Interfacial and bulk properties of hole transporting materials in perovskite solar cells: spiro-MeTADversusspiro-OMeTAD

Article abstract of DOI:10.1039/d0ta00623h

Two spiro-MeTAD compounds (1and2) were synthesized, characterized by experimental and quantum mechanical methods, and used as hole transporting materials (HTMs) in perovskite solar cells (PSCs). The new compounds differ from spiro-OMeTAD only by the presence of methyl substituents as compared to methoxy groups. This modification results in the absorption band blue shifting by ～20 nm as compared to spiro-OMeTAD, increased glass transition temperature for 2, and reduced ionization potentials by 0.02-0.12 eV. Hole mobilities five times larger were obtained for spiro-MeTAD/spiro-MeTAD, which is maintained in the presence of additives. Despite this improvement,J-Vmeasurements in PSCs resulted in a power conversion efficiency (PCE) of 17.2% and 17.05% for 1and2HTMs, respectively, as compared to 19.24% for spiro-OMeTAD. Photoluminescence measurements of perovskite:HTM layers indicate much stronger quenching in the case of spiro-OMeTAD/spiro-MeTAD. These results point to the dominant importance of the perovskite:HTM interfacial properties as compared to the HTM hole-transport properties in the bulk. Given that improved hole-mobility and energy-level alignment are the main targets of the current research efforts in this domain, our results alert to the necessity to prioritize the improvement of perovskite-HTM interaction properties.

Full text of DOI:10.1039/d0ta00623h

View Article Online
View Journal
Journal of
Materials Chemistry A
Materials for energy and sustainability
Accepted Manuscript
This article can be cited before page numbers have been issued, to do this please use: X. Sallenave, M.
Shasti, E. H. Anaraki , D. Volyniuk, J. V. Grazulevicius, S. M. Zakeeruddin, A. Mortezaali, M. Grätzel, A.
Hagfeldt and G. Sini, J. Mater. Chem. A, 2020, DOI: 10.1039/D0TA00623H.
Volume
Number 12
March 2018
Pages 4883-5230
6
This is an Accepted Manuscript, which has been through the
Royal Society of Chemistry peer review process and has been
accepted for publication.
28
Journal of
Materials Chemistry A
Materials for energy and sustainability
rsc.li/materials-a
Accepted Manuscripts are published online shortly after acceptance,
before technical editing, formatting and proof reading. Using this free
service, authors can make their results available to the community, in
citable form, before we publish the edited article. We will replace this
Accepted Manuscript with the edited and formatted Advance Article as
soon as it is available.
You can find more information about Accepted Manuscripts in the
Information for Authors.
Please note that technical editing may introduce minor changes to the
text and/or graphics, which may alter content. The journal’s standard
Terms & Conditions and the Ethical guidelines still apply. In no event
shall the Royal Society of Chemistry be held responsible for any errors
or omissions in this Accepted Manuscript or any consequences arising
from the use of any information it contains.
ISSN 2050-7488
COMMUNICATION
Zhenhai Wen et al.
An electrochemically neutralized energy-assisted low-cost
acid-alkaline electrolyzer for energy-saving electrolysis
hydrogen generation
rsc.li/materials-a

Page 1 of 34
Journal of Materials Chemistry A
TOC
View Article Online
DOI: 10.1039/D0TA00623H
Improving Perovskite-HTM interface properties in solar cells is more important than just improving
the HTM hole mobility.
Spiro-OMeTAD Spiro-MeTAD
1
2
-2
-3
-4
-5
TO
S
F
p
:
iro
s
1
2
-Opir
M
o-O
e
M
fitting
T
eT
A
AD
D
1
1
1
1
0
0
0
0
Spiro-OMeTAD
1
2
TOF
1
2
CELIV (without AD):
spiro-OMeTAD
1
+
CELIV
AD
2
MPPT (%)
CELIV (with AD):
no AD
spiro
1
2
(
-
a
er 2 months)
A
D
(AD=additives)
0
200
400
600
800
1000
1/2
1/2
1/2
Electric field (V /cm
)
1

Journal of Materials Chemistry A
Page 2 of 34
Interfacial and Bulk Properties of HView
o
Art
l
icle Online
e
DOI: 10.1039/D0TA00623H
Transporting Materials in Perovskite Solar Cells:
Spiro-MeTAD versus Spiro-OMeTAD
1
2,3
2
4
Xavier Sallenave , Mona Shasti , Elham Halvani Anaraki , Dmytro Volyniuk , Juozas Vidas
4
2
3*
2
Grazulevicius , Shaik M. Zakeeruddin , Abdollah Mortezaali , Michael Grätzel , Anders Hagfeldt
2*
, and Gjergji Sini1*
1
Laboratoire de Physicochimie des Polymères et des Interfaces, EA 2528 Université de Cergy-
Pontoise, 5 mail Gay-Lussac, Cergy-Pontoise Cedex, 95031, France.
2
École Polytechnique Fédérale de Lausanne, Institute of Chemical Science and Engineering (ISIC),
CH-1015 Lausanne, Switzerland
3
Department of Physics, Alzahra University, Tehran 1993893973, Iran
4
Kaunas University of Technology, Department of Polymer Chemistry and Technology, Radvilenu
pl. 19, LT-50254, Kaunas, Lithuania
Keywords: dft, energy disorder, interfaces, solar cells, perovskites, interaction energy
1

Page 3 of 34
Journal of Materials Chemistry A
Abstract
View Article Online
DOI: 10.1039/D0TA00623H
Two spiro-MeTAD compounds (1 and 2) were synthesized, characterized by experimental and
quantum mechanical methods, and used as hole transporting materials (HTMs) in perovskite solar
cells (PSC). The new compounds differ from spiro-OMeTAD only by the presence of methyl
substituents as compared to methoxy groups. This modification results in absorption-band blue shift
by ~20 nm as compared to spiro-OMeTAD, increased glass transition temperature for 2, and reduced
ionization potentials by 0.02-0.12 eV. Hole mobilities larger by five times were obtained for spiro-
MeTAD /spiro-MeTAD, which is maintained in the presence of additives. Despite this improvement,
J-V measurements in PSCs resulted in power conversion efficiency (PCE) of 17.2% and 17.05% for
1
and 2 HTMs, respectively, as compared to 19.24ꢀ% for spiro‐OMeTAD. Photoluminescence
measurements of perovskite:HTM layers indicate much stronger quenching in the case of spiro-
OMeTAD/spiro-MeTAD. These results point to the dominant importance of the perovskite/HTM
interfacial properties as compared to HTM hole-transport properties in the bulk. Given that improved
hole-mobility and energy-level alignment are the main targets of the current research efforts in this
domain, our results alert on the necessity to give priority to improving perovskite-HTM interaction
properties.
2

Journal of Materials Chemistry A
Page 4 of 34
1
. INTRODUCTION
View Article Online
DOI: 10.1039/D0TA00623H
Efficient and low cost solar-energy harvesting by means of photovoltaic devices constitutes one of
the main objectives for present research. Among different solar cell concepts, dye sensitized solar
cells (DSSC) and perovskite based solar cells (PSC) have shown increasing power conversion
1
2
efficiencies (PCE) recently reaching 14.3% and 25.2% , respectively. In both types of solar cells,
the prevalently used hole transporting material (HTM) is 2,2′,7,7′-tetrakis(N,N-di-p-
methoxyphenylamine)-9,9′-spirobifluorene (spiro-OMeTAD). While several phenomena jointly
determine the solar cell performances, here we focus on the impact of the HTM properties in PSC.
The success of spiro-OMeTAD is supposed to reside at least on the good energy level matching
between HOMO of this HTM with that of several dyes and of the perovskite absorber. However, this
-6
-5
2
3
semiconductor suffers from low hole mobility of roughly 10 -10 cm /Vs in the amorphous phase,
-3
2
4
reaching only ~10 cm /Vs in single crystals, which constitutes an important drawback for its use
in high efficiency photovoltaic devices. This issue has motivated intensive research efforts in order
to replace spiro-OMeTAD by more efficient HTMs, which dominance has indeed been already
5
-7
overcome. However, a general mechanism that could explain this success has not been identified
so far, due to multiple differences between the new molecules and spiro-OMeTAD. Consequently,
a seemingly random optimization process is still required with every new HTM, rendering their
further development unpredictable, cost intensive, and inefficient. Importantly, the research efforts
have been- and are primarily focused on the improved hole-mobility of the new HTMs, hence
ignoring other crucial parameters concerning the HTM-perovskite interactions at the interface.
As a means to render more efficient the research for new HTMs, a detailed analysis of the pros and
cons of spiro-OMeTAD properties is needed. While detailed analyses on the hole mobility of
amorphous spiro-OMeTAD are missing, interesting insights come from two recent studies: (i) a
3

Page 5 of 34
Journal of Materials Chemistry A
-3
2
theoretical upper-limit hole mobility of roughly 4.10 cm /Vs was found along one crystallographic
View Article Online
DOI: 10.1039/D0TA00623H
4
direction in spiro-OMeTAD single crystals, which is due to large electronic couplings along this
direction, whereas vanishingly small electronic couplings as low as only a few meV were found along
the other directions. Geometrical deformations, strong dipole moments, and steric hindrance were
found to constitute the bottlenecks preventing spiro-OMeTAD from high hole mobility. Accordingly,
very small electronic couplings and hole mobilities can be expected with regard to the hole mobility
of spiro-OMeTAD in the amorphous phase. (ii) Important energy disorder in hole transport levels
was found in amorphous spiro-OMeTAD, which, in the frame of a Gaussian distribution, was
characterized by a standard deviation (-parameter) of 0.16 eV.8
The origin of the important  value of spiro-OMeTAD can in part stem from the distribution of dipole
4
moments ranging between 0-5.3 Debye generated by the presence and random orientations of eight
methoxy groups. Indeed, methoxy substituents in triphenylamine compounds were found to reduce
the hole mobility as compared to the non-substituted one, due to the enhancement of the dipole
moment and their disorder.9-13 On the other hand, several methoxy- and methyl substituted
triphenylamine-based compounds were found to transport holes faster than non-substituted ones in
spite of important dipole moments,14, 15 with better effect in the case of methyl substituents. Detailed
analysis of these contradictory results pointed to the competition between two factors playing
16
1
6
important roles for the hole mobility: (i) the permanent molecular dipole moments impacting
negatively the hole transport, and (ii) the strength of intermolecular interaction. The latter factor was
found to strongly reduce the intra- and intermolecular geometrical randomness and their contribution
1
6
on the energy disorder. Interestingly, it was found that the methyl substituents are able to enhance
the intermolecular interactions, hence to reduce the geometrical disorder, without modifying the
permanent dipole moments of the parent molecules. On the contrary, while the methoxy groups
enhance the intermolecular interactions more importantly as compared to methyl groups, they
4

Journal of Materials Chemistry A
Page 6 of 34
increase the dipole moments of the parent molecules, thus destroying part of the beneficial effect
View Article Online
DOI: 10.1039/D0TA00623H
stemming from the stronger intermolecular interactions.
Motivated by these results, and aiming to obtain new HTMs exhibiting improved hole transport
properties, but very similar structural, electrochemical, and optical properties as compared to spiro-
OMeTAD, here we present two new molecules called spiro-MeTAD-1 and spiro-MeTAD-2, reported
hereafter as 1 and 2, differing from spiro-OMeTAD by only replacing the methoxy groups with
methyl groups linked in different positions and different numbers. The basic idea was to reduce the
energy disorder (value of -parameter) by removing the detrimental effect of the dipole moments
induced by the methoxy groups. We show that the new compounds indeed exhibit smaller , higher
hole mobility by at least five times as compared to spiro-OMeTAD, while simultaneously preserving
similar electronic and structural characteristics as the original molecule. Interestingly, the use of these
new compounds in the device structure of FTO/compactTiO /meso-porousTiO /perovskite/HTM/Au
2
2
results in PCE of 17.2% and 17.05% for 1 and 2-based devices, respectively, as compared to 19.2%
for spiro-OMeTAD. In view of the better transport properties found for the new HTM compounds as
compared to spiro-OMeTAD, the slightly smaller PCEs of the new HTMs point to the dominant
importance of the perovskite/HTM interfacial properties.
In spite of the slightly smaller PCE of the new HTMs as compared to spiro-OMeTAD, their use in
PS solar cells seems very promising. Last but not least, the new compounds are less costly than spiro-
OMeTAD with respect to laboratory procedure, which may result in decreased cost efficiency of their
use in PV devices.
The paper is organized as follows: we firstly focus on the intrinsic and bulk properties of the new
compounds, after which we compare their performances in perovskite solar cells with respect to spiro-
OMeTAD.
5

Page 7 of 34
Journal of Materials Chemistry A
2
. EXPERIMENTAL SECTION
View Article Online
DOI: 10.1039/D0TA00623H
2
.1. Materials. Synthesis. 1-iodo-3,5-dimethylbenzene (98%), Copper iodide CuI (>99%), potassium
phosphate K PO (97%), lithium amide LiNH (95%) were purchased from Alfa Aesar. Sodium tert-
3
4
2
butoxide (97%), palladium acetate Pd(OAc) (98%), tri-tert-butylphosphine P(tBu) (98%) were
2
3
purchased from Aldrich. 2,2’,7,7’-tetrabromo-9,9’-spirobifluorene (>98%) and p,p’-ditolylamine
>97%) were purchased from TCI Chemicals. DMF and toluene were dried by distillation from
(
calcium hydride CaH and sodium metal, respectively. Column chromatography and TLC were
2
performed on silica gel 60 (230-400 mesh, Merck) and silica gel 60 F254 plates (Merck), respectively.
Devices. PbI and PbBr were purchased from TCI Chemicals. Cesium iodide CsI was purchased
2
2
from abcr GmbH. FAI was purchased from Dyesol. Anhydrous DMF and DMSO were purchased
from Acros Chemicals.
.2. Characterizations. H and 13C NMR spectra were recorded on a Bruker DPX-400 NMR
1
2
1
13
spectrometer in 5 mm tubes with CDCl as solvent in all cases. The H and C chemical shifts were
3
referenced to solvent peaks, 7.26 and 77.2 ppm, respectively. High-resolution mass spectroscopy was
performed by the CNRS IMAGIF in Gif-sur-Yvette. Thermogravimetric analysis (TGA) was carried
out on a TA Instrument Q50 TGA under argon flow at a heating rate of 20°C/min. The temperature
of thermal degradation (Td5%) was measured at the point of 5% weight loss. Differential scanning
calorimetry analysis (DSC) was performed on a TA Instruments Q100 calorimeter, calibrated with
indium and flushed with argon. Samples were scanned from 0 °C to 200 °C at a heating rate of
2
0°C/min then rapidly cooled to 0 °C (quenching) and heated at the same rate to 200 °C. UV-Vis
measurements were performed at room temperature in spectrometric-grade THF solutions and on
glass subtrates using a Jasco (V-570) spectrometer. Cyclic voltamperometry (CV) was carried out
using an Autolab (AU128N, FRA2) electrochemical analyser with Pt disc used as working and
counter electrodes and Ag wire as pseudo-reference. The three electrodes were immersed in a CH Cl
2
2
6

Journal of Materials Chemistry A
Page 8 of 34
-3
solution with 0.1 M tetrabutylammonium hexafluorophosphate (n-Bu PF ) and 10 M 1 or 2
4
6
View Article Online
DOI: 10.1039/D0TA00623H
+
derivatives. Ferrocenium/ferrocene (Fc /Fc) redox potential was measured in order to calibrate the
pseudo-reference electrode.
2
.3. Ionization potentials and hole mobility measurements. To obtain ionization potentials (IPPES
)
of the compounds in solid-state, photoelectron emission spectra were taken under similar setup of
1
7
electron photoemission spectrometry in air (PESA) as it was described elsewhere. The vacuum-
deposited films onto fluorine doped tin oxide glass substrates were fabricated for PESA
3
measurements. The samples under electric field of 10 V/cm were excited by monochromatic light of
different photon energies utilizing coupled monochromator (CM110 1/8m) and deep UV deuterium
lamp ASBN-D130-CM. Electron photoemission current at a counter electrode under different
excitations were measured by the 6517B Keithley electrometer. IPPES values were taken from
corresponding photoelectron emission spectra as it was described previously.18
To investigate hole-transporting properties of the compounds in diode-like indium tin oxide
(
ITO)/tested film/Al structures, two methods, i.e. time-of-flight (TOF) and photo charge carrier
extraction by linearly increasing voltage (Photo-CELIV) were used. For the TOF measurements, 1.3-
.1 µm thick films were deposited by thermo-vacuum evaporation technique exploiting vacuum
3
equipment from Kurt J. Lesker in-built in an MB EcoVap4G glove box. For the Photo-CELIV
measurements, 0.4-0.6 µm thick films with and without additives were spin-coated utilizing the spin
coater SPIN 150. The conditions of the preparation of the samples with additives was the same as it
is described below for the preparation of devices. The Nd:YAG laser EKSPLA NL300 (355 nm, 3-6
ns) was used as the excitation source in setups of both TOF and Photo-CELIV. The corresponding
current transients were recorded by oscilloscope Tektronix TDS 3032C. Different voltages were
applied to the samples in the TOF experiments using the Keithley 6517B electrometer. Hole mobility
2
at different electric fields was calculated by the formula µTOF=d /V·t having layer thickness (d),
tr
7

Page 9 of 34
Journal of Materials Chemistry A
applied voltage (V), and transit time (t ) taken from the TOF transients. The triangle pulse with pulse
tr
View Article Online
DOI: 10.1039/D0TA00623H
duration (tpulse) of 10 µs and different amplitudes (U) was applied to the samples in the Photo-CELIV
experiments using the functional generator Tektronix AFG3011C. Hole mobilities were calculated at
different electric fields by the formula µCELIV=2d /A·tmax2 having the slope A=U/tpulse of triangle pulse
2
1
9
and time (t ) taken at the moment when maximum was seen in the CELIV photocurrent transients.
max
2
.4. Synthesis of target molecules. Synthesis of compound 4. To a solution of 1-iodo-3,5-
dimethylbenzene (2.1 g, 9 mmol) in dry DMF (25 mL) were added LiNH (414 mg, 18 mmol), K PO
4
2
3
(1.91 g, 9 mmol) and CuI (171 mg, 0.9 mmol) under argon. The reaction mixture was heated at 130
°
C for 24 h. After cooling, the solution was filtered under celite and the solid washed with CH Cl
2
2
(
100 mL). The organic phase was washed with water (4×25 mL) and dried with MgSO . After
4
evaporation of solvent, the residue was purified by column chromatography (SiO ) with cyclohexane/
2
1
CH Cl (1/1) as eluent to obtain white solid (740 mg, 75%). H NMR (CDCl )  6.71 (s, 4H), 6.60
2
2
3
1
3
(
s, 2H), 2.29 (s, 12H). C NMR (CDCl )      
3
General procedure for the synthesis of 1 and 2. To a solution of 2,2’,7,7’-tetrabromo-9,9’-
spirobifluorene (0.5 mmol) in dry toluene (15 mL) were added diphenylamine 3 or 4 (2.5 mmol),
sodium tert-butoxide (288 mg, 3 mmol) and Pd(OAc) (6 mg, 0.025 mmol). The solution was
2
degassed with argon for 30 min. Then P(tBu) (10 mg, 0,05 mmol) was added and the reaction mixture
3
was refluxed (115 °C) for 12 h. After cooling, the solution was poured into 30 mL of water. The
aqueous layer was extracted with ethyl acetate (2×30 mL).The organic phase was washed with brine
(
3×20 mL) and water (2×20 mL) and dried with MgSO . After evaporation of solvent, the residue
4
was purified by column chromatography (SiO ) with cyclohexane/ CH Cl (7/3) as eluent.
2
2
2
1
1
: white solid, 85% yield. H NMR (400 MHz, CDCl ):   7.39 (d, J = 8.2 Hz, 4H), 7.15-
3
1
3
7
.02 (m, 18H), 7.02-6.82 (m, 18H), 6.69 (s, 4H), 2.32 (s, 24H, CH ); C NMR (101 MHz, CDCl ):
3 3
8

Journal of Materials Chemistry A
Page 10 of 34

 150.2, 146.9, 136.3, 131.7, 129.7, 124.2, 123.4, 120.1, 119.5, 77.3, 20.8; HRMS (ESI+) calculated
View Article Online
DOI: 10.1039/D0TA00623H
for M+: 1096.5438, found: 1096.5465.
1
2
: white solid, 78% yield. H NMR (400 MHz, CDCl ):   7.36 (d, J = 7.8 Hz, 4H), 7.08-
3
1
3
6
1
.95 (m, 4H), 6.93-6.72 (m, 28H), 2.28 (s, 48H, CH ); C NMR (101 MHz, CDCl ):   149.6, 147.5,
3
3
46.6, 138.6, 124.4, 123.2, 122.3, 122.0, 121.9, 119.9, 119.7, 118.6, 77.2, 21.6; HRMS (ESI+)
calculated for M+: 1208.6713, found: 1208.6744.
2
.5. Device fabrication. After rinsing with deionised water, acetone, and ethanol, the FTO coated-
glass substrate were further cleaned with UV ozone treatment for 15 min. Then, 30 nm TiO compact
2
layer was deposited on FTO via spray pyrolysis using O as carrier gas at 450 °C from a precursor
2
solution of titanium diisopropoxide in anhydrous ethanol with the molarity of 0.2 M. Then, a
mesoporous TiO layer was deposited by spin coating for 15 s at 5000 rpm with a ramp of 2000 rpm/s,
2
using a 30 nm particle paste (Dyesol) diluted in ethanol with weight ratio of 1:6. After the spin
coating, the substrates were sintered at 450 °C for 30 min under dry air flow.
The perovskite composition used for this study is Cs FA Pb(I Br ) The perovskite precursor
0
.1
0.9
0.9
0.1 3.
solution was prepared by dissolving of FAI, PbI , CsI and PbBr powder in anhydrous DMF:DMSO
2
2
4
:1 (v:v) and the molarity of solution is fixed 1.35 M. This solution was kept at 70 °C for 30 min.
The perovskite solution was spin coated in a two steps program at 2000 and 6000 rpm for 10 and 20
s, respectively. During the second step, 200 μL of chlorobenzene was poured on the spinning substrate
1
0 s prior to the end of the program. The substrates were then annealed (usually at 100 °C) for 1 h in
a dry box. The substrates were cooled down for few minutes and a spiro-OMeTAD solution (74 mM
in chlorobenzene) was spin coated at 4000 rpm for 20 s. Spiro-OMeTAD was doped with Li-TFSI
salt and TBP. The molar ratio of additives for spiro-OMeTAD was: 0.5, and 3.3 for Li-TFSI, and
TBP, respectively. For spiro-MeTAD labelled as 1 and 2, the molarity of solution was chosen as 37
mM with the same molar ratio of additive. We also investigate the effect of molar ratio of the
9

Page 11 of 34
Journal of Materials Chemistry A
additives. For spiro-MeTAD labelled as 1-2 and 2-2, the molar ratio of additives was selected as: 1
View Article Online
DOI: 10.1039/D0TA00623H
and 6.6 for Li-TFSI, and TBP with the same solution molarity of 37 mM, respectively. Finally, 70-
8
0 nm of gold top electrode was thermally evaporated under high vacuum.
A spectrophotometer (Varian Cary5) was used to measure UV–Vis absorption spectra of the samples.
Current-voltage characteristics are measured by a digital source meter (Keithley 2400) with voltage
scan rate of 10 mV/s using a 450 W xenon light source under AM 1.5G. IPCE spectra were measured
2
under constant white light bias with an intensity of 10 mW/cm through a Gemini-180 double
monochromator.
2
0
21, 22
2
.6. Computational Details. Density functional methods (DFT) by employing the B3LYP
and
2
3

B97X-D functionals in conjunction with the 6-31G* basis set were used in this study. Symmetry
unconstrained geometry optimizations were carried out on the isolated molecules and on four type of
dimers constructed for each of compounds spiro-OMeTAD and 1 (Annex I, SI). Frequency
calculations for the isolated molecules and for the dimers indicated in all cases presence of only
positive frequencies, thus assuring that the obtained geometries correspond to minimum points on the
potential surfaces of individual molecules or dimers. The intermolecular interaction energy
calculations were performed with the B97X-D functional and the same basis set. The interaction
energies calculated on the bases of the total energies of the isolated molecules and of the dimers were
subsequently corrected for the zero point energy (ZPE), and for the basis set superposition error
(BSSE) following the counterpoise method. The B3LYP functional was used for the rest of the
calculations. The optical properties of the monomers and dimers were calculated by means of the
time dependent density functional theory method (TDDFT).24-28 The vertical ionization potentials
(IP) were calculated as energy difference between neutral and cation radical species at the neutral
state geometry. The intramolecular reorganization energies ( ) were calculated in the frame of the
i
2
9
potential surface method. The electronic couplings between the molecular orbitals were calculated
3
0
according to the direct interaction method. All calculations were performed with Gaussian 09
1
0

Journal of Materials Chemistry A
Page 12 of 34
3
1
program. A more detailed description on the choice of the theoretical method and the theoretical
View Article Online
DOI: 10.1039/D0TA00623H
model compounds and dimers is given in Annex I in the Supporting Information (SI).
3
3
. RESULTS AND DISCUSSION
.1 Synthesis and thermal characteristics
The synthetic route for the studied compounds is outlined in scheme 1. p,p’-ditolylamine 3 is a
commercially available product. Diphenylamine intermediate 4 was synthetized by a Cu-catalysed
amination of 1-iodo-3,5-dimethylbenzene. Spiro derivatives 1 and 2 containing substituted methyl
functional groups were synthesized by the Buchwald-Hartwig amination reaction, carried out
between 2,2’,7,7’-tetrabromo-9,9’-spirobifluorene and the respective bis(methylphenyl)amines
containing p-Me or m,m-Me substituents (3 or 4 respectively). The two spiro-MeTAD compounds
1
13
were purified by column chromatography and characterized by NMR spectroscopy ( H and C) and
mass spectrometry. All the analytical data were consistent with the proposed structures. The two
compounds have a good solubility in all of the common organic solvents.
X
X
Y
Y
Br
Br
Br
Br
X
X
N
N
X
X
X
X
+
X
X
X
X
Y
Y
Y
Y
X=H, Y=Me (1)
X=Me, Y=H (2)
H
N
X
Y
X
Y
X
X
N
N
X
X
X=H, Y=Me (3)
X=Me, Y=H (4)
Y
Y
X
X
Scheme 1. Synthesis of compounds 1-2 (Reagents and conditions: sodium tert-butoxide, Pd(OAc)2,
P(tBu) , toluene, 115 °C, 12 h.)
3
1
1

Page 13 of 34
Journal of Materials Chemistry A
It is worth noting that following the same synthesis path, the yields of compound 1 and 2 are roughly
View Article Online
DOI: 10.1039/D0TA00623H
2
0-30% larger as compared to that of spiro-OMeTAD. This is due to the smaller amount of solvent
and shorter time needed for the separation of the new compounds by chromatographic methods as
compared to spiro-OMeTAD. While we are not able to evaluate the impact of this aspect at the
industrial level, we believe that the same trend might be observed in both laboratory- and industrial
synthesis conditions.
Thermal properties of 1 and 2 were investigated by thermogravimetric analysis (TGA) and differential
scanning calorimetry (DSC) under argon atmosphere. The thermal decomposition (T ) and the glass
d
transition (T ) temperatures are summarized in Table 1. 1 and 2 showed high thermal stabilities with
g
Td5% values superior to 400 °C, which are high enough for applications in optoelectronic devices.
Confirmed by DSC analysis, compounds 1 and 2 are revealed as amorphous materials. All compounds
exhibit high glass transition temperatures (T ) ranging from 96 to 143 °C. In particular, compound 2
g
shows a significant increase in T (143 °C) compared to spiro-OMeTAD (110 °C), which may lead
g
to better thermal stability of the device over time.
Table 1. The thermal characteristics of spiro-OMeTAD and compounds 1 and 2.
Molecules
Td5% (°C)
450
Tg (°C)
96
1
2
443
143
spiro-OMeTAD
424
110
3
3
.2 Characteristics of pristine compounds
.2.1 Geometries and frontier orbitals of isolated compounds
The geometries of compounds 1 and 2 obtained in the absence of medium effects (“gas phase”) are
shown in Figure 1a. The geometry of spiro-OMeTAD is also shown for comparison. Except for the
substitutions on the diphenylamine moieties, quite similar global geometries were obtained.
1
2

Journal of Materials Chemistry A
Page 14 of 34
(
a)
View Article Online
DOI: 10.1039/D0TA00623H
1
2
Spiro-OMeTAD
(b)
2
1
Spiro-OMeTAD
0
.0
eV-
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
-
-
-
-
-
-
-
-
-
3.71
3.66
3.55
Figure 1. (a) Geometries of spiro-OMeTAD (with orientation of methoxy groups arbitrary chosen)
and of compounds 1 and 2 obtained at the B97XD/6-31G* level (“gas phase”). (b) Energy diagram
corresponding to frontier orbitals of spiro-OMeTAD and compounds 1 and 2 obtained at B3LYP/6-
3
1G* level in gas phase. The HOMO and LUMO representations are also shown. The left-right
ordering of the compounds follows the increasing HOMO energy.
The HOMOs and LUMOs of spiro-OMeTAD and compounds 1 and 2 also show quite similar shapes,
with the dominant contribution coming from the spirofluorene moieties (Figure 2). Both HOMO and
1
3

Page 15 of 34
Journal of Materials Chemistry A
LUMO levels of spiro-OMeTAD are higher than those of compounds 1 and 2 (Table 2), which is due
View Article Online
DOI: 10.1039/D0TA00623H
to the -donor effect of the methoxy groups in the former case. Nevertheless, this effect is more
pronounced in the case of HOMOs due to their important delocalization over the peripheral di-phenyl
moieties, thus resulting in smaller HOMO-LUMO gap in the case of spiro-OMeTAD as compared
to compounds 1 and 2. It is worth noting that the delocalization of the HOMOs over both fluorene
moieties (Figure 2) is expected to be strongly affected by the geometrical deformations occurring in
4
the real materials, in which case HOMO localization on only one of the arms is to be expected.
3.2.2 Impact of substituents on the strength of intermolecular interactions
The replacement of methoxy groups of spiro-OMeTAD by one or more methyl groups is
expected to influence the strength of intermolecular interactions in the bulk of these materials.
Indeed, changes in the intra- and intermolecular geometrical randomness, directly impacting
1
6
the energy disorder and hole mobility, have already been reported. Additionally and
importantly, the methyl vs methoxy substituents are expected to impact on the perovskite-
HTM interfacial interactions.
In order to obtain some insight on these aspects, four model dimers were optimized for spiro-
OMeTAD and compound 1 (Figure 2). The average dissociation energy in the case of 1 is
larger than for spiro-OMeTAD (23 kcal/mol and 12 kcal/mol respectively) indicating that the
intermolecular interactions are strengthened upon methyl substitutions. In the following we
show that the stronger intermolecular interactions for compounds 1 and 2 as compared to
spiro-OMeTAD impact several properties, such as polarization energies in the bulk of these
materials, hole transport properties, and characteristics of layer-deposition from solutions.
1
4

Journal of Materials Chemistry A
Page 16 of 34
Dimer-1
Dimer-2
Dimer-3
Dimer-4 View Article Online
DOI: 10.1039/D0TA00623H
Spiro-OMeTAD
1
Figure 2. Geometries of four dimer types constructed with spiro-OMeTAD and compound 1,
obtained at the B97XD/6-31G* level (“gas phase”). For clearer representation, one molecule of
each dimer is highlighted in yellow. Note that in each dimer type, the intermolecular interactions in
the case of spiro-OMeTAD and of compound 1 involve strictly equivalent molecular moieties, which
allows for a pertinent comparison between the corresponding dissociation energies.
3
.2.3 Physical and optical properties
Cyclic voltammetry (CV) was employed to investigate the redox properties and to estimate the
SS
ionization potentials (IP ) of compounds. The ionization potentials were additionally determined by
photoelectron spectroscopy in air (IPPES). The results of these measurements on compounds 1 and 2
and spiro-OMeTAD are shown in Figure 3 and in Table 2.
SS
SS
Table 2. Ionization potentials (IP ) (and electron affinities (EA )) from cyclic voltammetry (and
optical band gap Egopt), ionization potentials from photoelectron emission spectra (IPPES), and HOMO
(HOMO) and LUMO (LUMO) energies (B3LYP/6-31G* level in “gas” phase) corresponding to
compounds 1, 2 and Spiro-OMeTAD.
Compound
Eox vs Fc/Fc+
IPSS
EASS
IPPES
HOMO
LUMO
1
5

Page 17 of 34
Journal of Materials Chemistry A
(eV)
(eV)
(eV)
(eV)
(eV)
View Article Online
DOI: 10.1039/D0TA00623H
(V)
spiro-OMeTAD
0.04
0.08
4.84
4.88
4.86
1.84
1.82
1.80
4.94
4.92
4.90
-4.35
-4.49
-4.55
-0.80
-0.83
-0.84
1
2
0.055
IPSS and EASS values were estimated by using the ferrocene (Fc) ionization potential value (4.8 eV) as external standard,
according to the following equations: E opt
= 1240/λonset, I
SS
= 4.8+E1/2 vs Fc, E
SS
= I
SS
–E
opt. E
opt
designs the optical
g
p
A
p
g
g
band gaps estimated from the edges of electronic absorption spectra.
(b)
3.5
3.0
2.5
2.0
1.5
1.0
0.5
0.0
spiro-OMeTAD
1
2
liner fitting
4
.8 5.0 5.2 5.4 5.6 5.8 6.0 6.2 6.4
Photon energy (eV)
-3
Figure 3. (a) Cyclic voltammograms (CV) of compounds 1, 2 and spiro-OMeTAD in 10 M CH Cl
2
2
solutions at sweep rate of 100 mV/s, 0.1 M n-Bu PF . (b) Photoelectron emission spectra of the film
4
6
for vacuum deposited compounds.
SS
PES
Both IP and IP datasets provide similar trends between the spiro-MeTAD compounds, with 1 <
2
. Interestingly, the two experimental methods indicate opposite trends between spiro-OMeTAD and
SS
PES
spiro-MeTAD IP values: IP value of spiro-OMeTAD (4.84 eV) is smaller- whereas IP
4.94 eV) is larger as compared to spiro-MeTAD ones. The theoretical -HOMO values (representing
the ionization energies deduced in the frame of Koopmans’ theorem) are in better agreement with the
value
(
1
6

Journal of Materials Chemistry A
Page 18 of 34
SS
IP values, although an inversion can be observed in the latter values between compounds 1 and 2.
View Article Online
DOI: 10.1039/D0TA00623H
These discrepancies point to the crucial role of the polarization energy in different media, which are
expected to vary significantly between the isolated compounds (absence of contributions from the
polarization energy during theoretical calculations), the solution state with presence of electric field
around the electrode in CV, and the condensed packed solid films obtained by vacuum evaporation
(
PES measurements). The decreasing IPPES in the order spiro-OMeTAD > 1 > 2 might correlate with
the increasing intramolecular interaction energies in the same order, in turn enhancing the
intermolecular polarization energy in the vacuum deposited layers. It is worth mentioning that both
SS
PES
IP and IP values for vacuum deposited pristine spiro-OMeTAD without additives (4.84 eV and
4
.94 eV respectively) are smaller than the reported value of 5.1 eV for solution prepared Spiro-
OMeTAD with LiTFSI and TBP, again indicating the strong dependence of this parameter on the
environment conditions and additives type.32
Despite the above discrepancies, the two experimental datasets indicate global variation by only 0.04
eV, accordingly suggesting small variation on the V values in solar cell applications.
oc
Optical properties of compounds 1 and 2 were examined in dilute THF solution and in the thin film
solid state. The experimental absorption spectra are shown in Figure 4, whereas the max wavelengths
are collected in Table 3. All compounds exhibit similar optical behaviour in dilute solution, with a
strong absorption band located in the UV spectral domain (380-400 nm). It should be noted that the
absorption peaks of the two compounds 1 and 2, both in solution and in the solid state, are narrower
than those of spiro-OMeTAD allowing a better transparency of the HTM layer. The optical band gaps
(
Egopt) of 1-2 and spiro-OMeTAD were estimated from the edge of the first absorption peak in solution
Table 3). The smaller bandgap of spiro-OMeTAD as compared to compounds 1 and 2 seems to
(
correlate with the similar trend of the corresponding HOMO-LUMO gaps (see Figure 1).
Photoluminescence spectra of HTMs in chlorobenzene are shown in Figure 4c. The trend of the
emission peaks at 420 nm for spiro-OMeTAD33, 34 and at 407 nm for compounds 1 and 2 is in line
1
7

Page 19 of 34
Journal of Materials Chemistry A
with the similar trend of the absorption onset wavelengths and with the one between the
View Article Online
DOI: 10.1039/D0TA00623H
corresponding HOMO-LUMO gaps.
(b)
(a)
1
1
1
1,00E+00
,00E-01
8,00E- 00 1. 8
,00E-01
.6
9
0
,8
.8
0
1
2
7
0 ., 66
6,00E-01
0
0
spiro-OMeTAD 5,00E-01
0
0 ., 44
4,00E-01
.4
0
3
,00E-01
.2
0
0 ., 22
2,00E-01
0
1
,00E-01
0
0
0,00E+00
0
3
00
350
400
450
500
300
350
400
450
500
Wavelength (nm)
Wavelength (nm)
(c)
Figure 4. UV spectra of (a) dilute THF solutions (normalized at Abs = 1), and (b) in solid state (spin
coated on glass) of compounds 1, 2 and spiro-OMeTAD. (c) Photoluminescence spectra of 1, 2 and
.
-5
spiro-OMeTAD (1 10 M) in chlorobenzene.
Table 3. Optical characteristics of compounds 1, 2 and spiro-OMeTAD. The values in parentheses
correspond to the theoretical values (B3LYP/6-31G* level in gas phase). The two theoretical values
in the case of spiro-OMeTAD correspond to two different combinations between O-CH bond
3
orientations.
1
8

Journal of Materials Chemistry A
Page 20 of 34
Compound
max
onset
max
Egopt
View Article Online
DOI: 10.1039/D0TA00623H
(
eV)
(
solution, nm)
385 (389)
386 (386)
(solution, nm) (solid, nm)
1
404
404
412
386
386
378
3.06
3.06
3.0
2
spiro-OMeTAD
386 (393-404)
3
.2.4 Charge transport properties
Figure 5 shows electric field dependence of hole-drift mobility of compounds 1, 2 and spiro-
OMeTAD at room temperature (RT), measured by the time-of-flight (TOF) method in vacuum
deposited layers, and by the CELIV method in spin-coated layers with and without additives
(Experimental section).
-3
2
TOF hole mobilities. Hole mobilities of roughly five times faster (ca. 5·10 cm /Vs) were recorded
5
at an electric field of 3.6·10 V/cm for the layers of compounds 1 and 2, as compared to spiro-
OMeTAD (9·10-4 cm /Vs). Moreover, current-time signals shown in Figure S1 (Supporting
information, SI) suggest less dispersive hole transport for compounds 1 and 2, which comforts the
conclusion that the new compounds globally exhibit improved hole-transport properties as compared
to spiro-OMeTAD.
2
It is worth noting that the hole mobility found for the vacuum-evaporated “home-made” spiro-
OMeTAD is at least 1-2 orders of magnitude larger than values found in literature for spin coated
-5
2
-6
2
3
35
layers (4·10 cm /Vs and 1.7·10 cm /Vs from SCLC and FET measurements, respectively) and
-3
2
only one order of magnitude smaller than the FET value of 1.3·10 cm /Vs found for single crystals
3
5
of spiro-OMeTAD. While these comparisons suggest enhanced degree of ordering in our layers
1
9

Page 21 of 34
Journal of Materials Chemistry A
obtained by vacuum evaporation, the pertinent information remains that the transport properties of
View Article Online
DOI: 10.1039/D0TA00623H
the methyl-containing compounds 1 and 2 are improved as compared to spiro-OMeTAD.
TOF:
spiro-OMeTAD
-2
-3
-4
-5
1
1
1
1
0
0
0
0
1
2
fitting
CELIV (without AD):
spiro-OMeTAD
1
2
CELIV (with AD):
spiro-OMeTAD
1
2
0
200
400
600
800
1000
1/2
1/2
1/2
Electric field (V /cm )
Figure 5. Electric field dependencies of hole-drift mobility of compounds 1, 2 and spiro-OMeTAD
in vacuum deposited layers (measured by the TOF method) and spin-coated layers with and without
additives (AD) (measured by the CELIV method) at room temperature (RT).
Table 4. Hole transport parameters of compounds 1, 2 and spiro-OMeTAD. The hole mobilities are
.
4
approximate values taken at the electric field of 3.6 10 V/cm. Theoretical parameters calculated for
compounds 1, 2 and spiro-OMeTAD at the B3LYP/6-31G* and B97WD/6-31G* levels in gas phase
are also given, such as HOMO and LUMO energies, intramolecular reorganization energies (h),
HOMO-HOMO electronic couplings (tH-H) deduced from a selected set of optimized dimers (average
over 4 dimers for each compound, Figure 2), and dipole moments of isolated compounds.
μ
h
tH-H
(eV)
Dip. Mom.
(Debye)
0.001
σ
μ0
Compound
2
2
(
cm /Vs)
(
eV)
(cm /Vs)
(eV)
0.213
1
5.10-3
0.084
0.090
0.003
2
0

Journal of Materials Chemistry A
Page 22 of 34
2
5.10-3
9.10-4
0.086
0.106
0.104
0.360
0.375
0.133
-
0.05
View Article Online
DOI: 10.1039/D0TA00623H
spiro-OMeTAD
0.004
0 - 5.3*
*
range of dipole moments obtained by considering a “random” orientation of the methoxy groups
(selection of relative orientations).
In order to obtain some insight on the reasons for the observed improvements, temperature dependent
TOF hole-mobility measurements were performed, and different intra- and intermolecular properties
for compounds 1, 2 and spiro-OMeTAD were calculated. Additionally, four model dimers were
optimized for each compound (Figure 2) and the HOMO-HOMO electronic couplings between
adjacent molecules were calculated.
The energetic disorder parameter (σ Table 4) was deduced within the Gaussian disorder
model, according to the following equation:36
ˆ
3
1
2
2
2
2
2

  exp[( ) ]exp[C(ˆ   )E ]
0
(1)
Here, σ is the standard deviation of the hopping site manifold, Σ is the positional disorder, C is a
constant, and ˆ = σ/kT.
Details on different methods are given in the Figures S2, S3, and S4 (SI) and only selected
numerical results are given in Table 4. The important result is that the energetic disorder
parameter σ decreases importantly from 0.106 eV to 0.084 eV in the order spiro-OMeTAD >
2
= 1 (Table 4), which is consistent with the increasing hole-mobilities in the same order.
Interestingly, the largest  value, which characterizes the intramolecular charge transport

2
properties, was obtained for spiro-OMeTAD (0.36 cm /Vs, Table 4). Given the similar
electronic couplings shown in Table 4, the larger  of spiro-OMeTAD simply translates to

lower intramolecular reorganization energies  as compared to 1 and 2 compounds. However,
h
the highest σ and the lowest mobility values found for spiro-OMeTAD suggest that the
2
1

Page 23 of 34
Journal of Materials Chemistry A
differences in the hole mobilities of these compounds are dominated by disorder phenomena.
View Article Online
DOI: 10.1039/D0TA00623H
Indeed, the almost zero dipole moments of the methylated compounds (Table 4) indicate
absence of the detrimental effect from the dipole-moment disorder. This effect is in strike
contrast to the large dipole moments as high as 5.3 Debye calculated for spiro-OMeTAD,
which has been found to negatively impact the charge-transport properties of organic
semiconductors.9-12 Additionally, stronger intermolecular interactions were calculated for the
1
compound as compared to spiro-OMeTAD (Table 4), which effect has been shown to
reduce the geometrical contribution to the disorder parameter of 1 and 2 compounds as
compared to methoxy-substituted ones.16
The global hole mobility improvement for methyl-substituted compounds wrt spiro-OMeTAD
results consequently from a delicate balance between two opposite factors, the positive effect
being dominant: (i) detrimental increase in intramolecular reorganization energy ( ), and (ii)
h
positive impact from the decrease in energy disorder parameter (), in turn stemming from
the decrease in dipole moment.
CELIV hole mobilities. The electric field dependences of CELIV hole mobilities in spin-coated layers
of pure spiro-OMeTAD and spiro-MeTAD compounds are shown in Figure 5. Similar global trends
wrt TOF results can be observed, with hole mobilities increasing in the order spiro-OMeTAD < 1 <
2
.
In view of the large utilization of additives in perovskite solar cells, we performed hole mobility
measurements of spiro-OMeTAD, 1, and 2 compounds in spin-cast layers with- and without
additives. The results shown in Figure 5 indicate hole-mobility improvement for spiro-OMeTAD and
1
1
compounds, but missing impact in the hole mobility of 2. While the results for spiro-OMeTAD and
compounds correspond to the common-knowledge based expectations, the missing sensitivity of 2
hole-mobility wrt the presence of additives is intriguing. We suspect these differences to correlate
2
2

Journal of Materials Chemistry A
Page 24 of 34
with the increasing intermolecular interaction energies, probably making the HTM-additives mixture
View Article Online
DOI: 10.1039/D0TA00623H
less homogeneous in the case of strong HTM-HTM interactions.
3
.3 Device characteristics
A schematic energy level diagram of the solar cell device is illustrated in Figure 6. After light
absorption by the perovskite layer, photogenerated electrons and holes are moved to the photoanode
(
TiO layer) and hole conductor layer, respectively. The HOMO level of HTMs should be higher than
2
the valence band of the perovskite layer for hole extraction. As shown in CV results, the HOMO
levels of both spiro-MeTAD compounds 1 and 2 are a little deeper than spiro-OMeTAD.
Figure 6. Schematic energy level diagram of the solar cell device. The ionization potentials of spiro-
OMeTAD is as reported in Ref32
As shown in figure S7, the Scanning electron microscopy (SEM) images in cross-section view reveal
that roughly 126-148 nm thick HTL has covered the perovskite film in devices based on new
compounds while the thickness of spiro-OMeTAD is roughly 190-196 nm in control devices.
According to owing larger hole mobility, this thickness is appropriate for hole extraction process in
new compounds-based devices, as proven in J-V characteristics.
2
3

Page 25 of 34
Journal of Materials Chemistry A
Figure 7 shows photoluminescence spectra of perovskite-HTM systems for different HTMs
View Article Online
DOI: 10.1039/D0TA00623H
deposited on the perovskite layer. The spectra show a band-edge transition of 773 nm related to the
perovskite layer. The band-intensity corresponding to spiro-OMeTAD sample is lower than for other
HTMs, indicating lower hole recombination or rapid extraction of holes across the interface.37
Figure 7. The PL spectra of the samples. The notations 1 and 1-2 correspond to two different
concentrations of additives, with the concentration in the case of 1-2 being twice as large as
the one for 1. Same differences apply in the case of 2 and 2-2.
Current-voltage characteristic curves are shown in Figure 8 under sun illumination and in dark. The
data is summarized in Table 5. The control sample with spiro-OMeTAD as HTM showed a power
conversion efficiency (PCE) of 19.2% and 16.9% in backward and forward scan direction,
respectively, with an open circuit voltage (V ) of 1.128 V, a short current density (J ) of 22.86 mA
OC
SC
−
2
cm and a fill factor (FF) of 74.6% under reverse scan conditions. For 1-based devices we observe a
2
small hysteresis with VOC of 1.109 V, J of 22.57 mA/cm , FF of 68.7% resulting in a PCE of 17.2%
SC
in backward scan. Spiro-OMeTAD-, 1-, and 2-based devices show hysteresis which can be due to
inappropriate ambient atmospheric condition during the film deposition. Although the highest PCE
of 18.0% in backward voltage scan can be obtained for devices labelled as 2-2 with VOC of 1.124 V,
2
JSC of 22.86 mA/cm and FF of 70.2%, the hysteresis effect is increased as well with increasing
2
4

Journal of Materials Chemistry A
Page 26 of 34
concentration of additives (molar ratios of additives in 2-2 being two times larger as compared to 2).
View Article Online
DOI: 10.1039/D0TA00623H
Increasing the amount of additives has thus a negative impact on the photovoltaic performance of 1-
2
and 2-2 -based devices. As shown, hysteresis is increased while PCE is decreased compared to 1
and 2 -based devices. Interestingly, the performances of spiro-OMeTAD PSCs remain basically
unaffected when the additive amounts are increased by the same amount (two times).
Figure 8. J-V curves of the best cells recorded with reverse (from VOC to J ) scan direction.
SC
Table 5. Photovoltaic parameters of different HTM based-devices under backward and forward
scan direction.
2
HTM
Scan
direction
Rev
J (mA/cm )
V (V)
PCE (%) FF (%)
Integrated
Jsc (IPCE)
22.04
SC
OC
Spiro-
OMETAD
1
22.86
23.19
22.57
21.29
22.18
1.128
1.085
1.109
1.099
1.042
19.24
16.99
17.2
74.6
67.5
68.7
65.7
68
For
Rev
21.87
19.02
For
15.38
15.72
1
-2
Rev
2
5

Page 27 of 34
Journal of Materials Chemistry A
For
Rev
For
Rev
For
22.45
22.42
22.91
22.86
23.28
0.984
1.119
1.086
1.124
1.074
10.28
17.05
15.24
18.04
12.18
46.5
68
View Article Online
DOI: 10.1039/D0TA00623H
2
21.66
62.1
70.2
48.7
2
-2
20.54
Photovoltaic parameters from different devices in backward direction are shown in Figure 9. 1-based
devices shows higher JSC and lower Voc compared to 2-based devices. 2-based devices have the
lowest fill factor, that we suspect is resulting from the inhomogeneity of film and interface problems.
The average PCEs of 1-based devices were 17% and the PCEs for 2-based devices were 15%. With
increasing additive concentration, 2-2 based devices show better performance in J , V , FF and
SC
OC
PCE than 2. These results revealed that the performance of 1 based device outperforms 2.
2
6

Journal of Materials Chemistry A
Page 28 of 34
View Article Online
DOI: 10.1039/D0TA00623H
Figure 9. Statistical plot of J , V , FF and PCE of different devices.
SC
OC
Incident photon-to-current efficiency (IPCE) curves for devices are shown in Figure 10. The
2
integrated current obtained by IPCE is 22.04, 21.87, 19.02, 21.66, and 20.54 mA/cm for spiro-
OMeTAD, 1, 1-2, 2 and 2-2 based devices, respectively. As shown in this figure, and from the
integrated current obtained from IPCE, the obtained current decreased dramatically with increasing
amount of additives. Intriguingly, this effect is less pronounced for 2 as compared 1.
Figure 10. Incident photon-to-current efficiency (IPCE) curves for HTM.
2
7

Page 29 of 34
Journal of Materials Chemistry A
Figure 11 shows the maximum power conversion efficiency as function of time from maximum
View Article Online
DOI: 10.1039/D0TA00623H
power point tracking measurements under AM1.5 simulated solar light. These are initial values
measured after device preparation and after 2 months of storing devices at dark in dry-box with
humidity of less than 2% RH. The values are tabulated in Table 6. Maximum power point tracking
(MMPT) for spiro-OMeTAD, 1 and 2-2 -based devices are 18.26, 15.08 and 15.33, respectively.
As a final and general comment, we observe that the trends observed in the PV performances of the
studied compounds are at odds wrt the larger hole mobility of 1 in presence of additives as compared
to both other compounds (Figure 5), thus pointing to other dominant factors. Additionally, these
SS
PES
trends could hardly be related to differences in the corresponding HOMO levels, as the IP or IP
values for 1 and 2 compounds vary by only 0.02 eV and by a maximum of only 0.04 eV with spiro-
OMeTAD. Impact of the additives in the HTM layer morphology, or differences in interfacial
properties could thus be in play, such as, amongst others, eventual energy level-shifts, the quality and
strength of perovskite-HTM contact and interactions, polarity mismatch, etc.
Table 6. Maximum power point tracking (MPPT) for HTM devices for initial measurements and after
2
months.
HTMs
Spiro-
1
1-2
2
2-2
OMeTAD
18.26
MPPT(%)
15.076
15.01
12.68
------
14.79
14.43
15.33
-----
MPPT(%) After 17.93
month
2
2
8