Department of Physics Clarendon Laboratory Parks Road Oxford OX1 3PU e-mail: [email protected]
Photovoltaics and Optoelectronic Devices Group
Organic-inorganic perovskite thin film formation for high efficiency solar cells: A new paradigm for low cost solar energy
Henry J. Snaith
Solar energy resource
Terrestrial sun light
Global Power Demand
PV instillations
Perovskite is calcium titanium oxide or calcium titanate, with the chemical formula CaTiO3. The mineral was discovered by Gustav Rose in 1839 and is named after Russian mineralogist Count Lev Alekseevich Perovski (1792–1856).” All materials with the same crystal structure as CaTiO3, namely ABX3, are termed perovskites:
Perovskites
1892: 1st paper on lead halide perovskites
Structure deduced 1959: Kongelige Danske Videnskabernes Selskab, Matematisk-Fysike
Meddelelser (1959) 32, p1-p17 Author: Moller, C.K.
Title: The structure of cesium plumbo iodide Cs Pb I3
1978*: Hybrid Pb and Sn halide perovskites
Conducting Layered Organic-inorganic Halides Containing <110>-Oriented Perovskite Sheets D. B. Mitzi, S. Wang, C. A. Feild, C. A. Chess, A. M. Guloy IBM T. J. Watson Research Center, Post Office Box 218, Yorktown Heights, NY 10598, USA. Department of Chemistry and Texas Center for Superconductivity, University of Houston, Houston, TX 77204-5641, USA. Science 10 March 1995: Vol. 267 no. 5203 pp. 1473-1476 DOI: 10.1126/science.267.5203.1473
Abstract Single crystals of the layered organic-inorganic perovskites, [NH2C(I=NH2]2(CH3NH3)m SnmI3m+2, were prepared by an aqueous solution growth technique. In contrast to the recently discovered family, (C4H9NH3)2(CH3NH3)n-1SnnI3n+1, which consists of (100)-terminated perovskite layers, structure determination reveals an unusual structural class with sets of m <110>-oriented CH3NH3SnI3 perovskite sheets separated by iodoformamidinium cations. Whereas the m = 2 compound is semiconducting with a band gap of 0.33 ± 0.05 electron volt, increasing m leads to more metallic character. The ability to control perovskite sheet orientation through the choice of organic cation demonstrates the flexibility provided by organic-inorganic perovskites and adds an important handle for tailoring and understanding lower dimensional transport in layered perovskites.
Electrolyte (2006) Solid-State (2008)
1st Solar Cell Reports
Perovskites – Solar Cells
Solid-state perovskite “sensitized” solar cells
NN
NN
O O
OO
O
O
CH3CH3
OO
CH3 CH3
CH3 CH3
CH3 CH3
Spiro-OMeTAD
3*CH3NH3I + 1*PbCl2 CH3NH3PbI3-xClx
First devices
0.0 0.2 0.4 0.6 0.80
5
10
15
C
urre
nt D
ensi
ty (m
Acm
-2)
Applied Bias (V)
Jsc = 13.8 mA/cm2 Eff = 7.1 % Voc = 0.75 V FF = 0.69
Initial operation under full sun illumination
Charge Transport
Charge extraction must faster in perovskite “sensitized” than Dye (D102) Sensitized. Is the perovskite also conducting charge in the solar cells?
Replace TiO2 with Al2O3
Lets see what happens when we get rid of the porous TiO2…..
Replace TiO2 with Al2O3
1st set of devices:
Filed 3 patents on the 11th May 2012
Nature 485, Pages:486–489 Received 06 February 2012 Accepted 08 March 2012 Published online 23 May 2012
Precisely the opposite of our discover, nevertheless we promptly submitted our manuscript to Science on the 31st May…..
……The resulting solid-state dye-sensitized solar cells consist of CsSnI2.95F0.05 doped with SnF2, nanoporous TiO2 and the dye N719, and show conversion efficiencies of up to 10.2 per cent (8.51 per cent with a mask)………
Perovskite solar cells
Meso-Al2O3 η =10.9%
Meso-TiO2 η =7.6%
Planar Junction η =1.8%
Thickness dependence of meso-Al2O3
500 nm 500 nm 500 nm
A simple paradigm shift….
Diffusion length Estimation
Perovskite Species D (cm2s-1) LD (nm)
CH3NH3PbI3-xClx Electrons 0.042 ± 0.016 1094 ± 210
Holes 0.054 ± 0.022 1242 ± 250
CH3NH3PbI3 Electrons 0.017 ± 0.011 117 ± 38
Holes 0.011 ± 0.007 96 ± 29
B
C
S. Stranks et al. Science 2013 Also see Xing et al. Science 2013
LD > 1 μm in CH3NH3PbI3-xClx LD ~ 100 nm in CH3NH3PbI3
Why didn’t the thin films work? Answer: poor film formation
G. Eperon et al. Advanced Materials 2014 V. Burkolov et al. Applied Physical Review 2014
On mesoporous
alumina
On flat substrate
Dual source evaporation
Representation of layered perovskite (RNH3)2PbI4 and dual source evaporation process AX salt powder + BX2 salt powder = ABX3 film
Era et al. Chem. Mater. 1997, 9, 8-10
Vapour deposition of n-i-p heterojunction
Glass FTO
n-type contact Perovskite
p-type contact Ag/Au
Cross section of films and devices
Evaporated
Solution coated
Efficient Planar Heterojunction Solar Cells
Subtleties of solution processing:
Challenge, but not impossible to obtain highly uniform thin film
Efficient stable sustained output
Current voltage curve Stabilized power output
Plethora of techniques for thin film formation
(b)
(a)
(c)
(d)
a) N. J. Jeon, J. H. Noh, Y. C. Kim, W. S. Yang, S. Ryu, S. I. Seok, Nat Mater 2014, 13, 897-903. b) Z. Xiao, Q. Dong, C. Bi, Y. Shao, Y. Yuan, J. Huang, Advanced Materials 2014, 26, 6503-6509.
Organic p- and n-type contacts
a) b)
“Inverted” architecture essential for some tandem applications Capitalise upon all the processing and device architecture know-how from OPV
Aluminum
TiOX
P3HT:PCBM
PEDOT:PSS ITO
Glass
c.f. A. J. Heeger et al. Adv. Mater. 2006, 18, 572–576
PEDOT:PSS-Perovskite-PC60BM planar heterojunction devices
Jsc Eff Voc FFRegular 17.8 11.8 1.02 0.66Inverted 15.9 10.0 0.96 0.63
Jsc Eff Voc FFPET 14.4 6.4 0.88 0.51
Glass 14.4 6.3 0.92 0.47
P. Docampo et al. Nature Communications 2013
Laminated electrode
J. Troughton et al, 2015
Image courtesy of FutureTimeline.net
Do people want brown Buildings? Quite a hard sell….
Perovskite film (with thumb print)
Semi-transparent Solar Cells……
G. Eperon et al. ACS Nano (2014)
Semi-transparent solar cell operation
5 10 15 20 25 300
2
4
6
8
Pow
er c
over
sion
effi
cien
cy (%
)
Average visible transmittancethrough full device (%)
a
0.0 0.2 0.4 0.6 0.80
4
8
12
16
Cur
rent
den
sity
(mA
cm-2)
Voltage (V)
7.5-12.5% AVT 12.5-17.5% AVT 17.5-22.5% AVT 22.5%+ AVT
bUsing Semi-transparent thin gold electrodes
G. Eperon et al. ACS Nano (2013) 0 10 20 30 40 50 60
0
2
4
6
8
10
12 PCE Mean Maximum
Pow
er c
onve
rsio
n ef
ficie
ncy
for 1
ligh
t pas
s (%
)
Average visible transmittance of active layer (%)
c
“colour-tinted” semi-transparent perovskite solar cells
400 500 600 700 8000
20
40
60
80
100
120
Tran
smitt
ance
(%)
Wavelength (nm)
5mg/ml D102 in spiro-OMeTAD Semi-transparent perovskite cell +D102 Semi-transparent perovskite cell
b
0.0 0.2 0.4 0.6 0.80
2
4
6
8
10
12
14
Jsc(mAcm-2) Voc(V) FF η(%)Controls 12.0 0.85 0.63 6.4 D102-spiro 12.2 0.87 0.61 6.5
Cur
rent
den
sity
(mA
cm-2)
Voltage (V)
Semi-transparent perovskite cells + D102 Semi-transparent perovskite cells
c
No gain, but no loss G. Eperon et al. ACS Nano (in-press)
Ordered microstructure with templates
1D Photonic Crystal Solar Cell
W. Zhang et al. Nanoletters 2015 (in collaboration with Hernán Míguez CSIC)
Lead-free: CH3NH3SnI3 Perovskite
a=b= 8.7912 Å and c = 4.4770 Å
N. Noel et al. EES 2014 Also see: Hayase and co workers 2014; Kanitzidis and co workers Nat Photo 2014, K-Y Jen and coworkers 2014
Solar Cell results
N. Noel et al. EES 2014
Band gap ~ 1.23 eV; Voc ~ 0.88 V Eg-Voc ~ 0.35eV ?????
Perovskite solar cells: The rest of the world Pa
pers
pub
lishe
d
Year
Perovskites Certified 20% efficiency on lab based cells (small area)
>30% by end 2016?
Is this the future?
Production of silicon and silicon wafers Expensive, high-energy process generating high levels of waste material
Coke reduction in arc furnace at
1800 °C
Disolve in HCI at 300 °C + distillation
Chemical refinement
Siemens process at 900 °C
Modified Siemens process
Sand SiO2 + C
Metallurgical Grade
Silicon (MG Silicon)
Hydrogen Chloride
HCI HCI Hydroge
n High purity Trichlorosila
ne HSiCl3 High purity
polysilicon ∼ 9N
Polysilicon ∼ 6-7N Upgraded MG silicon >5N
Various Gasses
Electronic-grade
Solar-grade
Solar grade
Polysilicon
Melting Czochralski
pulling
Cutting/
squaring
Squared
ingot
Wire sawing
Cleaning
Wafer
Wings, top and tail recycling/etching
Slurry recycli
ng
from sand silicon to
from silicon wafer to
Production of perovskite cell Simpler, lower cost, lower embodied energy, massively reduced environmental impact, lowest LCOE
Incoming coated glass
Deposit titanium dioxide Deposit perovskite Finished panel
with back contact Deposit hole
transport layer
+ + = Yellow
precursor salt
White precursor
salt
Organic solvent
Perovskite liquid
formulation
from salts perovskite to
from perovskite liquid perovskite solar panel
to
Even simpler than conventional thin film
But…
Tuning the band gap of 3D perovskites
• When will a 3D perovskite form?
• When the A, B and X components fit together neatly in the crystal lattice.
• Assuming ionic radii of RA etc, For a close packed cubic perovskite the structure is possible, provided:
Tuning the band gap with cation size
Tuning the band gap with mixed anions
5.9 6.0 6.1 6.2 6.3 6.4
1.4
1.6
1.8
2.0
2.2
2.4
Tetragonal
y=1
g ()
Pseudocubic lattice parameter a* ( )
y=0
Cubic
G. Eperon et al. EES 2014 (Also See Noh et al. nanoletters 2012 for methylamonium trihalogenplumbates)
Formamidinium trihalogenplumbate (iodide-bromide mixed halide)
Perovskite/Silicon Tandem
Figure courtesy of M. McGehee, Standford Uni
Utility Scale PV
Combining Perovskites and Si in a tandem architecture could lead to >30% efficient modules
Why are organic-inorganic perovskites
such good solar cell materials???
Low energetic disorder
Technology Charge carrier lifetime
(micro seconds)
Urbach Energy (meV)
GaAs 1 7
c-Si 500* 11
MAPbI3 or MAPbI3-xClx >1 15
CIGS 0.25 25
Organics 0 001 50
(g)
PLQE and lasing!!
0 500 1000 1500 200030
40
50
60
70
PLQ
E (%
)
Excitation power (mW/cm²)
Very High Photo Luminescent quantum yield Negligible non-radiative decay
740 760 780 800 8200.0
0.5
1.0
1.5
2.0
Cou
nts
(x10
6 )
Fluence (µJ/cm2) 100 4 (scaled x25) PL Spectrum
Wavelength (nm)
Even room temperature lasing of as cast films within a cavity
Felix Deschler et al. JPCL 2014
Grain boundaries in thin film solar cells
L. M . Woods et al. NREL, Photovoltaic Solar Energy Conversion; 6-10 July 1998; Vienna, Austria
CdTe
A lot of electrons get trapped at the grain boundary, which introduces a lot of losse
“CIGS the wonder material” has ~ 100meV barrier at grain boundaries
Perovskites: “more wonderful”
Solution cast
Vapour Deposited
MAPbI3 potential barrier ~ 60 meV in dark MAPbI3 barrier ~ 15 meV in light MAPbI3-xClx barrier even lower MAPbI3-xClx is “almost like a singe crystal”
D. Cahen, G. Hodes and co-workers
Commercialisation:
Device and mini-module development Target: Develop stable and efficient materials stack
Develop processing methodology to deliver
Efficient modules at high yield
Deliver 1st product in 2017
Test and reliability laboratory Climatic testing to IEC61646 at 20*30 cm mini-
module scale
-85̊C/85% RH 1000hrs
+85 to -40̊C cycling 200 cycles
“Full Spectrum” Light soaking to AM1.5G 3000hrs
(not IEC)
High UV exposure
Perovskite Phase Stability (II) FAPbI3 Trigonal and Hexagonal phases possible at RTP
• Black (desired) 3D trigonal phase stable at 150oC in bulk and film
Koh, T. M. et al. J Phys Chem C (2013)
• 54 cycles -40 to +85oC (6hour cycle)
54 cycles
0
20
40
60
80
100
120
0 200 400 600 800 1000 1200
Normalised perovskite Colour
Intensity (%)
Stressing Time (hours)
Control(140)Control(115)A
B
C
D
Moisture sensitivity and Encapsulation Development
Interlayer assembly only
Encapsulation selection using 1000hr 85oC/85% baseline
Perovskite layer degradation by moisture ingress after early lamination failure
350hrs 0 hrs
Moisture ingress accelerates degradation
Interlayer desiccation Cover Glass
Interlayer
Perovskite Film
Edge Seal Gap Module Glass
Full sun light soaking 60⁰ C
0
0,2
0,4
0,6
0,8
1
1,2
1,4
1,6
0 500 1000 1500 2000
Nor
mal
ised
Pm
ax
Hours elapsed
Solar cells aged under load with no UV filter
Future direction for perovskite solar cells:
Acknowledgements
Funding EPSRC, ERC & FP7, Oxford John Fell Fund, Oxford Martin School, Royal Society.
Research group Collaborators: Perovskites: Takuru Murakami Tsutomu Miyasaka Oxford: Michael Johnston Laura Herz Robin Nicholas Victor Burlakov Alan Goriely Swansea: David Worsley, Tristan Watson et al. Milan: Annamaria Petrozza Giulia Grancini et al. Cambridge: Richard Friend Felix Deschler Michael Price et al.
+ Mingzhen liu Tomas Leijtens
Mike
Sam Giles
Konrad
James
Pablo
Antonio
Nakita
Films heated at 80C in air
[1] Schuettfort et al. Nano Lett. 2009, 9, 3871–3876 [2] Dabera et al. ACS Nano 2013, 7, 556–565 [3] Dissanayake et al. Nano Lett. 2011, 11, 286–290
63
Polymer wrapping with poly(3-hexylthiophene) (P3HT)
Solubilizing the SWNTs[1]
Making SWNTs more p-type[2,3]
Carbon Nanotube Functionalization
ref. [7]
Device construction
SWNT devices
HTL architecture Jsc [mA/cm2]
Voc [V] FF max. PCE
[%] av. PCE [%]
P3HT/SWNT (HiPCO) only 20.8 0.85 0.42 7.4 2.8±2.7
P3HT/SWNT(HiPCO)-PMMA 21.5 1.04 0.63 14.2 10.9±1.9
P3HT/SWNT(CG200)-PMMA 22.7 1.02 0.66 15.3 10.4±2.6
Thermal Stressing 80⁰C in air 96 hrs
PC PMMA PC PMMA
Not an IEC test!
68
Water Stability
Direct exposure to a stream of running water (60 s)
Exposing a perovskite with spiro-OMeTAD directly to water
Severin N. Habisreutinger [email protected]