Article

Cite This: Chem. Mater. 2019, 31, 6945−6949 pubs.acs.org/cm

Molecular Passivation of MoO3: Band Alignment and Protection of

Charge Transport Layers in Vacuum-Deposited Perovskite Solar
Cells
Daniel Pérez-del-Rey,† Lidón Gil-Escrig,‡ Kassio P. S. Zanoni,†,§ Chris Dreessen,† Michele Sessolo,†
Pablo P. Boix,*,† and Henk J. Bolink*,†
†
Instituto de Ciencia Molecular, Universidad de Valencia, C/Catedrático J. Beltrán 2, 46980 Paterna, Spain
‡
Helmholtz-Zentrum Berlin für Materialien und Energie GmbH, Kekuléstraße 5, 12489 Berlin, Germany
§
Istituto de Física de São Carlos, Universidade de São Paulo, Av. Trab. São Carlense 400, 13566 São Carlos, Brazil
See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

*
S Supporting Information
Downloaded via UNIV OF WASHINGTON on July 24, 2024 at 18:39:42 (UTC).

ABSTRACT: Vacuum deposition of perovskite solar cells can

achieve efficiencies rivaling solution-based methods and allows
for more complex device stacks. MoO3 has been used to
enhance carrier extraction to the transparent bottom electrode
in a p-i-n configuration; here, we show that by inserting an
molecular interlayer with a lone pair nitrogen atom and low
ionization potential, it can also be used on the top of a
perovskite absorber in a n-i-p configuration. This strategy
enables the first vacuum-deposited perovskite solar cells with
metal oxides as charge transporting layers for both electrons
and holes, leading to power conversion efficiency >19%.

1. INTRODUCTION layers to ensure that the photogenerated charge carriers are

efficiently extracted to an external circuit.9
Halide perovskites have a wide range of chemical and physical
Semiconducting metal oxides are widely studied in solution-
properties that make them suitable for new generation
processed perovskite devices and have also been recently
photovoltaics.1 These encompass long charge carrier diffusion applied as charge transport layers in highly efficient vacuum-
lengths2 and high absorption coefficients.3 Furthermore, their deposited perovskite solar cells.10 The use of TiO2 in n-i-p
compositional flexibility allows fine-tuning of the photophysical structure instead of doped C60 reduces parasitical absorption
properties.4,5 The use of perovskites in high efficiency devices and has a positive impact on device stability and material
requires directing charge carriers to the appropriate electrode. costs.10 However, the system still relies on organic doped
This is generally achieved by placing selective charge transport materials as hole transport layer (HTL) in the top contact with
layers below and above the perovskite layer. Interfaces between the associated complexity and stability-related issues. Hence, it
the perovskite and the charge extraction material and between is important to find metal oxide alternatives that can reduce
the charge extraction material and the electrodes are thus parasitic absorption in p-i-n devices and can be incorporated to
crucial for an efficient charge carrier collection.6 In this aspect, a n-i-p stack without compromising the rest of the deposited
perovskite-based solar cells can have two main device layers.
architectures depending on whether the photogenerated Molybdenum(VI) oxide (MoO3) has been recurrently
holes are extracted through the light incident side and employed as hole injection/extraction layers in organic light-
photogenerated electrons extracted through the metal side emitting diodes (OLEDs)11 and in organic, quantum dot, and
(p-i-n) or vice versa (n-i-p), just by changing the deposition perovskite solar cells12,13 due to its very deep work function
order of the transport layers. and low-lying conduction band.14 MoO3 has the advantage of
creating an optimal ohmic contact with ITO. Its protective
Vacuum deposition is a widely adopted technique in the
effect against moisture when used in this architecture has also
industry with proven scalability capabilities. More interestingly
been reported.15 Together with its low parasitic absorption and
for perovskite optoelectronics, it is a solvent-free process,
making it more environmentally friendly.7 Vacuum processing
also allows for an accurate control on multiple-layer stacks.8 Special Issue: Jean-Luc Bredas Festschrift
Vacuum-deposited high efficiency solar cells generally rely on a Received: April 9, 2019
combination of intrinsic organic materials as charge selective Revised: June 3, 2019
layers and doped materials as charge transport/extraction Published: June 4, 2019

© 2019 American Chemical Society 6945 DOI: 10.1021/acs.chemmater.9b01396

Chem. Mater. 2019, 31, 6945−6949
Chemistry of Materials Article

Figure 1. (a) Schematic device architecture, (b) dark and (c) illuminated J−V curves (AM1.5G) of p-i-n devices according to panel a for three
different hole extraction architectures: MoO3 (not annealed)/TaTm, MoO3 (annealed)/TaTm, and MoO3 (not annealed)/TPBi/TaTm. Dashed
lines in panel c stand for reverse scan.

the possibility of being sublimed, it appears to be a good TaTm and C60 was 0.5 Å s−1 and 0.2−0.3 Å s−1 for the thinner BCP,
candidate for use as a hole extraction material in our system. TPBi, CBP, and TCTA layers. For the perovskite deposition, MAI
However, the low selectivity of MoO3 requires it to be used in and PbI2 were coevaporated by measuring the deposition rate of each
combination with more selective contacts. material in a different sensor and obtaining the total perovskite
thickness in a third one, leading to a 590 nm-thick perovskite. MoO3,
Here, we developed a method to modify the MoO3/electron Ag, and Au were evaporated in a second vacuum chamber using
blocking layer interface, obtaining an improved ohmic contact aluminum boats as sources by applying currents ranging from 2.0 to
suitable for p-i-n configuration, yet with the advantage of 4.5 A. For the solar cell characterization, the J−V characteristics were
avoiding any annealing step. In the n-i-p structure, this obtained using a solar simulator by Abet Technologies (model 10500
interfacial modification also acts as a protection layer which with an AM1.5G xenon lamp as the light source). Before each
prevents MoO3 diffusion into the underlying layers. As a result, measurement, the exact light intensity was determined using a
we fabricated the first high efficiency vacuum-deposited calibrated Si reference diode equipped with an infrared cutoff filter
perovskite solar cell with both charge extraction layers based (KG-3, Schott). The J−V curves were recorded between −0.2 and 1.2
on metal oxides. V with 0.01 V steps, integrating the signal for 20 ms after a 10 ms
delay. This corresponds to a speed of about 0.3 V s−1. The layout used
to test the solar cells has four equal areas (0.0651 cm2, defined as the
2. EXPERIMENTAL METHODS overlap between the ITO and the top metal contact) and measured
Photolithographically patterned ITO coated glass substrates were through a shadow mask with 0.0264 cm2 aperture. Absorption spectra
purchased from Naranjo Substrates. N4,N4,N4′,N4′-tetra([1,1′- were collected using a fiber optics-based Avantes Avaspec2048
biphenyl]-4-yl)-[1,1′:4′,1′-terphenyl]-4,4′-diamine (TaTm) was pro- Spectrometer. The work functions were determined by Kelvin probe
vided by Novaled GmbH. TiO2 nanoparticle suspensions were measurements using an Ambient Pressure Photoemission Spectros-
prepared in IMEC as previously reported.10 Fullerene (C60) was copy system from KP Technology.
purchased from Sigma Aldrich. PbI2 was purchased from Tokyo
Chemical Industry CO (TCI), while CH3NH3I (MAI), MoO3, and 3. RESULTS
bathocuproine (BCP) were purchased from Lumtec.
ITO prepatterned substrates were cleaned following a standard To evaluate the overall capability of MoO3 as a hole extraction
procedure in which they are cleaned with soap, water, deionized layer, the following p-i-n configuration was selected as it is
water, and isopropanol in a sonication bath, followed by UV similar to a previously tested one with an organic charge
treatment for 20 min. The TiO2 dispersion was deposited in air by extraction material (F6-TCNNQ):9 ITO/MoO3 (7 nm)/
spin-coating at 3000 rpm for 30 s and annealed at 100 °C for 30 min, TaTm (10 nm)/CH3NH3PbI3 (590 nm)/C60 (25 nm)/BCP
leading to a 50−80 nm-thick compact layer. For the deposition of the (7 nm)/Ag, as represented in Figure 1.
other materials through vacuum deposition, the samples were In Figure 1b and 1c, respectively, the typical current−voltage
transferred to a vacuum chamber integrated into a nitrogen-filled curves in the dark and under AM1.5G are depicted for three
glovebox (H2O and O2 < 0.1 ppm) and evacuated to a pressure of
10−6 mbar. The vacuum chamber was equipped with six temperature- different hole extraction layers. The devices with nonannealed
controlled evaporation sources (Creaphys) fitted with ceramic (as deposited) MoO3 show a high short circuit current density,
crucibles. The sources were directed upward with an angle of Jsc (21 mA/cm2), a reasonable open circuit voltage, Voc (1.03
approximately 90° with respect to the bottom of the evaporator. The V), but a rather poor fill factor, FF (73%). This may be due to
substrate holder to evaporation source distance was approximately 20 either a low layer conductivity, to an energy level mismatch
cm. Three quartz crystal microbalance (QCM) sensors were used: with the electron blocking material, or to a combination of
two monitoring the deposition rate of each evaporation source and a both. It has been reported that annealing the MoO3 layer
third one close to the substrate holder monitoring the total deposition increases the amount of oxygen vacancies and creates gap
rate. For thickness calibration, we individually sublimed the charge states which can contribute to the charge transport through a
transport materials. A calibration factor was obtained by comparing
the thickness inferred from the QCM sensors with that measured with shift in the Fermi level.16,17 We also find that an annealing
a mechanical profilometer (Ambios XP1). Then, the materials were process of the MoO3 layers for 10 min at 100 °C in nitrogen
sublimed at temperatures ranging from 60 to >300 °C, and the leads to an increase in FF and also in Jsc and Voc values. This is
evaporation rate was controlled by separate QCM sensors obtaining most likely the result of the work function (WF) closer to the
precisely the deposited thickness. In general, the deposition rate for vacuum level, as shown in Figure S1. Despite the higher power
6946 DOI: 10.1021/acs.chemmater.9b01396
Chem. Mater. 2019, 31, 6945−6949
Chemistry of Materials Article

conversion efficiencies obtained, the temperature requirement Table 2. Average PV Parameters of n-i-p Solar Cells with
for this annealing step is undesired. Therefore, we also MoO3 on Top of TaTm and TaTm/TPBi
considered an alternative approach to interfacial tuning by
PCE Jsc
adding an interlayer with a lower ionization energy (IE) than sample (%) Voc (V) (mA/cm2) FF (%)
the WF of the MoO3, as was recently reported for similar TaTm/MoO3 13.0 0.898 20.7 70.0
devices.18 We used 2,2′,2″-(1,3,5-benzinetriyl)-tris(1-phenyl-1- TaTm/TPBi (2 nm)/MoO3 18.8 1.084 21.3 81.4
H-benzimidazole) (TPBi) in a 2 nm thin interlayer between TaTm/TPBi (5 nm)/MoO3 11.0 1.015 21.2 51.0
the not annealed MoO3 and the hole transport/electron
blocking molecule TaTm. These devices have performance
very similar to that in the device with the annealed MoO3 This is most likely due to diffusion of MoO3 into TaTm, as
(Figure 1 and Table 1). The insertion of this interlayer is previously reported for various organic materials.19 While this
confirmed by contact angle measurements and does not lead to diffusion of MoO3 has the potential to improve the band
additional absorption (Figures S2 and S3). alignment without further annealing, it can also significantly
increase recombination losses due to the partial or complete
Table 1. Average Solar Cell Parameters for p-i-n Solar Cells oxidation of the TaTm electron blocking layer. The much
with the MoO3 Layer at the Front Contact higher dark current (hence, high conductivity) observed for
the TaTm/MoO3 containing device compared to the one with
PCE Jsc FF the TPBi thin interlayer (see Figure S5a) is in agreement with
samples (%) Voc (V) (mA/cm2) (%) this hypothesis. Such a reduced electron blocking and hence
MoO3 (not annealed) 15.8 1.034 21.0 73.0 increased charge recombination would indeed lead to a lower
MoO3 (annealed) 19.3 1.077 21.7 82.5 Voc without affecting the FF too much, as we observe. The
MoO3 (not annealed)/TPBi 19.0 1.080 21.4 82.0 device with a 2 nm TPBi interlayer in between the TaTm layer
and the MoO3 layer shows very good performance with high
Some other molecules were also sublimed and evaluated as Voc, Jsc, and FF, leading to a power conversion efficiency of
interlayers in between not annealed MoO3 and TaTm, namely, almost 19%. Although a similarly thin layer of TPBi also
TCTA (tris(4-carbazoyl-9-ylphenyl)amine) and CBP (4,4′- improved the performance of the p-i-n devices, it is rather
bis(N-carbazolyl)-1,1′-biphenyl). These molecular layers also surprising that such a thin layer is preventing the diffusion of
improve the device behavior but not to the extent that MoO3 MoO3 into the TaTm layer. Therefore, we also prepared a
annealing or TPBi insertion does (Figure S4), which is in similar device but now with a slightly thicker TPBi layer of 5
agreement with a lower difference between their IE (CBP = 6.0 nm. In these devices, the Voc is similar to what is observed for
eV and TcTa = 5.7 eV) and the HOMO level of TaTm, as was the device with a 2 nm layer, but the FF is much worse,
discussed also in ref 18. indicating a significant extraction barrier due to a strong
Because the TPBi interlayer enables the use of MoO3 as hole increase in the series resistance. This seems to indicate that the
extraction layer without the need for its annealing, this MoO3 interacts only with the surface of the TPBi layer, as
extraction layer can now be directly implemented in a n-i-p intrinsic TPBi is a poor hole transporter and undoped TPBi
structure with both contacts based on metal oxides. The needs to be very thin to effectively tunnel holes through. To
chosen device structure for the n-i-p devices is TiO2 (50−80 verify if MoO3 diffusion and the accompanied reduction in
nm)/C60 (10 nm)/CH3NH3PbI3 (590 nm)/TaTm (10 nm)/ electron blocking behavior can also be reduced by using
(TPBi 2 nm, optional)/MoO3 (5 nm)/Au (Figures 2a and 2b). thicker TaTm layers, such devices were prepared as well. In
In Figure 2c, the typical current−voltage curves in the dark these solar cells, the Voc recovers, as displayed in Figure S5b.
and under AM1.5G are depicted for the n-i-p devices with However, the MoO3 diffusion is a sensitive process and
three different hole extraction layers, and the parameters are difficult to control, and thicker TaTm n-i-p solar cells suffer
listed in Table 2. The devices with TaTm and the not annealed from poorer reproducibility and larger parasitic absorption. To
(as deposited) MoO3 show a rather low Voc but a decent FF. evaluate this TPBi interaction with MoO3, we substituted TPBi

Figure 2. Schematic device architectures of n-i-p devices with (a) and without (b) TPBi interlayer. (c) J−V curves under AM1.5G illumination
corresponding to devices with 0, 1, and 5 nm-thick TPBi interlayers. Dashed lines stand for reverse scan.

6947 DOI: 10.1021/acs.chemmater.9b01396

Chem. Mater. 2019, 31, 6945−6949
Chemistry of Materials Article

with TCTA. As in the p-i-n devices, it leads to similar device Pablo P. Boix: 0000-0001-9518-7549
performance (Figure S4). However, in the n-i-p configuration, Henk J. Bolink: 0000-0001-9784-6253
devices with a 2 nm-thick TCTA interlayer have low Voc values
Notes
similar to the TPBi-free devices, which is attributed to the
The authors declare no competing financial interest.

■
diffusion of MoO3 into both TCTA and TaTm, increasing
recombination and reducing the Voc (Figure S6). The
molecular weights of TCTA and TPBi are very similar, and ACKNOWLEDGMENTS
therefore, it is unlikely that this is the reason for the observed The authors would like to thank Alicia Forment for the contact
difference in device performance. The reactivity of MoO3 with angle measurements. Financial support is acknowledged from
organic compounds is well-known, particularly with a variety of the Spanish Ministry of Economy and Competitiveness
pyridine-like nitrogen atoms with a lone pair such as (MINECO) via the Unidad de Excelencia Mariá de Maeztu
bipyridine, where the ligand passivates MoO3 surface and MDM-2015-0538, MAT2017-88821-R, PCIN-2015-255,
blocks further Mo−oxo bond formation or with triazole family PCIN-2017-014, and the Generalitat Valenciana (Prometeo/
molecules forming a different layered structure.20,21 TPBi 2016/135). K.P.S.Z. acknowledges the financial support from
contains a lone pair which may lead to chemical modification Fundaçaõ de Amparo à Pesquisa do Estado de São Paulo
of the MoO3 surface, creating possibly the before mentioned (FAPESP 2018/05152-7). P.B. and M.S. thank the MINECO
oxygen vacancies that contribute to the charge transport, as for their RyC contracts. P.B. acknowledges the financial
other works have shown with pyridine-like nitrogen support from the Conselleria d’Educació, Investigació, Cultura
atoms.17,20,22 For this reason, a 2 nm layer of bathocuproine i Esport Valenciana (SEJI2017/2017/012).

■
(BCP, 2,9-dimethyl-4,7-diphenyl-1,10-phenanthroline) was
also inserted in the n-i-p device stack. The performance of REFERENCES
the resulting solar cells is very similar as in the case of the TPBi (1) Xu, W.; Cho, H.; Kim, Y.-H.; Kim, Y.-T.; Wolf, C.; Park, C.-G.;
interlayers (Figure S6), indicating that there is an interaction Lee, T.-W. Organometal Halide Perovskite Artificial Synapses. Adv.
between MoO3 and these lone pair nitrogen containing Mater. 2016, 28 (28), 5916−5922.
molecules that prevents the diffusion of MoO3. (2) Shi, D.; Adinolfi, V.; Comin, R.; Yuan, M.; Alarousu, E.; Buin, A.;
Considering both the p-i-n and n-i-p results, we hypothesize Chen, Y.; Hoogland, S.; Rothenberger, A.; Katsiev, K.; et al. Low
that TPBi reacts with the MoO3 surface through an electron Trap-State Density and Long Carrier Diffusion in Organolead
transfer, facilitating the charge transport. Not only that, but Trihalide Perovskite Single Crystals. Science (Washington, DC, U. S.)
also this TPBi-MoO3 surface species protects TaTm in the 2015, 347 (6221), 519−522.
inverted structure from the diffusion of MoO3 by forming a (3) De Wolf, S.; Holovsky, J.; Moon, S.-J.; Löper, P.; Niesen, B.;
Ledinsky, M.; Haug, F.-J.; Yum, J.-H.; Ballif, C. Organometallic Halide
passivation layer, which makes it possible to use this material in Perovskites: Sharp Optical Absorption Edge and Its Relation to
n-i-p configuration without compromising the performance or Photovoltaic Performance. J. Phys. Chem. Lett. 2014, 5 (6), 1035−
reproducibility of the device. 1039.
(4) Cheetham, A. K.; Rao, C. N. R.; Feller, R. K. Structural Diversity
4. CONCLUSIONS and Chemical Trends in Hybrid Inorganic?Organic Framework
In conclusion, the insertion of a thin layer of a wide bandgap Materials. Chem. Commun. 2006, No. 46, 4780.
(5) Li, W.; Wang, Z.; Deschler, F.; Gao, S.; Friend, R. H.; Cheetham,
conjugated molecule containing lone pair nitrogen atoms leads
A. K. Chemically Diverse and Multifunctional Hybrid Organic-
to efficient hole extraction in p-i-n and n-i-p architectures, in Inorganic Perovskites. Nat. Rev. Mater. 2017, 2 (3), 16099.
part due to the decoupling of the dipole moment and (6) Fakharuddin, A.; Schmidt-Mende, L.; Garcia-Belmonte, G.; Jose,
additionally preventing evaporation damage of MoO3 on the R.; Mora-Sero, I. Interfaces in Perovskite Solar Cells. Adv. Energy
organic charge transporting molecules. The efficient charge Mater. 2017, 7 (22), 1700623.
extraction and resulting device efficiency in the n-i-p (7) Á vila, J.; Momblona, C.; Boix, P. P.; Sessolo, M.; Bolink, H. J.
configuration is particularly promising as it can be used in Vapor-Deposited Perovskites: The Route to High-Performance Solar
tandem cells and low temperature demanding substrates. Cell Production? Joule 2017, 1 (3), 431−442.

■
(8) Á vila, J.; Momblona, C.; Boix, P.; Sessolo, M.; Anaya, M.;
ASSOCIATED CONTENT Lozano, G.; Vandewal, K.; Míguez, H.; Bolink, H. J. High Voltage
Vacuum-Deposited CH 3 NH 3 PbI 3 − CH 3 NH 3 PbI 3 Tandem
* Supporting Information
S Solar Cells. Energy Environ. Sci. 2018, 11 (11), 3292−3297.
The Supporting Information is available free of charge on the (9) Momblona, C.; Gil-Escrig, L. L.; Bandiello, E.; Sessolo, M.;
ACS Publications website at DOI: 10.1021/acs.chemma- Hutter, E. M.; Lederer, K.; Blochwitz-Nimoth, J. B.-N.; Bolink, H. J.;
ter.9b01396. et al. Efficient Vacuum Deposited P-i-n and n-i-p Perovskite Solar
Cells Employing Doped Charge Transport Layers. Energy Environ. Sci.
Kelvin probe and contact angle measurements of the 2016, 9 (11), 3456−3463.
MoO3 films with and without TPBi and device data with (10) Pérez-del-Rey, D.; Boix, P. P.; Sessolo, M.; Hadipour, A.;
alternative interfacial molecules (PDF) Bolink, H. J. Interfacial Modification for High-Efficiency Vapor-Phase-

■
Deposited Perovskite Solar Cells Based on a Metal Oxide Buffer
Layer. J. Phys. Chem. Lett. 2018, 9 (5), 1041−1046.
AUTHOR INFORMATION (11) You, H.; Dai, Y.; Zhang, Z.; Ma, D. Improved Performances of
Corresponding Authors Organic Light-Emitting Diodes with Metal Oxide as Anode Buffer. J.
*E-mail: [email protected]. Appl. Phys. 2007, 101 (2), 26105.
*E-mail: [email protected]. (12) Shrotriya, V.; Li, G.; Yao, Y.; Chu, C.-W.; Yang, Y. Transition
Metal Oxides as the Buffer Layer for Polymer Photovoltaic Cells.
ORCID Appl. Phys. Lett. 2006, 88 (7), 73508.
Kassio P. S. Zanoni: 0000-0003-4586-6126 (13) Gao, J.; Perkins, C. L.; Luther, J. M.; Hanna, M. C.; Chen, H.-
Michele Sessolo: 0000-0002-9189-3005 Y.; Semonin, O. E.; Nozik, A. J.; Ellingson, R. J.; Beard, M. C. N-Type

6948 DOI: 10.1021/acs.chemmater.9b01396

Chem. Mater. 2019, 31, 6945−6949
Chemistry of Materials Article

Transition Metal Oxide as a Hole Extraction Layer in PbS Quantum

Dot Solar Cells. Nano Lett. 2011, 11 (8), 3263−3266.
(14) Meyer, J.; Hamwi, S.; Kröger, M.; Kowalsky, W.; Riedl, T.;
Kahn, A. Transition Metal Oxides for Organic Electronics: Energetics,
Device Physics and Applications. Adv. Mater. 2012, 24 (40), 5408−
5427.
(15) Sanehira, E. M.; Tremolet de Villers, B. J.; Schulz, P.; Reese, M.
O.; Ferrere, S.; Zhu, K.; Lin, L. Y.; Berry, J. J.; Luther, J. M. Influence
of Electrode Interfaces on the Stability of Perovskite Solar Cells:
Reduced Degradation Using MoOx/Al for Hole Collection. ACS
Energy Lett. 2016, 1 (1), 38−45.
(16) Dasgupta, B.; Goh, W. P.; Ooi, Z. E.; Wong, L. M.; Jiang, C. Y.;
Ren, Y.; Tok, E. S.; Pan, J.; Zhang, J.; Chiam, S. Y. Enhanced
Extraction Rates through Gap States of Molybdenum Oxide Anode
Buffer. J. Phys. Chem. C 2013, 117 (18), 9206−9211.
(17) Greiner, M. T.; Helander, M. G.; Tang, W.-M.; Wang, Z.-B.;
Qiu, J.; Lu, Z.-H. Universal Energy-Level Alignment of Molecules on
Metal Oxides. Nat. Mater. 2012, 11, 76.
(18) Kotadiya, N. B.; Lu, H.; Mondal, A.; Ie, Y.; Andrienko, D.;
Blom, P. W. M.; Wetzelaer, G.-J. A. H. Publisher Correction:
Universal Strategy for Ohmic Hole Injection into Organic Semi-
conductors with High Ionization Energies. Nat. Mater. 2018, 17 (6),
563.
(19) White, R. T.; Thibau, E. S.; Lu, Z.-H. Interface Structure of
MoO3 on Organic Semiconductors. Sci. Rep. 2016, 6, 21109.
(20) Zapf, P. J.; Haushalter, R. C.; Zubieta, J. Hydrothermal
Synthesis and Structural Characterization of a Series of One-
Dimensional Organic/Inorganic Hybrid Materials of the [(MoO3)-
n(2,2′-Bipy)m] Family: [MoO3(2,2′-Bipy)], [Mo2O6(2,2′-Bipy)],
and [Mo3O9(2,2′-Bipy)2]. Chem. Mater. 1997, 9 (9), 2019−2024.
(21) Hagrman, P. J.; LaDuca, R. L.; Koo, H.-J.; Rarig, R.; Haushalter,
R. C.; Whangbo, M.-H.; Zubieta, J. Ligand Influences on the
Structures of Molybdenum Oxide Networks. Inorg. Chem. 2000, 39
(19), 4311−4317.
(22) Vasilopoulou, M.; Douvas, A. M.; Georgiadou, D. G.; Palilis, L.
C.; Kennou, S.; Sygellou, L.; Soultati, A.; Kostis, I.;
Papadimitropoulos, G.; Davazoglou, D.; et al. The Influence of
Hydrogenation and Oxygen Vacancies on Molybdenum Oxides Work
Function and Gap States for Application in Organic Optoelectronics.
J. Am. Chem. Soc. 2012, 134 (39), 16178−16187.

6949 DOI: 10.1021/acs.chemmater.9b01396

Chem. Mater. 2019, 31, 6945−6949