Tuning the interface properties of
Pervoskite Solar Cells with Graphene
and other two-dimensional materials
Aldo Di Carlo
CHOSE – Centre for Hybrid and Organic Solar Energy
University of Rome «Tor Vergata»
AND
National University of Science and Technology “MISiS”, Moscow, Russia
In Collaboration with
- Italian Institute of Technology: F. Bonaccorso group
- TEI Crete: E. Kymakis Group
Perovskite Solar Cells Architectures
Z. Son et al. J. of Photonics for Energy, 6(2), 022001 (2016)
Typical (mesoscopic) PSC
HTL
ETL
G. Divitini et al., Nature Energy (2016) 1, 15012
Several deposition methods
1) One-step procedure
Spin coating perovskite Drying RT Annealing 90-130°C
1a) 3 CH3NH3I + PbCl2 → CH3NH3PbI3-xClx + …
1b) Solvent engineering CH3NH3I + Pbl2 → CH3NH3PbI3
Nam Joong Jeon et al. Nature 517,476 (2015)
2) Two-step procedure
CH3NH3I Perovskite
PbI2 deposition
dipping layer
J. Burschka, et al. Nature, 2013, 499, 316
All in air, little humidity control
Two-step deposition on small area
0
Current Density (mA/cm )
2
-5 PCE=17 %
c-TiO2 | mp-TiO2 | MAPI | SPIRO
Small area cells PCE (%) =16.8 ± 0.15
-10
-15
c-TiO2 | mp-TiO2 | MAPI | P3HT
-20
PCE (%) =15.0 ± 0.9 (now max 18%)
-25
0.0 0.2 0.4 0.6 0.8 1.0 1.2 N. Yaghoobi Nia et al. ChemSusChem 10, 854 (2017)
Voltage (V)
Efficiency and stability are controlled by the interfaces
E. Palomares* et al. Chem Mat (2015) DOI: 10.1021/acs.chemmater.5b03902
Several strategies can be used to tune interface properties. Both chemical and
physical methods have been applied so far
Can we use Graphene and Related 2D Materials to properly change
interface properties ?
Perovskite and Graphene ???
Von Klitzing
Novoselov
I had to convince the Graphene Board …. Very difficult task !
Interface Engineering with Graphene and other 2D materials
Graphene based Interface Engineering
Graphene and other related two-dimensional (2D) materials (GRM) are
very suited to control interface properties:
• High values of charge mobility, Young’s
modulus, thermal conductivity, optical
transmittance
• More than 2000 two-dimensional
materials (G, GO, MoS2, WSe2, etc. etc.)
• Fine tuning of the electronic/optical
properties with functionalization and edge
modifications
• Size and stack control
• Optimized for printing processes
• Dispersed in several solvents
For this reason an increasing number of papers are devoted to use GRM
in photovoltaics and in particular in Perovskite Solar Cells
Interface engineering with 2D materials
Graphene in c-TiO2 rGO in m-TiO2
J. Tse-Wei Wang et al G. S. Han et al.,
Nano Lett., 2014, 14 (2), pp 724–730 ACS Appl. Mater. Interfaces, 2015,7, 23521
Graphene into m-TiO2 GO-Li* on m-TiO2 Perovskite/GO/HTM funct-RGO RGO
Perovskite/MoS2/HTM in the HTM as HTM
A. Agresti, B. Taheri et al., A. Agresti et al., A. Capasso et al., S. Casaluci et al., A. L. Palma et al.,
ChemSusChem 2016 Adv. Funct. Mater. 2016, Adv. Energy Mater. 2016, Proc. IEEE Nanotech. 2015 Nano Energy
ACS Energy Lett. (2017) T. Gatti et al. Adv. Func. Mat. 22, 349 (2016)
2016
Graphene at the PSC photoanode
Graphene + mTiO2
Spray-coating deposition technique
Perovskite
Spray coating of electron transport layer m-TiO2 on
Large-area Substrate
18NRT PASTE ETHANOL
SPRAY mTiO2 SOLUTION :
Optimization of TiO2 Spray Deposition (Coating Temperature)
b) 260 100
95
Thickness 90
240
Roughness 85
220 80
Roughness (nm)
75
Thickness (nm)
200 70
65
180 60
55
50
160 45
40
140 35
30
120 25
20
100 15
16
ng
px C
px C
px C
px C
ste 120°C
0
4
ste 80°
ste 30°
ste 30°
ste 30°
ati
=1
=1
=1
=1
=1
14
co
px
ray
ray
ray
ray
Spray mTiO2
ray
in
12 Spin mTiO2
Sp
Sp
Sp
Sp
Sp
32%
Sp
10
PCE (%)
8
6
Thickness and roughness of TiO2 films,
4
by spray coating deposition at 30°C,
2
80°C, 120 °C and by spin coating one. 0
°C
°C
C
°C
0°
0°
30
80
25
10
12
in
Sp
Deposition Temperature
The TiO2 film Spray coated at 30°C showed the lowest roughness of about 29 nm
Graphene flakes into the m-TiO2
1.05
1.0
1.00
Normalized Voc
Normalized PCE
0.9
0.95
0.8
0.90
0.7 0.85
0.6 0.80
0.0 0.5 1.0 1.5 2.0
Graphene concentration (%)
The optimum concentration of graphene
solution appears to be at 1 % v/v.
Graphene from F. Bonnaccorso group IIT (Italy) A. Agresti et al. ChemSusChem (2016) 9, 2516
Impact of Graphene on PSC.
SPIN COATING SPRAY COATING
18 20
Efficiency (%) with MPPT
SMALL AREA
18
17
Efficiency (%)
16
16
14
15 12
10
14 0.0 0.5 1.0 1.5 2.0 2.5 3.0
Time [min]
13
TiO2 TiO2 + Gr TiO2 TiO2 + Gr
area of the cell = 1 cm2
LARGE AREA
Graphene at the PSC photoanode
Graphene + mTiO2
GO-Li
Graphene at the PSC photoanode
GO-Li: Lithium neutralized GO on m-TiO2 layer
(E. Kymakis group, TEI Crete)
• Functionalization of GO with Li alkali metal, resulting in
the effective reduction of the WF of GO down to -
4.3eV
• Dispersion of GO-Li in EtOH and H2O (1:3) solution (0.3
mg/ml)
GO-Cl GO GO-Li
GO-Cl (WF increase) GO-Li (WF decrease)
E. Kymakis et al. Adv. Funct. Mater. (2013), 23, 2742, Adv. Funct. Mater. (2015), 25, 2213
Graphene in small area PSCs
12 devices for each type of PSC
• GO-Li showed the highest increase in
JSC values by penalizing the VOC;
• G into m-TiO2 improved the JSC by
retaining the VOC.
• G+GO-Li provided a trade-off between
improved JSC and VOC above 1V
A. Agresti et al. ACS Energy Lett. 2017, 2, 279−287
Graphene in small area PSCs
• Linear trends -> a good energy level matching
50 • The charge injection into TiO2 improves for G
40
mTiO2+Graphene doped mTiO2
Jsc [mA/cm ]
2
30 Graphene stabilize permanent ferrorelectric
mTiO2
20 dipole improving charge injection
Volonakis and Giustino JPCL, 2015, 6, 2496,
10
0
0.00 0.05 0.10 0.15 0.20
2
Light Power [W/cm ]
1.0
mTiO2 + Graphene
Normalized Voc
0.9 mTiO2
• PSC with G have faster rise-time than Ref.
PSC.
0.8
• VOC rise time is correlated to:
• the charge transfer from the active to
0.7 transport layer
0 1 2 3 • the active layer regeneration
Time [s]
A. Agresti et al. ChemSusChem (2016) DOI: 10.1002/cssc.201600942
Picosecond time-resolved PL
Investigated ETL TR-PL Emission
REF p E
λ=600nm e T λ=600nm
No r L
Graphene
Side A o +
Side B
v G
No Graphene
With
With Graphene
Graphene
TT
Biccari et al., Adv. En. Mater. 2017, 22,1701349
Time-integrated PL measurements
Perovskite crystalline structure Temperature-dependent PL
C- Cubic (T>300K)
T- Tetragonal (150K<T<300K)
O- Orthorhombic (T<150K)
Crystallinity of MAPI at low temperature is dependent from the ETL, which inhibits
phase change into the orthorhombic form.
Biccari et al., Adv. En. Mater. 2017, 22,1701349
Anode and Cathode with Graphene
G in mTiO2 improve charge-transport dynamics
Graphene + mTiO2
GO
GO interlayer:
• increase wettability of PSK
• Improve the interconnection between
perovskite and spiro-OMeTAD film, by
enhancing the charge-collection efficiency
W. Li, et al. J. Mater. Chem. A 2014, 2, 20105.
A. Agresti et al. ChemSusChem (2016) 9, 2516
Interface Engineering with G and GO
New18.2%
Max 19.6%
A. Agresti et al. ChemSusChem (2016) 9, 2516
Shelf life and light soaking stability
1.0 mTiO2+G Light soaking
Normalized PCE
1SUN, MPPT
0.8
Ref. Light soaking reveal better stability
0.6 of PSC with graphene doped
GO mTiO2 with respect to reference
0.4 cells. On the other hand all the
cells with GO interlayer show a
0.2 mTiO2+G/GO worse performance
0.0
0 2 4 6 8 10 12 14 16
time (h)
• MAPI crystals can undergo a hydration reaction triggered by
prolonged illumination. (Kamat et al. JACS 2016, 137, 1530)
• This can produce hydrogen iodine (HI) which can reduce the
GO (S. Pei et al. Carbon 2010, 48, 4466).
• G in mTiO2 reduces the trapping of the charges improving the Z. Fan et al. Joule 1, 548–562 (2017).;
stability of the cell (Ann et al arXiv:1604.07912) B. R. Southerland Joule 1, 421–430
A. Agresti et al. ChemSusChem (2016) DOI: 10.1002/cssc.201600942
From GO to MoS2 interlayer
MoS2: High mobility, chemical inertness
Liquid-phase exfoliation (LPE) of MoS2 in NMP
Solvent exchange with IPA (compatible with
perovskites)
MoS2 flakes are deposited by spin-coating on
perovskite
A. Capasso et al. Adv. Energy Mat. (2016) DOI: 10.1002/aenm.201600920
TiO2/Perovskite/MoS2/Spiro
Shelf-life
Reference PSCs without MoS2 have shown a η of 14.2% diminishing to η = 9.3%
after an endurance test of 550 h. MoS2-based PSCs instead reached a η of
13.3% after device fabrication which decreased to η = 12.4% after 550 h, proving
the enhanced stability of the cell due to the presence of MoS2 layer acting as an
active buffer layer.
A. Capasso et al. Adv. Energy Mat. (2016) DOI: 10.1002/aenm.201600920
Y. Busby et al, submitted (2017)
hBN reduces degradation of MAPI
The Thermal Stability Characterization of MAPbI3 with Surface Protection by a BN Thin Layer
(A) BN-perovskite-BN heterostructure fabrication process. (B) Schematic of the BN-perovskite-BN
heterostructure. (C) The BN-perovskite-BN heterostructure on a gas cell bottom chip.
(D) MAPbI3 with surface protection shows robust thermal stability. The thermal energy (85C in vacuum)
has not affected the crystalline structure of the MAPbI3 in 30 min of heating. Scale bar, 2 nm1.
Z. Fan et al. Joule 1, 548–562 (2017)
rGO stabilize Au/CuSCN interface
Au
rGO
CuSCN
N. Arora et al., Science 0.1126/science.aam5655 (2017).
From cells to modules
0.1 cm2
10 cm2
100 cm2
?
Fabrication of large area modules
Modules, fabricated on a 10x10 cm2 substrate area, consist in 8 series-connected
solar cells (active area 6.32 cm2) with an overall active area of 50.56 cm2, while the
module aperture ratio is approximately 76%
mTiO2+G, GO-Li and PbI2 are spin-coated, while SPIRO-OMeTAD is blade coated
A. Agresti et al. ACS Energy Letters (2017)
Graphene based module
Voltage (V) 1 SUN illumination condition
1 2 3 4 5 6 7 8 9 50 cm2 active area
0
10 cm
Ref. Module (A)
-20
Current (mA)
TiO2+G / Go-Li (D) 10 cm
-40
-60
-80
PCE = 11.6%
-100
PCE = 12.6%
-120
Electrical parameters
Module type
VOC (V) I (mA) FF (%) PCE(%) ΔPCE(%)
Ref 8.72 -112.8 59.4 11.6 -
mTiO2+G 8.23 -118.1 62.4 11.9 +3%
mTiO2/GOLi 8.46 -121.6 61.4 12.5 +8%
mTiO2+G/GOLi 8.6 -114.8 64.6 12.6 +9%
A. Agresti et al. ACS Energy Lett. 2017, 2, 279−287
Scaling up Graphene/Perovskite module
100 mm 125 mm
Active Area 50 cm2 Active Area 100 cm2
Perovskite module: Monolithic integration
Scribing of the FTO. Patterning of C-TiO2, Perovskite
Easy: laser (CO2 etc.) and HTM. Very critical
Small area Large area
Au
HTL
Perovskite
TiO2/Perovskite
ETL
FTO
GLASS
FF = 66% FF = 78%
F. Matteocci et al. Prog. Photovoltaics (2014) DOI: 10.1002/pip.2557
Improved P1 P2 P3 process
P1: Nd:YVO4, λ=1064 nm, 15 ns pulsed laser on FTO 44 μm wide scribing
SA: 50 μm wide spacing area (CCD alignment)
P2: Nd:YVO4, λ=355 nm, 10 ps pulsed laser on TiO2/pero/HTM 213 μm wide etching
SA: 50 μm wide spacing area (CCD alignment)
P3: Nd:YVO4, λ=532 nm, 10 ps pulsed, 25 µm wide scribing
5 cells module
(on 5x5 cm substrate) Previous report
Active Area: 14,52 cm2 Active Area = 5.02 cm2
P3 P2 P1
Aperture Area: 15,28 cm2 Aperture Area = 6 cm2
AR = 83,7 %
25μm 213μm 44μm PCE = 6,6 %
Aperture Ratio: 95%
387μm Aperture PCE = 5.52 %
PCE = 9.5%
Aperture PCE = 9.03% S. J. Moon et al., IEEE J.
Photovolt, 5 (4), 1087 (2015)
A. L. Palma et al. IEEE J. Photovoltaics 2017 DOI 10.1109/JPHOTOV.2017.2732223
Conclusions
Graphene and related material are very effective for interface engineering in Perovskite
Solar Cells.
Graphene oxide and Graphene doped m-TiO2 can be used to
boost the efficiency of PSC up to 19.6%. Issues with GO stability
MoS2 represents a viable solution for stability of cathode
interlayer.
The graphene-based modules (50 cm2 active area) showed
improved PCE values up to 12.6% and enlarged long-term stability.
Acknowledgments:
Ministry of Education and Science of the
Russian Federation in the framework of
Increase Competitiveness Program of
NUST “MISiS” (No. К2-2017-025),
implemented by a governmental decree
dated 16th of March 2013, N 211.
PERSEO - MIUR www.cheops-project.eu
CHOSE team @ ROME