Naphthalene Diimide-Based Molecules for Efficient and Stable Perovskite Solar Cells

Article abstract of DOI:10.1002/ejoc.202000287

Naphthalene-diimide (NDI)-based molecules have shown an interesting behaviour within the field of perovskite solar cells, thanks to their promising application as electron transporting materials. In this paper, three novel NDI-containing molecules are synthesized and fully characterized, more specifically two polymers and an analogue small molecule. Each one of the NDI units contains an amine, either tertiary or quaternary, which is a moiety known for improving the conductivity. The newly synthesized compounds are suitable for the use as n-type buffer layers on the top of PC₆₁BM, thanks to their appropriate energy levels and their solubility in polar solvents. The photovoltaic performances of the NDI-containing cells are highly comparable to those of the reference cells, which contain bathocuproine (BCP) as buffer layer. Furthermore, the stability of the NDI-containing cells is higher than that of the BCP reference.

Full text of DOI:10.1002/ejoc.202000287

European Journal of Organic Chemistry
Accepted Article
Accepted Article
Title: Naphthalene Diimide-Based Molecules for Efficient and Stable
Perovskite Solar Cells
Authors: Silvia Valero, Andrea Cabrera-Espinoza, Silvia Collavini,
Jorge Pascual, Nevena Marinova, Ivet Kosta, and Juan Luis
Delgado
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: Eur. J. Org. Chem. 10.1002/ejoc.202000287
Link to VoR: https://doi.org/10.1002/ejoc.202000287
01/2020

10.1002/ejoc.202000287
European Journal of Organic Chemistry
FULL PAPER
Naphthalene Diimide-Based Molecules for Efficient and Stable
Perovskite Solar Cells
Silvia Valero,[a] Andrea Cabrera-Espinoza,[a] Silvia Collavini,*[a] Jorge Pascual,[b] Nevena Marinova,[a]
Ivet Kosta,[c] and Juan Luis Delgado*[a,d]
[a]
Dr. S. Valero, A. Cabrera-Espinoza, Dr. S. Collavini, Dr. N. Marinova, Prof. J. L. Delgado
POLYMAT, University of the Basque Country UPV/EHU, Avenida de Tolosa 72 & Faculty of Chemistry, P. Manuel Lardizabal 3, 20018
Donostia-San Sebastián, Spain
E-mail: silvia.collavini@polymat.eu, juanluis.delgado@polymat.eu
http://www.polymat.eu/en/groups/hybrids-materials-for-photovoltaics
Dr. Jorge Pascual
Helmholtz-Zentrum Berlin für Materialien und Energie GmbH,
Hahn-Meitner-Platz 1, 14109 Berlin, Germany
Dr. I. Kosta
CIDETEC, Basque Research and Technology Alliance (BRTA), P. Miramón 196, 20014
Donostia-San Sebastián, Spain
Prof. J. L. Delgado
[b]
[c]
[d]
Ikerbasque, Basque Foundation for Science,
Bilbao, Spain
Supporting information for this article is given via a link at the end of the document.
Abstract: Naphthalene-diimide (NDI) -based molecules have shown
an interesting behaviour within the field of perovskite solar cells,
thanks to their promising application as electron transporting
materials. In this paper, three novel NDI-containing molecules are
synthesized and fully characterized, more specifically two polymers
and an analogue small molecule. Each one of the NDI units contains
an amine, either ternary or quaternary, which is a moiety known for
improving the conductivity. The newly synthesized compounds are
suitable for the use as n-type buffer layers on the top of PC61BM,
thanks to their appropriate energy levels and their solubility in polar
solvents. The photovoltaic performances of the NDI-containing cells
are highly comparable to those of the reference cells, which contain
bathocuproine (BCP) as buffer layer. Furthermore, the stability of the
NDI-containing cells is higher than that of the BCP reference.
organic ETMs considered fullerene and derivatives for the
purpose.[5, 8] Their success is due to their excellent properties,
such as extremely efficient electron transport, suitable energy
alignment with perovskites, stability, and low processing
temperature. Moreover, they are well-known for reducing the
undesirable phenomenon of hysteresis thanks to their ability to
passivate trap states, and they can be processed at low
temperatures.[9] The most widely used fullerene derivative in
solar cells is Phenyl-C61-butyric acid methyl ester (PC61BM).[10-12]
It has the tendency to create a charge-transporting network
thanks to the phenyl moiety, which is inclined to undergo π–π
stacking with other PC61BM molecules. The highest occupied
molecular orbital (HOMO) and the lowest unoccupied molecular
orbital (LUMO) levels of PC61BM, which are -5.9 eV and -3.9 eV
respectively, make it suitable for extracting electrons and
blocking holes coming from methylammonium lead iodide
perovskite (CH3NH3PbI3, or MAPbI3), whose valence and
conduction bands are, respectively, -5.4 eV and -3.9 eV.[13]
However, PC61BM is not flawless. Thickness of PC61BM layer is
critical: the films made using this molecule are often too thin to
completely cover the perovskite, due to the small molecular size
of PC61BM and the low viscosity of the PC61BM solution, which
together with the poor solubility of the molecule in the
Introduction
In the last decade, perovskite solar cells (PSCs) went through a
surprisingly fast development, with their power conversion
efficiency (PCE) improving from 3.8% to the latest registered
2]
value of 25.2%.[1,
Each one of the components of the
perovskite device is under constant and extensive study in order
to get the best out of its performances. Together with the other
device layers, several charge transporting materials, both
organic and inorganic, have been employed during the years.[3-5]
Among all the structures that have been tested as ETMs, TiO2
has been widely used and it is still the dominant ETM in PSCs.[1]
However, it has a major drawback: it requires a high processing
temperature (~500 ºC), which implies the need of higher costs
but also impedes the development of flexible devices.[6]
Moreover TiO2 suffers UV instability, leading to the formation of
trap states, which can affect both efficiency and stability of PSCs.
To overcome these limitations, a variety of ETMs, including
inorganic, organic, or polymers, has been tested in the last
years.[7]
processing solvents could lead to the generation of small cracks
15]
in the films.[14,
These cracks can act as a shunting pass,
leading to losses in performance. On the other hand, an
excessive thickness confers too high resistivity to the solar cell.
Therefore, the morphology and the electrical properties of
PC61BM-based electron transporting layer (ETL) need to be
tuned in order to obtain high-performance PSCs. One efficient
way to improve the performances of PC61BM is to employ a
buffer layer between the fullerene and the metal cathode. The
most known buffer layer for PC61BM is bathocuproine (BCP).[11]
In the present work, an ammonium-bearing naphthalene diimide
polymer (P1) was synthesized, together with its ternary amine
polymer counterpart (P0) and its small molecule analogue (M1)
as a comparison. Naphthalene diimides (NDIs) are the smallest
component of the family of rylene diimides and consist in a
naphthalene core with imide substituents at the 1,4,5,8 positions
(Figure 1).[16]
Organic semiconductors are easy to prepare and manipulate
and their properties can be finely tuned through chemical
synthesis.[3] Until now, the vast majority of studies concerning
1
This article is protected by copyright. All rights reserved.

10.1002/ejoc.202000287
European Journal of Organic Chemistry
FULL PAPER
Figure 1: General structure of NDIs.
NDIs are chemically and functionally versatile: their naphthalene
core could be substituted at the positions 2, 3, 6, and 7 giving
derivatives with variable properties, and the imide unit could be
functionalised to modulate the solubility or to tether a specific
functional fragment. NDIs are electron-deficient aromatic
compounds, due to the strong electron-withdrawing diimide
groups. They are planar molecules with the ability to stack both
in an intra- and intermolecular arrangement, leading to potential
pathways for electron conduction. Moreover, they have good
electron mobility, excellent thermal and oxidative stability, and
they can be easily solution processed. All these properties,
together with their easy and low-cost synthesis make them
promising candidates for photovoltaic devices.[17] In fact, NDIs
have been used several times as ETMs for PSCs with
[18-22]
interesting results, both as small molecules
polymers.[23-25]
and
Regarding the latters, their film-forming property is generally
better than that of small molecules, thus providing better
processing window for the preparation of high-quality films,
based on scalable coating methods. The studies carried out until
now have highlighted that one of the features of an ideal NDI
polymer is a continuous conjugation, to guarantee good electron
mobility and reduce recombination.[25]
Last but not least, NDI-based ETLs have demonstrated to be
cheaper than the reference ETL, PC61BM.[26]
Figure 2: Structure of the target compounds P0, P1, and M1.
The novel molecules chosen for this study were incorporated as
buffer layers in order to improve the performances of PC61BM.
The high polarity of these novel compounds make them suitable
for the purpose, with no risk of dissolving the underlying PC61BM
layer during deposition. To the best of our knowledge, this is the
first time the potential of these useful structures is used to
improve the performances of the well-known PC61BM. The
performances of the NDI-containing devices were compared to
those of a cell containing BCP, as a reference. To conclude this
study, the stability of the cells was evaluated.
quaternary ammonium halides can effectively passivate trap
states in perovskites, with consequent reduction of the charge
recombination and improvement of the charge extraction at the
perovskite/ETM interface.[25, 27]
In addition, both tertiary and quaternary ammonium groups
improve the conductivity of polymeric n-type organic
semiconductors.[28] More specifically, regarding the quaternary
ammonium, as an immediate consequence the halogen anions
contribute to transfer electrons to the ETL, increasing its
conductivity and therefore the short-circuit current density (JSC
)
of the cell. Furthermore, the presence of a tertiary amine in most
of the doping agents induces self-doping of the layer, thanks to
the lone pair electron transfer to the ETL.[29]
Results and Discussion
Syntheses of the Target Compounds
To determine the role of the polymeric nature of the molecule, a
small molecule bearing the same chain as P1, named M1, was
synthesized. The synthesis of M1 is reported in Scheme 1.
To obtain molecule M1, the commercially-available 1,4,4,8-
naphthalenetetracarboxylic dianhydride (NDA) was treated with
3-(dibutylamino)propylamine in dimethylformamide (DMF) at
140 ºC under microwave irradiation to get molecule M0. This
was then treated with iodoethane in CH3CN at room temperature
in the absence of light for 5 days.
The structure of the target compounds P0, P1, and M1 is
reported in Figure 2. The backbone of polymers P0 and P1 is
composed of a benzene unit combined with different NDI-based
moieties. The benzene unit connected to the 2, 6-positions of
the NDI core should help obtain a more extended conjugation
than the one that would be generated through the bonding of the
sterically hindered NDI positions. Furthermore, to ensure
sufficiently high solubility for device fabrication and good film
formation, branched alkyl chains (2-ethylhexyl) were introduced
in the benzene unit. Polymers P0 and P1 hold respectively an
amino- or an ammonium-functionalized side chain at the imide
positions of the NDI core. It has been proven that both amino or
The preparation of polymers P0 and P1 started with the
synthesis of the monomers M2 and M3, depicted in Figure 3.
These monomers are engineered to be the proper reagents for
2
This article is protected by copyright. All rights reserved.

10.1002/ejoc.202000287
European Journal of Organic Chemistry
FULL PAPER
Scheme 1: Synthetic pathway to molecule M1.
ethylhexyl)benzene (M4), which was brominated leading to
compound M5, which after a Miyaura borylation gave monomer
M2.
The synthesis of monomer M3 is shown in Scheme S2. The
preparation of intermediate M6 was necessary, and it was
carried out following a reported procedure.[30] The reaction
between NDA and 1,3-dibromo-5,5-dimethylhydantoin (DBDMH)
in H2SO4 gave intermediate M6, which was then imidated with 3-
(dibutylamino)propylamine in acetic acid (AcOH) leading to
monomer M3.
The synthesis of polymer P0 was performed through Pd-
catalyzed Suzuki coupling copolymerization of the diboronic
ester monomer M2 with the NDI- based monomer M3 in THF
using NaHCO3 as base (Scheme 2). The resulting mixture was
purified through Soxhlet extraction in MeOH, CH3CN and
acetone giving polymer P0 as a product. Polymer P1 was then
obtained by quaternizing the precursor polymer P0 with an
excess of iodoethane in a mixture of DMF:THF solution at room
temperature in darkness for 5 days. Both syntheses are reported
in Scheme 2.
Figure 3: Structures of the monomers M2 and M3
the subsequent Pd-catalyzed Suzuki cross coupling reactions
that allowed the obtaining of the polymers. The synthetic
strategy to prepare monomer M2 is shown in Scheme S1. First,
the Kumada coupling of 3,4-dibromobenzene with 2-
(ethylhexyl)magnesiumbromide (RMgBr) provided 1,4-bis(2-
The fully detailed syntheses of all compounds can be found in
the Experimental Section.
Scheme 2: Syntheses of polymers P0 and P1
3
This article is protected by copyright. All rights reserved.

10.1002/ejoc.202000287
European Journal of Organic Chemistry
FULL PAPER
1
Characterization of the target compounds
Figure 5 shows the H-NMR spectra of polymers P0 and P1. It
can be observed how the signal between 2.20-2.70 ppm
(highlighted in blue) attributed to the hydrogen atoms closer to
the amine in polymer P0, is shifted between 2.80-3.74 ppm in
polymer P1, due to the deshielding effect produced by the
protonated amine. Moreover, in that region another signal
appeared, which is assigned to the proton close to the
quaternized amine (-NCH2CH3) in polymer P1. These are strong
evidences of the conversion of polymer P0 into polymer P1.
The molecular weight of polymers P0 was estimated by size-
exclusion chromatography (SEC) analysis in THF compared to a
polystyrene standard (Table 1). It showed a Mn value of
4350 g mol-1, leading to Xn values of 5. These values were
comparable to those obtained for other NDI-based polymers
described in literature.[32-36] Polymer P1 could not be analyzed
through SEC due to its insolubility in THF. However, it is safe to
assume that polymer P1 would have a molecular weight of the
same order as that of its neutral precursor P0. This assumption
is based on the process used to obtain P1, which only implies
the use of iodoethane without resorting to further polymerization
processes.
The chemical structure of both polymers P0 and P1 were
confirmed by 1H-NMR spectroscopy. Moreover, through the
mass spectrometry analysis of P0, it was possible to find peaks
which differ from each other by a value that corresponds to the
building block (m/z calculated for [C58H86N4O4]+: 902.664), as
shown in Figure 4.
7371.1
P0
6468.9
8273.7
5567.7
901.2
902.2
902.6
4000
6000
Mass (m/z)
Figure 4. MALDI-TOF of polymer P0.
8000
10000
Table 1. Molecular weights of polymer P0.
M
(g mol-1)
M
(g mol-1)
X
n
PDI
n
w
P0
4350
6953
5
1.60
Polymer P1 could not be analyzed by MALDI-TOF, most likely
due to the difficulty of polycationic polymers to be ionized.[31]
Figure 5. 1H-NMR spectra (300 MHz, CD2Cl2, 298 K) of polymers P0 (top) and P1 (bottom).
Optical properties
The optical properties of the compounds were inspected
resorting to UV-Vis absorption measurements. Figure S1 shows
the absorption spectrum of M1 recorded in DCM. As it can be
seen, the compound shows two intense absorption bands below
400 nm attributed to the π-π* transition (Table S1).[37] The
4
This article is protected by copyright. All rights reserved.

10.1002/ejoc.202000287
European Journal of Organic Chemistry
FULL PAPER
Table 2: Electrochemical properties and energy levels of compound M1. All
bandgap value could be estimated from the absorption onset of
the longest absorption wavelength and is as well reported in
Table S1.
the values are expressed in eV.
Red1
E1/2
LUMO
Bandgap
HOMO
M1
-0.94
-3.86
3.14
-7.00
The absorption spectra of polymers P0 and P1 in DCM are
depicted in Figure S2. The registered data are reported in Table
S2. The polymers displayed very low absorption in the visible
range, which is a favourable feature for their application in
Similar to what registered for the small molecule, the novel
polymers show two reversible reduction processes, due to the
NDI core (Figure 7). Also in this case, the LUMO value of the
ammonium-containing polymer P1 is lower than that of the
neutral counterpart, and again it could be explained with the
anion π-interactions between the NDI and the iodine anions. P1,
similarly to M1, suffered two non-reversible oxidation processes,
while polymer P0 does not exhibit any oxidation. This effect
could be explained by the lone pair electrons on the nitrogen
atom and its tendency to transfer onto the electron-deficient
NDIs and to lead to the formation of charge transfer complexes,
which was also observed in previous reports.[28, 44]
perovskite-based devices. Both polymers exhibit
a similar
absorption pattern with four distinct absorption bands. A broad
peak lies at the low-energy region (400-520 nm) and is related to
an intramolecular charge transfer (ICT) process between the
electron donor (benzene) and the NDI electron-deficient unit,
which was in accordance with previously reported NDI-donor
copolymers.[38-41] The ICT absorption peak of the ammonium-
functionalized copolymer P1 is slightly red-shifted compared to
that of P0, which could be due to the electrostatic perturbation
given by ions in close proximity to the optically active segment,
solvatochromic effects arising from the polar solvent, or
multichain aggregation.[28] The other three absorption bands lie
in the high-energy region (249-380 nm) and can be assigned to
π-π* transitions of the backbones. Among these, the absorption
band centred in between 249 and 253 nm is the most intense
and can be assigned to π-π* transitions of the benzene unit. The
other two absorption bands are characteristics of π-π*
transitions of the NDI core, similar to those registered for the
small molecule M1.
P0
P1
Electrochemical properties
The electrochemical properties of the compounds were
investigated by Cyclic Voltammetry (CV) measurements, which
were carried out in a 0.1 M solution of tetra-n-butylammonium
fluoride (TBAHFP) in anhydrous DCM. The LUMO values (Table
2 and 3) were obtained from the half-wave potential of the first
reduction (E1/2Red1) during CV measurements using the Fc/Fc+
redox couple as the internal standard. Therefore, the LUMO is
calculated according to the following equation:
-3
-2
-1
0
1
Voltage vs Fc/Fc+ (V)
Figure 7. Cyclic voltammograms of polymers P0 and P1 in DCM/TBAHFP (0.1
M) at a scan rate of 100 mV s-1.
Table 3. Electrochemical properties and energy levels of polymers P0 and P1.
All the values are expressed in eV.
Red1
Red1
LUMO (eV) = -4.8 – E1/2
.
E1/2
LUMO
-3.53
-3.75
Bandgap
2.41
HOMO
-5.94
P0
P1
-1.27
-1.05
M1 underwent two reversible reduction processes, which
indicated an initial radical anion and subsequent dianion
formation and are typical of the NDI derivatives (Figure 6).[35]
The low LUMO value registered for molecule M1 could be due to
2.35
-6.10
Figure 8 summarizes the energy levels of all the compounds
used in the device fabrication.[45] With respect to M1, the energy
levels are highly similar to those of the reference buffer layer
BCP, therefore the performances obtained with this molecule
are expected to be comparable to the reference. With these
LUMO values (summarized in Table 2 and 3), these novel NDI-
based small molecules and polymers are compatibles to be
used as buffer layers between the PC61BM ETL and the gold
cathode in inverted PSCs. Hence, these materials can behave
as suitable filters towards electrons, which is fundamental for
their synergetic action as buffer layer for PC61BM. On the
contrary, the hole-blocking synergetic action of the polymers is
expected to be weaker than those of BCP and M1, since their
higher-lying HOMOs. Furthermore, P0 is expected to behave
slightly worse than P1 due to both the higher LUMO and HOMO.
Nonetheless, the energy levels might be overall suitable for their
application in PSCs and highly compatible with PC61BM.
anion-π interactions between the electron deficient aromatic NDI
In addition, M1 suffered two
oxidation processes, which were non-reversible.
43]
core and the I- anions.[42,
M1
-2
-1
0
1
Voltage vs Fc/Fc+ (V)
Figure 6. Cyclic voltammogram of model compound M1 in DCM/TBAHFP (0.1
M) at a scan rate of 100 mV s-1.
5
This article is protected by copyright. All rights reserved.

10.1002/ejoc.202000287
European Journal of Organic Chemistry
FULL PAPER
split of the alkyl chains. This trend in the thermal properties was
also observed in other conjugated polymers with terminal amino
or ammonium side groups.[28, 46, 47]
The melting point molecule M1 was registered in order to get a
comparison with the polymers. The observed value was 218 ºC,
quite higher than the operational temperature of the devices. On
the contrary, the melting point of molecule M0 was observed to
be sensibly lower (89 ºC).
Energy Levels (eV)
-3.5
-3.5
-3.7
-3.9
-3.8
-3.9
-5.4
-4.5
-5.0
-5.1
DSC measurements of P0 revealed its amorphous nature,
showing a Tg values at 157.15 ºC. On the contrary, polymer P1
did not show any thermal feature in the second heating cycle
over the temperature range of 25-180 ºC (Figure S3B, Table S3).
-5.9
-5.9
-6.1
P1
-7.0 -7.0
Incorporation into PSCs
As stated before, thanks to their solubility, the three novel
molecules could be incorporated as buffer layers between the
ETL and the gold electrode. The cells were then characterized in
terms of performances and their stability was preliminarily
evaluated.
FTO NiOx PVSK PCBM BCP M1
P0
Au
Figure 8. Summary of the energy levels of the different compounds
considered in the study.
The chosen devices architecture was the inverted one: FTO
glass/NiOx/CH3NH3PbI3/PC61BM/buffer layer/Au. The buffer
layer could be either one of the novel molecules (P0, M1, P1) or
BCP as a reference. An additional reference was given by
preparing devices with no buffer layer, so that PC61BM directly
contacts with the gold cathode. Further details about the
preparation of the cells are reported in the Experimental Section.
The statistics of the photovoltaic parameters extracted from the
cells, such as JSC, open-circuit voltage (VOC), fill factor (FF), and
PCE, are reported in Figure 9. The best curves of each condition,
with the corresponding parameters, are reported in Figure S4.
EQE experiments are reported in figure S5 and the photovoltaic
parameters extracted from both these experiments and the best
curves are reported in table S4
Thermal properties
To better evaluate their suitability for an application in perovskite
device, the thermal behaviour of polymers P0 and P1 is
investigated by means of TGA and DSC analyses. The
thermograms of these copolymers are showed in Figure S3A.
Both polymers show thermal stability under operation
temperature of PSCs; in fact their Td values, which correspond
to their 2% weight losses, are, respectively, 240.20 ºC and
188.34 ºC. The main decomposition process of copolymer P0
could be related to the loss of the amino groups. In case of
polymer P1, the first mass loss was attributed to the loss of the
ethyl iodide groups, followed by the loss of the amino groups
similarly to what happens to P0. Thereafter, both P0 and P1
undergo the second major degradation process owing to the
1200
1000
30
25
20
15
10
5
800
600
400
200
0
0
18
16
14
12
10
8
6
4
2
0
80
70
60
50
40
30
20
-2
No buffer layer BCP M1 P0 P1
Figure 9: Statistics of the photovoltaic parameters (short-circuit current JSC, open-circuit voltage VOC, fill factor FF, and power conversion efficiency PCE) of the
different NDI-containing devices and their references.
6
This article is protected by copyright. All rights reserved.

10.1002/ejoc.202000287
European Journal of Organic Chemistry
FULL PAPER
Table 4: Best and Average photovoltaic parameters of the different NDI-containing devices and their references.
JSC (mA cm-2)
VOC (mV)
FF (%)
PCE (%)
Best
15.2
15.9
21.6
19.3
21.8
19.8
9.5
837
583
45
37
68
63
60
59
33
31
65
60
5.73
3.48
No Buffer Layer
Average
Best
1066
969
15.57
11.82
13.73
11.42
1.79
BCP
M1
P0
Average
Best
1045
984
Average
Best
566
Average
Best
7.9
284
0.76
21.4
19.4
1027
985
14.25
11.43
P1
Average
The most important information that could be extracted from
these statistics is that the results that were obtained with
molecule M1 and polymer P1 are highly comparable to those
obtained with the reference BCP buffer layer. As a matter of fact,
BCP gave an average 11.82% PCE and both M1 and P1 gave,
respectively, average values of 11.42% and 11.43%. This
demonstrates once again the great potential of halogen-
ammonium salts for the improvement of the performances of the
perovskite devices. The improvement of JSC led by the novel
structure is clear for the case of M1, where the average current
increases from 19.3 mA cm-2 of registered with BCP to 19.8 mA
cm-2. This is confirmed by the results reported in Table 4, which
shows both the best and the average photovoltaic parameters.
On the contrary, the performances given by P0 as buffer layer
were not satisfactory. The morphology of the devices was
studied in order to find out the reason of these lower
performances. The FESEM images of the layers are shown in
Figure S6. The pinholes of the P0 layer, which can be observed
in the FESEM top view, could have explained the bad results,
nevertheless they can be seen also in the case of P1 even if to a
slightly lesser extent, and so this hypothesis has to be rejected.
The stability was evaluated in two different ways. The first trial
was performed leaving the cells on a hotplate at 80 °C and
keeping track of the performances (Figure 10A). In both cases
the NDI-containing cells showed higher stability than the
reference; the better results obtained with P1 should be ascribed
to the polymeric nature of the compound and its consequent
tendency to generate better layers.
The second trial was performed leaving the cells in a dark
drawer for a prolonged period of time, in order to evaluate their
resistance to air (Figure 10B). In this case, only P1 could give
better results than the reference after 10 days. Its polymeric
nature must be once again at the root of this improved stability.
A1.0
BCP
M1
P1
0.8
0.6
0.4
0.2
0.0
On the contrary, M1 presented
a continuous layer. The
thickness of the layers was also investigated, as reported in
Figure S7. The layers are sometimes too thin to be differentiated
from PC61BM, therefore the thicknesses were calculated from
the difference between the measured overall ETL and the bare
PC61BM, which was estimated to be of an average value of 31 ±
4 nm. The obtained thicknesses are reported in Table 5. The
standard error was calculated dividing the standard deviation by
the square root of the data sample size. The estimated average
thickness of P0 is in fact about half of the average thickness
measured for the other compounds, and this might be the
reason of its lower performances.
0
5
10
15
20
25
Time (h)
1.0
BCP
M1
P1
B
0.9
0.8
0.7
0.6
0.5
0.4
Due to its poorer performance, P0 was not further considered for
stability tests.
Table 5: Estimated average thicknesses of the buffer layers. The standard
error was calculated dividing the standard deviation by the square root of the
data sample size.
0
2
4
6
8
10
Time (Days)
PCBM+Buffer Layer
Estimated Buffer Layer
Figure 10: A) Normalized PCE of the cells that underwent thermal stability test
(nm)
Thickness (nm)
and B) Normalized PCE of the cells that underwent long-term stability test.
BCP
M1
P0
53±6
54±2
42±7
51±9
22
23
11
20
P1
7
This article is protected by copyright. All rights reserved.

10.1002/ejoc.202000287
European Journal of Organic Chemistry
FULL PAPER
working electrode, a platinum wire counter electrode and a silver wire
pseudoreference electrode. All the potential values were reported versus
the redox potential of the ferrocene/ferrocenium redox couple. HOMO
and LUMO values were obtained from the onset potentials of cyclic
voltammetry measurements. The potential of Fc/Fc+ is 0.2 V vs. the
normal hydrogen electrode (NHE), which has an absolute potential of -
4.6 eV vs. vacuum.[48] Therefore, the energy levels are HOMO/LUMO
(eV) = -4.8 – E1/2 Ox1/Red1 and the optical band gap = LUMO – HOMO.
The morphologies of the films were analyzed with an ULTRA plus ZEISS
field-emission scanning electron microscope (FESEM).
Conclusions
In summary, three novel n-type NDI-based molecules, M1, P0,
and P1, bearing ternary or quaternary ammonium moieties, were
designed and synthesized. The compounds were successfully
purified as confirmed by a full characterization through different
spectroscopic techniques. The analysis of the electrochemical
properties of the novel molecules shows that they meet the
requirements to be used as buffer layers for the ETL in PSCs. In
fact, the compounds showed suitable energy levels with low-
lying LUMO levels of about -4 eV and HOMO levels around -6
eV or lower, which would ensure synergetic action with PC61BM.
Moreover, thermal analysis indicated that polymers are thermally
stable with no phase transition under operation temperatures of
PSCs. The choice of using them as buffer layers for the ETL
finds its root in their proper solubility in polar solvents, which
provides an easy processing. The introduction of these
molecules into PSCs gave similar results to those obtained with
the reference buffer layer, BCP, both in terms of photovoltaic
performances and stability. Hence, we could conclude that NDIs
are a very promising candidate to improve the performances of
PSCs.
The absorptance spectra were obtained from the transmittance and
reflectance spectra, which were measured with
spectrophotometer with an integrating sphere.
a JASCO V-570
The J–V characteristics were measured using a 450 W xenon light
source (Oriel). The light intensity was calibrated with a Si photodiode
equipped with an IR-cutoff filter (KG3, Schott) and it was recorded during
each measurement. Current-voltage characteristics of the cells were
obtained by applying an external voltage bias while measuring the
current response with a digital source meter (Keithley 2400). The voltage
scan rate was 20 mV s-1, and no device preconditioning was applied
before starting the measurement, such as light soaking or forward
voltage bias applied for a long time. The cells were masked with a black
metal mask of 0.16 cm2 to estimate the active area and reduce the
influence of the scattered light. The external quantum efficiency (EQE)
was measured as a function of wavelength from 300 to 850 nm with a
step of 10 nm using a home-built small-spot EQE system. A Stanford
Research SR830 Lock-In amplifier is used to measure the electrical
response of the device under test and evaluated in TracQ-Basic software.
Experimental Section
Synthetic Procedures
M0:1,4,5,8-Naphthalenetetracarboxylic dianhydride (500 mg, 1.87 mmol)
Methods and Materials
All commercially available reagents and solvents for synthesis and
photovoltaic study were used as received without further purification.
Anhydrous solvents were dried using a SPS purification system. Column
chromatography was carried out using silica gel (40 to 60 μm, 60 Å) as
stationary phase. Thin layer chromatography (TLC) was performed on
precoated silica gel and observed under UV light.
was
suspended
in
5
mL
of
DMF together
with
3-
(dibutylamino)propylamine (834 μL, 3.70 mmol). The mixture was left to
react under microwave at 140 °C for 10 min. The reaction mixture was
then poured into 500 mL of water and the precipitate was filtered off,
washed 3 times with water and dried under reduced pressure. The
product was kept in darkness (900 mg, 80%).
The 1H-NMR and 13C-NMR spectra were taken on Bruker NMR
spectrometers (300 MHz for 1H NMR, 75 MHz and 100 MHz for 13C
NMR) at room temperature. Chemical shifts (δ) are reported in ppm.
Abbreviations of coupling patterns are as follows: s = singlet, d = doublet,
t = triplet, m = multiplet.
Matrix Assisted Laser Desorption Ionization (MALDI) experiments,
coupled to a Time-of-flight (TOF) analyzer, were recorded on Bruker
REFLEX spectrometer.
Electrospray ionization (ESI) measurements were carried out on HPLC
Agilent 1200 Series system coupled to a hybrid quadrupole-time of flight
(LC-QTOF) mass spectrometer Agilent 6520 from Agilent Technologies
(Santa Clara, CA, USA).
The molecular weight distribution of the polymers was determined by size
exclusion chromatography (SEC, Waters). The setting consisted of a
pump, a differential refractometer (Waters 2410) and three columns in
series (Styragel HR2, HR4 and HR6; with a pore size from 102 to 106 Å).
The analyses were performed at 35 °C and THF was used as solvent at
a flow rate of 1 mL min-1. Calibration was relative to polystyrene
standards. A series of polystyrene standards in the range of 580-
3,848,000 g mol-1 were used to prepare a universal calibration curve.
The thermogravimetric analysis (TGA) was performed on a TA Q500
using a 10 °C min-1 heating rate under a nitrogen flow.
Differential scanning calorimetry (DSC) was performed on a DSC3+
Mettler Toledo at a scan rate of 20 °C min-1 in the nitrogen atmosphere.
Absorption measurements were recorded on a Perkin-Elmer Lambda 950
spectrometer. All the compounds were measured at various
concentrations (10-5-10-7 mol l-1) to affirm linear behaviour. The
measurements of films of compounds were performed on samples that
were spin-coated from the respective solvent onto FTO/SnO2 substrates.
The optical band gap was estimated from the onset of the absorption
edge (λonset) in the longer wavelength region.
1H-NMR (300 MHz, CDCl3), δ (ppm): 8.75 (s, 4H), 4.22 (t, 4H), 2.58 (t,
4H, 4H), 2.41 (t, 8H), 1.97-1.82 (4H), 1.47-1.22 (16H), 0.89 (t, 12H).
13C-NMR (75 MHz, CDCl3), δ (ppm): 162.96, 131.02, 126.81, 126.78,
53.82, 51.94, 39.86, 29.38, 25.71, 20.86, 14.20.
MS (ESI-TOF): calculated for C36H52N4O4 [M+H]+: 605.406, found:
605.406.
M1: M0 (400 mg, 661 μmol) was dissolved in 2 mL of CH3CN and placed
in a closed tube under nitrogen atmosphere. After that, iodoethane (3.00
mL. 37.5 mmol) was added to the solution. The mixture was allowed to
react at room temperature at dark for 5 d. The product, which precipitates
as orange crystals, was collected by filtration, washed several times with
Et2O and dried over vacuum. The solid was recrystallized from DCM to
obtain 310 mg of M1 (47% yield).
1H-NMR (300 MHz, DMSO), δ (ppm): 8.74 (s, 4H), 4.15 (t, 4H), 3.33-3.22
(8H), 3.22-3.10 (8H), 2.15-2.01 (4H), 1.64-1.50 (8H), 1.37-1.22 (8H), 1.18
(t, 6H), 0.91 (t, 12H).
13C-NMR (75 MHz, DMSO), δ (ppm): 162.91, 130.53, 126.42, 126.25,
57.16, 55.02, 53.42, 37.53, 23.07, 20.31, 19.16, 13.49, 7.38.
MS (ESI-TOF): calculated for C40H62N4O4 [M]+: 662.477, found: 662.476.
M2: 1,4-Dibromo-2,5-bis(2-ethylhexyl)benzene (13, 240 mg, 521 μmol),
bis(pinacolato)diboron (389 mg, 1.53 mmol) and potassium acetate (400
mg, 4.08 mmol) were dissolved in 8 mL of anhydrous 1,4-dioxane. The
resulting suspension was degassed with an argon stream before and
after the addition of PdCl2(dppf) (11.0 mg, 15.3 μmol), and the mixture
was stirred at 80 °C for 3 d . After cooling down to room temperature, 40
mL of water were added and the aqueous phase was extracted with
DCM (2 x 40 mL). The combined organic phases were washed with
water (2 x 150 mL), brine (1 x 50 mL) and then dried over anhydrous
Na2SO4, concentrated under reduced pressure and purified by flash
column chromatography using hexane:AcOEt (20:1) as eluent to give
monomer M2 as a white solid (249 mg, 84%).
Cyclic voltammograms were carried out on a Princeton Applied Research
Parstat 2273 in a 3-electrode single compartment cell with carbon disc
8
This article is protected by copyright. All rights reserved.

10.1002/ejoc.202000287
European Journal of Organic Chemistry
FULL PAPER
1H-NMR (300 MHz, CDCl3), δ (ppm): 7.48 (s, 2H), 2.85-2.70 (4H), 1.59-
1.45 (2H), 1.38-1.11 (40H), 0.91-0.77 (12H).
Finally, the yellow product was dried in vacuum (3.80 g, 89%), which was
used directly in the next step without further purification.
13C-NMR (75 MHz, CDCl3), δ: 144.73, 137.61, 83.41, 42.03, 39.56, 32.02,
28.63, 25.48, 25.00, 23.35, 14.30, 10.84.
1H-NMR (300 MHz, DMSO), δ (ppm): 8.79 (s, 2H).
P0: In a Schlenk flask three cycles of argon-vacuum were applied. Next,
monomer M6 (549 mg, 721 μmol), monomer M4 (400 mg, 721 μmol),
NaHCO3 (2.42 g, 28.8 mmol) and Pd(PPh3)4 (16.7 mg, 14.4 μmol,) were
dissolved in 10 mL of a mixture of THF:water (8:3) previously degassed.
The reaction mixture was stirred and refluxed for 6 d. After that, 30 mL of
chloroform and 30 mL of water were added to the mixture and the
phases were separated. The aqueous phase was extracted with
chloroform (2 x 30 mL) and the combined organic phases were washed
with brine (1 x 100 mL), dried over Na2SO4 and the solvent removed
under reduced pressure. The crude product was dissolved in 0.7 mL of
chloroform and dripped into 200 mL of MeOH. The orange precipitate
was filtered off and washed with MeOH (4 d), acetone (3 d) and CH3CN
(3 d) using a Soxhlet extractor to afford P0 (90.7 mg, 14%). SEC: Mn =
4350, Mw = 6953, PDI = 1.6.
MS (MALDI-TOF): calculated for C34H60B2NaO4 [M+Na]+: 577.462, found:
577.459.
M3: To a solution of 2,6-dibromo-1,4,5,8-naphthalenetetracarboxylic
dianhydride (M9, 500 mg, 1.17 mmol) in 15 mL of AcOH, 3-
(dibutylamino)propylamine (660 μL, 2.92 mmol) was added slowly at
room temperature and stirred at 130 °C for 45 min. Upon completion, the
reaction mixture was poured into cooled water and the aqueous phase
was neutralized with Na2CO3 and extracted with chloroform until no color
could be seen in the organic phase. The combined organic extracts were
washed with brine, dried over Na2SO4, filtered and the solvent removed
under reduced pressure. The crude product was purified by flash column
chromatography using the mixture CHCl3:EtOH (95:5) with 0.5% of NEt3
as eluent to yield the product (200 mg, 23%).
1H-NMR (300 MHz, CDCl3), δ (ppm): 8.99 (s, 2H), 4.28-4.18 (4H), 2.58 (t,
4H), 2.40 (t, 8H), 1.95-1.81 (4H), 1.46-1.21 (16H), 0.90 (t, 12H).
13C-NMR (100 MHz, CDCl3), δ (ppm): 160.87, 160.85, 139.14, 128.41,
127.82, 125.49, 124.23, 53.77, 51.89, 40.45, 29.27, 25.41, 20.85, 14.28.
MS (MALDI-TOF) calculated for C36H51Br2N4O4 [M+H]+: 761.228, found:
1H-NMR (300 MHz, CD2Cl2), δ (ppm): 8.82 (s, 2H), 7.18 (s, 2H), 4.49-
3.82 (4H), 2.68-2.22 (16H), 2.02-1.71 (4H), 1.69-1.52 (2H), 1.52-1.28
(16H), 1.18-0.80 (28H), 0.77-0.55 (12H).
MS (MALDI-TOF): The building block (C58H86N4O4 [M]+: 902.664) of P0
was detected. (Figure S5)
761.202.
P1: Copolymer P0 (130 mg, 144 μmol) was dissolved in 2 mL of THF.
Next, iodoethane (690 μL, 8.64 mmol) and 1 mL of DMF were added,
and the solution was stirred for 5 d at room temperature in darkness.
After that, the reaction was put into water and the resulting precipitate
was filtered and washed with AcOEt for 1 d using a Soxhlet extractor.
The solid was dissolved in DCM and the remained solid was filtered off.
The filtrated was evaporated under reduced pressure to give polymer P1
as an orange solid (100 mg, 57%).
[49]
M4:
This product was prepared in two steps: (1) preparation of the
Grignard reagent (RMgX); (2) the reaction of RMgX with 1,4
dibromobenzene by Kumada coupling.
1) Grignard reaction. To a solution of magnesium (1.65 g, 67.8 mmol) in
4 mL of anhydrous Et2O was added one crystal of iodine and 2-
ethylhexyl bromide (9.82 g, 50.9 mmol) together with 4 mL of anhydrous
Et2O slowly (30 min). The mixture was refluxed under argon atmosphere
for 1 h.
1H-NMR (300 MHz, CD2Cl2), δ (ppm): 8.97-8.60 (2H), 7.42-7.10 (2H),
4.52-4.05 (4H), 3.74-2.76 (16H), 2.61-2.12 (4H), 1.91-1.68 (6H), 1.56-
1.33 (16H), 1.18-0.78 (34H), 0.78-0.48 (12H).
2) Kumada reaction. To an ice-cooled solution of 1,4-dibromobenzene
(4.00 g, 16.9 mmol) and Ni(dppp)Cl2 (279 mg, 514 μmol) in 5 mL of
anhydrous Et2O was added slowly the solution with the Grignard reagent
(previously synthesized) together with 5 mL of Et2O to transfer that
reagent completely. The reaction mixture was stirred at 40 ºC under
argon atmosphere during 2 d. After this time, 50 mL of HCl (1 M) was
added to the reaction mixture and the aqueous phase was extracted with
Et2O (3 x 50 mL). The combined organic phases were washed with water
(2 x 150 mL) and brine (1 x 150 mL), dried over Na2SO4 and
concentrated under reduced pressure. The resulting yellowish oil was
purified by vacuum distillation (43 mbar, 200 °C) to give a colorless oil
(3.82 g, 74%), which was used directly in the next step without further
purification.
Perovskite Solution Preparation
580.0 mg of PbI2 and 200.0 mg of MAI were dissolved in 0.7 mL of DMF
and 0.3 mL of DMSO. The solution was stirred in a glove box at 70 ºC for
12 h and then filtered using a 0.22 µm PTFE syringe filter.
Device Fabrication
The FTO/glass substrate was cleaned by successive sonication in
detergent (2% Hellmanex water solution), deionized water, isopropyl
alcohol and ethanol for 20 min and then dried by flowing dry oxygen.
After a 30 min UV-ozone treatment, 100 µL of the NiOx precursor solution
(25.4 mg ml-1 Ni(OAc)2·4H2O in 2-methoxyethanol and 6.04 µL of
ethanolamine) was spread over an FTO substrate and spin coated for
40 s at 4000 rpm with a ramp of 2000 rpm s-1. The substrates were
immediately annealed at 300 ºC for 10 min and followed by 500 °C for 30
minutes in air. For the deposition of perovskite films and further
processing, the substrates were transferred into a glove box. 70 µL of
perovskite solution was deposited on top of the NiOx layer by a two-step
spin-coating process at 1000 rpm with a ramp of 200 rpm s-1 and 4000
rpm with a ramp of 3000 rpm s-1 for 5 and 45 s, respectively. During the
spinning, 200 μL of chlorobenzene were dropped on the middle of the
film 35 s before the end of the procedure. The substrates were then
annealed at 60 ºC for 1 min and then 100 °C for 10 min. 50 µL of the
PC61BM solution (20 mg ml-1 in chlorobenzene) was dynamically spun
onto the perovskite layer at 2000 rpm for 45 s with acceleration of
1000 rpm s-1, then annealed at 100 ºC for 10 min. For the buffer layer,
BCP was fully dissolved in EtOH, P0 in EtOAc:MeOH (1:1), M1 and P1 in
MeOH to concentration of 1 mg ml-1. 50 µL of each buffer solution was
dynamically spin-coated on top of PC61BM layer at 4000 rpm for 30 s with
a ramp of 2000 rpm s-1.
This compound was previously described in literature and showed
identical spectroscopic data as those reported therein. [49]
1H-NMR (300 MHz, CDCl3), δ (ppm): 7.31 (s, 2H), 2.58 (d, J = 7.2 Hz,
4H), 1.73-1.60 (2H), 1.36-1.21 (16H), 0.92-0.84 (12H).
M5:[50] Bromine (1.51 mL, 29.1 mmol) was added dropwise in the dark to
a mixture of 1,4-bis(2-ethylhexyl)benzene (12, 3.83 g, 12.7 mmol) and
iodine (32.8 mg, 129 μmol) at 0 °C. The mixture was stirred for 48 h at
room temperature. After that time, 20 mL of DCM and 20 mL of NaOH
(20%) were added and the phases were separated. The aqueous phase
was extracted with DCM (2 x 20 mL) and the combined organic phases
were washed with 60 mL of NaOH (20%), Na2S2O3 (2 x 60 mL) and brine
(1 x 60 mL), dried over Na2SO4 and concentrated under reduced
pressure. The crude product was purified by flash column
chromatography using hexane as eluent to afford compound M5 as a
colourless oil (3.43 g, 58%)
1H-NMR (300 MHz, CDCl3), δ (ppm): 7.31 (s, 2H), 2.58 (d, J = 7.2 Hz,
4H), 1.73-1.60 (2H), 1.36-1.21 (16H), 0.92-0.84 (12H).
M6:[30] 1,4,4,8-Naphthalenetetracarboxylic dianhydride (2.70 g, 9.99
mmol) was suspended in 25 mL of H2SO4 (96%) and the mixture was
stirred for 5 min. Next, 1,3- dibromo-5,5-dimethylhydantoin (DBDMH,
4.28 g, 15.0 mmol) was added to the mixture in portions over a period of
1 h at room temperature. The reaction mixture was then stirred during 48
h at 50 °C. After that time, the mixture was poured into crushed ice and
the precipitated solid was filtered and washed several times with water.
9
This article is protected by copyright. All rights reserved.

10.1002/ejoc.202000287
European Journal of Organic Chemistry
FULL PAPER
[24] S. Shao, Z. Chen, H. H. Fang, G. H. ten Brink, D. Bartesaghi, S.
Adjokatse, L. J. A. Koster, B. J. Kooi, A. Facchetti, and M. A. Loi, J.
Mater. Chem. A, 2016, 4, 2419-2426.
Acknowledgements
A. C.-E. and S. C. acknowledge the Polymat Foundation for a
PhD and a Postdoctoral research position, respectively. J. L. D.
acknowledges Ikerbasque, the Basque Foundation for Science,
for an “Ikerbasque Research Fellow” contract, the Polymat
Foundation and the MINECO of Spain for CTQ 2016-81911-
REDT and RTI2018-101782-B-I00 grants.
[25] C. Sun, Z. Wu, H.-L. Yip, H. Zhang, X.-F. Jiang, Q. Xue, Z. Hu, Z. Hu, Y.
Shen, M. Wang, F. Huang, and Y. Cao, Adv. Energy Mater., 2016, 6,
1501534.
[26] J. H. Heo, D. S. Lee, D. H. Shin, and S. H. Im, J. Mater. Chem. A, 2019,
7, 888-900.
[27] X. Zheng, B. Chen, J. Dai, Y. Fang, Y. Bai, Y. Lin, H. Wei, Xiao C. Zeng,
and J. Huang, Nat. Energy, 2017, 2, 17102.
[28] Z. Wu, C. Sun, S. Dong, X.-F. Jiang, S. Wu, H. Wu, H.-L. Yip, F. Huang,
and Y. Cao, J. Am. Chem. Soc., 2016, 138, 2004-2013.
[29] B. Russ, M. J. Robb, B. C. Popere, E. E. Perry, C.-K. Mai, S. L. Fronk,
S. N. Patel, T. E. Mates, G. C. Bazan, J. J. Urban, M. L. Chabinyc, C. J.
Hawker, and R. A. Segalman, Chem. Sci., 2016, 7, 1914-1919.
[30] M. Sasikumar, Y. V. Suseela, and T. Govindaraju, Asian J. Org. Chem.,
2013, 2, 779-785.
Keywords: Naphthalene Diimides • Perovskite Solar Cells •
PC61BM • Buffer Layer • Stability
[1]
[2]
A. Kojima, K. Teshima, Y. Shirai, and T. Miyasaka, J. Am. Chem. Soc.,
2009, 131, 6050-6051.
NREL,
Best
Research-Cell
Efficiencies
Chart,
[31] L. Li, MALDI MS: A Practical Guide to Instrumentation, Methods and
Applications (Ed: Franz Hillenkamp and Jasna Peter‐Katalinić), 2007,
245–297.
https://www.nrel.gov/pv/assets/pdfs/best-research-cell-
efficiencies.20191104.pdf.
[3]
[4]
[5]
[6]
[7]
[8]
[9]
S. F. Völker, S. Collavini, and J. L. Delgado, ChemSusChem, 2015, 8,
3012-3028.
[32] P. M. Alvey, R. J. Ono, C. W. Bielawski, and B. L. Iverson,
Macromolecules, 2013, 46, 718-726.
J. Urieta-Mora, I. García-Benito, A. Molina-Ontoria, and N. Martín,
Chem. Soc. Rev., 2018, 47, 8541-8571.
[33] N. B. Kolhe, A. Z. Ashar, K. S. Narayan, and S. K. Asha,
Macromolecules, 2014, 47, 2296-2305.
S. Collavini and J. L. Delgado, Sustainable Energy Fuels, 2018, 2,
2480-2493.
[34] S. Vasimalla, S. P. Senanayak, M. Sharma, K. S. Narayan, and P. K.
Iyer, Chem. Mater., 2014, 26, 4030-4037.
S. Sun, T. Buonassisi, and J.-P. Correa-Baena, Adv. Mater. Interfaces,
2018, 5, 1800408.
[35] P. M. Alvey and B. L. Iverson, Org. Lett., 2012, 14, 2706-2709.
[36] H. Huang, N. Zhou, R. P. Ortiz, Z. Chen, S. Loser, S. Zhang, X. Guo, J.
Casado, J. T. López Navarrete, X. Yu, A. Facchetti, and T. J. Marks,
Adv. Funct. Mater., 2014, 24, 2782-2793.
J. Lian, B. Lu, F. Niu, P. Zeng, and X. Zhan, Small Methods, 2018, 2,
1800082.
T. Gatti, E. Menna, M. Meneghetti, M. Maggini, A. Petrozza, and F.
Lamberti, Nano Energy, 2017, 41, 84-100.
[37] A. Luzio, D. Fazzi, D. Natali, E. Giussani, K.-J. Baeg, Z. Chen, Y.-Y.
Noh, A. Facchetti, and M. Caironi, Adv. Funct. Mater., 2014, 24, 1151-
1162.
E. Castro, J. Murillo, O. Fernandez-Delgado, and L. Echegoyen, J.
Mater. Chem. C, 2018, 6, 2635-2651.
[38] Y. Kim, J. Hong, J. H. Oh, and C. Yang, Chem. Mater., 2013, 25, 3251-
3259.
[10] J. C. Hummelen, B. W. Knight, F. LePeq, F. Wudl, J. Yao, and C. L.
Wilkins, J. Org. Chem., 1995, 60, 532-538.
[39] F. Doria, I. Manet, V. Grande, S. Monti, and M. Freccero, J. Org. Chem.,
2013, 78, 8065-8073.
[11] J.-Y. Jeng, Y.-F. Chiang, M.-H. Lee, S.-R. Peng, T.-F. Guo, P. Chen,
and T.-C. Wen, Adv. Mater., 2013, 25, 3727-3732.
[40] H. Chen, Y. Guo, Z. Mao, G. Yu, J. Huang, Y. Zhao, and Y. Liu, Chem.
Mater., 2013, 25, 3589-3596.
[12] D. Luo, W. Yang, Z. Wang, A. Sadhanala, Q. Hu, R. Su, R. Shivanna,
G. F. Trindade, J. F. Watts, Z. Xu, T. Liu, K. Chen, F. Ye, P. Wu, L.
Zhao, J. Wu, Y. Tu, Y. Zhang, X. Yang, W. Zhang, R. H. Friend, Q.
Gong, H. J. Snaith, and R. Zhu, Science, 2018, 360, 1442.
[13] J. K. Mwaura, M. R. Pinto, D. Witker, N. Ananthakrishnan, K. S.
Schanze, and J. R. Reynolds, Langmuir, 2005, 21, 10119-10126.
[14] W. Zhao, S. Li, H. Yao, S. Zhang, Y. Zhang, B. Yang, and J. Hou, J.
Am. Chem. Soc., 2017, 139, 7148-7151.
[41] T. Jia, C. Sun, R. Xu, Z. Chen, Q. Yin, Y. Jin, H.-L. Yip, F. Huang, and
Y. Cao, ACS Appl. Mater. Interfaces, 2017, 9, 36070-36081.
[42] B. L. Schottel, H. T. Chifotides, and K. R. Dunbar, Chem. Soc. Rev.,
2008, 37, 68-83.
[43] Q. Song, F. Li, Z. Wang, and X. Zhang, Chem. Sci., 2015, 6, 3342-3346.
[44] C. Duan, W. Cai, B. B. Y. Hsu, C. Zhong, K. Zhang, C. Liu, Z. Hu, F.
Huang, G. C. Bazan, A. J. Heeger, and Y. Cao, Energy Environ. Sci.,
2013, 6, 3022-3034.
[15] J. Seo, S. Park, Y. Chan Kim, N. J. Jeon, J. H. Noh, S. C. Yoon, and S.
I. Seok, Energy Environ. Sci., 2014, 7, 2642-2646.
[45] C.-C. Chueh, C.-Z. Li, and A. K. Y. Jen, Energy Environ. Sci., 2015, 8,
1160-1189.
[16] M. Al Kobaisi, S. V. Bhosale, K. Latham, A. M. Raynor, and S. V.
Bhosale, Chem. Rev., 2016, 116, 11685-11796.
[46] F. Huang, L. Hou, H. Wu, X. Wang, H. Shen, W. Cao, W. Yang, and Y.
Cao, J. Am. Chem. Soc., 2004, 126, 9845-9853.
[17] X. Zhan, A. Facchetti, S. Barlow, T. J. Marks, M. A. Ratner, M. R.
Wasielewski, and S. R. Marder, Adv. Mater., 2011, 23, 268-284.
[18] J. H. Heo, S.-C. Lee, S.-K. Jung, O. P. Kwon, and S. H. Im, J. Mater.
Chem. A, 2017, 5, 20615-20622.
[47] P. B. Balanda, M. B. Ramey, and J. R. Reynolds, Macromolecules,
1999, 32, 3970-3978.
[48] C. M. Cardona, W. Li, A. E. Kaifer, D. Stockdale, and G. C. Bazan, Adv.
Mater., 2011, 23, 2367-2371.
[19] S. K. Jung, J. H. Heo, D. W. Lee, S. H. Lee, S. C. Lee, W. Yoon, H.
Yun, D. Kim, J. H. Kim, S. H. Im, and O. P. Kwon, ChemSusChem,
2019, 12, 224-230.
[49] T. Schwalm and M. Rehahn, Macromolecules, 2007, 40, 3921-3928.
[50] P. Wessig, M. Gerngroß, D. Freyse, P. Bruhns, M. Przezdziak, U.
Schilde, and A. Kelling, J. Org. Chem., 2016, 81, 1125-1136.
[20] S. K. Jung, J. H. Heo, D. W. Lee, S. C. Lee, S. H. Lee, W. Yoon, H.
Yun, S. H. Im, J. H. Kim, and O. P. Kwon, Adv. Funct. Mater., 2018, 28,
1800346.
[21] T. Nakamura, N. Shioya, T. Shimoaka, R. Nishikubo, T. Hasegawa, A.
Saeki, Y. Murata, R. Murdey, and A. Wakamiya, Chem. Mater., 2019,
31, 1729-1737.
[22] S.-K. Jung, J. H. Heo, B. M. Oh, J. B. Lee, S.-H. Park, W. Yoon, Y.
Song, H. Yun, J. H. Kim, S. H. Im, and O. P. Kwon, Adv. Funct. Mater.,
2020, 30, 1905951.
[23] W. Wang, J. Yuan, G. Shi, X. Zhu, S. Shi, Z. Liu, L. Han, H.-Q. Wang,
and W. Ma, ACS Appl. Mater. Interfaces, 2015, 7, 3994-3999.
10
This article is protected by copyright. All rights reserved.

10.1002/ejoc.202000287
European Journal of Organic Chemistry
FULL PAPER
Naphthalene Diimide-Based Molecules for Efficient and Stable Perovskite Solar Cells
Key Topic: Polymers as Additives for Perovskite Solar Cells
Naphthalene Diimides (NDI) are promising candidates to improve the performances and the stability of perovskite solar cells. In this
paper, three novel NDI-based molecules are synthesized, characterized, and incorporated into solar devices as buffer layers between
the electron transporting layer and the cathode. The evaluation of the performances and of the stability of the cells highlighted the
beneficial effects brought by these molecules.
Institute and/or researcher Twitter usernames:
@bercpolymat
@upvehu
@JuanLDelgadoC
11
This article is protected by copyright. All rights reserved.