interface modification as a route to improving performance of...

Interface Modification as a Route to

Improving Performance of Vacuum-

Based Perovskite Solar Cells

Speaker：Run Xu (徐闰）Cooperators: Fei Xu and Linjun Wang

E-mail：[email protected]

School of Materials Science and Engineering, Shanghai University

mailto:[email protected]

Contents：

➢ Part I: Research Background and Motivation

➢ Part II: Growth mechanism of dual-source evaporated perovskite films

➢ Part III: Dual-source evaporated perovskite solar cells

➢ Part IV: Summary

Part I

Research Background and Motivation

• Development of PSCs

•

➢ developing rapidly

➢ >20%

➢ “Next Big Thing in Photovoltaics”

CH3NH3PbI3 (MAPbI3, MPI)

DSSC

PSC

• Crystal & Band Structure

Fig1.2 Crystal structure

➢ MX6 octahedral

➢ A fills the octahedral holes

Eg=~1.5eV

CB: Pb p electron

VB: I p electron

➢ direct bandgap

➢ ideal value

Fig1.3 Band structure

• Preparation Methods

Fig1.5 (a)one-step (b)two-step (c) dual-source evaporation

333233 PbINHCHPbIINHCH

Solution-based:

➢ Large Area ？

One-step,MAPbI3

One-step,MAPbCl3-xI3

Two-step,PbI2

Two-step,MAPbI3

Vacuum-based

Solution-based

Fig1.6 Morphology comparison

Vacuum-based:

➢ large area fabrication，easy control of filmthickness, high purity, smooth surface。

➢ expensive equipment ， MA erosion,Interface Control

➢ 1. Snaith: first report, PCE=15.4%, excellentmorphology (Nature , 2013, 501, 395-398.)

➢ 2. Bolink: OSCs, PCE=12% (Nat. Photonics 2014, 8,128-132.); flexible,7.7% (Energy Environ. Sci., 2014, 7,2968-2973.)

➢ 3. Yabing Qi: large area, PCE=9.9% (Energy Environ.Sci. 2014, 7, 3989-3993.)

➢ 4. Burn: band engineering, highest PCE=16.5% (Nat.Photonics 2015, 9, 106−112.)

Most focus on device structure, bandengineering, high efficiency, large areaand flexible PCEs fabrication

• Research Status on Vacuum-based

• Research Status on Vacuum-based

perovskite processing source material

HTL ETLVOC

(V)

JSC(mA/cm2

)

FF(%)

PCE(%)vacuum solution

coevaporation / PbI2+MAI PEDOT:PSS/PolyTPD PCBM 1.05 16.12 67 12.04




coevaporation PbI2+MAI PEDOT:PSS/PCDTBTPCBM/LiF

1.05 21.9 72 16.5

coevaporation / PbI2+MAI spiro-MeOTAD TiO2 1.1 18 70 13

coevaporation / PbCl2+MAI PEDOT:PSS PCBM 0.97 17.3 63 10.5

coevaporation / PbCl2+MAI spiro-MeOTAD C60 0.78 14.4 69 7.8

coevaporation / PbCl2+MAI NiO PCBM 0.79 14.2 65 7.26

coevaporation / PbCl2+MAI spiro-MeOTAD TiO2 1.07 21.5 68 15.4

hybriddeposition

/ PbCl2+MAI spiro-MeOTAD TiO2 1.09 17 54 9.9

hybriddeposition

/ PbCl2+MAI spiro-MeOTAD TiO2 1.01 12.82 66 8.64

/ one-step PbCl2+MAI spiro-MeOTAD TiO2 1.13 22.75 75 19.3

/one-step(moisture)

PbCl2+MAI PEDOT:PSSPCBM/PFN

1.05 20.3 80.2 17.1

/ one-step PbI2+MAI PEDOT:PSS PCBM 1.1 20.9 79 18.2

/one-step(hot-casting)

PbI2+MAI PEDOT:PSS PCBM 0.94 22.4 83 17.4

/ one-step PbI2+MAI PEDOT:PSS PCBM 0.89 18.85 80 13.37

/solventengineering

PbI2+MAI NiO PCBM 1.06 20.2 81.3 17.3

Table1.1 Device performance comparison of vacuum-based and solution-based PCEs

Vacuum-based: 7%~13% < Solution-based:15%~20%

• Why We Choice Vacuum-based Methods?

Why the PCE of vacuum-based PSCs is lower thanthat of solution-based PSC?

Lack of research on films growth mechanism

Out of favor（papers 30 vs 2000）

Part II

Growth Mechanism of Dual-source Evaporated Perovskite Films

• Motivation

1. The vacuum-based methods can stop film growth at anytime and therefore we can study phase evolution during filmgrowth.

2. In optoelectronics devices, the interface chemistry and theelectronic structure play a critically important role in chargeseparation, collection and recombination.

3. The study on growth mechanism helps us to reveal theorigin of the poor performance of PSCs by the vacuumevaporation method

• Perovskite Films Deposition

optimal deposition condition

evaporation characteristic of MAI

Fig2.1 Schematic diagram of dual-source evaporation system Fig2.2 Rate of MAI and PbI2

Fig2.3 XRD patterns of perovskite films deposited under different MAI partial pressure

vacuum system

sample preparation

Fig2.4 Images of the perovskite films

unstable

vapor source

MAI partial pressure:

1.2x10-5Torr

rate of PbI2: 0.4 Å/s

• Evolution of Films Structure

Fig2.5 XRD patterns of perovskite films with different thickness deposited on ITO substrates

Fig2.6 Evolution of films structure with films thickness

➢ At the initial growth stage, PbI2 phase is rich.

➢ The interfacial PbI2 is gradually converted tothe MAPbI3 phase during growth

• Evolution of Films Morphology

➢ films thickness <20nm, 2D layer

➢ films thickness >20nm, 3D island

➢ Stranski–Krastanow-like evolution mode

➢ The morphology evolves along with films component

Fig2.7 Surface SEM images of perovskite films

(a) 5nm (b) 10nm (c) 20nm

(d) 50nm (e) 150nm

• Evolution of Films Chemstry

➢ XPS results agree well with XRD results

Fig2.8 High resolution XPS core level spectra of (a) Pb 4f and (b) I 3d5/2 (c) N 1s of the perovskite films with different thicknesses on

deposited ITO substrates. (d) The atomic ratio of I/Pb of the perovskite films with different thicknesses deposited on ITO substratescalculated by XRS and EDS.

(c)

(d)

• Why PbI2 Phase is Rich at Interface?

Fig2.9 X-ray diffraction patterns of PbI2 film,MAI-rich MAPbI3 films, PbI2-rich MAPbI3films and MAI films.

Fig2.10 Surface SEM images of the corresponding samples

SampleMAI partial

pressure (Torr)Rate of PbI2

(Å/s)thickness or time

PbI2 / 0.4 5nm

PbI2-rich 1.2×10−5 0.5 20nm

MAI-rich 1.3×10−5 0.4 20nm

MAI 1.2×10−5 / 1hour

Table2.1 Sample deposition conditions

(a) MAI (b) MAI-rich

(c) PbI2-rich (d) PbI2

➢ MAI: 3D island, extremely low binding energy

➢ PbI2: 3D layer, perfect wetting property

➢ PbI2 phase at interface can not be eliminated byincrease MAI partial pressure

• Effect of PbI2 at Interface on Carrier Transportation

Fig2.11 Valence band spectra Fig2.12 Energy shift as a function of films thickness

Both the electrons and holes transportinto the electrode are blocked by theinterfacial PbI2 layer inevitably formedat the initial growth stage.

Fig2.13 Energy level diagram of the valence andconduction band edges of PbI2 and MAPbI3

Part III

Dual-source evaporated perovskite solar cells

• Motivation

1. To prove the results above that the PbI2,inevitably exiting at interface during growth,is harmful to device performance, because ofits blocking effect.

2. The disadvantageous effect caused by PbI2 atinterface can be eliminated by interfacemodification through two simple ways, resultingin a dramatic improvement of PCE.

• Device Performance

Fig3.1 Dark J-V curves of Device A0 Fig3.2 Light J-V curves of Device A0

Fig3.3 XRD patterns of Device A0 on FTO/TiO2/PCBM Fig3.4 Cross-sectional SEM images of Device A0

Device A0: optimal deposition condition (MAI partial pressure:1.2x10-5Torr, rate of PbI2:0.4 Å/s )

• Effect of PbI2 at Interface on Device Performance

Fig3.7 XRD patterns of Device A0~A20

Fig3.8 Dark J-V curves of Device A0~A20

Fig3.6 Cross-sectional SEM images of Device A20

DeviceJsc(mA/c

m2)Voc(V) FF(%) PCE(%)

A0 14.56 0.991 62.3 8.99

A5 11.45 0.989 47.2 5.34

A10 7.24 0.953 45.7 3.15

A20 2.78 0.909 36.1 0.91

Fig3.9 Light J-V curves of Device A0~A20

Table3.1 Performance of Device A0~A20

➢ PbI2 phase at interface is harmful to device performance

Fig3.5 Device structure of Device A0~A20

• Increase MAI Partial Pressure to Eliminate PbI2

Fig3.10 XRD patterns of Device B

Device B: Increase MAI partial pressure from 1.2 x 10 -5 Torr to 1.3 x 10 –5 Torr.

Fig3.11 Surface SEM images of Device B

Fig3.12 Light J-V curves of Device B

Increasing MAI partial pressure caneliminate PbI2 finally, but not improvedevice performance due to the poormorphology caused by excess of MAI.

• Interface Modification to Eliminate PbI2

Fig3.13 Scheme diagram of two interface modification methods Fig3.14 XRD patterns evolution of Device C

Fig3.15 XRD patterns evolution of Device D Fig3.16 SEM images evolution of Device D

(a) PbI2 (b) exposed for 20min

(c) exposed for 40min (d) exposed for 60min

The PbI2 phase at interface is eliminated successfully through these two interface modification methods.

• Enhancement of Device Performance

Device Jsc(mA/cm2) Voc(V) FF(%) PCE(%)

A0 14.56 0.991 62.3 8.99

C 20.85 0.977 57.4 11.69

D 17.52 1.024 73.7 13.22

Thickness(nm) Jsc(mA/cm2) Voc(V) FF(%) PCE(%)

200 13.89 0.994 57.9 7.99

260 17.52 1.024 73.7 13.22

350 19.83 1.027 70.5 14.35

450 18.35 1.028 65.3 12.31

600 16.59 1.031 57.3 9.80

Fig3.17 Light J-V curves of Device C&D

Fig3.18 Cross-sectional SEM images of Device D

Fig3.19 Light J-V curves of Device D as a function of films thickness

Table3.2 Performance of Device C&D

Table3.3 Performance of Device D as a function of films thickness

I. Growth mechanism of dual-source evaporated perovskite films1. Stranski–Krastanow-like evolution mode.2. PbI2 phase at interface, carrier blocking, harmful to device performance.

II. Dual-source evaporated perovskite solar cells1. Elimination of PbI2 at interface is necessary for high PCE.2. Interface modification, elimination of PbI2, enhancement of performance.

• Conclusions

Thanks for your attention！

Run XuSchool of Materials Science and Engineering

Shanghai UniversityMay 14， 2017

• PCEs Device Fabrication Process

• PSC related Publications (corresponding Author)

1. Haitao Xu, Yanglin Wu, Jian Cui, Jiang Cai, Chaowei Ni, Fuzong Xu, Feng Hong, Zebo Fang, Wenzhen Wang, Jiabin Zhu, Linjun Wang,

Run Xu and Fei Xu. Interface Modification as a Route to Improving Performance of Vacuum-Based Perovskite Solar Cells[J]. 2017, (Under

Review)

2. Haitao Xu, Yanglin Wu, Jian Cui, Chaowei Ni, Fuzong Xu, Jiang Cai, Feng Hong, Zebo Fang, Wenzhen Wang, Jiabin Zhu, Linjun Wang,

Run Xu and Fei Xu. Formation and Evolution of the Unexpected PbI2 Phase at Interface during the Growth of Evaporated Perovskite Films

[J]. Physical Chemistry Chemical Physics, Vol.18, 2016, pp. 18607~18613

3. Haitao Xu, Yanglin Wu, Fuzong Xu,Jiabin Zhu, Chaowei Ni, Wenzhen Wang, Feng Hong, Run Xu, Fei Xu, Jian Huang and Linjun Wang.

Grain Growth Study of Perovskite Thin Films Prepared by Flash Evaporation and its Effect on the Solar Cell Performance [J]. RSC

Advances, Vol.6, 2016, pp. 48851~48857

4. Runan Cao, Fei Xu, Jiabin Zhu, Sheng Ge, Wenzhen Wang, Haitao Xu, Run Xu, Yanglin Wu, Zhongquan Ma, Feng Hong and Zuimin

Jiang. Unveiling the Low-temperature Pseudodegradation of Photovoltaic Performance in Planar Perovskite Solar Cell by Optoelectronic

Observation [J]. Advanced Energy Materials, Vol. 6, 2016, pp. 1600814.

5. Fei Xu, Jiabin Zhu, Runan Cao, Sheng Ge, Wenzhen Wang, Haitao Xu, Run Xu, Yanglin Wu, Zhongquan Ma, Feng Hong and Zuimin

Jiang. Elucidating the Evolution of the Current-voltage Characteristics of Planar Organometal Halide Perovskite Solar Cells to an S-shape at

Low Temperature [J]. Solar Energy Materials and Solar Cells, Vol. 157, 2016, pp. 981~988.

6. Wenzhen Wang, Haitao Xu, Jiang Cai, Jiabin Zhu, Chaowei Ni, Feng Hong, Zebo Fang, Fuzong Xu, Siwei Cui, Run Xu, Linjun Wang, Fei

Xu and Jian Huang. Visible Blind Ultraviolet Photodetector Based on CH3NH3PbCl3 Thin Film [J]. Optics Express, Vol. 24, 2016, pp.

8411~8419.

7. Sheng Ge, Haitao Xu,Wenzhen Wang, Runan Cao, Yanglin Wu, Wenqiang Xu, Jiabin Zhu, Fei Xue, Feng Hong, Run Xu, Fei Xu, Linjun

Wang and Jian Huang. The Improvement of Open Circuit Voltage by the Sputtered TiO2 Layer for Efficient Perovskit Solar e Cell [J].

Vacuum, Vol. 128, 2016, pp. 91~98

8. Wenzhen Wang, Haitao Xu, Wenqiang Xu, Yanglin Wu, Runan Cao, Jiabin Zhu, Zebo Fang, Feng Hong, Run Xu, Fei Xu, Linjun Wang,

Jian Huang and Yicheng Lu. The Control of Surface Texture for Planar CH3NH3PbI3-xClx Film and Its Effect on Photovoltaic Performance

[J]. Journal of Materials Science: Materials in Electronics, Vol. 27, 2016, pp. 9384~9390.

9. Ye Yuan, Run Xu, Haitao Xu, Feng Hong, Fei Xu and Linjun Wang. Nature of the Band Gap of Halide Perovskites ABX3(A=CH3NH3,

Cs; B=Sn, Pb; C=Cl, Br, I): First-Principles Calculations [J]. Chinese Physics B, Vol. 24, 2015, pp. 116302.

• Device Structure

Fig1.4 (a) mesoporous (b) planar

➢ TCO: FTO，ITO

➢ ETL: TiO2, ZnO，PCBM

➢ Mesoporous layer: TiO2, Al2O3

➢ MAPbI3 light harvester layer

➢ HTL: spiro-OMeTAD、PTAA、P3HT

➢ (a) originates from ssDSSCs

complex structure, high energy costing, high temperature process,

narrow choices of substrates

➢ (b) originates from OCs

simple structure, low energy costing, low temperature process,

wide choices of substrates

Part IV

Perovskite films and solar cells by single-source flash evaporation

• Introduction

1. In dual-source vacuum evaporation, there aremany disadvantages, such as difficult control ofthe evaporation rates of MAI, high consumptionof the MAI source, and potential erosion of thevacuum components.

2. The growth mechanism of perovskite filmsprepared by flash evaporation are still lacking.

• 薄膜沉积工艺

Fig4.1 Process flow for the deposition of perovskite thin films by flash evaporation

• Films Characterization

➢ At a high pressure(10 Pa), as-deposited films with pure perovskitephase can be easily achieved using a precursor with an ideal MAI toPbI2 molar ratio of 1.0.

➢ At a low pressure(5 x 10-3 Pa), to achieve pure MAPbI3 films withoutPbI2 phase, the MAI concentration in the precursor must be increasedto compensate for the loss of MAI.

Fig4.2 XRD patterns of perovskite films deposited at different pressure and molar ratios

Fig4.3 UV-Vis transmittance of perovskite films deposited at different pressure and molar ratios

• Films Characterization

Fig4.4 Surface SEM images of perovskite films deposited at different pressure and molar ratios

Fig4.5 Grain size as a function of MAI concentration

➢ At a high pressure(10 Pa), voids can be clearly observed.

➢ At a low pressure(5 x 10-3 Pa), surfaces covered with dense grains andwithout any voids or pinholes can be observed.

➢ The size of the perovskite grains in the films increases linearly from0.2µm to around 1µm with increasing MAI concentration.

➢ The presence of PbI2 phase in the films can suppress the grain growth.

• Films Growth Mechanism

Fig4.6 The schematic of the evaporation process at a (a) high and (b) low pressure

Collision-determined transport process

➢ At a low pressure(10 Pa), distributions of MAI and PbI2 particles are in limitedspaces due to loss of kinetic energy during collision, insufficient diffusion, rod-like grains with pinholes and voids.

➢ At a low pressure(5 x 10-3 Pa), negligible collision, the limited-space distributionof PbI2 and the full-space distribution nature of MAI, loss of MAI, less energyloss, dense films.

• Device Performance

Fig4.7 Light J-V curves of PSCs fabricated at different pressure and molar ratios

evaporation

pressureMAI:PbI2 JSC(mA/cm2) VOC(V) FF(%) PCE(%)

5×10-3 Pa

1.5:1 5.56±1.32 0.81±0.05 0.40±0.07 1.82±0.61

1.7:1 12.13±1.14 0.96±0.03 0.59±0.04 7.41±0.74

1.9:1 13.62±1.33 0.99±0.02 0.65±0.04 8.13±1.05

2.0:1 15.81±1.25 0.99±0.04 0.64±0.03 10.01±1.41

2.1:1 12.84±1.21 0.98±0.05 0.58±0.03 7.54±0.89

10 Pa 1.0:1 9.19±1.81 0.98±0.02 0.52±0.03 4.90±1.18

Fig4.8 PCE as a function of grain size

Table4.1 Device performance of PSCs fabricated at different pressure and molar ratios

➢ Low growth pressure is critical to the device performance.

➢ PCE of PSCs increases with increasing grain size of the perovskite films.

interface modification as a route to improving performance of...

Documents