Tightly Bound Double-Caged [60]Fullerene Derivatives with Enhanced Solubility: Structural Features and Application in Solar Cells

Article abstract of DOI:10.1002/asia.201700194

A series of novel highly soluble double-caged [60]fullerene derivatives were prepared by means of lithium-salt-assisted [2+3] cycloaddition. The bispheric molecules feature rigid linking of the fullerene spheres through a four-membered cycle and a pyrrolizidine bridge with an ester function CO₂R (R=n-decyl, n-octadecyl, benzyl, and n-butyl; compounds 1 a–d, respectively), as demonstrated by NMR spectroscopy and X-ray diffraction. Cyclic voltammetry studies revealed three closely overlapping pairs of reversible peaks owing to consecutive one-electron reductions of fullerene cages, as well as an irreversible oxidation peak attributed to abstraction of an electron from the nitrogen lone-electron pair. Owing to charge delocalization over both carbon cages, compounds 1 a–d are characterized by upshifted energies of frontier molecular orbitals, a narrowed bandgap, and reduced electron-transfer reorganization energy relative to pristine C₆₀. Neat thin films of the n-decyl compound 1 a demonstrated electron mobility of (1.3±0.4)×10^?3 cm² V^?1 s^?1, which was comparable to phenyl-C₆₁-butyric acid methyl ester (PCBM) and thus potentially advantageous for organic solar cells (OSC). Application of 1 in OSC allowed a twofold increase in the power conversion efficiencies of as-cast poly(3-hexylthiophene-2,5-diyl) (P3HT)/1 devices relative to the as-cast P3HT/PCBM ones. This is attributed to the good solubility of 1 and their enhanced charge-transport properties — both intramolecular, owing to tightly linked fullerene cages, and intermolecular, owing to the large number of close contacts between the neighboring double-caged molecules. Test P3HT/1 OSCs demonstrated power-conversion efficiencies up to 2.6 % (1 a). Surprisingly low optimal content of double-caged fullerene acceptor 1 in the photoactive layer (≈30 wt %) favored better light harvesting and carrier transport owing to the greater content of P3HT and its higher degree of crystallinity.

Full text of DOI:10.1002/asia.201700194

CHEMISTRY
AN ASIAN JOURNAL
www.chemasianj.org
Accepted Article
Title: Tightly bound double-caged [60]fullerene derivatives with
enhanced solubility: structural features and application in solar
cells
Authors: Victor A. Brotsman, Vitaliy A. Ioutsi, Alexey V. Rybalchenko,
Vitaliy Yu. Markov, Nikita M. Belov, Natalia S. Lukonina,
Sergey I. Troyanov, Ilya N. Ioffe, Vasiliy A. Trukhanov, Galina
K. Galimova, Artur A. Mannanov, Dmitry N. Zubov, Erhard
Kemnitz, Lev N Sidorov, Tatiana V. Magdesieva, Dmitry Yu.
Paraschuk, and Alexey A Goryunkov
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Chem. Asian J. 10.1002/asia.201700194
Link to VoR: http://dx.doi.org/10.1002/asia.201700194
A Journal of
A sister journal of Angewandte Chemie
and Chemistry – A European Journal

10.1002/asia.201700194
Chemistry - An Asian Journal
FULL PAPER
Tightly bound double-caged [60]fullerene derivatives with
enhanced solubility: structural features and application in solar
cells
Victor A. Brotsman,a Vitaliy A. Ioutsi,a Alexey V. Rybalchenko,a Vitaliy Yu. Markov,a Nikita M. Belov,a
Natalia S. Lukonina,a Sergey I. Troyanov,a Ilya N. Ioffe,a Vasiliy A. Trukhanov,b Galina K. Galimova,b
Artur A. Mannanov,b Dmitry N. Zubov,c Erhard Kemnitz,d Lev N. Sidorov,a Tatiana V. Magdesieva,a
Dmitry Yu. Paraschuk,*b Alexey A. Goryunkov*a
Abstract: A series of novel highly soluble double-caged
[60]fullerene derivatives were prepared by means of lithium salt-
assisted [2+3]-cycloaddition. The bispheric molecules feature rigid
linking of the fullerene spheres via a four-membered cycle and a
pyrrolizidine bridge with an ester function CO2R (R=n-decyl, n-
octadecyl, benzyl, and n-butyl: compounds 1a–d, respectively), as
demonstrated by NMR spectroscopy and X-ray diffraction. Cyclic
voltammetry studies revealed three closely overlapping pairs of
reversible peaks due to consecutive one-electron reductions of
fullerene cages, as well as an irreversible oxidation peak attributed
to abstraction of an electron from the nitrogen lone electron pair.
Due to charge delocalization over both carbon cages, compounds
1a–d are characterized by upshifted energies of frontier molecular
orbitals, narrowed band gap, and reduced electron transfer
reorganization energy compared to pristine C60. Neat thin films of the
n-decyl compound 1a demonstrated the electron mobility (1.3 ±
0.4)×10–3 cm2V–1s–1, comparable to PCBM and thus potentially
advantageous for organic solar cells (OSC).
intermolecular, due to large number of close contacts between the
neighboring double-caged molecules. Test P3HT/1 OSCs
demonstrated power conversion efficiencies up to 2.6% (1a).
Surprisingly low optimal content of double-caged fullerene acceptor
1 in the photoactive layer (ca 30 wt%) favored better light harvesting
and carrier transport due to larger content of P3HT and its higher
degree of crystallinity.
Introduction
In the past years, fullerene-based acceptors became an
established standard for organic solar cells (OSCs). Energy
conversion efficiency of the fullerene-based cells has been
improved to 11%[1,2] via optimization of the donor and acceptor
components and fabrication protocols. In particular, important
progress have been achieved in development of novel low
bandgap conjugated polymers[3] and in appropriate tuning of
fullerene cages via derivatization.[4] The desired increase in
open-circuit voltage of OSCs can be achieved via up-shifting of
the fullerene LUMO levels by means of exohedral
functionalization with electron-donating groups, or with other
addends of non-acceptor nature via broadening of the HOMO–
LUMO gap.[5–10] However, excessive functionalization is known
to suppress the electron mobility. For example, OSCs based on
fullerene tris-adducts demonstrate a two-order drop in the
electron mobility and disappointingly low current density
compared to mono- and bisadducts.[11] An alternative approach
toward optimization of fullerene-based acceptor phases consists
in replacement of traditional C60 and C70 with higher
fullerenes,[12] endohedral fullerenes,[13] heterofullerenes,[14] or
homofullerenes,[15,16] but the resulting benefit of increased
optical density is largely outweighed by their higher cost and
limited availability.
Yet further effort to improve the properties of the photoactive
layers in OSCs is associated with novel, comparatively low
studied family of fullerene derivatives – the molecules consisting
of two or more covalently linked fullerene cages.[17] One can
expect that covalent linking of fullerene cages can result in a
beneficial effect on the electron mobility, as well as on the
ordering and morphological stability of the bulk heterojunction.
However, reduced solubility of oligomeric fullerenes[18]
considerably complicates OSC fabrication. Nevertheless, there
are several examples of successful application of double-caged
An application of compounds 1 in OSC allowed a two-fold
increase in power conversion efficiencies of as-cast P3HT/1 devices
compared to the as-casted P3HT/PCBM ones. This is attributed to
good solubility of 1 and their enhanced charge transport properties,
both intramolecular, due to tightly linked fullerene cages, and
[a]
V. A. Brotsman, Dr. V. A. Ioutsi, A. V. Rybalchenko, Dr. V. Yu.
Markov, N. M. Belov, Dr. N. S. Lukonina, Dr. Ilya N. Ioffe, Prof. Dr.
S. I. Troyanov, Prof. Dr. L. N. Sidorov, Prof. Dr. T. V. Magdesieva,
and Dr. A. A. Goryunkov*
Chemistry Department
Lomonosov Moscow State University
Leninskie Gory, 1-3, 119991, Moscow, Russia
E-mail: aag@thermo.chem.msu.ru
[b]
V. A. Trukhanov, G. K. Galimova, A. A. Mannanov, and Prof. Dr.
Dmitry Yu. Paraschuk*
Faculty of Physics & International Laser Center
Lomonosov Moscow State University
Leninskie Gory, 1-62, 119991, Moscow, Russia
E-mail: paraschuk@gmail.com
[c]
[d]
D. N. Zubov
Institute of Nanotechnology of Microelectronics RAS, Leninsky
Prospekt, 32A, 119991, Moscow, Russia
Prof. Dr. E. Kemnitz
Institut für Chemie
Humboldt Universität zu Berlin
Brook-Taylor-Str. 2, 12489 Berlin, Germany
Supporting information for this article is given via a link at the end of
the document.
For internal use, please do not delete. Submitted_Manuscript
This article is protected by copyright. All rights reserved.

10.1002/asia.201700194
Chemistry - An Asian Journal
FULL PAPER
fullerene derivatives as acceptor material. They include a
cyclopropanated derivative of furane-bridged dimer C120O (PCE
0.8%),[19] pyrazolino-pyrrolidino-cross-linked fullerene dimers
(PCE 0.9%),[20] benzene-linked fulleropyrrolidines (PCE up to
0.7%),[21] and fullerene dendrimers.[22] The best performance
values for that class of molecules was reported for dimeric
analogs of phenyl-C61-butyric acid methyl ester (PCBM) with
PCE 2.6–3.7%.[23–25] Addition of fullerene dimers to PCBM
improves the OSC performance, morphological stability and
lifetimes.[26] To achieve further performance enhancement, one
can consider improvement of electron transport by means of
tighter linking of fullerene cages.
Furthermore, the reaction products were observed to contain
related triple-caged compounds.[28,29] The linking moieties in
compounds 1 withdraw four sites on each of the cages from the
π-system and form the addition pattern known as cis-1.
Structurally similar cis-1 fullerene dimers have also been
obtained earlier via methanofullerene thermolysis[30] and solid-
state reaction of C60 with phthalazine.[31] However, compounds 1
have an advantage of bearing alkyl chains of variable length
attached to the pyrrolizidine bridge via an ester moiety. Those
chains enable tuning of physicochemical behavior, e.g. the
solubility, which is crucial for OSCs fabrication. Indeed, n-decyl
double-caged fullerene derivative 1a has demonstrated
enhanced solubility and reversible electrochemical reduction.[27]
Thus, double-caged fullerene derivatives 1 can be expected to
provide improvements both in carrier mobility and in bulk
heterojunction morphology. In the present paper, we report
synthesis of novel compounds of type 1 with enhanced solubility,
their structural elucidation according to NMR spectroscopy and
X-ray crystallography, electrochemical characterization, and
testing in polymer solar cells.
Results and Discussion
Synthesis and structural features. Lithium salts are known
to assist diastereoselective concerted 4π+2π cycloaddition of
azomethine ylides to a fullerene double bond that gives
fulleroproline esters with reasonable yield.[32] Unexpectedly, it
was found that ca 3-fold excess of paraformaldehyde enables
bis-hydroxymethylation of the glycine ester leading to ultimate
formation of double-caged fullerene derivatives with 58–71%
yield.[27] Those double-caged compounds consist of two C60
cages linked via rigid cyclobutane and carboxylated pyrrolizidine
moieties (1). Thus, ethyl, tert-butyl and n-decyl double-caged
fullerene derivatives 1 have been obtained and characterized.
Unfortunately, the advantageousness of geometric and
electronic structural features of those compounds in the context
of organic electronics is devalued by their insufficient solubility
(n-decyl derivative 1a being an exception).[27]
Figure 1. Structure of double-cage fullerene derivatives 1 and fulleroprolines
2, where R= n-decyl, n-octadecyl, benzyl, and n-butyl (1a–d and 2a–d,
respectively) are studied in this work.
Recently, we have presented a synthetic approach toward
novel pyrrolizidine/cyclobutane bridged double-caged fullerene
[27]
derivatives 1
that can be regarded as products of fusing a
(Figure 1).
second fullerene cage to fulleropyrrolidine
2
Table 1. Preparation of the double-caged fullerene derivatives 1 and fulleroproline esters 2a and solubility of 1.
GlyOR, R=
Yieldsb / %
1
S(1)c /
mg mL–1
2
a
n-decyl
49
12
35
b
c
n-octadecyl
47
44
11
15
45
20
benzyl
d
n-butyl
40
18
10
[a] Toluene solution, 1 eq. C60, 1.2 eq. GlyOR·TsOH, 10 eq. paraformaldehyde, 2.4 eq. Et3N, 1.2 eq. LiClO4, reflux for 3 h. [b] Yields as isolated. [c] Solubility (±5) of
double-caged fullerene derivatives 1 in oDCB at 25 oC.
For internal use, please do not delete. Submitted_Manuscript
This article is protected by copyright. All rights reserved.

10.1002/asia.201700194
Chemistry - An Asian Journal
FULL PAPER
In order to obtain better soluble double-caged compounds,
we have produced a novel series of compounds 1 with ester
alkyl chains and an ester benzyl group (1a–d, R=n-decyl, n-
octadecyl, benzyl, n-butyl), as well as the corresponding
fulleroproline esters (2a–d). The new compounds were prepared
using the same lithium salt assisted 1,3-dipolar cycloaddition
approach (see Table 1). Briefly, the respective glycine ester, 5-
fold molar excess of paraformaldehyde, Et3N, and LiClO4 were
added to a solution of C60 in toluene and refluxed for 3 h.
Separation by means of column chromatography afforded 1a–d
and 2a–d products with 95+% purity with a yield of 40–50% and
10–20%, respectively. Molecular composition of the isolated
products has been verified by MALDI and high-resolution ESI
mass spectra (see SI, Figure S7). The negative ion MALDI mass
spectra were quite typical for the similar double-caged
derivatives and fulleroproline esters.[28,29,27] They were
constituted by the molecular ([M]–) and fragment ([M–C60]– and
C60–) peaks in the cases of 1a–d, and the predominant molecular
peaks in the cases of 2a–d.
2.850(4), 2.933(4) and 3.260(4) Å (Figure 2c) that can be
regarded as additional channels of electronic communication.
Considered separately, the fullerene cages form nearly a two-
layer hexagonal close pack along the [105] direction in the
crystal lattice. The shortest intermolecular C···C distances
between double-caged molecules range from 3.03 to 3.32 Å and
thus can mediate π–π interactions. Each molecule is surrounded
by nine neighbouring molecules, whereas the nearest
surrounding in the solvent-free PCBM reaches six or seven
only.[40]
Structural elucidation of the compounds 1a–d and 2b–d was
performed using 13C and 1H NMR spectroscopy. Signal
1
assignment was supported by H–13C HMBC and HSQC NMR
data analysis, and the NMR data for the previously known
derivatives 1a and 2a were consistent with the earlier report.[27]
1
Briefly, H NMR spectra of the double-caged derivatives contain
an AB system due to the two equivalent CH2 groups of the
2
pyrrolizidine moiety (δH of 5.65–5.69 and 5.62–5.66 ppm, JHH
=
11–12 Hz, 4H) and the set of alkyl signals. The 13C NMR spectra
of the compounds consist of the set of signals due to the ester
moiety (C=O and OCH2 resonances were observed at δC of 170
and 65.8–67.8 ppm, respectively), the pyrrolizidine moiety
(resonances of two equivalent CH2 groups and sp3-C atom
bearing C(O)OR were observed at δC of 71.8–71.9 and 109.2–
109.6 ppm), and due to the two equivalent fullerene cages. The
56 x 2 sp2 atoms of the latter give rise to 49–53 discernible
signals in the range of 155–133 ppm while the remaining sp3
atoms are observed at 71–73 and 77 ppm (cyclobutane bridge)
and 69 and 85 ppm (pyrrolizidine bridge).
The structure of the double-caged benzyl ester 1c has been
also conclusively verified by means of single crystal X-ray
diffraction analysis (Figure 2).The intercage cyclobutane bonds
show noticeably different lengths of 1.553(4) and 1.601(4) Å
(Figure 2c), which is comparable with other cyclobutane-bridged
fullerene dimers (give values from 1.56 to 1.65).[33–36]
Interestingly, the pyrrolizidine moiety is found to be twisted in the
crystalline phase: pyramidalization angle (POAV)[37] of the
nitrogen atom is about 18° and ROOC–C–N–CH2 dihedral
angles have considerably different values of 97.6 and –138.9o
(Figure 2d), in agreement with the observations for some related
compounds,[38,39] while DFT optimization of the molecule rather
favors by ca 10 kJ mol–1 the Cs-symmetrical conformation with
nitrogen POAV angle of 5°. Due to tight linkage between the
cages, there are several additional non-bonded short contacts of
Figure 2. Molecular structure of compound 1c: general view (a) and details of
the linkage between the cages (b, c) (R denotes CO2Bn group).
As it had been anticipated, variation of the ester alkyl/benzyl
chain in compounds 1 was found to have pronounced effect on
the solubility (see Table 1). While many double-caged molecules
like C120, C120O, C120(CH2), C120(CH2)2, etc. show a drop in
solubility compared to pristine C60,[18] introduction of longer ester
alkyl chains in the compounds
1 increases their room
temperature solubility in o-DCB from 10 (n-butyl) to 45 (n-
octadecyl) mg mL–1. Within the synthetic scheme employed,
solubility enhancement is readily achieved by simply selecting
the desired precursory glycine ester.
Thermal stability of the reported compounds can be
exemplified by the thermogravimetry and differential scanning
calorimetry (TG/DSC) study of 1a in the temperature range of
o
40–500 C under inert atmosphere (see Figure S45). Between
o
300–500 C, there is a notable weight loss (–11%) that agrees
well with abstraction of the carboxydecyl moiety (CO2C10H21),
the associated endothermic effect peaking at 420 oC.
For internal use, please do not delete. Submitted_Manuscript
This article is protected by copyright. All rights reserved.

10.1002/asia.201700194
Chemistry - An Asian Journal
FULL PAPER
Figure 3. Registered (a) and deconvoluted (b) CV curves for 2a and 1a (3.0 and 2.1 mmol –1, 0.15 M Bu4NBF4, oDCB, 100 mV s–1, RT). PBE/TZ2P Kohn–Sham
MO energy levels (c) for C60, 1d, and 2d as well as LUMO and HOMO density distribution in 1d (d, e).
Electrochemical
and
spectroscopic
behavior.
three electrochemically reversible one-electron reduction steps
separated by 0.4–0.5 V. Compared to the parent C60, the first
reduction peak is shifted by –0.07 V. Oxidation of fulleroproline
esters is observed as an irreversible process at ca 1.0 V vs Fc+/0
and corresponds to abstraction of an electron from the nitrogen
lone electron pair,[43,44] the first oxidation potential value being
negatively shifted by ca 0.2 V with respect to that of PCBM.
Noteworthy, the second irreversible oxidation process is
observed at ca +1.2 V which is close to the first oxidation
potential of PCBM (+1.24 V vs Fc+/0). It has been reported that
controlled-potential electrolysis of fulleropyrrolidines and
methanofullerenes at their first oxidation potentials results in
retro-cycloaddition with methano bridge cleavage.[44,45]
Electrochemical behavior of the double-caged compounds 1 and
the related fulleroproline esters 2 was studied by means of cyclic
voltammetry (CV). Representative CV curves for the n-decyl
derivatives 2a and 1a are presented in Figure 3a (see SI for
highly similar CV curves of the rest of the compounds). Formal
redox potential values for the whole set of compounds are
summarized in Table 2. One can see that the effect of the alkyl
moiety on the reduction and oxidation potentials is negligible for
both types of compounds, and the redox behavior is thus
determined by the fullerene backbone.
The CV curves of fulleroproline esters 2a–d are similar to
the vast majority of fullerene monocycloadducts.[41–43] They show
[a]
Table 2. Formal reduction potentials (Epa+Epc)/2 and anodic peak values [Epa] for compounds 1, 2, PCBM and C60
.
(Epa+Epc)/2[b] vs Fc+/0 / V
Process
[c]
1a
1b
1c
1d
2a
2c
2d
PCBM[c]
C60
+/2+
0/+
[+1.22]
[+0.96]
–1.10
–1.19
–1.53
–1.62
–2.23
–2.31
[+1.23]
[+0.95]
–1.11
–1.20
–1.54
–1.63
–2.25
–2.32
[+1.20]
[+0.96]
–1.09
–1.17
–1.52
–1.60
–2.22
–2.30
[+1.18]
[+0.96]
–1.08
–1.17
–1.51
–1.60
–2.22
–2.26
–
[+0.91]
–1.13
–1.52
–2.06
–
[+1.15]
[+0.92]
–1.14
–1.52
–2.04
–
[+1.19]
[+0.96]
–1.14
–1.52
–2.05
–
–
[+1.24]
–1.14
–1.52
–2.03
–
–
[+1.29]
–1.07
–1.46
–1.93
–
0/–
–/2–
2–/3–
3–/4–
4–/5–
5–/6–
–
–
–
–
–
–
–
–
–
–
[a] Pt, oDCB, 0.15 M Bu4NBF4, vs Fc+/0, 100 mV s–1. [b] All potential values were determined according to semi-integration technique; peak potential values of
irreversible processes are given in parentheses. [c] The values are given according to Ref.[16]
For internal use, please do not delete. Submitted_Manuscript
This article is protected by copyright. All rights reserved.

10.1002/asia.201700194
Chemistry - An Asian Journal
FULL PAPER
Double-caged compounds 1a–d demonstrate two irreversible
oxidation peaks at the same positions as fulleroproline esters 2a–d
(ca 1.0 and 1.2 V vs Fc+/0). Similarly to fulleroproline esters,
oxidation is followed by compound degradation. The first oxidation is
localized on the nitrogen lone pair (which constitutes the HOMO,
Figure 3e), and the second involves the fullerene conjugated π-
system. With regard to reduction, however, compounds 1a–d
demonstrate more complicated electrochemical behavior than 2a–d.
The CV curves show three overlapping pairs of reversible one-
It has been shown that single-caged fullerene derivatives
have small electron transfer reorganization energies due to cage
rigidity and highly delocalized frontier MOs.[52–56] High rigidity of
the double-caged compounds 1 ensures low electron transfer
reorganization energy, which is beneficial in the context of
organic electronic devices. We estimated the reorganization
energy (λi) for n-type carrier transport as vertical ionization
energy of the anionic form minus vertical electron affinity of the
neutral molecule. It was found that the λi value for the double-
caged compound 1a is twice lower than in PCBM and C60 (0.06
eV vs 0.12 and 0.10 eV, respectively), since even lower per
atom charge density in compounds 1 causes less geometric
relaxation upon charging (see Figure 3d).
electron peaks, a typical feature of two coupled fullerene cages[30,46]
.
To verify one-electron nature of all reduction steps, the CV curves
were deconvoluted using semi-integration techniques[47] to simulate
the steady-state voltammetric curves (see Figure 3b). Indeed, three
peaks of equal intensity were obtained for single-caged
fulleroprolines, and three pairs of equal peaks for double-caged
compounds. The first reduction potential values of the compounds 1
fall in that for pristine C60 and PCBM, the potential values for
fulleroproline esters 2 and PCBM coincide. The potential difference
between the first and the second reduction steps for compounds 1 is
80–90 mV.
Reversibility of electrochemical behavior of compounds 1 is
mainly attributed to multiple linkage of its carbon cages. Indeed,
fullerene [2+2]-cycloadducts (C60)2 and (C60)C70 are known to readily
dissociate upon reduction to the anionic state,[33,48,49] and stability of
various dimeric anions correlates with the number of bonds
connecting
the
fullerene
cages
(C120<C120O<C120O2,
C120OS<C120(CH2)2).[30,50] Stable pyrrolizidine bridge that keeps the
two cages of compounds 1 close together even prevents dissociation
of the second, cyclobutane bridge upon negative charging. According
to our DFT calculations, conformations of compound 1a with closed
cyclobutane linker is more favorable by 29, 33, and 33 kJ mol–1 in the
dianionic, monoanionic and neutral states, respectively. Furthermore,
negative charging of 1a slightly decreases the closest non-bonded
contact between the cages from 2.834 Å for the neutral molecule to
2.825 and 2.822 Å for mono- and dianions, correspondingly. This is
possibly due to the fact that the LUMO of 1 that becomes occupied in
the anions forms certain σ-bonding density in the space between the
cages (see Figure 3d). Contrariwise, filling of the antisymmetrized
LUMO+1 with its slight σ-antibonding character upon 3-rd and 4-th
reductions mediates repulsion between the fullerene cages. Note that
the same observations have been reported for the C120O and cis-1-
C120(CH2)2 double-caged molecules.[51,30]
Figure 4. UV-Vis-NIR absorption spectra of compounds 1a and 2a in toluene
solution as indicated in the legend (285–900 nm). The enlarged regions of the
spectra demonstrating absorption edge are shown in the insets for compounds
1a (a) and 2a (b).
UV-Vis absorption spectra of compounds 1a and 2a are
shown in Figure 4 (see SI for the spectra of all compounds).
While fulleroproline esters exhibit peaks at 312 and 432 nm,
typically of monocycloadducts of C60,[32,57] the spectrum of the
double-caged compound shows features that are specific of cis-
1 fullerene bisadducts:[30,31,58] a shoulder at 324 nm and a
broader band at 432 nm. Noteworthy, the presence of two
fullerene chromophores results in higher molar extinction
coefficients in 1a compared to 2a.
Table 3. Electronic properties of the synthesized fullerene derivatives and related compounds
Experimental data[a], E in eV
DFT data[b], E in eV
Compound
Ref[c]
EC
EC
EC
Opt
EHOMO
ELUMO
EG
1.8
EG
EHOMO
ELUMO
EKS
1a–d
2a–d
C60
–5.6
–5.7
–6.0
–5.8
–4.9
–3.8
–3.7
–3.9
–3.7
–2.6
1.65
1.72
1.77
1.71
1.87
–5.5 [–5.5]
–5.6 [–5.7]
–6.0
–4.3 [–3.2]
–4.2 [–3.1]
–4.4
1.2 [2.3]
1.4 [2.5]
1.6
t.w.
t.w.
1.9
2.1
2.1
2.3
t.w., [5,59]
t.w., [16,24,60–62]
t.w., [16,63]
PCBM
–5.7
–4.2
1.5
P3HT
–
–
–
[a]Electrochemical (EC) HOMO and LUMO energies were calculated using the following equations E(HOMO)EC = –e(Eoxonset+4.8) eV and E(LUMO)EC = –
e(Eredonset+4.8) eV. Optical (Opt) gap EG=1240/λonset [b]PBE/TZ2P values; B3LYP/Def2-TZVP//PBE/TZ2P values are given in square parentheses. [c]t.w. – this
.
work.
For internal use, please do not delete. Submitted_Manuscript
This article is protected by copyright. All rights reserved.

10.1002/asia.201700194
Chemistry - An Asian Journal
FULL PAPER
Table 3 presents a summary of the electronic properties of
compounds 1 and 2 together with the reference data for PCBM,
C60, and P3HT. The optical band gap values (EGopt) were
estimated from the respective absorption onsets observed at ca
750 and 720 nm, respectively (Figure 4, insets). Intercage
interactions in the double-caged molecules results in a slight
narrowing of the band gap by 0.12 and 0.07 eV compared to
with pristine C60 and fulleroproline esters, respectively. The
electrochemical estimates of the frontier MO levels energy were
based on the onsets of the first reduction and oxidation peaks of
the compounds. While the LUMO energies of both compounds
1a–d and 2a–d remain quite close to that of PCBM, the HOMO
levels are shifted 0.2 eV higher since they are associated with
the nitrogen lone pair that is specific to the present compounds.
With the LUMO position comparable to PCBM, the novel
compounds are applicable as prospective acceptor materials for
photovoltaic applications.
optimal donor-to-acceptor weight ratio reported for P3HT/dimeric
fullerene blends varied between 0.5:1 and 2:1.[20,21] In the
present work, we varied the key fabrication parameters including
the ratio of the BHJ components, the active layer thicknesse,
and the annealing protocol. We used n-decyl double-caged
compound 1a as a test system due to its balanced alkyl chain
length, which provides sufficient solubility and miscibility with the
polymer phase while avoiding such possible adverse effects of
longer alkyl chains as higher risk of phase segregation, reduced
electronic coupling with P3HT, and electron mobility drop in the
blends.[65–67] For comparison, we also tested reference OSCs
based on blends of P3HT with PCBM or single-caged n-decyl
fulleroproline 2a that were taken in a typical donor-to-acceptor
weight ratio of 1.25:1.[4,68]
The electron mobility (μe) of neat thin films of PCBM and
double-caged derivative 1a was measured using space charge
limited current (SCLC) technique in the electron-only devices
(see Figures S37, S38 and Table S1 in SI). The μe value for
PCBM was (1.4 ± 0.5) × 10–3 cm2 V–1 s–1, in good agreement
with the previously reported data.[26,68–70] Neat thin films of
double-caged derivative 1a demonstrated virtually the same μe
value (1.3 ± 0.4) × 10–3 cm2 V–1 s–1. Therefore, one can expect
that PCBM and 1a will give the close electron collection
efficiencies in OSC. To our knowledge, this μe value is the
highest among any double-caged or oligomeric fullerene
derivatives studied so far.[26,71,24]
Photovoltaic
performance.
Solubility of fullerene
derivatives is an important characteristic that influences
morphology of polymer-fullerene blends and their photovoltaic
performance. The optimal solubility of the acceptor compound in
ortho-dichlorobenzene (oDCB) to ensure its better dispergation
in the P3HT polymer matrix and consequent improvement of the
bulk heterojunction (BHJ) morphology was reported to be in the
range of 25–80 mg mL–1.[64] Compounds 1, in contrast to many
other double-caged molecules,[18] demonstrate very good
solubility in organic solvents due to their ester alkyl chains. In
particular n-decyl, n-octadecyl, and benzyl double-caged
fullerene derivatives 1a–c show solubility in oDCB of above 20
mg mL–1 and are therefore suitable for OSC fabrication.
In the first set of the experiments, the optimal donor-to-
acceptor ratio (D/A) was determined by testing as-cast OSC
devices with P3HT/1a active layers of equal thickness (110–130
nm) in the D/A range of 5:1 to 1:1. Figure 5 shows the trends in
the key photovoltaic parameters and J–V curves for the
fabricated OSC devices (see Tables S2, S3 and Figures S39,
S41 in SI for further details).
Compared to the single cage fullerene derivatives, the
double-caged molecules have significantly different molecular
volume and shape. This may affect the BHJ morphology and
hence the optimal processing protocol of OSC. For example, the
Figure 5. Left: Photovoltaic properties of P3HT/1a-based OSC as a function of D/A ratio: (a) open-circuit voltage Voc, (b) short-circuit current density Jsc, (c) fill
factor FF, and (d) power conversion efficiency PCE. Right: Current density–voltage characteristics of as-prepared and pre-annealed (110 oC, 5 min) P3HT/1a and
P3HT/PCBM devices.
For internal use, please do not delete. Submitted_Manuscript
This article is protected by copyright. All rights reserved.

10.1002/asia.201700194
Chemistry - An Asian Journal
FULL PAPER
Figure 6. (a) Optical absorption spectra of as-cast thin films of P3HT/1a (1.25:1 and 2.5:1 w/w, film thicknesses of 150 and 200 nm, respectively) and
P3HT/PCBM (1.25:1 w/w, 150 nm). (b) EQE spectra of as-cast and annealed P3HT/1a (2.5:1 w/w) and P3HT/PCBM (1.25:1) OSC. (c) The annealing effect on the
molecular order degree of the P3HT phase of thin films of P3HT/1a (2.5:1 w/w) and P3HT/PCBM (1.25:1 w/w). AFM height images of as-cast thin films of 2.5:1
P3HT/1a (d) and 1.25:1 P3HT/PCBM (е).
For 1a, the highest PCE values of 1.9–2.2% were found for
the D/A ratio of 2.0 and 2.5 (see Table 4). Note that a high
optimal D/A ratio of 2:1 was reported for OSC devices based on
P3HT and pyrazolino-pyrrolidino-cross-linked fullerene dimers,
but their PCE was notably lower (0.91 %).[20] Remarkably, 1a
provides a considerable gain in PCE compared to as-cast
PCBM-based reference device due to a two-fold increase in the
short-circuit current density (Jsc), while both the fill factor (FF)
and the open-circuit voltage (Voc) values are comparable (Figure
5, left). Larger content of 1a in the BHJ results in a significant
decrease of Voc and FF, possibly due to higher carrier
recombination losses and non-optimal BHJ morphology. The
lower content of 1a, in its turn, provides lower Jsc values,
possibly due to unbalanced electron and hole carrier mobilities
in the BHJ layer. Such effects of the D/A ratio on the carrier
mobility were studied for P3HT/PCBM OSC.[68]
performance was studied in the range of 60 to 200 nm at a fixed
D/A ratio of 2.5:1 (see Figure S43). The highest PCE of ca 2.2%
was observed at 120 nm. The devices with thinner or thicker
active layers demonstrated 1.5- to 4-fold drop in the PCE due to
lower Jsc and FF values. This drop could be assigned to less
efficient photocurrent generation and higher charge
recombination rate as a result of impaired morphology. The
observed effect of the photoactive layer thickness on the OSC
performance as well as optimal thickness of the P3HT/1a layer
are close to those reported for P3HT/PCBM ones.[72]
The above effects of D/A ratio and layer thickness on the
OSC performance can be associated not only with the BHJ
morphology but also with BHJ optical absorption (the latter being
not fully independent of the former). In order to study this issue,
we measured the absorption coefficients in P3HT/1a and
P3HT/PCBM films (Figure 6a). In the range 2.0–2.8 eV, the
P3HT/1a thin film shows a slightly higher absorptivity than the
The effect of the photoactive layer thickness on the OSC
For internal use, please do not delete. Submitted_Manuscript
This article is protected by copyright. All rights reserved.

10.1002/asia.201700194
Chemistry - An Asian Journal
FULL PAPER
P3HT/PCBM film of the same D/A weight ratio (1.25:1), and
upon increase of the P3HT/1a ratio to 2.5:1 the absorption
coefficient grew considerably. Correspondingly, in the fullerene
absorption region (3.0–4.0 eV), the trend was opposite. Note
that the P3HT/1a thin film demonstrated a more diffuse
lineshape in the fullerene absorption region compared to a
pronounced band at 3.7 eV for P3HT/PCBM, which is an
agreement with the solution-phase UV spectra: peak at 312 nm
in PCBM vs. shoulder at 324 nm in 1a (Figure 4).
There is also a redshift of the 2.5:1 P3HT/1a spectrum
compared to that of the 1.25:1 film indicating higher crystallinity
of the P3HT phase in the former. It is known, that the fullerene
compounds tend to disturb crystallization of P3HT in the course
of film formation,[68] and P3HT crystallinity in the 2.5:1 blend
probably benefits from lower content of 1a. This assumption was
supported by the Raman data for as-cast thin films of P3HT/1a
(2.5:1) and P3HT/PCBM (1.25:1). These data revealed,
respectively, 70 and 45% degree of the P3HT molecular order
(Figure 6c) that was calculated from the ratio of the amorphous
and semicrystalline polymer phase in the blends.[73,74] Thus, the
2.5:1 P3HT/1a blend ratio is more optimal from the viewpoint of
better light harvesting and higher P3HT crystallinity.
performance degradation: there was a significant drop in the Voc
and FF values that prevailed over an increase in Jsc (see Tables
S2, S3 in SI). Upon annealing of the final devices at a lower
temperature (110 oC), the performance degradation was less
pronounced, and in several cases we even observed slight
improvement in PCE (Table S2). Pre-annealing at 110 oC of only
the photoactive layer provided more significant improvement in
the PCE: the value of 2.6% was achieved via an increase in Jsc
and FF. Finally, a combination of pre- and post-annealing
treatment of the 1a-based devices was found to be more or less
equivalent to sole post-annealing. These annealing data are
indicative of some effect of the top electrode material on the
annealing process of the adjacent part of the active layer.
Noteworthy, both pre- and post-annealing of the reference
P3HT/PCBM devices at 160 oC rather provided a significant gain
in PCE (from ca 1.2% to 3.5%) due to an increase in Jsc and FF,
in good agreement with literature data (e.g. [4]). The effect of
annealing on Jsc and EQE was much higher in the reference
PCBM devices than in the 1a-based ones (Figure 6b). In
particular, the 1a-based device initially demonstrated
a
maximum on the EQE curve at 35%, which increased to 45%
upon annealing, while the reference PCBM device showed
about a double increase from 23 to 45%. Those results possibly
reflect the fact that the crystallinity of P3HT in the pre-annealed
1a-based device was already quite high, in contrast with rather
amorphous P3HT phase in the reference P3HT/PCBM film. This
conclusion is in accordance with the optical absorption and
Raman data (Figure 6a,c).
Thermal annealing of the P3HT/PCBM thin films is a
common processing step to optimize the BHJ morphology and
the device performance [1,72,75]. In view of that, we attempted to
further improve the OSC performance by means of (a) post-
annealing of the final OSC devices, (b) pre-annealing of the
photoactive layer and (c) combining the two approaches. Post-
annealing of the final P3HT/1a devices at 140 oC resulted in
Table 4. Photovoltaic parameters and AFM morphology data of as-cast and annealed OSCs.
Annealing protocol[b]
Acceptor
1a
D/A[a]
2.5:1
VOC / V
JSC / mA cm–2
FF / %
PCE / %
Rq / nm
hmax / nm
Treatment
T / o
C
t / min
–
pre
post
–
–
–
5
0.603±0.006
0.600±0.003
0.514±0.002
0.575±0.004
0.571±0.002
0.325±0.004
0.547±0.005
0.610±0.002
0.592±0.014
0.481±0.010
0.646±0.006
0.627±0.005
6.11±0.08
7.11±0.08
6.42±0.10
3.47±0.08
3.61±0.30
5.97±0.15
4.41±0.09
5.11±0.06
5.92±0.07
5.40±0.11
3.21±0.08
8.27±0.18
60.5±1.0
61.0±0.6
53.1±0.9
51.3±0.8
53.5±0.3
49.3±0.3
53.4±0.3
61.5±2.4
55.5±1.8
55.7±1.6
59.1±3.0
67.3±1.5
2.22±0.05
2.61±0.06
1.75±0.03
0.99±0.02
1.08±0.09
0.94±0.01
1.15±0.03
1.91±0.07
1.94±0.06
1.44±0.06
1.23±0.06
3.52±0.11
7.2±0.2
6.5±0.2
5.7±0.2
5.6±0.2
5.3±0.2
5.3±0.2
–
35.5±1.1
33.1±1.3
28.2±1.2
29.2±1.5
28.5±1.8
25.6±0.9
–
110
110
–
5
1b
2.5:1
–
pre
post
pre/post
–
110
110
110
–
5
5
5/5
–
1c
2.5:1
6.0±0.5
–
28.7±3.0
–
pre
post
–
110
110
–
5
5
5.3±0.2
1.19±0.1
0.98±0.2
26.1±1.4
8.1±0.9
5.1±0.3
PCBM
1.25:1
–
pre
160
5
[a] Donor-to-acceptor (D/A) weight ratio. [b] Annealing temperature, time, and treatment protocol.
For internal use, please do not delete. Submitted_Manuscript
This article is protected by copyright. All rights reserved.

10.1002/asia.201700194
Chemistry - An Asian Journal
FULL PAPER
Morphology of the as-cast and pre-annealed thin films of
P3HT/1a and P3HT/PCBM blends with optimal D/A ratios were
further studied by means of optical and atomic force microscopy.
The optical microscopy analysis of the films was not revealed
any considerable macroscopic phase segregation in both the as-
cast and annealed samples (see Figure S46 in SI). The AFM
analysis of the as-cast P3HT/1a thin films demonstrated more
pronounced phase separation compared to the as-cast
P3HT/PCBM films (Figure 6d). The root-mean-square
roughness (Rq) and the maximal domain size (hmax) were 7.2
and 36 nm, respectively, compared to much flatter P3HT/PCBM
film surface with Rq and hmax values of 1.2 and 8 nm (Table 4).
Annealing of the 1a-based devices at a moderate temperature
performance than the rest of the double-caged compounds
tested. Hence, there is a significant impact of the ester alkyl
chain on the optoelectronic properties of the BHJ layer. The
highest soluble and, as was expected, the best dispersible in
P3HT n-octadecyl derivative 1b has demonstrated the lowest
PCE value associated with poor Jsc and EQE. One can
hypothesize that the protruding n-octadecyl chains can decrease
the BHJ electron mobility, possibly via shielding of the fullerene
cages and increased intermolecular distance. Furthermore,
those longer chains likely require additional thermal energy to
self-organize, as suggested by the particularly pronounced
results of annealing of the 1b-based device. On the other hand,
more compact benzyl group in 1c also proved somewhat
disadvantageous. Analogous observations have been previously
o
(110 C) resulted in a partial decrease in Rq and hmax to 5.7 and
28 nm, respectively, i.e. it did not induce additional phase
segregation. Still, the annealed P3HT/PCBM became much
smoother with Rq of 1.0 nm and hmax of 5.1 nm. Thus, P3HT/1a
blends demonstrate disadvantageously coarser nanoscale
morphology than the reference P3HT/PCBM thin films that
develop sufficiently large interfacial area.
made for phenyl-C71-butyric esters (PC70BR, Rm=n-CmH2m+1,
m=1–9) where the n-heptyl ester has resulted in the highest
PCE.[65]
Conclusions
In contrast to the photovoltaic performance of the double-
caged fullerene derivative (1a), the related n-decyl fulleroproline
(2a) demonstrated disappointingly low Voc, Jsc, and FF values
and the PCE was only 0.5% (Table S3). AFM analysis of the as-
cast and annealed P3HT/2a thin films revealed somewhat
smoother active layer surface compared to 1a-based device (Rq
and hmax values of 3.9–4.5 and 21–23 nm, respectively, Table
S6 and Figures S47, S49, S50). Still, the advantages of the
fused double-caged structure of 1a turn out to completely
outweigh the effects of coarser morphology.
The double-caged molecules 1b and 1c were tested using
the same parameters as optimized with 1a. Our attempt to vary
the D/A ratio in 1c-based OSC device only confirmed the optimal
ratio of 2.5:1. As with 1a, as-cast, pre-annealed, and post-
annealed devices were studied (see Tables 4 and S4 for the
results). The as-cast devices demonstrated higher Jsc values
compared to the reference as-cast P3HT/PCBM-based device,
but they were 43% (1b) and 16% (1c) less than those of the as-
cast 1a-based device. Pre-annealing at a moderate temperature
of 110 oC provided slight increase in the OSC via moderate
growth of Jsc, but that was mostly countered by a decrease in Voc
and FF. As a result, the as-cast and annealed devices have
reached almost identical maximal PCE of 1.0 (1b) and 1.9%
(1c). Interestingly, prolonged annealing of the 1b-based device
led to much higher growth of Jsc than in the cases of 1a and 1c.
The trends in the EQE spectra of the three compounds were
mostly similar (Figure S44).
Restricted solubility of oligomeric fullerenes is known to
hamper their applicability in organic electronics. Improvement in
solubility can be achieved via additional functionalization of the
fullerene cages. Unfortunately, such modifications can adversely
affect the electron mobility as well as ordering and morphology
of the bulk heterojunction, so it is necessary to restrict the
number of addends and to choose the most efficient ones in
terms of solubility and miscibility enhancement.
The double-caged fullerene derivatives of the present study
constitute an example of such functionalization. Variation of the
ester alkyl chain significantly improves solubility without affecting
the frontier MOs, the LUMO position being advantageously
slightly upshifted with respect to C60. Rigid pyrrolizidine and
cyclobutane moieties provide tight linkage of the two fullerene
cages thus enhancing intramolecular electronic communication
and ensuring small reorganization energy. Furthermore, neat
thin films of the double-caged compounds demonstrate electron
mobility comparable with that of PCBM.
Optimization of the solar cells based on P3HT and the
present double-caged compounds has yielded the PCE up to
2.6% observed for the compound bearing a n-decyl ester
function. Surprisingly low optimal content of double-caged
fullerene acceptor in the photoactive layer favors better light
harvesting and carrier transport due to larger content of P3HT
and its higher degree of crystallinity. Thus, as-cast OSC devices
show twice better performance compared to the as-cast PCBM-
based reference devices. Noteworthy, the optimal content of the
double-caged acceptors in the BHJ solar cells was found to be
surprisingly low (ca 28 wt.%, D/A ratio of 2.5:1). To the best of
our knowledge, this is the lowest value among the fullerene
derivatives employable as acceptors in OSC. Unfortunately, we
cannot presently relate that with confidence to any particular
effect of morphological of other effects. To clarify this, more
detailed studies of the active layer structure and properties are
needed that are beyond the current work.
Optical microscopy did not reveal any macroscopic phase
segregation of P3HT/1b and P3HT/1c in both as-cast and
annealed devices. AFM analysis revealed comparably coarse
morphology: the RMS roughness Rq and the maximal domain
size hmax were 5.3–6.0 and 26–29 nm, respectively, with rather
low effect of annealing on the nanoscale morphology (Table 4,
S6 and Figures S48, S49).
Thus, despite essential similarity in the electronic structure,
absorption spectra, and morphology of the thin films, the n-decyl
derivative 1a has been found to show much better OPV
All in all, the reported results suggest good potential of the
For internal use, please do not delete. Submitted_Manuscript
This article is protected by copyright. All rights reserved.

10.1002/asia.201700194
Chemistry - An Asian Journal
FULL PAPER
n-Octadecyl
(RS)-1''-
novel doubled-cage fullerene derivatives with customizable alkyl
chain as acceptor materials for organic electronic devices.
Compared to widely used single-caged fullerene derivatives, the
novel compounds enhance performance of the as-cast OSC
devices. Further studies in chemical modification of the double-
caged compounds and consideration of further oligomeric
acceptors with three or more fused fullerene units promise a
good opportunity of further improvement of organic solar cells
performance.
azabicyclo[3'',3'',0]octan[3'',4'':1,9;6'',7'':1',9']cyclobuta[2,12:2',12']di(C60
-
Ih)fullerene-5''-carboxate (1b). Yield 47%. tR 9.6 min (Cosmosil
Buckyprep 4.6 mm I.D. × 25 cm, toluene, 1 mL min–1). λmax (toluene, nm):
328, 432. m/z (MALDI-TOF-MS, negative ion mode): 1791.3
(C142H41NO2–, M–, 20%), 1071.3 (C82H41NO2–, {M-C60}–, 11%), 720.0
–
(C60
,
100%). HRMS (ESI) calculated for {M+H}+, C142H42NO2,
1792.3210, found 1792.3204. 1H NMR (600 MHz, 1,2-C6D4Cl2, ppm): δH
5.69 (2H, d, 2JHH 11.8 Hz, CH2N), 5.66 (2H, d, 2JHH 11.8 Hz, CH2N), 4.12
3
(2H, t, JHH 6.4 Hz, CH2O), 1.44 (2H, m, CH2), 1.34 (2H, m, CH2), 1.27
(2H, m, CH2), 1.22 (20H, m, CH2), 1.13 (4H, m, CH2), 1.04 (6H, m, CH2),
0.88 (3H, t, 3JHH 7.1 Hz, CH3). 13C NMR (150 MHz, 1,2-C6D4Cl2, ppm): δC
169.94 (C=O), 154.56, 153.04, 149.40, 149.14, 148.93, 148.48, 148.28,
148.19, 148.02, 147.22, 147.06, 147.01, 146.93, 146.45, 146.39, 146.31,
146.16, 145.71, 145.66, 145.58, 145.09, 145.07, 144.91, 144.90, 144.85,
144.80, 144.61, 144.53, 144.48, 144.30, 144.27, 144.19, 144.11, 144.04,
143.90, 143.86, 143.55, 143.48, 142.97, 142.83, 142.59, 142.49, 142.37,
142.29, 141.45, 141.32, 141.17, 138.69, 137.77, 137.36, 137.23, 132.77
(sp2-С, cage С60), 109.64 (ССО2), 84.73 (sp3-C, cage С60), 77.20 (sp3-C
of cyclobutan moiety, cage С60), 71.88 (CH2N), 70.88 (sp3-C of
cyclobutan moiety, cage С60), 69.46 (sp3-C, cage С60), 66.15 (CH2O),
32.31 (CH2), 30.34 (CH2), 30.21 (CH2), 30.17 (CH2), 30.12 (CH2), 29.91
(CH2), 29.83 (CH2), 29.54 (CH2), 28.91 (CH2), 26.68 (CH2), 23.10 (CH2),
14.42 (CH3).
Experimental Section
Chromatographically purified fullerene C60 (Fullerene center,
99.98%) was used as starting material for synthesis of double-caged
fullerene derivatives 1 and fulleroproline esters 2. The PCBM fullerene
derivative (HPLC purity 99.5%) used as a reference acceptor material in
the control series of bulk heterojunction solar cells was purchased from
Solenne BV. Two samples of regioregular poly(3-hexylthiophene-2,5-diyl)
donor polymer (P3HT) received from Sigma-Aldrich (Mw=54–75 kDa and
Mw/Mn≤2.5) and Rieke Metals Inc (model no. 4002-EE, Mw=57 kDa and
Mw/Mn=2.4)
were
used
as
donor
materials.
Poly(3,4-
ethylenedioxythiophene)poly(styrenesulfonate), PEDOT:PSS HTL Solar
(Baytron P VP Al 4083) used as hole transport layer was obtained from
H.C. Starck. Solvents were dried by the usual methods and distilled
before use. Unless otherwise specified, reagents were used as received
without further purification.
Benzyl
(RS)-1''-
azabicyclo[3'',3'',0]octan[3'',4'':1,9;6'',7'':1',9']cyclobuta[2,12:2',12']di(C60
-
Ih)fullerene-5''-carboxate (1c). Yield 44%. tR 13.4 min (Cosmosil
Buckyprep 4.6 mm I.D. × 25 cm, toluene, 1 mL min–1). λmax (toluene, nm):
328, 432. m/z (MALDI-TOF-MS, negative ion mode): 1629.1
–
(C131H11NO2–, M–, 40%), 909.1 (C71H11NO2–, {M-C60}–, 6%), 720.0 (C60
,
Synthesis. The double-caged fullerene derivatives
1
and
100%). HRMS (ESI) calculated for {M+H}+, C131H12NO2, 1630.0863,
fulleroproline esters 2 were prepared via lithium salt assisted [2+3]-
cycloaddition.[27] Typically, triethylamine (46 μL, 0.34 mmol) was added to
a suspension of glycine ester p-toluenesulfonate (0.34 mmol) in CH2Cl2
(2 mL). Obtained solution was added to fullerene C60 (200 mg, 0.28
mmol) dissolved in the toluene (150 mL). Then paraformaldehyde (84
mg, 2.80 mmol), lithium perchlorate (36 mg, 0.34 mmol) and
triethylamine (46 μL, 0.34 mmol) were added, and the mixture was
refluxed for 3–5 h. The compounds 1 and 2 were isolated by means of
column chromatography (sequentially employing toluene/hexane 2:1 v/v
mixture, toluene, and then toluene/ethyl acetate, 5:1 v/v). Finally,
products were HPLC purified and characterized by means of MALDI MS
and NMR spectroscopy (see supporting information for further details).
found 1630.0857. 1H NMR (600 MHz, 1,2-C6D4Cl2, ppm): δH 7.10 (3Н, m,
2
para-ArH and meta-ArH), 6.97 (2Н, m, ortho-ArH), 5.65 (2H, d, JHH 11.5
Hz, CH2N), 5.62 (2H, d, 2JHH 11.5 Hz, CH2N), 5.13 (2H, s, CH2С6Н5). 13
C
NMR (150 MHz, 1,2-C6D4Cl2, ppm): δC 170.1 (C=O), 154.47, 152.93,
149.19, 149.11, 148.91, 148.4, 148.24, 148.15, 147.99, 147.19, 146.98,
146.93, 146.39, 146.37, 146.28, 146.15, 145.61, 145.55, 145.37, 145.04,
144.85, 144.82, 144.78, 144.77, 144.56, 144.51, 144.44, 144.26, 144.18,
144.16, 143.07, 143.98, 143.88, 143.83, 143.81, 143.51, 143.41, 142.94,
142.78, 142.56, 142.51, 142.46, 142.31, 142.26, 141.38, 141.27, 141.12,
138.54, 137.86, 137.36, 137.18, 132.79 (sp2-С, cage С60), 134.95 (Ph,
ipso-С), 129.25 (Ph, ortho-С), 128.70 (Ph, meta-С), 125.63 (Ph, para-С),
109.18 (ССОО), 84.78 (sp3-Ccage), 77.15 (sp3-C of cyclobutane moiety,
cage С60), 71.89 (CH2N), 71.16 (sp3-C of cyclobutan moiety, cage С60),
69.31 (sp3-Ccage), 67.81 (CH2С6Н5).
n-Decyl
(RS)-1''-
azabicyclo[3'',3'',0]octan[3'',4'':1,9;6'',7'':1',9']cyclobuta[2,12:2',12']di(C60
-
Ih)fullerene-5''-carboxate (1a). Yield 49%. tR 11.2 min (Cosmosil
Buckyprep 4.6 mm I.D. × 25 cm, toluene, 1 mL min–1). λmax (CH2Cl2, nm;
n-Butyl
(RS)-1''-
ε / L cm–1 mol–1): 328 (67580), 432 (10290), 716 (250). m/z (MALDI-TOF-
azabicyclo[3'',3'',0]octan[3'',4'':1,9;6'',7'':1',9']cyclobuta[2,12:2',12']di(C60
-
–
,
M–, 42%), 959.2
MS, negative ion mode): 1679.2 (C134H25NO2
Ih)fullerene-5''-carboxate (1d). Yield 40%. tR 13.1 min (Cosmosil
Buckyprep 4.6 mm I.D. × 25 cm, toluene, 1 mL min–1). λmax (toluene, nm):
328, 432. m/z (MALDI-TOF-MS, negative ion mode): 1595.1
(C73H23NO2, {M–C60}–, 12%), 720.0 (C60–, 100%). 1H NMR (600 MHz,
1,2-C6D4Cl2, ppm): δH 5.69 (2H, d, 2JHH 11.8 Hz, CH2N), 5.65 (2H, d, 2JHH
11.8 Hz, CH2N), 4.12 (2H, t, 3JHH 6.4 Hz, CH2O), 1.35 (2H, m, CH2), 1.20
(4H, m, CH2), 1.11 (4H, m, CH2), 1.01 (6H, m, CH2), 0.83 (3H, t, 3JHH 7.0
Hz, CH3). 13C NMR (150 MHz, 1,2-C6D4Cl2, ppm): δC 169.92 (C=O),
154.55, 153.02, 149.36, 149.12, 148.91, 148.46, 148.25, 148.17, 148.00,
147.19, 147.03, 146.98, 146.91, 146.43, 146.36, 146.27, 146.14, 145.70,
145.63, 145.57, 145.06, 144.88, 144.83, 144.78, 144.59, 144.51, 144.44,
144.27, 144.17, 144.09, 144.00, 143.87, 143.83, 143.52, 143.45, 142.96,
142.80, 142.56, 142.46, 142.34, 142.28, 141.42, 141.29, 141.15, 138.65,
137.75, 137.33, 137.21, 132.75 (sp2-С, cage С60), 109.59 (ССОО), 84.71
(sp3-C, cage С60), 77.18 (sp3-C of cyclobutan moiety, cage С60), 73.21
(sp3-C of cyclobutan moiety, cage С60), 71.85 (CH2N), 69.44 (sp3-C, cage
С60), 66.12 (CH2O), 32.18 (CH2), 29.85 (CH2), 29.78 (CH2), 29.69 (CH2),
29.50 (CH2), 28.88 (CH2), 26.61 (CH2), 23.03 (CH2), 14.40 (CH3). IR
(KBr, cm–1): 2922, 2852, 1742, 1462, 1429, 1382, 1298, 1260, 1191,
1126, 1071, 797, 726, 526.
–
(C128H13NO2–, M–, 11%), 875.1 (C68H13NO2–, {M–C60}–, 9%), 720.0 (C60
,
100%). HRMS (ESI) calculated for {M+H}+, C128H14NO2, 1596.1019,
found 1596.1018. 1H NMR (600 MHz, 1,2-C6D4Cl2, ppm): δH 5.67 (2H, d,
2JHH 11.4 Hz, CH2N), 5.65 (2H, d, JHH 11.4 Hz, CH2N), 4.11 (2H, t, JHH
2
3
3
6.6 Hz, CH2O), 1.36 (2H, m, CH2), 1.08 (2H, m, CH2), 0.68 (3H, t, JHH
7.3 Hz, CH3). 13C NMR (150 MHz, 1,2-C6D4Cl2, ppm): δC 170.01 (C=O),
154.56, 153.03, 149.31, 149.13, 148.91, 148.46, 148.26, 148.16, 148.00,
147.19, 147.01, 146.99, 146.93, 146.41, 146.37, 146.28, 146.13, 145.71,
145.63, 145.57, 145.06, 144.89, 144.85, 144.83, 144.78, 144.59, 144.52,
144.45, 144.28, 144.25, 144.18, 144.09, 144.02, 143.88, 143.84, 143.53,
143.44, 142.94, 142.81, 142.57, 142.55, 142.47, 142.33, 142.27, 141.42,
141.29, 141.14, 138.63, 137.78, 137.36, 137.21, 132.74 (sp2-С, cage
С60), 109.55 (ССОО), 84.70 (sp3-C, cage С60), 77.16 (sp3-C of
cyclobutan moiety, cage С60), 71.82 (CH2N), 70.86 (sp3-C of cyclobutan
For internal use, please do not delete. Submitted_Manuscript
This article is protected by copyright. All rights reserved.

10.1002/asia.201700194
Chemistry - An Asian Journal
FULL PAPER
moiety, cage С60), 69.43 (sp3-C, cage С60), 65.67 (CH2O), 30.76 (CH2),
19.50 (CH2), 13.65 (CH3).
[13] R. B. Ross, C. M. Cardona, F. B. Swain, D. M. Guldi, S. G.
Sankaranarayanan, E. V. Keuren, B. C. Holloway, M. Drees, Adv. Funct.
Mater. 2009, 19, 2332–2337.
Single crystal X-ray crystallography. Single crystals of 1c were
formed via slow evaporation of solvent from toluene solution.
Synchrotron X-ray data were collected at 100 K on a MAR225 CCD
detector on BL14.3 at the BESSY II electron storage ring (Berlin,
Germany) (λ = 0.8950 Å). The structure weas solved and anisotropically
[14] W. Cambarau, U. F. Fritze, A. Viterisi, E. Palomares, M. von Delius,
Chem. Commun. 2015, 51, 1128–1130.
[15] S. H. Park, C. Yang, S. Cowan, J. K. Lee, F. Wudl, K. Lee, A. J. Heeger,
J. Mater. Chem. 2009, 19, 5624–5628.
refined using the SHELX package.[76] C131H11O2N∙0.48C7H8∙0.43C6H14
,
[16] V. A. Brotsman, V. A. Ioutsi, A. V. Rybalchenko, V. P. Bogdanov, S. A.
Sokolov, N. M. Belov, N. S. Lukonina, V. Y. Markov, I. N. Ioffe, S. I.
Troyanov, et al., Electrochim. Acta 2016, 130–142.
monoclinic, C2/c, a = 40.551(3), b = 16.4821(5), c = 21.648(2) Å, β =
110.12(1)°, V = 13585.8(18) Å3, Z = 8, R1(F)/wR2(F2) = 0.079/0.206 for
9149/13522 reflections and 1364 parameters. CCDC 1531487 contains
the supplementary crystallographic data for this paper. These data can
be obtained free of charge from The Cambridge Crystallographic Data
Centre via www.ccdc.cam.ac.uk/data_request/cif.
[17] F. Giacalone, N. Martín, Chem. Rev. 2006, 106, 5136–5190.
[18] J. L. Segura, N. Martín, Chem. Soc. Rev. 2000, 29, 13–25.
[19] L. Weber, S. Sensfuss, U. Ritter, P. Scharff, Chem. Lett. 2008, 37,
750–751.
Quantum Chemical Calculations. Molecular geometry of the each
structure in question was optimized at the AM1 level of theory with the
use of the Firefly 8.1 software[77] (partly based on the GAMESS (US)
source code)[78] which is followed by full DFT optimization using the
PRIRODA v.6 software that implements efficient RI approximation[79]
(PBE exchange-correlation functional[80] and a built-in basis set of triple
zeta quality with (11s6p2d)/[6s3p2d] contraction scheme for first row
atoms and (5s1p)/[3s1p] for hydrogen atoms were used). Further
refinement of the electron structure was performed using the Firefly 8.1
employing the hybrid functionals B3LYP[81,82] as well as Def2-TZVP basis
set.[83]
[20] J. L. Delgado, E. Espildora, M. Liedtke, A. Sperlich, D. Rauh, A.
Baumann, C. Deibel, V. Dyakonov, N. Matrin, Chem. Eur. J. 2009, 15,
13474–13482.
[21] S. Erten-Ela, C. Villegas, J. L. Delgado, N. Martín, New J. Chem. 2015,
39, 1477–1482.
[22] K. Hosomizu, H. Imahori, U. Hahn, J.-F. Nierengarten, A. Listorti, N.
Armaroli, T. Nemoto, S. Isoda, J. Phys. Chem. C 2007, 111, 2777–
2786.
[23] J. Liu, X. Guo, Y. Qin, S. Liang, Z.-X. Guo, Y. Li, J. Mater. Chem. 2012,
22, 1758–1761.
[24] J. Ge, J. Liu, X. Guo, Y. Qin, H. Luo, Z.-X. Guo, Y. Li, Chem. Phys. Lett.
2012, 535, 100–105.
[25] Y. Morinaka, M. Nobori, M. Murata, A. Wakamiya, T. Sagawa, S.
Yoshikawa, Y. Murata, Chem. Commun. 2013, 49, 3670–3672.
[26] B. C. Schroeder, Z. Li, M. A. Brady, G. C. Faria, R. S. Ashraf, C. J.
Takacs, J. S. Cowart, D. T. Duong, K. H. Chiu, C.-H. Tan, et al., Angew.
Chem. Int. Ed. 2014, 53, 12870–12875.
Acknowledgements
This work was partially supported by the Russian Foundation
for Basic Research (grant no. 15-33-21083-mol_a_ved, 14-03-
31815-mol_a, and 15-03-05083). We also thank Dr. T. B.
Shatalova for TG/DSC analysis. This work was done using
equipment purchased under the Lomonosov Moscow State
University Program of Development and MSU Supercomputer
Center facilities.[84]
[27] V. A. Ioutsi, V. Y. Markov, N. M. Belov, M. G. Apenova, A. A.
Goryunkov, M. E. Maksimova, O. M. Nikitin, S. V. Kovalev, V. E.
Shevchenko, T. V. Magdesieva, et al., New J. Chem. 2013, 37, 804–
809.
[28] V. Y. Markov, N. A. Samokhvalova, P. S. Samokhvalov, V. A. Ioutsi, P.
A. Khavrel’, N. S. Ovchinnikova, L. N. Sidorov, J. Anal. Chem. 2010, 65,
1495–1503.
[29] V. Y. Markov, A. Y. Borschevsky, L. N. Sidorov, Int. J. Mass Spectrom.
2012, 325–327, 100–112.
Keywords: fullerene dimers; polymer solar cells; cyclic
voltammetry; organic photovoltaics; molecular ordering;
morphology; X-ray diffraction
[30] N. Dragoe, H. Shimotani, M. Hayashi, K. Saigo, A. de Bettencourt-Dias,
A. L. Balch, Y. Miyake, Y. Achiba, K. Kitazawa, J. Org. Chem. 2000, 65,
3269–3273.
[31] Y. Murata, N. Kato, K. Komatsu, J. Org. Chem. 2001, 66, 7235–7239.
[32] V. A. Ioutsi, A. A. Zadorin, P. A. Khavrel, N. M. Belov, N. S.
Ovchinnikova, A. A. Goryunkov, O. N. Kharybin, E. N. Nikolaev, M. A.
Yurovskaya, L. N. Sidorov, Tetrahedron 2010, 66, 3037–3041.
[33] G.-W. Wang, K. Komatsu, Y. Murata, M. Shiro, Nature 1997, 387, 583–
586.
[1]
[2]
A. Polman, M. Knight, E. C. Garnett, B. Ehrler, W. C. Sinke, Science
2016, 352, aad4424-aad4424.
M. A. Green, K. Emery, Y. Hishikawa, W. Warta, E. D. Dunlop, D. H.
Levi, A. W. Y. Ho-Baillie, Prog. Photovolt. Res. Appl. 2017, 25, 3–13.
A. J. Heeger, Chem. Soc. Rev. 2010, 39, 2354–2371.
Y.-Y. Lai, Y.-J. Cheng, C.-S. Hsu, Energy Environ. Sci. 2014, 7, 1866.
H. Dong, H. Zhu, Q. Meng, X. Gong, W. Hu, Chem. Soc. Rev. 2012, 41,
1754–1808.
[3]
[4]
[5]
[34] D. V. Konarev, S. S. Khasanov, A. Otsuka, G. Saito, R. N. Lyubovskaya,
J. Am. Chem. Soc. 2006, 128, 9292–9293.
[35] A. A. Goryunkov, I. N. Ioffe, P. A. Khavrel, S. M. Avdoshenko, V. Y.
Markov, Z. Mazej, L. N. Sidorov, S. I. Troyanov, Chem. Commun. 2007,
704–706.
[6]
[7]
Y. Li, Chem. Asian. J. 2013, 8, 2316–2328.
Y. Zhang, Y. Matsuo, C.-Z. Li, H. Tanaka, E. Nakamura, J. Am. Chem.
Soc. 2011, 133, 8086–8089.
[36] N. S. Ovchinnikova, A. A. Goryunkov, P. A. Khavrel, N. M. Belov, M. G.
Apenova, I. N. Ioffe, M. A. Yurovskaya, S. I. Troyanov, L. N. Sidorov, E.
Kemnitz, Dalton Trans. 2011, 40, 959–965.
[8]
[9]
Y. Matsuo, Chem. Lett. 2012, 41, 754–759.
Y. He, H.-Y. Chen, J. Hou, Y. Li, J. Am. Chem. Soc. 2010, 132, 1377–
1382.
[37] R. C. Haddon, J. Phys. Chem. A 2001, 105, 4164–4165.
[38] C. Newton, J. van Ameijde, G. W. J. Fleet, R. J. Nash, D. J. Watkin,
Acta Crystallogr. Sect. E Struct. Rep. Online 2004, 60, o1463–o1464.
[39] E. T. S. Kamala, R. Murugan, S. Nirmala, L. Sudha, S. S. Narayanan,
Acta Crystallogr. Sect. E Struct. Rep. Online 2008, 64, o1958–o1959.
[40] G. Paternò, A. J. Warren, J. Spencer, G. Evans, V. G. Sakai, J.
Blumberger, F. Cacialli, J. Mater. Chem. C 2013, 1, 5619.
[10] Y. He, Y. Li, Phys. Chem. Chem. Phys. 2011, 13, 1970–1983.
[11] H. Kang, C.-H. Cho, H.-H. Cho, T. E. Kang, H. J. Kim, K.-H. Kim, S. C.
Yoon, B. J. Kim, ACS Appl. Mater. Interfaces 2012, 4, 110–116.
[12] F. B. Kooistra, V. D. Mihailetchi, L. M. Popescu, D. Kronholm, P. W. M.
Blom, J. C. Hummelen, Chem Mater 2006, 18, 3068–3073.
For internal use, please do not delete. Submitted_Manuscript
This article is protected by copyright. All rights reserved.

10.1002/asia.201700194
Chemistry - An Asian Journal
FULL PAPER
[41] T. Suzuki, Y. Maruyama, T. Akasaka, W. Ando, K. Kobayashi, S.
Nagase, J. Am. Chem. Soc. 1994, 116, 1359–1363.
[64] P. A. Troshin, H. Hoppe, J. Renz, M. Egginger, J. Y. Mayorova, A. E.
Goryachev, A. S. Peregudov, R. N. Lyubovskaya, G. Gobsch, N. S.
Sariciftci, et al., Adv. Funct. Mater. 2009, 19, 779–788.
[65] W. S. Shin, J.-C. Lee, J.-R. Kim, H. Y. Lee, S. K. Lee, S. C. Yoon, S.-J.
Moon, J. Mater. Chem. 2011, 21, 960–967.
[42] L. Echegoyen, Acc. Chem. Res 1998, 31, 593–601.
[43] I. A. Nuretdinov, V. V. Yanilkin, V. I. Morozov, V. P. Gubskaya, V. V.
Zverev, N. V. Nastapova, G. M. Fazleeva, Russ. Chem. Bull. Int. Ed.
2002, 51, 263–268.
[66] H.-M. Kim, S.-T. Kim, M. H. Seo, J.-W. Ka, K.-S. Lee, Y.-Y. Noh, N. S.
Baek, T.-D. Kim, J. Nanosci. Nanotechnol. 2014, 14, 2515–2519.
[67] R. Tao, T. Umeyama, K. Kurotobi, H. Imahori, ACS Appl. Mater.
Interfaces 2014, 6, 17313–17322.
[44] O. Lukoyanova, C. M. Cardona, M. Altable, S. Filippone, Á. Martín
Domenech, N. Martín, L. Echegoyen, Angew. Chem. Int. Ed. 2006, 45,
7430–7433.
[45] M. H. Hall, P. Shevlin, H. Lu, A. Gichuhi, C. Shannon, J. Org. Chem.
2006, 71, 3357–3363.
[68] M. T. Dang, L. Hirsch, G. Wantz, J. D. Wuest, Chem. Rev. 2013, 113,
3734–3765.
[46] N. Dragoe, H. Shimotani, J. Wang, M. Iwaya, A. de Bettencourt-Dias, A.
L. Balch, K. Kitazawa, J. Am. Chem. Soc. 2001, 123, 1294–1301.
[47] A. J. Bard, L. R. Faulkner, Electrochemical Methods: Fundamentals
and Applications, Wiley, New York, 2001.
[69] I. Riedel, E. von Hauff, J. Parisi, N. Martin, F. Giacalone, V. Dyakonov,
Adv. Funct. Mater. 2005, 15, 1979–1987.
[70] V. D. Mihailetchi, J. K. J. van Duren, P. W. M. Blom, J. C. Hummelen, R.
A. J. Janssen, J. M. Kroon, M. T. Rispens, W. J. H. Verhees, M. M.
Wienk, Adv. Funct. Mater. 2003, 13, 43–46.
[48] K. Komatsu, G.-W. Wang, Y. Murata, T. Tanaka, K. Fujiwara, K.
Yamamoto, M. Saunders, J. Org. Chem. 1998, 63, 9358–9366.
[49] K. Komatsu, K. Fujiwara, T. Tanaka, Y. Murata, Carbon 2000, 38,
1529–1534.
[71] T. Heumueller, W. R. Mateker, A. Distler, U. F. Fritze, R. Cheacharoen,
W. H. Nguyen, M. Biele, M. Salvador, M. von Delius, H.-J. Egelhaaf, et
al., Energy Env. Sci 2016, 9, 247–256.
[50] L. Dunsch, P. Rapta, A. Gromov, A. Staško, J. Electroanal. Chem. 2003,
547, 35–43.
[72] S. van Bavel, E. Sourty, G. de With, K. Frolic, J. Loos, Macromolecules
2009, 42, 7396–7403.
[51] A. L. Balch, D. A. Costa, W. R. Fawcett, K. Winkler, J. Phys. Chem.
1996, 100, 4823–4827.
[73] W. C. Tsoi, D. T. James, J. S. Kim, P. G. Nicholson, C. E. Murphy, D. D.
C. Bradley, J. Nelson, J.-S. Kim, J. Am. Chem. Soc. 2011, 133, 9834–
9843.
[52] H. Imahori, K. Hagiwara, T. Akiyama, M. Aoki, S. Taniguchi, T. Okada,
M. Shirakawa, Y. Sakata, Chem. Phys. Lett. 1996, 263, 545–550.
[53] D. M. Guldi, M. Prato, Acc. Chem. Res. 2000, 33, 695–703.
[54] H. Imahori, H. Yamada, D. M. Guldi, Y. Endo, A. Shimomura, S. Kundu,
K. Yamada, T. Okada, Y. Sakata, S. Fukuzumi, Angew. Chem. Int. Ed.
2002, 41, 2344–2347.
[74] J. Razzell-Hollis, S. Limbu, J.-S. Kim, J. Phys. Chem. C 2016, 120,
10806–10814.
[75] X. Yang, J. Loos, S. C. Veenstra, W. J. H. Verhees, M. M. Wienk, J. M.
Kroon, M. A. J. Michels, R. A. J. Janssen, Nano Lett 2005, 5, 579–583.
[76] G. M. Sheldrick, Acta Cryst. 2008, A64, 112–122.
[77] A. A. Granovsky, Firefly v. 8.1.1 (Formerly PC GAMESS), 2015.
[78] M. W. Schmidt, K. K. Baldridge, J. A. Boatz, S. T. Elbert, M. S. Gordon,
J. J. Jensen, S. Koseki, N. Matsunaga, K. A. Nguyen, S. Su, et al., J
Comput Chem 1993, 14, 1347–1363.
[55] S. Fukuzumi, K. Ohkubo, H. Imahori, D. M. Guldi, Chem. Eur. J. 2003,
9, 1585–1593.
[56] Y. Kawashima, K. Ohkubo, S. Fukuzumi, J. Phys. Chem. A 2013, 117,
6737–6743.
[57] M. Prato, M. Maggini, C. Giacometti, G. Scorrano, G. Sandona, G.
Farnia, Tetrahedron 1996, 52, 5221–5234.
[79] D. N. Laikov, Chem. Phys. Lett. 1997, 281, 151–156.
[80] J. P. Perdew, K. Burke, M. Ernzerhof, Phys. Rev. Lett. 1996, 77, 3865–
3868.
[58] K. Kordatos, S. Bosi, T. Da Ros, A. Zambon, V. Lucchini, M. Prato, J.
Org. Chem. 2001, 66, 2802–2808.
[81] A. D. Becke, J. Chem. Phys. 1993, 98, 5648–5652.
[82] P. J. Stephens, F. J. Devlin, C. F. Chabalowski, M. J. Frisch, J. Phys.
Chem. 1994, 98, 11623–11627.
[59] S. Leach, M. Vervloet, A. Desprès, E. Bréheret, J. P. Hare, T. J. Dennis,
H. W. Kroto, R. Taylor, D. R. M. Walton, Chem. Phys. 1992, 160, 451–
466.
[83] F. Weigend, R. Ahlrichs, Phys. Chem. Chem. Phys. 2005, 7, 3297–
3305.
[60] K. Yoshimura, K. Matsumoto, Y. Uetani, S. Sakumichi, S. Hayase, M.
Kawatsura, T. Itoh, Tetrahedron 2012, 68, 3605–3610.
[61] S. H. Yoo, J. M. Kum, S. O. Cho, Nanoscale Res. Lett. 2011, 6, 545.
[62] Y. He, G. Zhao, B. Peng, Y. Li, Adv. Funct. Mater. 2010, 20, 3383–
3389.
[84] V. Sadovnichy, A. Tikhonravov, V. Voevodin, V. Opanasenko, in
Contemp. High Perform. Comput. Petascale Exascale, CRC Press,
Boca Raton, USA, 2013, pp. 283 –307.
[63] C. Wang, F. S. Kim, G. Q. Ren, Y. Q. Xu, Y. Pang, S. A. Jenekhe, L.
Jia, J. Polym. Sci. Polym. Chem. 2010, 48, 4681–4690.
For internal use, please do not delete. Submitted_Manuscript
This article is protected by copyright. All rights reserved.