Dye-sensitized solar cells and complexes between pyridines and iodines. A NMR, IR and DFT study

Article abstract of DOI:10.1016/j.saa.2012.08.006

Interactions between triiodide (I3-) and 4-tert-butylpyridine (4TBP) as postulated in dye-sensitized solar cells (DSC) are investigated by means of ¹³C NMR and IR spectroscopy supported by DFT calculations. The charge transfer (CT) complex 4TBP·I₂ and potential salts such as (4TBP)₂I⁺, I3- were synthesized and characterized by IR and ¹³C NMR spectroscopy. However, mixing (butyl)₄N ⁺, I3- and 4TBP at concentrations comparable to those of the DSC solar cell did not lead to any reaction. Neither CT complexes nor cationic species like (4TBP)₂I⁺ were observed, judging from the ¹³C NMR spectroscopic evidence. This questions the previously proposed formation of (4TBP)₂I⁺ in DSC cells.

Full text of DOI:10.1016/j.saa.2012.08.006

Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 98 (2012) 247–251
Contents lists available at SciVerse ScienceDirect
Spectrochimica Acta Part A: Molecular and
Biomolecular Spectroscopy
journal homepage: www.elsevier.com/locate/saa
Dye-sensitized solar cells and complexes between pyridines and iodines. A NMR,
IR and DFT study
⇑
⇑
Poul Erik Hansen , Phuong Tuyet Nguyen, Jacob Krake, Jens Spanget-Larsen, Torben Lund
Department of Science, Systems and Models, Roskilde University, P.O. Box 260, DK-4000 Roskilde, Denmark
g r a p h i c a l a b s t r a c t
h i g h l i g h t s
"
"
"
Triiodide and pyridines are
important electrolyte components in
DSCs.
Complex formation between these
species has been previously
suggested.
Observed NMR shifts indicate that
no such interaction takes place in
CH3CN solution.
CH3
H3C
CH3
-
+
I3
N
a r t i c l e i n f o
a b s t r a c t
Interactions between triiodide (Iꢀ3 ) and 4-tert-butylpyridine (4TBP) as postulated in dye-sensitized solar
cells (DSC) are investigated by means of 13C NMR and IR spectroscopy supported by DFT calculations. The
charge transfer (CT) complex 4TBPꢁI2 and potential salts such as (4TBP)2I+, I3ꢀ were synthesized and char-
acterized by IR and 13C NMR spectroscopy. However, mixing (butyl)4N+, Iꢀ3 and 4TBP at concentrations
comparable to those of the DSC solar cell did not lead to any reaction. Neither CT complexes nor cationic
species like (4TBP)2I+ were observed, judging from the 13C NMR spectroscopic evidence. This questions
the previously proposed formation of (4TBP)2I+ in DSC cells.
Article history:
Received 11 April 2012
Received in revised form 1 August 2012
Accepted 7 August 2012
Available online 19 August 2012
Keywords:
Dye-sensitized solar cells
Charge transfer complex
Nitrogen additives
Ó 2012 Elsevier B.V. All rights reserved.
(4-tert Butylpyridine)2I+
Introduction
TiO2 to more negative potentials [4,7–9]. The nitrogen additives
have also been proposed to decrease the dark current of the DSC
Dye-sensitized solar cells (DSC) have been extensively studied
in the last decade as a promising renewable energy source because
of their potential inexpensive manufacturing technology compared
to silicon cells [1–5]. In a typical DSC the electrolyte is comprised
of an Iꢀ/Iꢀ3 redox couple (0.5 M/0.05 M) in combination with a
nitrogen containing additive (0.5 M), e.g., 4-tert-butylpyridine
(4TBP) or 1-methylbenzimidiazole dissolved in, e.g., 3-methoxy-
propionitrile [6]. The nitrogen additive increases the open circuit
voltage Voc of the DSC by shifting the conduction band of the
by the formation of complexes between Iꢀ3 and, e.g., 4TBP as shown
in Eqs. (1) and (2) [8,10,11]
4TBP þ Iꢀ¡4TBP ꢁ I2 þ Iꢀ
ð1Þ
ð2Þ
3
2ð4TBPÞ þ Iꢀ3 ¡ð4TBPÞ2Iþ þ 2Iꢀ
The complexes and ions formed according to Eqs. (1) and (2) are
suggested to diffuse slower towards the TiO2 photoanode than Iꢀ it-
3
self with the result that the back electron transfer from the TiO2
conduction band to Iꢀ3 is reduced and Voc increased [8,11]. Forma-
tion of the bis(4-tert-butylpyridine)iodonium cation, (4TBP)2I+,
according to Eq. (2) has also been proposed by Lund and co-workers
⇑
Corresponding authors. Tel.: +45 46742472 (T. Lund).
E-mail addresses: poulerik@ruc.dk (P.E. Hansen), tlund@ruc.dk (T. Lund).
1386-1425/$ - see front matter Ó 2012 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.saa.2012.08.006

248
P.E. Hansen et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 98 (2012) 247–251
[12] to reduce the thiocyanate ligand exchange rate in the N719
ruthenium dye by 4TBP at elevated temperatures.
information is provided as Supplementary data, referred to in the
ensuing text as S1–S9.
The reason for suggesting Eq. (2) is most likely the paper by
Jones et al. [13], which in turn was inspired by the work of Reid
and Mulliken [14], as formulated in Eq. (3):
Experimental
Synthesis
IClꢀ2 þ pyridine¡pyridineIþ þ 2Clꢀ
ð3Þ
Bis(4-tert-butylpyridine)iodonium and bis(pyridine)iodonium
tetrafluoroborate
Jones et al. [13] reported an equilibrium constant for this reaction
equal to 4.4 ꢂ 104 L molꢀ1 in 6 M HCl, determined by using the
Benesi–Hildebrand approach. However, such acidic conditions have
little relevance for the solar cell. In fact, it is difficult to see how Eq.
(3) could be relevant under such conditions.
(4TBP)2I+, BF4ꢀ (1) and (pyridine)2I+, BFꢀ4 (2) were synthesized by
a modified literature method [22]. AgBF4 (1.946 g, 100 mmol), 2 g
TLC silica gel and CH2Cl2 (60 ml) was added to a 100 ml round
bottom flask. 4-tert-Butylpyridine or pyridine (20 mmol) was
added at room temperature under magnetic stirring followed by
addition of I2 (10 mmol). After 2 h at room temperature the silica
gel was removed by filtration and washed by 2 ꢂ 10 ml CH2Cl2.
The combined dichloromethane was removed by rotary evapora-
tion and the orange product was washed with cold diethyl ether
and isolated by vacuum filtration. Yield of light yellow prisms
2.85 g (59%). Recrystallisation from dichloromethane/diethyl ether
gave white prisms with a melting point between 131–168 °C. Ele-
mental analysis for 1: Predicted based on C18H26N2IBF4: 44.66% C,
5.41% H, 5.79% N. Found: 44.24% C, 5.33% H, 5.69% N. m.p. 174 °C
(starts to decompose) 1H NMR data see S8; for 13C NMR see Table 1.
Other experimental evidence for the reaction between Iꢀ3 and
4TBP and other pyridines is limited and only based on UV–Vis mea-
surements of dilute solutions of 4TBP and Iꢀ3 [11]. In contrast, the
complex formation between I2 and heteroaromatic compounds
such as a variety of pyridines (Pyr) is well established [14–18]. In
non-polar solvents, the charge transfer (CT) complexes (Pyr I2)
are formed exclusively [14,17] whereas in polar solvents the CT
complex is formed with various amounts of ionic products such
as PyrI+ or Pyr2I+ [17].
The parent (pyridine)2I+ salt has been characterized by X-ray
crystallography [19]. Cationic species such as pyridineI+ or (pyri-
dine)2I+ were formed in stretched polyethylene by addition of ex-
cess iodine to pyridine and characterized by IR polarization
spectroscopy [20]. CT complexes and ionic product formation be-
tween alkyl substituted pyridines and iodine has been investigated
by NMR using solvent mixtures of CDCl3 and nitrobenzene [15].
The chemical shift changes induced by the formation of iodonium
salts were distinctly different from those found for the simple CT
complex. For DSCs, however, it is the reaction between Iꢀ3 and
nitrogen additives which is of interest. As the equilibrium constant
Bis(4-tert-butylpyridine)iodonium triodide
(4TBP)2I+, Iꢀ3 (3) was synthesized as described by Hassel and
Hope [19]. 4-tert-Butylpyridine (5 mmol) in methanol (25 ml)
was mixed with I2 (5 mmol) dissolved in methanol (25 ml). Precip-
itation of a red compound was observed immediately. The solvent
was decanted and the remaining solvent was removed by vacuum
evaporation. Yield 83%.
for the reaction Iꢀ þ I2¡Iꢀ in acetonitrile and similar solvents is
4-tert-Butylpyridine diiodine CT complex
3
very high, K ꢃ 107 Mꢀ1 [21], the distinction between I2 and Iꢀ3 is
essential.
(A) 4TBPꢁI2 (4) was synthesized by stirring a mixture of iodine
(15.7 mmol) and 4-tert-butylpyridine (15.7 mmol) in
230 ml hexane and 150 ml dichloromethane, leaving the
mixture overnight. Orange crystals precipitated after slow
evaporation of the solvent at room temperature. Elemental
analysis for 4: Predicted based on C9H13NI2 27.78% C, 3.34%
H and 3.60% N. Found 27.84% C, 3.07% H and 3.61% N.
(B) 4TBPꢁI2 (4) 4-tert-butylpyridine (10 mmol) was dissolved in
ethanol (140 ml) in a round bottom flask. Iodine (10 mmol)
was dissolved in ethanol (10 ml) and slowly added drop wise
under magnetic stirring at room temperature. The yellow
precipitate was isolated by vacuum filtration and dried in
a vacuum oven at 40 °C. Yield 0.98 g (25%). The product
began to decompose at 96 °C.
In view of the large interest in establishing the interactions be-
tween the components in the DSC solar cells we wish to provide
tools for investigation of the interactions between I2 or Iꢀ3 and
DSC nitrogen additives. In this context, we want to characterize
the 4-tert-butylpyridine (4TBP) reaction products with iodine and
triiodide using solid and liquid state NMR and IR spectroscopy. In
particular, we want to test the proposal of Lindquist and Kusama
that Iꢀ3 may react with 4TBP according to Eqs. (1) and (2).ꢀAs refer-
ence materials, the species (4TBP)2I+, BF4ꢀ (1), (4TBP)2I+, I3 (3) and
4TBPꢁI2 (4) have been synthesized (see Fig. 1). To tie back to previ-
ous results, (pyridine)2I+, BF4ꢀ (2) was also synthesized and investi-
gated. The experimental results are supported by DFT calculations
of 13C nuclear shieldings and vibrational transitions. Additional
NMR spectroscopy
CH3
C
X-
The liquid state NMR spectra were recorded on a Varian Mer-
cury 300 at 300 MHz for 1H and 75 MHz for 13C NMR using the
methyl signal of acetonitrile as reference. The solid state spectra
were recorded on a Varian Inova Sparc 400 instrument. Spectra
were recorded with spinning speeds of both 6000 Hz and
12000 Hz.
+
H3C
CH3
R
N
N
R
I
R
X-
N
-
1
2
3
C(CH3)3 BF4
IR spectroscopy
-
I2
H
BF4
IR absorbance spectra were recorded at room temperature on a
Perkin-Elmer Spectrum 2000 FT-IR spectrophotometer in the
4000–370 cmꢀ1 range with a resolution of 1 cmꢀ1 (average of 10
scans). Crystalline samples of 1, 2, 3, and 4 were measured in
-
C(CH3)3
I3
4
Fig. 1. Structures of compounds.

P.E. Hansen et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 98 (2012) 247–251
249
Table 1
13C chemical shifts and complexation shifts for complexes between I2 and pyridine or 4TBP in the liquid and the solid state.
Carbon
C-2
2 (CD3CN)
2 (solid)
150.8
4TBP (CD3CN)
150.7
1 (CD3CN)
1 (solid)
3 (CD3CN)
3 (Solid)
4 (CD3CN) Aa
4 (Solid) Ba
147.6
4 (Solid) Aa
150.9 (1.3)b
150.2 (ꢀ0.5)
149.6
147.9c (ꢀ2.8)
[150.2]d (ꢀ0.5)
123.9 c(2.1)
[126.0]d (4.2)
165.4 (4.5)
[n.o]d
150.9
[146]d
127.5
148.85c (ꢀ1.8)
[150.15]d (ꢀ0.5)
123.3c (1.5)
148.3
127.2
165.7
36.4
[150.7]d
126.4
C-3
C-4
C
129.1 (5.1)
128.2
143.2
121.8
160.9
35.3
126.0 (4.2)
168.4 (7.5)
36.6 (1.3)
126.7
165.7
36.4
[126.1]d (4.3)
163.9c (3.0)
143.6 (7.4)
168.8g
168.0
[165.7]d
37.0
[168.4]d (7.5)
35.9 (0.6)
–
–
37.1e
36.1 (0.8)
[36.0]f
31.2e
[n.o.]d
CH3
30.8
30.3 (ꢀ0.5)
30.5 (ꢀ0.3)
32.0
30.6 (ꢀ0.2)
32.1
32.2
[30.4]f
[30.3]d (ꢀ0.5)
[30.3] (ꢀ0.5)
a
4 is synthesized by methods A or B (see Experimental).
Values in brackets are the difference in chemical shift between the complex and the non-complexed heterocycle. Values used for pyridine are: C-2 (149.6 ppm), C-3
b
(124.2 ppm) and C-4 (136.2 ppm).
c
Major set of resonances.
Value in bracket the less intense set of resonances. For integration see Table S8.
The more intense resonance.
d
e
f
The less intense resonance. The ratio 2:1.
g
The resonance showed a multiplet splitting possibly due to through space couplings to 19F.
Fig. 2. Top: Observed IR absorbance spectrum of the red solid formed by the reaction between 4-tert-butylpyridine (4TBP) and diiodine in methanol solution (Sections 4-tert-
Butylpyridine diiodine CT complex and bis(4-tert-butylpyridine)iodonium triodide). Bottom: Graphical representation of vibrational transitions predicted for the bis(4-tert-
butylpyridine)iodonium cation (full line) and the 4-tert-butylpyridine/diiodine CT complex (4) (dashed line). The computed wavenumbers are scaled by the factor
a = 0.975
[19] and a Lorentz line shape with HWHM = 3 cmꢀ1 is assumed. The relative IR intensities predicted for 4 are multiplied by a factor of 3, and peak wavenumbers for 4 are
indicated by asterisks (ꢄ).
KBr tablet using standard procedures. The recorded spectra are
provided as Supplementary data S1–S4. The spectrum shown in
Fig. 2(top) can be assigned to a mixture of 3 and 4; a full version
of this spectrum is provided as S3. In addition, the IR spectrum of
4TBP was recorded in CHCl3 solution (S5).
Calculations
All calculations were performed with the Gaussian09 software
package [23] by using the B3LYP density functional [24,25] with the
DGDZVP all electrons basis set [26,27] as previously discussed [20].

250
P.E. Hansen et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 98 (2012) 247–251
Table 2
13C titration data for 4TBP and I2 in CD3CN.
Carbon
4TBP
4TBP: I2 (10:1)
4TBP: I2 (2:1)
4TBP: I2 (1:1)
C-2
C-3
C-4
C
150.7
121.8
160.9
35.3
150.3 (ꢀ0.4)a
122.0 (0.2)
161.4 (0.6)
35.4 (0.1)
148.9 (ꢀ1.8)
122.9 (1.1)
163.7 (2.8)
35.8 (0.5)
148.7b (ꢀ2.0) [150.1]c (ꢀ0.6)
123.5 (1.7) [126.0] (4.2)
164.2b
35.9 (0.6) [35.9]d (0.6)
30.5 (ꢀ0.3) [30.3] (ꢀ0.5)
CH3
30.8
30.7 (ꢀ0.1)
30.6 (ꢀ0.2)
a
b
c
Values in brackets are titration shifts.
Resonance is split into two.
Values in square brackets are due to a minor species.
Assumed overlapping with major resonance.
d
In one series of calculations, molecular equilibrium geometries
iodonium ion (4TBP)2I+ and the main component is therefore iden-
tified as 3. The shoulder at 164 ppm is identified as the C-4 peak in
the charge transfer complex 4 (see below). Two sets of 13C reso-
nances of 3 and 4 were also observed in 13C NMR in CD3CN (see
Table 1) and DMF-d7 (see Table S7) with the 168 ppm peak (3)
much smaller than the 164 peak (4). The ratio ꢅ95:5 between 3
and 4 was obtained from the 1H NMR spectrum in CD3CN shown
in S8. Finally, the IR spectrum of the red powder in the solid state
shows two sets of characteristic bands which can be assigned to
overlapping contributions from 3 and 4, in excellent consistency
with the predicted vibrational transitions (Fig. 2).
The charge transfer complex 4 may be synthesized either by pro-
tocol A or B (see 4-tert-butylpyridine diiodine CT complex section).
In synthesis A, reaction between pyridine and iodine was per-
formed in ethanol instead of methanol and I2 was slowly added
drop wise to the 4TBP solution with formation of an orange–yellow
precipitate. Both this product and the orange crystals isolated from
the synthesis B reaction in hexane/CH2Cl2 showed a sharp peak in
solid state 13C NMR at 165 ppm. The IR of both products showed
two characteristic bands close to 1012 and 1068 cmꢀ1 (S4). Elemen-
tal analysis of the orange crystals showed a 1:1 ratio between 4TBP
and I2. Based on this evidence, both products were identified as 4.
and vibrational transitions (harmonic approximation) were com-
puted in the gas phase. The predicted fundamental vibrational
transitions for (4TBP)2I+ and 4TBPꢁI2 (4) are visualized in
Fig. 2(bottom).
In a second series, corresponding calculations were performed
within the polarized continuum model (PCM) [28] in order to sim-
ulate the influence of an acetonitrile solvent (Gaussian09:
scrf = (solvent = acetonitrile)) [23]. These calculations were carried
out primarily to investigate the applicability of the free energies of
reaction,
D
G, derived from the results of the harmonic analyses in
values and
Gaussian09 (298.15 K, 1 atm) [23]. Predicted
D
G
equilibrium constants K for several reactions relating to the
4TBP–iodine system in acetonitrile solution are provided as Sup-
plementary data (S6).
Judging from comparison with experimental estimates, the pre-
dicted equilibrium constants K seem to be reliable within one or
two orders of magnitude. For example, the computed K for the equi-
librium Iꢀ¡I2 þ Iꢀ in acetonitrile solution is of the order 10ꢀ5 (S6),
3
compared with experimental estimates in the range 10ꢀ6–10ꢀ7
[29,30]. Similarly, the predicted K for the equilibrium 4TBP + I2
¡
4TBPꢁI2 is close to 101 (S6), compared with empirical values in the
range 102–103 [7,31]. Hence, the reaction data computed at the
present, modest level of sophistication may be useful in a qualita-
tive manner. The computed K values for the reactions indicated in
Eqs. (1) and (2) are close to 10ꢀ4 and 10ꢀ11, respectively (S6).
NMR nuclear shieldings were calculated in the gas phase in the
GIAO approximation [31].
Nuclear shieldings
13C NMR nuclear shieldings are calculated using DFT calcula-
tions (see Experimental). The following species are calculated:
4TBP, 4, (4TBP)2I+, and 4TBPI+ (see Table 3). The experimental 13C
shifts of 1 and 3 are nearly identical and the shifts of the (4TBP)2-
I+ ion are therefore essentially independent of whether BF4ꢀ or Iꢀ3 is
the counter ion. It is obvious that the calculated trends are slightly
different for 4TBPꢁI2 and (4TBP)2I+ as compared to 4TBPI+. The data
for the former have the correct signs as compared to the experi-
mental results (Table 1), but the values are overestimated for
(4TBP)2I+.
Results and discussion
Structures
The elemental analysis of 1 and 2 clearly showed a ratio of two
pyridines to each iodine. The 13C NMR spectra of 1 in both CD3CN
and DMF-d7 showed very clear cut spectra showing that complex
formation leads to shifts in CD3CN of 7.5 ppm (C-4), ꢀ0.5 ppm
(C-2) and 4.2 ppm (C-3) relative to 4-tert-butylpyridine. Very sim-
ilar numbers are found for the pyridine complex, 2 (see Table 1).
The shifts for 1 in solution can be compared with those observed
by solid state NMR and from Table 1 it is seen that very good agree-
ment is found.
Titrations with I2 and Iꢀ
3
Titration of 4TBP with I2 in CD3CN leads to complexation shifts
as seen in Table 2 due to the formation of the CT complex 4 by Eq.
Table 3
Calculated 13C nuclear shielding
r
in ppm.a
Schuster and Roberts found for the iodonium salts of pyridine
and its 2,6- and 2,4,6-trimethyl derivatives the following changes
of the 13C-shifts in nitromethane relative to the starting pyridines:
C-2 (0.6, 2.9 and 2.2 ppm), C-4 (6.8, 6.6 and 9.2 ppm) and C-3 (5.1,
5.3 and 5.0 ppm). These values are comparable to those observed
for salts 1 and 2. The iodonium ion (4TBP)2I+ may therefore be
identified by the large C-4 shift at 168 ppm.
The elemental analysis of 3 showed a ratio close to one pyridine
for two iodine molecules. The solid state 13C spectrum of the red
powder had a sharp peak at 168 ppm with a small shoulder at
164 ppm. The shift at 168 ppm shows the presence of the
Carbon
4TBP
4TBPꢁI2
(4TBP)2I+
4TBPI+
C-2
C-3
C-4
C
24.9
57.2
17.7
145.0
148.6
28.5 (3.6)b
54.9 (ꢀ2.3)
12.9 (ꢀ4.8)
144.5 (ꢀ0.5)
148.9 (0.3)
29.0 (4.1)
21.3 (-3.6)
51.1 (ꢀ6.1)
2.2 (ꢀ15.5)
143.1 (ꢀ1.9)
149.5 (0.9)
46.0 (ꢀ11.2)
ꢀ6.9 (ꢀ24.6)
140.8 (ꢀ4.2)
149.4 (0.8)
CH3
a
The nucleus shielding can be converted to chemical shifts with the equation
+ 176.38 (n = 15 r2 = 0.9957).
d = ꢀ0.9735
r
b
Values in brackets are complexation shifts calculated as the difference between
the complex and 4TBP. Notice the different signs as compared to Table 1 because in
this table we are operating with nuclear shielding.

P.E. Hansen et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 98 (2012) 247–251
251
(3). The NMR spectra were recorded in CD3CN in order to mimic
the 3-methoxypropionitrile of the solar cell, and low concentra-
tions of 4TBP (0.30 M) are used in order to eliminate the influence
of ring current effects. For the ratios I2:4TBP of 0.1 and 0.5 only one
set of resonances are seen, indicating that the forward and reverse
reactions of Eq. (3) are fast compared with the NMR time scale,
with the result that only the weighted 13C-shifts of 4TBP and 4
are observed. When an equivalent amount of I2 is added to 4TBP
in CD3CN, a major and a minor set of 13C resonances are seen (Ta-
ble 2). The 13C chemical shifts of the minor component is identical
to the 13C shifts of 3 in CD3CN. When the I2: 4TBP ratio is increased
from 0 to 1 the C-4 shift changes from 160 to 165 ppm which is
very close to the C-4 shift 165 ppm of the orange charge transfer
product 4 in CD3CN (Table 2). The 165 ppm shift of the 1:1 titration
is therefore likely to be close to the end titration value indicating
that more than 90% of the 4TBP is converted to the CT complex
4. A mole fraction of 4 > 0.9 corresponds to a CT equilibrium con-
stant K3 of Eq. (3) > 400 in reasonable agreement with the
K = 400 value found the reaction between 4-picoline and I2 in
nitrobenzene [32]. Attempts to titrate with excess I2 were unsuc-
cessful due to precipitation of 3.
DSC mechanisms based on the reactions indicated in Eqs. (1)
and (2).
Acknowledgements
The authors thank Rita Buch and Annette Christensen for their
help in recording NMR spectra and Eva M Karlsen for the IR spec-
tra. Jørgen Skibsted is acknowledged for kindly recording the 13C
NMR solid state spectra and Ole Hammerich for his help.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.saa.2012.08.006.
References
[1] A. Hagfeldt, G. Boschloo, L.C. Sun, L. Kloo, H. Pettersson, Chem. Rev. 110 (2010)
6595–6663.
[2] M. Grätzel, Acc. Chem. Res. 42 (2009) 1788–1798.
[3] A. Hagfeldt, M. Grätzel, Acc. Chem. Res. 33 (2000) 269–277.
[4] M.K. Nazeeruddin, A. Kay, I. Rodicio, R. Humphrybaker, E. Muller, P. Liska, N.
Vlachopoulos, M. Grätzel, J. Am. Chem. Soc. 115 (1993) 6382–6390.
[5] B. Oregan, M. Grätzel, Nature 353 (1991) 737–740.
[6] J.M. Kroon, N.J. Bakker, H.J.P. Smit, P. Liska, K.R. Thampi, P. Wang, S.M.
Zakeeruddin, M. Grätzel, A. Hinsch, S. Hore, U. Wurfel, R. Sastrawan, J.R.
Durrant, E. Palomares, H. Pettersson, T. Gruszecki, J. Walter, K. Skupien, G.E.
Tulloch, Prog. Photovoltaics 15 (2007) 1–18.
In contrast, titrating 4TBP with (C4H9)4N+, I3ꢀ leads to no change
of the 13C chemical shifts in CD3CN or in 4-methoxypropionitrile
(Table 2). The presence of TiO2 or heating the NMR samples for
24 h at 85 °C does not make any difference either.
[7] G. Boschloo, L. Haggman, A. Hagfeldt, J. Phys. Chem. B 110 (2006) 13144–
13150.
Heterocyclic bases and triiodide
[8] H. Kusama, H. Arakawa, J. Photochem. Photobiol. A 160 (2003) 171–179.
[9] H. Kusama, H. Orita, H. Sugihara, Sol. Energy Mater. Sol. Cells 92 (2008) 84–87.
[10] H. Kusama, H. Arakawa, Sol. Energy Mater. Sol. Cells 85 (2005) 333–344.
[11] Z. Kebede, S.E. Lindquist, Sol. Energy Mater. Sol. Cells 57 (1999) 259–275.
[12] P.T. Nguyen, A.R. Andersen, E.M. Skou, T. Lund, Sol. Energy Mater. Sol. Cells 94
(2010) 1582–1590.
[13] B. Jones, G.J. Moody, J.D.R. Thomas, Inorg. Chem. 9 (1970) 114–119.
[14] C. Reid, R.S. Mulliken, J. Am. Chem. Soc. 76 (1954) 3869–3874.
[15] J.D. Schuster II, J. Roberts, Org. Chem. 44 (1979) 2658–2662.
[16] L.F. Audrieth, E.J. Birr, J. Am. Chem. Soc. 55 (1933) 668–673.
[17] T. Tassaing, M. Besnard, J. Phys. Chem. A 101 (1997) 2803–2808.
[18] R.D. Bailey, G.W. Drake, M. Grabarczyk, T.W. Hanks, L.L. Hook, W.T.
Pennington, J. Chem. Soc. Perkin Trans. 2 (1997) 2773–2779.
[19] O. Hassel, H. Hope, Acta Chem. Scand. 15 (1961) 407–416.
[20] E.M. Karlsen, J. Spanget-Larsen, Chem. Phys. Lett. 473 (2009) 227–232.
[21] L.M. Peter, J. Phys. Chem. C 111 (2007) 6601–6612.
What happens when a heterocyclic base and triiodide are mixed
as in the DSC solar cell? From Eq. (2) one could get the impression
that Iꢀ3 mixed with 4TBP can lead to formation of (4TBP)2I+ cations.
However, the 13C NMR spectra show clearly that no interaction
takes place as no changes in the chemical shifts are observed at
all, in sharp contrast to mixing 4TBP with I2 (Table 2). This obser-
vation is consistent with the prediction that the reaction in Eq. (2)
is strongly endothermic:
D
G ꢃ þ15 kcal molꢀ1, K ꢃ 10ꢀ11 (S6).
Conclusion
13C NMR is a powerful tool in elucidating the interactions be-
tween species in the solar cell as demonstrated in the present case
with 4TBP, but this could be any heterocyclic aromatic compound.
It is demonstrated that the different species (4TBP)2I+, I3ꢀ and
4TBPꢁI2 can be identified. However, it is also shown that Iꢀ3 does
not interact significantly with 4TBP in acetonitrile or 4-methoxy-
propinitrile, not even in the presence of titanium dioxide or at ele-
vated temperature. The investigation is supported by the results of
DFT calculations, which are very useful for obtaining information
on species that cannot immediately be observed (such as 4TBPI+)
and in obtaining approximate thermodynamic data. The combined
experimental and theoretical results of the present investigation
demonstrate the need for a revision of the previously suggested
[22] J.M. Chalker, A.L. Thompson, B.G. Davis, Org. Synth. 87 (2010) 288–298.
[23] M. Frisch et al., Gaussian 09, Revision A02, Gaussian Inc., Wallingford, CT,
2009.
[24] A.D. Becke, J. Chem. Phys. 98 (1993) 5648–5652.
[25] C.T. Lee, W.T. Yang, R.G. Parr, Phys. Rev. B 37 (1988) 785–789.
[26] N. Godbout, D.R. Salahub, J. Andzelm, E. Wimmer, Can. J. Chem. 70 (1992) 560–
571.
[27] C. Sosa, J. Andzelm, B.C. Elkin, E. Wimmer, K.D. Dobbs, D.A. Dixon, J. Phys.
Chem. 96 (1992) 6630–6636.
[28] J. Tomasi, B. Mennucci, R. Cammi, Chem. Rev. 105 (2005) 2999–3093.
[29] J. Datta, A. Bhattacharya, K.K. Kundu, Bull. Chem. Soc. Jap. 61 (1988) 1735–
1742.
[30] G. Boschloo, A. Hagfeldt, Acc. Chem. Res. 42 (2009) 1819–1826.
[31] E.K. Plyler, R.S. Mulliken, J. Am. Chem. Soc. 81 (1959) 823–826.
[32] G. Posskin, P. Huyskens, Bull. Soc. Chim. Fr. (1976) 337–344.