Tetraaryl ZnII porphyrinates substituted at β-pyrrolic positions as sensitizers in dye-sensitized solar cells: A comparison with meso-disubstituted push-pull ZnII porphyrinates

FULL PAPER

DOI: 10.1002/chem.201300219

Tetraaryl Zn^IIPorphyrinates Substituted at b-Pyrrolic Positions as Sensitizers

in Dye-Sensitized Solar Cells: A Comparison with meso-Disubstituted

Push–Pull Zn^IIPorphyrinates

Gabriele Di Carlo,^[a]Alessio Orbelli Biroli,^[b]Maddalena Pizzotti,*^[a]

Francesca Tessore,^[a]Vanira Trifiletti,^[c]Riccardo Ruffo,^[c]Alessandro Abbotto,^[c]

Anna Amat,^[d]Filippo De Angelis,^[d]and Patrizia R. Mussini^[a]

Abstract: A facile and fast approach,

based on microwave-enhanced Sonoga-

shira coupling, has been employed to

obtain in good yields both mono- and,

for the first time, disubstituted push–

pull Zn^IIporphyrinates bearing a varie-

ty of ethynylphenyl moieties at the b-

show comparable or better efficiencies

when acting as sensitizers in DSSCs.

This behavior is apparently not attrib-

utable to more intense B and Q bands,

but rather to more facile charge injec-

tion. This is suggested by the DFT elec-

tron distribution in a model of a b-

monosubstituted porphyrinic dye inter-

acting with a TiO₂surface and by the

positive effect of the b substitution on

the incident photon-to-current conver-

sion efficiency (IPCE) spectra, which

In contrast, meso-substitution produces

IPCE spectra with two less intense and

well-separated peaks. The positive

effect exerted by a cyanoacrylic acid

group attached to the ethynylphenyl

substituent has been analyzed by a

photophysical and theoretical ap-

proach. This provided supporting evi-

dence of a contribution from charge-

transfer transitions to both the B and

Q bands, thus producing, through con-

jugation, excited electrons close to the

carboxylic anchoring group. Finally, the

straightforward and effective synthetic

procedures developed, as well as the

efficiencies observed by photoelectro-

chemical measurements, make the de-

scribed b-monosubstituted Zn^IIpor-

phyrinates extremely promising sensi-

tizers for use in DSSCs.

pyrrolic position(s). Furthermore,

a

comparative experimental, electro-

chemical, and theoretical investigation

has been carried out on these b-mono-

or disubstituted Zn^IIporphyrinates and

meso-disubstituted push–pull Zn^IIpor-

phyrinates. We have obtained evidence

that, although the HOMO–LUMO

energy gap of the meso-substituted

push–pull dyes is lower, so that charge

transfer along the push–pull system

therein is easier, the b-mono- or disub-

stituted push–pull porphyrinic dyes

show

a significant intensity over a

broad wavelength range (350–650 nm).

Keywords: cyclic voltammetry

density functional calculations

·

porphyrins · push–pull systems · so-

lar cells

Introduction

Dye-sensitized solar cells (DSSCs) have emerged as an inno-

vative and competitive alternative to conventional solar

cells based on silicon, due to their potential low cost and in-

teresting conversion efficiencies.^[1]Until recently, the most

efficient DSSCs were mainly based on polypyridyl rutheni-

um complexes as dyes,^[2]reaching up to 11% light conver-

sion efficiency.^[3]However, the poor absorption of these

dyes in the near-infrared region of the solar spectrum and

the relatively high cost and environmental issues of rutheni-

um complexes encouraged the search for more efficient and

possibly cheaper dyes.

Inspired by the process of solar energy collection by the

photosynthetic cores of bacteria and plants, which incorpo-

rate a porphyrinic center as a light-harvesting chromophore,

porphyrinic structures have been considered as interesting

dyes and some of them have been synthesized and investi-

gated for applications in DSSCs.^[1,4]Porphyrins, by virtue of

their strong electronic absorption bands up to the near-infra-

red region and their long-lived p* singlet excited states of

[a] Dr. G. Di Carlo, Prof. M. Pizzotti, Dr. F. Tessore, Prof. P. R. Mussini

Dipartimento di Chimica dellꢀUniversitꢁ degli Studi di Milano

Unitꢁ di Ricerca dellꢀINSTM

Via C. Golgi 19, 20133 Milano (Italy)

Fax : (+39)02-503-14405

E-mail: maddalena.pizzotti@unimi.it

[b] Dr. A. Orbelli Biroli

Istituto di Scienze e Tecnologie Molecolari del CNR (CNR-ISTM)

Via C. Golgi 19, 20133 Milano (Italy)

[c] Dr. V. Trifiletti, Dr. R. Ruffo, Prof. A. Abbotto

Department of Materials Science

Solar Energy Research Center MIB-SOLAR and

INSTM Research Unit, University of Milano Bicocca

Via Cozzi 53, 20125 Milano (Italy)

[d] Dr. A. Amat, Dr. F. De Angelis

Computational Laboratory for Hybrid/Organic Photovoltaics

(CLHYO)

Istituto di Scienze e Tecnologie Molecolari del CNR (CNR-ISTM)

Via Elce di Sotto 8, 06123 Perugia (Italy)

Supporting information for this article is available on the WWW

under http://dx.doi.org/10.1002/chem.201300219.

Chem. Eur. J. 2013, 19, 10723 – 10740

ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

10723

appropriate LUMO energy, are now viewed as interesting

sensitizers for use in DSSCs. In particular, meso-disubstitut-

ed push–pull Zn^IIporphyrinates, characterized by a strong

and directional electron excitation process along the push–

pull system, have been extensively investigated.^[4c]Recently,

more than 12% efficiency has been reached by an optimally

engineered meso-disubstituted push–pull Zn^IIporphyrinate

acting as a light-harvesting dye.^[5]Although some meso-di-

degree of steric hindrance that would prevent aggregation

of the dye when adsorbed on TiO₂.^[13]Specifically, we have

investigated the effects on the DSSC efficiency of the pres-

ence of one, or two adjacent, carboxylic acid groups or of a

cyanoacrylic acid group on the ethynylphenyl substituent in

the b position anchoring the dye on the TiO₂surface. The

effects on the electrochemical properties and absorption

spectra produced by the presence of different groups on the

b-ethynylphenyl moiety have also been experimentally in-

vestigated and their electronic origin has been studied by

means of DFT and TDDFT calculations.

Moreover, we have devised a new synthetic procedure, by

which several issues associated with the synthesis of b-disub-

stituted push–pull tetraaryl Zn^IIporphyrinates involving two

opposite b positions of the porphyrinic ring may be circum-

vented. Application of this procedure provides a facile syn-

thesis of such push–pull porphyrinic structures that, as far as

we are aware, have not been synthesized previously.

ACHTUNGTRENNUNG

substituted push–pull Zn^IIporphyrinates, investigated as

dyes in recent years, display interesting light-conversion effi-

ciencies,^[4c]their syntheses require multiple steps and are

characterized by very low overall yields.^[6,7]

b-Substituted tetraaryl Zn^IIporphyrinates, which incorpo-

rate a tetraaryl porphyrinic core as a starting material, are

synthetically more attractive, since this core can be easily

obtained by means of a one pot reaction between pyrrole

and the appropriate aryl aldehyde. Moreover, the attach-

ment to this core of an appropriate functionalization bearing

a carboxylic group, for instance after selective bromination

of the b-pyrrolic position, can be achieved by a one-step re-

action. Recently, a new synthesis of Zn^IIporphyrinates sub-

stituted at the b-pyrrolic position by an ethynyl moiety was

reported, based on a modified Horner–Emmons condensa-

tion, giving overall yields ranging from moderate to high.^[8]

Further advantages may be cited: firstly, the more sterically

hindered architecture of a tetraaryl porphyrinic core should

ensure a lower degree of p-stacking aggregation of the dye

when adsorbed on TiO₂, and secondly, enhanced solubility

in most common organic solvents should facilitate the purifi-

cation process, allowing the production of very pure por-

phyrinic dyes.

Finally, we have compared the light-conversion efficien-

cies, under the same fabrication conditions, of DSSCs based

on the various synthesized b-monosubstituted Zn^IIpor

ACHTUNGTRENNUNG

A

E

N

ACHTUNGTRENNUNG

AHCTUNGTRENNUNG

substituted push–pull Zn^IIporphyrinates that have previous-

ly been reported as efficient dyes for use in such cells.^[4c]

Results and Discussion

Synthesis: Partridge et al.^[8]have recently reported a modi-

fied Horner–Emmons condensation for the synthesis of sev-

eral tetraaryl Zn^IIporphyrinates bearing an ethynyl group at

the b-pyrrolic position. Their synthetic route involved a con-

densation reaction between a 2-formylporphyrin and highly

reactive halo-phosphonates to give various halo-vinyl

groups, followed by an elimination reaction to obtain an

ethynyl substituent at the b-pyrrolic position. Although this

protocol gave the products in moderate to high yields, it is

not compatible with various functional groups, such as car-

boxylic or cyanoacrylic acids. Thus, syntheses of such deriva-

tives would require additional steps, making them more ex-

pensive and laborious. Therefore, for the synthesis of a

series of tetraaryl Zn^IIporphyrinates substituted at the b-

pyrrolic position, as investigated in this work (Figure 1), we

have devised and optimized a synthetic route based on mi-

crowave-enhanced Sonogashira coupling. Traditional cou-

pling reactions between various ethynylaryl derivatives and

the tetraaryl Zn^IIporphyrinate brominated at the b-pyrrolic

position have previously been reported to proceed with rela-

tively low yields^[11,14](Scheme 1).

Surprisingly, relatively few reports have been published

on DSSCs based on b-substituted tetraaryl Zn^IIpor

N

ACHTUNGTNERrNUNG in-

ACHTUNGTRENNUNG

such systems of up to 7.1% have been reported.^[10]More-

over, to the best of our knowledge, b-disubstituted push–

pull tetraaryl Zn^IIporphyrinates have never been synthe-

sized and investigated as dyes for use in DSSCs, most proba-

bly due to synthetic difficulties.

An electron-transfer process induced by light absorption,

from the tetraaryl porphyrinic core to an ethynylaryl moiety

bearing a strong electron-acceptor group attached at the b-

pyrrolic position, was first highlighted by some of us as part

of a nonlinear optical investigation.^[11]Further evidence of

such electron transfer from the tetraaryl porphyrinic core to

the carboxylic group of an ethynylaryl moiety linked to the

b-pyrrolic position was subsequently provided by spectro-

scopic and computational studies,^[12]thus suggesting that b-

substituted porphyrinic structures of this kind may serve as

interesting light sensitizers in DSSCs. This evidence, togeth-

er with the promising light-conversion efficiencies hitherto

reported for some DSSCs involving b-substituted tetraaryl

Zn^IIporphyrinates as dyes,^[10]and the rather simple synthesis

of such systems,^[3,12]prompted us to investigate a series of b-

We therefore first optimized the procedure for bromina-

tion of the porphyrinic core, obtaining in good yields

(around 50–60%) both the monobromo (3a) and the dibro-

mo (2a) intermediates. The functionalization process in-

volves an initial complexation step of the core, furnishing in

quantitative yield the Zn^IIporphyrinate 1a, which is more

reactive, in this specific case, than the corresponding free

substituted

[tetrakis(3,5-di-tert-butylphenyl)porphyrinate]

Zn^IIspecies. It was anticipated that these would show in-

creased solubility in organic solvents and a significant

10724

www.chemeurj.org

ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Chem. Eur. J. 2013, 19, 10723 – 10740

Sensitizing Efficacies of Tetraaryl Zn^IIPorphyrinate Dyes

FULL PAPER

Figure 1. Zn^IIporphyrinates investigated in this work.

base for the following bromination reaction. This latter step

was accomplished by refluxing 1a in CCl₄and adding 1.1 or

1.8 equiv of N-bromosuccinimide in order to obtain larger

quantities of 3a or 2a, respectively. This is a quite remark-

mine atoms are attached at the b-pyrrolic positions of [tetra-

kis(3,5-di-tert-butylphenyl)porphyrinate]Zn^II. By reactions of

3a with appropriate ethynylphenyl reagents, we obtained in

good yields (40–65%) a series of b-ethynylphenyl monosub-

stituted Zn^IIporphyrinates 1, 3, and 4 (Scheme 1).

ACHTUNGTRENNUNG ACHTUNGTRENNUNG

able result, since the bromination of tetraaryl Zn^IIpor

A

R

The Sonogashira couplings were carried out by adding 3a

and the requisite ethynylphenyl reagent to a mixture of

ly only low yields have been reported owing to the forma-

tion of a series of different b-brominated products.^[15]How-

ever, it must be pointed out that we have been unable,

N,N-dimethylformamide and diethylamine (or tri

ACHTUNGTRENNUNGethACHTUNGTRENNUNGyl-

A

ACHTUNGTRENNUNG

1

purely on the basis of H NMR characterization, to discrimi-

and CuI. The reaction mixture was then subjected to micro-

wave-irradiation for 1 h at 1208C according to a previous

report on the Sonogashira coupling of several aryl-halogen-

ated substrates.^[17]In this way, we were able to attach ethyn-

yl groups at the highly crowded b-pyrrolic position of Zn^II

porphyrinates.

nate between dibromination at the 2,12- or 2,13 positions, al-

though we have clear chromatographic and ¹H NMR evi-

dence that mainly one dibrominated product was obtained

(Figure S1 in the Supporting Information). In the following,

the b-dibrominated tetraaryl Zn^IIporphyrinate 2a refers to

the more symmetrical [2,12-dibromo-5,10,15,20-tetrakis(3,5-

di-tert-butylphenyl)porphyrinate]Zn^II.

In a previous study, Yeh et al.^[14]reported the synthesis of

1 in just 30% yield by refluxing a solution of 3a and 4-eth-

To maximize the yield of the classical Sonogashira cou-

pling,^[16]a microwave-assisted synthetic approach has been

applied to circumvent the low reactivity of 3a, in which bro-

G

N

U

ACTHUNGTRENNUNG ACHUTNGRENNUaG mine for

CTHUNGTRENNUNG

Chem. Eur. J. 2013, 19, 10723 – 10740

ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemeurj.org

10725

M. Pizzotti et al.

Scheme 1. i) NBS, CCl₄, 5 h, reflux; ii) 4-ethynylbenzoic acid, [Pd

yl-N,N-dimethylaniline, [Pd(PPh₃)₄], CuI, HNEt₂/DMF (3:1), 1 h, MW (1208C); iv) dimethyl ester of 4-ethynylphthalic acid, [Pd

DMF (3:1), 1 h, MW (1208C); v) NaOH in THF/CH₃OH/H₂O (1:1:0.1) 1 h, reflux; vi) 4-ethynylbenzaldehyde, [Pd(PPh₃)₄], CuI, HNEt₂/DMF (3:1), 1 h,

ACHTUNGTRENNUNG

A

ACHTUNGTRENNUNG

AHCTUNGTRENNUNG

MW (1208C); vii) 2-cyanoacetic acid, piperidine, CH₃CN/CHCl₃, 16 h (808C); viii) 4-ethynylbenzaldehyde, 4-ethynyl-N,N-dimethylaniline, [Pd

CuI, HNEt₂/DMF (3:1), 1 h, MW (1208C).

ACHTUNGTRENNUNG(PPh₃)₄],

odology has provided access to 1 without recourse to AsPh₃,

with a significant reduction in the reaction time to 1 h and

an increased yield of up to 65%.

an easy and simple means of obtaining porphyrinic push–

pull structures.

The related meso-disubstituted push–pull Zn^IIporphyri-

nates 6, 7, and 8 were synthesized starting from [5,15-

diiodo-10,20-bis(3,5-di-tert-butylphenyl)porphyrinate]Zn^II

(6a), which was prepared as previously reported in the liter-

ature.^[7]As summarized in Scheme 2, 6 and 8 were obtained

by the above-described one-step procedure, while 7 was pre-

pared by the traditional multi-step synthesis (see Experi-

mental Section).^[7]

In conclusion, we have been able to synthesize dyes 1, 3,

and 4 in good overall yields by a simple and efficient three-

step synthetic procedure. Moreover, starting from 2a and

applying the same three-step process, we have also synthe-

sized the hitherto unknown b-disubstituted push–pull tet-

raaryl Zn^IIporphyrinates 2 and 5 (Figure 1). Compounds 2

and 5 were obtained in acceptable yields (25% and 23%,

respectively) by one pot reactions carried out by treating

solutions of 2a in N,N-dimethylformamide and diethylamine

(or triethylamine) with equimolar amounts of both 4-ethyn-

yl-N,N-dimethylaniline and 4-ethynylbenzoic acid or 4-ethy-

Energy levels and electronic distributions of molecular orbi-

tals: We carried out a computational DFT and TDDFT in-

vestigation on all of the Zn^IIporphyrinates considered in

this work in order to compare the energies and electronic

distributions of their ground- and excited-state levels. A sim-

ilar theoretical investigation was recently reported for 1 and

in particular for 4.^[12]

Figure 2 shows a schematic energy diagram with particular

emphasis on the HOMO–LUMO energy gaps for 1–8. Iso-

density plots of the electronic distributions for the first three

nylbenzaldehyde, under conditions similar to those de

ACHTUNGTRNEsNUNG crib-

ACHTUNGTRENNUNGed above for the synthesis of 1, 3, and 4 (see Experimental

Section). A similar one pot synthetic methodology has pre-

viously been reported by some of us^[7]for the synthesis of

meso-disubstituted push–pull Zn^IIporphyrinates. It has now

been extended to the synthesis of b-disubstituted push–pull

tetraaryl Zn^IIporphyrinates, confirming its applicability as

10726

www.chemeurj.org

ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Chem. Eur. J. 2013, 19, 10723 – 10740

Sensitizing Efficacies of Tetraaryl Zn^IIPorphyrinate Dyes

FULL PAPER

Scheme 2. ix) 4-Ethynylbenzoic acid, 4-ethynyl-N,N-dimethylaniline, [PdACHTNUGTRENNUNG

[PdCl₂ACHTUNGTRENNUNG(PPh₃)₂], CuI, HNEt₂, THF, 24 h, 708C; xi) 4-ethynyl-N,N-dimethylaniline, [Pd HCTUNGTRENNNUG

AHCTUNGTRENNUNG

It can be seen that the b-monosubstituted tetraaryl Zn^II

porphyrinate 4 shows significant variations in both the

HOMO and LUMO energy levels with respect to 1 and 3.

The HOMO–LUMO energy gap decreases from 2.59 and

2.57 eV for 1 and 3, respectively, to 2.33 eV for 4. This can

be attributed to a decrease in the LUMO energy level for

the latter, due to the presence of the cyanoacrylic acid

group, allowing a lower-energy HOMO–LUMO electron ex-

citation. The HOMO–LUMO energy gap appears to be un-

affected by the addition of a second adjacent carboxylic acid

group on the phenyl ring of the ethynylphenyl substituent,

as in 3 (Figure 2).

The key assumption underpinning the four-orbital Gouter-

man model is that in tetraaryl Zn^IIporphyrinates the

HOMO and HOMOꢀ1 are near-degenerate and the

LUMO and LUMO+1 are degenerate.^[18]In 1, 3, and even

4, substitution at the b position does not markedly affect the

relative energies of the two HOMOs, but the degeneracy of

the two LUMOs is broken. This is especially so for 4, for

which the difference between the LUMO and LUMO+1

energy levels is around 0.5 eV due to the lower energy and

concomitantly increased electron density located on the b-

ethynylphenyl substituent characterizing the LUMO (Fig-

ures 2 and 3). Moreover, the LUMO+2, which is far away in

Figure 2. Energy levels of the main occupied/unoccupied molecular orbi-

tals of 1 on TiO₂and of all of the Zn^IIporphyrinates investigated. The

LUMO of the dye is mainly located on the LUMO+7 and LUMO+12 of

1 on TiO₂, the contributions being 23% and 45%, respectively.

occupied HOMOs and unoccupied LUMOs are shown in

Figure 3.

Chem. Eur. J. 2013, 19, 10723 – 10740

ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemeurj.org

10727

M. Pizzotti et al.

Figure 3. Isodensity plots (cut-off 0.03), as computed by a DFT approach using THF solution, of the electronic distributions of the first occupied/unoccu-

pied orbitals of 1–8.

energy from the LUMO+1 in the Gouterman model, be-

comes closer, particularly in the case of 4, due to the signifi-

cant electron density on the b-ethynylphenyl substituent.^[12]

These deviations from the Gouterman model are much less

significant for 1 and 3 (Figures 2 and 3). It thus appears that

the presence of an electron-withdrawing group such as cya-

noacrylic acid on the substituent in the b position leads to a

significant departure from the Gouterman model due to the

stabilization of both the LUMO and LUMO+2.^[12]

Compared to the corresponding b-monosubstituted tet-

raaryl Zn^IIporphyrinates 1 and 4, the HOMO and LUMO

energy levels of the b-disubstituted push–pull tetraaryl Zn^II

porphyrinates 2 and 5 show shifts towards less negative en-

ergies of the HOMOs, leading to significant decreases in the

HOMO–LUMO energy gaps (2.30 eV for 2 versus 2.59 eV

for 1; 2.06 eV for 5 versus 2.33 eV for 4). This can be ascri-

bed to high electron densities located on the push–pull ethy-

nylphenyl substituent bearing the dimethylamino group

(Figures 2 and 3).

The presence in 5 of a cyanoacrylic acid group produces,

as in 4, a loss of the degeneracy between the LUMO and

LUMO+1 due to a decrease in the energy of the former. A

decrease in the energy of the LUMO+2 brings it closer to

the LUMO+1. It must be pointed out that the HOMOꢀ2,

which does not have a spatial analogue in the Gouterman

model, is largely substituent-based in 1, 3, and 4, but in 2

and 5 does not show a similar electron density on the “pull”

ethynylphenyl substituent bearing the carboxylic or cyano-

AHCTUNGTERGaNNUN crylic acid group (Figure 3).

10728

www.chemeurj.org

ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Chem. Eur. J. 2013, 19, 10723 – 10740

Sensitizing Efficacies of Tetraaryl Zn^IIPorphyrinate Dyes

FULL PAPER

The HOMO–LUMO energy gaps of the b-disubstituted

push–pull tetraaryl Zn^IIporphyrinates 2 and 5 were found to

be higher than those of the corresponding meso-disubstitut-

ed push–pull Zn^IIporphyrinates 6 and 8 (2.02 eV for 6

versus 2.30 eV for 2; 1.84 eV for 8 versus 2.06 eV for 5),

since the latter show both much higher HOMO energy

levels and significantly less negative LUMO energies. More-

over, the presence on the phenyl ring of the “pull” ethynyl-

phenyl substituent of a cyanoacrylic acid group as in 8, in-

stead of a carboxylic acid group as in 6, produces a decrease

in the HOMO–LUMO energy gap from 2.02 eV to 1.84 eV,

with a concomitant significant perturbation of the energy

the b-disubstituted push–pull porphyrinate 5. It also decreas-

es the energy of the LUMO+2, bringing the latter closer to

the LUMO+1, which is not affected by the presence of the

cyanoacrylic acid group, being centered on the porphyrinic

ring. Hence, the original degeneracy of the LUMO and

LUMO+1 levels of the Gouterman model is destroyed, but

a new near-degeneracy is introduced by the approach of the

LUMO+1 and LUMO+2 levels. Regarding the electronic

distributions, the presence of a “push” substituent shifts the

electron distribution of the HOMO towards the substituent,

while the presence of a cyanoacrylic acid group significantly

increases the electron distributions of the LUMO and the

LUMO+1 towards the “pull” substituent and the anchoring

group, thus suggesting more facile electron injection into the

semiconductor.

levels of

6

since the HOMOꢀ1, HOMOꢀ2 and the

LUMO+1, LUMO+2 become near-degenerate (Figure 2).

The LUMOs of 1, 3, and 4 are characterized by some

electron density located on the “pull” ethynylphenyl sub-

stituent in the b position, particularly in the case of 4 bear-

ing a cyanoacrylic acid group. In the LUMO+2 and

HOMOꢀ2, more significant electron density is located on

this “pull” substituent in the b position. As a consequence,

the HOMO!LUMO and particularly the HOMO!

LUMO+2 electron transitions are associated with electron

transfer from the porphyrinic core to the “pull” ethynyl-

phenyl substituent, as suggested previously,^[11,12]while the

HOMOꢀ2!LUMO electron transition produces, except in

4, an unexpected electron transfer from the “pull” ethynyl-

phenyl substituent to the porphyrinic core.

To gain insight into the interaction between the Zn^IIpor-

phyrinates, acting as electron-donor dyes, and the TiO₂semi-

conductor, we also performed DFT calculations on 1 adsor-

bed on the surface of an approximately 2ꢃ2 nm TiO₂clus-

ter.^[20]To the best of our knowledge, the only previous theo-

retical study modeling a porphyrinic dye/TiO₂interaction

has been that recently performed by He et al.^[21]using a very

simple semiconductor model. Our DFT investigation of this

joint system has shown that the LUMOs are mainly located

on the semiconductor. The first LUMOs to make a signifi-

cant contribution in

1 are the LUMO+7 and the

LUMO+12, contributing 23% and 45%, respectively (Fig-

ure 4(a)). The LUMO+7 of this joint system is found

0.13 eV below the LUMO of the isolated 1, not interacting

with the TiO₂, while the LUMO+12 is computed to lie at es-

sentially the same energy as that of the isolated 1.

In the b-disubstituted push–pull tetraaryl Zn^IIpor

ates 2 and 5, the electron density of the HOMO is located

on the “push” ethynylphenyl substituent bearing the dimeth-

ylamino group. While the LUMO of 2 is mainly located on

ACHUTGTNRNEUGpN hyCAHTUNGTRNErNUGN in-

ACHTUNGTRENNUNG

AHCTUNGTRENNUNG

A

the porphyrinic core with a minor contribution from the

“pull” ethynylphenyl substituent, the presence of the cya-

noacrylic acid group in 5 leads to a LUMO almost entirely

located on the substituent. The LUMO+1 of both com-

pounds is located on the porphyrinic core, while a significant

electron density in the LUMO+2 of 2 and, to a lesser extent

of 5, is located on the “pull” ethynylphenyl substituent.

The density of states (DOS) for the first 100 unoccupied

states has been described using a Gaussian convolution with

s=0.2 for the whole joint system as well as the projection

of the electronic states of 1 and of the semiconductor states

(Figure 4(b)).

As can be seen, in the low-energy range, below ꢀ3.10 eV,

no contribution from 1 to the joint DOS of the system is ap-

parent; only the unoccupied orbitals of the semiconductor

manifold contribute to the total DOS. The DOS of 1, when

linked to TiO₂, is mainly located between ꢀ3.00 and

ꢀ2.40 eV; at ꢀ2.79 eV, at which the LUMO+12 is located,

the accumulated contributions from the LUMO to the

LUMO+12 account for the entire electron density on the

LUMO of the isolated dye 1. Above the LUMO+12, there

is a series of orbitals that make small contributions to 1 up

to the LUMO+30 at ꢀ2.58 eV, which is almost entirely lo-

cated on 1 (99%). The broadening of the partial DOS for

dye 1 suggests quite strong coupling between the LUMO of

1 and the TiO₂manifold of unoccupied states, which, togeth-

er with the good alignment of the energy levels, is suggestive

of efficient electron injection from 1 to the semiconduc-

tor.^[20]

For the meso-disubstituted push–pull

6 and 7, the

LUMO+2 is characterized by a significant electron density

on the “pull” ethynylphenyl substituent, while in the

LUMO the electron density located on the “pull” ethynyl-

phenyl substituent is insignificant.

In 8, compared to 6 and 7, both the LUMO and

LUMO+1 are associated with a significant electron density

located on the “pull” ethynylphenyl substituent, due to the

presence of the cyanoacrylic acid group, as previously ob-

served for b-disubstituted push–pull tetraaryl Zn^IIpor

ACHTUNGTRENNUNG

ACHTUNGTRENNUNGrinACHTUNGTRENNUNG

In conclusion, b-mono- and push–pull disubstitution of

the porphyrinic ring produces a marked deviation from the

Gouterman simple model of the energy levels, which is most

pronounced for a push–pull structure and in the presence of

a cyanoacrylic acid group. The latter acts mainly on the

HOMO–LUMO energy gap by decreasing the energy of the

LUMO, in both the b-monosubstituted porphyrinate 4 and

Experimental and calculated electronic absorption spectra

and experimental emission spectra: Electronic absorption

spectra of b-monosubstituted tetraaryl Zn^IIporphyrinates 1,

Chem. Eur. J. 2013, 19, 10723 – 10740

ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemeurj.org

10729

M. Pizzotti et al.

510 nm) are computed, giving rise to a broad band in this

spectral region, in contrast to a single B band for 1, 2, and 3

(Table 1).

Zn^IIporphyrinates 1–3 show comparable energies of the

B bands, but the two Q bands of 2, characterized by a push–

pull architecture, are slightly red-shifted (by about 7–9 nm)

with respect to those of 1 and 3, due to a major contribution

from the HOMO!LUMO transition and to

a lower

HOMO–LUMO gap (Table 1). On the other hand, 5 shows

significant red-shifts of both the B and Q bands, due to the

push–pull structure and the presence of the electron-with-

drawing cyanoacrylic acid group, which significantly desta-

bilizes/stabilizes the HOMO/LUMO set (Figures S2 and S3;

Table 1).

If we consider the configuration interactions that give rise

to the B and Q bands, we notice that in 1 and 3 it is mainly

the HOMO, HOMOꢀ1, and LUMO+1, characterized by

close energy levels (Figure 2) and an electron density mainly

located on the porphyrinic core and LUMO, with a limited

electron density on the ethynylphenyl substituent (Figure 3),

that contribute to the pertinent transitions, in full agreement

with the “Gouterman four-orbital model”. The electronic

transition HOMOꢀ2!LUMO contributes only marginally

to the B band (Table 1). Since the HOMOꢀ2 of 1 and 3

shows a significant electron density located on the ethynyl-

phenyl substituent (Figure 3), the B band of these dyes is

apparently associated with a small degree of charge transfer

from the substituent to the porphyrinic core, which is at var-

iance with the simple “Gouterman four-orbital model”.

Whereas 1 and 3 depart only slightly from this model, a

much greater discrepancy is evident when we consider the

configuration interactions that give rise to the B and Q

bands of 4, bearing a cyanoacrylic acid group. In 4, the

HOMO, HOMOꢀ1, and LUMO+1 still show an electron

density mainly located on the porphyrinic core, but the

LUMO shows an increased electron density located on the

ethynylphenyl substituent, as do the HOMOꢀ2 and

LUMO+2 (Figure 3). In this case, the HOMOꢀ2 partici-

pates through a strong contribution of 70% to the main

transition of the B band computed at 451 nm (Table 1). In

this case, however, the involvement of the HOMOꢀ2 does

not imply an electron transfer from the “pull” ethynylphenyl

substituent to the porphyrynic core, since the LUMO is also

located on the “pull” substituent. On the other hand, the

LUMO+2, which becomes suitably close in energy to the

LUMO and LUMO+1 to be significantly involved in the

configuration interactions, is involved in the three calculated

B absorption transitions at 410 nm (23% contribution from

the transition HOMO!LUMO+2), 451 nm (21% contribu-

tion from the transition HOMOꢀ1!LUMO+2), and

509 nm (23% contribution from the transition HOMOꢀ1!

LUMO+2). Moreover, the LUMO is also involved in the B

band calculated at 451 nm (70% HOMOꢀ2!LUMO). It

also transpired that transitions involving the LUMO are

more relevant for 4 than for 1 or 3, if we consider the con-

figuration interactions giving rise to the transitions involving

Figure 4. a) Electronic distributions of the LUMO+7 and LUMO+12 of

the TiO₂system, where the LUMO of the dye 1 is mainly located, with

contributions of 23% and 45%, respectively. b) Density of unoccupied

states for 1@TiO₂; total: black line, TiO₂: dotted line; dye: dash-dot-dash

line. The inset shows an expansion of the partial density of states on 1.

3, and 4, b-disubstituted push–pull tetraaryl Zn^IIporphyri-

nates 2 and 5, and meso-disubstituted push–pull Zn^IIpor-

phyrinates 6, 7, and 8, each in THF solution, are presented

in Figures S2 and S3 in the Supporting Information. The cor-

responding optical data, both experimental and computed

by TDDFT calculations, are summarized in Table 1.

The b-mono- or disubstituted push–pull tetraaryl Zn^IIpor-

phyrinates 1–5 in THF solution show the typical absorption

spectra of porphyrinic systems, with a strong B absorption

band at about 430–460 nm and two weaker Q absorption

bands at about 566–577 and 604–616 nm, as expected for a

decrease in the porphyrin microsymmetry as considered in

the “Gouterman four-orbital simple model”.^[18]Porphyrinate

4, and more especially 5, bearing a cyanoacrylic acid group,

show a red-shifted shoulder of a broad B band (Figures S2

and S3), which may be rationalized in terms of a significant

departure from the “Gouterman four-orbital simple model”

due to the presence of an electron-withdrawing group on

the ethynylphenyl substituent, which, as previously suggest-

ed,^[11,12]is the origin of the broadening of the B band. The

computed absorption spectra are consistent with this hy-

pothesis since, for both 4 and 5, three transitions close in

energy and intensity in the region of the B band (410–

the two

Q

bands of

4

calculated at 580 nm (71%

10730

www.chemeurj.org

ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Chem. Eur. J. 2013, 19, 10723 – 10740

Sensitizing Efficacies of Tetraaryl Zn^IIPorphyrinate Dyes

FULL PAPER

Table 1. Experimental and computed electronic absorption spectra and experimental emission spectra in THF solution of the Zn^IIporphyrinates investi-

gated in this work.

Dye

Abs. B bands

Computed Abs. B bands^[a]

Abs. Q bands

Computed Abs. Q bands^[a]

Em. Bands

l_a(nm) [loge]

l_a(nm) [intensity (a.u.)]

l_a(nm) [loge]

l_a(nm) [intensity (a.u.)]

l_e(nm)

1

438 [5.37]

439 [5.18]

438 [5.49]

453 [1.18]

(40% H-1!L)

(35% H!L+1)

(19% H-2!L)

566 [4.20]

562 [0.04]

(49% H-1!L)

(47% H!L+1)

617

667

604 [3.90]

575 [0.08]

546 [0.09]

(72% H!L)

(26% H-1!L+1)

2

3

4

456 [1.85]

455 [1.88]

(40% H-2!L)

(34% H-1!L+1)

(21% H-3!L)

573 [4.15]

(33% H-2!L)

(22% H-1!L)

(22% H-1!L+1)

627

679

613 [3.94]

566 [4.31]

604 [4.02]

626 [0.76]

562 [0.04]

578 [0.09]

A

(40% H-1!L)

(36% H!L+1)

(13% H-2!L)

(50% H-1!L)

(47% H!L+1)

618

668

(74% H!L)

(24% H-1!L+1)

438 [5.30]

460 (sh)

410 [1.17]

451 [1.07]

509 [0.75]

(42% H-1!L+1)

(23% H!L+2)

(70% H-2!L)

(21% H-1!L+2)

(38% H!L+1)

(27% H-1!L)

568 [4.37]

606 [4.21]

580 [0.32]

617 [0.16]

(71% H-1!L)

(21% H!L+1)

617

669

A

(23% H-1!L+2)

5

449 [5.11]

464 (sh)

414 [1.01]

450 [0.64]

497 [1.07]

(56% H-2!L+2)

(18% H-1!L+1)

(63% H-3!L)

577 [4.36]

556 [0.12]

569 [0.13]

(57% H!L+2)

(19% H!L+1)

(36% H-2!L)

(15% H!L+1)

(15% H-1!L)

626

683

(30% H-2!L+1)

(49% H-1!L+2)

(21% H-2!L+1)

(21% H-2!L)

616 [4.25]

671 [4.69]

662 [4.61]

668 [4.58]

683 [0.74]

697 [1.37]

708 [1.43]

764 [1.69]

(94% H!L)

6

7

8

456 [5.23]

451 [5.32]

458 [5.26]

428 [1.59]

432 [1.85]

460 [1.38]

(59% H-1!L+1)

(15% H!L+2)

A

680

668

688

(64% H-1!L+1)

(12% H-2!L)

A

(44% H-1!L+2)

(44% H-2!L+1)

(98% H!L)

[a] The composition of the transitions involved in each computed absorption band is reported in parentheses.

HOMOꢀ1!LUMO) and in particular at 617 nm (90%

HOMO!LUMO) (Table 1).

on the “push” ethynylphenyl substituent bearing the di

ACTHNGUTRENNUG CAHTUNGTRENNUNG

ACHTUNGTRENNUNGmeth-

In summary, charge-transfer processes from the porphyr-

inic core to the ethynylphenyl substituent and, to a lesser

extent, in the opposite sense, may make significant contribu-

tions to both the B and Q bands. This implies that the elec-

tronic properties of the substituents on the phenyl ring of

the ethynylphenyl substituent in the b position affect both

the extent and direction of the charge-transfer associated

with the B and Q absorption bands. While for 1 and 3 both

the B and Q bands originate from electronic transitions

mainly located on the porphyrinic core, for 4 the transitions

contributing to the B and Q bands may largely involve orbi-

tals with a significant electron density located on the ethy-

nylphenyl substituent in the b position, so that the “Gouter-

man four-orbital simple model” is strongly perturbed.

In the push–pull Zn^IIporphyrinates 2 and 5, the HOMO

energy levels show, as expected, an electron density located

electron density located on the “pull” ethynylphenyl sub-

stituent, while for both the LUMO and LUMO+1 the elec-

tron density is mainly located on the porphyrinic core

(Figure 3). Since, as for 1 and 3, the electronic transition

HOMO!LUMO+2 does not contribute at all to either the

B or Q bands (Table 1), it follows that the classical charge-

transfer process along the push–pull system is not so rele-

vant. Moreover, the configuration interactions that give rise

to the B band of 2 only involve transitions located within

the porphyrinic core (Table 1 and Figure 3).

In the case of 5, however, the configuration interactions

of both the B and Q bands involve orbitals such as the

LUMO and LUMO+2, with high electron density located

on the ethynylphenyl substituent bearing the cyanoacrylic

group (Figure 3). As a consequence, 5 shows significant

push–pull character. In fact, the notable contributions from

Chem. Eur. J. 2013, 19, 10723 – 10740

ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemeurj.org

10731

M. Pizzotti et al.

transitions involving charge transfer between the “push”

and “pull” substituents, such as HOMO!LUMO (94%)

and HOMO!LUMO+2 (57%), to the calculated absorp-

tion Q bands at 683 and 556 nm, respectively, confirm a sig-

nificant charge-transfer process along the push–pull system

(Table 1). Moreover, the evidence for a significant contribu-

tion to the B absorption band of charge-transfer transitions

such as HOMOꢀ2!LUMO+2 (56% for the band calculat-

ed at 414 nm), HOMOꢀ1!LUMO+2 (49% for the band

calculated at 497 nm), and HOMOꢀ2!LUMO (21% for

the band calculated at 497 nm) supports the view that

charge-transfer processes also contribute significantly to the

calculated B bands. Therefore, the excited states associated

with the B and more particularly the Q bands should largely

correspond to the transfer of electrons to the “pull” ethynyl-

phenyl substituent in the b position bearing the anchoring

cyanoacrylic acid group, favoring electron injection into the

TiO₂.

In summary, the presence of the cyanoacrylic acid group

produces significant conjugation effects on both the energy

levels and electronic distributions of orbitals involved in

transitions that contribute to both the B and Q bands of

Zn^IIporphyrinate 4 when compared to 1 and 3, of 5 when

compared to 2, and of 8 when compared to 6 and 7. In par-

ticular, for all of these Zn^IIporphyrinates, both the B and Q

bands show increased charge-transfer character, with elec-

tron transfer to the “pull” substituent anchored on the TiO₂.

Finally, it must be pointed out that when the push–pull

system involves the 2,12- (or 2,13-) b positions of two oppo-

site pyrrolic rings, as in 2 and 5, the conjugation between

the “push” and “pull” substituents increases on going from

2 to 5, but it remains less significant than that in 6, 7, or 8,

in which the push–pull system involves the 5,15-meso posi-

tions of a 10,20-diaryl porphyrinic core (Table 1; Figures 2

and 3). Hence, for the meso-5,15- push–pull Zn^IIpor

AHCTUNGTREGaNNNU tes 6, 7, and particularly 8, the porphyrinic ring acts as an

ACHTUNGTRENNUNGphyACHTUNGTRENNUNGrin-

When we consider, for comparison, the meso push–pull

Zn^IIporphyrinates 6, 7, and 8, we observe strong red-shifts

of both the B band (by about 17, 12, and 19 nm, respective-

ly) and the single Q band (by about 58, 49, and 55 nm, re-

spectively), with respect to the corresponding b-disubstitut-

ed push–pull Zn^IIporphyrinate 2 and to a lesser extent 5

(Figure S3 and Table 1), as expected for a stronger push–

pull system.^[19,22]Although the intensity of the B band is not

too different, the single Q bands of 6–8 are markedly more

intense than the two Q bands of 2 and 5 (Figure S3 and

Table 1).

As when comparing 1 and 3, the presence in 7 of two ad-

jacent carboxylic acid groups located on the phenyl ring of

the “pull” phenylethynyl substituent produces insignificant

effects on both the B and Q bands, with only a slight in-

crease in the intensity of the B band when compared to that

of 6 (Figure S3 and Table 1). In 6–8, the HOMO is charac-

terized by a high electron density located on the “push”

ethynylphenyl substituent bearing the dimethylamino group,

as is the HOMOꢀ2 (Figure 3). Moreover, for both 6 and 7,

the electron density located on the “pull” ethynylphenyl

substituent is much larger in the LUMO+2 than in the

LUMO. Since in 6 the HOMO!LUMO+2 charge-transfer

transition contributes only marginally to the calculated B

band at 428 nm (15%) and in 7 the HOMOꢀ2!LUMO

charge-transfer transition contributes even less to the calcu-

lated B band at 432 nm (12%), it follows that, despite the

push–pull structure, in 6 and particularly in 7 the B band

mainly originates from transitions involving the porphyrinic

core.

efficient linker between the “pull” and “push” substituents,

whereas, for the b-disubstituted push–pull Zn^IIporphyrinate

2, and to a lesser extent 5, the porphyrinic ring is less effi-

cient as a linker. All of the investigated Zn^IIporphyrinates

show a significant fluorescence; the emission spectra of 1–5

in THF solution show two emission bands at around 620

and 680 nm, with the emission at higher energy always

stronger, in accordance with the emission spectra reported

for structurally similar tetraaryl Zn^IIporphyrinates substitut-

ed at the b-pyrrolic position.^[11,9a]The emission bands of 1–5

have rather similar energies (Table 1), with only small red-

shifts for 2 and 5, as expected for some push–pull character,

while 6, 7, and 8 show only one emission band, at roughly

the same energy, as previously reported for such meso push–

pull structures.^[7]

Electrochemical investigation: The electrochemical proper-

ties of the b-substituted or disubstituted push–pull tetraaryl

Zn^IIporphyrinates 1–5 and of the meso-disubstituted push–

pull Zn^IIporphyrinates 7 and 8 have been investigated by

cyclic voltammetry (CV). We also studied the dimethyl

esters of 3 and 7, designated as 3b (Scheme 1) and 7b, re-

spectively. Normalized CV patterns obtained at 0.2 Vs^ꢀ1on

a glassy carbon (GC) electrode in DMF + 0.1m TBAP (tet-

rabutylammonium perchlorate) are presented in Figures S5

(a, b) and S6 in the Supporting Information. The corre-

sponding key features are summarized in Table 2, together

with the electrochemical and DFT computed HOMO and

LUMO energies and, for the sake of comparison, the corre-

sponding HOMO!LUMO energy gaps. The electrochemi-

cal properties of the meso-disubstituted push–pull Zn^IIpor-

phyrinate 6 have already been fully investigated by some of

us,^[23]and its normalized CV pattern is included in the Sup-

porting Information and its key features in Table 2 for the

sake of comparison.

In 8, however, bearing a cyanoacrylic acid “pull” group,

the LUMO and LUMO+1, but not the LUMO+2, show

high electron density located on the “pull” ethynylphenyl

substituent (Figure 3). In this case, both the HOMO!

LUMO (98%) and HOMOꢀ2!LUMO+1 (44%) transi-

tions contribute significantly to the calculated Q and B

bands at 764 and 460 nm, respectively (Table 1), as expected

for a significant degree of charge transfer along the push–

pull system.

At least two oxidation and two reduction peaks were ob-

served for all of the studied Zn^IIporphyrinates in the poten-

tial window investigated, all of them being reversible or

quasi-reversible from both the electrochemical and chemical

10732

www.chemeurj.org

ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Chem. Eur. J. 2013, 19, 10723 – 10740

Sensitizing Efficacies of Tetraaryl Zn^IIPorphyrinate Dyes

FULL PAPER

Table 2. Key CV features of the investigated Zn^IIporphyrinates, and HOMO and LUMO energy levels derived therefrom. Oxidation peaks localized on

the dimethylamino groups are italicized. HOMO and LUMO energy levels have been calculated in THF solution by the DFT approach.

Experimental

Computed

E8_IIIc

E8_IIc

E8_Ic

E8_Ia

E8_IIa

E8_IIIa

L

H

[eV]

H-L gap

[eV]

L

[eV]

H

[eV]

H-L gap

[eV]

[V] (Fc⁺jFc) [V] (Fc⁺jFc) [V] (Fc⁺jFc) [V] (Fc⁺jFc) [V] (Fc⁺jFc) [V] (Fc⁺jFc) [eV]

1

2

3

3b

4

5

6

7

7b

8

ꢀ2.32

ꢀ2.13

ꢀ1.92

ꢀ2.14

ꢀ2.05

ꢀ2.10

ꢀ2.03

ꢀ2.07

ꢀ2.05

ꢀ1.738

ꢀ1.674

ꢀ1.752

ꢀ1.68

0.382

0.307

0.339

0.37

0.324

0.298

0.227

0.17

0.63

0.46

0.60

0.63

0.48

0.36

0.40

0.35

0.44

ꢀ3.06 ꢀ5.18 2.12

ꢀ3.13 ꢀ5.11 1.98

ꢀ3.05 ꢀ5.14 2.09

ꢀ3.12 ꢀ5.17 2.05

ꢀ3.11 ꢀ5.12 2.02

ꢀ3.15 ꢀ5.10 1.95

ꢀ3.29 ꢀ5.03 1.74

ꢀ3.24 ꢀ4.97 1.73

ꢀ3.34 ꢀ5.01 1.67

ꢀ3.30 ꢀ5.00 1.70

ꢀ2.78 ꢀ5.37 2.59

ꢀ2.83 ꢀ5.14 2.30

ꢀ2.81 ꢀ5.38 2.57

ꢀ2.23

ꢀ1.695

ꢀ1.655

ꢀ1.515

ꢀ1.56

ꢀ3.08 ꢀ5.42 2.33

ꢀ3.09 ꢀ5.15 2.06

ꢀ2.96 ꢀ4.98 2.02

ꢀ3.01 ꢀ5.00 1.99

0.51

0.47

ꢀ2.49

ꢀ2.38

ꢀ1.46

0.21

0.20

ꢀ1.89

ꢀ1.50

ꢀ3.18 ꢀ5.01 1.84

points of view. As a consequence, CV peak analysis could

be conveniently conducted in terms of formal potentials

E8’=(E_rev,aꢀE_rev,c)/2 (approaching standard potentials E8 but

neglecting activity coefficients). The experimental E8’ values

also provide the most reliable means of calculating the elec-

trochemical HOMO and LUMO energy levels and the

HOMO–LUMO energy gap (Table 2) according to the fol-

lowing equations:^[23]

Moreover, both the first oxidation and first reduction

peaks for meso push–pull Zn^IIporphyrinates, such as 6, 7,

and 8, appear at significantly milder potentials (less positive

and less negative, respectively), compared to those for 2 and

5. The difference is more remarkable for the first reduction

peaks (2 versus 6: 0.15 V; 5 versus 8: 0.14 V) than for the

first oxidation peaks (2 versus 6: 0.08 V; 5 versus 8: 0.10 V),

the latter being localized on the dimethylamino group

(Table 2).

Comparison of the CV patterns of a b-disubstituted push–

pull tetraaryl Zn^IIporphyrinate and a b-monosubstituted tet-

raaryl Zn^IIporphyrinate (2 versus 1 or 5 versus 4) shows

that in the presence of the competing dimethylamino redox

site, as in 2 and 5, implying localization of the first oxidation

on the “push” substituent, the second oxidation involving

the porphyrinic core shifts to slightly more positive poten-

tials. This is consistent with a Coulombic repulsive interac-

tion with the charge being generated in the first oxidation

step.

In the absence of the dimethylamino group, and therefore

of a push–pull system (as in 1, 3, 3b, and 4), the E8’_IaꢀE8’_Ic

gap becomes larger (Table 2), approaching (as for 3, 3b, and

4) or reaching (as for 1) the boundaries of the well-known

Kadish relationship,^[27,28]which holds for symmetrical tetra-

phenylporphyrin (TPP) and diphenylporphyrin (DPP) tem-

plates when the porphyrinic electronic core is not signifi-

cantly perturbed by the substitution:

E_LUMOðeVÞ ¼ ꢀe ꢁ ½ðE^ꢂ0_Ic=VðFc^þjFcÞ þ 4:8 VðFc^þjFc vs zeroÞꢃ

ð1Þ

E_HOMOðeVÞ ¼ ꢀe ꢁ ½ðE^ꢂ0_Ia=VðFc^þjFcÞ þ 4:8 VðFc^þjFc vs zeroÞꢃ

ð2Þ

where Fc⁺/Fc is the ferrocenium/ferrocene redox couple

adopted as a reference for intersolvental comparison of elec-

trode potentials.^[24]

The presence of a dimethylamino group, as in 2, 5, 6, 7,

7b, and 8, consistently results in a first oxidation peak at a

significantly lower positive potential (easier oxidation), with

a second oxidation peak closely following. This feature,

which is in accordance with observations made previously

by some of us^[23,25]and by other research groups^[11,26]for a

series of Zn^IIporphyrinates bearing one substituent with a

dimethylamino group, can be explained in terms of the first

oxidation peak being localized on the dimethylamino group

with the second involving oxidation of the porphyrinic core.

If we compare the Zn^IIporphyrinates having the same

push–pull system attached at the 2,12-b or the 5,15-meso po-

sitions, respectively, that is, 2 versus 6 and 5 versus 8, we

notice a significantly larger gap between the first oxidation

and first reduction peaks for 2 and 5 (0.24–0.25 V), corre-

E^ꢂ0_IaꢀE^ꢂ0¼ 2:25 ꢄ 0:15 V

ð3Þ

Ic

It is noteworthy that this larger gap is not only a result of

the positive shift of the first oxidation peak in the absence

of the dimethylamino group, but also of the negative shift of

the first reduction peak (Table 2). The second Kadish rela-

tionship, typical of the above templates, concerning the dif-

ference between first and second reduction peaks:

sponding to

a larger electrochemical HOMO–LUMO

energy gap (Table 2), as confirmed by DFT calculations

(Figure 2). This confirms that for a push–pull system attach-

ed at the 2,12-b-pyrrolic positions the HOMO–LUMO

energy gap is widened and therefore there is less efficient

charge transfer.

E^ꢂ0_IcꢀE^ꢂ0¼ 0:42 ꢄ 0:03 V

ð4Þ

IIc

also appears to be satisfied by 1, 3, and 4 and approached

by 3b (Table 2).

Chem. Eur. J. 2013, 19, 10723 – 10740

ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemeurj.org

10733

M. Pizzotti et al.

These observations are in agreement with TDDFT theo-

retical interpretation of the experimental absorption spectra,

which predicts a weak effect of monosubstitution at the b

position, particularly for 1 and 3, on the energy levels and

probably due to a detrimental slow aggregation of the por-

phyrinic dye on the TiO₂surface. Nine different electrolytes

were investigated in order to optimize the photovoltaic effi-

ciencies of the DSSCs, varying the solvent and the concen-

trations of the electrolyte components and additives. The

best reproducible efficiencies were obtained using the elec-

trolytes VT3P and VT7 (Table 3). The main difference was

electronic distributions of the orbitals involving the por

rinic core.

As regards the effects due to the nature of the “pull”

group, switching from a carboxylic acid group to a cyano-

acrylic acid group (Table 2) results in slightly less negative

ACHTUNGTNERpNUNG hy-

A

ACHTUNGTRENNUNG

Table 3. Photovoltaic characteristics of DSSCs incorporating 1–8 and of

first reduction potentials (ꢀ1.70 V for 4 versus ꢀ1.74 V for

1; ꢀ1.66 V for 5 versus ꢀ1.67 V for 2; ꢀ1.50 V for 8 versus

ꢀ1.52 V for 6). This trend may be ascribed to an improved

conjugation efficiency rather than to a higher inductive

effect, since also the first oxidation potentials are at slightly

lower values (0.32 V for 4 versus 0.38 V for 1; 0.30 V for 5

versus 0.31 V for 2; 0.20 V for 8 versus 0.23 V for 6).

a conventional DSSC containing a benchmark dye.

Dye^[a]

Electrolyte

J_sc

[mAcm^ꢀ2

V_oc

[mV]

FF

PCE

[%]

G

]

1

2

3

4

5

6

7

8

VT3P^[a]

VT7^[b]

VT3P^[a]

VT7^[b]

A6141^[c]

8.9

528

563

586

598

586

600

482

519

757

0.65

0.62

0.65

0.60

0.58

0.59

0.53

0.61

0.65

3.1

4.1

4.0

4.6

4.7

3.9

3.0

4.2

7.8

11.6

10.5

12.9

14.0

11.1

11.7

13.3

15.9

An interesting observation concerns the Zn^IIporphyri-

nates bearing two carboxylic acid groups on the phenyl ring

of the “pull” ethynylphenyl substituent in the b position. In-

triguingly, no significant difference can be perceived be-

tween the first reduction potentials of 1 (E8’_Ic=ꢀ1.74 V)

and 3 (E8’_Ic=ꢀ1.75 V), in spite of an expected effect due to

the additivity of the Hammett constants of the carboxylic

acid group that express its inductive properties.^[29]Even

more impressive is the case of the meso push–pull Zn^IIpor-

N719

[a] 0.6m 1-methyl-3-propylimidazolium iodide, 0.02m I₂, and 0.1m LiI in

3-methoxypropionitrile; [b] 0.6m 1-methyl-3-propylimidazolium iodide,

0.02m I₂, 0.1m LiI, and 0.05m 4-tert-butylpyridine in 3-methoxypropioni-

trile; [c] 0.6m N-butyl-N-methylimidazolium iodide, 0.03m I₂, 0.10m gua-

nidinium thiocyanate, and 0.5m 4-tert-butylpyridine in acetonitrile/valero-

nitrile (85:15).

phyrinates, that is,

6

(E8’_Ic=ꢀ1.52 V) and

7 (E8’_Ic=

ꢀ1.56 V), since the presence of two carboxylic acid groups

results in a slightly less favored reduction. This apparent

anomaly might be explained in terms of some interactions,

such as reciprocal hydrogen-bonding interactions, between

the two adjacent carboxyl groups, weakening their inductive

effects. This latter assumption is supported by the observa-

tion that the corresponding dimethyl esters 3b and 7b, in

which hydrogen-bonding interactions between the two adja-

cent carboxylic groups are no longer possible, display the

expected enhanced electron-attracting inductive effect, with

a significant positive shift of the first reduction potential

(ꢀ1.68 V for 3b versus ꢀ1.74 V for 1; ꢀ1.46 V for 7b versus

ꢀ1.52 V for 6).

the presence of a small amount of 4-tert-butylpyridine

(TBP) in the latter. We also investigated the effect of the

addition of an equimolar amount, with respect to the por-

phyrinic dye, of chenodeoxycholic acid (CDCA) as a deag-

gregating co-adsorbent agent on the photovoltaic efficien-

cies of the DSSCs.^[30]Only a slight increase in efficiency was

discerned.

The best photovoltaic performances for each porphyrinic

dye are listed in Table 3, in comparison to that of the bench-

mark dye N719, measured with the standard electrolyte

A6141^[31]under the same experimental conditions. The over-

all conversion efficiencies (PCE) were obtained from the

equation:

Finally, our CV investigation has confirmed the predic-

tions of the DFT calculations, that is, that the b-mono- or

disubstituted push–pull tetraaryl Zn^IIporphyrinates 1–5

comply with the energy conditions necessary for their use as

dyes in a DSSC, that is, they all have electrochemical

LUMO levels at higher energy than that of the TiO₂con-

duction band and electrochemical HOMO levels at lower

energy than that of the iodine/iodide redox couple (Fig-

ure S4 in the Supporting Information).

PCE ¼ J_scꢁ V_ocꢁ FF

ð5Þ

where J_scis the short-circuit current density, V_ocis the open-

circuit voltage, and FF is the fill factor. Figure 5(a) shows

the photocurrent–voltage curves of DSSCs based on the por-

phyrinic dyes 1, 3, and 4, while Figure 5(b) shows the curves

for dyes 2, 5, 6, 7, and 8, characterized by a push–pull

system.

Photoelectrochemical investigation: The Zn^IIporphyrinates

shown in Figure 1 were evaluated as dyes in DSSCs, pre-

pared with an opaque active 10 mm TiO₂layer on FTO as a

photoanode. The TiO₂layers on FTO were immersed for

2 h at room temperature in the dark in a solution of the por-

phyrinic dye in EtOH/THF (9:1). Longer times were tested,

but were found not to be beneficial, leading only to modest

dye coatings with inferior photovoltaic efficiencies, most

In general, the presence of a cyanoacrylic acid group on

the “pull” substituent produces a higher photovoltaic effi-

ciency due to higher photocurrent values (Table 3). More-

over, under our experimental conditions, the photovoltaic

efficiencies of the DSSCs with b-monosubstituted tetraaryl

Zn^IIporphyrinates 1, 3, and 4 acting as dyes are quite com-

parable to those based on b-disubstituted push–pull Zn^II

porphyrinates 2 and 5 (4.1 for 2 versus 3.1 for 1; 4.6 for 4

10734

www.chemeurj.org

ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Chem. Eur. J. 2013, 19, 10723 – 10740

Sensitizing Efficacies of Tetraaryl Zn^IIPorphyrinate Dyes

FULL PAPER

Figure 6. Incident monochromatic photon-to-current conversion efficien-

cies (IPCE) of DSSCs incorporating (a) 2–5, and (b) 6 and 8.

Figure 5. Current–voltage characteristics of DSSCs incorporating (a) 1, 3,

and 4, and (b) 2 and 5–8.

versus 4.7 for 5) or even slightly better than those based on

the corresponding push–pull meso-disubstituted Zn^IIpor-

phyrinates such as 6, 7, and 8 (4.0 for 3 versus 3.0 for 7; 4.6

for 4 versus 4.2 for 8). The latter push–pull porphyrinic sys-

tems have been extensively studied and are reported to pro-

duce highly efficient DSSCs.^[4c]

The well-reproducible efficiencies of DSSCs prepared by

our experimental protocol for device fabrication incorporat-

ing porphyrinic dyes with the same push–pull system linked

either through the 2,12-b-pyrrolic positions or the 5,15-meso

positions appeared to be quite consistent (Table 3). This is

despite the fact that the latter are characterized by a smaller

HOMO–LUMO energy gap (Figure 2) and should therefore

show more efficient electron transfer along the push–pull

system, and also have a lower LUMO energy level in com-

parison to push–pull systems with the same “pull” substitu-

ent (6 versus 2 and 8 versus 5).

in Figure 6(a) for b-monosubstituted 3 and 4 and b-disubsti-

tuted push–pull Zn^IIporphyrinates 2 and 5. Dyes 4 and 5,

characterized by the presence of the electron-withdrawing

cyanoacrylic acid group on the “pull” substituent in the b

position, exhibit IPCE values higher than 50–60% over a

very broad spectral range (350–650 nm). Remarkable pla-

teaus are seen between 400 and 550 nm, at levels of about

80% for 5 and 70% for 4, along with two well-separated

peaks, at values of about 75% for 5 and 60–65% for 4, cor-

responding to the two Q absorption bands at around 550

and 620 nm, respectively. The IPCE spectrum of 2 is also

quite high (above 50%) over the same very broad spectral

range, but shows a somewhat lower (60–65%) and less ex-

tended plateau at around 450 nm, and two well-separated

peaks (about 55%) corresponding to the Q absorption

bands.

The porphyrinic dye 3 shows an IPCE spectrum with

lower values and without a diffuse plateau but with a broad

peak (about 60–65%) and two well-separated peaks (about

50–55%) corresponding to the B and Q bands, respectively

(Figure 6(a)).

In order to gain a better understanding of the origin of

the above trend in the efficiencies of DSSCs based on dyes

1–8, we investigated the incident monochromatic photon-to-

current conversion efficiencies (IPCEs). These are indicated

Chem. Eur. J. 2013, 19, 10723 – 10740

ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemeurj.org

10735

M. Pizzotti et al.

The IPCE spectra of the meso-disubstituted porphyrinates

6 and 8, depicted in Figure 6(b), are quite different, showing,

as previously reported for 6,^[7]a distinct maximum at about

450 nm corresponding to the B absorption band, IPCE

values of around 50–60%, and a less intense peak (IPCE

value around 40–50%) at about 670–680 nm corresponding

to the single Q band, typical of these push–pull systems. The

two maxima are well separated, with IPCE values in the

mid-range (530–610 nm) being rather low (around 20–30%).

The porphyrinic dye 8, bearing the acceptor cyanoacrylic

acid group, nevertheless shows, as in the case of 5, higher

IPCE values.

The calculated integrated current under the IPCE curves

nicely matched the measured photocurrents^[32]when ex-

posed to a solar simulator using a black tape shading mask

(0.40 mm²) on top of the DSSCs under standard global

AM 1.5 solar conditions.^[33]

It thus appears that the difference between the IPCE

spectra of DSSCs based on b-mono or disubstituted push–

pull Zn^IIporphyrinates 2–5 and meso push–pull disubstitut-

ed Zn^IIporphyrinates 6 and 8 is quite significant. In particu-

lar, under our experimental conditions, the IPCE peak cor-

responding to the single rather intense Q absorption band

of 6 and 8 was invariably lower in intensity compared to the

two peaks corresponding to the two rather weak Q absorp-

tion bands of 2–5. Moreover, while the IPCE spectra of 6

and 8 show distinct peaks at around 450–470 nm correspond-

ing to the B absorption band, those of 2, 4, and 5 show in-

tense and rather diffuse plateaus over a wide wavelength

range (400–550 nm), particularly in the case of 5.

also in part to the Q bands, of electronic transitions centered

on the porphyrinic core, as pointed out in the section devot-

ed to the analysis of the calculated absorption spectra.

This interpretation is also supported by the higher J_sc

values of b-substituted DSSCs based on dyes 4 and 5 when

compared to those based on dyes 1, 2, and 3 (Table 3). It ap-

pears that the critical factor in determining injection effi-

ciency in b-substituted DSSCs is the chemical nature of the

anchoring group, with the cyanoacrylic acid group exhibiting

a better performance than the carboxylic group of a benzoic

acid moiety, confirming the findings of a recent report.^[4b]

The relevance of the injection efficiency in determining

the absorbed photon-to-current conversion in the investigat-

ed DSSCs is confirmed by the increased IPCE and J_scvalues

and by the performance of the DSSCs when the “pull” sub-

stituent in the b position bears a cyanoacrylic acid group as

in 4 and 5 (Figure 6(a) and Table 3). This is despite a con-

comitant shift towards lower energies of the excited states

(Figure 2), which should cause smaller overlap of the

LUMOs with the acceptor states of TiO₂.^[4b]

In order to confirm a limited difference between b-mono-

and disubstituted Zn^IIporphyrinates when acting as dyes in

DSSCs, we have investigated the electrochemical properties

of the best performing porphyrinic dyes 4 and 5 by means of

electrochemical impedance spectroscopy (EIS), comparing

the data for the b-substituted porphyrinates with those for

the corresponding push–pull meso-disubstituted por

ACHTUNGTRENNUNGphyACHTUNGTRENNUNGrin-

AHCTUNGTREGaNNNU te 8. The impedance spectra were analyzed in terms of Ny-

quist plots (Figure 7), in which the imaginary part of the im-

Similar IPCE spectra have been reported for DSSCs

based on structurally related porphyrinic dyes such as b-

monosubstituted tetraaryl Zn^IIporphyrinates with an ethe-

nylphenyl substituent instead of an ethynylphenyl substitu-

ent. The presence of a cyanoacrylic acid group on the “pull”

substituent, as in 4 and 5, produces IPCE spectra with

values higher than 50% over a broad wavelength range

(400–680 nm), with a plateau of about 75% between 430

and 520 nm and two peaks with values around 60–70% cor-

responding to the two Q bands.^[9a]

Assuming that the efficiency of light harvesting is close to

unity, as reported for many porphyrinic dyes,^[4b]this evi-

dence would suggest more efficient electron injection into

the TiO₂manifold of unoccupied states when the porphyrin-

ic dye is substituted at the b position, as indicated by previ-

ous results.^[4a]

Figure 7. Electrochemical impedance spectroscopy (EIS) for DSSCs

based on porphyrinic dyes 4, 5, and 8. Continuous lines represent the

data fitting by using the equivalent circuit shown in the inset.

It is noteworthy that the high IPCE values shown by 4

and 5 are related to the B and Q absorption bands, charac-

terized by a significant contribution from charge-transfer

transitions involving, in the excited state, transfer of elec-

trons to the “pull” ethynylphenyl substituent bearing the cy-

anoacrylic acid group, an excitation process that should fa-

cilitate charge-transfer injection to the TiO₂surface. In fact,

when the cyanoacrylic acid group is absent, as in 2 and 3,

while the absorption spectra are rather similar to those of 4

and 5, the IPCE spectra show lower values. For these dyes,

this may be ascribed to the contribution to the B band, and

pedance is plotted as a function of the real part over the fre-

quency range. The equivalent circuit model describing the

cell impedance is shown in the inset in Figure 7. Under soft

illumination and open-circuit voltage conditions, the recom-

bination resistance (R_rec), the chemical capacitance (C_m), and

the apparent electron lifetime t_n(t_n=R_recC_m) were calculat-

ed.^[33]The DSSCs based on 4 and 5 showed similar chemical

capacitances, which reflect charge-carrier accumulation on

the TiO₂film and the density of states in the band gap.

However, 5 exhibited a slightly higher recombination resist-

10736

www.chemeurj.org

ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Chem. Eur. J. 2013, 19, 10723 – 10740

Sensitizing Efficacies of Tetraaryl Zn^IIPorphyrinate Dyes

FULL PAPER

ance, which implies an increased electron lifetime and, thus,

a more effective suppression of the dark current (Table 4).

The meso derivative 8 showed the highest recombination re-

sistance, that is, lower recombination losses and dark cur-

rent, but also the lowest chemical capacitance. The net

result was a shorter electron lifetime than in the b-disubsti-

tuted Zn^IIporphyrinates.

suggested by the previously reported higher performances

of DSSCs based on N719^[3]or on the meso-disubstituted

push–pull porphyrinic dye 6.^[4c]

The higher efficiency of DSSCs based on dyes with a

push–pull system attached at the 2,12-b positions of the por-

phyrinic ring is a rather unexpected result, since electron

transfer by excitation along the push–pull system is more ef-

ficient when it occurs through the 5,15-meso positions, thus

suggesting a significant influence of the positioning of the

anchoring group.

Table 4. Parameters calculated from EIS data plots of DSSCs incorporat-

ing 4, 5, and 8.

Moreover, as a general feature, the presence of a cyano-

AHCTUNGTREGaNNUN crylic acid group on the “pull” ethynylphenyl substituent of

Dye

R_rec[Wcm²]

C_m[Fcm^ꢀ2

]

t_n[ms]

4

5

8

14.0

18.2

22.8

1.3ꢃ10^ꢀ3

1.2ꢃ10^ꢀ3

0.8ꢃ10^ꢀ3

18.2

21.8

18.2

all of the porphyrinic dyes investigated was found to pro-

duce superior DSSC performances, probably due to more ef-

ficient electron injection into the TiO₂, since the cyanoacryl-

ic acid group favors the localization of excited electrons on

the “pull” ethynylphenyl substituent bearing the anchoring

group.

Conclusions

Interestingly, the IPCE spectra of the b-mono- or disubsti-

tuted porphyrinic dyes are more intense, particularly in the

presence of a cyanoacrylic acid group, over a broad wave-

length range (350–650 nm), while those of the meso-disubsti-

tuted push–pull porphyrinic dyes are less intense and show

the limitation of two well-separated and less intense peaks,

corresponding to the B and the single Q bands, with weak

IPCE values in the mid-range (530–610 nm) of only 20–

30%.

In fact, DSSCs based on dye 4, and more especially dye 5,

show an IPCE spectrum with broad and intense coverage

over the wavelength range 350–650 nm, similar to that of

dyes based on ruthenium complexes such as N719.^[3]In

order to obtain such a broad and intense coverage with por-

phyrinic dyes substituted at the meso positions, the addition

of a co-sensitizer^[5]or the introduction of two thienylene-vi-

nylene units to bridge the porphyrinic ring and the anchor-

ing cyanoacrylic acid group^[34]is necessary.

A comparative investigation has been carried out on b-

monosubstituted or disubstituted push–pull tetraaryl Zn^II

porphyrinates and meso-disubstituted push–pull Zn^IIpor-

phyrinates as sensitizers for use in DSSCs. The aim was to

identify possible differences in their photon-to-electron con-

version efficiencies and to compare their performances in

DSSCs fabricated according to a standard protocol.

In order to facilitate such a comparison, we have devised

a new facile synthesis of b-disubstituted push–pull Zn^IIpor-

phyrinates involving two opposite b-pyrrolic positions of the

porphyrinic ring, a push–pull porphyrinic system that, to the

best of our knowledge, has not hitherto been synthesized.

Through comparative spectroscopic, electrochemical, and

theoretical DFT investigations, we have obtained evidence

that the HOMO–LUMO energy gap is lower in the case of

the meso-disubstituted push–pull porphyrinic dyes, not only

when compared to the b-monosubstituted but also the

b-di

sub

E

In summary, the increased light-harvesting properties aris-

ing from b substitution of the porphyrinic ring cannot be as-

cribed to more intense B and particularly Q bands, but

probably to more facile charge injection into the TiO₂sur-

face, consistent with the DFT electron distribution of the b-

monosubstituted porphyrinic dye 1 interacting with the TiO₂

surface. Finally, it is worth noting the significant effects on

the HOMO–LUMO energy gap, the IPCE spectra, and the

performance of the DSSCs exerted by the presence of a cya-

noacrylic acid group on the “pull” ethynylphenyl substituent

for all of the dyes investigated, confirming the findings of

other research groups.^[4b,9b]In this work, we have investigat-

ed in more detail the electronic origin of these effects, as

was partially carried out in a recent theoretical study.^[12]

Our electrochemical and TDDFT investigations have

clearly shown that the positive effect produced by the cya-

noacrylic acid group is due more to increased conjugation of

the “pull” ethynylphenyl substituent with the porphyrinic

ring than to an inductive effect. This increased conjugation

lowers the energy levels of the LUMO and LUMO+2

(Figure 2), and consequently, upon excitation, increases the

that the porphyrinic ring is more effective as a linker be-

tween the “push” and “pull” ethynylphenyl substituents

when the push–pull system involves the 5,15-meso positions

(dyes 6, 7, and 8) as opposed to the 2,12-b-pyrrolic positions

(dyes 2 and 5) of the porphyrinic ring. A more efficient elec-

tron transfer by excitation along the push–pull system to-

wards the anchoring carboxylic acid group for dyes 6, 7, and

8 was confirmed by the higher intensity and lower energy of

their single Q bands, which originate mainly from the

HOMO–LUMO transition, as compared to the rather weak

Q bands at lower energy of dyes 2 and 5, which also mainly

originate from the HOMO–LUMO transition.

The performances of DSSCs based on b-disubstituted

push–pull dyes 2 and 5 were found to be only slightly superi-

or to those of DSSCs based on b-monosubstituted dyes 1

and 4, respectively. Unexpectedly, however, they were also

superior to those of DSSCs based on structurally related

meso-disubstituted push–pull dyes 6 and 8. The DSSCs were

fabricated according to a standard protocol in order to allow

a valid comparison. Our protocol is probably not optimal, as

Chem. Eur. J. 2013, 19, 10723 – 10740

ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemeurj.org

10737

M. Pizzotti et al.

5 mV, modulation amplitude: 50 mV) were also recorded for each com-

pound as both oxidative and reductive scans. The working electrode was

made of glassy carbon (AMEL, diameter 1.5 mm) cleaned with synthetic

diamond powder (Aldrich; diameter 1 mm) on a wet cloth (STRUERS

DP-NAP); the counter electrode was a platinum disk or wire. An aque-

ous saturated calomel electrode served as the reference, but the poten-

tials were ultimately referred to the Fc⁺/Fc (ferrocenium/ferrocene)

couple (the intersolvental redox potential reference currently recom-

mended by IUPAC)^[25,38]by both external and internal standardization.

To prevent leakage of water and chloride into the working solution, a

compartment filled with the operating medium and closed with a porous

frit was interposed between the reference electrode and the cell.

electron density in the proximity of the anchoring carboxylic

group, as confirmed by the IPCE spectra.

TDDFT calculations have also shown that such an elec-

tron transfer involves, upon excitation, not only the Q ab-

sorption bands at lower energy but also the B band, con-

firming the suggestion previously made by some of us to

justify the significant photoconversion of the B absorption

band of the push–pull porphyrinic dye 6.^[7]The lower signifi-

cance of the inductive process is supported by the lack of

effect of introducing a second carboxylic acid group on the

“pull” ethynylphenyl substituent on the electrochemical,

electronic, and photoelectrochemical properties of the inves-

tigated porphyrinic dyes.

Finally, our investigation has shown that b-monosubstitut-

ed or disubstituted push–pull Zn^IIporphyrinates, obtained

by our simple and efficient three-step synthesis, are capable

of serving as effective dyes in DSSCs. Indeed, the DSSC

performances are comparable to those obtained using meso

push–pull Zn^IIporphyrinates, which have previously been

reported to be the most promising porphyrinic dyes,^[4c]but

which can only be obtained by multi-step syntheses in very

low overall yields. Therefore, for technological applications,

this work provides an incentive to focus further attention on

this promising class of green and stable b-substituted por-

phyrinic dyes.

Preparation of DSSCs: DSSCs were prepared by adapting a procedure

reported in the literature.^[31]FTO glass plates (Solaronix 7 Wsq^ꢀ1) were

cleaned in a detergent solution and in EtOH, and then rinsed with pure

water and EtOH. FTO plates were treated with a 40 mm aqueous solu-

tion of TiCl₄for 30 min at 708C and then rinsed with water and EtOH.

An active monolayer of 10 mm was screen-printed using a 400 nm nano-

particle paste (Dyesol 18NR-AO); the coated films were dried at 1258C

for 6 min and then thermally treated under an air flow at 3258C for

10 min, 4508C for 15 min, and 5008C for 15 min. The sintered layer was

again treated with 40 mm aqueous TiCl₄at 708C, rinsed with EtOH, and

heated at 5008C. After cooling to 808C, the TiO₂-coated plate was im-

mersed in a 0.2 mm solution of the dye in EtOH/THF (9:1) containing

chenodeoxycholic acid (CDCA, 0.2 mm) for 2 h at room temperature in

the dark. In the case of the dye N719, the TiO₂-coated plate was im-

mersed in a 0.5 mm solution of the dye in EtOH for 20 h. The thickness

of the layer was measured by means of a VEECO Dektak 8 Stylus Profil-

er. The counter electrode was prepared according to the following proce-

dure. A 1 mm hole was made in an FTO plate using diamond drill bits.

The electrode was then sequentially cleaned with a detergent solution,

10% HCl, and acetone. After heating at 4008C, the electrode was cooled

and a drop of a 5ꢃ10^ꢀ3m solution of H₂PtCl₆in EtOH was added. The

thermal treatment at 4008C was then repeated. The TiO₂electrode with

adsorbed dye and Pt counter electrode were then assembled in a sealed

Experimental Section

sandwich-type cell by heating with

a hot-melt ionomer-class resin

General: All reagents and solvents used in the syntheses were purchased

from Sigma–Aldrich and used as received, except Et₃N, Et₂NH (freshly

distilled over KOH), THF (freshly distilled from Na/benzophenone

under nitrogen atmosphere), and 4-ethynylphthalic anhydride (purchased

from ABCR). Silica gels for gravimetric chromatography (Geduran Si 60,

63–200 mm) and for flash chromatography (Geduran Si 60, 40–63 mm)

were purchased from Merck. Glassware was flame-dried under vacuum

before use when necessary. [5,15-Diiodo-10,20-bis(3,5-di-tert-butylphe-

nyl)porphyrinate]Zn^II[7](6a), [5-(4’-carboxyphenylethynyl)-15-(4’’-N,N-di-

methylaminophenylethynyl)-10,20-bis(3,5-di-tert-butylphenyl)porphyrina-

te]Zn^II[7](6), 4-ethynylbenzoic acid,^[35]and 4-ethynylbenzaldehyde^[36]were

prepared as reported in the literature. Microwave-assisted reactions were

(Surlyn, 25 mm thickness) as a spacer between the electrodes. The elec-

trolyte solution was introduced through the hole by vacuum backfilling

and the hole was sealed with Surlyn. A reflective foil was taped to the

rear side of the counter electrode to reflect unabsorbed light back to the

photoanode.

Photoelectrochemical measurements: Photovoltaic J–V curves were ob-

tained using a 500 W xenon light source (ABET Technologies Sun 2000

class ABA Solar Simulator). The power of the simulated light was cali-

brated to AM 1.5 (100 mWcm^ꢀ2) using a reference Si cell photodiode

equipped with an IR cut-off filter (KG-5, Schott) to reduce the mismatch

in the region 350–750 nm between the simulated light and the AM 1.5

spectrum. Measurements were made after 3 h and after 1, 3, 5, and

7 days of ageing in the dark. J–V curves were obtained by applying an ex-

ternal bias to the cell and measuring the generated photocurrent with a

Keithley model 2400 digital source meter. Incident photon-to-current

conversion efficiency (IPCE) spectra were recorded as a function of exci-

tation wavelength in AC mode with a chopping frequency of 1 Hz and a

bias of white light (0.3 sun).

1

performed using a Milestone MicroSYNTH instrument. H NMR spectra

were recorded on a Bruker Avance DRX-400 or a Bruker AMX 300

spectrometer from solutions in CDCl₃containing, when necessary, a drop

of [D₅]pyridine, or from solutions in [D₈]THF (Cambridge Isotope Labo-

ratories, Inc.). Mass spectra were obtained on a VG Autospec M246 mass

spectrometer with an LSIMS ion source. Electronic absorption spectra

were recorded on a JASCO V-530 spectrometer; emission spectra were

Electrochemical impedance spectroscopy (EIS): Impedance spectra were

obtained using an EG&G PARSTAT 2263 galvanostat potentiostat. The

measurements were performed in the frequency range from 100 kHz to

0.1 Hz using an ac stimulus of 10 mV under 250 Wm^ꢀ2(0.25 sun) illumi-

nation under open-circuit conditions (no applied bias voltage).

recorded on

a Jobin–Yvon Fluorolog-3 spectrometer equipped with

double monochromators and a Hamamatsu-928 photomultiplier tube

(PMT) as a detector. Details of the syntheses are reported in the Sup-

porting Information.

Electrochemical measurements: Voltammetric studies were performed in

a 4 cm³cell, using 5ꢃ10^ꢀ4–1ꢃ10^ꢀ3m solutions in dimethylformamide (Al-

drich, 99.8%) with 0.1m tetrabutylammonium perchlorate (TBAP;

Fluka) as the supporting electrolyte. The solutions were deaerated by N₂

bubbling. The ohmic drop was compensated for by the positive feedback

technique.^[37]The experiments were carried out using an AUTOLAB

PGSTAT potentiostat (EcoChemie, The Netherlands) run by a PC with

GPES software. Cyclic voltammetry (CV) investigations were carried out

at scan rates typically ranging from 0.05 to 2 Vs^ꢀ1, with ohmic drop com-

pensation; differential pulse voltammetry (DPV) curves (step potential:

Theoretical DFT and TDDFT calculations: DFT and TDDFT calcula-

tions were performed on all of the investigated Zn^IIporphyrinates using

Gaussian 09.^[39]All of the structures were freely optimized in vacuo using

B3LYP^[40]and the 6-311G* basis set.^[41]Single point calculations, includ-

ing the solvent effects (THF) by means of the CPCM^[42]conductor-like

solvation model, were performed on the optimized structures in order to

compute the energies and electronic distributions of the first three occu-

pied/unoccupied molecular orbitals. The absorption spectra were simulat-

ed in THF solution by computing the first singlet-singlet excitations; the

10738

www.chemeurj.org

ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Chem. Eur. J. 2013, 19, 10723 – 10740

Sensitizing Efficacies of Tetraaryl Zn^IIPorphyrinate Dyes

FULL PAPER

computed spectra were analyzed in terms of the molecular orbital com-

positions of the main B and Q transitions.

[12] J. C. Earles, K. C. Gordon, A. W. I. Stephenson, A. C. Partridge,

D. L. Officer, Phys. Chem. Chem. Phys. 2011 , 13 , 159 7 –1 605.

[13] T. M. R. Viseu, G. Hungerford, M. I. C. Ferreira, J. Phys. Chem. B

2002 , 106 , 185 3 –1 861.

[14] C.-W. Lee, H.-P. Lu, C.-M. Lan, Y.-L. Huang, Y.-R. Liang, W.-N.

Yen, Y.-C. Liu, Y.-S. Lin, E. W. G. Diau, C.-Y. Yeh, Chem. Eur. J.

2009 , 15 , 140 3 –1 412.

[15] a) H. J. Callot, Bull. Soc. Chim. Fr. 1974, 1492; b) H. J. Callot, Tetra-

hedron Lett. 1973 , 14 , 4987.

[16] K. Sonogashira, Y. Tohda, N. Hagihara, Tetrahedron Lett. 1975 , 16 ,

446 7 –4 470.

To gain insight into the interactions between the Zn^IIporphyrinates and

the semiconductor, the adsorption of porphyrin 1 on a TiO₂surface was

also investigated. Following the strategy devised by some of us in a previ-

ous investigation,^[43]a cluster of 1@

ACTHNUTRGEN(UNG TiO₂)82 was optimized in the gas

phase with the ADF program package^[44]employing the BPW91 ex-

change-correlation functional. To take into account the solvent effect, we

carried out a single point calculation in THF at the same B3LYP/6-

311G* level of calculation as used for the isolated dyes.

[17] M. Erdꢅlyi, A. Gogoll, J. Org. Chem. 2001 , 66 , 416 5 –4 169.

[18] M. Gouterman, G. Wagniere, L. C. Snyder, J. Mol. Spectrosc. 1963 ,

11 , 10 8 –1 27.

[19] F. De Angelis, S. Fantacci, A. Sgamellotti, M. Pizzotti, F. Tessore, A.

Orbelli Biroli, Chem. Phys. Lett. 2007 , 447 , 1 0 –1 5.

Acknowledgements

[20] a) C. Anselmi, E. Mosconi, M. Pastore, E. Ronca, F. De Angelis,

Phys. Chem. Chem. Phys. 2012 , 14 , 1596 3 –1 5974; b) M. Pastore, F.

De Angelis, Phys. Chem. Chem. Phys. 2012 , 14 , 92 0 –9 28.

[21] H. He, A. Gurung, L. Si, A. Sykes, Chem. Commun. 2012 , 48 , 761 9 –

7621.

[22] a) R. Kumble, S. Palese, V. S.-Y. Lin, M. J. Therien, R. M. Hoch-

strasser, J. Am. Chem. Soc. 1998 , 120 , 1148 9 –1 1498; b) V. S.-Y. Lin,

M. J. Therien, Chem. Eur. J. 1995 , 1 , 64 5 –6 51; c) V. S.-Y. Lin, S. G.

DiMagno, M. J. Therien, Science 1994 , 264 , 110 5 –1 111.

[23] P. R. Mussini, A. Orbelli Biroli, F. Tessore, M. Pizzotti, C. Biaggi, G.

Di Carlo, M. G. Lobello, F. De Angelis, Electrochim. Acta 2012 , 85 ,

50 9 –5 23.

[24] G. Gritzner, J. Kuta, Pure Appl. Chem. 1984 , 56 , 46 1 –4 66.

[25] T. Morotti, M. Pizzotti, R. Ugo, S. Quici, M. Bruschi, P. Mussini, S.

Righetto, Eur. J. Inorg. Chem. 2006 , 174 3 –1 757.

[26] S. M. LeCours, S. G. DiMagno, M. J. Therien, J. Am. Chem. Soc.

1996 , 118 , 1185 4 –1 1864.

M.P. and G.D.C. thank the Fondazione Tronchetti Provera for financial

support and Prof. Renato Ugo for useful discussions. A.O.B., F.D.A., and

A.A. are grateful to EU Research Projects for financial support through

FP7-NMP-2009 project 246124 “SANS”. We thank Dr. Davide Motta for

his valued collaboration in carrying out the electrochemical measure-

ments.

[1] a) A. Hagfeldt, G. Boschloo, L. Sun, L. Kloo, H. Pettersson, Chem.

Rev. 2010 , 110 , 659 5 –6 663; b) A. Hagfeldt, M. Grꢄtzel, Acc. Chem.

Res. 2000 , 33 , 26 9 –2 77; c) M. Grꢄtzel, J. Photochem. Photobiol. C

2003 , 4 , 14 5 –1 53.

[2] a) B. OꢀRegan, M. Gratzel, Nature 1991, 353, 737–740; b) M. K. Na-

zeeruddin, A. Kay, I. Rodicio, R. Humphry-Baker, E. Muller, P.

Liska, N. Vlachopoulos, M. Gratzel, J. Am. Chem. Soc. 1993 , 115 ,

638 2 –6 390.

[27] K. M. Kadish, E. V. Caemelbecke, J. Solid State Electrochem. 2003,

7, 254–258.

[28] M. Zerner, M. Gouterman, Theor. Chim. Acta 1966 , 4 , 4 4 –6 3.

[29] C. Hansch, A. Leo, R. W. Taft, Chem. Rev. 1991 , 91 , 16 5 –1 95.

[30] G. Boschloo, L. Hꢄggman, A. Hagfeldt, J. Phys. Chem. B 2006 , 110 ,

1314 4 –1 3150.

[3] M. K. Nazeeruddin, F. De Angelis, S. Fantacci, A. Selloni, G. Viscar-

di, P. Liska, S. Ito, B. Takeru, M. Grꢄtzel, J. Am. Chem. Soc. 2005 ,

127 , 1683 5 –1 6847.

[4] a) W. M. Campbell, A. K. Burrell, D. L. Officer, K. W. Jolley, Coord.

Chem. Rev. 2004 , 248 , 136 3 –1 379; b) M. J. Griffith, K. Sunahara, P.

Wagner, K. Wagner, G. G. Wallace, D. L. Officer, A. Furube, R.

Katoh, S. Mori, A. J. Mozer, Chem. Commun. 2012 , 48 , 414 5 –4 162;

c) L. L. Li, E. W. G. Diau, Chem. Soc. Rev. 2013 , 42 , 29 1 –3 04.

[5] A. Yella, H. W. Lee, H. N. Tsao, C. Yi, A. K. Chandiran, M. K. Na-

zeeruddin„ E. W. G. Diau, C. Y. Yeh, S. M. Zakeeruddin, M. Grꢄt-

zel, Science 2011 , 334 , 62 9 –6 34.

[6] a) C. F. Lo, S. J. Hsu, C. L. Wang, Y. H. Cheng, H. P. Lu, E. W. G.

Diau, C. Y. Lin, J. Phys. Chem. C 2010 , 114 , 1201 8 –1 2023; b) T.

Bessho, S. M. Zakeeruddin, C. Y. Yeh, E. W. G. Diau, M. Grꢄtzel,

Angew. Chem. 2010 , 122 , 679 6 –6 799; Angew. Chem. Int. Ed. 2010 ,

49 , 664 6 –6 649; c) Y. C. Chang, C. L. Wang, T. Y. Pan, S. H. Hong,

C. M. Lan, H. H. Kuo, C. F. Lo, H. Y. Hsu, C. Y. Lin, E. W. G. Diau,

Chem. Commun. 2011 , 47 , 891 0 –8 912; d) C. L. Wang, C. M. Lan,

S. H. Hong, Y. F. Wang, T. Y. Pan, C. W. Chang, H. H. Kuo, M. Y.

Kuo, E. W. G. Diau, C. Y. Lin, Energy Environ. Sci. 2012 , 5 , 693 3 –

6940.

[7] A. Orbelli Biroli, F. Tessore, M. Pizzotti, C. Biaggi, R. Ugo, S. Cara-

mori, A. Aliprandi, C. A. Bignozzi, F. De Angelis, G. Giorgi, E. Li-

candro, E. Longhi, J. Phys. Chem. C 2011 , 115 , 2317 0 –2 3182.

[8] A. W. I. Stephenson, P. Wagner, A. C. Partridge, K. W. Jolley, V. V.

Filichev, D. L. Officer, Tetrahedron Lett. 2008 , 49 , 563 2 –5 635.

[9] a) Q. Wang, W. M. Campbell, E. I. Bonfantani, K. W. Jolley, D. L.

Officer, P. J. Walsh, K. Gordon, R. Humphry-Baker, M. K. Nazeer-

uddin, M. Gratzel, J. Phys. Chem. B 2005 , 109 , 1539 7 –1 5409; b) S. J.

Lind, K. C. Gordon, S. Gambhir, D. L. Officer, Phys. Chem. Chem.

Phys. 2009 , 11 , 559 8 –5 607.

[10] W. M. Campbell, K. W. Jolley, P. Wagner, K. Wagner, P. J. Walsh,

K. C. Gordon, L. Schmidt-Mende, M. K. Nazeeruddin, Q. Wang, M.

Graetzel, D. L. Officer, J. Phys. Chem. C 2007 , 111 , 1176 0 –1 1762.

[11] E. Annoni, M. Pizzotti, R. Ugo, S. Quici, T. Morotti, M. Bruschi, P.

Mussini, Eur. J. Inorg. Chem. 2005 , 385 7 –3 874.

[31] a) X. Jiang, T. Marinado, E. Gabrielsson, D. P. Hagberg, L. Sun, A.

Hagfeldt, J. Phys. Chem. C 2010 , 114 , 279 9 –2 805; b) J.-H. Yum, S.-

R. Jang, R. Humphry-Baker, M. Grꢄtzel, J.-J. Cid, T. Torres, M. K.

Nazeeruddin, Langmuir 2008 , 24 , 563 6 –5 640.

[32] 2: J_sc=10.6 mAcm^ꢀ2

8.9 mAcm^ꢀ2

_oc=584 mV,

11.4 mAcm^ꢀ2, V_oc=598 mV, FF=0.61, h=4.2%.

,

V_oc=566 mV, FF=0.62, h=3.8%; 3: J_sc=

,

V

FF=0.65, h=3.4%; 4: J_sc=

[33] S. Ito, M. K. Nazeeruddin, P. Liska, P. Comte, R. Charvet, P. Pꢅchy,

M. Jirousek, A. Kay, S. M. Zakeeruddin, M. Grꢄtzel, Prog. Photo-

voltaics 2006 , 14 , 58 9 –6 01.

[34] E. M. Barea, R. Caballero, L. Lopez-Arroyo, A. Guerrero, P. de La

Cruz, F. Langa, J. Bisquert, ChemPhysChem 2011 , 12 , 96 1 –9 65.

[35] E. Yashima, T. Matsushima, Y. Okamoto, J. Am. Chem. Soc. 1997 ,

119 , 634 5 –6 359.

[36] A. Lembo, P. Tagliatesta, D. M. Guldi, J. Phys. Chem. A 2006 , 110 ,

1142 4 –1 1434.

[37] A. J. Bard, L. R. Faulkner, Electrochemical Methods: Fundamentals

and Applications, Wiley, New York, 2002, pp. 648–650.

[38] G. Gritzner, Pure Appl. Chem. 1990 , 62 , 183 9 –1 858.

[39] Gaussian 09 (RevisionA.02), M. J. Frisch, G. W. Trucks, H. B. Schle-

gel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V.

Barone, B. Mennucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X.

Li, H. P. Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Son-

nenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa,

M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven,

J. A. Montgomery, Jr., J. E. Peralta, F. Ogliaro, M. Bearpark, J. J.

Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov, R. Kobayashi, J.

Normand, K. Raghavachari, A. Rendell, J. C. Burant, S. S. Iyengar,

J. Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene, J. E. Knox,

J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E.

Chem. Eur. J. 2013, 19, 10723 – 10740

ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemeurj.org

10739

M. Pizzotti et al.

Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W.

Ochterski, R. L. Martin, K. Morokuma, V. G. Zakrzewski, G. A.

Voth, P. Salvador, J. J. Dannenberg, S. Dapprich, A. D. Daniels, ꢆ.

Farkas, J. B. Foresman, J. V. Ortiz, J. Cioslowski, D. J. Fox, Gaussian,

Inc., Wallingford, CT, 2009.

deau, N. E. Davis, R. C. Binning, Jr., L. Radom, J. Chem. Phys.

1995 , 103 , 610 4 –6 113.

[42] M. Cossi, N. Rega, G. Scalmani, V. Barone, J. Comput. Chem. 2003 ,

24 , 66 9 –6 81.

[43] F. De Angelis, S. Fantacci, E. Mosconi, M. K. Nazeeruddin, M. Grꢄt-

zel, J. Phys. Chem. C 2011 , 115 , 882 5 –8 831.

[44] G. te Velde, F. M. Bickelhaupt, E. J. Baerends, C. Fonseca Guerra,

S. J. A. van Gisbergen, J. G. Snijders, T. Ziegler, J. Comput. Chem.

2001 , 22 , 93 1 –9 67.

[40] a) A. D. Becke, J. Chem. Phys. 1993 , 98 , 564 8 –5 652; b) C. Lee, W.

Yang, R. G. Parr, Phys. Rev. B 1988 , 37 , 78 5 –7 89; c) B. Miehlich, A.

Savin, H. Stoll, H. Preuss, Chem. Phys. Lett. 1989 , 157 , 20 0 –2 06.

[41] a) A. D. McLean, G. S. Chandler, J. Chem. Phys. 1980 , 72 , 563 9 –

5648; b) R. Krishnan, J. S. Binkley, R. Seeger, J. A. Pople, J. Chem.

Phys. 1980 , 72 , 65 0 –6 54; c) L. A. Curtiss, M. P. McGrath, J.-P. Blau-

Received: January 21, 2013

Revised: May 3, 2013

Published online: June 21, 2013

10740

www.chemeurj.org

ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Chem. Eur. J. 2013, 19, 10723 – 10740

Article Doi

Tetraaryl ZnII porphyrinates substituted at β-pyrrolic positions as sensitizers in dye-sensitized solar cells: A comparison with meso-disubstituted push-pull ZnII porphyrinates

DOI: 10.1002/chem.201300219

Source and publish data:

Authors:

Article abstract of DOI:10.1002/chem.201300219

Full text of DOI:10.1002/chem.201300219

Products guided by the article

R&D Labs maybe for 1449421-56-0

Relevant to this article

Hot Product