reduction processes were observed.34 The first reduc-
tion process, observed in the range of -1.08 to -1.12 V versus
NHE, is attributed to the reduction of the low-lying LUMO
orbital of the carboxy moieties. The second and third reduction
processes are attributed to the reversible formation of the
porphyrin radical anion and diradical anion species, respec-
tively.35 Only two reduction processes were observed for
m-ZnTCPP-[E], and it is possible that the formation of the
diradical anion species of this compound occurs outside the
potential window under the conditions employed. Cyclic vol-
tammetry and differential pulse voltammetry scans for m-ZnTC-
(PEP)P-[E] are shown in Figure 10a.
3.5.2. Bound to TiO2 and ZnO. Analogous to the solution
studies, a reversible first oxidation was observed in the range
of +1.09 to +1.10 V versus NHE on TiO2 and +1.04 to +1.07
V versus NHE on ZnO surfaces, due to the formation of the
porphyrin radical cation species. Formation of the diradical
cationic species was irreversible on both TiO2 and ZnO surfaces.
For DSSCs applications, the stability of the sensitizer is
dependent on the reversibility of the first oxidation process
following charge injection into the metal oxide semiconductor
from the photoexcited chromophore. Table 3 summarizes the
redox behavior of p-ZnTCPP-[E], m-ZnTCPP-[E], m-ZnTCP2P-
[E], and m-ZnTC(PEP)P-[E] bound to both TiO2/ITO and ZnO/
ITO mesoporous thin films.
The photocurrent action spectra of all sensitizers concurred
reasonably well with their corresponding absorption spectra. The
IPCE for m-ZnTCPP-[S] at 430 nm (59%), corresponding to
the Soret absorption band, represented a 3-fold increase with
respect to the para-substituted analogue p-ZnTCPP-[S] at the
corresponding wavelength. IPCE values at the lower energy Q
bands of m-ZnTCPP-[S] and m-ZnTCP2P-[S] were in the range
of 16-34%, which is a greater than 20-fold increase with respect
to the IPCE values observed in the same region for p-ZnTCPP-
[S] (Table 4).
For all porphyrins, the Soret band showed the highest IPCE
value. This was expected as its extinction coefficient is 1 order
(36) Bonhote, P.; Gogniat, E.; Tingry, S.; Barbe, C.; Vlachopoulos, N.;
Lenzmann, F.; Compte, P.; Gra¨tzel, M. J. Phys. Chem. B 1998, 102, 1498.
(37) Liu, G.; Jaegermann, W.; He, J.; Su¨ndstrom, V.; Sun, L. J. Phys. Chem. B
2002, 106. 5814.
(38) Redmond, R.; O’Keeffe, A.; Burgess, C.; MacHale, C.; Fitzmaurice, D. J.
Phys. Chem. 1997, 97, 11081.
(34) Wolberg, A.; Manassen, J. J. Am. Chem. Soc. 1970, 92, 2982.
(35) Clack, D. W.; Hush, N. S. J. Am. Chem. Soc. 1965, 87, 4238.
9
4664 J. AM. CHEM. SOC. VOL. 129, NO. 15, 2007

Porphyrins for Semiconductor Sensitization
A R T I C L E S
for the meta-substituted systems,39 a 3-, 8-, and 9-fold increase
in Isc was observed for m-ZnTC(PEP)P-[S], m-ZnTCPP-[S], and
m-ZnTCP2P-[S], respectively, in comparison to that of the para-
substituted p-ZnTCPP-[S]. The higher IPCE values observed
for m-ZnTCP2P-[S] in the Q-band region in comparison to those
of m-ZnTCPP-[S] as well as its greater Isc suggest a reduced
charge recombination for this system, possibly because of the
increased distance of the porphyrin ring from the TiO2 surface
(Figure 9) and/or a better surface coverage blocking the electrons
in the TiO2 from reacting with the electrolyte. The lack of any
aggregate absorption band in the photocurrent action spectrum
of p-ZnTCPP-[S] suggests that aggregates do not contribute to
any current generation.
4. Conclusions
Figure 11. Photocurrent action spectra of tetra(triethylammonium)-
Four para- and meta-Zn(II) tetra(carboxyphenyl)porphyrins
were studied in solution and bound to metal oxide (TiO2, ZnO,
and ZrO2) nanoparticle films to determine the effect of the spacer
length and anchoring group position on their photoelectrochemi-
cal and photophysical properties. Both COOH and COOEt3NH
derivatives were employed for the binding studies as well as
solution studies. Solution phase electrochemistry studies were
performed on the methyl ester derivatives (COOMe). In
m-ZnTCPP, m-ZnTCP2P, and m-ZnTC(PEP)P the anchoring
groups are in meta position on the meso-phenyl rings. The meta
substitution favored a planar binding mode to the metal oxide
surfaces, as determined by a combination of studies that included
IR, UV, and solar cell efficiencies on TiO2. Fluorescence
emission was studied on ZrO2. All studies indicated that only
p-ZnTCPP aggregated, suggesting close packing of the dye
molecules on the semiconductor surface. Aggregation effects
were not observed for the meta porphyrins. These observations
are consistent with the work of others on p-ZnTCPP and
m-ZnTCPP. The photoelectrochemical behavior of the para- and
meta-substituted porphyrin sensitizers suggests that the binding
geometry, as well as the distance of the sensitizer from the metal
oxide surface, dramatically influence their efficiencies. The
greater efficiency of the rigid planar meta-substituted systems
was explained in terms of a greater charge injection into the
TiO2 semiconductor from rings that lie flat, and closer, to the
surface. A kinetic study of the injection and recombination
processes for all four compounds described here is clearly of
interest, and this work is forthcoming.
carboxyporphyrin salts.
Table 4. Photoelectrochemical Properties of
Tetra(triethylammonium)carboxyporphyrin Salts
IPCE (%)
Isc
Voc
(V)
porphyrin
(mA cm–2
)
ff 430 nm
570 nma
600 nmb
(1e) p-ZnTCPP-[S]
(2e) m-ZnTCPP-[S]
(3e) m-ZnTCP2P-[S]
0.39 0.44 0.54 18.50 1.44 (0.08) 0.86 (0.05)
3.33 0.51 0.41 58.60 29.40 (0.50) 16.30 (0.28)
3.72 0.50 0.42 56.90 34.50 (0.61) 21.10 (0.37)
(4e) m-ZnTC(PEP)P-[S] 1.36 0.43 0.45 25.30 9.00 (0.36) 4.81 (0.19)
a In parentheses are the Q(1,0) vs Soret peak intensity ratios. b In
parentheses are the Q(0,0) vs Soret peak intensity ratios.
of magnitude larger than for the Q bands. However, the larger
increase in IPCE observed for the meta-substituted systems in
the Q-band region, relative to the Soret absorption, suggests
that the lower energy S1 transition of the porphyrin chromophore
has a greater quantum efficiency for charge injection when
oriented in a planar geometry with respect to the TiO2 surface.
Q(1,0)/Soret peak intensity ratios of 0.08, 0.50, 0.61, and 0.36
were observed for the sensitizers p-ZnTCPP-[S], m-ZnTCPP-
[S], m-ZnTCP2P-[S], and m-ZnTC(PEP)P-[S], respectively. The
IPCE can also be expressed in terms of the light harvesting
efficiency (LHE), the quantum yield of charge injection (φinj),
and the charge collection efficiency at the back electrode (ηc)
in eq 5
IPCE ) (LHE)φinjηc
(5)
Acknowledgment. E.G. thanks DOE (DE-FG02-01ER15256)
for funding. J.R. and E.G. thank the Rutgers Research Council
for travel funds. D.C. thanks the OISE office of NSF for an
REU supplement (NIRT-0303829) for an undergraduate summer
stipend and international travel funds. E.G. thanks Professor
Mendelsohn for generously making available IR instrumentation.
D.C. thanks Dr. Ha¨ggman, Dr. Edvinsson, and Dr. Boschloo
for help with the solar cell measurements.
where ηc is dependent upon both the rate of dye regeneration
by the oxidation of electrolyte and the rate of charge recombina-
tion from the reduced TiO2 particles to the oxidized sensitizer.
According to the extinction coefficients and binding efficiencies
calculated for all four sensitizers, and assuming equivalent
values for φinj and ηc, this equation predicts only a slight increase
in IPCE for the meta-substituted systems. Considering the
comparable solution photophysical properties of all four sen-
sitizers, and their similar ground and excited-state redox
behavior, such large increases in the IPCE values was rational-
ized in terms of greater φinj for the porphyrin chromophore when
oriented in a planar geometry. This interpretation is consistent
with that made by Campbell et al. when they observed a 5-fold
increase in Isc for m-ZnTCPP-[A] with respect to that of
p-ZnTCPP-[A].9a Although little increase of Voc was observed
Supporting Information Available: Preparation of the TiO2
and ZrO2 films; TiO2 Raman, PXRD, and AFM data; p-H2-
TCPP-[E], m-H2TCPP-[E], p-ZnTCPP-[E], m-ZnTCPP-[E], p-
ZnTCPP-[A], and m-ZnTCPP-[A] NMR data; FT-IR-ATR
spectra of 1d and e, 2d and e, 3d and e neat and bound to TiO2.
This material is available free of charge via the Internet at
http://pubs.acs.org.
(39) For an effect of dye distance on Voc see, for instance, ref 6d.
JA068218U
9
J. AM. CHEM. SOC. VOL. 129, NO. 15, 2007 4665