Electron-Poor Thiophene 1,1-Dioxides: Synthesis, Characterization, and Application as Electron Relays in Photocatalytic Hydrogen Generation

Article abstract of DOI:10.1002/chem.201500543

The synthesis and characterization of electron-poor thiophene 1,1-dioxides bearing cyanated phenyl groups are reported. The electron-accepting nature of these compounds was evaluated by cyclic voltammetry, and highly reversible and facile reductions were observed for several derivatives. Moreover, some of the reduced thiophene dioxides form colorful anions, which were investigated spectroelectrochemically. Photoluminescence spectra of the electron-deficient sulfones were measured in CH₂Cl₂, and they emit in the blue-green region with significant variation in the quantum yield depending on the aryl substituents. By expanding the degree of substitution on the phenyl rings, quantum yields up to 34% were obtained. X-ray diffraction data are reported for two of the thiophene 1,1-dioxides, and the electronic structure was probed for all synthesized derivatives through DFT calculations. The dioxides were also examined as electron relays in a photocatalytic water reduction reaction, and they showed potential to boost the efficiency.

Full text of DOI:10.1002/chem.201500543

DOI: 10.1002/chem.201500543
Full Paper
&
Sulfur Heterocycles
Electron-Poor Thiophene 1,1-Dioxides: Synthesis,
Characterization, and Application as Electron Relays in
Photocatalytic Hydrogen Generation
[a]
[a]
[a]
[a]
Chia-Hua Tsai, Danielle N. Chirdon, Husain N. Kagalwala, Andrew B. Maurer,
[b]
[b]
[a]
[a]
Aman Kaur, Tomislav Pintauer, Stefan Bernhard, and Kevin J. T. Noonan*
Abstract: The synthesis and characterization of electron-
poor thiophene 1,1-dioxides bearing cyanated phenyl
groups are reported. The electron-accepting nature of these
compounds was evaluated by cyclic voltammetry, and highly
reversible and facile reductions were observed for several
derivatives. Moreover, some of the reduced thiophene diox-
ides form colorful anions, which were investigated spectroe-
lectrochemically. Photoluminescence spectra of the electron-
deficient sulfones were measured in CH Cl , and they emit in
the blue-green region with significant variation in the quan-
tum yield depending on the aryl substituents. By expanding
the degree of substitution on the phenyl rings, quantum
yields up to 34% were obtained. X-ray diffraction data are
reported for two of the thiophene 1,1-dioxides, and the elec-
tronic structure was probed for all synthesized derivatives
through DFT calculations. The dioxides were also examined
as electron relays in a photocatalytic water reduction reac-
tion, and they showed potential to boost the efficiency.
2
2
2
+
Introduction
Among available electron acceptors, methyl viologen (MV
,
Figure 1) is a popular material that has been successfully used
[
5]
[3]
Stable and efficient electron-transport materials are far less
abundant than hole-transporting substances, but they are in
high demand for both small-molecule and polymer technolo-
in both water reduction and electrochromics. Reduction of
2
+
MV causes a dramatic color change from yellow to the deep
blue of the MVC radical cation. This radical is fairly stable and
+
[1]
gy. They are particularly valuable for energy-related applica-
reduction occurs readily and reversibly. Despite these desirable
qualities, expansion from viologens towards neutral acceptors
would be advantageous, because neutral species may be solu-
ble in a wider variety of solvents to allow exploration of new
device environments and fabrication techniques. Additionally,
the mechanism of solid-state and solution-based charge trans-
port may be different for neutral structures as compared to
the ionic viologen.
[
2]
tions such as homogenous, photocatalytic water reduction
[3]
and electrochromic devices.
Photocatalytic water reduction in homogenous systems gen-
erally relies on a photosensitizer to absorb light and donate
[4]
the excited electron to a reduction catalyst. Electron accept-
ors with highly reversible electrochemistry can be used in this
process to shuttle electrons efficiently from the photosensitizer
to the catalyst. Electrochromic devices also require versatile ac-
ceptors. These compounds must undergo significant color
changes upon reduction and must be reproducibly switched
[3]
from the neutral to anionic state.
+
+
[
a] C.-H. Tsai, D. N. Chirdon, H. N. Kagalwala, A. B. Maurer, Prof. S. Bernhard,
Prof. K. J. T. Noonan
2
+
Figure 1. Methyl viologen (MV ) and 2,5-diphenylthiophene 1,1-dioxide
DPTD).
Department of Chemistry, Carnegie Mellon University
(
4400 Fifth Avenue, Pittsburgh, Pennsylvania, 15213 (USA)
E-mail: noonan@andrew.cmu.edu
[
b] A. Kaur, Prof. T. Pintauer
Thiophene 1,1-dioxide attracted our attention as a scaffold
Department of Chemistry and Biochemistry, Duquesne University
[
6]
600 Forbes Avenue, 308 Mellon Hall
for uncharged acceptors (Figure 1). These heterocycles are ac-
[
7]
Pittsburgh, Pennsylvania, 15282 (USA)
cessible from well-known thiophene precursors and have sig-
nificantly higher electron affinity than their unoxidized coun-
+
[
] These authors contributed equally to this work.
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201500543. CCDC-1400136 and CCDC-
[6]
terparts. Oligothiophenes bearing the dioxide moiety have al-
ready been explored as acceptors in organic photovoltaics to
1
400137 contain the supplementary crystallographic data for this paper.
[
6,8]
replace commonly used fullerenes. Thiophene dioxide mate-
rials are also highly emissive, and various derivatives have
These data can be obtained free of charge from The Cambridge Crystallo-
graphic Data Centre via www.ccdc.cam.ac.uk/data_request/cif.
Chem. Eur. J. 2015, 21, 11517 – 11524
11517
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
[9]
been explored in organic light-emitting diodes and cell-imag-
[
10]
ing applications.
Appending strongly electron-withdrawing groups (EWGs) to
the p system of organic materials is often used as a strategy to
[1a–c,11]
enhance electron affinity.
We recently observed signifi-
cantly more facile reduction of the thiophene dioxide when p-
C H NO or p-C H CN groups are attached at the 2- and 5-posi-
6
4
2
6
4
[
12]
tions. Herein, the effect of p-accepting cyano functionality
was explored in 2,5-diphenylthiophene dioxides (DPTDs) by in-
stalling cyano groups on various positions of the phenyl ring.
Redox potentials of the new cyanated compounds follow
a clear trend, and some derivatives produce highly colored
anions upon reduction. Both the stability and color of the re-
duced compounds led us to explore these spectroelectrochem-
ically and as electron relays in photocatalytic water reduction.
Figure 2. Solid-state molecular structure of 2,4-CNDPTD. Thermal ellipsoids
at 50%. Selected bond lengths [Š] and angles[8]: S1=O1 1.4357(9), S1=O2
1
1
1
.4332(10), S1ÀC1 1.7819(12), S1ÀC4 1.7856(11), C1ÀC2 1.3405(17), C2ÀC3
.4606(18), C3ÀC4 1.3400(16), S2=O3 1.4314(10), S2=O4 1.4383(9), S2ÀC21
.7835(11), S2ÀC24 1.7795(14), C21ÀC22 1.3386(19), C22ÀC23 1.4679(19),
C23-C24 1.3441(17); C1-S1-C4 93.93(6), C21-S2-C24 94.21(6).
Results and Discussion
Using our previously reported Stille cross-coupling methodolo-
[12]
gy,
a series of compounds was prepared to explore the
impact of both the position and number of cyano groups on
the DPTD structure (Scheme 1). Analogues to the previously
[
12]
reported 4-CNDPTD were prepared with the cyano group in
the ortho and meta positions (2-CNDPTD and 3-CNDPTD) from
2
-bromo and 3-bromobenzonitrile.
Figure 3. Solid-state molecular structure of 2,4,5-CNDPTD. Thermal ellipsoids
at 50%. Selected bond lengths [Š] and angles [8]: S1=O1 1.4204(16), S1=O2
1
1
.4248(16), S1ÀC1 1.7843(16), S1ÀC4 1.7793(15), C1ÀC2 1.319(2), C2ÀC3
.471(2), C3ÀC4 1.322(2); C1-S1-C4 92.54(7).
both 2,4-CNDPTD and 2,4,5-CNDPTD (1.42–1.44 Š) are compa-
rable to those of the previously reported DPTD (S=O 1.418(5)
[
13]
and 1.427(5) Š).
The CÀC bond lengths of the thiophene dioxide illustrate
the lack of aromaticity for this ring system with strong differen-
tiation of single and double bonds within the five-membered
structure (e.g., C1ÀC2 1.3405(17), C2ÀC3 1.4606(18), C3ÀC4
Scheme 1. Preparation of functionalized DPTDs.
1
.3400(16) Š for 2,4-CNDPTD). This is in sharp contrast to 2,5-
Derivatives with multiple cyano groups (2,4-CNDPTD and
,4,5-CNDPTD) were prepared from 4-bromoisophthalonitrile
and 5-bromobenzene-1,2,4-tricarbonitrile, both of which were
derived from their respective carboxylic acids (see Supporting
Information). The coupling of these multiply cyanated arene
rings to the dioxide heterocycle proceeded in moderate yields
diphenylthiophene, in which all CÀC bond lengths of the thio-
[13]
2
phene ring are within 0.03 Š (1.36–1.39 Š). The loss of aro-
maticity due to sulfur oxidation is also indicated by the longer
SÀC bond lengths in the thiophene dioxides (1.7819(12) and
1.7856(11) Š for 2,4-CNDPTD) as compared to 2,5-diphenylth-
[
13]
iophene (1.718(6) and 1.726(6) Š).
(
67% for 2,4-CNDPTD and 42% for 2,4,5-CNDPTD). All newly
The crystallization behavior of these two compounds is
quite interesting. For 2,4-CNDPTD (Figure 2), two different ori-
entations are observed in the unit cell. In one case, the dihe-
dral angles between the 2,4-CN aryl rings and the central het-
erocycle are 24.9 and 30.18. In the other case, the outer rings
are twisted in different directions, with dihedral angles of 33.1
and À27.08.
synthesized DPTDs were characterized by NMR spectroscopy
1
13
(
H and C) and mass spectrometry.
Crystallography
Crystals suitable for X-ray diffraction were grown for 2,4-
CNDPTD by slow evaporation of a chloroform solution
(
In 2,4,5-CNDPTD (Figure 3), the 2,4,5-tricyano aryl rings are
significantly more twisted from the central thiophene dioxide
with dihedral angles of 62.7 and À59.58. Interestingly, the elec-
tron-deficient nature of the arene ring results in short distan-
Figure 2), whereas 2,4,5-CNDPTD crystallized from dichloro-
methane (Figure 3). In both compounds, the SO moiety is per-
2
pendicular to the heterocyclic ring; the S=O bond lengths for
Chem. Eur. J. 2015, 21, 11517 – 11524
www.chemeurj.org
11518
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
ces to adjacent oxygen atoms in the crystal lattice. The lone
pair–p interaction observed for the oxygen atom of the sul-
fone to an adjacent tricyanobenzene has a distance of only
the heterocycle had a profound impact on photolumines-
cence: a quantum yield of 34.1% was recorded for 2,4-
CNDPTD. However, adding a third cyano group at the 5-posi-
tion (2,4,5-CNDPTD), which is not in resonance with the het-
erocycle, does not improve the quantum yield. Clearly, cyano
substituents that participate in the singlet excited state with
the thiophene dioxide affect the vibrational modes or relative
energies of states in such a way as to slow vibrational decay
and other deactivation pathways. The basic calculations pre-
sented here do not provide further insight into the effect of
the cyano groups, since the frontier orbitals for all derivatives
vary only in the amount of participation of the cyano groups,
and TDDFT calculations predict the same major types of transi-
tions for all compounds. Past literature does, however, describe
similar drastic changes in quantum yield on adding cyano
3
.329 Š.
DFT calculations and photophysics
DFT calculations were performed to evaluate the frontier orbi-
tals of the dioxides. Calculated HOMOs for all new derivatives
extend over the thiophene dioxide heterocycle and its phenyl
[
12]
substituents, as in previously reported DPTDs (Figure 4). The
LUMOs also mirror those of earlier analogues, residing mainly
on the heterocycle with heavy contributions from its p anti-
bonding orbitals.
[
14]
groups to terthiophenes.
Electrochemistry
Redox behavior was probed experimentally by cyclic voltam-
metry in acetonitrile under an inert atmosphere. Oxidation is
observed within the solvent window for the parent DPTD but
not for the cyano derivatives in which EWGs stabilize the
HOMO. At least two reductions are seen for all the DPTDs, and
these processes shift anodically, becoming easier as the quanti-
ty of cyano groups increases (Table 1). With three cyano
groups per phenyl ring, 2,4,5-CNDPTD boasts a reduction
shifted by +0.93 V relative to DPTD (Figure 5). This impressive
shift brings the only third reduction of the series into view and
makes the first two reductions of 2,4,5-CNDPTD easier than
that of the well-known acceptor fullerene (E of C : À0.44 V
Figure 4. Frontier orbitals calculated for the parent DPTD and representative
cyano derivative 2,4,5-CNDPTD.
Absorbances of the electron-deficient thiophene dioxides
just reach into the visible region (Supporting Information). For
all derivatives, lmax spans a small range of 16 nm (370–386 nm)
and shows no clear trend related to the substitution pattern
red
60
[
15]
vs. SCE).
(
Table 1). The absorption was assigned as a p–p* transition by
Beyond the large influence of the number of cyano groups,
their positioning has some effect on potential, as evidenced by
comparison of 2-CNDPTD, 3-CNDPTD, and 4-CNDPTD
(Figure 6). In this series, reduction is facilitated by placing the
cyano group on the phenyl substituent in the ortho (2-
means of time-dependent (TD) DFT calculations (Supporting In-
formation). All of the compounds luminesce in dichlorome-
thane with a maximum near 480 nm.
Interestingly, significant variation in the luminescence quan-
tum yield was observed in the series. DPTD and 3-CNDPTD
emit fairly weakly (1.8 and 3.5%, respectively) compared to 2-
CNDPTD and 4-CNDPTD (8.0 and 10.6%, respectively), each of
which has a cyano group in conjugation with the thiophene
dioxide. Addition of a second cyano group in conjugation with
CNDPTD, Ered =À0.83 V) or para position (4-CNDPTD, E
À0.77 V) rather than the meta position (3-CNDPTD E
=
=
red
red
À0.90 V). The electrochemical effect of the cyano group in 3-
CNDPTD is diminished due to the inability of the aryl meta po-
sition to provide resonance stabilization of the thiophene diox-
Table 1. Photophysical and electrochemical properties of thiophene 1,1-dioxides.
[
a]
[a]
max,em
[a]
[b]
[b]
[c]
[c]
[c]
Entry
l
max,abs
e
l
F
HOMO
[eV]
LUMO
[eV]
E
ox
E
red1
DEred1
E
red2
DEred2
À1 À1
[nm]
[cm
m
]
[nm]
[%]
[V vs. SCE]
[V vs. SCE]
[mV]
[V vs. SCE]
[mV]
[d]
[e]
[f]
DPTD
383
370
374
386
377
377
16200
14700
15700
18500
18200
18700
493
474
478
489
479
477
1.8
8.0
3.5
10.6
34.1
33.9
À5.942
À6.545
À6.686
À6.670
À7.189
À7.707
À2.477
À3.097
À3.134
À3.365
À3.906
À4.436
1.85
À1.16
66.7
86.2
66.7
60.3
77.1
61.8
À1.73
–
2
3
4
2
2
-CNDPTD
-CNDPTD
-CNDPTD
,4-CNDPTD
,4,5-CNDPTD
–
–
–
–
–
À0.83
À1.23
109.7
–
82.4
74.4
59.6
[
e]
e]
[g]
À0.90
À0.77
À0.48
À0.23
–
[
[e]
À1.16
À0.71
À0.37
[h]
[
2 2
a] Absorption and emission spectra were recorded in CH Cl and quantum yields were measured versus quinine sulfate. [b] Frontier orbital energies were
obtained from DFT calculations with the B3LYP functional and 6-31G(d,p) basis set (Supporting Information). [c] Potentials were measured with a three-
+
electrode system in degassed MeCN solutions containing 0.1m NnBu
.40 V vs. SCE). [d] Irreversible. [e] Quasireversible. [f] The process is irreversible and is much lower in intensity than the first reduction. [g] Process is not
well defined. [h] 2,4,5-CNDPTD exhibits a third reversible reduction at À1.81 V with DEred =90.1 mV.
4 6
PF and 1.9 mm analyte. An internal ferrocene reference was employed (Fc/Fc
0
Chem. Eur. J. 2015, 21, 11517 – 11524
www.chemeurj.org
11519
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Figure 7. LUMOs for 2-CNDPTD, 3-CNDPTD, and 4-CNDPTD, calculated by
using the B3LYP/6-31G(d,p) basis set. These images identify the site of re-
duction while exploring the effect of cyano positioning.
tion) for comparison with the o-C H CN substituent. Interest-
6
4
ingly, the 2-MeDPTD showed only quasireversible reductions,
that is, steric effects alone are not responsible for the electro-
chemical differences observed in the different DPTD deriva-
tives. The reversibility for 2-CNDPTD persists even in a nonpolar
solvent (CH Cl ), and this eliminates the possibility that reversi-
Figure 5. Cyclic voltammograms of 2,4-CNDPTD (solid) and 2,4,5-CNDPTD
(
dashed) showing the modulation of reduction with increasing cyano con-
À1
tent. Voltammograms were recorded at 0.1 Vs in degassed acetonitrile
4 6
with 0.1m NnBu PF supporting electrolyte and 1.9 mm analyte.
2
2
bility results from increased solvent stabilization of the reduced
state of o-CN derivatives. Therefore, the impact of an o-CN
group is likely a combination of sterics, electronics, and other
factors such as intramolecular forces. Hammett parameters
could not be used to help explain the change in reduction be-
havior from ortho to para, since constants for ortho substitu-
[
16]
ents are not available. Previous work on triaryl amines sug-
gests that steric effects of ortho substituents can affect redox
potential, though a change in the reversibility of the triaryl
[
17]
amine oxidation was not observed.
Regardless of the exact link between reversibility and cyano
positioning, all derivatives without o-CN clearly undergo signif-
icant degradation in the reduced form. The nature of this pro-
cess was examined by using 3-CNDPTD as a representative ex-
ample. Generation of degradation products in 3-CNDPTD is
substantiated by the appearance of multiple, low-intensity oxi-
dative processes following the initial cathodic sweep of the
first reduction. The degradation products limit the reversibility
of the first reduction irrespective of whether the second reduc-
tion is scanned. Meanwhile, concentration studies revealed
Figure 6. Cyclic voltammograms of 2-CNDPTD (black), 3-CNDPTD (black
dashed), and 4-CNDPTD (gray dashed). All voltammograms were recorded
À1
at 0.1 Vs in degassed acetonitrile with NnBu
and 1.9 mm analyte.
4 6
PF supporting electrolyte
ide, where reduction is expected to occur. This lack of reso-
nance is illustrated by the absence of cyano participation with
the heterocycle in the calculated LUMO of 3-CNDPTD
that cathodic peak potential and reversibility (i /i ) of the first
pa pc
reduction show initial dependence on analyte concentration
before leveling off at higher concentrations (Figure 8).
(
Figure 7). However, the calculated LUMO energies for 2-
CNDPTD, 3-CNDPTD, and 4-CNDPTD do not accurately reflect
the relationship between experimental potentials.
Although CÀS bond cleavage and dimerization are common
[
18]
causes of irreversibility in organic sulfones, monomolecular
[
18a,19]
Cyano positioning also controls the reversibility of reduction
processes including CÀS cleavage can be ruled out here.
(
Figure 6). With m-CN, a first, quasireversible reduction is fol-
Reductive dimerization or multistep coupling could still be en-
visioned, but further experimentation such as EPR spectrosco-
py would be needed to substantiate this. As expected, no cor-
relation with concentration is seen for the reversible reductions
of 2-CNDPTD.
lowed by a poorly defined process. Two quasireversible reduc-
tions are observed for 4-CNDPTD, whereas completely reversi-
ble reductions are achieved for all derivatives having cyano
groups in the ortho position (2-CNDPTD, 2,4-CNDPTD, and
2
,4,5-CNDPTD).
Along with complexity, the electrochemistry reveals possible
applicability. Compounds having multiple cyano groups per
phenyl ring (2,4-CNDPTD and 2,4,5-CNDPTD) rival the impres-
Simple electronic features of the ortho position cannot ex-
plain its role in reversibility, since the ortho- and para-substitut-
ed compounds have similar LUMOs (Figure 7). A model com-
pound was synthesized to probe how the sterics of an ortho-
substituted phenyl ring may influence reduction reversibility.
An o-C H Me group was installed at the 2- and 5-positions of
sive reductions of viologens (methyl viologen dichloride: E
=
red
[
20]
À0.687, À1.121 V vs. SCE). This indicates a strong possibility
for charge-transport or electron-relay functionality.
The first reduction of derivatives having o-CN functionality
also causes an intense visible color change from yellow to blue
6
4
the thiophene dioxide (2-MeDPTD; see Supporting Informa-
11520
Chem. Eur. J. 2015, 21, 11517 – 11524
www.chemeurj.org
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Figure 8. Concentration dependence of the reversibility (ipa/ipc) of the first re-
duction. The dashed line represents the behavior of an ideal reversible re-
duction, filled squares correspond to 2-CNDPTD, and empty squares corre-
spond to 3-CNDPTD. Reductions were measured by cyclic voltammetry in
4 6
degassed acetonitrile with NnBu PF supporting electrolyte.
at the electrode surface. To investigate this for possible electro-
chromic applications, absorbance spectra were recorded
during cyclic voltammetry. Spectral features representing the
color change of the first reduction generally appear as the po-
tential of that process is traversed, but in the case of 2-
CNDPTD, the changes are low in intensity, and their detection
lags behind the reduction (Supporting Information). The spec-
tral changes are similar in all o-CN derivatives and include par-
tial loss of the absorbance of the neutral species near 350 nm
and the appearance of two new signals for the reduced anion:
a major peak below 750 nm and a smaller, low-energy feature
Figure 9. Spectroelectrochemistry of 2,4-CNDPTD. Bottom: absorbance
spectrum and difference spectra showing the change in absorbance both
during and after cyclic voltammetry. Top: corresponding cyclic voltammo-
gram with bars identifying the range of potentials traversed during collec-
tion of each difference spectrum.
The blue color that appears in the o-CN derivatives is also
visible during the first reduction of 3-CNDPTD and 4-CNDPTD
but is too fleeting for detection. A more unique color change
occurs in the previously reported parent DPTD, reduction of
which creates a red appearance corresponding to emergence
of an absorbance near 547 nm (Figure 10). The blueshift in this
absorbance is due to the simpler p system of DPTD. Ultimately,
the red color of DPTD, like the blue color of 3-CNDPTD and 4-
CNDPTD, is short-lived due to redox instability.
(
Figure 9). The new signals are redshifted as the p system is ex-
tended, and the major absorbance ranges from 621 nm for 2-
CNDPTD to 647 nm for 2,4-CNDPTD and 709 nm for 2,4,5-
CNDPTD (Figure 10).
Despite the many similarities of the o-CN derivatives, the
spectral behavior on passing the second reduction and revers-
ing the voltammogram differentiates 2,4-CNDPTD and 2,4,5-
CNDPTD from 2-CNDPTD. In the last-named, progression of
the voltammogram ultimately leads to increased absorbance
near 458 nm that remains prominent while the signal of the
singly reduced species (lmax =621 nm) fades. Such dynamic
spectroelectrochemistry suggests slight degradation over time,
which is not detected by electrochemistry but has been en-
[
18c]
countered previously for p-nitrophenyl methyl sulfone.
However, for 2,4-CNDPTD and 2,4,5-CNDPTD, the second re-
duction does not cause a spectral change, and instead the in-
tense absorbances of the first reduction persist throughout the
voltammogram and for several seconds after its completion.
Radical anions formed from 2,4-CNDPTD and 2,4,5-CNDPTD
thus appear to be stable on a timescale longer than that of
cyclic voltammetry. Such charge stability in these derivatives
combined with the intensity and low turn-on voltage of their
color change make them interesting candidates for electro-
chromic devices.
Figure 10. Absorbance spectra representing the height of the redox-induced
color change for all compounds investigated by spectroelectrochemistry.
The spectra were taken past the second reduction wave and show a mixture
of absorbances from the original species (ꢀ350–400 nm) as well as from
the reduced species (ꢀ450–725 nm).
Chem. Eur. J. 2015, 21, 11517 – 11524
www.chemeurj.org
11521
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Photocatalytic dihydrogen generation
(TEA) as the sacrificial donor, and K PtCl as a precursor to the
2
4
[
21]
colloidal platinum catalyst. In THF, this system produces ap-
Given their rich electrochemical properties, the DPTDs were in-
proximately 570 mmol of H in 70 h. However, as evidenced in
2
vestigated as electron relays (ERs) for photocatalytic water re-
Figure 11 (no-relay trace), the rate of H production is signifi-
2
duction. Setups for photogeneration of H2 from H O can in-
2
clude three components (Scheme 2A): a photosensitizer, a sacri-
Figure 11. Comparison of the amount of H
atives as ERs. 2-CNDPTD was found to be the most efficient ER, generating
over 1600 mmol H even after an illumination period of 70h. Reactions con-
2
generated by using DPTD deriv-
2
2 4
ditions: 0.1 mm photosensitizer, 0.2 mm ER, and 30 mm K PtCl in THF/water/
TEA (8/1/1) solution.
cantly diminished after only 5 h. The addition of DPTDs to act
as ERs decreased the rate of H formation initially, but afforded
2
a significantly more robust water reduction system with larger
amounts of H generated over time (Figure 11).
2
The thiophene dioxide clearly facilitates the electron-transfer
process from the photosensitizer to the proton reduction cata-
lyst, and control experiments in the absence of photosensitizer
did not produce any H , which confirms that the DPTDs do not
2
act as the light absorbers. All the compounds bearing cyano
groups generated more H as compared to the unsubstituted
2
DPTD, indicating that the former act as more resilient ERs.
Moreover, the cyanated compounds seem to impart robust-
ness to the overall system since some of the reactions contin-
Scheme 2. A) Three-component setup for dihydrogen generation with an iri-
dium photosensitizer, triethylamine, and a platinum catalyst. B) Four-compo-
nent setup including a DPTD electron relay. C) Thermodynamics of an elec-
tron-transfer cascade observed after reductive quenching of the photosensi-
tizer in a four-component setup. Displayed energy levels are based on the
experimental first reduction potentials of the components. The rate of elec-
tron transfer (ET) between the partners greatly influences the hydrogen-evo-
lution kinetics of the system.
ued to produce H even after 70 h of illumination. CN binding
2
to the colloidal platinum could result in efficient electron trans-
fer to the reduction catalyst and facilitate the overall relay pro-
[23]
cess.
Like the electrochemical behavior, the yield of H was dra-
2
matically influenced by the structure of the DPTD chosen as
the ER. In THF, 2-CNDPTD underwent 1605 ER TON and pro-
duced the largest amount of H2 for all experiments. Mean-
while, electronically similar 3-CNDPTD and 4-CNDPTD pro-
duced only 1067 and 717 TON, respectively. The derivatives
lacking o-CN groups exhibit only quasireversible reductions in
cyclic voltammetry, which is likely to lead to degradation
during ER redox activity. Interestingly, 2,4-CNDPTD and 2,4,5-
CNDPTD, which have completely reversible reductions, still do
not perform as well as 2-CNDPTD. Two factors likely contribute
to the excellent behavior observed for 2-CNDPTD. First, the
smaller size of 2-CNDPTD compared to derivatives with multi-
ple cyano groups may allow faster diffusion from the photo-
ficial donor, and a proton reduction catalyst. Hydrogen genera-
tion can also involve an electron relay as a fourth component,
which shuttles electrons from the photosensitizer (PS) to the
catalyst (Scheme 2B). To date, a variety of iridium catalysts
have proven to be efficient photosensitizers in hydrogen gen-
eration (up to 2400 photosensitizer turnover numbers in
[
21]
THF), and recently, an iridium catalyst with aryl silyl substitu-
[
22]
ents exhibited exceptional performance for water reduction.
We explored a popular three-component system that uses
[
Ir(ppy) (dtbbpy)]PF6 (ppy=2-phenylpyridine, dtbbpy=4,4’-di-
2
tert-butyl-2,2’-bipyridine) as the photosensitizer, triethylamine
Chem. Eur. J. 2015, 21, 11517 – 11524
www.chemeurj.org
11522
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
III
sensitizer to the catalyst. Also, the LUMO of 2-CNDPTD may be
best positioned between the LUMOs of the photosensitizer
and catalyst to streamline electron transfer.
known to reductively quench the Ir excited state (k on the
q
7
8
À1 À1 [25]
order of 10 –10 m s ) and is present at a comparatively
higher concentration (0.7m) than the ER (0.2 mm), which likely
makes reductive quenching the dominant process. Since the
ER acts as an electron shuttle between the photochemically re-
duced photosensitizer and platinum, the position of its LUMO
energy level is critical for efficient electron transfer. Although
a low-lying energy level favorably assists electron injection
from the HSOMO of the photosensitizer, it could also give rise
to competitive quenching pathways (with TEA), energy-level
mismatch with the catalyst, or even the possibility of energy
transfer rather than the desired electron transfer, which would
lead to an overall decrease in efficiency. In this regard, the
LUMO energy level of 2-CNDPTD in THF seems to be optimally
positioned such that both transfer from the photosensitizer
and to the catalyst have sufficient driving force without too
great of an energy change (Scheme 2C). In MeCN, 2-CNDPTD
still performs well (971 TON), but 2,4-CNDPTD produces more
The effect of coordinating solvents like MeCN on photosen-
sitizer stability and overall catalytic performance has been well
[
24]
documented.
Unlike traditional systems, the current four-
component system yielded substantial TONs for some of the
DPTDs in MeCN, which confirmed their viability as ERs. The
trends in H2 yield are similar in both solvents (Figure 12),
H (1073 TON), which suggests that more factors are at play in
2
a highly coordinating and polar solvent.
Figure 12. Comparison of H
and MeCN (front). All reactions involved 0.1 mm photosensitizer, 0.2 mm ER,
and 30 mmol K PtCl in solvent/water/TEA (8/1/1) solution.
2
generation, PS TON, and ER TON in THF (back)
Conclusion
2
4
The synthesis and characterization of a series of electron-poor
thiophene 1,1-dioxides have been reported. The phenyl sub-
stituents appended to the dioxide structure have a marked in-
fluence on the electrochemical and photophysical behavior of
the compound. In particular, phenyl substituents with ortho
cyano groups exhibit significantly more stable electrochemical
behavior than their meta and para congeners. Increasing the
cyano content of the phenyl rings facilitates electrochemical
reduction while increasing radical-anion stability. Many of the
diphenyl thiophene 1,1-dioxides produce colored radical
anions, which were analyzed by spectroelectrochemistry. The
except for 2,4-CNDPTD, which maintains essentially the same
performance (1062 TON in THF and 1073 TON in MeCN). Ex-
periments are ongoing to evaluate why 2,4-CNDPTD maintains
the same efficiency while all the other derivatives exhibit
a drop in relay performance in MeCN.
Dynamic quenching studies were carried out in MeCN with
III
the Ir photosensitizer (Supporting Information) to obtain
a better mechanistic understanding of H generation. Stern–
2
Volmer plots were generated for all derivatives and revealed
that cyano-bearing compounds quench the photosensitizer
emission better than the parent DPTD. Given the electron-
poor nature of the dioxides, oxidative quenching would be ex-
pected to take place in this instance. As suspected, the elec-
tron-accepting ability was found to increase with increasing
cyano content, with 2,4,5-CNDPTD having the largest quench-
2,4-CNDPTD and 2,4,5-CNDPTD derivatives exihibit stability
beyond the timescale of electrochemical cycling. The electro-
chemical properties of the synthesized dioxides led to their ex-
ploration in photocatalytic hydrogen generation. The DPTDs
function well as ERs and can significantly enhance the amount
of H generated with an iridium photosensitizer, triethylamine,
2
9
À1 À1
and a platinum proton reduction catalyst. Compound 2-
CNDPTD outperformed all other derivatives in THF, while 2,4-
CNDPTD was the best ER in MeCN. Studies on the enhanced
stability with an electron-withdrawing substituent in the ortho
position are currently underway.
ing constant of k =7.4”10 m
s
(Supporting Information).
q
Moreover, the k values follow the trend of the first reduction
q
potentials Ered1 and the calculated SOMO energy levels of the
anion (Supporting Information). These results strongly suggest
that lowering of the LUMO level as a consequence of cyano
substitution facilitates electron transfer from the excited state
of the Ir photosensitizer. From these trends, one would expect
2
,4,5-CNDPTD to be the best ER for photocatalysis. However, Acknowledgements
as observed from the H -evolution results (Figure 12), 2-
2
CNDPTD was the most active ER in THF, and 2,4-CNDPTD per-
formed slightly better in MeCN.
K.J.T.N. is grateful to the ARO for a Young Investigator Award
(63038-CH-YIP). NMR Instrumentation at Carnegie Mellon was
partially supported by the NSF (CHE-0130903 and CHE-
1039870). S.B. acknowledges support from the NSF under
award number CHE-1362629. D.N.C. gratefully acknowledges
the support of the US DOE Office of Science Graduate Research
Fellowship.
The observed quenching trend does not match the trend
found for ER performance due to the complexity of water re-
duction (four components). The photocatalytic system involves
two additional participants, that is, the sacrificial agent (TEA)
and the water reduction catalyst (K PtCl /colloidal Pt). TEA is
2
4
Chem. Eur. J. 2015, 21, 11517 – 11524
www.chemeurj.org
11523
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
golani, G. Gigli, Adv. Mater. 2003, 15, 2060–2063; e) R. He, J. Xu, Y. Xue,
D. Chen, L. Ying, W. Yang, Y. Cao, J. Mater. Chem. C 2014, 2, 7881–7890;
f) J. Zhang, S. Dong, K. Zhang, A. Liang, X. Yang, F. Huang, Y. Cao, Chem.
Commun. 2014, 50, 8227–8230; g) J. Liu, J. Zou, W. Yang, H. Wu, C. Li,
B. Zhang, J. Peng, Y. Cao, Chem. Mater. 2008, 20, 4499–4506.
Keywords: cyanides · electrochemistry · electron relays ·
photocatalysis · sulfur heterocycles
[
1] a) J. E. Anthony, A. Facchetti, M. Heeney, S. R. Marder, X. Zhan, Adv.
Mater. 2010, 22, 3876–3892; b) M. Stolar, T. Baumgartner, Phys. Chem.
Chem. Phys. 2013, 15, 9007–9024; c) X. Zhao, X. Zhan, Chem. Soc. Rev.
[
10] I. Palamà, F. Di Maria, I. Viola, E. Fabiano, G. Gigli, C. Bettini, G. Barbarel-
la, J. Am. Chem. Soc. 2011, 133, 17777–17785.
[
[
11] H. Usta, A. Facchetti, T. J. Marks, Acc. Chem. Res. 2011, 44, 501–510.
12] C.-H. Tsai, D. N. Chirdon, A. B. Maurer, S. Bernhard, K. J. T. Noonan, Org.
Lett. 2013, 15, 5230–5233.
13] P. Pouzet, I. Erdelmeier, D. Ginderow, J.-P. Mornon, P. M. Dansette, D.
Mansuy, J. Heterocycl. Chem. 1997, 34, 1567–1574.
14] F. Demanze, J. Cornil, F. Garnier, G. Horowitz, P. Valat, A. Yassar, R. Lazzar-
oni, J.-L. BrØdas, J. Phys. Chem. B 1997, 101, 4553–4558.
15] D. Dubois, K. M. Kadish, S. Flanagan, R. E. Haufler, L. P. F. Chibante, L. J.
Wilson, J. Am. Chem. Soc. 1991, 113, 4364–4366.
16] C. Hansch, A. Leo, R. W. Taft, Chem. Rev. 1991, 91, 165–195.
17] T. P. Bender, J. F. Graham, J. M. Duff, Chem. Mater. 2001, 13, 4105–4111.
18] a) J.-F. Bergamini, P. Hapiot, C. Lagrost, L. Preda, J. Simonet, E. Volanschi,
Phys. Chem. Chem. Phys. 2003, 5, 4846–4850; b) J. Simonet, in Sulfur-
Containing Functional Groups, 2nd ed. (Eds.: S. Patai, Z. Rappoport),
Wiley, Chichester, 1993, pp. 439–493; c) J.-F. Pilard, O. Fourets, J. Simo-
net, L. J. Klein, D. G. Peters, J. Electrochem. Soc. 2001, 148, E171–E175;
d) E. Constantinescu, M. Hillebrand, E. Volanschi, M. Andrei, G. Ivanescu,
O. Maior, J. Electroanal. Chem. 1995, 395, 211 –220; e) L. Preda, V. Lazar-
escu, M. Hillebrand, E. Volanschi, Electrochim. Acta 2006, 51, 5587–
2
011, 40, 3728–3743; d) J. Zaumseil, H. Sirringhaus, Chem. Rev. 2007,
1
07, 1296–1323; e) C. R. Newman, C. D. Frisbie, D. A. da Silva Filho, J.-L.
BrØdas, P. C. Ewbank, K. R. Mann, Chem. Mater. 2004, 16, 4436–4451.
2] a) W. T. Eckenhoff, R. Eisenberg, Dalton Trans. 2012, 41, 13004–13021;
b) Z. Han, R. Eisenberg, Acc. Chem. Res. 2014, 47, 2537–2544; c) S. Be-
rardi, S. Drouet, L. Francàs, C. Gimbert-SuriÇach, M. Guttentag, C. Rich-
mond, T. Stoll, A. Llobet, Chem. Soc. Rev. 2014, 43, 7501–7519.
3] a) R. J. Mortimer, Chem. Soc. Rev. 1997, 26, 147–156; b) R. J. Mortimer,
Electrochim. Acta 1999, 44, 2971–2981; c) R. J. Mortimer, A. L. Dyer, J. R.
Reynolds, Displays 2006, 27, 2–18.
[
[
[
[
[
[
[
[
[
[
4] J. Kiwi, K. Kalyanasundaram, M. Grätzel, Struct. Bonding (Berlin) 1982, 49,
3
7–125.
5] a) K. Kalyanasundaram, J. Kiwi, M. Grätzel, Helv. Chim. Acta 1978, 61,
720–2730; b) J. Kiwi, M. Grätzel, Nature 1979, 281, 657–658; c) I.
2
Okura, N. Kim-Thuan, J. Mol. Cat. 1979, 6, 449–451; d) N. Mourtzis, P. C.
Carballada, M. Felici, R. J. M. Nolte, R. M. Williams, L. de Cola, M. C. Feit-
ers, Phys. Chem. Chem. Phys. 2011, 13, 7903–7909; e) P. C. Carballada, N.
Mourtzis, M. Felici, S. Bonnet, R. J. M. Nolte, R. M. Williams, L. De Cola,
M. C. Feiters, Eur. J. Org. Chem. 2012, 6729–6736.
6] G. Barbarella, M. Melucci, Thiophene-S,S-dioxides as a Class of Electron-
Deficient Materials for Electronics and Photonics, in Handbook of Thio-
phene-Based Materials: Applications in Organic Electronics and Photonics
5595.
[
[
[
19] A. J. Bard, L. R. Faulkner, Electrochemical Methods: Fundamentals and Ap-
plications, Wiley, New York, 2001.
[
[
20] C. L. Bird, A. T. Kuhn, Chem. Soc. Rev. 1981, 10, 49–82.
21] P. N. Curtin, L. L. Tinker, C. M. Burgess, E. D. Cline, S. Bernhard, Inorg.
Chem. 2009, 48, 10498–10506.
(
Eds.: I. F. Perepichka, D. F. Perepichka), Wiley, Chichester, 2009,
pp. 255–292.
7] a) G. Barbarella, L. Favaretto, G. Sotgiu, M. Zambianchi, L. Antolini, O.
Pudova, A. Bongini, J. Org. Chem. 1998, 63, 5497–5506; b) S. Rozen, Y.
Bareket, J. Chem. Soc. Chem. Commun. 1994, 1959–1959; c) E. Amir, S.
Rozen, Angew. Chem. Int. Ed. 2005, 44, 7374–7378; Angew. Chem. 2005,
[
[
[
22] D. R. Whang, K. Sakai, S. Y. Park, Angew. Chem. Int. Ed. 2013, 52, 11612–
11615; Angew. Chem. 2013, 125, 11826–11829.
23] I. N. Mills, H. N. Kagalwala, D. N. Chirdon, A. C. Brooks, S. Bernhard, Poly-
hedron 2014, 82, 104–108.
24] a) R. Ziessel, J. Hawecker, J.-M. Lehn, Helv. Chim. Acta 1986, 69, 1065–
1
17, 7540–7544.
[
[
8] a) F. Zhang, D. Wu, Y. Xu, X. Feng, J. Mater. Chem. 2011, 21, 17590–
1
1
1084; b) L. L. Tinker, N. D. McDaniel, P. N. Curtin, C. K. Smith, M. J. Ire-
7600; b) P. Sonar, J. P. F. Lim, K. L. Chan, Energy Environ. Sci. 2011, 4,
558–1574; c) N. Camaioni, G. Ridolfi, V. Fattori, L. Favaretto, G. Barbar-
land, S. Bernhard, Chem. Eur. J. 2007, 13, 8726–8732.
[
25] E. D. Cline, S. E. Adamson, S. Bernhard, Inorg. Chem. 2008, 47, 10378–
ella, J. Mater. Chem. 2005, 15, 2220–2225.
10388.
9] a) G. Gigli, G. Barbarella, L. Favaretto, F. Cacialli, R. Cingolani, Appl. Phys.
Lett. 1999, 75, 439–441; b) G. Barbarella, L. Favaretto, G. Sotgiu, M.
Zambianchi, A. Bongini, C. Arbizzani, M. Mastragostino, M. Anni, G.
Gigli, R. Cingolani, J. Am. Chem. Soc. 2000, 122, 11971–11978; c) G. Bar-
barella, L. Favaretto, G. Sotgiu, L. Antolini, G. Gigli, R. Cingolani, A. Bon-
gini, Chem. Mater. 2001, 13, 4112–4122; d) M. Mazzeo, V. Vitale, F. Della
Sala, D. Pisignano, M. Anni, G. Barbarella, L. Favaretto, A. Zanelli, R. Cin-
Received: February 9, 2015
Revised: May 9, 2015
Published online on June 30, 2015
Chem. Eur. J. 2015, 21, 11517 – 11524
www.chemeurj.org
11524
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim