scalable fabrication of high efficiency hybrid perovskite

Scalable Fabrication of High Efficiency Hybrid Perovskite Solar Cells by Electrospray

Printing

Yuanyuan Jiang

Dissertation submitted to the faculty of the Virginia Polytechnic Institute and State

University in partial fulfillment of the requirements for the degree of

Doctor of Philosophy

In

Mechanical Engineering

Shashank Priya Co-Chair

Weiwei Deng Co-Chair

Jiangtao Cheng Member

Cheng Chen Member

Zheng Li Member

May 15th, 2019

Blacksburg, VA

Keywords: Perovskite Solar Cell, Electrospray Printing, Scalable Process

Copyright (2019)


Printing

Yuanyuan Jiang

ABSTRACT

Perovskite solar cells have attracted much attention both in research and industrial

domains. An unprecedented progress in development of hybrid perovskite solar cells

(HPSCs) has been seen in past few years. The power conversion efficiencies of HPSCs has

been improved from 3.8% to 24.2% in less than a decade, rivaling that of silicon solar cells

which currently dominate the solar cell market. Hybrid perovskite materials have

exceptional opto-electrical properties and can be processed using cost-effective solution-

based methods. In contrast, fabrication of silicon solar cells requires high-vacuum, high-

temperature, and energy intensive processes. The combination of excellent opto-electrical

properties and cost-effective manufacturing makes hybrid perovskite a winning candidate

for solar cells.

As power conversion efficiencies of HPSCs improves beyond that of the

established solar cell technology and their long-term stability increases, one of the crucial

hurdles in the path to commercialization remaining to be adequately addressed is the cost-

effective scalable fabrication. Spin-coating is the prevailing method for fabrication of

HPSCs in laboratories. However, this technique is limited to small areas and results in

excessive material waste. Two types of scalable manufacturing methods have been

successfully demonstrated to fabricate HPSCs: (i) meniscus-assisted coating such as

doctor-blade coating and slot-die coating; and (ii) dispersed deposition based on the

coalescence of individual droplets, such as inkjet printing and spray coating. Electrospray

printing belongs to the second category with advantages of high material utilization rate

and patterning capability along with the scalability and roll-to-roll compatibility.

In Chapter 3 of this dissertation, electrospray printing process is described for

manufacturing of HPSCs in ambient conditions below 150 C. All three functional layers

were printed using electrospray printing including perovskite layer, electron transport layer,

and hole transport layer. Strategies for successful electrospray printing of HPSCs include

formulation of the precursor inks with solvents of low vapor pressures, judicial choice of

droplet flight time, and tailoring the wetting property of the substrate to suppress coffee

ring effects. Implementation of these strategies leads to pin-hole free, low surface

roughness, and uniform perovskite layer, hole transport layer and electron transport layer.

The power conversion efficiency of the all electrospray printed device reached up to 15.0%,

which is among the highest to date for fully printed HPSCs.

The most efficient HPSCs rely on gold and organic hole-transport materials (HTMs)

for achieving high performance. Gold is also chosen for its high stability. Unfortunately,

the high price of gold and high-vacuum along with high-temperature processing

requirements for gold film is not suitable for the large-scale fabrication of HPSCs. Carbon

is a cheap alternative electrode material which is inert to hybrid perovskite layer. Due to

the ambipolar transport property of hybrid perovskite, perovskite itself can act as a hole

conductor, and the extra hole transport layer can be left out. Carbon films prepared by

doctor-blade coating method have been reported as the top electrode in HPSCs. The

efficiencies of these devices suffer from the poor interface between the doctor-blade coated

carbon and the underlying perovskite layer. In Chapter 4, electrospray printing was applied

for the fabrication of carbon films and by optimizing the working distance during

electrospray printing, the interface between carbon and the underlying perovskite layer was

greatly improved compared to the doctor-blade coated carbon film. The resulting HPSCs

based on the electrospray printed carbon electrode achieved higher efficiency than that

based on doctor-blade method and remarkably, this performance is close to that of gold

based devices.

In Chapter 5, preliminary results are provided on the laser annealing of hybrid

perovskite films to further advance their scalable manufacturing. All layers of HPSCs

require thermal annealing at temperature over 150 C for about half an hour or longer. The

time-consuming conventional thermal annealing complicates the fabrication process and is

not suitable for continuous production. High temperature over150 C is also not compatible

with flexible substrates such as PET. Laser annealing is a promising method for

overcoming these issues. It has several other advantages including compatibility with

continuous roll-to-roll printing, minimal influence on non-radiated surrounding area, and

rapid processing. Laser annealing can be integrated with the electrospray process to realize

the continuous fabrication of hybrid perovskite film. Rapid laser annealing process with

optimized power density and scanning pattern is demonstrated here for annealing

perovskite films. The resulting hybrid perovskite film is highly-crystalline and pin-hole

free, similar to that obtained from conventional thermal annealing.


Printing

Yuanyuan Jiang

GENERAL AUDIENCE ABSTRACT

Hybrid perovskite solar cell (HPSC) is a promising low-cost and high efficiency

photovoltaic technology. One of the big challenges for it to be commercially competitive

is scalable fabrication method. This dissertation focuses on developing electrospray

printing technology for HPSCs. This is a scalable method with high material usage rate

that naturally lead to large scale fabrication of HPSCs. Electrospray printing parameter

space was systematically studied and optimized to synthesize high-quality perovskite films

and other functional layers including hole transport layer and electron transport layer. All

electrospray printed high-efficiency perovskite solar cell devices were successfully

demonstrated under the ambient condition and low temperature. Another achievement of

this thesis is the electrospray printing of carbon film to replace the costly gold electrode in

perovskite solar cells. Laser annealing technique is demonstrated for HPSCs, which is

compatible with continuous fabrication and integrates easily with electrospray printing.

VI

Dedication

To my lovely family members.

I love you all no matter what.

VII

Acknowledgements

First of all, I would like to express my deepest gratitude to my advisor Dr. Shashank

Priya. Without his kind support and warm encouragement, I don’t think I will be able to

finish my Ph.D. study. His profound knowledge, sharp mind, and broad vision inspired me

deeply.

I want to express my sincere gratitude to my previous advisor Dr. Weiwei Deng.

His continued mentorship and guidance made this dissertation possible. His passion for

research and creative mind inspired me profoundly. He also has a kind heart. I feel blessed

to be a student of Dr. Deng.

Also, I want to thank my committee members: Dr. Jiangtao Cheng, Dr. Cheng Chen,

and Dr. Zheng Li for their guidance, support, and encouragement during the course of this

research.

I want to say thanks to Dr. Congcong Wu, a research assistant professor at our

group, who helped me so much during the past 4 years. I cannot say enough thanks to him.

Thanks to all my previous and current lab-mates and colleagues for their help and

comradery: Ting Yang, Liurui Li, Fan Gao, Dr. Kai Wang, Dr. Dong Yang, Yuchen Hou,

Dr. Xiaotian Li, Dr. Hairui Liu, Dr. Yongke Yan, Dr. Prashant Kumar, Han-byul Kang, Dr.

Min Gyu Kang, and the list goes on.

Thanks to all the lab-mates and colleagues at SUSTech for their warm help during

my brief stay at SUSTech.

VIII

Thanks to Dr. Lee Williams from NanoSonic, Inc. for his help during the course of

this research.

Thanks to all friends, colleagues, the department faculty and staff at VT and Penn

State for making this a wonderful experience.

Special thanks to my roommate, Jiamin.

Finally, thanks to the funding provided by:

National Science Foundation, US (CMMI 1549917: Scalable Electrospray

Processing of High-Efficiency Perovskite Solar Cells)

US Army (0402436 UP 47FZ0)

IX

Table of Contents

List of Figures .................................................................................................................. XII

List of Tables ................................................................................................................ XVII

Chapter 1 Introduction ...................................................................................................... 1

1.1 Solar energy and solar cells ...................................................................................1

1.2 Perovskite Materials ...............................................................................................3

1.3 Material properties .................................................................................................6

1.3.1 Crystal phase of ABX3 perovskite ................................................................ 6

1.3.2 Photoelectric properties ................................................................................ 7

1.3.3 Morphological properties .............................................................................. 8

1.4 Device physics .......................................................................................................9

1.4.1 Device structure ............................................................................................ 9

1.4.2 Operating principle of solar cell ................................................................. 11

1.4.3 Device characterization ............................................................................... 13

1.5 Scalable fabrication of PSCs ................................................................................17

1.5.1 Meniscus assisted coating ........................................................................... 19

1.5.2 Droplets based deposition methods ............................................................ 21

1.5.3 Other methods ............................................................................................. 27

1.6 Summary ..............................................................................................................28

Chapter 2 Experimental ................................................................................................... 29

X

2.1 Solution preparation .............................................................................................29

2.2 Device preparation ...............................................................................................31

2.2.1 Deposition of films by electrospray printing .............................................. 31

2.2.2 Deposition of films by spin-coating ............................................................ 34

2.3 Characterization ...................................................................................................34

Chapter 3 All Electrospray Printed Perovskite Solar Cells ............................................. 38

3.1 Introduction ..........................................................................................................38

3.2 Results and discussion .........................................................................................40

3.2.1 The stable operation of the electrospray in the cone-jet mode ................... 41

3.2.2 Printing of uniform wet perovskite precursor thin film .............................. 43

3.2.3 Fast drying induced high-quality perovskite film ....................................... 51

3.2.4 Electrospray printing of electron transport layer and hole transport layer . 56

3.2.5 The performance of all electrospray printed PSCs ..................................... 60

3.3 Conclusion ...........................................................................................................64

3.4 Experimental section ............................................................................................65

3.5 Supporting Information ........................................................................................67

Chapter 4 Hole-conductor-free perovskite solar cells based on electrospray printed

carbon electrode ................................................................................................................ 71

4.1 Introduction ..........................................................................................................71


XI

4.2.1 The Stable operation of the electrospray of carbon .................................... 75

4.2.2 Characterization of electrospray printed carbon film ................................. 77

4.2.3 Working distance and interface .................................................................. 79

4.2.4 Hole-conductor-free PSCs based on carbon electrode ................................ 81

4.3 Conclusion ...........................................................................................................85

Chapter 5 Laser annealing for the scalable fabrication of perovskite films .................... 87

5.1 Motivation ............................................................................................................87


5.3 Conclusion ...........................................................................................................94

Chapter 6 Summary and outlook ..................................................................................... 95

References ........................................................................................................................ 98

XII

List of Figures

Figure 1-1 Generations of solar cells and corresponding highest research cell efficiencies

[4,5]. .................................................................................................................................... 2

Figure 1-2 Schematic description of the crystal structure of ABX3 perovskite [9]. ........... 4

Figure 1-3 NREL solar efficiency chart [5]. ....................................................................... 6

Figure 1-4 Various architectures of PSCs. ........................................................................ 10

Figure 1-5 General solar cell device configuration. .......................................................... 11

Figure 1-6 Band structures of each layer in a solar cell device and the sub-steps of the

photovoltaic process.......................................................................................................... 12

Figure 1-7 Equivalent circuit of the single-junction solar cell. ........................................ 14

Figure 1-8 An example of J-V curve of solar cells under illumination [43]..................... 16

Figure 1-9 Scalable solution deposition methods for the roll-to-roll fabrication of PSCs,

including blade coating (panel a), slot-die coating (panel b), spray coating (panel c), and

inkjet printing (panel d) [47]. ............................................................................................ 18

Figure 1-10 (a) Schematic illustration of the pneumatic spray process [65], (b) Photographs

of ultrasonic spray-coating [66], (c) typical electrospray and details of the jet break-up. 22

Figure 1-11 1. Different types of cone-jet mode: a) pulsed cone-jet b) and c) multi-jet. 2.

Variations in meniscus forms for cone-jet mode (a – e). 3. Sequence of stages during

dripping mode. 4. Sequence of stages during micro-dripping mode. Taken from reference

[68]. ................................................................................................................................... 25

Figure 1-12 Various kinds of nozzle arrays for the scalable electrospray deposition [72].

........................................................................................................................................... 26

Figure 2-1 The process to prepare carbon ink for electrospray. ....................................... 30

XIII

Figure 2-2 Picture of the actual electrospray system. ....................................................... 32

Figure 2-3 Schematic diagram of the Conductive AFM system, taken from [79]. .......... 36

Figure 3-1 Diagram of an electrospray system. ................................................................ 40

Figure 3-2 Silica nozzle with 30 µm tip inserted into the conductive silicone gasket. .... 42

Figure 3-3 Taylor cone without (a), (b) and with(c) anti-wetting coating on the outer surface

of the nozzle. ..................................................................................................................... 43

Figure 3-4 SEM top-view images of perovskite films printed at different conditions. The

scale bar is 1 micron. ........................................................................................................ 46

Figure 3-5 Diagram of the solvent evaporation during the electrospray process. ............ 47

Figure 3-6 Schematic of serpentine path motion of electrospray nozzle relative to the

substrate [90]. .................................................................................................................... 48

Figure 3-7 Optical microscope images of perovskite film printed using one, two and three

paths. ................................................................................................................................. 49

Figure 3-8 (a) Schematic illustration for the directional microscale solution flow towards

perovskite film printed earlier. (b) Optical image of perovskite films obtained from nearly

0 contact angle and (c) 18 contact angle. ....................................................................... 50

Figure 3-9 Optical Microscope images (a, b) and AFM topographies (c, d) of Electrospray

printed perovskite films dried using anti-solvent method (a, c) and vacuum-assisted flash

drying (b, d). ..................................................................................................................... 52

Figure 3-10 Top-view (left column), corresponding cross-section view (middle column) of

SEM images and AFM surface scanning images (right column) of perovskite films

electrospray printed at flow rate of 0.43, 0.57, 0.72 µL/min. ........................................... 53

XIV

Figure 3-11 XRD of the perovskite films using electrospray printed and spin-coated

perovskite. ......................................................................................................................... 54

Figure 3-12 PCEs of the PSCs using electrospray printed and spin-coated perovskite. ... 55

Figure 3-13 Top-view and corresponding cross-sectional SEM images of TiO2 films

electrospray printed at working distance of 4 (a), 3 (b), 2 (c) mm and spin-coated (d). The

scale bar is 600 nm. ........................................................................................................... 57

Figure 3-14 Conductivity comparison between the spin-coated TiO2 film and electrospray

printed at 2 mm. ................................................................................................................ 58

Figure 3-15 PCEs of the PSCs using electrospray printed and spin-coated perovskite films.

........................................................................................................................................... 58

Figure 3-16 PCE of PSCs using four different fabrication conditions for Spiro-OMeTAD:

electrospray printing with CB or DCB as solvent, and spin-coating with CB or DCB as

solvent. The inset shows the corresponding top-view SEM images. The scale bar is 400

nm. .................................................................................................................................... 60

Figure 3-17 (a) Cross-section view SEM image of an all electrospray printed photovoltaic

device. (b) J-V curves for the champion cell of the all-printed, all spin-coated devices and

the device with only perovskite layer electrospray printed. Inset: The PCE distribution

histogram of the all-printed devices. ................................................................................. 62

Figure 3-18 Top-view SEM images of spin-coated perovskite films. .............................. 67

Figure 3-19 Typical JV curves of the device using perovskite ES printed at different flow

rates and spin coated. The HTL and ETL are spin coated for all devices. ....................... 67

XV

Figure 3-20 Typical JV curves of the device using TiO2 ES printed at working distances of

4, 3, 2 mm and TiO2 spin coated. The perovskite film is ES printed and HTL is spin coated

for all devices. ................................................................................................................... 68

Figure 3-21 Typical JV curves of the device using Spiro-MeOTAD ES printed with

chlorobenzene and 1,2-dichlorobenzene as the solvent and Spiro-MeOTAD spin-coated

with chlorobenzene and 1,2-dichlorobenzene. The perovskite film and TiO2 are both ES

printed for all devices. ....................................................................................................... 68

Figure 3-22 The stability of the all electropsray printed device without encapsulation in

ambient conditions with a relative humidity of ~20% at room temperature. ................... 69

Figure 4-1 Details of the three types of C-PSCs. a) Energy level and charge transfer

behavior in C-PSCs. b) Meso C-PSCs developed by Ku et al. [110]. c) Assembled C-PSCs

developed by Wei et al. [115]. d) Paintable C-PSCs developed by Wei et al. and other

groups [115–117]. ............................................................................................................. 72

Figure 4-2 Electrospray of carbon ink without (left) and with the additive (right). ......... 76

Figure 4-3 XRD patterns of perovskite film, perovskite/carbon-ES and perovskite/Carbon-

doctor blade. ...................................................................................................................... 78

Figure 4-4 Top-view SEM image of the doctor blade coated (left) and electrospray printed

(right) carbon film. ............................................................................................................ 78

Figure 4-5 Cross-section SEM images of carbon films printed at different working distance

and the carbon film that was doctor blade coated. ............................................................ 81

Figure 4-6 Diagram of the architecture (a) and energy diagram (b) of carbon electrode

based hole-conductor-free PSCs. ...................................................................................... 82

XVI

Figure 4-7 JV curves of hole-conductor-free PSCs based on different carbon electrodes.

........................................................................................................................................... 83

Figure 4-8 The steady state photoluminescence of perovskite and carbon film structure. 84

Figure 4-9 Time-resolved photoluminescence lifetimes for perovskite/carbon films on

glass slides. ....................................................................................................................... 85

Figure 5-1 Diagram for the concept of roll-to-roll fabrication of PSCs with integration of

electrospray printing and laser annealing. ........................................................................ 88

Figure 5-2 Scanning pattern of the laser. .......................................................................... 89

Figure 5-3 Optical microscope image of perovskite film annealed using laser at focal plane,

with scan speed of 2000 mm/s and power percentage of 0.05%. ..................................... 90

Figure 5-4 Optical images of perovskite films laser annealed at different power density.

........................................................................................................................................... 91

Figure 5-5 X-ray diffraction (XRD) pattern of laser-crystallized perovskite films at

different laser power densities and perovskite film that was thermal annealed. .............. 92

Figure 5-6 SEM images of perovskite films annealed by thermal heating (left) and laser

(right). ............................................................................................................................... 93

Figure 5-7 AFM images of perovskite films annealed by thermal heating (left) and laser

(right). ............................................................................................................................... 93

XVII

List of Tables

Table 1-1 The calculated and experimental lattice constants, and band gap of different

CH3NH3PbX3 perovskites; taken from [21]. ....................................................................... 8

Table 2-1 Typical ES printing parameters for TiO2, perovskite, Sprio-MeOTAD and carbon

ink. .................................................................................................................................... 33

Table 2-2 Typical spin-coating parameters for TiO2, perovskite, and Sprio-MeOTAD. . 34

Table 3-1 Photovoltaic metrics of the all electrospray printed, all spin-coated devices and

the device with only the perovskite layer electrospray printed. ........................................ 61

Table 3-2 Overview of PCE of the reported PSCs with all three layers (ETL, perovskite,

and HTL) fabricated using scalable methods. ................................................................... 63

Table 4-1 Representative C-PSCs from literature. ........................................................... 75

Table 4-2 Typical ES printing parameters of carbon ink. ................................................. 77

Table 4-3 Sheet resistance of carbon films prepared by doctor blade coating or electrospray

printing at different working distance. .............................................................................. 79

Table 4-4 Photovoltaic metrics of the hole conductor free C-PSCs based on electrospray

printed carbon (C-ES-1.5) and based on doctor blade coated carbon (C-DC) and PSCs

based on gold electrode. .................................................................................................... 83

Table 5-1 Parameters of the modified RGL-FM fiber laser marker. ................................ 88

Table 5-2 Laser annealing parameters. ............................................................................. 91

1

Chapter 1 Introduction

1.1 Solar energy and solar cells

Climate change, created predominantly by the usage of fossil fuels, is becoming one of the

prominent challenges for our planet. As a result, we are observing global warming and extreme

weather events. In conjunction, the global energy demand is growing driven by the economic

growth and population. Intergovernmental Panel on Climate Change (IPCC) recently released a

report stating that in order to avoid worst impacts of climate change, we should resort to renewable

energy based electricity in the range of 70%-85% by 2050 [1]. This transition from fossil fuel to

renewables based economy is vital for human society. Solar energy represents an attractive source

of renewable energy. The magnitude of solar energy striking the earth’s surface in two hours is

more than all the energy consumed by the world per year [2]. Excluding the part compromised due

to geographical inaccessibility, which sets technical limitations on the amount of sunlight that can

be harvested with current technology, solar power is still far more than enough to meet the global

energy demand [3].

One of the direct applications of solar energy are solar cells that absorb and convert solar

energy into electricity via photovoltaic process. The working principle of solar cell devices is

described in section 1.4. Three generations of solar cells have been developed as shown in Figure

1-1 [4]. The 1st generation is based on silicon includes poly-crystalline silicon, mono-crystalline

silicon solar cells and the hybrid of crystalline and amorphous silicon solar cells. They are the

oldest and the most widely used technology because of their relatively high power conversion

efficiency (PCE), but the manufacturing process for the 1st generation solar cells is expensive and

2

energy extensive. Another drawback of the 1st generation is they are rigid because of the wafer

nature. The 2nd generation solar cell is composed of thin films which become flexible, including

microcrystalline, amorphous silicon, copper indium gallium selenide (CIGS) and cadmium

telluride (CdTe) solar cells. However, the high efficient types still require costly manufacturing

process. The 3rd generation solar cell refers to the technologies developed in recent years but has

not been commercialized yet. These cells have low manufacturing cost as they can be processed

via solution based approaches. The 3rd generation solar cell features the nanocrystal-based solar

cells, polymer-based solar cells, dye-sensitized solar cells, and perovskite solar cells (PSCs).

Figure 1-1 Generations of solar cells and corresponding highest research cell efficiencies

[4,5].

3

Currently, the solar energy market is still dominated by the silicon-based solar cells

because of their high PCE and long-term stability. Solar cells based on the amorphous Si, CdTe,

and CIGS photoactive layers have relatively lower material stability as well as they utilize rare

elements. The PCE of organic solar cells and dye-sensitized solar cells is relatively low, but HPSCs

provide ultra-high-performance matching that of silicon and other semiconductor-based cells. The

performance of HPSCs has attracted significant attention and an unprecedented development has

been observed. The power conversion efficiencies of HPSCs has increased from 3.8% [6] to 24.2%

[5] in less than a decade, rivaling those of silicon solar cell. Hybrid perovskite has excellent photo-

physical properties and it can be processed using cost-effective solution-based methods. In contrast,

fabrication of silicon solar cell requires high-vacuum, high-temperature, and energy intensive

process. This combination makes perovskite a promising candidate for future solar cells [7].

1.2 Perovskite Materials

The term “perovskite” refers to calcium titanate compound, which was discovered in 1839 by

the German mineralogist Gustav Rose and was named by the Russian mineralogist Lev Perovski.

Later on, the term “perovskite” has been used to describe all the compounds with the same crystal

structure as calcium titanate, generally represented as XIIA2+VIB4+X2−3 [8].

The ideal cubic structure has B-cation surrounded by an octahedron of X anions, and the

A-cation is in 12-fold cuboctahedral coordination as shown in Figure 1-2 [9]. Some distortions

may exist in the ideal cubic form of perovskite resulting in orthorhombic, rhombohedral,

hexagonal, and tetragonal forms. By suitably choosing A, B, and X ions, the crystal and electronic

4

structure can be tuned to design materials with different functionality that have application-

oriented properties.

Figure 1-2 Schematic description of the crystal structure of ABX3 perovskite [9].

Hybrid perovskite materials applied in photovoltaics are generally organic-inorganic lead

halide hybrid perovskites, where A is a monovalent cation (such as methylammonium (MA+),

formamidinium (FA+) or caesium (Cs+)), B is a divalent metallic cation (such as Pb2+ or Sn2+) and

X is a halide (I−, Br− or Cl−). These materials have suitable optical-electrical characteristics such

as broadband light absorption, high ambipolar mobility, long charge-carrier diffusion lengths,

bandgap tunability and defect tolerance [9–11]. In addition, lead halide perovskites can be

synthesized from low-cost and earth-abundant raw materials using solution-based processes at

moderate temperatures (below 150 C).

5

One of the earliest application of perovskite in solar cells was reported by Miyasaka et al.

[6], who incorporated perovskite as the “dye” in the dye-sensitized solar cells (DSSCs) with a thin

layer of perovskite on mesoporous TiO2 as electron-collector. This cell demonstrated an PCE of

3.8%. In earlier DSSCs, because the electrolyte was a corrosive liquid, the cells were stable only

for a few minutes.

Two years later, Park et al. improved the electrolyte in the same dye-sensitized structure

and achieved a PCE of 6.5% [12]. A breakthrough was provided when Snaith et al. and Lee et al.

introduced the solid-state hole-transport material, Spiro-OMeTAD, which is much more

compatible with perovskite than the previous liquid electrolytes, and as a result the PCE was

greatly improved to 10% [9,13]. Further improvment in PCE of over 10% was obtained when TiO2

was replaced with an inert scaffold [14]. This led to the hypothesis that a scaffold is not needed

for electron extraction and the assumption of ambipolar transport property of the organic-inorganic

lead halide hybrid perovskite materials. Soon after that a large number of investigators conducted

studies on both the planar and sensitized PSCs. Burschka et al. [15] developed the two-step

solution method to fabrication perovskite film and the PSCs based on the sensitized architecture

that demonstrated a PCE over 15%. Malinkiewicz et al. and Liu et al. presented the thermal co-

evaporation process to fabricate perovskite film and the PSCs based on the planar structure

achieving more than 12% and 15% PCE [16,17]. Another perovskite architecture with the hole

transport layer (HTL) and electron transport layer (ETL) was demonstrated by Docampo et al. [18]

which is referred to as the “inverted” structure. Many new deposition methods kept emerging and

higher efficiencies were obtained. Yang et al. [19] reported an efficiency of 19.3% based on a

planar structure. KRICT made the record efficiency of 20.1% in 2014, which was renewed to

6

21.0% in 2015 by researchers at EPFL. In 2016, the record was further pushed to 22.1% by

researchers from KRICT and UNIST. Certified efficiencies of 23.3% and 24.2% were reported in

2018 by researchers from CAS and in 2019 by researchers from KRICT as shown in Figure 1-3

[5].

Figure 1-3 NREL solar efficiency chart [5].

1.3 Material properties

1.3.1 Crystal phase of ABX3 perovskite

The hybrid perovskite tends to have polymorphs with cubic (α phase), tetragonal (β phase),

and orthorhombic (γ phase) structure, depending upon the composition or more specifically the

size of A, B, and C. The value of Goldschmidt tolerance factor “t” can be used to determine the

crystal structure of a specific ABX3 perovskite. The Goldschmidt tolerance factor “t” is

represented in equation 1-1 [20].

7

𝑡 = (𝑅𝐴 + 𝑅𝑋)/√2(𝑅𝐵 + 𝑅𝑋) (1-1)

where RA, RX, and RB are the ionic radii of A cation, X anion, and B cation, respectively.

Empirically, 𝑡 was found to be between 0.8 and 1.0 for black-phase perovskites. Phase transition

from one crystal structure to another is not uncommon for hybrid perovskite when a specific

temperature is reached. For example, for the popular CH3NH3PbI3 halide perovskite, phase

transition from α to β to γ occurs at 330 K and 160 K, respectively [21]. Table 1-1 summarizes the

calculated lattice constants and band properties of different phases of the hybrid perovskite.

1.3.2 Photoelectric properties

The hybrid perovskite materials usually have high carrier mobility due to the unique

electronic band structure and have high light extinction coefficient due to its direct bandgap nature.

Samuel et al. found that the mixed halide perovskite has much longer diffusion length than the

pristine perovskite. For example, the diffusion length of triiodide (CH3NH3PbI3) can be on ~100

nm scale, while for the mixed halide (CH3NH3PbI3-xClx) perovskite the diffusion length can be

over 1 µm [22]. These investigations were conducted using the photoluminescence-quenching

measurements and transient absorption. CH3NH3PbI3 single crystals could achieve extremely low

trap density on the order of 109 ~ 1010 cm-3 as reported by Shi et al. [23], which is the same level

as that of the best photovoltaic-performance silicon material. The charge carrier diffusion lengths

of over 10 μm has been identified in these materials [23]. Such high mobility ensures that the light-

generated charges can be sufficiently driven to be collected as current, instead of losing their

energy as heat within the PSCs, which will greatly enhance the PCE of the device.

8

Table 1-1 The calculated and experimental lattice constants, and band gap of different

CH3NH3PbX3 perovskites; taken from [21].

lattice constant (Å ) Relative energy (meV) Band gap (eV)

Phase symmetry PBE Experimental PBE PBE+SOC HSE+SOC PBE PBE+SOC HSE+SOC Experimental

CH3NH3PbI3

α 𝑷𝒎�̅�𝒎 a=6.39 a=6.31,6.28 0 0 0 1.53 0.46 1.14

β 𝑰𝟒/𝒎𝒄𝒎 a=8.80 a=8.85,8.88

-93 -127 -100 1.57 0.77 1.60 1.52,1.5

c=12.99 c=12.64,12.67

γ 𝑷𝒃𝒏𝒎

a=8.84 a=8.84

-38 -55 -50 1.46 0.59 1.43 b=8.77 b=8.56

c=12.97 c=12.58

CH3NH3PbBr3

α 𝑷𝒎�̅�𝒎 a=6.04 a=5.94,5.95 0 0 0 1.93 0.89 1.92 2.23,2.32,2.29,2.35

β 𝑰𝟒/𝒎𝒄𝒎 a=8.28

-60 -59 -92 1.89 1.13 2.11 c=12.25

γ 𝑷𝒃𝒏𝒎

a=8.32

-21 -19 -44 1.81 0.91 1.86 b=8.29

c=12.15

CH3NH3PbCl3

α 𝑷𝒎�̅�𝒎 a=5.78 a=5.70 0 0 0 2.40 1.33 2.57 3.11

β 𝑰𝟒/𝒎𝒄𝒎 a=7.93

-52 -53 -65 2.47 1.57 2.77 c=11.71

γ 𝑷𝒃𝒏𝒎

a=7.94

-12 -16 -18 2.27 1.31 2.47 b=7.95

c=11.66

1.3.3 Morphological properties

To achieve high PCE for PSCs, one of the most important prerequisites is to prepare high-

quality perovskite films. The morphology of the hybrid perovskite is highly dependent on the

9

crystallization process, which is further dependent on various factors including chemical

composition, deposition technics, surfaces effect, and processing solvents and additives [24,25].

Typically, perovskite thin films with high surface coverage, smooth surface and larger crystal

domains are preferable for higher device performance.

1.4 Device physics

1.4.1 Device structure

The microstructure of perovskite layer will greatly influence its performance and also

dictate the selection of the materials for each component and subsequently the deposition method.

Two major state-of-the-art device structures have been developed so far, i.e., meso-superstructure

and “planar heterojunction” [16,26–29]. Depending on which layer comes into contact with

sunlight, it could further be classified as n-i-p and p-i-n. Figure 1-4 shows the diagram of the most

commonly reported device structures for PSCs.

The n-i-p mesoscopic structure typically consists of a mesoporous TiO2 layer or Al2O3

layer. In mesoscopic structure, perovskite infiltrates into the mesoporous structure to form an

intermixed junction, which will facilitate electron transfer and separation and help suppress the

photocurrent-voltage (J-V) hysteresis. But these mesoporous layers usually require high-

temperature annealing (~500 °C), which complicates the fabrication of PSCs and is not compatible

with flexible substrates like polyethylene terephthalate (PET).

10

Figure 1-4 Various architectures of PSCs.

The mesoporous structure has been found to be not necessary for electron extraction, rather

the planar structure as shown in Figure 1-4 is able to provide all the features [30,31]. It eliminated

the mesoporous layer and lead to more facile fabrication. The development of low-temperature (<

150 °C) deposition process of ETL like TiO2 and SnO2 [32,33] enables the fabrication of flexible

solar cells on PET substrates. In the planar p-i-n structure, the most commonly used ETL is Phenyl-

C61-butyric acid methyl ester (PC61BM)) and HTL is Poly(3,4ethylenedioxythiophene)

Polystyrene sulfonate (PEDOT:PSS)). Both could be processed using solution-based method and

enjoy the advantages of facile fabrication and flexible substrate compatibility. However, one of

the problems is the hygroscopic nature of PEDOT:PSS that would compromise the long-term

stability of PSCs due to its instability toward water [34,35]. To improve the device performance,

11

research efforts has been dedicated towards new directions such as interfacial engineering, device

engineering, and finding alternative HTLs.

1.4.2 Operating principle of solar cell

To form a functional perovskite solar cell, the perovskite active layer is sandwiched

between the HTL and ETL, and two electrodes are deposited to make contact with the two charge

transport layers on either side. One of the two electrodes has to be transparent to allow sunlight to

come in and absorbed by the pervoskite layer. The general device configuration is shown in Figure

1-5, representing a planar n-i-p device structure.

Figure 1-5 General solar cell device configuration.

The most important component in solar cells is the photoactive layer that absorbs the

sunlight and “converts” the photon energy to the potential and kinetic energy of electrons [36,37].

This process is influenced by many factors, such as the density of state (DOS) distribution, band

gap, and electrical potential energy. All these factors can affect the lifetime of the activated charge

carriers [38]. The activated electron can move freely in the material and leaving a hole (defined as

an absence of an electron in a particular place in the electron orbits) in the lower energy level. The

12

movement of free electrons and holes significantly depends on the energy level alignment of the

charge transport layer and the electrode.

Figure 1-6 Band structures of each layer in a solar cell device and the sub-steps of the

photovoltaic process.

Figure 1-6 shows the schematic of the band structures of each layer in a solar cell device. The

general photovoltaic process contains the following sub-steps [37,39]:

• The generation of light-generated carriers;

The hybrid perovskite active layer absorbs sunlight and generates charge carriers. Hybrid

perovskite is a semiconductor with a band gap. When the photon has energy higher than the band

gap, it will activate bound electrons, namely, the electron will go from the valence band to

conduction band. The electron leaves behind a hole in the valence band which is free to move.

13

• The collection of the light-generated carriers to generate current;

The collection is the process of light absorption generated charge carriers been swept out

and collected through an external circuit. In parallel to collection, recombination can also occur,

which causes energy loss and lowers the efficiency of solar cells. Activated electrons are in a meta-

stable state and will eventually stabilize to a lower energy position in the valence band and combine

with a hole in that band. Increasing the built-in potential can enhance the charge collection process.

The built-in potential could be modified by adding functional buffer layers, modifying the interface,

etc. Refining the materials composition and structural inhomogeneity, and reducing the trap-states

are also effective ways to improve the collection of charge carriers.

• The dissipation of power in the load and in parasitic resistances;

Several energy barriers exist before the generated charge carriers could reach the electrode

to contribute to the outer circuit. From the photoactive layer to charge transport layers (HTL and

ETL), and from the charge transport layers to the electrodes, barrier exists, and power dissipation

happens. Sometimes, in more complicated cases, multiple buffer layers are used, which may raise

the internal resistance and the recombination.

1.4.3 Device characterization

Solar cells can be simplified to the equivalent circuit according to the one-diode model[40]

when trying to understand its electrical behavior. Figure 1-7 presents the equivalent circuit of the

single-junction solar cell. IL is the light-induced current, ID is the dark current, and Rs and Rsh are

the parasitic series and shunt resistances. The ID is modeled using the Shockley equation for an

14

ideal diode and can be expressed by equation 1-2 [40,41], where n is the ideality factor, I0 is the

saturation current, VT is the thermal voltage given by equation 1-3.

Figure 1-7 Equivalent circuit of the single-junction solar cell.

𝐼𝐷 = 𝐼0[exp (𝑉+𝐼𝑅𝑠

𝑛𝑉𝑇) − 1] (1-2)

𝑉𝑇 =𝑘𝑇𝑐

𝑞 (1-3)

The current collected by outside circuit can be expressed by equation 1-4.

𝐼 = 𝐼𝐿 − 𝐼0 [exp (𝑉+𝐼𝑅𝑠

𝑛𝑉𝑇) − 1] −

𝑉+𝐼𝑅𝑠

𝑅𝑠ℎ (1-4)

The electrical properties can be modulated when applying different bias and light intensity.

Hereafter, we introduce some commonly used methods to characterize solar cell devices.

J-V curve under illumination

15

The photovoltaic performance of solar cells could be characterized by performing a voltage

sweep on the device and measuring the current to obtain a current density to voltage (J-V) curve.

The J-V curve is measured under illumination with AM1.5G solar spectrum (1000 W/m2) at a

temperature of 25 °C. AM1.5G represents air mass 1.5 global, an standard reference spectra

defined to allow the performance comparison of photovoltaic devices from different manufacturers

and research laboratories [42]. Figure 1-8 shows an example J-V curve under illumination. The

PCE is defined as the percentage of solar energy being converted into electricity and is calculated

using the equation 1-5:

PCE = 𝑉𝑜𝑐𝐽𝑠𝑐𝐹𝐹

𝑃𝑖𝑛=

𝑃𝑀𝑃𝑃

𝑃𝑖𝑛 (1-5)

where Voc is the open-circuit voltage, namely the potential when the current is zero and Jsc is the

short-circuit current referring to the maximum current that can run through the device. The FF is

affected by the resistances present in the cell indicating the ease of charge collection and the

amount of leakage current in the device. FF can be calculated as the ratio between maximum power

point (PMPP) and the theoretically maximum obtainable power density (Pmax):

FF =𝑃𝑀𝑃𝑃

𝑃𝑀𝑎𝑥=

𝑃𝑀𝑃𝑃

𝑉𝑜𝑐𝐽𝑠𝑐 (1-6)

16

Figure 1-8 An example of J-V curve of solar cells under illumination [43].

External quantum efficiency (EQE)

The EQE describes the ratio of incident photons to the converted electrons in a

photoactive device. Typically, two types of EQ are generally used [44]:

EQE, describes the ratio of the charge carrier numbers collected by the device to the

number of incident photons shining on the device.

IQE, i.e., internal quantum efficiency, which describes the ratio of the charge carriers

collected by the device to the number of the absorbed photons by the device.

EQE and IQE can be quantified using Equations 1-7 and 1-8.

、

17

EQE = 𝑒𝑙𝑒𝑐𝑡𝑟𝑜𝑛𝑠/𝑠𝑒𝑐

𝑝ℎ𝑜𝑡𝑜𝑛𝑠/𝑠𝑒𝑐=

𝑐𝑢𝑟𝑟𝑒𝑛𝑡/(𝑐ℎ𝑎𝑟𝑔𝑒 𝑜𝑓 𝑜𝑛𝑒 𝑒𝑙𝑒𝑐𝑡𝑟𝑜𝑛)

(𝑡𝑜𝑡𝑎𝑙 𝑝𝑜𝑤𝑒𝑟 𝑜𝑓 𝑝ℎ𝑜𝑡𝑜𝑛𝑠)/(𝑒𝑛𝑒𝑟𝑔𝑦 𝑜𝑓 𝑜𝑛𝑒 𝑝ℎ𝑜𝑡𝑜𝑛) (1-7)

IQE = 𝑒𝑙𝑒𝑐𝑡𝑟𝑜𝑛𝑠/𝑠𝑒𝑐

(𝑎𝑏𝑠𝑜𝑟𝑏𝑒𝑑 𝑝ℎ𝑜𝑡𝑜𝑛𝑠/𝑠𝑒𝑐=

𝐸𝑄𝐸

1−𝑅𝑒𝑓𝑙𝑒𝑐𝑡𝑖𝑜𝑛−𝑇𝑟𝑎𝑛𝑠𝑚𝑖𝑠𝑠𝑖𝑜𝑛 (1-8)

From EQE and IQE, the information of light response at different wavelength can be extracted,

which reflects the photovoltaic properties of the device. By applying a standard light illumination,

the integrated photocurrent density from the device can be obtained, which needs to be consistent

with that extracted from the J-V curve.

1.5 Scalable fabrication of PSCs

The PCE of PSCs has increased to 24.2% [5], catching up with that of the silicon solar cell,

as reviewed in section 1.1. The scalable fabrication of PSCs is an important topic towards their

commercialization [45–48]. Spin-coating is the most commonly used method for synthesis of

perovskite thin-film in labs. However, this method suffers from two main limitations, first, the film

area is usually under 10 cm2, and second, 90% of the precursor solution is wasted. Two categories

of scalable methods for fabricating perovskite films have been successfully demonstrated: (i)

meniscus-assisted coating such as doctor-blade coating and slot-die coating[49–52]; (ii) dispersed

deposition based on overlapping of individual impacting droplets, such as inkjet printing [53,54]

and spray coating [55–57], which can conformally deposit films on curved surface and are more

tolerant to the roughness or non-flatness of the substrate. The common scalable solution deposition

methods for the roll-to-roll fabrication of PSCs are shown in Figure 1-9.

18

Figure 1-9 Scalable solution deposition methods for the roll-to-roll fabrication of PSCs,

including blade coating (panel a), slot-die coating (panel b), spray coating (panel c), and

inkjet printing (panel d) [47].

19

1.5.1 Meniscus assisted coating

Meniscus assisted coating uses a moving meniscus liquid edge of precursor ink to spread

it across the substrate to form perovskite thin films.

Doctor-blade coating

Doctor-blade coating (Figure 1-9a) uses a blade to spread the precursor ink on the substrate

to coat thin films. A meniscus is formed between the substrate and the blade. The film thickness

is generally controlled by the gap between the blade and the substrate, the concentration of the

precursor ink, and the coating speed. The efficiency of small-area PSCs based on a doctor-blade

coated active layer has been progressively improved over 20% [50,58,59]. To regulate the crystal

growth during the coating process to form large-grained dense perovskite film, the substrate

temperature was held at ~150 °C and a trace amount of surfactants were also added into the

precursor ink to suppress the solution flow dynamics during drying as reported by Deng et al. [59].

Small area (~ 0.1 cm2) devices based on doctor-blade coated perovskite film has reached PCE over

20%. Using this process, perovskite solar cell module of area up to 57.2 cm2 was demonstrated

with efficiency up to 14.6%.

Doctor-blade coating is scalable and roll-to-roll compatible process. The material usage

rate is higher than that of spin-coating, but material waste still happens. One of the drawbacks is

the temperature has to be kept around 150 °C in order to form high-quality perovskite film, which

complicates the scalable fabrication process and is not suitable for devices on flexible substrates.

Slot die coating

20

For slot-die coating (Figure 1-9b), uniform flow of solution is delivered from the metering

system into the slot-die head. The solution comes out of the thin slit of the slot-die head and applied

to the substrate. A meniscus is also formed between the slit and the substrate. Because the solution

is pre-metered, so the material usage rate is very high with almost no waste. Liquid with a broad

range of viscosities can be coated using slot die coating. The thickness of the wet film can be

prescribed and precisely controlled by adjusting the flow rate of the solution fed into the die and

the coating speed [60]. In the slot-die coating of perovskite films, strategies such as heating the

substrate and gas flow were used to facilitate the drying of the wet film to get fully covered

perovskite film [61–63]. Research on slot-die coating of perovskite film is less active than that of

doctor-blade coating and the best efficiency of slot-die coated PSCs is lower. This is mostly

because slot-die coating usually requires a large amount of solution to fill the slot-die head and

therefore it is not suitable in the research phase. However, once the precursor ink has been

optimized, it starts to play an important role in the scale-up of the fabrication process [47].

Other novel meniscus-assisted coating methods such as soft-cover deposition have also

been developed recently. For the soft-cover deposition, a soft-cover with high surface wettability

is used to spread the perovskite precursor solution on a large area preheated substrate. A meniscus

forms between the cover and the substrate. To facilitate drying of the solvent, the substrate is

heated at the boiling point of the solvents (~ 210 °C) for a short period of time before dropping the

precursor solution, which is not compatible with flexible substrate. There will be limit on the scale

of the soft-cover in order to form uniform film and is not compatible with roll-to-roll fabrication.

21

For meniscus-assisted coating of perovskite film, the key in order to form high-quality

perovskite thin film with large crystal size, smooth surface, without pin-holes is to carefully control

the drying and micro flow.

1.5.2 Droplets based deposition methods

Another category of film deposition technique involves the generation and overlapping of

droplets, including spray coating and inkjet printing. According to the mechanism of generating

droplets, spray coating can be classified into fast gas flow assisted pneumatic spray, ultrasonic

vibration assisted ultrasonic spray and electric repulsion assisted electrospray.

Pneumatic and ultrasonic spray

Pneumatic spray (Figure 1-10a) employs compressed gas—usually air—to atomize and

direct the liquid to be deposited. The strong gas flow in the pneumatic spray may blow away active

ingredients in the liquid phase to cause excessive material waste which makes it not optimal for

large scale fabrication. Ultrasonic spray ((Figure 1-10b) employs high-frequency sound waves to

atomize the liquid to be deposited [64]. The high-frequency sound waves are converted into

mechanical energy that is transferred to the liquid. The liquid is then atomized into an ultrafine

mist of droplets when existing the ultrasonic nozzle. Every ultrasonic nozzle operates at a specific

resonant frequency, which dictates the size of the generated droplet. The higher is the frequency

of ulrasonic nozzle, the smaller is the median droplet size. The droplets generated from the

ultrasonic nozzle are generally in the range of 10 to 100 microns. The gas flow in ultrasonic spray

is for directing the mist of droplets instead of atomizing the droplets, therefore is much weaker and

this reduces the material waste.

22

Figure 1-10 (a) Schematic illustration of the pneumatic spray process [65], (b) Photographs

of ultrasonic spray-coating [66], (c) typical electrospray and details of the jet break-up.

Two issues in spray coating of perovskite film are the “wetness” or the degree of

evaporation of the droplets when reaching the substrate and the subsequent drying. Barrows et al.

[56] optimized the processing parameter space of spray coating, including the substrate

temperature and solvent vapor pressure to increase the film coverage to 85% and a PCE of 11%

was obtained for devices based on the spray-coated perovskite.

23

Re-dissolution of the pre-deposited perovskite film could also happen in spray coating

when the droplet reaches the pre-deposited film that is already dried. As a consequence, the

perovskite film is usually rough with low surface coverage. Naturally, the performance of the PSCs

based on the corresponding film is very poor. Heo et al. took advantage of the overlapping of wet

droplets on the pre-deposited film and reported a technique for re-dissolution and crystal grain

growth to prepare smooth perovskite films with large crystal grains. By choosing solvents with

appropriate vapor pressure, they balanced the inward flux of the spray solution with the outward

flux of the evaporating solvent. Consequently, the moistened underlying polycrystalline perovskite

film with small crystal grains re-dissolved and merged into larger crystalline grains through

recrystallization. The efficiency based on this kind of perovskite film reached an average power

conversion efficiency of 16.08% [65].

Electrospray

Electrohydrodynamic spray, or electrospray, is a liquid atomization technique that can

generate quasi-monodispersed fine droplets. This method was first discovered by Zeleny in 1915

[67]. Typically, an electrospray system (Figure 1-10c) can be implemented by feeding a fluid with

appropriate electrical conductivity through a small capillary that is charged to a few kV relative to

a nearby ground electrode. At the tip of the capillary, liquid meniscus can form four typical modes

of operation: dripping mode, cone-jet mode, micro-dripping mode and spindle mode [68]. Among

these modes, the most well-known and widely applicable mode is the cone-jet mode, commonly

known as Taylor-cone for Taylor. He first determined the angle of the cone and demonstrated that

surface tension and electric stress can be balanced at any point on a liquid cone surface [69]. An

electrified fine jet will issue from the Taylor-cone, which undergoes the Rayleigh-Plateau

24

instability and breaks up into two types of droplets: the primary droplets and satellite droplets [70].

The primary droplets will dominate the mass and the charge when forming the core of the spray

region while the satellite droplet will form the shroud due to their small inertia and strong initial

repulsions.

What makes electrospray deposition process special is that droplet size can be easily

controlled by electric charge level, flow rate, conductivity of the liquid and geometry of the

capillary. This results in a flexible droplet dimension from several nanometers to hundreds of

microns with narrow size distribution. Such properties enable the electrospray to find numerous

applications in the field of biotechnology and nanotechnology, among which the most famous

example is Electrospray Ionization Mass Spectrometry (ESI-MS). ESI-MS was first introduced by

Yamashita and Fenn in 1984. The later was awarded Nobel Prize in Chemistry in 2002 for using

electrospray to introduce the gas phase multiple charged macromolecule ions, originally in solution,

for subsequent analysis in a mass spectrometer.

25

Figure 1-11 1. Different types of cone-jet mode: a) pulsed cone-jet b) and c) multi-jet. 2.

Variations in meniscus forms for cone-jet mode (a – e). 3. Sequence of stages during dripping

mode. 4. Sequence of stages during micro-dripping mode. Taken from reference [68].

Electrospray also has very high material usage rate, because it uses an electric field to

generate uniform charged droplets [70]. The coulombic attraction force between the droplets and

the substrate suppresses droplet rebounce hence material waste is minimized [71]. The through-

put of electrospray could be greatly improved using multiplex electrospray nozzle arrays as shown

in Figure 1-12 [72].

26

Figure 1-12 Various kinds of nozzle arrays for the scalable electrospray deposition [72].

Such advantages enable the electrospray to find more application apart from ESI-MS, such

as thin film deposition, uniform nanoparticle synthesis for drug delivery, battery and solar cell.

Recently, it has also been introduced for the fabrication of PSCs [73–75]. Kavadiya et al. optimized

the two-step fabrication of perovskite film by electrospray. They found enhanced device stability

with electrosprayed perovskite film over the spin-coated ones with PCE reaching 12% [75]. In all

these prior studies, the PCEs of the PSCs with electrosprayed perovskite layer still lags behind

their spin-coated counterparts, mainly due to the non-optimal perovskite film quality. Electrospray

shares the same issues with pneumatic spray and ultrasonic spray when it comes to film quality.

Additionally, only the perovskite layer was fabricated using electrospray in these studies, while

the electron transport layer (ETL) and hole transport layer (HTL) were still prepared by

27

conventional spin-coating. It is highly desirable to apply one type of printing process to fabricate

all functional layers.

Inkjet printing

In inkjet printing (Figure 1-9d), the piezoelectric print head is used to generate droplets

with fine control of the size and trajectory of the droplet. The print head usually consists of many

miniaturized nozzles placed together within a short distance from each other. The application of

inkjet printing for fabrication of PSCs remains sparse [53,76]. High lateral resolution patterning

ability is the strength of inkjet printing. Complete fabrication of PSCs by inkjet printing is suitable

when there is a requirement for fine patterning, such as aesthetically pleasing and artistic solar

cells.

1.5.3 Other methods

Screen printing

In screen printing, the ink is transferred to the substrate using a screen with patterned mesh.

Photosensitive polymers are usually used to block the unwanted area of the mesh screen and the

unblocked mesh will allow ink to cross the screen and reach the substrate when a printing squeegee

is moved on the surface of the screen. In this way, the designed pattern could be printed. The

printed film thickness is determined by the mesh size and the thickness of the block layer. For

screen printing, the deposition area can be as large as several square meters, and the material

utilization can be as high as 100% for a continuous process. Screen printing is more suitable for

depositing thicker films (~1–10 μm), and it has been used to fabricate mesoporous scaffolds and

carbon back electrodes in PSCs [45,77].

28

1.6 Summary

Perovskite materials have excellent opto-electrical properties and can be fabricated using

cost-effective method. These two features make perovskite materials promising candidates for

solar cells. In just a few years, the efficiency of PSCs has improved from 3.8% [6] to 24.2% [5],

which is very close to that of the single crystalline silicone solar cell. In order to commercialize

PSCs, many scalable fabrication techniques have been applied to the preparation of perovskite

films, such as doctor-blade coating, spray coating, inkjet printing, and etc. The highest efficiency

of the PSCs based on perovskite films prepared using scalable method has achieved over 20% [59].

However, most of reported scalable PSCs based only on the perovskite layer that are prepared

using scalable methods, with other layers such as ETL, HTL still prepared using spin-coating and

electrode by thermal evaporation. In this dissertation, we are striving to apply the electrospray

printing, which is scalable and materials-saving, to realize the fully scalable fabrication of PSCs

including all the function layers.

29

Chapter 2 Experimental

Chapter 2 will describe the experimental techniques utilized to synthesize and characterize

PSCs. This chapter is divided into the solution preparation, device making, and characterization

three parts.

2.1 Solution preparation

Perovskite precursor solution

CH3(NH)2I (FAI), CH3NH3Br (MABr) powders and PbI2 were added in a mixture of γ-

Butyrolactone (GBL) and1-Methyl-2-pyrrolidinone (NMP) (7:3 v/v) to make 1 M

FA0.85MA0.15PbI2.85Br0.15 solution. All chemicals are purchased from Sigma-Aldrich and used as

received without further purification. The small amount of MA and Br added aims to increase the

stability of the FAPbI3 perovskite phase, because the pure FAPbI3 is not stable and tend to

transform to non-perovskite phase. The same perovskite precursor solution is used for both

electrospray and spin-coating.

TiO2 ETL precursor solution

Ti-Nanoxide T-L paste (Solaronix) was mixed with DI water and tert-butyl alcohol,

followed by stirring for 2 h and ultrasonic dispersing for 30 min to form a uniform TiO2

nanoparticle dispersion. This dispersion is directly used for spin-coating. When used for

electrospray printing, an extra small amount of DMSO was added to stabilize the Taylor cone-jet

mode.

Spiro-OMeTAD precursor solution

30

72 mg of Spiro-OMeTAD was dissolved in 1 ml chlorobenzene, with the addition of 23 μL

Li-TFSI/acetonitrile (170 mg/mL), 20 μL of 4-tert-butylpyridine (TBP) to form the Spiro-

OMeTAD solution. This solution is directly used for spin-coating. A small change of solvent was

made when used for electrospray printing, 1-2 Dichlorobenzene was used instead of chlorobenzene

and was diluted to 7.2 mg/ml, in order to stabilize the Taylor cone-jet mode and form uniform

films. It will be detailed in chapter 3.

Carbon ink

The carbon ink was prepared by dispersing the dried carbon paste (Guangzhou Seaside

Technology Co., Ltd) in chlorobenzene at a concentration of 40 mg/ml and 10 vol% LiTFSI

solution (10 mg/ml) was added to increase the electric conductivity of the ink, extra graphite (US

Research Nanomaterials, Inc.) could also be added into the ink as designed. The process is depicted

in Figure 2-1.

Figure 2-1 The process to prepare carbon ink for electrospray.

31

2.2 Device preparation

All processes for device fabrication were performed in ambient conditions. The glass/FTO

substrate was successively washed with Hellmanex III, acetone, ethanol, and DI water. The sheet

resistance of FTO is 12-14 ohm/sq and the thicknesses of glass and FTO are 2.2 mm and 200 nm,

respectively. The average transmittance of glass/FTO in the visible region is around 82-84.5 %.

The devices in this dissertation are based on the planar n-i-p structure. The ETL, perovskite layer,

and HTL was sequentially deposited onto the cleaned FTO glass by spin-coating or electrospray.

The last step is to deposit the top electrode. Gold thin film prepared using thermal evaporation is

the most commonly used top electrode for n-i-p structure PSCs. In this dissertation, we tried both

the thermally evaporated gold electrode and carbon electrode prepared using doctor-blade coating

or electrospray printing.

2.2.1 Deposition of films by electrospray printing

A typical electrospray system can be implemented by feeding a liquid with sufficient

electrical conductivity through a small capillary that is charged to a few kV relative to a nearby

ground electrode. The liquid encompasses a conical shape (termed Taylor-cone [78]) with an

electrified fine jet issuing from the cone. The jet undergoes the Rayleigh-Plateau instability [70]

and breaks up into ultrafine and quasi-monodispersed droplets.

32

Figure 2-2 Picture of the actual electrospray system.

All the electrospray printing processes took place in a fume hood in the ambient

atmosphere. Figure 2-2 shows the picture of the actual electrospray system, which consists of an

electrospray nozzle on a syringe driven by a syringe pump, a high voltage DC power supply, a

cone-jet visualization subsystem using a camera with a microscope lens, and a substrate motion

control subsystem using a CNC 3D motion stage. Prior to printing, the prepared solution was

loaded into a syringe and the flow rate was controlled by the syringe pump. The substrate and the

syringe were mounted on a computer controlled motorized linear stage and a serpentine path

motion of electrospray nozzle relative to the substrate was designed. TiO2, perovskite and Spiro-

OMeTAD films were printed sequentially on the pre-cleaned FTO glass. Table 1 summarizes the

parameters such as working distance, flow rate and printing speed for printing TiO2, perovskite

and Spiro-OMeTAD, and carbon. The TiO2 layer was printed first, and the printed film was

annealed at 150 °C for 30 min. The perovskite film was printed on the annealed TiO2 layer, and

33

after the wet perovskite precursor film was printed, the sample underwent flash vacuum drying in

a vacuum chamber which induces rapid sample drying within a few seconds. The dried perovskite

film was annealed on a hotplate at 150 °C for 15 min. Next, the Spiro-OMeTAD layer was printed

on top of the perovskite film. Similarly, the sample with wet Spiro-OMeTAD was dried by flash

vacuum. Finally, for the gold-based device, an 80 nm thick gold layer was evaporated as the top

electrode. The active area of each device was 0.1 cm2. And for the carbon-based device, the carbon

electrode was deposited on top of perovskite by doctor-blade coating or electrospray coating. For

carbon-based device, the hole transport layer is eliminated the carbon electrode itself also serve

some function of hole transport. During the electrospray printing of Carbon, the stage temperature

was kept at 100 oC. And after printing, the film was placed on a hot plate of 120 oC for 30 min.

Table 2-1 Typical ES printing parameters for TiO2, perovskite, Sprio-MeOTAD and

carbon ink.

Working

distance (mm)

Flow rate

(µL/min)

Offset

(mm)

Printing speed

(mm/min)

TiO2 2 0.3 0.2 333

perovskite 1.2 0.4 0.2 900

Spiro-MeOTAD 1.2 3 0.2 700

Carbon 1.5 15 0.2 300

34

2.2.2 Deposition of films by spin-coating

All the solutions were filtered using a 0.2 µm filter before spin-coating. All process of spin-

coating for ETL, perovskite, and ETL were very similar, the only difference is the spin-coating

speed and whether in dynamic or static mode of spin-coating. The spin-coating conditions for

different layers are summarized in table. The dynamic spin-coating refers to the process that firstly

let the spin-coater go to the desired speed and then drop the solution onto the substrate while it is

spinning, this method is required for those solutions with high-vapor-pressure solvents. On the

other hand, static spin-coating refers to the process that the solution is first dropped on to the

substrate and then start the spin-coater.

Table 2-2 Typical spin-coating parameters for TiO2, perovskite, and Sprio-MeOTAD.

Precursor solution Spin-coating speed Dynamic or Static Annealing

ETL (TiO2) / H2O, 4000 Dynamic 150 °C for 30 min

ETL (SnO2) /H2O 3000 Static 150 °C for 30 min

Perovskite /GBL, NMP 2000 Static 150 °C for 15 min

Sprio-MeOTAD

/Chlorobenzene

4000 Dynamic \

2.3 Characterization

The characterization techniques could be those used to characterize the various layers in

PSCs and those used to characterize of the PSCs.

Optical profiler

35

Optical profiler is used to observe the surface structure of thin film and measure the

thickness of thin film. Optical profiler is a non-contact 3D imaging instrument based on white light

interference. The light was split by beam splitter, and part of the light is directed to the sample,

and the other part toward the camera as a reference. The light reflected from a different area of the

sample will have different phases due to the texture of the sample surface. The reflected light will

recombine with the reference light and an interference pattern is formed which could be analyzed

by the software to determine the topography of the sample. In order to measure film thickness, the

film needs to be scratched down to the substrate and then measure the step-height.

Scanning electron microscopy (SEM)

SEM was utilized to characterize the microstructure of the sample. The scanning electron

microscope (SEM) produces images by scanning the sample with a high-energy beam of electrons.

As the electrons interact with the sample, they produce secondary electrons, backscattered

electrons, and characteristic X-rays. These signals are collected by one or more detectors to form

images. SEM images in this dissertation were obtained from scanning electron microscope (Zeiss

1550). Both surface topography and cross-section were characterized. The cross-section image

could be used to determine the film thickness, which is more accurate than using the optical profiler.

Atomic force microscope (AFM)

AFM is a way to characterize the surface topography. An AFM uses a cantilever with a

very sharp tip to scan over a sample surface. The cantilever will be deflected towards or away from

the surface depending on the distance between the tip and the surface. A laser beam is used to

detect cantilever deflections and is recorded by the PSPD. By using a feedback loop to control the

36

height of the tip above the surface—thus maintaining constant laser position—the AFM can

generate an accurate topographic map of the surface features. Conductive AFM is used to obtain

electrical properties in nanoscale of the prepared samples. By applying a bias between the

conductive cantilever and the sample as shown in Figure 2-3, with an assistance of a current

amplifier to measure the current flows between the two, the electric information could be obtained

during scanning along with topography information. AFM images in this dissertation were

obtained from AFM (Parker XE7) in the dark.

Figure 2-3 Schematic diagram of the Conductive AFM system, taken from [79].

X-ray diffraction (XRD)

XRD analysis was used to determine the crystalline structure of the prepared films. X-rays

diffract through the crystal and into a detector. The beam and detector are rotated through a range

of angles. The angles at which the crystals diffract the beam into the detector correspond to planes

37

of the crystals. Each crystal has a characteristic pattern of diffraction angles and corresponding

intensity of the diffracted beam. XRD measurements in this dissertation were performed at a

scanning rate of 5 ˚/min on an X-ray diffractometer (Philips Xpert Pro).

J-V curve under illumination

Photovoltaic performance of the solar cells was analyzed under one sun (AM 1.5 G, 100

mW/cm2) illumination with a solar simulator (150W Sol 2ATM, Oriel), and the current-voltage

characteristics of each cell were recorded with a digital source meter (Keithley 2400). Photovoltaic

metrics for PSCs such as Voc, Jsc, FF and PCE could all be extracted from the JV curve under

illuminated condition.

Photoluminescence (PL)

Photoluminescence spectroscopy is a widely used technique for characterization of the

optical and electronic properties of semiconductors and molecules. In a typical PL experiment, the

material is excited with a light source and electron is excited into higher energy state and then

relaxes to a lower energy state through the emission of a photon. The relaxation processes can be

studied using Time-resolved fluorescence spectroscopy to find the decay lifetime of the

photoluminescence. PL spectroscopies in this dissertation were measured using FLS 1000 from

Edinburgh Instruments.

38

Chapter 3 All Electrospray Printed Perovskite Solar Cells

Most of the results presented in this chapter were published in Yuanyuan Jiang, Congcong

Wu, Liurui Li, Kai Wang, Zui Tao, Fan Gao, Weifeng Cheng, Jiangtao Cheng, Xin-Yan Zhao,

Shashank Priya, Weiwei Deng, “All electrospray printed perovskite solar cells” Nano Energy. 53

(2018) 440–448.

3.1 Introduction

Hybrid organic-inorganic lead halide perovskites have emerged as promising photovoltaic

materials because of their outstanding optoelectronic characteristics such as the broadband light

absorption, high ambipolar mobilities, and long charge-carrier diffusion lengths [9,10]. In addition,

lead halide perovskites can be synthesized using low-cost and earth-abundant raw materials

through solution-based processes at moderate temperatures (below 150 C), rendering them

promising candidates for low-cost thin film photon energy conversion devices. Within eight years,

the power conversion efficiencies (PCEs) of perovskite solar cells (PSCs) have leapfrogged from

3.8% to a certified 24.2% [13,80–82]. To make PSCs commercially competitive, one of the

remaining challenges is to develop a facile and cost-effective scalable printing process. Two

categories of scalable methods for fabricating perovskite films have been successfully

demonstrated: (i) continuous deposition such as slot-die coating [51,52] and doctor blade coating

[49,58,59,83]; (ii) dispersed deposition based on overlapping of individual impacting droplets,

such as inkjet printing [53] and spray coating [56], which can conformally deposit films on curved

surface and are more tolerant to the roughness or non-flatness of the substrate. Most spray

39

deposition methods such as airbrush and blow dryer require strong gas flow to atomize, spread, or

dry the liquid [56,84]. The strong gas flow may blow away active ingredients in the liquid phase

to cause excessive material waste. Electrospray, in contrast, does not require the assistance of gas

flow because it uses an electric field to generate uniform charged droplets [70]. The Coulombic

attraction force between the droplets and the substrate suppress droplet rebound hence material

waste is minimized [71]. Because electrospray is also scalable and roll-to-roll compatible, recently

there has been growing interest in the application of electrospray deposition for perovskite thin

films using one-step or two-step method [73–75]. Kavadiya et al. optimized the two-step

fabrication of perovskite film by electrospray and discovered the enhanced device stability with

electrosprayed perovskite film over the spin-coated ones and the PCE reaches 12% [75]. In all

these prior studies, the PCEs of the PSCs with electrosprayed perovskite layer still lag noticeably

behind their spin-coated counterparts. Additionally, only the perovskite layer was fabricated using

electrospray in these studies, while the electron transport layer (ETL) and hole transport layer

(HTL) were still prepared by conventional spin-coating. It is highly desirable to apply one type of

printing process to fabricate all functional layers.

Here we report a universal electrospray process to print the perovskite layer, HTL and ETL

of the PSCs in ambient air environment at modest temperatures. Specifically, the precursor ink

formulation and printing parameters were judicially chosen to make the time of flight shorter than

the evaporation time of the droplet, thereby promoting the formation of continuous wet precursor

films that lead to high-quality dry functional films. Smooth and pin-hole free 300 to 500 nm

perovskite, 90 nm TiO2-based ETL and 140 nm Spiro-OMeTAD based HTL were successfully

40

electrospray printed in ambient air below 150 C. The all-electrospray printed devices exhibited a

PCE up to 15.0%.

3.2 Results and discussion

A typical electrospray system can be implemented by feeding a liquid with sufficient

electrical conductivity through a small capillary that is charged to a few kV relative to a nearby

ground electrode as shown in Figure 3-1. The liquid encompasses a conical shape (termed Taylor-

cone [78]) with an electrified fine jet issuing from the cone. The jet undergoes the Rayleigh-Plateau

instability [70] and breaks up into ultrafine and quasi-monodispersed droplets.

Figure 3-1 Diagram of an electrospray system.

The key to print high-quality polycrystalline perovskite film lies in successfully addressing

three critical issues: (1) the stable operation of the electrospray in the cone-jet mode that ensures

41

uniform droplet diameters, (2) the printing of even and wet perovskite precursor thin film, and (3)

the proper drying and conversion of the wet film into polycrystalline perovskite film. These three

issues are discussed in details below.

3.2.1 The stable operation of the electrospray in the cone-jet mode

The perovskite precursor solutions are challenging to electrospray in the cone-jet mode

because of at least two reasons. First, emitters made by typical metal such as stainless steels are

not suitable for spraying perovskite precursor solution without introducing impurities because the

Pb2+ ions in the solution will have red-ox reactions with the metal. Second, the electric conductivity

of the perovskite precursor solution is as high as 4.3 S/m, which leads to a very fine jet diameter

that chokes the flow and restricts the maximum flow rate (under 1 L/min in this work). The low

flow rate is prone to perturbations from mechanical vibration, liquid supply instability, and the

wetting behavior of the emitter.

To address these two challenges, we used a silica nozzle with 30 m tip outer diameter

(OD) as the emitter because silica is inert to Pb2+ ions. The small nozzle effectively increases the

viscous damping that stabilizes the cone-jet. The electric contact was made by using a laser

machined conductive silicone gasket that connects the silica emitter with the perovskite precursor

solutions. The assembly of the silicone nozzle and the gasket is shown in Figure 3-2.

42

Figure 3-2 Silica nozzle with 30 µm tip inserted into the conductive silicone gasket.

Another strategy was to apply anti-wetting coating onto the outer surface of the nozzle to

suppress the wetting of the precursor solution onto the nozzle. As shown in Figure 3-3, during

electrospray, without the anti-wetting coating the Taylor cone would keep growing bigger from as

shown from Figure 3-3a to Figure 3-3b until the Taylor cone finally burst out and a large amount

of solution discharged rapidly, the droplets could be very large in contrast to the uniform deposition

of small-sized droplets when in stable Taylor cone-jet mode. And this kind of wetting induced

cone growth and expulsion cycle occurred with a frequency of once every several seconds, which

greatly disrupt the otherwise uniform film. Wetting-induced cone-jet instability would further

43

complicate the deposition of perovskite film process because the perovskite could deposit onto the

outer surface of the nozzle due to wetting, which could be re-dissolved, making the flow rate less

predictable as well as leading to variations in the perovskite concentration of the solution, and

consequently non-uniformity of the resulting film.

Figure 3-3 Taylor cone without (a), (b) and with(c) anti-wetting coating on the outer

surface of the nozzle.

3.2.2 Printing of uniform wet perovskite precursor thin film

For printing methods based on individual droplets to eventually achieve smooth and dense

dry film without pin-holes, it is essential to form continuous wet precursor films to avoid the re-

dissolving pitfall described in section 1.5.2. To that end, the droplets should remain fluidic upon

impacting the substrate, allowing the droplets to merge with each other and become part of a

continuous wet film. Otherwise, the droplet would dry prematurely, and the substrate will only

collect nanoparticles, resulting in a rough and porous film with many voids as shown in left image

in Figure 3-4. The degree of wetness of the droplets when they about to impact the substrate can

be quantitatively described using Damkhöler number (Da) of evaporation as [85]:

𝐷𝑎 =𝑡𝑓

𝑡𝑒 , (3-1)

44

where 𝑡𝑓 is the droplet fight time and 𝑡𝑒 is the droplet evaporation time. Many ES processing

parameters will directly or indirectly influence 𝑡𝑓 and 𝑡𝑒 . Droplet diameter d0 is one of the

important intermediate parameters, which can be introduced from scaling laws [86].

6/13

00

k

QCd d

, (3-2)

where Cd is the scaling constant of order one, Q flow rate, 0 the vacuum permittivity, ρ the density

of the liquid, k the electrical conductivity, γ the surface tension. Equation 3-2 suggests that the

droplet size is determined by the liquid properties (surface tension γ and the electrical conductivity

k) and electrospray process condition liquid flow rate Q.

Approximately, 𝑡𝑓 can be calculated from:

𝑡𝑓 =𝑊𝐷

𝑉𝑡 , (3-3)

where 𝑊𝐷 is the working distance between the nozzle and the substrate, and 𝑉𝑡 is the

droplet terminal velocity. When the droplet size is very small (< 10 µm), which is the case for

perovskite precursor solution and other liquids used in this dissertation, terminal velocity can be

reached in a few µs and the Reynolds number is guaranteed to be small as an order of unity. The

droplet terminal velocity at low Reynolds number flow is:

0

0

3 d

Eqvt

, (3-4)

where q0 is the electric charge carried by every droplet, E is the electric field, which is

approximately V/WD, and µ is the viscosity of the ambient air. The q0 of a droplet generated by

electrospray is:

45

q0 =I

Q×p

6d0

3, (3-5)

which is close to the Rayleigh charge limit [87]. The current I can also be approximately calculated

from the scaling laws [86]:

2/1kQCI I , (3-6)

where CI is another scaling constant of order one. From Equations 3-2 to 3-6, an expression for 𝑡𝑓

can be obtained.

On the other hand, 𝑡𝑒 is often derived from the classic d-squared law [88]:

)(/2

0 TKdt ee , (3-7)

where Ke(T) is the evaporation rate, which can be calculated from the mass transfer models for an

isolated spherical droplet [89]:

)1

1(8

s

g

diffeY

YLnDK

, (3-8)

where Ddiff is the mass diffusivity of the vapor molecule to the ambient environment, ρg is the

solvent vapor density, Y∞ is the vapor partial concentration far from the droplet, and Ys is the vapor

partial concentration at the surface of the droplet. Ys can be evaluated from Pv/P0 (the ratio of vapor

pressure of the solvent to ambient pressure). Typically, Y∞ is zero, and Ys =Pv/P0<<1, therefore

Equation 3-8 becomes:

0

8P

PDK vg

diffe

. (3-9)

Combining Equations 3-2 to 3-9, and using the scaling constants from literature [86], we can

express Da as:

46

0

3/52/3

26/1

3/2

0

)(28

P

PD

VQ

WDkDa vgdiff

. (3-10)

As shown in Equation 3-10, it is clear that Da is affected by many variables, which can be

organized into five groups, separated using parentheses. The first three groups are the scaling

constant, the critical liquid properties (surface tension and electrical conductivity k), and the key

ES operating conditions (working distance WD, voltage V, and flow rates Q). The fourth group

includes properties that are weak functions of temperature T. For example, Ddiff scales with T1.5,

which suggests that when the temperature is raised from 25 C to 100 C, Ddiff is only expected to

have a modest increase of 40%. This is in sharp contrast to the last group, in which the vapor

pressure Pv changes rapidly with respect to the temperature T.

Figure 3-4 SEM top-view images of perovskite films printed at different conditions. The

scale bar is 1 micron.

At a typical perovskite precursor flow rate Q of 700 nL/min and k = 4.3 S/m, the estimated

initial droplet diameter is ~130 nm (the detail numbers used for estimating initial droplet diameter

are shown in supplement), which is prone to rapid drying according to Equation (3-7). To ensure

47

wet deposition, we have to decrease the evaporation Ke. According to Equation 3-9, Ke can be

decreased by using solvents have low vapor pressure Pv, therefore we chose the mixture of γ-

Butyrolactone (GBL) and 1-Methyl-2-pyrrolidinone (NMP) as the solvent, both of which have

very low vapor pressure (1.5 mm Hg for GBL and 0.29 mm Hg for NMP at 20 oC) and as a result

the droplet evaporation time te increased ~10-4 s. On the other hand, in order to decrease the droplet

flight time tf , the working distance (WD) between the nozzle and the substrate is decreased to 1.2

mm. With all these strategies, we can ensure that Da is less than 1, as schematically illustrated in

Figure 3-5. Different Da will lead to different outcomes of either nano-porous perovskite film (Da

> 1) or dense perovskite film (Da < 1), as shown in Figure 3-4.

Figure 3-5 Diagram of the solvent evaporation during the electrospray process.

To print continuous wet perovskite films with large areas, electrospray is combined with

the relative motion of the substrate and the nozzle using an X-Y-Z 3D motorized stage to realize

electrospray printing. We employed a serpentine motion path as shown in Figure 3-6, every path

48

will overlap with adjacent paths and finally form a film of a large area with uniform thickness

except the very edge. Since the relative motion of the nozzle could go unlimited in the printing

direction, electrospray printing process is continuous and scalable.

Figure 3-6 Schematic of serpentine path motion of electrospray nozzle relative to the

substrate [90].

The thickness of the dry perovskite film can be controlled by precursor solution

concentration and printing parameters, as expressed in Equation (3-11):

δ = 𝑄

𝑉𝑌offset , (3-11)

where δ is the film thickness, is the volume concentration of the precursor, Yoffset is the

distance between two adjacent printing paths and V is the stage moving speed. Hence, the film

thickness can be precisely prescribed.

49

Figure 3-7 shows a series of electrospray printed perovskite wet films with different paths

overlapping. These images were taken using the optical profiler and the different colors indicate

different thickness. For a single path printing (Figure 3-7a), the working distance determined the

width of film printed. Here a working distance of 1.2 mm was used, hence the width of the single

path is around 1 mm. And about half of the width has uniform thickness indicated by the uniform

yellow color in the center. When a second path (Figure 3-7b) or a third path (Figure 3-7c) with an

Yoffset was printed, an overlap between paths happened and the central uniform area grows.

Figure 3-7 Optical microscope images of perovskite film printed using one, two and three

paths.

From Figure 3-7, a kind of asymmetry was observed in the printed film, and it became

more severe when the number of paths increased. This asymmetry was caused Marangoni effect

[91] as shown in Figure 3-8a, the earlier printed part would experience solvent evaporation first

and will cause a surface tension gradient to induce a micro-flow from the freshly printed area to

the previously printed area. We found that by increasing the contact angle between the perovskite

50

solution and the substrate, this kind of asymmetry could be mitigated. For the as-prepared TiO2

film surface after annealing at 150 °C for 30 min, the contact angle is nearly 0, while after the

TiO2 surface is post-processed simply by exposing to the vacuum environment, the contact angle

is increased to ~18. The mechanism of this change of contact angle is unknown. Nevertheless,

the contact angle profoundly affects the printing outcome. For a small contact angle, the solution

can spread on the substrate, leading to an inhomogeneous film, as indicated by the obvious change

in shades of brown color (Figure 3-8b). When the contact angle is increased to ~18, the variation

of color becomes much less, suggesting that the thickness of the film is more uniform (Figure

3-8c). Small contact angle corresponds to more uneven evaporation that causes stronger

Marangoni flow to convect the liquid from the freshly printed region to the semi-dried edge. As

the contact angle increases, the Marangoni flow becomes weaker and the film becomes evener.

Figure 3-8 (a) Schematic illustration for the directional microscale solution flow towards

perovskite film printed earlier. (b) Optical image of perovskite films obtained from nearly

0 contact angle and (c) 18 contact angle.

51

3.2.3 Fast drying induced high-quality perovskite film

After successfully printed uniform and wet perovskite precursor films, the drying method

or more specifically speed of drying and subsequent crystallization was the key in order to obtain

the final smooth and compact perovskite film. Firstly, we tried the anti-solvent method that has

been reported [92–94]. Anti-solvent works by adding a secondary solvent (a liquid in which the

solute is insoluble) to the solution to reduce the solubility of the solute and consequently to

generate a supersaturated driving force and fast drying of the wet film [94]. Toluene was used as

anti-solvent here to induce fast drying of the freshly electrospray printed perovskite film. The film

obtained was shown in Figure 3-9a and c. Grains could be seen from the optical microscope image.

And AFM topography shows the surface roughness is around 40 nm, which is a bit too rough for

perovskite device to gain high efficiency. To decrease the roughness, we tried another method,

vacuum flash drying [95]. The freshly printed film was placed into a small vacuum chamber to

induce rapid drying and crystallization. This process took less than 20 seconds. The resulting film

is as shown in Figure 3-9b and d. No obvious grains were observed from the optical image and the

surface roughness measured using AFM was only around 11 nm, which is much more smooth than

that from the anti-solvent method. This indicates that the speed of solvent extract is faster when

using vacuum flash drying, because of which the crystallization of perovskite film is faster and the

resulting film is more smooth. Therefore, we chose the vacuum flash drying as the film drying

method in this dissertation.

52

Figure 3-9 Optical Microscope images (a, b) and AFM topographies (c, d) of Electrospray

printed perovskite films dried using anti-solvent method (a, c) and vacuum-assisted flash

drying (b, d).

There exists a certain film thickness range for forming pin-hole free perovskite films using

vacuum flash drying. As shown in the cross-section view SEM images (Figure 3-10), the thickness

of films printed at the flow rate of 0.43, 0.57 and 0.72 µL/min are 300, 400 and 500 nm respectively,

which is consistent with the values calculated using Equation (3-11). The top-view SEM images

show that the perovskite grain size increases with flow rates. At low flow rate (0.43 µL/min), the

vacuum flash drying of the thin wet film is completed within seconds, achieving a high degree of

53

Figure 3-10 Top-view (left column), corresponding cross-section view (middle column) of

SEM images and AFM surface scanning images (right column) of perovskite films

electrospray printed at flow rate of 0.43, 0.57, 0.72 µL/min.

supersaturation, which can promote homogeneous and fast crystal nucleation [96]. However, the

crystal size is small because of the short time for crystal growth and the small film thickness. When

the flow rate is increased to 0.57 µL/min, the longer drying process allows more time for crystal

growth which induces larger crystal domains. However, when the flow rate is further increased to

0.72 µL/ min, the dried film becomes rougher with pin-holes, which is likely due to the further

slowed drying that decreases the degree of supersaturation so that heterogeneous nucleation or

54

secondary nucleation emerges [96]. The root means square roughness (Rq) of the perovskite films

printed at the flow rate of 0.43, 0.57, and 0.72 µL/min is 7 nm, 9 nm, and 21 nm respectively, as

indicated by the AFM images in Figure 3-10, in accord with the morphology revealed by SEM

images.

The crystalline structure of the electrospray printed perovskite is examined by the X-ray

diffraction (XRD). The XRD spectra of electrospray printed film showed the typical black phase

of FA0.85MA0.15PbI2.85Br0.15 perovskite diffraction [97], which is the same with that of spin-coated

samples (Figure 3-11), indicating that excellent crystalline perovskite can be formed through the

electrospray process.

Figure 3-11 XRD of the perovskite films using electrospray printed and spin-coated

perovskite.

55

To demonstrate the photovoltaic performance of the electrospray printed perovskite layer,

we compared the PCE of the electrospray printed cell and spin-coated cell, as shown in Figure

3-12. The ETL and HTL in the electrospray printed devices were still fabricated by spin-coating.

The corresponding J-V curves are shown in Figure 3-19. The highest PCE (champion cell: PCE

15.9%) for the device using the electrospray printed perovskite layer at the flow rate of 0.57

µL/min was very close to that of spin-coated PSC (champion cell: PCE 16.1%). This is attributed

to the large grain size and smooth surface of perovskite films printed at 0.57 µL/min. Notably, the

electrospray printing uses significantly less ink than spin-coating because of high material

utilization rate. For example, only 1 µL precursor solution is needed to print a 500 nm thick

perovskite film with an area of 20 mm15 mm, while at least 20 µL solution is required to coat

the same area for spin-coating.

Figure 3-12 PCEs of the PSCs using electrospray printed and spin-coated perovskite.

56

3.2.4 Electrospray printing of electron transport layer and hole transport

layer

The same principle of keeping Da < 1 is also applied in the electrospray printing of TiO2

based ETL and Spiro-OMeTAD based HTL. For ETL, the choice of solvent is limited because the

TiO2 nanoparticles were pre-dispersed in deionized water and the addition of other solvent

destabilize the TiO2 nanoparticles suspension. Da < 1 can still be guaranteed by reducing tf by

adjusting the working distance. The printed TiO2 films became smoother as tf is shortened Figure

3-13). With the shortest tf, the electrospray printed TiO2 layer is nearly as smooth as the spin-

coated TiO2. This again confirms the importance of forming wet films for obtaining smooth dry

films in the electrospray printing process. The J-V curves (Figure 3-14) of the spin-coated and

electrospray printed FTO/TiO2/Au structure are virtually indistinguishable, indicating similar

conductivity for electrospray printed and spin-coated TiO2. The PCE data of devices fabricated

using spin-coated and electrosprayed TiO2 films as ETL are shown in Figure 3-15(see J-V curve

in Figure 3-20). Here, perovskite films were electrospray printed and Spiro-OMeTAD were spin-

coated for all devices. Not surprisingly, the PCE increases as the TiO2 film becomes smoother.

57

Figure 3-13 Top-view and corresponding cross-sectional SEM images of TiO2 films

electrospray printed at working distance of 4 (a), 3 (b), 2 (c) mm and spin-coated (d). The

scale bar is 600 nm.

58

Figure 3-14 Conductivity comparison between the spin-coated TiO2 film and electrospray

printed at 2 mm.

Figure 3-15 PCEs of the PSCs using electrospray printed and spin-coated perovskite films.

59

For the HTL, in formulating the Spiro-OMeTAD solution, we chose 1,2-dichlorobenzene

(DCB) instead of chlorobenzene (CB) as the solvent, because DCB has much lower vapor pressure

(1.36 mm Hg at 25 C) than that of CB (12 mm Hg at 25 C). Thus DCB droplet provides a longer

evaporation time to ensure wet deposition. Indeed, Spiro-OMeTAD particles were present in the

electrospray printed Spiro-OMeTAD film using CB solution due to rapid evaporation, but when

CB was replaced by DCB, the electrospray printed Spiro-OMeTAD film is smooth and

indistinguishable from that of the spin-coated Spiro-OMeTAD films (inset of Figure 3-16). Figure

3-16 shows the PCE data of four groups of devices with Spiro-OMeTAD processed under four

conditions: (i) spin-coated Spiro-OMeTAD dissolved in CB, (ii) spin-coated Spiro-OMeTAD

dissolved in DCB, (iii) electrospray printed Spiro-OMeTAD dissolved in CB, (ⅳ) electrospray

printed Spiro-OMeTAD dissolved in DCB. Here, both the ETL (TiO2) and perovskite layer were

electrospray printed using the optimal parameters established earlier. The PCEs of the devices

using spin-coated Spiro-OMeTAD with CB or DCB as the solvent are similar, but for the

electrospray printed Spiro-OMeTAD, the PCE is improved from 13.0% (CB as solvent) to 15.0%

(DCB as solvent) (see Figure 3-21 for J-V curves), suggesting that the improved surface

morphology of the printed Spiro-OMeTAD boosts the PCE of devices.

60

Figure 3-16 PCE of PSCs using four different fabrication conditions for Spiro-OMeTAD:

electrospray printing with CB or DCB as solvent, and spin-coating with CB or DCB as

solvent. The inset shows the corresponding top-view SEM images. The scale bar is 400 nm.

3.2.5 The performance of all electrospray printed PSCs

Finally, we printed all three functional layers (ETL, perovskite, and HTL) using

electrospray process based on the PSC architecture of FTO/TiO2/FA0.85MA0.15PbI2.85Br0.15/Spiro-

OMeTAD/Au. The cross-sectional view SEM image (Figure 3-17a) of the complete device showed

all three layers are homogeneous and dense. Figure 3-17b shows the J-V curves of the top

performers of the all-printed device, the all spin-coated device, and the device with only the

perovskite layer was electrospray printed. The photovoltaic metrics of these devices are

summarized in Table 1. The all electrospray printed device showed an open-circuit voltage (VOC)

of 1.06 V, a short-circuit current (JSC) of 21.9 mA/cm2, a fill factor (FF) of 64.1% and PCE of

15.0%, which reflect only a modest performance drop from the all-spun devices. The electrospray

61

printed perovskite based device showed higher Jsc than the spin-coated perovskite based device,

which is owing to the perovskite grains by electrospray printing are larger and more uniform than

that from spin-coating (Figure 3-18). To examine the reproducibility of the all-printed device, we

have conducted J-V measurements on one batch of devices and the inset of Figure 3-17 shows the

PCE distribution, with an average efficiency of 13.6%. The stability of the all-printed devices

without encapsulation was also studied. The data (Figure 3-22) show that the all-electrospray

printed devices are fairly stable with only 8% efficiency loss after being stored for 30 days in

ambient condition with a relative humidity of ~20% at room temperature.

Table 3-1 Photovoltaic metrics of the all electrospray printed, all spin-coated devices and

the device with only the perovskite layer electrospray printed.

Process Voc (V) Jsc (mA/cm2) Fill Factor

(%)

Champion

Efficiency (%)

Average

Efficiency (%)

All electrospray printing 1.06 21.9 64.1 15.0 13.6

All-spin coating 1.10 20.7 70.6 16.1 15.0

Perovskite layer printed

by electrospray

1.09 25.0 58.2 15.9 14.7

62

Figure 3-17 (a) Cross-section view SEM image of an all electrospray printed photovoltaic

device. (b) J-V curves for the champion cell of the all-printed, all spin-coated devices and

the device with only perovskite layer electrospray printed. Inset: The PCE distribution

histogram of the all-printed devices.

Prior investigations [58,59,98,99] have shown great promise in the scalable fabrication of

high-quality perovskite photoactive layer, enabling PCE up to 20% for PSCs with printed

photoactive layer, yet only a limited number of studies have applied the scalable deposition process

for all three layers (ETL, perovskite, and HTL) [51,52,62,100]. Especially for those reporting high

PCEs, spin-coated [98] or thermal evaporated [58,59] charge transport layers are still required.

To the best of our knowledge, the PCE of our all electrospray printed champion cell is the highest

for PSCs with all three functional layers fabricated using scalable methods in air and at moderate

temperature (up to 150 C), as shown in Table 2. Zheng et al. reported a blow-drying method to

fabricate mesoporous TiO2, methylammonium lead halide (CH3NH3PbI3) perovskite and Spiro-

OMeTAD layers [84], however, the processing conditions are more demanding, such as N2

63

environment and high-temperature annealing (450 C) for the TiO2 layer. Recently Zuo et al. [8]

demonstrated one-step roll-to-roll air processed PSCs on a flexible substrate achieving a

respectable PCE of 11.16%. They also fabricated devices on a glass substrate with only the

perovskite slot die coated, and the corresponding PCE is 15.57%, which is comparable with the

15.9% PCE of the device with only perovskite layer printed by electrospray in this work (Table 1).

Table 3-2 Overview of PCE of the reported PSCs with all three layers (ETL, perovskite,

and HTL) fabricated using scalable methods.

Scalable Method Special conditions* Best PCE Reference

Infiltration \ 12.8% [101]

Brush printing \ 9.1% [102]

Slot-die \ 12.6% [103]

Doctor blade Humidity controlled (15%-25%) 10.7% [100]

Slot-die \ 12.0% [51]

Slot-die \ 14.7% [62]

Screen-printing High-temperature (500 C) 13.3% [104]

Blow drying High-temperature (450 C), glove box 17% [84]

Electrospray printing \ 15.0% this work

* The special conditions needed including humidity control, environment control, and temperature. The

omission of a certain condition means this condition is not required during the preparation of PSCs.

It is insightful to compare the electrospray printing process to doctor blading, which is

highly successful in making micrometer-sized perovskite crystals and reaching one of the highest

PCEs to date for PSCs. Electrospray printing generates polycrystalline film with a characteristic

64

grain size of only ~300 nm, which is comparable to one-step spin-coating but much smaller than

the state-of-art of doctor blading [58]. Interestingly, more than two orders of magnitude increase

in grain boundary density did not dramatically deteriorate the PCE (15.0% for all-electrospray

printed, 15.9% for perovskite layer electrospray printed only vs 20% for best reported doctor

blading data), signifying one important and desirable trait of perovskite, which is the strong defect

tolerance [105].

In terms of printing speed, doctor blading has an optimal substrate moving speed of ~10

µm/s in the most commonly operated evaporation region. Recently the speed of ~50 mm/s has

been reached in the Landau-Levich region by adding surfactant to suppress surface flow instability

[59]. The printing speed for the electrospray is ~10 mm/s, which is much faster than the doctor

blading operated in the evaporating region and comparable with that in the Landau-Levich region.

The throughput of electrospray printing can be dramatically scaled-up because large arrays of 91

to 331 emitters have been demonstrated [72]. Furthermore, the electrospray printing may

complement the doctor blading when the substrate has relatively high surface roughness or is non-

flat along the direction that is perpendicular to the substrate moving direction.

3.3 Conclusion

We have succeeded in fabricating efficient PSCs devices using electrospray to print all

three layers (ETL, perovskite, and HTL) in air and below 150 C. Choosing short working

distances and solvents with low vapor pressure such as GBL, NMP and DCB ensures that the

droplet evaporation time is longer than the droplet flying time, which enables wet film deposition.

Such electrospray printing process results in pin-hole free, homogeneous and smooth perovskite

65

films. The results demonstrate that electrospray is able to print uniform functional films from sub-

100 nm to 500 nm range, making it a powerful tool to coat or pattern functional layers of perovskite

optoelectronic devices. The PCE of the all-printed devices reached up to 15.0%. Our results

demonstrate that the electrospray printed PSCs can provide performance on par with spin-coated

cells in terms of active layer morphology and overall device performance. In addition, electrospray

printing offers benefits of roll-to-roll compatibility and nearly zero material waste. More

importantly, electrospray can be massively multiplexed [106], providing a feasible route for large-

scale manufacturing of PSCs and paving the way for roll-to-roll printing on flexible electrodes.

3.4 Experimental section

CH3(NH)2I, CH3NH3Br powders and PbI2 were added in a mixture of γ-Butyrolactone

(GBL) and1-Methyl-2-pyrrolidinone (NMP) (7:3 v/v) to make 1 M FA0.85MA0.15PbI2.85Br0.15

solution. All chemicals are purchased from Sigma-Aldrich and used as received without further

purification. Mix Ti-Nanoxide T-L paste (Solaronix) with DI water and tert-butyl alcohol,

followed by stirring for 2 h and ultrasonic dispersing for 30 min. Dissolve 36 mg of Spiro-

OMeTAD in 1 ml 1,2-Dichlorobenzene, with addition of 23 μL Li-TFSI/acetonitrile (170 mg/mL),

75 μL of [tris(2-(1H-pyr- azol-1-yl)-4-tert-butylpyridine) cobalt (III) bis (trifluoro-

methylsulphonyl) imide] (FK209) /acetonitrile (100 mg/mL) and 10 μL of 4-tert-butylpyridine

(TBP). Before electrospray printing, the Spiro-OMeTAD solution was diluted to 7.2 mg/ml. The

glass/FTO substrate was successively washed with Hellmanex III, acetone, ethanol, and DI water.

The sheet resistance of FTO is 12-14 ohm/sq and the thicknesses of glass and FTO are 2.2 mm and

200 nm, respectively. The average transmittance of glass/FTO in the visible region is around 82-

84.5 %. The electrospray printing process took place in a fume hood in the ambient atmosphere.

66

Figure 3-1 illustrates the electrospray printing apparatus, which consists of an electrospray emitter

on a syringe driven by a syringe pump, a high voltage DC power supply, a cone-jet visualization

subsystem, and a substrate motion control subsystem. Prior to printing, the prepared solution was

loaded into a syringe and the flow rate was controlled by the syringe pump. The substrate and the

syringe were mounted on a computer controlled motorized linear stage and a serpentine path

motion of electrospray nozzle relative to substrate as shown in Figure 3-6 was designed. TiO2,

perovskite, and Spiro-OMeTAD film were printed sequentially on the pre-cleaned FTO glass.

Table S1 summarizes the parameters such as working distance, flow rate and printing speed for

printing TiO2, perovskite, and Spiro-OMeTAD. The TiO2 layer was printed first, and the sample

was annealed at 150 °C for 30 min. The perovskite film was printed on the annealed TiO2 layer,

and after the wet perovskite precursor film was printed, the sample underwent flash vacuum drying

in a vacuum chamber which induces rapid sample drying within a few seconds. The sample with

dry perovskite film was annealed on a hotplate at 150 °C for 15 min. Next, the Spiro-OMeTAD

layer was printed on top of the perovskite film. Similarly, the sample with wet Spiro-OMeTAD

was dried by flash vacuum. Finally, an 80 nm thick gold layer was evaporated as the top electrode.

The active area of each device was 0.1 cm2.

X-ray diffraction (XRD) analyses were performed at a scanning rate of 5 ˚/min on an X-

ray diffractometer (Philips Xpert Pro). Scanning electron microscopy images were obtained from

scanning electron microscope (Zeiss 1550) operated at an accelerating voltage of 5 kV.

Photovoltaic performance of the solar cells was analyzed under one sun (AM 1.5 G, 100 mW/cm2)

illumination with a solar simulator (150W Sol 2ATM, Oriel), and the current-voltage

67

characteristics of each cell were recorded with a digital source meter (Keithley 2400). AFM images

were obtained from atomic force microscope (Parker XE7) in the dark.

3.5 Supporting Information

Figure 3-18 Top-view SEM images of spin-coated perovskite films.

Figure 3-19 Typical JV curves of the device using perovskite ES printed at different flow

rates and spin coated. The HTL and ETL are spin coated for all devices.

68

0.0 0.2 0.4 0.6 0.8 1.00

5

10

15

20

Curr

ent density (

mA

/cm

2)

Voltage (V)

4 mm

3 mm

2 mm

SC

Figure 3-20 Typical JV curves of the device using TiO2 ES printed at working distances of

4, 3, 2 mm and TiO2 spin coated. The perovskite film is ES printed and HTL is spin coated

for all devices.

Figure 3-21 Typical JV curves of the device using Spiro-MeOTAD ES printed with

chlorobenzene and 1,2-dichlorobenzene as the solvent and Spiro-MeOTAD spin-coated with

chlorobenzene and 1,2-dichlorobenzene. The perovskite film and TiO2 are both ES printed

for all devices.

69

Figure 3-22 The stability of the all electropsray printed device without encapsulation in

ambient conditions with a relative humidity of ~20% at room temperature.

70

Supplementary Note 1:

The values used to estimate the droplet size 𝒅𝟎 for equation (4), evaporation rate 𝑲𝒆 for

equation (3) and the droplet flight time te for equation (1) are as followed.

Property value unit

Dielectric constant (𝜅)* 39 /

Vacuum permittivity (𝜖0) 8.85E(-12) F/m

Liquid flow rate (𝑄) 0.7 µL/min

Liquid electrical conductivity (𝛾)** 4.3 S/m

Mass diffusivity (𝐷𝑑𝑖𝑓𝑓)*** 0.1 cm2/s

Solvent vapor density (𝜌𝑔) 3.675E(-3) g/ml

Liquid density (𝜌)* 1.13 g/ml

Vapor pressure of the solvent (𝑃𝑣)* 200 Pa

Ambient pressure (𝑃0) 100 000 Pa

Surface tension (𝜎)* 44.6E(-3) N/m

Molar volume of the liquid (𝑣𝑚)* 76.28E(-6) m3

Universal gas constant (𝑅) 8.314 /

Temperature (T) 298 K

*The liquid properties were estimated using the corresponding properties of GBL.

** Conductivity of the liquid was measured with a piece of homemade equipment.

*** Estimated value.

71

Chapter 4 Hole-conductor-free perovskite solar cells based on

electrospray printed carbon electrode

4.1 Introduction

Interfacial deterioration by chemical reaction between the perovskite layer and the metal

electrode (e.g., Ag, Al) has been reported to be one of biggest the origins of the initial degradation

of PSCs [107][108][107,108]. Even though less active noble metals such as gold are chosen as

counter electrodes for the majority of the state-of-the-art perovskite solar cell devices, the chemical

reaction between perovskite and Au at the interface is still observed. Apart from that, the prices of

those noble metal electrode materials are very high and the preparation process usually requires

high-vacuum thermal evaporation. Therefore, in order to commercialize PSCs, it is important to

find alternative cheap and perovskite-inert electrode materials that also can be prepared using a

scalable method for high-efficiency PSCs. Carbon-based materials have risen as a promising

candidate to replace noble metal electrodes for PSCs. They have suitable Femi level (5.0 eV) and

are inert to perovskite and water resistant [109,110]. At the same time, they have low prices and

can be prepared using solution-based methods.

Acording to the way the carbon film incorporated into the PSC structure, the carbon-based

PSCs (C-PSCs) reported so far can be roughly generalized into three different categories including

meso C-PSCs, assembled C-PSCs, and paintable C-PSCs [111]. Meso C-PSCs (Figure 4-1b)

[104,112,113] are prepared by infiltrating perovskite solution into the pre-deposited triple-layer

architecture including mesoporous ETL, space layer, and carbon electrode. A mixed structure of

perovskite and carbon is formed from this method. The carbon layer for meso C-PSCs is prepared

72

by doctor-blade coating or screen printing using a carbon paste containing graphite and carbon

black, followed by annealing at high temperature (e.g. 400 °C). The PCE of the meso C-PSCs

reaches over 12.8% [112]. By incoporating hole electron material such as Co3O4 [104] or single

walled carbon nanotubes [114], the PCE were increase to 13.27% and 14.7% respectively. The

limitation of this method is that the high-temperature annealing (~ 400 °C) of carbon film and the

mesoporous ETL and space layer, which is not compatible with flexible substrates. And the

filtration of the perovskite into the carbon-containing mesoporous structure to form a mixture is a

different mechanism for preparing perovskite from the planar structure. Methods to optimize the

perovskite film quality to improve the efficiency of mesoporous C-PSC is still needed.

Figure 4-1 Details of the three types of C-PSCs. a) Energy level and charge transfer

behavior in C-PSCs. b) Meso C-PSCs developed by Ku et al. [110]. c) Assembled C-PSCs

73

developed by Wei et al. [115]. d) Paintable C-PSCs developed by Wei et al. and other

groups [115–117].

The carbon electrode for assembled C-PSCs (Figure 4-1c) is by transferring a pre-

deposited carbon layer onto the perovskite layer or PbI2 layer and then convert it to perovskite

[118,119]. It is a another way to solve the conflict between the high-temperature annealing of

carbon and perovskite cannot sustain such high temperature. Instead of direct deposit carbon on

perovskite, assembled C-PSCs separate the preparation of perovskite and carbon and then transfer

the already high-temperature annealed carbon onto the perovskite layer or PbI2 layer. One key

issue with the assembled PSCs is the interface between the perovskite layer and carbon layer, even

though methods like hot press was applied to tighten the contact between the perovskite layer and

carbon layer, the efficiency of the embedment PSCs still is relatively low [123].

One of the breakthrough in C-PSCs is the report of low-temperature processible carbon

film[122,123], the required annealing temperature of carbon decreased from 400 C to 100 C.

Simple planar structure is feasible. The commercially available carbon material is in paste or paint

form (the slurry usually contains graphite, carbon black, and polymer binders). It can be directly

coated on a perovskite layer using either a painting process or a doctor blade technique, followed

by annealing at low temperature (e.g. 100 °C). The highest efficiency of the low temperature

carbon electrode based planar structure reaches 15.38% [124] by using a thermally evaporated C60

film as ETL. Zhang et al. reported a solvent-exchange method to make a self-adhesive carbon film

and based on a assembled structure, the PSC achieved an efficiency of 19.2% [125], owing to the

improved interface and the incorporation of HTL Spiro-OMeTAD.

74

The device with highest PCE (19.2%) reported so far is based on a high temperature

processed mesoporous ETL and the high price Spiro-OMeTAD HTL [125]. The PCE of HTL free

C-PSCs processed at low-temperature still lags far behind that of Au-based PSCs. Table 4-1 shows

some representative C-PSCs reported in literature. The highest PCE for HTL free C-PSCs are

reported by Han et al.[126] with a power conversion efficiency of 15.6%, using a carbon

black/graphite electrode in a planar structure. In general, the unsatisfactory performance of HTM-

free C-PSCs is related to [111,119]: 1) relative higher resistance of carbon electrode compared to

the noble metal electrode; 2) mismatch of the carbon electrode fermi level and the Valence band

of perovskite; 3) insufficient charge separation; 4) poor interface contact between carbon electrode

and perovskite layer. The poor interface contact between the carbon electrode and the underlying

layer dominates the performance loss of the reported carbon-based PSCs.

Herein, we applied electrospray printing to prepare the C-PSCs. A tunable carbon ink is

used for the printing. The interface between the perovskite and the carbon could be controlled by

the ES-printing parameters such as working distance and substrate temperature. By optimizing the

printing parameters, the interface between the perovskite layer and carbon layer is improved

compared with that from doctor-blade coating which is the prevailing method to prepared carbon.

Consequently, the performance of the corresponding C-PSCs was improved. A PCE of 14.41%

was achieved for the HTL free C-PSCs prepared using electrospray printing at low temperature.

75

Table 4-1 Representative C-PSCs from literature.

Device structure Carbon

incorporation

form

HTL High

temperature

PCE Reference

c-TiO2/m-TiO2/ZrO2/MAPbI3/carbon Mesoporous free Y 12.8% [112]

c- TiO2/m-TiO2/ ZrO2/MAPbI3/

Co3O4/carbon

Mesoporous Co3O4 Y 13.27% [104]

c- TiO2/m-TiO2/Al2O3/ MAPbI3/

SWCNTs/carbon

Mesoporous SWCNTs Y 14.7% [127]

c- TiO2/m-TiO2/MAPbI3/carbon Assembled free Y 11.02% [118]

c- TiO2/m-TiO2/MAPbI3/spiro-

MeOTAD/carbon

Assembled Spiro-

MeOTAD

Y 19.2% [125]

C60/MAPbI3/Carbon Planar free Y 15.38% [122]

TiO2/MAPbI3/Carbon Planar free Y 12.06% [122]

c-TiO2/m-TiO2/FA0.8Cs0.2PbI2.64Br0.36

/PEO/carbon

Planar free Y 14.9% [128]

c-TiO2/m-TiO2/MAPbI3/carbon Planar free Y 13.5% [129]


4.2.1 The Stable operation of the electrospray of carbon

The carbon dispersion for electrospray was prepared by re-disperse the dried carbon paste

chunk into chlorobenzene and was ultrasonically mixed before use. The electrospray of the pristine

carbon dispersion was difficult. Due to the low conductivity of the carbon ink itself, no stable

Taylor cone could be formed as shown in the left image in Figure 4-2. This is in contrast to the

perovskite solution which the conductivity of solution is too high to electrospray because of the

ionic nature of the perovskite precursor solution. In carbon ink, there is almost no ions and as a

76

result the conductivity is low. To increase the conductivity, protonic solvents like acetic acid is

usually added in electrospray practice. But compatibility with perovskite is another issue we have

to take into consideration here, because acetic acid will react with perovskite and carbon is directly

printed on top of perovskite and usually the carbon ink droplet from electrospray will still more or

less contain certain amount of remaining solvent. So acetic acid is excluded in this case. We found

adding a small amount of Li-TFSI salt into the carbon ink makes the electrospray much easier.

Very stable Taylor-cone could be formed (right image in Figure 4-2).

Figure 4-2 Electrospray of carbon ink without (left) and with the additive (right).

The same serpentine path as described in chapter 3 that has been used to print perovskite,

ETL, and HTL was also used here. To obtain carbon films with good conductivity, the thickness

of carbon films usually need to be over 20 microns, which is much thicker than other component

layers in the PSCs. This high thickness requires higher flow rate printing and longer printing time.

The flow rate of electrospray of the carbon ink can be increased to 15 µL/min owing to the low

77

conductivity of the carbon ink. The printing parameters for carbon electrode is as shown in Table

4-1.

Table 4-2 Typical ES printing parameters of carbon ink.

Working distance (mm) 1.5

Flow rate (µL/min) 15

Offset (mm) 0.2

Printing speed (mm/min) 300

4.2.2 Characterization of electrospray printed carbon film

The crystalline structure of perovskite films with and without carbon film electrospray

printed or doctor-blade coated on are examined by XRD. As shown in Figure 4-3, all the peaks of

perovskite are preserved after the electrospray printing or doctor-blade coating of carbon on top,

proving that the printing of carbon did not damage the underlying perovskite layer. And there is a

strong extra peak observed in the sample with carbon films, which is ascribed to the crystalline

graphite in the carbon film. The doctor-blade coated carbon film and the electrospray printed

carbon film has identical XRD peaks.

SEM was used to observe the microstructure of the electrospray printed carbon film. As

shown in Figure 4-4, there were two distinct structures in the carbon film. The sheet-like phase

was believed to be graphite and the small particle phase is carbon black. Among these two phases,

graphite is the one that contribute more to the conductivity of the carbon film.

78

5 10 15 20 25 30 35 40

perovskite/Carbon-ESperovskite

perovskite

graphite

FTO

Inte

nsity (

a.u

.)

2 Theta (degree)

perovskite

perovskite/Carbon-ES

perovskite/Carbon-Doctor blade

Figure 4-3 XRD patterns of perovskite film, perovskite/carbon-ES and perovskite/Carbon-

doctor blade.

Figure 4-4 Top-view SEM image of the doctor blade coated (left) and electrospray printed

(right) carbon film.

79

4.2.3 Working distance and interface

One of the important electrospray parameter that greatly influences the quality of the

resulting carbon film is the working distance between the nozzle and substrate. From the discussion

in Chapter 3, we have known that working distance and the solvent vapor pressure are two factors

that influence the Da number, namely the degree of evaporation of the droplets and whether they

will still be wet or not when they reach the substrate. The flow rate of printing of carbon is large

(~ 15 µl/min), the solvent used was chlorobenzene which has high vapor pressure. Here we tune

the Da number by changing the working distance. A series carbon films were printed at different

working distance to optimize the quality of the resulting film and the interface between carbon and

the underlying perovskite layer.

Table 4-3 Sheet resistance of carbon films prepared by doctor blade coating or electrospray

printing at different working distance.

Carbon film (~30 µm) Sheet resistance (Ω/sq)

Doctor-blade coating 18

Electrospray printing-1.5 cm 20

Electrospray printing -2 cm 21

Electrospray printing -1 cm 22

80

The sheet resistance of carbon films printed at different working distance were measured and

recorded in Table 4-2, which showed very close values for carbon films printed at various working

distances and the carbon film that is doctor-blade coated when the film thickness is close.

SEM was used to detect the differences of these carbon films from the microstructure

perspective. As shown in cross-section image of the interface between the perovskite and carbon

in Figure 4-5, when working distance is 2 cm, the electrospray printed carbon film has many voids

in the interface and the film itself is more loosely packed. And when the working distance

decreased to 1.5 cm, the carbon film itself becomes more compact and more importantly the

interface between the carbon and perovskite becomes seamless. The interface between the doctor-

blade coated carbon and perovskite was also observed as a reference. As shown in Figure 4-5,

voids also present in the interface between perovskite and doctor-blade coated carbon film. The

seamless interface between perovskite and electrospray printed carbon at 1.5 cm is presumably

owing to the decreased working distance for electrospray. Working distance influences the degree

of wetness of droplets when reach the substrate. Lower working distance means wetter droplet. If

the carbon ink droplets are still wet when reach the substrate and join the pre-deposited part that

has not fully dried, the interface between the carbon and the underlying perovskite will be better.

But further lowering this working distance resulted in overflow of carbon ink on the substrate

because of the influx of solvent is larger than the outflux of solvent due to evaporation, so more

and more carbon ink was accumulated to finally form puddle of carbon ink which is not good for

forming uniform films.

81

Figure 4-5 Cross-section SEM images of carbon films printed at different working distance

and the carbon film that was doctor blade coated.

4.2.4 Hole-conductor-free PSCs based on carbon electrode

The interface did not influence much on the conductivity of the carbon film as shown in

table 4-2. To explore its influence on the performance of the final PSC, a series PSCs use

different carbon films as electrode were prepared. Carbon-based PSCs prepared here are hole

transport layer free and the whole architecture is as shown in Figure 4-6.

82

Figure 4-6 Diagram of the architecture (a) and energy diagram (b) of carbon electrode

based hole-conductor-free PSCs.

JV curves of the carbon electrode based HTL free C-PSCs under illumination were

measured and the results are shown in Figure 4-7. The photovoltaic metrics extracted from the JV

curves are recorded in Table 4-3. Interestingly, the perovskite solar cell device based on carbon

electrode electrospray printed at a working distance of 1.5 cm (denoted as Carbon-ES-1.5) showed

a PCE of 14.41 %, higher than that based on the carbon electrode that was doctor-blade coated

(denoted as Carbon-DC). This higher PCE mainly comes from the higher current density and fill

factor. For Carbon-ES-1.5 the current density reaches 24.35 mA/cm2, higher than 21.79 mA/cm2

of the Carbon-DC. Fill factor was also increased from 52.3% to 57.7% when the doctor blade

coated carbon electrode was replaced with the electrospray printed carbon electrode at optimal

working distance. What’s also worth noting is that carbon-ES-1.5 only showed slight PCE decrease

compared to the Gold based PSC (denoted as Au), which has a HTL between the gold electrode

and perovskite. All these results show that the carbon film printed at optimal working distance is

a very good alternative electrode for perovskite solar cell replacing gold.

83

0.00 0.25 0.50 0.75 1.000

5

10

15

20

25

C-ES-1.5

C-DC

Au

Voltage (V)

Curr

ent D

ensity (

mA

/cm

2)

Figure 4-7 JV curves of hole-conductor-free PSCs based on different carbon electrodes.

Table 4-4 Photovoltaic metrics of the hole conductor free C-PSCs based on electrospray

printed carbon (C-ES-1.5) and based on doctor blade coated carbon (C-DC) and PSCs

based on gold electrode.

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

C-ES-1.5 1.03 24.35 57.7 14.41

C-DC 0.98 21.79 52.3 11.14

Au 1.00 23.73 63.4 15.10

To further explore the charge transport behavior at the interface between the perovskite

and carbon, steady state and transient photoluminescence were measured. From the steady state

84

photoluminescence spectroscopy (Figure 4-8), the perovskite with carbon film printed by

electrospray at optimal working distance 1.5 cm (denoted as PVK/carbon-ES) has lower

photoluminescence peak. And from the transient photoluminescence shown in Figure 4-9, the

PVK/carbon-ES shows faster photoluminescence decay than that of PVK/Carbon-DC. Both results

indicate that Carbon-ES has strong charge extraction than Carbon-DC, which presumably owing

to the seamless interface between the perovskite layer and carbon, and contribute the higher PCE

of the perovskite solar cell devices based on Carbon-ES electrode.

Figure 4-8 The steady state photoluminescence of perovskite and carbon film structure.

85

Figure 4-9 Time-resolved photoluminescence lifetimes for perovskite/carbon films on glass

slides.

4.3 Conclusion

Carbon electrode was printed using electrospray for HTL free PSCs. The sheet resistance

of the printed carbon film is similar to that of doctor blade coated carbon film. By carefully

choosing printing parameters, specifically the working distance, to allow the good interact of the

carbon ink droplet, the interface between carbon and perovskite layer could be greatly improved.

Voids observed in doctor blade coated carbon film could be eliminated and the interface is almost

seamless. Stronger and faster PL quenching from perovskite to electrospray printed carbon

indicated enhanced hole extraction. Due to these advancements, the hole free PSCs based on this

electrospray printed carbon electrode showed a PCE of 14.41%, which is higher than the doctor

blade coated carbon film based one and is only slight lower than the gold based perovskite solar

86

cell with hole transport layer. Our results point to new opportunities for the scalable fabrication of

cost-efficient and flexible PSCs toward market deployment.

87

Chapter 5 Laser annealing for the scalable fabrication of

perovskite films

5.1 Motivation

In the previous chapters, we have successfully demonstrated the scalable fabrication of

various layers that compromising the PSCs using electrospray printing. There is still one key

process that hindering the large-scale fabrication of PSCs. Almost every layer requires thermal

annealing at temperature over 150 C for about half an hour or longer. The time-consuming

conventional thermal annealing complicates the fabrication process and is not suitable for

continuous fabrication. 150 C is also not compatible with flexible substrates such as PET. Laser

annealing stands out as a promising candidate to overcome all these obstacles for its advantages

of compatibility to continuous roll-to-roll printing, minimal influence to non-radiated surrounding

area, and rapid processing. It can be integrated with the electrospray process to realize the

continuous fabrication of perovskite film as shown in the Figure 5-1.

High energy excimer laser in UV wavelength regime has been applied in material

processing such as assisting the phase transition of amorphous silicon to crystalline silicon, or

amorphous metal oxides into highly crystalline structures [130]. However, perovskite materials

that contains organic component cannot sustain this kind of high energy UV radiation [131]. Apart

from that, UV laser can also interact with transparent conductive oxides like indium tin oxides

(ITO) and fluorine-doped tin oxide (FTO) which are used as electrode in PSCs. Therefore, we used

a modified 10W RGL-FM Series Fiber Laser Marker as the laser source, which is a low energy

NIR laser. The galvanometer system enhanced the precision of the laser, which enabled us to freely

88

change laser scanning parameters. The parameters of modified GRL-FM fiber laser marker are as

shown in the Table 5-1. In this chapter, we aim to develop a laser assisted annealing process for

high quality perovskite films.

Figure 5-1 Diagram for the concept of roll-to-roll fabrication of PSCs with integration of

electrospray printing and laser annealing.

Table 5-1 Parameters of the modified RGL-FM fiber laser marker.

Parameters Range

Scan speed 0-6000 mm/s

Average power 10 W

Peak power 2.5 KW – 10 KW

Frequency 20 KHz – 80 KHz

Focus diameter 12 µm

Energy per pulse 0.5 mJ

89


The freshly printed perovskite film was first vacuum flash dried and then was placed under

the laser source and the laser beam laterally scanned on the film surface and the transparent

perovskite precursor film gradually turned into dark brown, indicating the crystallization of the

perovskite film, which only takes around 2 minutes for a 20 × 20 mm film. Power density and

scan pattern are the key factors affecting the laser annealing result. Power density of laser pulse is

determined by factors such as energy per pulse, frequency of laser pulse, power percentage, laser

spot diameter, and scan speed as described in equation 5-1. The scan pattern we used is as shown

in Figure 5-2. The distance between two parallel scans is 6 microns. The energy per pulse is already

determined by the built-in laser. All other factors could be tuned to change the overall laser power

density.

Power density = 𝑒𝑛𝑒𝑟𝑔𝑦 𝑝𝑒𝑟 𝑝𝑢𝑙𝑠𝑒∗𝐹𝑟𝑒𝑞𝑢𝑒𝑛𝑐𝑦∗𝑝𝑜𝑤𝑒𝑟 𝑝𝑒𝑟𝑐𝑒𝑛𝑡𝑎𝑔𝑒

𝐿𝑎𝑠𝑜𝑟 𝑠𝑝𝑜𝑡 𝑑𝑖𝑎𝑚𝑒𝑡𝑒𝑟∗𝑠𝑐𝑎𝑛 𝑠𝑝𝑒𝑒𝑑 (5-1)

Figure 5-2 Scanning pattern of the laser.

90

The focus diameter of the laser is very small with diameter of only 12 microns, as a result,

even with very high scan speed (2000 mm/s) and low power percentage (0.05%), the perovskite

film got etched/burned away and the substrate get exposed as shown in Figure 5-3.

Figure 5-3 Optical microscope image of perovskite film annealed using laser at focal plane,

with scan speed of 2000 mm/s and power percentage of 0.05%.

Therefore, the stage was moved a bit away from the focal plane to enlarge the laser spot

and even out the energy. Here we chose a distance of 4 mm from the focal plane and the spot size

was increased to 200 microns. Here, the power density was tuned mainly by changing the power

percentage. Four different combinations of laser scanning parameters as shown in Table 5-2 were

tried. The optical image of the films treated with different laser densities were shown in Figure

5-4. When the laser power density was 0.337 or 0.253 W/cm2, very obvious scalding of perovskite

film was observed. When the laser power was decrease to 0.168 W/cm2, only slight laser scanning

marks could be seen from the optical image. And when the laser power was further decreased to

0.084 W/cm2, no visible mark could be observed from the optical image any more.

91

Table 5-2 Laser annealing parameters.

Sample Power percentage

(%)

Scan speed

(mm/s)

Power density

(w/cm2)

Damage to perovskite

1 4 600 0.337 Yes

2 3 600 0.253 Yes

3 2 600 0.168 Yes

4 1 600 0.084 No

XRD was also used to characterize the perovskite film after treated with different laser

power densities. As shown in Figure 5-5, all samples showed the typical peaks of the light-absorbing

perovskite phase. For sample 2 (0.253 W/cm2), obvious peak of PbI2 was observed indicating the

power density was too strong and the perovskite partly decomposed into PbI2. For sample 3 and 4,

the perovskite peaks coincided with that of perovskite film annealed by thermal heating, and no

PbI2 could be observed, indicating the rapid laser annealing (less than 2 min) is enough to finish

the perovskite crystallization just like the long time thermal annealing (around 15 min).

Figure 5-4 Optical images of perovskite films laser annealed at different power density.

92

5 10 15 20 25 30 35 40

0.253 mW/cm2

0.168 mW/cm2

150 oC for 15 min

PbI2

Perovskite

FTO

0.084 mW/cm2

Inte

nsity

2 Theta (degree)

Figure 5-5 X-ray diffraction (XRD) pattern of laser-crystallized perovskite films at

different laser power densities and perovskite film that was thermal annealed.

SEM images were used to compare the microstructure of perovskite film annealed by laser

scanning at optimal power density and by conventional thermal annealing and shown in Figure

5-6. The insets are the image of the corresponding perovskite film taken by camera. Both laser

annealed and thermal annealed perovskite films are of dark brown. SEM images show both films

were compactly packed films without any pin-holes which is desirable for perovskite solar cell

devices. AFM topographies (Figure 5-7) also showed similar results, the perovskite film by

thermal annealing and laser annealing has similar structure and roughness, 6 nm and 7 nm

respectively.

93

Figure 5-6 SEM images of perovskite films annealed by thermal heating (left) and laser

(right).

Figure 5-7 AFM images of perovskite films annealed by thermal heating (left) and laser

(right).

94

5.3 Conclusion

In summary, laser assisted annealing and crystallization of organic−inorganic hybrid

perovskite films has been successfully demonstrated. Highly crystalline and pin-hole free

perovskite film could be obtained using low energy NIR laser beam at appropriate laser power

density. This rapid laser annealing process could be integrated with the electrospray printing

suggesting a direct compatibility to practical roll-to-roll printing. Laser annealing could be further

scaled for large-area parallel processing using slit beams or beam arrays.

95

Chapter 6 Summary and outlook

Summary

The goal of this dissertation is to develop a scalable method for fabrication of high

efficiency HPSCs in the ambient environment and low temperature. The focus was on developing

electrospray printing process to provide capability for synthesizing fully printed cells. Electrospray

printing was chosen for its roll-to-roll compatibility, high-material usage rate, and patterning

capability.

The first part of this dissertation focused on the electrospray printing process of perovskite

films, including ensuring the stable operation of the electrospray in the Taylor cone-jet mode,

printing uniform wet thin film, and rapid drying of wet precursor film to obtain high-quality

perovskite film. The important findings on electrospray printing of thin films include:

In order to obtain stable electrospray of perovskite precursor solution, strategies

such as small-sized silica nozzle and anti-wetting coating on the outer surface of

the nozzle were developed.

During the continuous printing of wet films, the micro-flow induced by uneven

evaporation of the solvent is the cause of non-uniformity of the printed film.

Vacuum flashing was discovered as an effective method to change the substrate

surface property and suppress the uneven evaporation of solvent, and hence print

uniform films.

Working distance and solvent also greatly influence the morphology of the printed

film. Working distance decides the droplet flight time before reaching the substrate.

96

Vapor pressure of solvents affect the droplet evaporation time before the droplet

totally dries out. The ratio of these two quantities (Damkhöler number Da)

determines the degree of wetness of the droplets upon reaching the substrate. To

obtain smooth and pin-hole free films, wet droplets instead of dried particles are

needed, namely, Da has to be less than one.

These findings were next applied towards the electrospray printing of TiO2 (ETL), Spiro-

OMeTAD (HTL) and carbon (top electrode). All electrospray printed devices were successfully

developed, achieving efficiency of over 15%, which is among the highest to date for fully printed

PSCs.

Electrospray printed carbon film was used as electrode replacing the pricy gold film

electrode. In literature, the most commonly used method to prepare carbon film is doctor-blade

coating or screen printing. The HPSCs based on electrospray printed carbon showed improved

efficiency compared to that based on doctor-blade coated carbon, owing to the improved interface

between the carbon and underlying perovskite. High quality interface is achieved by carefully

tuning the electrospray printing working distance. Electrospray printed carbon-based HPSCs

showed only slight decrease of efficiency compared with that based on gold electrode. Considering

the low cost of carbon, solution processing capability, and the hole transport layer free cell

architecture, these results are of great significance for the scalable fabrication of PSCs.

In last part of this dissertation, laser annealing of perovskite film was developed to replace

the conventional thermal annealing, in an effort to push the scalability of the fabrication of PSCs.

Optimizing the power density and scanning pattern of the laser annealing, highly crystalline and

97

pin-hole free perovskite film were obtained. Laser annealed perovskite film is like the thermally

annealed film in terms of crystal phase as determined using XRD, microstructure obtained from

SEM, and surface topography observed using AFM. Laser annealing offers the promise of direct

roll-to-roll compatibility of PSC fabrication.

Outlook

Based on the results obtained in this disseration, there are related areas of research that can

be further investigated. Several suggestions are given below.

The morphology and crystal structure characterization of the laser annealed perovskite film

has been performed and shown to be similar to that of the thermal annealed one. Next, PSCs based

on the laser annealed perovskite films needs to be fabricated and compared. Laser annealing can

be further generalized for the fabrication of other layers in the perovskite solar cell. A more

thorough study on laser annealing process is needed in order to improve the perovskite film

properties such as by increasing the crystallite size.

Vacuum flash drying is an effective and rapid method for drying wet perovskite precursor

film. However, it presents some challenges when it comes to scalable fabrication. Replacement for

vacuum flash drying step is needed which could be achieved by formulating the perovskite

precursor solution. The ideal case would be when freshly printed film dries automatically and

rapidly. Patterning capability is one of the unique advantages of electrospray printing. Exploration

of the patterning step for PSCs or other photo-electronic devices would be another relevant step to

design unique geometries. Increase in the throughput of electrospray printing could be obtained by

developing multi-nozzle systems. This process remains to be developed.

98

References

[1] V. Masson-Delmotte, H.-O. Pörtner, J. Skea, IPCC report Global Warming of 1.5 C, 2018.

[2] F. Muhammad-sukki, A. Bakar, R. Ramirez-iniguez, S.H. Abu-bakar, S. Hajar, M. Yasin,

S.G. Mcmeekin, B.G. Stewart, Solar photovoltaic in Malaysia : The way forward, Renew.

Sustain. Energy Rev. 16 (2012) 5232–5244. doi:10.1016/j.rser.2012.05.002.

[3] J. Tsao, N. Lewis, G. Crabtree, Solar FAQs, US Dep. Energy. (2006) 1–24.

http://www.sandia.gov/~jytsao/Solar FAQs.pdf.

[4] T. Ibn-Mohammed, S.C.L. Koh, I.M. Reaney, A. Acquaye, G. Schileo, K.B. Mustapha, R.

Greenough, Perovskite solar cells: An integrated hybrid lifecycle assessment and review in

comparison with other photovoltaic technologies, Renew. Sustain. Energy Rev. 80 (2017)

1321–1344. doi:10.1016/j.rser.2017.05.095.

[5] NREL, Best Research-Cell Efficiencies Chart, Best Res. Effic. Chart. (2018) 2020.

[6] A. Kojima, K. Teshima, Y. Shirai, T. Miyasaka, Organometal Halide Perovskites as Visible-

Light Sensitizers for Photovoltaic Cells, J. Am. Chem. Soc. 131 (2009) 6050–6051.

doi:10.1021/ja809598r.

[7] J.-P. Correa-Baena, A. Abate, M. Saliba, W. Tress, T. Jesper Jacobsson, M. Grätzel, A.

Hagfeldt, The rapid evolution of highly efficient perovskite solar cells, Energy Environ. Sci.

10 (2017) 710–727. doi:10.1039/C6EE03397K.

[8] Anthony, J.W., Bideaux, R.A., Bladh, K.W. and Nichols, M.C., 2011. The handbook of

99

mineralogy. Mineralogical Society of America, Chantilly, VA.

[9] M.A. Green, A. Ho-Baillie, H.J. Snaith, The emergence of perovskite solar cells, Nat.

Photonics. 8 (2014) 506. http://dx.doi.org/10.1038/nphoton.2014.134.

[10] J. Huang, Y. Shao, Q. Dong, Organometal Trihalide Perovskite Single Crystals: A Next

Wave of Materials for 25% Efficiency Photovoltaics and Applications Beyond?, J. Phys.

Chem. Lett. 6 (2015) 3218–3227. doi:10.1021/acs.jpclett.5b01419.

[11] G. Xing, N. Mathews, S. Sun, S.S. Lim, Y.M. Lam, M. Grätzel, S. Mhaisalkar, T.C. Sum,

Long-Range Balanced Electron- and Hole-Transport Lengths in Organic-Inorganic

CH3NH3PbI3, Science (80-. ). 342 (2013) 344–347. doi:10.1126/science.1243167.

[12] J. Im, C. Lee, J. Lee, S. Park, N. Park, 6 . 5 % Efficient Perovskite Quantum-Dot-Sensitized

Solar Cell, (2011) 1–5.

[13] M.M. Lee, J. Teuscher, T. Miyasaka, T.N. Murakami, H.J. Snaith, Efficient Hybrid Solar

Cells Based on Meso-Superstructured Organometal Halide Perovskites, Science (80-. ). 338

(2012) 643–647. doi:10.1126/science.1228604.

[14] H.-S. Kim, C.-R. Lee, J.-H. Im, K.-B. Lee, T. Moehl, A. Marchioro, S.-J. Moon, R.

Humphry-Baker, J.-H. Yum, J.E. Moser, M. Grätzel, N.-G. Park, Lead Iodide Perovskite

Sensitized All-Solid-State Submicron Thin Film Mesoscopic Solar Cell with Efficiency

Exceeding 9%, Sci. Rep. 2 (2012) 591. doi:10.1038/srep00591.

[15] J. Burschka, N. Pellet, S.-J. Moon, R. Humphry-Baker, P. Gao, M.K. Nazeeruddin, M.

Grätzel, Sequential deposition as a route to high-performance perovskite-sensitized solar

100

cells, Nature. 499 (2013) 316. doi:10.1038/nature12340.

[16] O. Malinkiewicz, A. Yella, Y.H. Lee, G.M. Espallargas, M. Graetzel, M.K. Nazeeruddin,

H.J. Bolink, Perovskite solar cells employing organic charge-transport layers, Nat. Photon.

8 (2013) 128. doi:10.1038/nphoton.2013.341.

[17] M. Liu, M.B. Johnston, H.J. Snaith, Efficient planar heterojunction perovskite solar cells by

vapour deposition, Nature. 501 (2013) 395. doi:10.1038/nature12509.

[18] P. Docampo, J.M. Ball, M. Darwich, G.E. Eperon, H.J. Snaith, Efficient organometal

trihalide perovskite planar-heterojunction solar cells on flexible polymer substrates, Nat.

Commun. 4 (2013) 2761. doi:10.1038/ncomms3761.

[19] H. Zhou, Q. Chen, G. Li, S. Luo, T. Song, H.-S. Duan, Z. Hong, J. You, Y. Liu, Y. Yang,

Interface engineering of highly efficient perovskite solar cells, Science (80-. ). 345 (2014)

542–546. doi:10.1126/science.1254050.

[20] X.C. Liu, R. Hong, C. Tian, Tolerance factor and the stability discussion of ABO 3-type

ilmenite, J. Mater. Sci. Mater. Electron. 20 (2009) 323–327. doi:10.1007/s10854-008-9728-

8.

[21] Y. Huang, Q. De Sun, W. Xu, Y. He, W.J. Yin, Halide perovskite materials for solar cells:

A theoretical review, Wuli Huaxue Xuebao/ Acta Phys. - Chim. Sin. 33 (2017) 1730–1751.

doi:10.3866/PKU.WHXB201705042.

[22] S.D. Stranks, G.E. Eperon, G. Grancini, C. Menelaou, M.J.P. Alcocer, T. Leijtens, L.M.

Herz, A. Petrozza, H.J. Snaith, Electron-Hole Diffusion Lengths Exceeding 1 Micrometer

101

in an Organometal Trihalide Perovskite Absorber, 342 (2013) 341–345.

[23] D. Shi, V. Adinolfi, R. Comin, M. Yuan, E. Alarousu, A. Buin, Y. Chen, S. Hoogland, A.

Rothenberger, K. Katsiev, Y. Losovyj, X. Zhang, P.A. Dowben, O.F. Mohammed, E.H.

Sargent, O.M. Bakr, Low trap-state density and long carrier diffusion in organolead trihalide

perovskite single crystals, 347 (2015).

[24] T. Salim, S. Sun, Y. Abe, A. Krishna, A.C. Grimsdale, Y.M. Lam, Perovskite-based solar

cells: impact of morphology and device architecture on device performance, J. Mater. Chem.

A. 3 (2015) 8943–8969. doi:10.1039/c4ta05226a.

[25] M. Jung, S.-G. Ji, G. Kim, S. Il Seok, Perovskite precursor solution chemistry: from

fundamentals to photovoltaic applications, Chem. Soc. Rev. (2019).

doi:10.1039/c8cs00656c.

[26] J.-Y. Jeng, Y.-F. Chiang, M.-H. Lee, S.-R. Peng, T.-F. Guo, P. Chen, T.-C. Wen, Benefits

of very thin PCBM and LiF layers for solution-processed p–i–n perovskite solar cells, Adv.

Mater. 25 (2013) 3727–3732. doi:doi:10.1002/adma.201301327.

[27] J. Seo, S. Park, Y. Chan Kim, N.J. Jeon, J.H. Noh, S.C. Yoon, S. Il Seok, Benefits of very

thin PCBM and LiF layers for solution-processed p–i–n perovskite solar cells, Energy

Environ. Sci. 7 (2014) 2642–2646. doi:10.1039/C4EE01216J.

[28] J. Jeng, K. Chen, T. Chiang, P. Lin, T. Tsai, Y. Chang, T. Guo, P. Chen, T. Wen, Y. Hsu,

Nickel Oxide Electrode Interlayer in CH3NH3PbI3 Perovskite/PCBM Planar-Heterojunction

Hybrid Solar Cells, (2014) 4107–4113. doi:10.1002/adma.201306217

102

[29] M.M. Lee, J. Teuscher, T. Miyasaka, T.N. Murakami, H.J. Snaith, Efficient Hybrid Solar

Cells Based on Meso-Superstructured Organometal Halide Perovskites, Science (80-. ). 338

(2012) 643–647. doi:10.1126/science.1228604.

[30] G.E. Eperon, V.M. Burlakov, P. Docampo, A. Goriely, H.J. Snaith, Morphological control

for high performance, solution-processed planar heterojunction perovskite solar cells, Adv.

Funct. Mater. 24 (2014) 151–157. doi:10.1002/adfm.201302090.

[31] M. Saliba, K.W. Tan, H. Sai, D.T. Moore, T. Scott, W. Zhang, L.A. Estroff, U. Wiesner,

H.J. Snaith, Influence of thermal processing protocol upon the crystallization and

photovoltaic performance of organic-inorganic lead trihalide perovskites, J. Phys. Chem. C.

118 (2014) 17171–17177. doi:10.1021/jp500717w.

[32] K. Mahmood, S. Sarwar, M.T. Mehran, Current status of electron transport layers in

perovskite solar cells: materials and properties, RSC Adv. 7 (2017) 17044–17062.

doi:10.1039/C7RA00002B.

[33] Q. Jiang, L. Zhang, H. Wang, X. Yang, J. Meng, H. Liu, Z. Yin, J. Wu, X. Zhang, J. You,

Enhanced electron extraction using SnO2 for high-efficiency planar-structure

HC(NH2)2PbI3-based perovskite solar cells, Nat. Energy. 2 (2016) 16177.

http://dx.doi.org/10.1038/nenergy.2016.177.

[34] L. Groenendaal, F. Jonas, D. Freitag, H. Pielartzik, J.R. Reynolds, Poly(3,4-

ethylenedioxythiophene) and Its Derivatives : Past , Present , and Future, Adv. Mater. 12

(2000) 481–494.

103

[35] J. Ouyang, C.W. Chu, F.C. Chen, Q. Xu, Y. Yang, High-conductivity poly(3,4-

ethylenedioxythiophene):poly(styrene sulfonate) film and its application in polymer

optoelectronic devices, Adv. Funct. Mater. 15 (2005) 203–208.

doi:10.1002/adfm.200400016.

[36] B.M. Kayes, H.A. Atwater, N.S. Lewis, Comparison of the device physics principles of

planar and radial p-n junction nanorod solar cells, J. Appl. Phys. 97 (2005).

doi:10.1063/1.1901835.

[37] M. Green, Solar Cells: Operating Principles, Technology and System Applications, 1982.

[38] F.D.M. Haldane, Model for a Quantum Hall Effect without Landau Levels: Condensed-

Matter Realization of the “Parity Anomaly,” Phys. Rev. Lett. 61 (1988) 2015–2018.

doi:10.1103/PhysRevLett.61.2015.

[39] T. Markvart, L. Castañer, Principles of solar cell operation, Elsevier Ltd, 2017.

doi:10.1016/B978-0-12-809921-6.00001-X.

[40] S. Silvestre, Strategies for Fault Detection and Diagnosis of PV Systems, Elsevier Inc., 2018.

doi:10.1016/B978-0-12-812959-3.00007-1.

[41] C. Sah, R.N. Noyce, W. Shockley, Carrier generation and recombination in P-N junctions

and PeN junction characteristics, Proc IRE. 45 (1957) 1957.

[42] K. Emery, Measurement and Characterization of Solar Cells and Modules, 2003.

doi:10.1002/9780470974704.ch18.

[43] J. Wei, Measuring the I-V Characteristic of PN Junction Devices with HF2LI Lock-in

104

Amplifier, (2015). https://www.zhinst.com/blogs/jamesw/measuring-the-i-v-characteristic-

of-pn-junction-devices-with-hf2li-lock-in-amplifier/.

[44] A. Hagfeldt, U.B. Cappel, G. Boschloo, L. Sun, L. Kloo, H. Pettersson, E.A. Gibson, Dye-

Sensitized Photoelectrochemical Cells, Elsevier Ltd, 2012. doi:10.1016/B978-0-12-

385934-1.00015-5.

[45] T.M. Schmidt, T.T. Larsen-Olsen, J.E. Carlé, D. Angmo, F.C. Krebs, Upscaling of

Perovskite Solar Cells: Fully Ambient Roll Processing of Flexible Perovskite Solar Cells

with Printed Back Electrodes, Adv. Energy Mater. 5 (2015) 1–9.

doi:10.1002/aenm.201500569.

[46] Y.S. Jung, K. Hwang, Y.J. Heo, J.E. Kim, D. Vak, D.Y. Kim, Progress in Scalable Coating

and Roll-to-Roll Compatible Printing Processes of Perovskite Solar Cells toward

Realization of Commercialization, Adv. Opt. Mater. 1701182 (2018) 1–30.

doi:10.1002/adom.201701182.

[47] Z. Li, T.R. Klein, D.H. Kim, M. Yang, J.J. Berry, M.F.A.M. Van Hest, K. Zhu, Scalable

fabrication of perovskite, (2018) 1–20. doi:10.1038/natrevmats.2018.17.

[48] Y. Rong, Y. Hu, A. Mei, H. Tan, M.I. Saidaminov, S. Il Seok, M.D. Mcgehee, E.H. Sargent,

H. Han, Challenges for commercializing perovskite solar cells, (2018).

doi:10.1126/science.aat8235.

[49] Y. Deng, Q. Dong, C. Bi, Y. Yuan, J. Huang, Air-Stable, Efficient Mixed-Cation Perovskite

Solar Cells with Cu Electrode by Scalable Fabrication of Active Layer, Adv. Energy Mater.

105

6 (2016) 1–6. doi:10.1002/aenm.201600372.

[50] Y. Deng, E. Peng, Y. Shao, Z. Xiao, Q. Dong, J. Huang, Scalable fabrication of efficient

organolead trihalide perovskite solar cells with doctor-bladed active layers, Energy Environ.

Sci. 8 (2015) 1544–1550. doi:10.1039/C4EE03907F.

[51] K. Hwang, Y.S. Jung, Y.J. Heo, F.H. Scholes, S.E. Watkins, J. Subbiah, D.J. Jones, D.Y.

Kim, D. Vak, Toward large scale roll-to-roll production of fully printed perovskite solar

cells, Adv. Mater. 27 (2015) 1241–1247. doi:10.1002/adma.201404598.

[52] C. Zuo, D. Vak, D. Angmo, L. Ding, M. Gao, One-Step Roll-to-Roll Air Processed High

Efficiency Perovskite Solar Cells, Nano Energy. 46 (2018) 185–192.

doi:10.1016/j.nanoen.2018.01.037.

[53] Z. Wei, H. Chen, K. Yan, S. Yang, Inkjet printing and instant chemical transformation of a

CH3NH3PbI3/nanocarbon electrode and interface for planar perovskite solar cells, Angew.

Chemie - Int. Ed. 53 (2014) 13239–13243. doi:10.1002/anie.201408638.

[54] F. Mathies, T. Abzieher, A. Hochstuhl, K. Glaser, A. Colsmann, U.W. Paetzold, G.

Hernandez-sosa, Multipass inkjet printed planar methylammonium lead iodide perovskite

solar cells †, (2016) 19207–19213. doi:10.1039/c6ta07972e.

[55] R. Buntem, P. Kraisingdecha, W. Sadee, Fluoride and Antimony-Doped Tin Oxide Film by

Spray Pyrolysis, Adv. Mater. Res. 55–57 (2008) 513–516.

doi:10.4028/www.scientific.net/AMR.55-57.513.

[56] A. Barrows, A. Pearson, C. Kwak, A. Dunbar, A. Buckley, D. Lidzey, Efficient planar

106

heterojunction mixed-halide perovskite solar cells deposited via spray-deposition, Energy

Environ. Sci. 7 (2014) 1–7. doi:10.1039/C4EE01546K.

[57] W. Chang, D. Lan, K. Lee, X. Wang, Controlled Deposition and Performance Optimization

of Perovskite Solar Cells Using Ultrasonic Spray-Coating of Photoactive Layers, (2017)

1405–1412. doi:10.1002/cssc.201601711.

[58] W. Wu, Q. Wang, Y. Fang, Y. Shao, S. Tang, Y. Deng, H. Lu, Y. Liu, T. Li, Z. Yang, A.

Gruverman, J. Huang, Molecular doping enabled scalable blading of efficient hole-

transport-layer-free perovskite solar cells, Nat. Commun. 9 (2018) 1625.

doi:10.1038/s41467-018-04028-8.

[59] Y. Deng, X. Zheng, Y. Bai, Q. Wang, J. Zhao, J. Huang, Surfactant-controlled ink drying

enables high-speed deposition of perovskite films for efficient photovoltaic modules, Nat.

Energy. (2018). doi:10.1038/s41560-018-0153-9.

[60] X. Ding, J. Liu, Review Article : Transport Phenomena And Fluid Mechanics A Review of

the Operating Limits in Slot Die Coating Processes, 62 (2016). doi:10.1002/aic.

[61] J.-E. Kim, Y.-S. Jung, Y.-J. Heo, K. Hwang, T. Qin, D.-Y. Kim, D. Vak, Slot die coated

planar perovskite solar cells via blowing and heating assisted one step deposition, Sol.

Energy Mater. Sol. Cells. 179 (2018) 80–86. doi:10.1016/j.solmat.2018.02.003.

[62] T. Qin, W. Huang, J.E. Kim, D. Vak, C. Forsyth, C.R. McNeill, Y.B. Cheng, Amorphous

hole-transporting layer in slot-die coated perovskite solar cells, Nano Energy. 31 (2017)

210–217. doi:10.1016/j.nanoen.2016.11.022.

107

[63] K. Hwang, Y.-S. Jung, Y.-J. Heo, F.H. Scholes, S.E. Watkins, J. Subbiah, D.J. Jones, D.-Y.

Kim, D. Vak, Toward Large Scale Roll-to-Roll Production of Fully Printed Perovskite Solar

Cells, Adv. Mater. 27 (2015) 1241–1247. doi:doi:10.1002/adma.201404598.

[64] R.J. Lang, Ultrasonic Atomization of Liquids, J. Acoust. Soc. Am. 34 (1962) 6–8.

doi:10.1016/j.phpro.2015.03.006.

[65] S. Hyuk, Highly efficient CH3NH3PbI3−xClx mixed halide perovskite solar cells prepared

by re-dissolution and crystal grain growth via spray coating, J. Mater. Chem. A Mater.

Energy Sustain. 4 (2016) 17636–17642. doi:10.1039/C6TA06718B.

[66] D.K. Mohamad, J. Griffin, C. Bracher, A.T. Barrows, D.G. Lidzey, Spray-Cast Multilayer

Organometal Perovskite Solar Cells Fabricated in Air, (2016) 1–7.

doi:10.1002/aenm.201600994.

[67] J. Zeleny, The electrical discharge from liquid points, and a hydrostatic method of

measuring the electric intensity at their surfaces, Phys. Rev. 3 (1914) 69–91.

doi:10.1103/PhysRev.3.69.

[68] M. Cloupeau, B. Pruet-Foch, Electrohydrodynamic spraying functioning modes: A critical

review, Pergamon. 25 (1994).

[69] Disintegration of water drops in an electric field, Proc. R. Soc. London. Ser. A. Math. Phys.

Sci. 280 (1964) 383 LP – 397.

http://rspa.royalsocietypublishing.org/content/280/1382/383.abstract.

[70] M. Cloupeau, B. Prunet-Foch, Electrostatic spraying of liquids in cone-jet mode, J.

108

Electrostat. 22 (1989) 135–159. doi:10.1016/0304-3886(89)90081-8.

[71] W. Deng, A. Gomez, The role of electric charge in microdroplets impacting on conducting

surfaces, Phys. Fluids. 22 (2010) 1–4. doi:10.1063/1.3431739.

[72] W. Deng, C.M. Waits, B. Morgan, A. Gomez, Compact multiplexing of monodisperse

electrosprays, J. Aerosol Sci. 40 (2009) 907–918. doi:10.1016/j.jaerosci.2009.07.002.

[73] S.C. Hong, G. Lee, K. Ha, J. Yoon, N. Ahn, W. Cho, M. Park, M. Choi, Precise Morphology

Control and Continuous Fabrication of Perovskite Solar Cells Using Droplet-Controllable

Electrospray Coating System, ACS Appl. Mater. Interfaces. 9 (2017) 7879–7884.

doi:10.1021/acsami.6b15095.

[74] P.-Y. Lin, Y.-Y. Chen, T.-F. Guo, Y.-S. Fu, L.-C. Lai, C.-K. Lee, Electrospray technique in

fabricating perovskite-based hybrid solar cells under ambient conditions, RSC Adv. 7 (2017)

10985–10991. doi:10.1039/C6RA27704G.

[75] S. Kavadiya, D.M. Niedzwiedzki, S. Huang, P. Biswas, Electrospray-Assisted Fabrication

of Moisture-Resistant and Highly Stable Perovskite Solar Cells at Ambient Conditions, Adv.

Energy Mater. 7 (2017) 1–9. doi:10.1002/aenm.201700210.

[76] S. Li, K. Jiang, M. Su, X. Cui, J. Huang, Materials Chemistry A, (2015) 9092–9097.

doi:10.1039/c4ta05675b.

[77] K. Cao, Z. Zuo, J. Cui, Y. Shen, T. Moehl, S.M. Zakeeruddin, M. Grätzel, M. Wang,

Efficient screen printed perovskite solar cells based on mesoscopic TiO2/Al2O3/NiO carbon

architecture, Nano Energy. 17 (2015) 171–179. doi:10.1016/j.nanoen.2015.08.009.

109

[78] J. Fernández de la Mora, The Fluid Dynamics of Taylor Cones, Annu. Rev. Fluid Mech. 39

(2007) 217–243. doi:10.1146/annurev.fluid.39.050905.110159.

[79] K. Park Atomic Force Microscope, Conductive AFM, Probing the Local Electronic

Structure of a Sample’s Surface, (n.d.). https://www.parksystems.com/index.php/park-spm-

modes/94-electrical-properties/233-conductive-afm.

[80] A. Kojima, K. Teshima, Y. Shirai, T. Miyasaka, Organometal Halide Perovskites as Visible-

Light Sensitizers for Photovoltaic Cells, J Am Chem Soc. 131 (2009) 6050–6051.

doi:10.1021/ja809598r.

[81] J. Burschka, N. Pellet, S.J. Moon, R. Humphry-Baker, P. Gao, M.K. Nazeeruddin, M.

Grätzel, Sequential deposition as a route to high-performance perovskite-sensitized solar

cells, Nature. 499 (2013) 316–319. doi:10.1038/nature12340.

[82] M.A. Green, Y. Hishikawa, E.D. Dunlop, D.H. Levi, J. Hohl-Ebinger, A.W.Y. Ho-Baillie,

Solar cell efficiency tables (version 51), Prog. Photovoltaics Res. Appl. 26 (2018) 3–12.

doi:10.1002/pip.2978.

[83] Y. Deng, E. Peng, Y. Shao, Z. Xiao, Q. Dong, J. Huang, Scalable fabrication of efficient

organolead trihalide perovskite solar cells with doctor-bladed, Energy Environ. Sci. 8 (2015)

1544–1550. doi:10.1039/C4EE03907F.

[84] J. Zheng, M. Zhang, C.F.J. Lau, X. Deng, J. Kim, Q. Ma, C. Chen, M.A. Green, S. Huang,

A.W.Y. Ho-Baillie, Spin-coating free fabrication for highly efficient perovskite solar cells,

Sol. Energy Mater. Sol. Cells. 168 (2017) 165–171. doi:10.1016/j.solmat.2017.04.029.

110

[85] B. Almería, Synthesis of Biodegradable Polymer Micro- and Nanoparticles for Controlled

Drug Delivery by Multiplexed Electrosprays, Yale University, 2012.

http://gradworks.umi.com/35/35/3535286.html.

[86] R.P.A. Hartman, D.J. Brunner, D.M.A. Camelot, J.C.M. Marijnissen, B. Scarlett, Jet break-

up in electrohydrodynamic atomization in the cone-jet mode, J. Aerosol Sci. 31 (2000) 65–

95. doi:10.1016/S0021-8502(99)00034-8.

[87] L.R. F.R.S., On the equilibrium of liquid conducting masses charged with electricity,

London, Edinburgh, Dublin Philos. Mag. J. Sci. 14 (1882) 184–186.

[88] S.E. Park, S. Kim, K. Kim, H.E. Joe, B. Jung, E. Kim, W. Kim, B.K. Min, J. Hwang,

Fabrication of ordered bulk heterojunction organic photovoltaic cells using nanopatterning

and electrohydrodynamic spray deposition methods, Nanoscale. 4 (2012) 7773–7779.

doi:10.1039/c2nr32165c.

[89] B. Abramzon, W.A. Sirignano, Droplet vaporisation model for spray combustion

calculations, Int. J. Heat Mass Transf. 9 (1989) 1605–1618.

[90] S. Novak, Electrospray deposition of chalcogenide glass films for gradient refractive index

and quantum dot incorporation, Clemson University, 2015.

[91] C. V. Sternling, L.E. Scriven, Interfacial turbulence: Hydrodynamic instability and the

marangoni effect, AIChE J. 5 (1959) 514–523. doi:10.1002/aic.690050421.

[92] H. Minemawari, T. Yamada, H. Matsui, J. Tsutsumi, S. Haas, R. Chiba, R. Kumai, T.

Hasegawa, Inkjet printing of single-crystal films, Nature. 475 (2011) 364–367.

111

doi:10.1038/nature10313.

[93] S. V Dalvi, M.D. Yadav, Ultrasonics Sonochemistry Effect of ultrasound and stabilizers on

nucleation kinetics of curcumin during liquid antisolvent precipitation, Ultrason. -

Sonochemistry. 24 (2015) 114–122. doi:10.1016/j.ultsonch.2014.11.016.

[94] X. Zheng, B. Chen, C. Wu, S. Priya, Room temperature fabrication of CH3NH3PbBr3 by

anti-solvent assisted crystallization approach for perovskite solar cells with fast response

and small J-V hysteresis, Nano Energy. 17 (2015) 269–278.

doi:10.1016/j.nanoen.2015.08.023.

[95] X. Li, D. Bi, C. Yi, J.-D. Décoppet, J. Luo, S.M. Zakeeruddin, A. Hagfeldt, M. Grätzel, A

vacuum flash-assisted solution process for high-efficiency large-area perovskite solar cells.,

Science. 353 (2016) 58–62. doi:10.1126/science.aaf8060.

[96] Y. Yu, S. Yang, L. Lei, Y. Liu, Nucleation mediated interfacial precipitation for

architectural perovskite films with enhanced photovoltaic performance, Nanoscale. (2017).

doi:10.1039/C6NR08741H.

[97] N.J. Jeon, J.H. Noh, W.S. Yang, Y.C. Kim, S. Ryu, J. Seo, S. Il Seok, Compositional

engineering of perovskite materials for high-performance solar cells, Nature. 517 (2015)

476–480. doi:10.1038/nature14133.

[98] M. He, B. Li, X. Cui, B. Jiang, Y. He, Y. Chen, D. O’Neil, P. Szymanski, M.A. EI-Sayed,

J. Huang, Z. Lin, Meniscus-assisted solution printing of large-grained perovskite films for

high-efficiency solar cells, Nat. Commun. 8 (2017) 16045. doi:10.1038/ncomms16045.

112

[99] J. Li, R. Munir, Y. Fan, T. Niu, Y. Liu, Y. Zhong, Z. Yang, Y. Tian, B. Liu, J. Sun, D.M.

Smilgies, S. Thoroddsen, A. Amassian, K. Zhao, S. (Frank) Liu, Phase Transition Control

for High-Performance Blade-Coated Perovskite Solar Cells, Joule. (2018) 1–18.

doi:10.1016/j.joule.2018.04.011.

[100] Z. Yang, C.C. Chueh, F. Zuo, J.H. Kim, P.W. Liang, A.K.Y. Jen, High-Performance Fully

Printable Perovskite Solar Cells via Blade-Coating Technique under the Ambient Condition,

Adv. Energy Mater. 5 (2015) 1–6. doi:10.1002/aenm.201500328.

[101] A. Mei, X. Li, L. Liu, Z. Ku, T. Liu, Y. Rong, M. Xu, M. Hu, J. Chen, Y. Yang, M. Grätzel,

H. Han, A hole-conductor-free, fully printable mesoscopic perovskite solar cell with high

stability, Science (80-. ). 345 (2014) 295–298. doi:10.1126/science.1254763.

[102] J.-W. Lee, S.-I. Na, S.-S. Kim, Efficient spin-coating-free planar heterojunction perovskite

solar cells fabricated with successive brush-painting, J. Power Sources. 339 (2017) 33–40.

doi:https://doi.org/10.1016/j.jpowsour.2016.11.028.

[103] Y.-S. Jung, K. Hwang, Y.-J. Heo, J.-E. Kim, D. Lee, C.-H. Lee, H.-I. Joh, J.-S. Yeo, D.-Y.

Kim, One-Step Printable Perovskite Films Fabricated under Ambient Conditions for

Efficient and Reproducible Solar Cells, ACS Appl. Mater. Interfaces. 9 (2017) 27832–

27838. doi:10.1021/acsami.7b05078.

[104] A. Bashir, S. Shukla, J.H. Lew, S. Shukla, A. Bruno, D. Gupta, T. Baikie, R. Patidar, Z.

Akhter, A. Priyadarshi, N. Mathews, S.G. Mhaisalkar, Spinel Co3O4 nanomaterials for

efficient and stable large area carbon-based printed perovskite solar cells, Nanoscale. (2018)

2341–2350. doi:10.1039/C7NR08289D.

113

[105] M. V Kovalenko, L. Protesescu, M.I. Bodnarchuk, Properties and potential optoelectronic

applications of lead halide perovskite nanocrystals, 750 (2017) 745–750.

doi:10.1126/science.aam7093.

[106] W. Deng, J.F. Klemic, X. Li, M.A. Reed, A. Gomez, Increase of electrospray throughput

using multiplexed microfabricated sources for the scalable generation of monodisperse

droplets, J. Aerosol Sci. 37 (2006) 696–714. doi:10.1016/j.jaerosci.2005.05.011.

[107] A. Guerrero, J. You, C. Aranda, Y.S. Kang, G. Garcia-Belmonte, H. Zhou, J. Bisquert, Y.

Yang, Interfacial degradation of planar lead halide perovskite solar cells, ACS Nano. 10

(2016) 218–224. doi:10.1021/acsnano.5b03687.

[108] Y. Han, S. Meyer, Y. Dkhissi, K. Weber, J.M. Pringle, U. Bach, L. Spiccia, Y.B. Cheng,

Degradation observations of encapsulated planar CH3NH3PbI3 perovskite solar cells at high

temperatures and humidity, J. Mater. Chem. A. 3 (2015) 8139–8147.

doi:10.1039/c5ta00358j.

[109] Y. Yang, J. Xiao, H. Wei, L. Zhu, D. Li, Y. Luo, H. Wu, Q. Meng, An all-carbon counter

electrode for highly efficient hole-conductor-free organo-metal perovskite solar cells, RSC

Adv. 4 (2014) 52825–52830. doi:10.1039/c4ra09519g.

[110] Z. Ku, Y. Rong, M. Xu, T. Liu, H. Han, Full Printable Processed Mesoscopic

CH3NH3PbI3/TiO2 Heterojunction Solar Cells with Carbon Counter Electrode, Sci. Rep. 3

(2013) 3132. doi:10.1038/srep03132.

[111] H. Chen, S. Yang, Carbon-Based Perovskite Solar Cells without Hole Transport Materials:

114

The Front Runner to the Market?, Adv. Mater. 29 (2017). doi:10.1002/adma.201603994.

[112] A. Mei, X. Li, L. Liu, Z. Ku, T. Liu, Y. Rong, M. Xu, M. Hu, J. Chen, Y. Yang, M. Grätzel,

H. Han, A hole-conductor-free, fully printable mesoscopic perovskite solar cell with high

stability, Science (80-. ). 345 (2014) 295–298. doi:10.1126/science.1254763.

[113] Z. Ku, Y. Rong, M. Xu, T. Liu, H. Han, Full printable processed mesoscopic CH3NH3PbI3

/TiO2 heterojunction solar cells with carbon counter electrode, Sci. Rep. 3 (2013).

doi:10.1038/srep03132.

[114] H. Li, K. Cao, J. Cui, S. Liu, X. Qiao, Y. Shen, M. Wang, 14.7% Efficient Mesoscopic

Perovskite Solar Cells Using Single Walled Carbon Nanotubes/Carbon Composite Counter

Electrodes, Nanoscale. 8 (2016) 6379–6385. doi:10.1039/c5nr07347b.

[115] Z. Wei, K. Yan, H. Chen, Y. Yi, T. Zhang, X. Long, J. Li, L. Zhang, J. Wang, S. Yang,

Cost-efficient clamping solar cells using candle soot for hole extraction from ambipolar

perovskites, Energy Environ. Sci. 7 (2014) 3326–3333. doi:10.1039/c4ee01983k.

[116] H. Zhou, Y. Shi, Q. Dong, H. Zhang, Y. Xing, K. Wang, Y. Du, T. Ma, Hole-conductor-

free, metal-electrode-free TiO2/CH3NH3PbI3 heterojunction solar cells based on a low-

temperature carbon electrode, J. Phys. Chem. Lett. 5 (2014) 3241–3246.

doi:10.1021/jz5017069.

[117] F. Zhang, X. Yang, H. Wang, M. Cheng, J. Zhao, L. Sun, Structure engineering of hole-

conductor free perovskite-based solar cells with low-temperature-processed commercial

carbon paste as cathode, ACS Appl. Mater. Interfaces. 6 (2014) 16140–16146.

115

doi:10.1021/am504175x.

[118] Z. Wei, K. Yan, H. Chen, Y. Yi, T. Zhang, X. Long, J. Li, L. Zhang, J. Wang, S. Yang,

Cost-efficient clamping solar cells using candle soot for hole extraction from ambipolar

perovskites, Energy Environ. Sci. 7 (2014) 3326–3333. doi:10.1039/C4EE01983K.

[119] H. Zhang, J. Xiao, J. Shi, H. Su, Y. Luo, D. Li, H. Wu, Y.-B. Cheng, Q. Meng, Self-

Adhesive Macroporous Carbon Electrodes for Efficient and Stable Perovskite Solar Cells,

Adv. Funct. Mater. (2018) 1802985. doi:10.1002/adfm.201802985.

[120] F. Meng, L. Gao, Y. Yan, J. Cao, N. Wang, T. Wang, T. Ma, Ultra-low-cost coal-based

carbon electrodes with seamless interfacial contact for effective sandwich-structured

perovskite solar cells, Carbon N. Y. 145 (2019) 290–296. doi:10.1016/j.carbon.2019.01.047.

[121] H. Wei, J. Xiao, Y. Yang, S. Lv, J. Shi, X. Xu, J. Dong, Y. Luo, D. Li, Q. Meng, Free-

standing flexible carbon electrode for highly efficient hole-conductor-free perovskite solar

cells, Carbon N. Y. 93 (2015) 861–868. doi:10.1016/j.carbon.2015.05.042.

[122] X. Meng, J. Zhou, J. Hou, X. Tao, S.H. Cheung, S.K. So, S. Yang, Versatility of Carbon

Enables All Carbon Based Perovskite Solar Cells to Achieve High Efficiency and High

Stability, Adv. Mater. 1706975 (2018) 1–6. doi:10.1002/adma.201706975.

[123] S. Mashhoun, Y. Hou, H. Chen, F. Tajabadi, N. Taghavinia, H.-J. Egelhaaf, C.J. Brabec,

Resolving a Critical Instability in Perovskite Solar Cells by Designing a Scalable and

Printable Carbon Based Electrode-Interface Architecture, Adv. Energy Mater. (2018)

1802085. doi:10.1002/aenm.201802085.

116

[124] X. Meng, J. Zhou, J. Hou, X. Tao, S.H. Cheung, S.K. So, S. Yang, Versatility of Carbon

Enables All Carbon Based Perovskite Solar Cells to Achieve High Efficiency and High

Stability, Adv. Mater. (2018). doi:10.1002/adma.201706975.

[125] H. Zhang, J. Xiao, J. Shi, H. Su, Y. Luo, D. Li, H. Wu, Y. Cheng, Q. Meng, Self‐Adhesive

Macroporous Carbon Electrodes for Efficient and Stable Perovskite Solar Cells, Adv. Funct.

Mater. 1802985 (2018) 1–8. doi:10.1002/ADFM.201802985.

[126] L. Zhuang, Synergy of ammonium chloride and moisture on perovskite crystallization for

efficient printable mesoscopic solar cells, Wuli Huaxue Xuebao/ Acta Phys. - Chim. Sin.

33 (2017) 647–648. doi:10.1016/j.lfs.2004.02.037.

[127] H. Li, K. Cao, J. Cui, S. Liu, X. Qiao, Y. Shen, M. Wang, 14.7% Efficient Mesoscopic

Perovskite Solar Cells Using Single Walled Carbon Nanotubes/Carbon Composite Counter

Electrodes, Nanoscale. 8 (2016) 6379–6385. doi:10.1039/c5nr07347b.

[128] Z. Wu, Z. Liu, Z. Hu, Z. Hawash, L. Qiu, Y. Jiang, L.K. Ono, Y. Qi, Highly Efficient and

Stable Perovskite Solar Cells via Modification of Energy Levels at the Perovskite / Carbon

Electrode Interface, 1804284 (2019) 1–7. doi:10.1002/adma.201804284.

[129] Q.Q. Chu, B. Ding, Q. Qiu, Y. Liu, C.X. Li, C.J. Li, G.J. Yang, B. Fang, Cost effective

perovskite solar cells with a high efficiency and open-circuit voltage based on a perovskite-

friendly carbon electrode, J. Mater. Chem. A. 6 (2018) 8271–8279. doi:10.1039/c7ta10871k.

[130] J.S. Im, H.J. Kim, M.O. Thompson, Phase transformation mechanisms involved in excimer

laser crystallization of amorphous silicon films, Appl. Phys. Lett. 63 (1993) 1969–1971.

117

doi:10.1063/1.110617.

[131] T. Jeon, H.M. Jin, S.H. Lee, J.M. Lee, H. Il Park, M.K. Kim, K.J. Lee, B. Shin, S.O. Kim,

Laser Crystallization of Organic-Inorganic Hybrid Perovskite Solar Cells, ACS Nano. 10

(2016) 7907–7914. doi:10.1021/acsnano.6b03815.

scalable fabrication of high efficiency hybrid perovskite

Documents