저 시-비 리- 경 지 2.0 한민

는 아래 조건 르는 경 에 한하여 게

l 저 물 복제, 포, 전송, 전시, 공연 송할 수 습니다.

다 과 같 조건 라야 합니다:

l 하는, 저 물 나 포 경 , 저 물에 적 된 허락조건 명확하게 나타내어야 합니다.

l 저 터 허가를 면 러한 조건들 적 되지 않습니다.

저 에 른 리는 내 에 하여 향 지 않습니다.

것 허락규약(Legal Code) 해하 쉽게 약한 것 니다.

Disclaimer

저 시. 하는 원저 를 시하여야 합니다.

비 리. 하는 저 물 리 목적 할 수 없습니다.

경 지. 하는 저 물 개 , 형 또는 가공할 수 없습니다.

공학박사 학위논문

Control of vacuum process for growing high quality

organic/inorganic perovskite films for solar cells

태양전지용 고품질 유/무기 혼합

페로브스카이트 성막을 위한 진공공정제어

2019년 8월

서울대학교 대학원

재료공학부

김 범 수

i

Abstract

Control of vacuum process for growing high quality

organic/inorganic perovskite films for solar cells

Kim Beom-Soo

Department of Materials Science and Engineering

The Graduate School

Seoul National University

Metal halide based perovskites have attracted large attention

delivering high power conversion efficiencies (PCEs) in solar cells and

light emitting diodes (LEDs) in last several years. One of the critical

issues in this field is the growing high-quality films reproducibly. Most

of the devices reported up to now have been fabricated using solution

processes. The solution process have brought to considerable increase

in PCEs, with easy accessibility for experiments combined with various

process engineering. However, the vacuum process has potentials for a

stable process with high reproducibility and controllability, considering

the fact that this process is already successfully employed in

commercialized organic light emitting diode (OLED) industry. In the

ii

vacuum process, various advantages can be realized for the fabrication

the perovskite films such as high purity, large area, controllable growth

parameters in real time, excluding the influences from atmospheric

factors and solvents. Additionally, conformal morphology to the sub-

layer can be readily obtained hence an integration with silicon solar

cells in tandem structure on the textured silicon surfaces is achievable.

Nevertheless, the vacuum processes for the perovskite is still minor

process compared to the solution process in the perovskite field,

presumably due to some difficulties in the process with large entry

barrier. In this thesis, we report all-vacuum processed perovskite solar

cells, effective method for growing the perovskite films and the

adsorption/growth mechanism for the precursor of the perovskite,

methylammonium iodide (MAI) in the vacuum process to improve the

process controllability and to understand the film growth mechanism.

In chapter 1, basic features of the perovskite in respect of the

structure, electrical and optical properties are introduced. In addition,

the general working principle for the perovskite solar cells and vacuum

process are explained.

In chapter 2, application of the vacuum process to the perovskite

sell is discussed. The perovskite solar cell is fabricated using all-

vacuum process including the co-deposited perovskite layer.

Molybdenum oxide (MoO3) and N,N′-Di(1-naphthyl)-N,N′-

iii

diphenyl-(1,1′-biphenyl)-4,4′-diamine (NPB) which are widely used

in vacuum processed organic light emitting diode and organic

photovoltaic, are successfully introduced into the perovskite solar cell.

The solar cells resulted in a high open circuit voltage (VOC) of 1.12 V,

which is comparably high value considering the high performing solar

cells reported up to that time. The fact that the MoO3/NPB layer is

widely used in OPVs/OLEDs implies that technologies developed for

vacuum processed OLEDs/OPVs can be successfully applied in the

perovskite solar cells to accelerate the development of the perovskite

solar cells.

In chapter 3, we report an effective method to control the

compositions of MAPbI3 perovskite via working pressure of MAI and

its effects on perovskite films in the vacuum co-deposition process.

Thermal heating of the MAI source in a vacuum chamber increases the

working pressure, which can be an indicator for the control of the ratio

of MAI in the perovskite film. The working pressure is systematically

varied and resulting physical properties of the perovskite films are

analysed. At the optimum working pressure, the perovskite solar cell

fabricated using an all-vacuum process showed a maximum power

conversion efficiency of 14.5% with high reproducibility, implying

successful formation of the film using the pressure-controlled vacuum

co-deposition.

iv

In chapter 4, we report the growth characteristics of MAI in vacuum

process. Growth characteristic is crucial for the fundamental

understanding the film-growth, as well as for the controllability in the

fabrication process due to the effects on the deposition parameters.

MAI shows surface dependent deposition property which is especially

varied in the presence of PbI2 layers. As the thickness of PbI2 layer

increase, the deposition rates of MAI increased rapidly. At a constant

working pressure, all of the deposition rates begin to decrease reaching

to similar values after certain times depending on the initial PbI2

thicknesses. This phenomena was analysed by atomic force microscopy

(AFM) discussing on the topology, surface potential, thickness and

density changes. Conclusively, these results suggest that the dipole

induced adsorption mechanism for MAI in vacuum process. We believe

that our model would give insights for understanding the growth

mechanism of MAI in vacuum, as well as the improvement of the

controllability since the mechanism directly affect the deposition

parameters.

Keywords: MAPbI3 perovskite solar cell, all-vacuum process, MAI

working pressure, MAI deposition, MAI growth in vacuum

Student Number: 2015 - 30180

v

Contents

Abstract ................................................................................... i

Contents .................................................................................. v

List of Tables ......................................................................... ix

List of Figures ........................................................................ x

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

1.1 Motivation and outline of thesis........................................ 1

1.1.1 Motivation ............................................................................... 1

1.1.2 Outline of thesis ....................................................................... 3

1.2 Perovskite solar cells ......................................................... 5

1.2.1 Materials of the perovskite ...................................................... 5

1.2.2 Optoelectronic properties of the perovskite ............................ 6

1.2.3 Various Methods to grow the perovskite thin film ................ 10

1.2.4 Basic characterization of perovskite solar cell ...................... 15

1.3 Vacuum process ............................................................... 16

1.3.1 General theory ....................................................................... 16

1.3.2 Equipments for vacuum process ........................................... 17

vi

Chapter 2. Vacuum processed perovskite solar cells ....... 21

2.1 Introduction .................................................................... 21

2.2 Preliminary data .............................................................. 23

2.3 Film characterizations..................................................... 26

2.3.1 Absorbance & XRD patterns ................................................. 27

2.3.2 Surface analyzes .................................................................... 30

2.4 Device performances ....................................................... 32

2.4.1 Effects of HTLs ..................................................................... 33

2.4.2 Reproducibility ...................................................................... 48

2.4.3 Hysteresis .............................................................................. 50

2.5 Summary ......................................................................... 53

Chapter 3. Composition-controlled organometal halide

perovskite via MAI pressure in vacuum co-deposition

process .................................................................................. 54

3.1 Introduction ................................................................... 54

3.2 Experiments ................................................................... 56

3.2.1 Perovskite evaporation system .............................................. 56

3.2.2 Film characterization ............................................................. 59

3.3 Results and discussion................................................... 60

vii

3.3.1 Working pressure and thickness ............................................ 60

3.3.2 Working pressure and structural property ............................. 63

3.3.3 Working pressure and absorbance ......................................... 65

3.3.4 Working pressure and morphologies ..................................... 68

3.3.5 Device performance............................................................... 71

3.3.6 Origin for low IPCEs ............................................................. 74

3.3.7 Reproducibility ...................................................................... 77

3.3.8 Compositin measurment by EDX .......................................... 80

3.4 Summary ....................................................................... 82

Chapter 4. Growth characteristics of MAI on PbI2 in

vacuum process.. .................................................................. 83

4.1 Introduction ................................................................... 83

4.2 Experiments ................................................................... 84

4.2.1 Film fabrication ..................................................................... 84

4.2.2 QCM analysis ........................................................................ 85

4.2.3 Film characterization ............................................................. 86

4.3 Results and discussion................................................... 86

4.3.1 Thickness change and XRD depending on sub-layers .......... 86

4.3.2 Study of deposition kinetics using QCM............................... 91

4.3.3 Hypothesis ............................................................................. 95

viii

4.3.5 AFM study ............................................................................. 98

4.3.6 Density and thickness change analysis ................................ 104

4.3.8 Surface potential analysis .................................................... 109

4.3.9 Growth mechanism of MAI in vacuum ............................... 113

4.4 Summary ..................................................................... 118

Chapter 5. Summary and conclusion .............................. 119

Bibliography ...................................................................... 122

초록 ..................................................................................... 131

CURRICULUM VITAE ................................................... 134

List of Publications ............................................................ 136

List of Presentations .......................................................... 137

List of Awards .................................................................... 141

ix

List of Tables

Table 2.1 Perovskite solar cell performance of the best performing

device and the average values of the 72 devices from 6 different

batches.. ................................................................................................ 47

Table 2.2 The solar cell parameters of the device with the MoO3/NPB

layer with different scan directions. ..................................................... 51

Table 3.1 Average photovoltaic parameters of the fabricated perovskite

solar cells. The values were extracted from 4–8 cells (23 cells for the

cells fabricated at 5.1 × 10−5 torr)... ...................................................... 78

Table 4.1 Thickness of deposited MAI on different substrates. All the

substrates are exposed to MAI for 60 min under 7×10-6 torr. .............. 89

x

List of Figures

Figure 1.1 The structure of organometal halide perovskite ................... 8

Figure 1.2 Schematic figure for the various methods to fabricate the

MAPbI3 perovskite films. ................................................................... 14

Figure 1.3 The picture of glove box integrated vacuum evaporator

systems used for the perovskite and other layers. ................................ 18

Figure 2.1 1 (a) The evaporator used for the preliminary experiment

(Bell-jar, Selcos) (b) Absorbance and (c) XRD data of the MAPb3-xClx

film. (d) Device fabricated with different structures using the

evaporator. ............................................................................................ 25

Figure 2.2 The absorbance of the 320 nm thick MAPbI3 film grown on

the glass substrate using vacuum co-evaporation.. ............................... 28

Figure 2.3 The X-ray diffraction (XRD) patterns of a 320 nm thick

MAPbI3film using vacuum co-evaporation... ....................................... 29

Figure 2.4 The demonstration of MAPbI3 perovskite in the preferred

orientation along (110) direction..... ..................................................... 31

Figure 2.5 The J−V characteristic of the device without the MoO3 layer

.............................................................................................................. 36

Figure 2.6 The J−V characteristic of the device without the NPB layer

.............................................................................................................. 38

Figure 2.7 X-ray diffraction (XRD) patterns of the fabricated

perovskite film on the different substrates .......................................... 41

xi

Figure 2.8 The schematic diagram of the energy levels of each layer in

the device without MoO3 layer. Note that the energy levels are

represented against the vacuum level. ................................................ 43

Figure 2.9 (a) The current density-voltage (J−V) characteristics and (b)

incident photon-to-electron conversion efficiency (IPCE) spectrum of

the best performing perovskite solar cell.. ........................................... 46

Figure 2.10 Histogram plots of the solar cell parameters; (a) Power

conversion efficiency (b) short-circuit current (c) open circuit voltage

(d) fill factor of the 72 devices from 6 different batches... .................. 49

Figure 2.11 . The current density-voltage (J−V) characteristics of a

device with the MoO3/NPB layer for different scan directions... ........ 51

Figure 3.1 Schematic Fig. of the vacuum co-deposition system for the

perovskite layer...... .............................................................................. 58

Figure 3.2 Thickness and deposition rate of the perovskite films under

different working pressures deposited for 1 h with the same deposition

rate of PbI2 as 0.5 Å/s. The first data point (2 × 10−6 torr) represents

PbI2 only film...... ................................................................................ 62

Figure 3.3 (a) XRD patterns of the perovskite films under different

pressures on the indium tin oxide/MoO3 (5 nm)/NPB (20 nm) substrate.

(b) Magnified image of the XRD patterns in red square region...... .... 64

Figure 3.4 (a) Absorbance of the perovskite film at various pressures

on a glass substrate. The thicknesses are indicated in the brackets. (b)

xii

Absorbance at 550 nm of the fabricated films. (c) Color of the

perovskite films. From the left, the films are PbI2 only, 0.6, 1.7, 3.2, 4.1,

5.1, 6.5, and 8.3 × 10−5 torr-deposited films....... ................................. 66

Figure 3.5 Reflectance of PbI2. The incident beam is tilted 7° toward

the substrate ......................................................................................... 67

Figure 3.6 AFM images of the perovskite films with different pressures

grown on an indium tin oxide/MoO3 (5 nm)/NPB (20 nm) substrate .. 70

Figure 3.7 Three-dimensional AFM image of the perovskite films

shown in Fig. 3.6. ................................................................................ 70

Figure 3.8 SEM image of the perovskite films shown in Fig. 3.6. .... 70

Figure 3.9 (a) Current density-voltage (J-V) characteristics of the

fabricated devices at the different pressures. (b) IPCE spectra of the

fabricated perovskite solar cells. .......................................................... 73

Figure 3.10 (a) IPCE spectrum selected from Fig. 3.9 for low,

optimized, high working pressure fabricated device with short (450 nm)

and long (650 nm) wavelength. (b) Simulated electric field in the

device of short (450 nm) (c) and long (650 nm) wavelength... .......... 76

Figure 3.11 (a) Hysteresis of the fabricated solar cell fabricated under

the optimized pressure, 5.1×10-5 torr. (b) J–V characteristics at the

different scan rates of the device fabricated under optimized pressure,

5.1×10-5 torr .......................................................................................... 79

Figure 3.12 Atomic ratio of iodine to lead measured by EDX for the

xiii

fabricated perovskite films. .................................................................. 79

Figure 4.1 (a) The thickness of MAI/PbI2 films on different initial PbI2

thicknesses (b) X-ray diffraction patterns of the MAI deposited on PbI2

films with different thicknesses... ......................................................... 90

Figure 4.2 (a) Schematic figure for comparative adsorption experiment

on QCMs (b) monitored deposited weight of MAI on QCM/PbI2(x nm)93

Figure 4.2 (c) The measured thickness change and adsorbed weight

and (d) monitored deposition rates and working pressure.. ................ 94

Figure 4.3 Schematic figure of the hypothesis to explain the deposition

parameter changes shown in Fig. 4.2(d). .............................................. 97

Figure 4.4 (a) 3D-topology (5μm × 5μm) of the MAI exposure films

with different times and initial PbI2 thickness and (b) RMS roughness

values extracted from 5μm × 5μm images.. ........................................ 99

Figure 4.5 (a) Topology (b) Surface potential and (c) phase images of

potentials of the MAI films (5 𝜇𝜇𝑚𝑚×5 𝜇𝜇𝑚𝑚) ......................................... 101

Figure 4.6 (a) Topology (b) Surface potential and (c) phase images of

potentials of the MAI films (1 𝜇𝜇𝑚𝑚×1 𝜇𝜇𝑚𝑚) ......................................... 102

Figure 4.7 (a) Topology (b) Surface potential and (c) phase images of

potentials of the MAI films (0.5 𝜇𝜇𝑚𝑚×0.5 𝜇𝜇𝑚𝑚) ................................... 103

Figure 4.8 (a) Thickness change and (b) density change of the

deposited films on PbI2 (x nm) films with MAI deposition time ....... 106

Figure 4.9 Schematic figure for the thickness change on PbI2 layer in

xiv

vacuum process .................................................................................. 108

Figure 4.10 (a) Surface potential images (5 𝜇𝜇𝑚𝑚 × 5 𝜇𝜇𝑚𝑚) and (b)

Average surface potential values measured from various scales ........ 112

Figure 4.11 (a) Schematic figure for why the MAI potential is changed

with thickness and (b) illustration how the orientation change of MAI

appear in SKPM measurement ........................................................... 115

Figure 4.12 Schematic illustration for growth of MAI on various PbI2

thickness with time ............................................................................. 117

１

Chapter 1

Introduction

1.1 Motivation and outline of thesis

1.1.1 Motivation

The hybrid organic/inorganic perovskite (e.g., MAPbI3) based solar

cells attract large attention because of high power conversion

efficiencies (PCEs) [1-5]. Most of the devices reported up to now have

been fabricated using solution processes. One of the critical issues of

the solution processed perovskite solar cells is the reproducibility

coming from the difficulties in controlling the uniformity [6-7], pin-

hole formation [7-8], sensitivity to moisture and air [9-10]. A variety of

approaches have been introduced such as sequential deposition [2,8,11],

solvent modifications [4,12-16], manipulation of the crystallization

time [17,18], and changing the annealing conditions [19-22], which are

mostly attempted in the modification of the solution processes. Vacuum

evaporation has a potential for high reproducibility and controllability

due to solvent free processes combined with controllable growth

parameters in a clean and inert environment [23-29]. Nevertheless, the

fully vacuum–processed perovskite solar cells have been rarely

２

reported comparing to the solution processed devices. Following Liu’s

successful application of the vacuum process to the perovskite solar

cell, several groups have reported the perovskite solar cells using

vacuum co-deposition process [23-25, 30-42]. Most of these groups

commented on difficulties of controlling the MAI in vacuum. It is

known to be difficult to calibrate the thicknesses of deposited MAI due

to the gas-like behaviors in vacuum chamber and the rough surfaces

with poorly adsorbed MAI films [24, 25, 30-33, 43-46]. These

problems hinder an accurate monitoring and controlling the MAI in

vacuum process. However, the vacuum process is still one of the most

promising method since its general advantages such as high purity

films, uniformity, clean environments excluding solvent/atmospheric

influences [23-29]. Additionally, commercially succeeded OLED

industries have used vacuum process implying it is stable and effective

process for optoelectronic devices.

Regarding these issues, in this thesis, researches for employing the

advantages of vacuum process to the perovskite solar cells,

methodology to control the process and growth mechanism of MAI in

vacuum process will be discussed.

３

1.1.2 Outline of thesis

In chapter 1, a brief introduction for the material, MAPbI3, and the

development of the methods for perovskite film-growth will be

described. Additionally, the general working mechanism of the

perovskite solar cell and the experimental methodologies to fabricate

the perovskite films in vacuum process will be introduced.

In chapter 2, the perovskite solar cell employing all-vacuum process

will be mainly discussed. The stable and efficient combination for hole

extraction layers MoO3/NPB, which is widely used in the vacuum

processed OLEDs/OPVs, were successfully applied into the vacuum

processed perovskite solar cell. This successful application of the

vacuum deposited MoO3/NPB as a hole extraction layer implies that

utilization of typical thermal-evaporation materials developed for the

OPVs/OLEDs to perovskite solar cells is feasible, thus leading to an

extension of the research areas

In chapter 3, an efficient method to control the vacuum process for

the perovskite and the resultant thin film properties are discussed. The

compositions of the perovskite were able to be adjusted with the

working pressures of MAI. Using this method, the properties of the

films with various compositions were significantly differed. The

４

thickness, optical properties, structural, morphological, compositional

properties with device performances are reported. At the optimum

pressure of MAI resulted in the maximum power conversion efficiency

of 14.5% with good reproducibility. Throughout this work, it is

expected that it would be useful for the readers to find their optimized

condition of the vacuum co-deposited perovskite films.

In chapter 4, the growth mechanism of MAI in vacuum process is

discussed. MAI shows surface dependent deposition characteristics

which is especially varied in the presence of PbI2 layers. As the

conclusion, we suggest the dipole induced adsorption mechanism for

MAI in vacuum process. This chapter would be interested in terms of

the newly suggested growth mechanism which varies the thickness or

deposition rate. In this regards, the researchers who work in this field

may find practical information for control MAI in vacuum since the

growth mechanism directly affect the deposition parameters.

５

1.2 Perovskite solar cells

1.2.1 Materials of the perovskite

In perovskite solar cells, the perovskite usually composed with

organic cat ion (A), metal (B), and halogen components(X) called

ABX3 type perovskite or organo-metal halide perovskite. In this

perovskite, MA+, Pb+, I- are widely used components as the organic cat

ion, metal and halogen parts respectively in the orgaometal halide

perovskite. MA+ plays the role of providing a charge balance in

perovskite structure without covalent interactions with other

components where Pb2+ and I- are bonded with the combination of

covalent and ionic bonds. In this perovskite structure, the lower part of

the conduction band (CB) is mainly composed by hybridization of s

orbital of Pb and p orbital of I where upper part of the valance band

(VB) is mainly comprised of p orbital of Pb [47]. The perovskite is

typically obtained from the reaction of the two precursors, PbI2 and

MAI. The formation activation energy of the perovskite is low enough

to occur in room temperature, hence the perovskite can be readily

formed by mixing the two components or employing the diffusion of

MAI into the PbI2 layer.

６

1.2.2 Optoelectronic properties of perovskite

The probability of exiting the molecules are given by Since the direct

p-p band transition of the perovskite, where the electron density is

localized, the perovskite shows the high absorption coefficients in the

visible light range with the optical band gap of ~1.5 eV [48]. The

absorption coefficient of the perovskite in the visible light is

comparable to amorphous silicon or GaAs. Also, since the optical band

gap can be tuned by substituting the elements of the perovskite, the

absorption edge can approach to Shockley−Queisser limit [49]. On the

other hand, it is reported that measuring accurate refractive index of the

perovskite is challenging. The grain size, roughness, pin-holes and

crystallinity which depend on the film growing process affect the

refractive index significantly [50, 51]. Therefore, the stable process for

the film fabrication is required to accurately measure the optical

property of the perovskite thin films. In the perovskite, it was debatable

whether the excitons or free carriers are dominantly generated since the

exciton binding energy is known to be about 13~16 meV [52-54],

which results in delocalized Wannier-like exciton characteristics.

Therefore, in perovskite solar cells the free carriers are mainly

considered. The crucial working principle of the perovskite absorber in

the solar cell is characterized by significantly long charge carrier

diffusion length [55-58]. Several mechanisms have been proposed to

７

explain the long carrier diffusion length both hole and electron. The one

of the crucial reasons for this long electron-hole diffusion lengths is

that the dominant point defects, such as iodine interstitial defects,

methylammonium interstitial defects and iodine vacancies generate

shallow traps against CB or VB resulting in easy de-trap process which

does not significantly hinder the carrier transporting [59-62]. This is

considered to be the reason of the superior electrical property (long life

time, diffusion length) of the perovskite. Another mechanism is

featured as the ferroelectricity of the perovskite. J. M. Frost et al.,

reported that the spontaneous polarization of the perovskite induces

muti-domains in the perovskite resulting in the separation the path for

the electron and holes respectively. A similar mechanism has been

previously proved by piezorsponse force microscopy (PFM) for the

BaFeO3 perovskite where the domains reduce the bimolecular

recombination resulting in long carrier diffusion length [46].

Therefore, it is possible to diffuse over a micrometer for the electrons

and holes through the CB and VB respectively. Additionally, it has been

８

Figure. 1.1 The structure of organometal halide perovskite. The most

widely used components are marked with bold in the parentheses

which are MA+, Pb2+, I- for the monovalent organic cat ion, divalent

metal and the halogen components respectively.

９

reported that MAPbI3 possesses inherently benign grain boundaries,

which may act as traps in the film [65]. With these characteristics, the

high absorption coefficients combined with long electron-hole diffusion

lengths attributed from the shallow taps of the point defects and benign

grain boundaries, MAPbI3 generates the charge carriers then transports

the carriers to the electron or hole transport layers under the lights.

These properties are suitable for the active layer in the photovoltaic

devices.

１０

1.2.3 Various methods to grow the perovskite films

MAPbI3 is formed via the reaction between MAI and PbI2. Since the

fabrication process directly affect the optoelectronic properties and it is

challenging to grow the high quility films, various method for the

growing the films have been reported as follows.

One-step Spin coating

The spin coating is carried out by the casting the mixed solvent

containing MAI and PbI2 on to the desired substrates. In this process,

the solvent should resolve both MAI and PbI2 and the widely used

solvent is dimethylformamide (DMF). Once the solvent drops onto the

substrates, the perovskite film is immediately formed since the DMF

evaporates quickly. Therefore, it is often indicated that this process

causes poor morphologies, as well as the lots of the voids inside of the

film. This possibly cause degrading of the film quality and the device

performance in the solar cells. To retard the crystallization, improved

methods such as a dropping toluene or use of dimethyl sulfoxide

(DMSO) which evaporates slowly comparing to DMF [4,13].

Sequential (two-step) spin coating

In the sequential spin coating, the PbI2 layer is deposited on the

substrate by spin-coating and PbI2 deposited substrate is dipped into the

１１

MAI solution. MAI is resolved in a solvent, such as IPA, therefore MA+

+ I- ions intercalate into PbI2 layer. In order to fully convert the pre-

deposited PbI2 layer into the perovskite, the MAI should diffuse into

PbI2. Therefore, generally substrate heating is usually required for this

process, around 100 ℃. Another stagey is employing a porous PbI2

layer to enhance the diffusivity of MAI.

Burshka et al reported this method to fabricate the perovskite solar

cell and the authors showed that the pin-hole free, compact perovskite

film which resulted in the PCE of 15% [2]. The formation of the

perovskite is possible since the MAI inter-diffuses into the PbI2 layer,

therefore the annealing process is typically required to form the

perovskite structure.

Vacuum process

Liu et al., reported vacuum processed perovskite by co-deposition of

PbCl2 and MAI [23]. In this process, MAI and the lead halide (PbI2 or

PbCl2) are co-evaporated in the vacuum chamber and the composition

is controlled by the deposition rate of the each source. It is reported that

the use of PbCl2 requires the post annealing process to form the

perovskite structure since the MAI should be supplied excessively to

satisfy the stoichiometry, and this excessive MAI should be removed

whereas the iodide perovskite using PbI2 can be formed without

annealing processes. Polander et al. reported fully vacuum–processed

１２

perovskite solar cells using various hole transporting layers (HTLs) and

found that 2,2',7,7'-tetra(N,N -di-p -tolyl)amino-9,9-spirobifluorene

(spiro-TTB) possessing HOMO level of -5.3 eV resulted in the higher

solar cell performance with VOC of 1.07 V and PCE of 10.9% than

spiro-2’,7,7’-tetrakis-(N,N-di-p-methoxyphenylamine)-9,9’-

spirobifluorene (spiro-OMeTAD), possessing the HOMO level of -5.0

eV which showed VOC of 0.85 V and PCE of 7.7%. [31]. O.

Malinkiewicz et al., reported the similar process, vacuum co-deposition,

with MAI and PbI2. The authors introduced the organic charge

transporting layers, PEDOT:PSS/poly(N,N′-bis(4-butylphenyl)-

N,N′-bis(phenyl)benzidine (poly-TPD) for the hole-extraction and

6,6)-phenyl C61-butyric acid methyl ester (PCBM) as the electron-

extraction layer [24].

Vapor state reaction

Q. Chen et al., reported the perovskite formation by “vapor-assisted

solution process” (VASP) [15]. The PbI2 layer was previously deposited

on the films by solution process and then the substrate was exposed to

MAI vapor in nitrogen-filled glove box. The MAI powder was spread

out around the PbI2 coated substrates with a Petridis covering on the

top and heated at 150 °C for desired film. In this work, the authors

reported a planar structure perovskite solar cell with PCE of 12.1%. ,

X. Zhu et al., reported that vapors state reaction of MAI with inorganic

１３

mixed halide, PbCl2 and CsCl2. In this work the PbCl2 without and

with Cs-substituted films was deposited on by a thermal evaporation

[17]. During PbCl2 deposition, the CsCl was simultaneously evaporated

obtain the PbCl2 films with different compositions. The mixed halide

films were transferred into the nitrogen-filled glovebox. The MAI

powder was placed into a 70 mm × 70 mm aluminum reactor to form a

∼0.5 mm thick uniform and compact thin film. Then, the samples were

placed on a layer of MAI powder, and heated to 150 °C for 20 min to

ensure transformation of PbCl2 and CsCl into perovskite, similar to the

methods “VASP” reported by Q. Chen [15]. Finally, the perovskite

samples were transferred into a petri dish, cooled for 30 min, washed

with 50 mL isopropanol, and dried by flowing nitrogen gas. In this

work, the authors demonstrated the PCE of 20.13 %.

１４

Figure 1.2 Schematic figure for the various methods to fabricate the

MAPbI3 perovskite films

１５

1.2.4 Basic characterization of perovskite solar cell solar cell

Current density-voltage characterization

The short-circuit current density is defined as the current extracted at

applied external voltage is zero and the open circuit voltage is defined

as the photo voltage where the extracted current is zero under

illumination. The maximum output power (PMAX) is defined by the

maximum area given by photo current × photo voltage (JMAX × VMAX ).

The fill factor, FF, is defined as follows.

SCOC

MAXMAX

JVJV

FF××

=

The power conversion efficiency (PCE, ηp ), consequently, is

defined as follows where Pinput correspond to the power of incident

light typically 100 mW/cm2 for the AM1.5G condition.

input

SCOC

input

MAXp P

FFJVPP ××

==η

Incident photon-to-current efficiency

The incident photon to current efficiency (IPCE) or external quantum

efficiency (EQE) is the ratio of number of incident photons over

number of collected electrons. Since the photo generated current is

dependent on the wavelength it is described as follows.

１６

)(#)(#

(%)λ

λphotonsincidentof

electronsofIPCE =

The short circuit current at AM 1.5G, therefore can be calculated

from the measured IPCE as follows.

sourcelighttheofnsityintelighttheI

dIIPCEqhcJ MAXSC

:

)()(

0

0, λλλλλ

∫=

1.3 Vacuum process

1.3.1 General theory

The vacuum deposition process is typically conducted below 10-4 torr

and the vacuum levels are categorized as low vacuum (~10-3 torr), high

vacuum (10-7 torr) and ultra-high vacuum (~10-9 torr). The sources are

heated in vacuum and when they reach to the vapor pressure, the

sources start to vaporize. The vaporization can be described as Hertz–

Knudsen vaporization equation as follows.

0.5 1/ (2 ) ( * )secdN dt C mKT p pπ − −= −

,where dN is number of evaporating atoms per surface area, C is

１７

constant that depends on the rotational degrees of freedom in the liquid

and the vapor, p* is vapor pressure of the material at temperature, p is

pressure of the vapor above the surface, k is Boltzmann’s constant, T is

absolute temperature and m mass of the vaporized species [66]. The

vaporized molecules would collide to the substrate, which results in the

deposition forming films on substrate.

1.3.2 Equipment for vacuum process

Evaporator

The nitrogen filled glove box integrated vacuum chambers (VEV-502,

VTS corp.) are used to fabricate perovskite films as well as other layers

in the solar cells. One of the chamber possess 7 sources where hole

transporting layers and the perovskite are deposited and another

chamber possess 3 sources where electron transporting layers and

electrode metal layers are deposited. All of the sources are able to be

loaded through the glove box without exposing to air. The distance

from the source to the substrate is approximately 40cm and the

substrate rotates which ensure the uniformity of the deposited films.

１８

Figure. 1.3. The picture of glove box integrated vacuum evaporator

(VTS corps.) systems used for the perovskite and other layers.

１９

Thickness monitor

The thickness and deposition rate can be measured by quartz crystal

microbalance (QCM) which measures the adsorbed mass by the

frequency change. The relation between adsorbed mass and frequency

change is represented by Sauerbrey equation [67].

202

q q

ff m c mA ρ µ

∆ = − ∆ = − ∆

, where f0 is resonance frequency (Hz), Δf is frequency change (Hz),

Δm is mass change (g), A is pezoelectrically active crystal area, ρq is

density of quartz (g/cm3) and μq is shear modulus of quartz of AT-cut

crystal (2.947 ×1011 g·cm-1·S-1). For the simplicity, c is used as

“sensitivity factor” for the QCM. The frequency change appears as the

adsorbed thickness or deposition rate through the thickness monitor

(CRTM-9000, ULVAC). The frequency change and adsorbed mass are

in the linear relation when the frequency change is less than ~5%,

hence the QCMs are replaced around the frequency change is ~2%.

Vacuum gauges

For the low vacuum, from 1 atm to 10-3 torr, the pirani gauges are used.

The high level vacuum is monitored by ionization gauge (VARIAN 572,

Agilent). It reads the current generated by the hot electrons from the

filament, which would collide with gas molecules between the filament

２０

and electrodes in the ionization gauge. The pressure can be measured

from the current as follows.

c

g g

IpS I

=

, where p is pressure Ic is ion collector current Ig is electron emission

current and Sg is gas sensitivity factor. Sg was set to default value, 1,

corresponding to nitrogen and air. However, if MAI molecule affect the

ion collector current the gas sensitivity factor would be changed,

presumably larger than 1 since its polarity [68], hence the working

pressure of MAI in this research are nominal values.

Vacuum pumps

The two types of pumps are used for the evaporator, rotary and

cryogenic pump. The rotary pump is used to reach the vacuum level

around 10-3 torr. Once the vacuum level reaches this value, the

cryogenic pump (C200, Suzuki Shokan) is used which adsorbs the gas

molecules with the temperature ~10 k. The temperature is maintained

by helium compressor (A200, Suzuki Shokan) and water cooling chiller.

This pump is quite robust enough to reach to the base pressure of ~10-7

torr under using the MAI evaporation more than a year.

２１

Chapter 2

Vacuum processed perovskite solar cells

2.1 Introduction

In the perovskite solar cells, the one of the critical issues, as

mentioned in the chapter 1.1.1, is the reproducibility. The main reason

for this is that the fabrication of the reproducible film is challenging

due to the characteristics of the perovskite films such as non-uniform

surface, voids in the films, pinhole formation and the sensitivity to the

air/moisture. Therefore, to enhance the reproducibility, as well as to

enhance the film quality, various attempts have been introduced such as

the solvent modifications [4, 12-16], varying the post treatments of the

films [17-22]. The vacuum process has a potential for high

reproducibility and controllability due to solvent free process combined

with controllable growth parameters in a clean and inert environment

[23-28]. Another important issue in the perovskite solar cells is HTLs

since the most widely used HTL, spiro-OMeTAD possesses a few meV

lower HOMO level than the VB edge of the MAPbI3 and MAPbI3-xClx

[69]. Nonetheless, there have been reported high performance

perovskite solar cells using this HTL, spiro-OMeTAD, there is still

room for further improvement in terms of the energy level matching.

２２

The level mismatching would cause a reduction of the VOC, due to the

lowering in built-in field and the thermionic loss during the hole

extractions as mentioned in the chapter 1.2.3. In this regard, there have

been various reports to find the HTL s in order to achieve high

performing perovskite solar cells [70-775]. However, most of the

devices shows the lower performances than spioro-OMeTAD used

devices except Poly(triarylamine) (PTAA) used device which shows the

PCE of 16.7% [4]. Polander et al. reported all vacuum−processed

perovskite solar cells using various HTLs and found that 2,2',7,7'-

tetra(N,N -di-p -tolyl)amino-9,9-spirobifluorene(spiro-TTB) possessing

HOMO level of -5.3 eV resulted in the higher solar cell performance

with VOC of 1.07 V than spiro-OMeTAD possessing the HOMO level of

-5.0 eV with VOC of 0.85 V. In this regard, one can note that the HOMO

level matched HTMs are able to result in the higher performances than

spiro-OMeTAD used perovskite solar cells.

In these regards, research on the stable process for the perovskite with

efficient charge transporting layer (CTL) can be two important issues in

the perovskite solar cell community. In this chapter, we report all-

vacuum processed perovskite solar cell employing efficient HTL

minimizing the energetic loss. All of the layers, including the

perovskite layer and other CTLs deposited without breaking the

vacuum. In addition, the widely used HTLs, MoO3 and NPB, in the

vacuum processed OLEDs and OPVs field are successfully applied into

２３

the perovskite solar cell system.

2.2 Preliminary data

The mixed halide perovskite, which is typically formed with the

reaction between MAI and PbCl2, consequently forms the MAPbI3-xClx

is the widely used material next to MAPbI3. In order to satisfy the

stoichiometry in the perovskite, the molar ratio of MAI:PbCl2 should be

approximately 3:1 which requires the excessive MA+ and Cl-

components should be removed after the deposition. For the removal

process, annealing the substrate at 150°C is carried out. The vacuum

co-deposition of the mixed halide perovskite was fabricated

preliminarily without the information for the deposition rate of the MAI,

since the poor adsorption property onto the substrate, which makes the

measuring the film thickness is difficult. In this experiment, the

deposition rate of PbCl2 was set to 0.5 Å/s which had been calibrated

before the deposition using a profilometry (KLA-Tencor Alpha-Step

IQ) where the deposition rate of MAI was 0.5 Å/s using a nominal

tooling factor, but maintained the deposition rate constant during the

deposition. After the deposition, the substrate was annealed in the

nitrogen ambient at 150℃ as reported elsewhere [clx] . The perovskite

layer was deposited in the bell-jar evaporator (Selcos) as shown in Fig.

２４

2.1a. The absorbance and X-ray diffraction (XRD) patterns of the

fabricated perovskite films correspond to the previously reported data

[76-78] (Fig. 2.1b and 2.1c). However, most of the device are short-

circuited as featured in J-V curve shown in Fig. 2.1d, regardless the

composition of the perovskite and the device structures. The origin was

not clear, but we presume that the technical problems may cause the

poorly performing device. For example, air exposure before/after the

perovskite layer deposition, short distance from the source to the

substrate and oil contamination from the diffusion pump are inevitable

for the bell-jar evaporator used in this experiment. Therefore, we

started to use the evaporator (VEV-502, VTS corp, Fig. 1.3), which

includes glove box integrated with two vacuum chambers, various

source and cryogenic pumps for high vacuum, for the perovskite layer,

other CTLs and metal.

２５

Figure 2.1 (a) The evaporator used for the preliminary experiment

(Bell-jar, Selcos) (b) Absorbance and (c) XRD data of the MAPb3-xClx

film. (d) Device fabricated with different structures using the

evaporator

２６

2.3 Film characterization

2.3.1 Film fabrication

Under the enhanced experimental condition using the evaporator

shown in Fig. 1.3 as mentioned above, the perovskite layer was formed

by co-deposition of MAI (Jida Ruibo Optoelectronics Tech.) and PbI2.

Unlikely to the mixed halide perovskite (e.g. MAPbI3-xClx), any post-

treatment such as heating is not required for the triiodide perovskite

using vacuum co-deposited.

A QCM (INFICON, 5Mhz) was located on the PbI2 source which

reads only the deposition rate of PbI2. This QCM was enclosed by the

aluminum foil to exclude the effect of the MAI, since its gas-like

behavior in the vacuum chamber. The other QCM was located at the

near the substrate which enables to read the deposition rate of MAI and

also PbI2. This QCM, therefore, is able to read the growth rate of the

perovskite, as well as the rate of MAI since the deposition rate of PbI2

is fixed constant using the other QCM located near the PbI2 source. The

base pressure of the vacuum chamber was about 1×10-6 torr and this

pressure was increased up to 2×10-5 Torr when MAI is deposited.

Regardless this increase of the vacuum pressure, the deposition rates of

the PbI2 and MAI are maintained constant during the co-deposition. To

２７

form the 320 nm thick of the perovskite, 150 nm thick of PbI2 was

deposited which was read by the QCM 1. The total deposition tame was

about 1 hour.

2.3.2. Absorbance and XRD patterns

Fig. 2.2(a) shows the absrobance of a 320nm thick MAPbI3 film

grown on a glass substrate, which were recorded by a VARIAN Cary

5000 UV–vis spectrophotometer. The absorbance corresponds to

previously reported data in visible light range, which implies that the

perovskite structure via vacuum co-evaporation described in the

chapter 2.4 was successfully formed. Fig 2.2(b) shows the XRD

patterns of the fabricated MAPbI3 film grown on a ITO (150 nm)/MoO3

(5 nm)/NPB (20 nm) substrate. The peaks at 14.06˚, 28.26˚ 31.73˚,

40.55˚ and 43.13˚ correspond to (110), (220), (310), (224) and (314)

planes of the tetragonal form of MAPbI3 structure [76-78]. It is worth

noting that the high intensity of (110) and (220) peaks represent that the

highly preferred oriented of the perovskite film on the NPB layer,

which will be used as a HTM later. The preferred orientation, [110]

direction, is demonstrated in Fig. 2.3.

２８

Figure 2.2 (a) The absorbance of the 320 nm thick MAPbI3 film

grown on the glass substrate using vacuum co-evaporation. (b) The X-

ray diffraction (XRD) patterns of a 320 nm thick MAPbI3 film grown

on the ITO/MoO3(5 nm)/NPB (20 nm) substrate using vacuum co-

evaporation.

２９

Figure 2.3. The demonstration of MAPbI3 perovskite in the preferred

orientation along (110) direction. This figure was drawn by Diamond

3.2 program.

３０

2.3.3 Surface ananlyzes

Fig. 2.4(a) shows the AFM (XE-100, Park Systems) image of the

MAPbI3 perovskite film grown on ITO/MoO3 (5 nm)/NPB (20 nm)

substrate with noncontact mode. The root mean squre (RMS) roughness

is 9.7 nm and the maximum peak to valley value is 70 nm from the 320

nm thick film. These values can be consiered small considering the

total thickness of the film. Fig. 2.4(b). shows the red line profile of Fig.

2.4(a), which demonstrates the rms value is about 10 nm again. This

surface can be considered smooth since the rms values reported in

MAPbI3 values are 10 nm ~ 170 nm in the solution perovskite, and it is

often stated that rms value of 10 nm as smooth in the published paper

[79,80]. In this regard, the fabricated perovskite by vacuum co-

evaporation as described in the chapter 2.2, was succesed with the

smooth morphology.

３１

Figure 2.4. (a) The atomic force microscopy (AFM) topographic

image of the 320nm thick MAPbI3 perovskite grown on the

ITO/MoO3(5 nm)/NPB(20 nm) substrate using the vacuum co-

evaporation. (b) The red line profile of the topographic image in Fig.

2.4a.

３２

2.4 Device performance

150 nm -thick ITO coated glass substrates were cleaned with acetone

and isopropyl alcohol. All the materials except the perovskite, were

thermally evaporated at the base pressure of <1×10-6 Torr without

breaking the vacuum. The vapor pressure was increased up to 2×10-5

torr when MAI is deposited similar to the previous reports [24, 31], as

mentioned in chapter 2.2. The deposition rate of MAI was monitored in

real time using a crystal thickness sensor, QCM, and was maintained

constant during the deposition, while the deposition rate of PbI2 (Alpha

Aesar) was kept constant with 0.5 Å/s during the co-deposition. The

perovskite layer was formed without any annealing process. The NPB

(Nichem), C60 (SES Research) and BCP (Nichem) layers were

deposited at a rate of 1 Å/s, and the MoO3 (Sigma Aldrich) and the Al

layers were deposited at the rates of 0.5 Å/s and 4 Å/s respectively. The

deposition rates were calibrated using a profilometry (KLA-Tencor

Alpha-Step IQ) before the deposition. The active area of each device

was 4 mm2 and 72 devices were fabricated from 6 different batches to

average the cell performance. After the evaporation, the devices were

encapsulated using an epoxy resin with glass cans in an N2

environment. The current density-voltage (J–V) characteristics were

measured under simulated AM 1.5G sunlight of 100 mW/cm2 using

３３

Newport (91160A) solar simulator, and recorded using a Keithley 237

source measurement unit at room temperature. The light intensity was

calibrated using a standard Si-solar cell (NREL). The scan step of J-V

characteristics was 0.02 V with 0.3 seconds of interval time for each

step, and the scan direction was from negative to positive voltage. The

incident photon–to–electron conversion efficiency (IPCE) was

measured using a 1000 W Xe lamp combined with a monochromator

and its intensity was calibrated with a Si photodiode. The IPCE was

measured without light bias under the short circuit condition.

2.4.1 Effects of HTLs

2.4.1.1 Function of the MoO3 layer

Transition metal oxides(TMOs), such as MoOx, ruthenium oxides

(RuOx), vanadium oxides (VOx), tungsten oxides (WO3) and rhenium

oxide (ReO3) with high work functions are widely used hole injection

layers (HILs) in OLEDs/OPVs [81]. These high electron affinity TMOs,

especially MoO3 which shows the fermi level of 6.01 eV in an ultra-

high vacuum deposition on the indium tin oxide (ITO) substrate, causes

the electron transfers from the ITO to the MoO3 with the positive level

of the vacuum level shift. Therefore the hole injection barrier, which

can be determined the difference between the Fermi level of the anode

and HOMO level of the HTM can be reduced. The hole injection

３４

barrier should be considered in the perovskite solar cells similarly to

the conventional OPV devices since the holes accumulated at the

junction of the anode/HTM would cause the positive electric field in

the forward bias. Therefore, the photo-generated holes flows against the

positive field in hole extraction barrier which significantly reduces fill

factor (FF). Also, forming an ohmic contact between the anode and

HTM results in increase the built-in field in the fabricated photovoltaic

device.

To support these functions of MoO3 in the perovskite solar cells,

the device without MoO3 was fabricated. The device structure is

ITO(150 nm)/NPB (20 nm)/perovskite (320 nm)/C60 (50 nm)/BCP (8

nm)/Al (100 nm). The schematic energy diagram for this structure is

shown in fig. 3.2. The J–V characteristic of the fabricated device is

shown in fig. 3.3. The device showed PCE of 4.2%, JSC of 15.0

mA/cm2, VOC of 0.84 V and FF of 0.33. From the J–V characteristic, it

can be seen the s-kink in the forward bias region which implies the

injection barrier at the ITO/NPB interface as previously discussed in

chapter 3.1. Also, the current density is almost zero from 1 V ~ 3 V in

the forward bias region because of the injection barrier between

ITO/NPB which is known as about 1.2 eV ~ 1.7 eV. As previously

mentioned in chapter 3.1, the MoO3 forms an ohmic contact increasing

the built-in field in the devices. Therefore, without the MoO3 layer the

built-in field would be reduced and the overall solar cell performance

３５

might be degraded. Regarding these factors above, the MoO3 is a

crucial layer in the fabricated perovskite solar cell for the hole

extraction layer to enhance the solar cell performances.

３６

Figure 2.5. The current density−voltage (J−V) characteristic of the

device without the MoO3 layer. The device structure is ITO(150

nm)/NPB (20 nm)/perovskite (320 nm)/C60 (50 nm)/BCP (8 nm)/Al

(100 nm).

３７

2.4.1.2 Function of the NPB layer

In MAPbI3, the free carriers are dominantly generated under the light

bias [52,53]. These generated charge carriers diffuse to either anode or

cathode. Therefore it would be needed the charge blocking layers to

collect the desired charges at the electrodes selectively. In order to

block the undesired charges, the energy levels should be taken into

account. In the perovskite solar cell system, since the CB of MAPbI3 is

-3.9 eV, the electron blocking layer should possesses the CB or LUMO

level lower than -3.9 eV. In this regard, the NPB layer, which possesses

the LUMO level of -2.2 eV, will act as an efficient electron blocking

layer enhancing the charge collection efficiency. Fig. 2.6 shows the

J−V characteristic of the device without NPB layer. The device

structure is ITO (150 nm)/MoO3(5 nm)/perovskite (320 nm)/C60 (50

nm)/BCP (8 nm)/Al (100 nm). The device showed the PCE of 6.4%,

JSC of 11.1 mA/cm2, VOC of 0.78 V and FF of 0.74.

３８

Figure 2.6. The current density−voltage (J−V) characteristic of the

device without the NPB layer. The device structure is ITO(150

nm)/MoO3(5 nm)/perovskite (320 nm)/C60 (50 nm)/BCP (8 nm)/Al

(100 nm).

３９

As previously mentioned in this chapter, the one of the possible reason

for the lowering these parameters can be caused by the absence of the

charge-blocking layer. However, since the perovskite layer is known to

be sensitive to the under layer there might be differences in the

crystallinity or the growth modes with the different sub-layers. Fig. 3.6

shows the X-ray diffraction (XRD) patterns of the perovskite film

grown on the ITO/MoO3 (5nm) and ITO/MoO3 (5nm)/NPB (20 nm)

substrates respectively. Both films show the (110) and (220) as the

main peaks at 14.06˚, 28.26˚. However the peak main intensities exhibit

a significant difference with the different substrates. The low peak

intensity on the MoO3 sub-layer indicates that the crystallinity of the

perovskite film is somehow but still oriented along [110] direction.

However, the relative very strong intensity of the NPB sub-layer show

that the high intensity of the main peak which implies the much higher

crystallinity than on the MoO3 sub-layer. The reason why the

crystallinity is degraded on the MoO3 layer is not clear yet. However, it

can be inferred that the NPB layer can be used for a higher crystallinity

in the perovskite film. After publication of this work, P. Schulz et al.,

reported that, the absence of the organic layer between MoO3 and the

perovskite results in forming MoO2. [82] In this work, it is discussed

that this MoO2 induces the interface trap states which hinders the

charge extraction hence resulting in the low FF and JSC. This work also

supports that the NPB layer is necessary for the successful fabrication

４０

of the perovskites solar cell.

４１

Figure 2.7. X-ray diffraction (XRD) patterns of the fabricated

perovskite film on the different substrates.

４２

2.4.2 Device with MoO3/NPB layer

As discussed previously, in chapter 2.4.1, both MoO3 and NPB layers

are needed to enhance the device performance. Therefore, the device

was desiged as ITO/MoO3 (5 nm)/NPB (20 nm)/Perovskite (320

nm)/C60 (50 nm)/BCP(8 nm)/Al(100 nm). The MoO3 and NPB layers

were used as a hole extraction layer, and the perovskite was used as the

light absorbing layer and charge transport layer with its high absorption

coefficient and ambipolar charge transport characteristic as discussed in

chapter 1.2.1. The C60 layer was used as the electron transporting layer

and the bathocuproine (BCP) layer was used as a buffer layer which

mainly acts 1) preventing Al diffusion in to C60 2) preventing a dipole

formation between C60 and Al layer and 3) formation of an ohmic

contact at their interface.[83-85] . The schematic diagram of the relative

energy levels of each layer in this device is shown in fig. 2.8.

４３

Figure 2.8. The schematic diagram of the relative energy levels of

each layer in the device without MoO3 layer. Note that the energy

levels are represented against the vacuum level.

４４

The 72 devices were fabricated and measured repeatedly from 6

different batches with identical structure. The best performing device

showed the JSC of 18.1 mA/cm2, the VOC of 1.12 V and the FF of 0.68

resulting in the PCE of 13.7% with the average values as JSC of 18.5

mA/cm2, the VOC of 1.05 V and the FF of 0.57 resulting in the PCE of

11.1%. The J−V characteristic and IPCE spectrum of the best

performing device is shown in Fig.2.9 and the parameters are

summarized in table 2.1.The calculated JSC by integrating the IPCE

spectrum using the following equation was 17.8mA/cm2, which agrees

well with the JSC obtained from J−V curve implying that the spectral

mismatch is less than 2%.

sourcelighttheofnsityintelighttheI

dIIPCEqhcJ MAXSC

:

)()(

0

0, λλλλλ

∫=

It is interesting to note that the device exhibits the high VOC of 1.12 V,

which is one of the highest VOC values reported up to now [2,3,5,8].

The high VOC indicates that MoO3/NPB structure works efficiently as a

hole extraction layer minimizing the voltage loss. This high VOC is

attributed to the good alignment of the HOMO level of NPB to the VB

edge of MAPbI3. The open circuit voltage is higher than the devices

with spiro-OMeTAD (VOC of 0.85 V and HOMO level of -5.0 eV) and

４５

spiro-TTB (VOC of 1.07 V and HOMO level of -5.3 eV) as the HTL in

the vacuum processed perovskite solar cells reported in the reference

24. One can note that the open circuit voltage increased as the HOMO

level of the HTM is getting closer to the VB edge of MAPbI3 in the

vacuum processed solar cells as expected.

４６

Figure 2.9. (a) The current density-voltage (J−V) characteristics and

(b) incident photon-to-electron conversion efficiency (IPCE) spectrum

of the best performing perovskite solar cell.

４７

PCE (%)

JSC (mA/cm2)

VOC (V) FF

Best device 13.7 18.1 1.12 0.68

Average 11.1 18.5 1.05 0.57

Table 2.1 Perovskite solar cell performance of the best performing

device and the average values of the 72 devices from 6 different

batches.

４８

2.4.2 Reproducibility

Fig. 2.10 displays the histograms of (a) PCE (b) JSC (c) VOC and (d) FF

of the 72 devices from 6 different batches. It is shown that the JSC and

VOC are distributed within 17.2 mA/cm2 ~ 19.5 mA/cm2 and 0.98 V ~

1.12 V respectively, exhibiting small deviation within approximately

7% from the average values. The FF values show a relatively wide

distribution, 0.39~0.68, hence causing the similar distribution of PCE

values, 7.1%~13.7%. The origin of this variation of the FF’s is not clear

yet. Still the variation of the performance of the solar cells fabricated

using the vacuum process is fairly narrow compared to the normal

solution processed perovskite solar cells

４９

Fig. 2.10. Histogram plots of the solar cell parameters; (a) Power

conversion efficiency (b) short-circuit current (c) open circuit voltage

(d) fill factor of the 72 devices from 6 different batches.

５０

2.4.3 Hysteresis

Fig. 5.1 shows the J–V characteristics with the different scan

directions. The scan step was 0.02 V and the step of the each voltage

was 300 ms. As shown in fig. 5.1, the vacuum processed perovskite

solar cell showed the hysteresis. The main difference caused from the

FF, which showed 0.62 in forward scan, short circuit to open circuit

direction, and 0.56 in the reverse scan, open circuit voltage to short

circuit direction. On the other hand, the JSC values were 18.1 mA/cm2

and 17.1 mA/cm2, VOC values were 1.09 V and 1.07 V for the forward

and reverse scans respectively, which showed the less difference than

FF. These different scan directions result in the differences of 6% in

JSC, 1.9% in VOC, 11% in FF and 17% in PCE. It has been reported that

in the mesoscopic structure the hysteresis is observed as well as in the

planar heterojunction structure the hysteresis might be observed. The

FF degradation in the reverse scan implies that the generated charge

carriers are not extracted to the external circuit and the reasons are not

clear yet. The plausible origins would be discussed in the next chapter.

,

５１

Fig. 2.11. The current density-voltage (J−V) characteristics of a

device with the MoO3/NPB layer for different scan directions. The

device structure is ITO (150 nm)/MoO3 (5 nm)/NPB (20 nm) MAPbI3

(320 nm)/C60 (50 nm)/BCP (8 nm)/Al (100 nm).

Table 2.2 The solar cell parameters of the device with the

MoO3/NPB layer shown in fig. 5.1 with different scan directions.

Scan directions

PCE (%)

JSC (mA/cm2)

VOC (V) FF

Forward 12.1 18.1 1.09 0.62

Reverse 10.3 17.1 1.07 0.56

５２

To investigate the effect of the HTLs on the hysteresis behaviour,

the devices without MoO3 or NPB layers were measured in different

scan directions of the J–V characteristic. As shown in fig. 5.2 the

device without MoO3 layer shows the hysteresis as well as in the case

of the device shown in fig. 5.1, with the MoO3/NPB layer. The solar

cell parameters were differed with difference scan directions, JSC of

2.3% of VOC, 10% of FF and 12.5% of PCE. In addition, the device

without NPB layer showed the hysteresis as well shown in fig. 5.3. In

the similar way to the device without MoO3 layer, the differences of the

JSC, VOC, FF and PCE were 4.5%, 12%, 1.3% and 7.8% respectively. It

is interesting to note that unlike to the previous cases, the FF showed

the least variation in different scan directions. In addition, the FF value

in the case of the without NPB, is the highest among the devices with

MoO3/NPB or NPB layer in the fabricated solar cell. The value 0.74

and 75 in the fig. 5.3, is even higher than the best performing device

with PCE of 13.7% which shows the FF of 0.68. Regarding these facts,

the higher fill factor gives smaller hysteresis, and the parameter

affecting the fill factor would be the one of the crucial reasons for the

hysteresis.

５３

2.7. Summary

In summary, we have demonstrated the fully vacuum-processed

perovskite solar cells with high efficiency and reproducibility. The best

performing device showed PCE = 13.7%, JSC = 18.1 mA/cm2, VOC =

1.12 V, FF = 0.67 and the reproducibility is higher than the normal

solution processed perovskite solar cells. Importantly the considerably

high VOC value was obtained by employing the MoO3/NPB layer as the

hole extraction layer. This high VOC is attributed to the alignment of

HOMO level of the NPB to the VB edge of MAPbI3. The fact that the

MoO3/NPB layer is widely used in OPVs/OLEDs implies that

technologies developed for vacuum processed OLEDs/OPVs can be

successfully applied in perovskite solar cells to accelerate the

development of the perovskite solar cells. In addition, since the vacuum

deposition process is free from the materials restriction by their

solvents, it is a promising tool to apply the various charge transport

materials in the layered structures.

５４

Chapter 3

Composition-controlled organometal halide

perovskite via MAI pressure in vacuum co-

deposition process

3.1 Introduction

The vacuum co-deposition is one of the promising methods for growing

perovskite films because the high purity of materials and easy

controllability without effects of solvents or the atmosphere ensure the

high quality of the films with good reproducibility [23–29]. In addition,

since the vacuum process does not affect under-layers, whereas the

solution process limits the under-layers because of the solvent effect of

the top layer, it can be readily employed to optimize the device

architectures with various charge transport layers. Unlike vacuum

sequential deposition, the vacuum co-deposition process does not

require a substrate heating process; thus, a room-temperature process is

easily achievable. However, considering there have been over

thousands papers published in the organic/inorganic hybrid perovskite

field, a few groups have reported the vacuum co-deposition process

[29–34]. The reasons are probably the difficulties of controlling MAI in

５５

the vacuum chambers, because MAI increases the vacuum pressure and

it is difficult to calibrate the thickness of MAI and, consequently, the

deposition rates because of the poor adsorption property of MAI

[24,25,30-32,43-48]. Effective control of deposition rates of MAI

would be key for the growth of the perovskite films, because the

compositions significantly affect the properties of the perovskite films.

With the solution processes, there have been various reports about the

effects of the compositions on the properties of the perovskite films,

such as the morphologies, grain sizes, electrical and optical properties,

and device performances [86-91]. In these works, the compositions

were controlled using different precursor ratios between MAI and PbI2

in one-step depositions or different concentrations of MAI solutions in

two-step depositions. However, in the vacuum co-deposition process,

only limited information about the effect of the composition on the

perovskite films is available [39,46,91]. Therefore, introducing an

effective method to control MAI, and establishing a framework for the

effects of the compositions on perovskite in a vacuum process would be

significant.

In this chapter, we report an effective method for controlling the

compositions of the perovskite films in a vacuum co-deposition process

using the working pressures monitored by a vacuum gauge. The

compositions of the perovskite were able to be adjusted with the

working pressures of MAI. Using this method, the properties of the

５６

films with various compositions were revealed; they showed significant

differences under the different working pressures. Furthermore, we

show the application of the fabricated perovskite films to solar cell

devices composed of molybdenum oxide MoO3, NPB and C60 layers

as the charge transport layers via the all-vacuum process. Optimum

pressure of MAI resulted in, the maximum power conversion efficiency

of 14.5% with good reproducibility.

3.2 Experimental

3.2.1 Perovskite evaporation system

Fig. 3.1 shows the schematic illustration for the vacuum evaporator

system for the perovskite films. Perovskite layers were grown by co-

deposition of PbI2 and MAI where the deposition rate of PbI2 was fixed

at 0.5 Å/s and the pressures of MAI was varied as a deposition

parameter. The deposition rates and vacuum pressures were monitored

by a crystal thickness sensor (INFICON, 5 MHz) and an ionization

gauge (Varian 572, Agilent Technologies), respectively. A 2 nm thick

PbI2 layer was deposited before heating up the MAI source. When MAI

was heated in the chamber, the vacuum pressure (working pressure)

was gradually increased, and it was heated until the vacuum pressure

５７

reached 0.6, 1.7, 3.2, 4.1, 5.1, 6.5, and 8.3 × 10−5 torr. The relative gas

sensitivity (gas correction factor) was set to 1.0 which corresponds to

the value for N2. The pressures were maintained within ±0.1 torr from

the setting pressures for 1 h during the deposition by controlling the

input power of the heating module of the MAI source manually. The

perovskite layers were formed without annealing.

５８

Fig. 3.1. Schematic Fig. of the vacuum co-deposition system for the

perovskite layer.

５９

3.2.2 Film characterization

The UV-vis absorption and reflectance spectra of the films were

recorded with a Varian Cary 5000 UV-vis spectrophotometer. AFM

topographic images of the perovskite film on the ITO/MoO3/NPB

substrate were taken using a PSIA XE-100 scanning probe microscope

in noncontact mode. XRD measurements were performed on a D8

Advance diffractometer (Bruker) using Cu Kα radiation. The structure

of the devices was ITO/MoO3 (5 nm)/NPB (20 nm)/perovskite/C60 (50

nm)/BCP (8 nm)/Al (100 nm), with an active area of 4 mm2. After the

deposition, the devices were encapsulated using an epoxy resin with

glass cans in a N2 environment. The J-V characteristics were measured

under 100 mW/cm2 of simulated AM 1.5G sunlight using a Newport

(91160A) solar simulator, and recorded using a Keithley 237 source

measurement unit at room temperature. The light intensity was

calibrated using a standard Si-solar cell (NREL). The general scan step

of the J-V characteristics was 0.02 V at intervals of 0.3 s for each step

(0.067 V/s), and the scan direction was from negative to positive

voltage. For the experiment involving changing the scan rates, the

various scan steps were 0.1, 0.08, 0.04, 0.005, and 0.001 V, which

resulted in scan rates of 0.333, 0.267, 0.133, 0.033, 0.017, and 0.003

V/s, respectively. The IPCE was measured using a 1000 W Xe lamp

combined with a monochromator, and its intensity was calibrated with

６０

a Si photodiode. The IPCE was measured without light bias under the

short-circuit condition using a lock-in amplifier with monochromatic

light from the chopped Xe lamp with 1 kHz. The EDX and SEM data

was measured using a MERLIN Compact (FE-SEM).

3.3 Results and discussion

3.3.1 Working pressure and the thickness

As shown in Fig. 3.1, the deposition rate of PbI2 was monitored by

sensor 1, and was fixed at 0.5 Å/s. When MAI is deposited, sensor 2

reads the rates of both PbI2 and MAI, while sensor 1 reads the rate of

PbI2. The vacuum pressure (working pressure) read by the ionization

gauge increases when MAI is thermally evaporated as previously

reported [24,30,31,39,46] The working pressure was maintained

constant by the controlling the temperature of the MAI crucible. Fig.

3.2 shows the thicknesses or the deposition rates of the perovskite films

fabricated under the different pressures with the fixed PbI2 deposition

rate for 1 h. Overall, the thicknesses increased from 180 to 320 nm as

the pressure increased. However, the thicknesses were maintained

constant when the working pressure increased from 4.1 × 10−5 torr to

６１

6.5 × 10−5 torr. The reason why there are plateaus in the figure is not

clear at this moment and requires further study.

６２

Fig. 3.2. Thickness and deposition rate of the perovskite films under

different working pressures deposited for 1 h with the same deposition

rate of PbI2 as 0.5 Å/s. The first data point (2 × 10−6 torr) represents

PbI2 only film

６３

3.3.2 Working pressure and structural property

The X-ray diffraction (XRD) patterns of the fabricated perovskite

films are shown in Fig. 3.3. The main peaks, (110) and (220) planes, of

the tetragonal phase (space group I4/mcm) of the perovskite appeared

at all the pressures as previously reported [34, 76-78]. The intensities of

the main peaks increased until the pressure reached 6.5 × 10−5 torr, and

then decreased at 8.3 × 10−5 torr, as well as the calculated crystal size

from the full-width at half-maximum (FWHM) as shown in Table S1.

In Fig. 2b, we show that the peak of the (001) plane of PbI2 appears at

the low pressures. The intensity of this peak of PbI2 gradually

decreased as the pressure increased, and this peak had almost

disappeared at 4.1 × 10−5 torr. Therefore, it can be inferred that at the

low pressures, the amount of MAI is insufficient to convert all the PbI2

to the perovskite and the perovskite and PbI2 are homogeneously

dispersed in the film.

６４

Fig. 3.3. (a) XRD patterns of the perovskite films under different

pressures on the indium tin oxide/MoO3 (5 nm)/NPB (20 nm)

substrate. (b) Magnified image of the XRD patterns in red square

region in Fig. 2a.

６５

3.3.3 Working pressure and absorbance

Fig. 3.4(a) shows the absorbance spectra of the fabricated perovskite

films on the glass substrates. The absorption edge of PbI2 is ~540 nm,

and the absorbance after this wavelength of the PbI2 film is attributed to

the reflectance of PbI2 as shown in Fig. 3.5 and in reference 37. Under

the low working pressures, 0.6–3.2 × 10−5 torr, the absorbance at 540

nm increased gradually causing the red shift of the absorption edges.

Accordingly, determining the absorption edges became difficult,

because the films were in homogeneously blended with PbI2 and

perovskite, as shown in Fig. 3.2. At the pressures of 4.1–6.5 × 10−5 torr,

the absorbances apparently increased for all wavelengths and the

absorption edges shifted to ~780 nm, indicating that perovskite with a

band gap of 1.55–1.60 eV was formed [92]. At the pressure of 8.3 ×

10−5 torr, the absorbance increased further owing to the increased

thicknesses. Fig. 3.4(b) shows the absorbance at 550 nm, near the

absorption edge of PbI2, at different pressures. As shown, the increase

in absorbance at 550 nm with the pressure indicated that the films were

optically changed from PbI2 to perovskite, and this tendency was

similar to the change in thickness shown in Fig. 3.2. Fig. 3.4(c) shows

the obvious color changes of the films caused by the shift of the

absorption edges.

６６

Fig. 3.4 (a) Absorbance of the perovskite film at various pressures on a glass substrate. The thicknesses are indicated in the brackets. (b) Absorbance at 550 nm of the fabricated films. (c) Color of the perovskite films. From the left, the films are PbI2 only, 0.6, 1.7, 3.2, 4.1, 5.1, 6.5, and 8.3 × 10−5 torr deposited films.

６７

400 500 600 700 8000

20

40

60

80

100 Glass/PbI2 (180 nm)

Refle

ctan

ce (%

)

Wavelength (nm)

Fig. 3.5. Reflectance of PbI2. The incident beam is tilted 7° toward the substrate

６８

3.3.5 Working pressure and morphological changes

Fig. 3.6 shows the atomic force microscopy (AFM) images of the

perovskite films grown on the same substrate shown in Fig. 3.2. As

shown in Fig. 3.6(a), the deposited PbI2 film shows a root mean square

(RMS) roughness of 7.5 nm and relatively high peak-to-valley value of

118 nm considering the thickness of the film, which is caused by

partially protruded areas. These values decreased with the incorporation

of MAI, and the RMS roughness and peak-to-valley values decreased

to 3–5 nm and 33–68 nm, respectively, in the low-pressure-fabricated

film shown in Fig. 3.7 and d. At the pressure of 4.1 × 10−5 torr, similar

to the tendencies of the thickness and absorbance, a radical change in

the morphology occurred with the increased particle sizes, and the

RMS values increased to ~12 nm. The particles seemed slightly sharper

at 5.1 × 10−5 torr with a negligible change in the RMS values, and the

particles apparently became large upon a further increase in the

pressure to 6.5 × 10−5 torr. At the pressure of 8.3 × 10−5 torr, the

particles appeared to have irregular shapes compared to the previous

cases and gave the highest RMS roughness values. As expected, an

overly excessive amount of MAI in the films caused the poor

morphologies of the perovskite films. The tendency of these changes in

the particle sizes corresponds to the calculated crystal sizes from the

XRD data, shown in Fig. 3.3. Additional information about the

６９

morphologies, RMS values, and three-dimensional images are

presented in Fig. 3.6, 3.7 and 3.8.

７０

Fig. 3.6. AFM images of the perovskite films with different pressures

grown on an indium tin oxide/MoO3 (5 nm)/NPB (20 nm) substrate:

(a) PbI2, (b) 0.6 (c) 1.7 (d) 3.2 (e) 4.1, (f) 5.1 (g) 6.5 and (h) 8.1 × 10−5

torr.

Fig. 3.7. 3D- AFM image of the perovskite films shown in Fig. 3.6

Fig. 3.8. Scanning electron microscopy image of the perovskite films

shown in Fig. 3.6. The scale bars represent 1 micrometer.

７１

3.3.6 Device performances

Fig. 3.9 shows the current density-voltage (J-V) characteristics of the

perovskite solar cells at the different pressures. The device structure is

indium tin oxide (ITO)/MoO3 (5 nm)/NPB (20 nm)/perovskite/C60 (50

nm)/bathocuproine (BCP) (8 nm)/Al (100 nm). The structure is

identical to that used in a previous work, where all the layers were

deposited by vacuum thermal deposition [30]. The MoO3 and NPB

layers are used as hole transport layers, the C60 layer is used as an

electron-transporting layer, and BCP is used as a buffer layer.

Interestingly, the performances differed significantly with the pressure

of MAI. As the pressure increased from 0.6 × 10−5 torr, the JSC

increased and showed a remarkable change at 4.1 × 10−5 torr, similar to

the previously discussed properties of the films. Therefore, it can be

inferred that perovskite was properly formed at certain pressures with

the appearance of considerable changes in the physical properties

discussed above, and the film works efficiently as an absorber in the

solar cells. Upon further increase of the pressure to 5.1 × 10−5 torr, the

JSC was maximized and decreased after this pressure. The open-circuit

voltage (VOC) values showed virtually no change until the pressure of

5.1 × 10−5 torr and then decreased after this pressure as well. The

values of fill factor (FF) showed a tendency similar to those of the JSC

and VOC, exhibiting the highest value at the pressure of 5.1 × 10−5 torr;

７２

hence, the PCE was maximized at this pressure. The best-performing

device showed a PCE of 14.1%, JSC of 19.7 mA/cm2, VOC of 1.05 V,

and FF of 0.69. Fig. 3.9(b) shows the incident photon-to-current

conversion efficiency (IPCE) spectra at the different pressures. The

different shapes in the IPCE spectra were probably caused by the

different optical properties, refractive indices, and morphologies

[50,51]. The values of experimental JSC and integrated JSC from the

IPCE spectra are tabulated in Table S3. At the pressure of 0.6 and 5.1 ×

10−5 the deviations of these values were less than 5% implying that the

spectral mismatch is less than 5%. However, at the other pressures,

especially at 3.2 × 10−5 torr and 8.3 × 10−5 , the experimental JSC values

showed relatively large deviations. The reason is not clear yet, but it

may be attributed to the changes in the recombination mechanisms in

the different compositions.

７３

Fig. 3.9 (a) Current density-voltage (J-V) characteristics of the fabricated devices at the different pressures. (b) IPCE spectra of the fabricated perovskite solar cells.

.

７４

3.3.7 Origin of low IPCE

Fig. 3.10 shows the IPCE spectrum of selected from Fig. 3.9

representing the low pressure, middle pressure and high pressure

fabricated device for 1.7, 5.1 and 6.5 × 10−5 torr respectively . As

shown, the low pressure fabricated device shows small IPCE around

long wavelength region, λ= 650 nm, whereas the high pressure one

shows relatively small IPCE around short wavelength region, λ= 450

nm . The simulated electric fiend intensity using transfer matrix method

based on the refractive indices of the materials composed the device is

shown in Fig. 3.10 (b) and (c). The case for λ=450 nm shows Beer-

Lamber regime and the λ=650 nm shows cavity interference regime due

to the absorption differences of the perovskite layer. It is know that the

low pressure, MAI deficient condition cause the perovskite as n-type.

Therefore, the holes generated from the near HTL side maybe collected

whereas the holes generated near around ETL from the long

wavelength, 650 nm may recombine during the transporting the n-

type perovskite layer [93]. On the contrary, the high pressure fabricated

device, which shows p-type property. Therefore, as shown in Fig.

3.10(b), the electrons generated near around HTL side by shot

wavelength, ~450 nm would recombine during transporting through the

p –type perovskite layer. These would be the plausible reason for the

low IPCE of the low and high pressure fabricated films under long and

７５

short wavelength respectively.

７６

Fig. 3.10 (a) IPCE spectrum selected from Fig. 3.9 for low, optimized, high working pressure fabricated device with short (450 nm) and long (650 nm) wavelength. (b) Simulated electric field in the device of short (450 nm) (c) and long (650 nm) wavelength.

.

７７

3.3.8 Reproducibility

In order to check the reproducibility of the vacuum-pressure-based

control, additional batches were fabricated under the optimized

conditions. As shown in Table 1, the average PCE of 23 devices from

three batches was 12.7% (standard deviation (s.d.) ±0.84%). In addition,

considering the other parameters such as JSC, VOC, and FF that exhibit

the reasonably small standard deviations shown in Table 1, the

vacuum-pressure-based control results in the reproducible results. Fig.

3.11(a) shows the hysteresis and Fig. 3.11(b) displays the scan-rate-

dependent behavior of the solar cell fabricated under the optimized

condition, 5.1 × 10−5 torr. As shown in Fig. 3.11(a), it is worth noting

that the J-V curve shows negligible changes with the different scan

rates and even slightly high performances at a slow scan rate, 3 mV/s,

which would be closer to the steady state for the measurement than the

fast scan rates [94]. With this scan rate, 3 mV/s, the best-performing

device showed increased photovoltaic parameters: JSC of 20.1 mA/cm2,

VOC of 1.00 V, and FF of 0.72 resulting in a PCE of 14.5%. The

average values and standard deviations of PCE, JSC, VOC, and FF with

this scan rate were 13.4% (s.d. ±0.78), 19.1 mA/cm2 (s.d. ±0.74), 1.02

V (s.d. ±0.02), and 0.65 (s.d. ±0.04), respectively, with the same

devices and batches shown in Table 3.1.

７８

Pressure

(×10−5 torr)

PCE

(%)

JSC

(mA/cm2)

VOC

(V) FF

0.2

(only PbI2) 0.04

(±0.002) 0.1 (±0.01) 0.91 (±0.003)

0.30 (±0.002)

0.6 1.8 (±0.3) 3.6 (±0.4) 1.07 (±0.02)

0.48 (±0.02)

1.7 3.9 (±1.5) 7.6 (±2.6) 1.08 (±0.03)

0.46 (±0.06)

3.2 4.9 (±0.2) 9.0 (±0.1) 1.05 (±0.003)

0.52 (±0.02)

4.2 10.7 (±1.2) 17.6 (±0.5) 1.07 (±0.02)

0.57 (±0.05)

5.1 12.7 (±0.8) 18.5 (±0.6) 1.05 (±0.02)

0.65 (±0.04)

6.5 6.2 (±1.8) 11.8 (±1.3) 0.95 (±0.01)

0.54 (±0.13)

8.3 0.4 (±0.02) 1.11 (±0.06)

0.89 (±0.02)

0.42 (±0.02)

(± standard deviation)

Table 3.1. Average photovoltaic parameters of the fabricated

perovskite solar cells. The values were extracted from 4–8 cells (23

cells for the cells fabricated at 5.1 × 10−5 torr).

７９

Fig. 3.11 (a) Hysteresis of the fabricated solar cell fabricated under

the optimized pressure, 5.1×10-5 torr. (b) J–V characteristics at the

different scan rates of the device fabricated under optimized pressure,

5.1×10-5 torr .

８０

3.3.9 Composition measurement by EDX

Fig. 3.12 displays energy-dispersive X-ray spectroscopy (EDX) data

of the fabricated perovskite films directly indicating how the working

pressures affect the compositions in the perovskite films. The ratio of I

to Pb increased with the pressure. At the low pressures, the ratio

showed relatively small changes and began to increase considerably at

the pressure of 4.1 × 10−5 torr, which is analogous to the other physical

properties discussed above. Therefore, it can be concluded that the

working pressure of MAI can be an efficient indicator to control the

composition of the perovskite in the vacuum co-deposition process.

Further study is required for the quantitative analysis of the growth

mechanism of the perovskite film under vacuum.

８１

Fig. 3.12. Atomic ratio of iodine to lead measured by EDX for the

fabricated perovskite films.

８２

3.4 Summary

In summary, we have demonstrated an effective method to control

the composition of MAI in perovskite films for vacuum co-deposition.

The increased working pressure of MAI in the vacuum chamber can be

a proper indicator for controlling the composition in the perovskite

films. The thicknesses, absorbances, XRD spectra, morphologies,

device performances, and compositions changed significantly with the

working pressure of MAI. At the optimum pressure, the device showed

a maximum efficiency of 14.5% with good reproducibility, and all of

the layers were deposited under vacuum. These results will provide a

useful guide for the fabrication of perovskite films using the vacuum

co-deposition process.

８３

Chapter 4

Growth characteristics of MAI on PbI2 in vacuum

process

4.1 Introduction

Following Liu’s successful application of the vacuum process to the

perovskite solar cell, several groups have reported the perovskite solar

cells using vacuum co-deposition process [23-25, 30-42]. Most of these

groups commented on difficulties of controlling the MAI in vacuum. It

is known to be difficult to calibrate the thicknesses of deposited MAI

due to the gas-like behaviors in vacuum chamber and the rough

surfaces with poorly adsorbed MAI films [24, 25, 30-33, 43-46]. These

problems hinder an accurate monitoring and controlling the MAI in

vacuum process. To resolve this issue, it is important to understand the

behavior of MAI in vacuum, how it adsorbs on different surfaces in

vacuum. In this paper, we report the adsorption mechanism and kinetics

of MAI in vacuum process. MAI shows surface dependent adsorption

characteristics and it is not fully explained by only chemisorption.

Initial PbI2 layer induces enhanced physisorption of MAI on MAI

surface, examined by comparative adsorption experiments on quartz

８４

crystal microbalances (QCMs) and scanning Kelvin probe microscopy

(SKPM). These results suggest that the ferroelectricity of perovskite

formed by reacting with adsorbed MAI triggers the adsorption of polar

molecules such as MAI to increase the adsorption rate, which is

proposed as the mechanism for the MAI adsorption in vacuum.

4.2 Experimental

4.2.1 Film fabrication

PbI2 (x-nm)/MAI films The ITO-coated glass were cleaned with acetone

and isopropyl alcohol. After PbI2 (Alpha aser) evaporated on the

substrates, MAI (Jida Rubibo Optoelectronic Tech. 2 times purifying

by sublimation graded) was heated until it reaches 7 × 10−5 torr,

monitored by ionization gauge (Varian 572, Agilent Technologies) and

then the substrates were exposed to MAI vapor for 5, 8 and 25 min. For

the accurate control the MAI, the same amount MAI source, 0.80g,

was carefully measured and loaded in a crucible through a nitrogen-

filled glove box. The temperature was controlled by PID control (SJ

Power, STP-1500) and after reaching the desired working pressure, the

temperature was manually controlled to maintain the working pressure

constant.

８５

4.2.2 QCM Analysis

The QCMs (INFICON, 5 MHz) were monitored by crystal oscillation

type deposition controller, (CRTM-9000, ULVAC). When the bare

QCMs were loaded for sensor 1 and 2, MAI was deposited and it was

calibrated to read same thickness by adjusting the tooling factor for

sensor 11 to be, 0.877 from the thickness difference shown in Fig. S1,

with while other deposition parameters such as Z-ratio and density

remained same (Z-ratio=1, density=1g/cm3). The PbI2 (Alfa Aesar)

layers were deposited ~1 Å/s under 2×10-6 torr only on the sensor 2.

The two sensors, exposed to MAI at the same time for 25 minutes with

different deposited PbI2 thicknesses on the QCM, sensor 2. The

adsorbed weights are calculated from the monitored thicknesses using

equation ∆f=Cf ∆m where ∆f is frequency change, Cf is sensitivity

factor for 5 MHz AT-cut quartz crystal, 56.6Hzμg-1cm2 and ∆m is

mass weight change per unit area, (g/cm2). In our experiment, the

∆f/f<0.005, therefore it is able to apply the equation above assuming

the variation from z ratio is negligible.

８６

4.2.3 Characterization

The thicknesses of the films measured by a profilometry (KLA-

Tencor Alpha-Step IQ). The thickness are measured 3~5 points in the

film and averaged in ambient condition. The topology, phase and

SKPM images were measured by the XE-100 (Park systems) under

ambient condition at room temperature using pt-coated ElectriTap300g

(Budget sensors) with force constant 40 N/m. The SKPM measurement

were carried out in non-contact mode, using scan parameters as VAC=2

V, resonance frequency of 17 kHz and scan rate for 0.5 Hz. X-ray

diffraction (XRD) measurements were carried out by D8 Advance

diffractometer (Bruker) using Cu Ka radiation. The steps were 0.04

degree with exposure time for 1 second for each step.

4.3 Results and discussion

4.3.1 Thickness change and XRD depending on sub-layers

Table 1 displays the thicknesses of vacuum-deposited MAI on different

substrates in a same batch for 60 min under 7×10-5 torr. Interestingly,

the thicknesses are remarkably varied depending on the presence or

absence of PbI2 layer whereas relatively small variations on the other

substrates. The measured thicknesses of MAI on PbI2 layer are ca. ~2

８７

times thicker than the substrates without PbI2 initial layers. It is

speculated that this variation of MAI on different substrates could be

the reason why it is difficult to calibrate the thickness of MAI

accurately, as well as to control the deposition of MAI and/or MAPbI3

in vacuum process as previously reported [24, 25, 30-33, 43-46]. The

deposited film thicknesses depend also on the thickness of the initial

PbI2 players as shown in Fig. 4.1. The initial thickness of PbI2 layers

were varied from 0 (ITO surface) to 200 nm and these substrates were

exposed to MAI vapor for 60 minutes at a pressure of 7×10-5 torr in a

same batch. The black and red lines indicate that the total thickness and

thickness changes (∆d) between the MAI/PbI2(x-nm) films and the

initial thicknesses of the PbI2 layer, respectively. Surprisingly, only 2

nm of PbI2 layer results in 2 times thicker film than the film on ITO

surface without PbI2 layer. The 2 nm of PbI2 layer is much thinner than

the adsorbed MAI thickness so that the amount of diffusion of MAI

into PbI2 forming the perovskite to increase the thickness would be

limited. As shown in Fig. 4.1b, apparent perovskite (110) peak at 14.2°

was observed on 20~200 nm of PbI2 layers whereas the MAI/PbI2-2nm

exhibits relatively small perovskite (310) peaks at 31.7°, inferring thin

perovskite layer is formed [95]. The ∆d increases with increasing the

thickness of the PbI2 layer up to 100 nm but decreases when the PbI2

layer is 200 nm thick. X-ray diffraction patterns in Fig. 1b show that

the PbI2 (001) peak at 12.6˚ appears only when the thickness of the

８８

initial PbI2 is 200 nm, indicating that unreacted PbI2 remains after the

exposure to MAI under the condition. It can be inferred that the 200 nm

of PbI2 is thick enough to form larger amount of perovskite resulting in

decrease in the thickness change. Also, the adsorbed amount of MAI is

not directly appeared in the ∆d since diffused MAI into PbI2 induces

small volume (or thickness) change compared to intrinsic MAI films.

Therefore, to analyze the adsorbed amount of MAI quantitatively, we

used two QCMs in vacuum as shown in Fig. 4.2. The peak at

12.7°corresponds to PbI2 (001). The peaks at 14.2°, 28.2°, 28.5°,

31.7°, 40.5° and 43.1° correspond to (110) (004) (220) (310) (224) and

(314) of tetragonal phase MAPbI3 perovskite. The peaks at 19.7°, 19.9°

and 26.3° correspond to (002), (101) and (102) MAI [95].

８９

Substrate ITO NPB C60 Au PbI2

Thickness of MAI (nm) 72 76 85 86 184

Table 4.1. Thickness of deposited MAI on different substrates. All

the substrates are exposed to MAI for 60 min under 7 10-6 torr. The

thicknesses of the NPB, C60, Au and PbI2 are 20 nm on

ITO(150nm)/glass substrate

９０

Figure 4.1 (a) The thickness of MAI/PbI2 films on different initial PbI2 thicknesses (b) X-ray diffraction patterns of the MAI deposited on PbI2 films with different thicknesses

９１

4.3.2 Study of deposition kinetics using QCM

The two QCM sensors were calibrated earlier to indicate a same

thickness when MAI is deposited on the bare QCMs (Au-coated) to

exclude influences from geometrical differences between the two

sensors. Then, one of the surface for the sensors was covered by the

PbI2 by thermal evaporation with different thicknesses before exposure

to MAI. Fig. 4.2(b) exhibits the monitored adsorbed weight with time

on the PbI2 (x nm) /QCM. Fig.4.2(b) shows that the final adsorbed

weights increase from 19.6 to 34.9, 48.3, 68.3 and 70.5 μg/cm2 as the

thickness of the PbI2 layer increases. For comparison, the ∆t of the

deposited films by a surface profilometer and adsorbed weight

measured from the QCMs are plot in Fig.4. 2(c). If we assume that the

same amounts of MAI are adsorbed on the QCMs and the substrates,

the average density of the MAI/PbI2(x-nm) films can be calculated and

the results are summarized in Table S1. The calculated density of MAI

intrinsic film deposited on ITO is 2.33 g/cm3, reasonably close to the

reported value, 2.31 g/cm3[96]. In contrast, the average density of the

film increased as the initial PbI2 thickness increases, reaching to 3.96

g/cm3 close to the density of perovskite (4.16 g/cm3) [97] when the

initial PbI2 thickness was 200 nm.

Fig. 4.2(d) shows the deposition rates of MAI on QCM/PbI2(x nm).

The deposition rates increase as the working pressure of MAI increases

９２

at the early stage of deposition, reaches the peak values when the

working pressure reaches the desired value (7×10-5 torr). As the

thicknesses of initial PbI2 layer increases, the deposition rates increase

rapidly. At the constant working pressure, the deposition rates decrease

and the rates decrease relatively quickly in as the thickness of PbI2 layer

decrease. Eventually the deposition rates becomes the same value after

different times for different PbI2 thicknesses. The reason why the

overall rates decrease under the constant working pressure is not clear

yet. The increase of the working pressure could be attributed from the

low sticking coefficient of MAI [24, 25, 30-33, 43-46].

９３

Fig. 4.2. (a) Schematic figure for comparative adsorption experiment

on QCMs (b) monitored adsorbed weight of MAI on QCM/PbI2(x nm)

９４

Fig. 4.2. (c) The measured thickness change and adsorbed weight

and (d) monitored deposition rates and working pressure

９５

4.3.4 Hypothesis

The results shown in the deposition rates can be summarized as

follows. 1) As the working pressure increases, the deposition rate

increase. 2) On PbI2, the deposition rate increase rapidly. 3) As the

thickness of PbI2 layer increase, the deposition rate increase rapidly. 4)

The deposition rates decrease under the constant working pressure. 5)

After certain times, all of the deposition rates reach a same value

regardless the thickness of PbI2 layers. Based on these results, it can be

hypothesized or speculated as follows. 1) As the working pressure

increase, the impingement rate of MAI would be increased which

means the number of striking MAI molecules per unit area is increased.

This explains reasonably why the deposition rates increase under

increasing working pressure since the deposition rate is proportional to

the impingement rate. 2) On PbI2, the reaction between MAI and PbI2

would occur. It can be speculated that this reaction enhances the

adsorption property since once the bonding (e.g. ionic bonding for the

perovskite) formed, the desorption would not be easily happened. This

reaction of MAI and PbI2 could be the reason why deposition rate of

MAI is faster on PbI2 layer. 3) As the thickness of PbI2 increase, the

diffusion length of MAI would can be increased. It can be speculated

that if the diffusion occurs quickly, the thicker PbI2 would induce a

faster deposition rates than the thinner one. 4) Once the perovskite

９６

formed on the PbI2 surface, the surface become ferroelectric due to the

perovskite layer. Since the MAI molecule is polar composed of

positively charged MA+ cation and I- anion, the adsorption can be

enhanced by dipole interaction. If this effect is decreased as the

thickness increase this could be the reason for the decreased deposition

rates under constant working pressure. 5) When the dipole effect

diminishes, the deposition rates reach same value regardless the

thickness and with or without PbI2 layers. Fig. 4.3 shows the

summarized hypothesis with schematic figures.

９７

Fig. 4.3. Schematic figure of the hypothesis to explain the deposition

parameter changes shown in Fig. 4.2(d).

９８

4.3.5 AFM study

The hypothesis discussed in 4.3.4 is involved with adsorption, reaction,

diffusion and thickness change, which is closely related to the surface

of the films. Therefore, we selected the five different MAI deposition

time, 0, 10, 20, 30 and 60 minutes with the 5 different initial PbI2 layers,

0, 2, 20, 100 and 200 nm and investigated the surface using AFM to

prove the hypothesis. Fig. 4.4(a) shows the morphology of the films

with different deposition time depending on the thickness of PbI2 layers

and Fig. 4.4(b) displays the corresponding RMS roughness values for

each film. As shown, roughness on ITO increase gradually from 1.7 nm

to 20 nm until 30 minutes and then maintained similar. For the 2 nm

PbI2 substrate, the initial roughness value is as same as the value of ITO,

1.7 nm. As shown in Fig. 4.7, the film shows uniformly covering the

entire ITO surface from magnified image (0.5μm × 0.5 μm). The

roughness increased rapidly at the MAI deposition time of 10 min, and

it increases to 35 nm until 30 min, showing similar value at 60 min. For

the 20 nm PbI2 substrate, the roughness values are increased from 1.7,

2.5, 24.3, 36.0 and 39.2 nm as the deposition time. For the 100 nm PbI2

substrate, the roughness changes from 4.5(value between 0, 2 nm and

200 nm of PbI2), 6.5, 15.8, 41 and 45 nm. For the 200 nm PbI2 substrate,

the roughness changes from 23, 10, 12, 22, and 30 nm.

９９

Fig. 4.4. (a) 3D-topology (5μm × 5μm) of the MAI exposure films

with different times and initial PbI2 thickness and (b) RMS roughness

values extracted from 5μm × 5μm images

１００

The value decrease at 10 min, then increases slowly until 20 min. At

30 min, it increased rapidly. If we assume that, the surface of MAI

grown on MAI is rougher than the perovskite it can be deduced that at

the certain times, the MAI growth of MAI would occur and the time

would be shorter as the thickness of PbI2 thinner. In this case, the times

when MAI on MAI growth occur are 0, 10, 20, 30 and 30 mins for 0, 2,

20, 100 and 200 nm of PbI2 layers. Fig. 4.5, 4.6 and 4.7 show that the

topology, phase image and surface potentials of the films with different

scales. The phase images correspond to the topology, without showing

clearly distinctive area inferring the surface is chemically homogeneous.

The changes of the surface potentials will be discussed later.

１０１

Fig. 4.5. (a) Topology (b) Surface potential and (c) phase images of potentials of

the MAI films (5 𝜇𝜇𝑚𝑚×5 𝜇𝜇𝑚𝑚)

１０２

Fig. 4.6. (a)Topology (b) Surface potential and (c) phase images of

potentials of the MAI films (1𝜇𝜇𝑚𝑚×1𝜇𝜇𝑚𝑚)

１０３

Fig. 4.7. (a) Topology (b) Surface potential and (c) phase images of

potentials of the MAI films (0.5 𝜇𝜇𝑚𝑚×0.5 𝜇𝜇𝑚𝑚)

１０４

4.3.6. Density and thickness analysis

Density of deposited films

Fig. 4.8(a) shows the thickness change measured by surface profiler

and Fig. 4.8(b) shows the density change of the deposited film. The

densities are calculated by the adsorbed mass from QCMs divided by

thickness changes measured by the surface profiler. It is noted that the

amount of deposited of MAI on the substrates and QCMs are same

since substrate holder and QCMs are located at the same approximately

heights, ~40 cm above the MAI source. Since the density of MAI is

lower than the perovskite and PbI2, the thickness change would be

smaller if the adsorbed MAI forms the perovskite. Once the diffusion is

restricted after a certain time, the MAI growth on MAI would occur as

speculated in 4.3.5. Therefore, until certain times, MAI would diffuse

into the PbI2 forming high density and after the certain times the MAI

would grow on MAI forming low-density films. As shown in Fig. 4.8,

the measured density of 0 and 2 nm of PbI2 substrates show similar

value for the reported MAI density, inferring the MAI growth on MAI

occur already at 10 min. This corresponds to the RMS roughness

rapidly increased at 10 min in the case of 2 nm of PbI2. For 20 nm

PbI2substrate, the density is roughly similar to the perovskite at 10 min,

１０５

then decrease reaching MAI density at 20 min and this also corresponds

to the RMS roughness rapidly increased at 20 min for 20 nm PbI2

substrate. Similarly, for the 100 and 200 nm PbI2 substrates, the

densities show values between PbI2 and perovskite, decreasing within

this range until 20 min. The densities reach similar value to MAI at 30

min, corresponding to the time when rapid RMS roughness increased

discussed in above section, 4.3.5. From these regards, we can define tc,

as the time when the diffusion of MAI is restricted on the surface. The

tc for each substrate is 0, 10, 20, 30 and 30 min for 0, 2, 20, 100 and

200 nm PbI2 respectively.

１０６

Fig. 4.8 (a) Thickness change and (b) density change of the deposited

films on PbI2 (x nm) films with MAI deposition time

１０７

Thickness change analysis

As shown in Fig. 4.9, the thickness change of deposited MAI on PbI2

layer in vacuum can be expressed as

. . 2( ) { ( ) ( .)}dep diffd d MAI d MAI d PbI Perov∆ = + − + →

,where Δd is thickness change, d(MAIdep.) is total thickness of

deposited MAI assuming the diffusion is not occurred, d(MAIdiff.) is the

distance of MAI diffusion length and d(PbI2→Perov.) is thickness

change due to forming perovskite from PbI2. As discussed, before tc,

the thickness increase due to forming the perovskite would be dominant.

On the other hand, after tc, the thickness increase due to MAI growth

would be dominant. Hence, the thickness change can be expressed

depending on the time considering tc as follows.

. . 2

. . 2

1) : ( ) { ( ) ( .)}

2) : ( ) { ( ) ( .)}c dep diff

c dep diff

t t d MAI d MAI d PbI Perovt t d MAI d MAI d PbI Perov< < − + →

> > − + →

１０８

Fig. 4.9. Schematic figure for the thickness change on PbI2 layer in

vacuum process

１０９

4.3.8 Surface potential analysis

Theory of SKPM

Using the pt-coated conductive tip, the surface potential can be

measured by SKPM. The DC and AC voltages are applied to the tip

resulting in

AC( ) sin( )DC CPDV V V V tω= − + ⋅

where VCPD is contact potential difference between the tip and sample.

The DC voltage nullifying VCPD is tracked by feed-back loop which

means the VCPD , contact potential. The electrostatic force generated by

the voltage inducing the capacitance is

2 2

22

22

1 1( ) 2 4

[ ] sin( )

1 cos(2 )4

12 DC

DC DC CPD AC

DC CPD AC

AC

F F F

dCF V V Vdz

dCF V V V td

d

zdCF

CF Vd

V

z

tdz

ω ω

ω

ω

ω

ω

+ +

= − +

= −

= −

= =

, where the second term with the frequency of ω is extracted by lock-

in amplifier, measuring the VCPD, surface potential.

１１０

Surface potential analysis

The surface potentials before MAI exposure correspond to the work

function values of ITO and PbI2 As discussed above, 2 nm of PbI2

substrate shows MAI-MAI growth showing rapid increase in roughness

and thickness compared to the thick-PbI2 substrates. In addition, as

shown in Fig. 4.5, 4.6 and 4.7, the phase images are corresponding to

the topology without notably distinctive area in phase image inferring

the surface of the films are homogeneous. Therefore, it can be inferred

that surface potential ~-0.3 eV in 10 min. on 2 nm of PbI2 layer

correspond to the MAI-MAI adsorption regime. The surface potential

of 100 and 200 nm substrates in 10 to 30 minutes of the exposure show

negative value than the 2 and 20 nm of PbI2 substrate and then the

values reach to similar, ~0.3 eV amorously to fact that the deposition

rate reached the similar value.

Validation of the surface potential values

The average surface potential values are summarized in Fig. 4.5. The

surface potentials before the MAI exposure, 0 min should be

correspond to the work function of ITO and PbI2. It is known that the

work function measured by SKPM gives higher value than UPS since

UPS measures the fermi-level edge and low energy cut-off secondary

electrons resulting in minimum value of work function [98]. The

work function of pt-coated tip is -5.0 eV [99], therefore the

１１１

corresponding work function can be calculated by

CPDWork functionof tip Work functionof sampleV

e−

=−

The calculated work function of ITO and PbI2 is -4.9 and ~ -5.6 eV

and these values are higher than the reported value by UPS [56, 57].

Also if we assume that the surface potential of the films at 60 minutes

correspond to the intrinsic work function of MAI films than the work

function is -4.6 eV which similar level higher value than reported by

UPS [100]. In addition, the surface potentials are measured at least 3

times with different scales then averaged without significant

differences during the measurement. Therefore, it is thought to be the

values are approximately correct and reproducible.

１１２

Fig. 4.10. (a) Surface potential images (5 𝜇𝜇𝑚𝑚 × 5 𝜇𝜇𝑚𝑚) and (b)

Average surface potential values measured from various scales

１１３

4.3.9. Growth mechanism of MAI in vacuum

Potential change of MAI

As discussed in the previous section, the surface potential shifted from

-0.3 V to + 0.3 V. As discussed in the AFM study and density/thickness

change analysis part, section 4.3.5 and 4.3.6, the surfaces of the

samples are PbI2, MAxPbI2+x and MAI depending on time or tc. For the

case of 20, 100 and 200 nm, the surface potential at 10 min would be

originated from the work function of the perovskite. However, in the

case of 2 nm of PbI2 layer, the diffusion would be limited as discussed

previously. Therefore, it can be deduced that the surface potential of

MAI is shifted from -0.3 V to + 0.3 V. Fig. 4.11(a) shows why the MAI

potential is changed with thickness. Since the perovskite is ferroelectric,

this effect may order the orientation of polar MAI facing positive

charge out of the surface. As the thickness increase, this dipole effects

would be decreased and diminished after certain time reaching similar

potentials as well as the similar deposition rates shown in Fig. 4.2(d).

Fig. 4.2(b) demonstrates how this change can be measured by SKPM.

As discussed, the ordered MAI molecules on the perovskite shows

positive change facing to the surface, which lowers the surface

potential. Therefore, the contact potential difference, VCPD shift to

negative value and as the thickness increase the surface potential

become close to its intrinsic value showing the VCPD shift to positive

１１４

value as shown in Fig. 4.2(b). From these results, we suggest dipole

induced adsorption as the mechanism for MAI growth in vacuum

process.

１１５

Fig. 4.11. (a) Schematic figure for why the MAI potential is

changed with thickness and (b) illustration how the orientation change

of MAI appear in SKPM measurement

１１６

Schematic figure for growth of MAI on PbI2 in vacuum

Fig. 4.12 shows the schematic illustration for the growth of MAI on the

PbI2 films based on the results discussed above. As explained, MAI

adsorbs faster on PbI2 than on without PbI2 (bare QCM or ITO) and the

adsorption rates gradually decrease. The fast and slow adsorption

regions are denoted by green and gray in Fig. 4.12. As the initial PbI2

thickness increases, the fast-adsorbed MAI region increased as shown

in Fig. 4.2(d), forming perovskite under MAI diffusion as shown in Fig.

4.1(b). The ∆d of MAI/PbI2 films increased as the initial PbI2 thickness

increases until 100 nm, and it decrease when PbI2 is 200 nm. This is

because the diffusion of MAI is not limited on the thick (200 nm) PbI2

forming larger amount of perovskite resulting in ∆ d decreases with

residual PbI2 as shown in Fig. 4.1(b).

１１７

Fig. 4.12. Schematic illustration for growth of MAI on various PbI2 thickness

with time

１１８

4.4. Summary

In summary, we have investigated the growth mechanism of MAI in

vacuum. As the thickness of PbI2 layer increase, the deposition rates of

MAI increased rapidly and at the constant working pressure, all of the

deposition rates started to decrease reaching to a same value after

certain times with different initial PbI2 thicknesses. This phenomena

was analysed by atomic force microscopy (AFM) discussing on the

topology, surface potential, thickness and density changes. The PbI2

layer enhances the deposition rates by surface reaction forming

perovskite. This effect last longer as the thickness of PbI2 layer increase

due to the diffusion length of MAI into the PbI2 would longer in the

thicker PbI2 layer. When the diffusion is finished, the MAI growth on

MAI regime starts where roughness and thickness changes rapidly

reaching the density of the deposited film to the MAI density. However,

the surface potential in this regime, MAI growth, is shifted -0.3 V to

+0.3 V especially for the 2 nm PbI2 case. Conclusively, we suggest that

a mechanism for the adosprtion of MAI as follows. Once the

ferroelectric perovskite, MAPbI3 is formed on the surface, the polar

MAI adsorption via dipole attraction resulting in the faster adsorption

rate on MAI surface. As the thickess of MAI increase, this effect

decrease resulting in the similar slow deposition rates.

We believe that this study will be of great interest to those working

１１９

on the vacuum processed perovskite to enhance the controllability of

MAI behavior in vacuum

Chapter 5

Summary and conclusion

In this thesis, the vacuum processed perovskite solar cells are

reported and as well as the efficient method to control the MAI in

vacuum process for improving the quality of the films. Moreover, the

growth mechanism of MAI was investigated; consequently, we suggest

dipole induced adsorption for MAI on PbI2 (perovskite) in vacuum.

Firstly, in chapter 2, full vacuum processed perovskite solar cell

employing MoO3/NPB, which are typically used in vacuum processed

OLEDs, will be discussed. . The best performing device showed PCE =

13.7%, JSC = 18.1 mA/cm2, VOC = 1.12 V, FF = 0.67 and the

reproducibility is higher than the normal solution processed perovskite

solar cells. Importantly the considerably high VOC value was obtained

by employing the MoO3/NPB layer as the hole extraction layer. This

high VOC is attributed to the alignment of HOMO level of the NPB to

the VB edge of MAPbI3. The fact that the MoO3/NPB layer is widely

１２０

used in OPVs/OLEDs implies that technologies developed for vacuum

processed OLEDs/OPVs can be successfully applied in perovskite solar

cells to accelerate the development of the perovskite solar cells.

In chapter 3, an effective method for controlling the compositions of

the perovskite films in using the working pressures will be discussed.

The compositions of the perovskite were able to be adjusted with the

working pressures of MAI. Using this method, the properties of the

films with various compositions were revealed; they showed significant

differences under the different working pressures. Furthermore, we

show the application of the fabricated perovskite films to solar cell

devices. Optimum pressure of MAI resulted in, the maximum power

conversion efficiency of 14.5% with good reproducibility.

In chapter 4, the growth mechanism of MAI in vacuum process is

discussed. MAI shows surface dependent growth characteristics. On

PbI2, the deposition rate of MAI increased. Since the diffusion of MAI

increase depending on the thickness of PbI2, this enhanced deposition

rate lasted longer. However, on MAI growth MAI regime, the

deposition rate is still faster than on the surface without PbI2, notably

showed on very thin (2 nm of PbI2) layer. From the surface analysis, we

suggest that the ferroelectricity of perovskite formed by reacting with

adsorbed MAI triggers the adsorption of polar molecules such as MAI

to increase the adsorption rate, which is proposed as the mechanism for

the MAI adsorption in vacuum.

１２１

The fact that the MoO3/NPB layer is widely used materials in

OPVs/OLEDs implies that a lots of technologies developed for vacuum

processed OLEDs/OPVs can be successfully applied in perovskite solar

cells to accelerate the development of the perovskite solar cells.

Additionally, effective methods to control the MAI is reported with

various physical properties of the perovskite films with controlled

compositions. This will be a good guide line for finding an optimum

condition in vacuum process. Lastly, the detail study on behaviour of

MAI in vacuum process would be of great interest to those working on

the vacuum processed perovskite to enhance the controllability of MAI

behaviour in vacuum.

１２２

Bibliography

[1] Kojima, A.; Teshima, K.; Shirai, Y.; Miyasaka, T. Organometal Halide Perovskites as Visible-Light Sensitizers for Photovoltaic Cells J. Am. Chem. Soc. 2009, 131, (17), 6050.

[2] Burschka, J.; Pellet, N.; Moon, S. J.; Humphry-Baker, R.; Gao, P.; Nazeeruddin, M. K.; Gratzel, M. Sequential Deposition as a Route to High-Performance Perovskite-Sensitized Solar Cells. Nature 2013, 499, 316–319.

[3] Bi, D.; Moon, S.-J.; Haggman, L.; Boschloo, G.; Yang, L.; Johansson, E. M. J.; Nazeeruddin, M. K.; Gratzel, M.; Hagfeldt, A. Using a Two-Step Deposition Technique To Prepare Perovskite (CH3NH3PbI3) for Thin Film Solar Cells Based on ZrO2 and TiO2 Mesostructures RSC Adv. 2013, 3, 18762– 18766

[4] Yang, W. S.; Park, B.-W.; Jung, E. H.; Jeon, N. J.; Kim, Y. C.; Lee, D. U.; Shin, S. S.; Seo, J.; Kim, E. K.; Noh, J. H.; et al. Iodide Management in Formamidinium-Lead-Halide−based

[5] National renewable energy laboratory, Best research-cell efficiency chart, https://www.nrel.gov/pv/assets/pdfs/best-research-cell-efficiencies-190416.pdf (update, April. 2019)

[6] Nanova, D.; Kast, A. K.; Pfannmoller, M.; Muller, C.; Veith, L.; Wacker, I.; Agari, M.; Hermes, W.; Erk, P.; Kowalsky, W. Unraveling the Nanoscale Morphologies of Mesoporous Perovskite Solar Cells and Their Correlation to Device Performance Nano Lett. 2014, 14, 2735– 2740

[7] Ball, J. M.; Lee, M. M.; Hey, A.; Snaith, H. J. Low-Temperature Processed Meso-Superstructured to Thin-Film Perovskite Solar Cells Energy Environ. Sci. 2013, 6, 1739– 1743

[8] Xiao, Z. G.; Bi, C.; Shao, Y. C.; Dong, Q. F.; Wang, Q.; Yuan, Y. B.; Wang, C. G.; Gao, Y. L.; Huang, J. S. Efficient, High Yield Perovskite Photovoltaic Devices Grown by Interdiffusion of Solution-Processed Precursor Stacking Layers Energy Environ. Sci. 2014, 7, 2619– 2623

[9] Noh, J. H.; Im, S. H.; Heo, J. H.; Mandal, T. N.; Seok, S. I. Chemical Management for Colorful, Efficient, and Stable Inorganic–Organic Hybrid Nanostructured Solar Cells Nano Lett. 2013, 13, 1764– 1769

[10] Niu, G. D.; Li, W. Z.; Meng, F. Q.; Wang, L. D.; Dong, H. P.; Qiu, Y. Study on the Stability of CH3NH3PbI3 Films and the Effect of Post-Modification by Aluminum Oxide in All-Solid-State Hybrid Solar Cells J. Mater. Chem. A 2014, 2, 705– 710

１２３

[11] Chen, Q.; Zhou, H.; Hong, Z.; Luo, S.; Duan, H. S.; Wang, H. H.; Liu, Y.; Li, G.; Yang, Y. Planar Heterojunction Perovskite Solar Cells Via Vapor-Assisted Solution Process J. Am. Chem. Soc. 2014, 136, 622– 5

[12] Shi, J.; Luo, Y.; Wei, H.; Luo, J.; Dong, J.; Lv, S.; Xiao, J.; Xu, Y.; Zhu, L.; Xu, X. Modified Two-Step Deposition Method for High-Efficiency TiO2/CH3NH3PbI3 Heterojunction Solar Cells ACS Appl. Mater. Interfaces 2014, 6, 9711– 9718

[13] Conings, B.; Baeten, L.; De Dobbelaere, C.; D’Haen, J.; Manca, J.; Boyen, H.-G. Perovskite-Based Hybrid Solar Cells Exceeding 10% Efficiency with High Reproducibility Using a Thin Film Sandwich Approach Adv. Mater. 2014, 26, 2041– 2046

[14] Im, J. H.; Lee, C. R.; Lee, J. W.; Park, S. W.; Park, N. G. 6.5% Efficient Perovskite Quantum-Dot-Sensitized Solar Cell Nanoscale 2011, 3, 4088– 4093

[15] Kim, H.-B.; Choi, H.; Jeong, J.; Kim, S.; Walker, B.; Song, S.; Kim, J. Y. Mixed Solvents for the Optimization of Morphology in Solution-Processed, Inverted-Type Perovskite/Fullerene Hybrid Solar Cells Nanoscale 2014, 6, 6679– 6683

[16] Zhao, Y. X.; Zhu, K. CH3NH3Cl-Assisted One-Step Solution Growth of CH3NH3PbI3: Structure, Charge-Carrier Dynamics, and Photovoltaic Properties of Perovskite Solar Cells J. Phys. Chem. C 2014, 118, 9412– 9418

[17] Wu, Y.; Islam, A.; Yang, X.; Qin, C.; Liu, J.; Zhang, K.; Peng, W.; Han, L. Retarding the Crystallization of PbI2 for Highly Reproducible Planar-Structured Perovskite Solar Cells via Sequential Deposition Energy Environ. Sci. 2014, 7, 2934– 2938

[18] Xiao, M.; Huang, F.; Huang, W.; Dkhissi, Y.; Zhu, Y.; Etheridge, J.; Gray-Weale, A.; Bach, U.; Cheng, Y. B.; Spiccia, L. A Fast Deposition-Crystallization Procedure for Highly Efficient Lead Iodide Perovskite Thin-Film Solar Cells Angew. Chem. 2014, 126 (37) 10056– 10061

[19] Eperon, G. E.; Burlakov, V. M.; Docampo, P.; Goriely, A.; Snaith, H. J. Morphological Control for High Performance, Solution-Processed Planar Heterojunction Perovskite Solar Cells Adv. Funct. Mater. 2014, 24 (1) 151– 157

[20] Dualeh, A.; Tétreault, N.; Moehl, T.; Gao, P.; Nazeeruddin, M. K.; Grätzel, M. Effect of Annealing Temperature on Film Morphology of Organic–Inorganic Hybrid Pervoskite Solid-State Solar Cells Adv. Funct. Mater. 2014, 24, 3250– 3258

[21] Saliba, M.; Tan, K. W.; Sai, H.; Moore, D. T.; Scott, T.; Zhang, W.; Estroff, L. A.; Wiesner, U.; Snaith, H. J. The Influence of Thermal Processing Protocol upon the Crystallization and Photovoltaic Performance of Organic-

１２４

Inorganic Lead Trihalide Perovskites J. Phys. Chem. C 2014, 118, 17171– 17177

[22] Xiao, Z.; Dong, Q.; Bi, C.; Shao, Y.; Yuan, Y.; Huang, J. Solvent Annealing of Perovskite-Induced Crystal Growth for Photovoltaic-Device Efficiency Enhancement Adv. Mater. 2014, 26, 6503– 6509

[23] Liu, M.; Johnston, M. B.; Snaith, H. J. Efficient Planar Heterojunction Perovskite Solar Cells by Vapour Deposition. Nature 2013, 501, 395–398.

[24] Malinkiewicz, O.; Yella, A.; Lee, Y. H.; Espallargas, G. M.; Graetzel, M.; Nazeeruddin, M. K.; Bolink, H. J. Perovskite Solar Cells Employing Organic Charge-Transport Layers. Nat. Photonics 2013, 8, 128–132.

[25] Momblona, C.; Gil-Escrig, L.; Bandiello, E.; Hutter, E. M.; Sessolo, M.; Lederer, K.; Blochwitz-Nimoth, J.; Bolink, H. J. Efficient vacuum deposited pin and nip perovskite solar cells employing doped charge transport layers Energy Environ. Sci. 2016, 9, 3456– 3463

[26] Ávila, J.; Momblona, C.; Boix, P. P.; Sessolo, M.; Bolink, H. J. Vapor-Deposited Perovskites: The Route to High-Performance Solar Cell Production?. Joule 2017 1, 431– 442.

[27] Shen, P.-S.; Chiang, Y.-H.; Li, M.-H. Guo, T.-F.; Chen, P. APL Mater. 2016, 4, 091509.

[28] Kim, H.; Lim, K.-G.; Lee, T.-W. Energy Environ. Sci., 2016, 9, 12-30.

[29] Ono, L. K.; Leyden, M. R.; Wang, S.; Qi, Y.; J. Mater. Chem. A, 2016, 4, 6693-6713.

[30] Kim, B.-S.; Kim, T.-M.; Choi, M.-S.; Shim, H.-S.; Kim, J.-J. Fully Vacuum–Processed Perovskite Solar Cells with High Open Circuit Voltage Using MoO3/NPB as Hole Extraction Layers. Organic Electronics 2015, 17, 102–106.

[31] Polander, L.E.; Pahner, P.; Schwarze, M.; Saalfrank, M.; Koerner, C.; Leo, K. Hole-Transport Material Variation in Fully Vacuum Deposited Perovskite Solar Cells. APL Mater. 2014, 2, 081503.

[32] Teuscher, J.; Ulianov, A.; Müntener, O.; Grätzel, M.; Tetreault, N. Control and Study of the Stoichiometry in Evaporated Perovskite Solar Cells ChemSusChem 2015, 8, 3847– 3852

[33] Kim, B.-S.; Choi, M.-H.; Choi, M.-S.; Kim, J.-J. Composition-Controlled Organometal Halide Perovskite via CH3NH3I Pressure in a Vacuum Co-Deposition Process J. Mater. Chem. A 2016, 4, 5663– 5668

[34] Lin, Q.; Armin, A.; Nagiri, R. C. R.; Burn, P. L.; Meredith, P. Electro-Optics of Perovskite Solar Cells Nat. Photonics 2015, 9, 106– 112

１２５

[35] Subbiah, A. S.; Halder, A.; Ghosh, S.; Mahuli, N.; Hodes, G.;Sarkar, S. K. Inorganic Hole Conducting Layers for Perovskite-Based Solar Cells J. Phys. Chem. Lett. 2014, 5, 1748−1753

[36] Ng, T.-W.; Chan, C.-Y.; Lo, M.-F.; Guan, Z. Q.; Lee, C.-S. Formation chemistry of perovskites with mixed iodide/chloride content and the implications on charge transport properties J. Mater. Chem. A 2015, 3, 9081– 9085.

[37] Gao, C.; Liu, J.; Liao, C.; Ye, Q.; Zhang, Y.; He, X.; Guo, X.; Mei, J.; Lau, W. Formation of organic–inorganic mixed halide perovskite films by thermal evaporation of PbCl2 and CH3NH3I compounds RSC Adv., 2015, 5, 26175.

[38] Zhao, D.; Ke, W.; Grice, C. R.; Cimaroli, A. J.; Tan, X.; Yang, M.; Collins, R. W.; Zhang, H.; Zhu, K.; Yan, Y. Annealing-free efficient vacuum-deposited planar perovskite solar cells with evaporated fullerenes as electron-selective layers Nano Energy 2016, 19, 88−97.

[39] Borchert, J.; Boht, H.; Fränzel, W.; Csuk, R.; Scheer, R.; Pistor, P. Structural investigation of co-evaporated methyl ammonium lead halide perovskite films during growth and thermal decomposition using different PbX2 (X = I, Cl) precursors J. Mater. Chem. A, 2015, 3, 19842

[40] Su, Z.; Hou, F.; Jin, F.; Wang, L.; Li, Y.; Zhu, J.; Chu, B.; Li, W. Hole transporting material-free and annealing-free thermal evaporated planar perovskite solar cells with an ultra-thin CH3NH3PbI3−XClX layer Organic Electronics 2015, 26, 104-108.

[41] Ke, W.; Zhao, D.; Grice, C.R.; Cimaroli, A.J.; Fang, G.; Yan, Y. Efficient fully-vacuum-processed perovskite solar cells using copper phthalocyanine as hole selective layers J. Mater. Chem. A, 2015, 3, 23888-23894.

[42] Tran, C. D. T.; Liu, Y.; Thibau, E. S.; Llanos, A.; Lu, Z.-H. Stability of Organometal Perovskites with Organic Overlayers AIP Adv. 2015, 5, 087185

[43] Chen, C. W.; Kang, H. W.; Hsiao, S. Y.; Yang, P. F.; Chiang, K. M.; Lin, H. W. Efficient and Uniform Planar‐Type Perovskite Solar Cells by Simple Sequential Vacuum Deposition Adv. Mater. 2014, 26, 6647−6652.

[44] D.Yang, Z. Yang, W. Qin, Y. Zhang, S,. Liu, C.Li. Alternating precursor layer deposition for highly stable perovskite films towards efficient solar cells using vacuum deposition J. Mater. Chem. A 2015, 3, 9401

[45] S.-Y Hsiao, H.-L. Lin, W.-H. Lee, W.-L. Tsai, K-M. Chiang, W.-Y. Liao, C.0ZR.-Wu. C.-Y. Chen, H.-W. Lin. Efficient All‐Vacuum Deposited Perovskite Solar Cells by Controlling Reagent Partial Pressure in High Vacuum Adv Mater. 2016, 28, 7013-7019.

[46 ] Wang, S.; Ono, L. K.; Leyden, M. R.; Kato, Y.; Raga, S. R.; Lee, M. V.

１２６

Qi, Y. Smooth Perovskite Thin Films and Efficient Perovskite Solar Cells Prepared by the Hybrid Deposition Method. J. Mater. Chem. A 2015, 3, 14631−14641.

[47] Du, H. M. Efficient carrier transport in halide perovskites: theoretical perspectives. J. Mater. Chem. A 2014, 2, 9091−9098.

[48] Yin, W.-J.; Shi, T.; Yan, Y. Superior Photovoltaic Properties of Lead Halide Perovskites: Insights from First-Principles Theory. J. Phys. Chem. C 2015, 119, 5253−5264.

[49] Rühle, S. Tabulated Values of the Shockley-Queisser Limit for Single Junction Solar Cells. Sol. Energy 2016, 130, 139– 147

[50] Shirayama, M.; Kadowaki, H.; Miyadera, T.; Sugita, T.; Tamakoshi, M.; Kato, M.; Fujiseki, T.; Murata, D.; Hara, S.; Murakami, T. N.; et al. Optical Transitions in Hybrid Perovskite Solar Cells: Ellipsometry, Density Functional Theory, and Quantum Efficiency Analyses for CH3NH3PbI3. Phys. Rev. Appl. 2016, 5, 014012.

[51] Guerra, J. A.; Tejada, A.; Korte, L.; Kegelmann, L.; Töfflinger, J. A.; Albrecht, S.; Rech, B.; Weingärtner, R. Determination of the Complex Refractive Index and Optical Bandgap of CH3NH3PbI3 Thin Films. J. Appl. Phys. 2017, 121, 173104,

[52] Miyata, A.; Mitioglu, A.; Plochocka, P.; Portugall, O.; Wang, J. T.-W.; Stranks, S. D.; Snaith, H. J.; Nicholas, R. J. Direct Measurement of the Exciton Binding Energy and Effective Masses for Charge Carriers in Organic-Inorganic Tri-Halide Perovskites Nat. Phys. 2015, 11, 582– 587

[53] Yang, Z.; Surrente, A.; Galkowski, K.; Bruyant, N.; Maude, D. K.; Haghighirad, A. A.; Snaith, H. J.; Plochocka, P.; Nicholas, R. J. Unraveling the Exciton Binding Energy and the Dielectric Constant in Single-Crystal Methylammonium Lead Triiodide Perovskite J. Phys. Chem. Lett. 2017, 8, 1851– 1855

[54] Yang, Y.; Yang, M.; Li, Z.; Crisp, R.; Zhu, K.; Beard, M. C. Comparison of Recombination Dynamics in CH3NH3PbBr3 and CH3NH3PbI3 Perovskite Films: Influence of Exciton Binding Energy J. Phys. Chem. Lett. 2015, 6, 4688– 4692

[55] Sherkar, T. S.; Momblona, C.; Gil-Escrig, L.; Ávila, J.; Sessolo, M.; Bolink, H. J.; Koster, L. J. A. Recombination in Perovskite Solar Cells: Significance of Grain Boundaries, Interface Traps, and Defect Ions ACS Energy Lett. 2017, 2, 1214– 1222

[56] Najafi, M.; Di Giacomo, F.; Zhang, D.; Shanmugam, S.; Senes, A.; Verhees, W.; Hadipour, A.; Galagan, Y.; Aernouts, T.; Veenstra, S.; Andriessen, R. Highly Efficient and Stable Flexible Perovskite Solar Cells with Metal Oxides Nanoparticle Charge Extraction Layers. Small 2018, 14 (12), 1702775

１２７

[57] Jeon, N. J.; Noh, J. H.; Yang, W. S.; Kim, Y. C.; Ryu, S.; Seo, J.; Seok, S. I. Compositional Engineering of Perovskite Materials for High-Performance Solar Cells Nature 2015, 517, 476– 480

[58] H. Zhou, Q. Chen, G. Li, S. Luo, T.-B. Song, H.-S. Duan, Z. Hong, J. You, Y. Liu, Y. Yang. Interface engineering of highly efficient perovskite solar cells Science 2014, 345, 542–546.

[59] Yin, W. J.; Shi, T.; Yan, Y. Unusual Defect Physics in CH3NH3PbI3 Perovskite Solar Cell Absorber Appl. Phys. Lett. 2014, 104, 063903

[60] Kang, B.; Biswas, K. Shallow Trapping vs. Deep Polarons in a Hybrid Lead Halide Perovskite, CH3NH3PbI3 Phys. Chem. Chem. Phys. 2017, 19 (40), 27184– 27190

[61] Baumann, A.; Väth, S.; Rieder, P.; Heiber, M. C.; Tvingstedt, K.; Dyakonov, V. Identification of Trap States in Perovskite Solar Cells J. Phys. Chem. Lett. 2015, 6, 2350– 2354

[62] Simpson, M. J.; Doughty, B.; Yang, B.; Xiao, K.; Ma, Y.-Z. Imaging Electronic Trap States in Perovskite Thin Films with Combined Fluorescence and Femtosecond Transient Absorption Microscopy J. Phys. Chem. Lett. 2016, 7, 1725– 1731

[63] Frost, J. M.; Butler, K. T.; Brivio, F.; Hendon, C. H.; van Schilfgaarde, M.; Walsh, A. Atomistic Origins of High-Performance in Hybrid Halide Perovskite Solar Cells Nano Lett. 2014, 14, 2584– 2590

[64] Liu, S.; Zheng, F.; Koocher, N. Z.; Takenaka, H.; Wang, F.; Rappe, A. M. Ferroelectric Domain Wall Induced Band Gap Reduction and Charge Separation in Organometal Halide Perovskites J. Phys. Chem. Lett. 2015, 6, 693– 699

[65] Yin, W.-J.; Shi, T.; Yan, Y. Unique Properties of Halide Perovskites as Possible Origins of the Superior Solar Cell Performance Adv. Mater. 2014, 26, 4653– 4658

[66] Ohring, M The Material science of thin films Academic press 1991

[67] Sauerbrey, G. Verwendung von Schwingquarzen zur Wägung dünner Schichten und zur Mikrowägung Z. Phys. A: Hadrons Nucl. 1959, 155, 206

[68] Bartmess, J. E.; Georgiadis, R. M. Empirical Methods for Determination of Ionization Gauge Relative Sensitivities for Different Gases Vacuum 1983, 33, 149– 153

[69] Schulz, P.; Edri, E.; Kirmayer, S.; Hodes, G.; Cahen, D.; Kahn, A. Interface Energetics in Organo-Metal Halide Perovskite-Based Photovoltaic Cells Energy Environ. Sci. 2014, 7, 1377-1381

１２８

[70] Jeon, N. J.; Lee, H. G.; Kim, Y. C.; Seo, J.; Noh, J. H.; Lee, J.; Seok, S. I. O-Methoxy Substituents in Spiro-OMetad for Efficient Inorganic-Organic Hybrid Perovskite Solar Cells J. Am. Chem. Soc. 2014, 136, 7837– 7840

[71] Qin, P.; Tetreault, N.; Dar, M. I.; Gao, P.; McCall, K. L.; Rutter, S. R.; Ogier, S. D.; Forrest, N. D.; Bissett, J. S.; Simms, M. J.; Page, A. J.; Fisher, R.; Grätzel, M.; Nazeeruddin, M. K. A Novel Oligomer as a Hole Transporting Material for Efficient Perovskite Solar Cells Adv. Energy Mater. 2015, 5, 1400980

[72] Lim, K.-G.; Kim, H. B.; Jeong, J.; Kim, H.; Kim, J. Y.; Lee, T.-W. Boosting the Power Conversion Efficiency of Perovskite Solar Cells Using Self-Organized Polymeric Hole Extraction Layers with High Work Function Adv. Mater. 2014, 26, 6461– 6466

[73] Kwon, Y. S.; Lim, J.; Yun, H.-J.; Kim, Y.-H.; Park, T. A Diketopyrrolopyrrole-Containing Hole Transporting Conjugated Polymer for Use in Efficient Stable Organic–Inorganic Hybrid Solar Cells Based on a Perovskite Energy Environ. Sci. 2014, 7, 1454– 1460

[74] Jeng, J. Y.; Chen, K. C.; Chiang, T. Y.; Lin, P. Y.; Tsai, T. D.; Chang, Y. C.; Guo, T. F.; Chen, P.; Wen, T. C.; Hsu, Y. J. Nickel oxide electrode interlayer in CH3NH3PbI3 perovskite/PCBM planar-heterojunction hybrid solar cells Adv. Mater. 2014, 26, 4107– 4113

[75] Krishna, A.; Sabba, D.; Li, H. R.; Yin, J.; Boix, P. P.; Soci, C.; Mhaisalkar, S. G.; Grimsdale, A. C. Novel Hole Transporting Materials Based on Triptycene Core for High Efficiency Mesoscopic Perovskite Solar Cells Chem. Science 2014, 5, 2702– 2709

[76] Malinkiewicz, O.; Roldán-Carmona, C.; Soriano, A.; Bandiello, E.; Camacho, L.; Nazeeruddin, M. K.; Bolink, H. J. Metal-Oxide-Free Methylammonium Lead Iodide Perovskite-Based Solar Cells: The Influence of Organic Charge Transport Layers Adv. Energy Mater. 2014, 4, 1400345

[77] Zhao, Y.; Zhu, K. Charge Transport and Recombination in Perovskite (CH3NH3)PbI3 Sensitized TiO2 Solar Cells J. Phys. Chem. Lett. 2013, 4, 2880– 2884

[78] Baikie, T.; Fang, Y.; Kadro, J. M.; Schreyer, M.; Wei, F.; Mhaisalkar, S. G.; Grätzel, M.; White, T. J. Synthesis and Crystal Chemistry of the Hybrid Perovskite (CH3NH3)PbI3 for Solid-State Sensitised Solar Cell Applications J. Mater. Chem. A 2013, 1, 5628– 5641

[79] Seo, J.; Park, S.; Chan Kim, Y.; Jeon, N. J.; Noh, J. H.; Yoon, S. C.; Seok, S. Il. Benefits of Very Thin PCBM and LiF Layers for Solution-Processed p–i–n Perovskite Solar Cells Energy Environ. Sci. 2014, 7, 2642– 2646

[80] Ding, Y.; Yao, X.; Zhang, X.; Wei, C.; Zhao, Y. Surfactant Enhanced

１２９

Surface Coverage of CH3NH3PbI3–xClx Perovskite for Highly Efficient Mesoscopic Solar Cells J. Power Sources 2014, 272, 351– 355

[81] Lee, J.-H.; Kim, J.-J. Interfacial Doping for Efficient Charge Injection in Organic Semiconductors Phys. Status Solidi A 2012, 209, 1399– 1413

[82] Schulz, P.; Tiepelt, J. O.; Christians, J. A.; Levine, I.; Edri, E.; Sanehira, E. M.; Hodes, G.; Cahen, D.; Kahn, A. High Work Function Molybdenum Oxide Hole Extraction Contacts in Hybrid Organic-Inorganic Perovskite Solar Cells ACS Appl. Mater. Interfaces 2016, 8, 31491– 31499

[83] Gommans, H.; Verreet, B.; Rand, B.; Muller, P.; Poortmans, J.; Heremans, P.; Genoe. On the Role of Bathocuproine in Organic Photovoltaic Cells J.Adv. Funct. Mater., 2008, 18, 3686–3691

[84] Song, Q. L.; Li, C. M.; Wang, M. L.; Sun, X. Y.; Hou, X. Y. Role of Buffer in Organic Solar Cells Using C60 as an Acceptor Appl. Phys. Lett. 2007, 90, 071109 1– 3

[85] Vogel, M.; Doka, S.; Breyer, C.; Lux-Steiner, M. C.; Fostiropoulosa, K. On the Function of a Bathocuproine Buffer Layer in Organic Photovoltaic Cells Appl. Phys. Lett. 2006, 89, 163501

[86] Wang, Q.; Dong, Q.; Xiao, Z.; Yuan, Y.; Huang, J. Large Fill-Factor Bilayer Iodine Perovskite Solar Cells Fabricated by Low-Temperature Solution-Process Energy Environ. Sci. 2014, 7, 2359– 2365

[87] Fu, Y.; Meng, F.; Rowley, M. B.; Thompson, B. J.; Shearer, M. J.; Ma, D.; Hamers, R. J.; Wright, J. C.; Jin, S. Solution Growth of Single Crystal Methylammonium Lead Halide Perovskite Nanostructures for Optoelectronic and Photovoltaic Applications J. Am. Chem. Soc. 2015, 137, 5810– 5818

[88] Im, J.-H.; Jang, I.-H.; Pellet, N.; Grätzel, M.; Park, N.-G. Growth of CH3NH3PbI3 cuboids with controlled size for high-efficiency perovskite solar cells Nat. Nanotechnol. 2014, 9, 927

[89] Shi, J.; Wei, H.; Lv, S.; Xu, X.; Wu, H.; Luo, Y.; Li, D.; Meng, Q. Control of Charge Transport in the Perovskite CH3NH3PbI3 Thin Film ChemPhysChem 2015, 16, 842– 847

[90] Sabba, D.; Dewi, H. A.; Prabhakar, R. R.; Baikie, T.; Chen, S.; Du, Y.; Mathews, N.; Boix, P. P.; Mhaisalkar, S. G. Incorporation Of Cl In Sequentially Deposited Lead Halide Perovskite Films For Highly Efficient Mesoporous Solar Cells Nanoscale 2014, 6, 13854– 13860

[91] Xie, Y.; Shao, F.; Wang, Y.; Xu, T.; Wang, D.; Huang, F. Enhanced Performance of Perovskite CH3NH3PbI3 Solar Cell by Using CH3NH3I as Additive in Sequential Deposition ACS Appl. Mater. Interfaces 2015, 7 (23) 12937– 12942

１３０

[92] Chen, H. W.; Sakai, N.; Jena, A. K.; Sanehira, Y.; Ikegami, M.; Ho, K. C.; Miyasaka, T. A Switchable high-sensitivity photodetecting and photovoltaic device with perovskite absorber J. Phys. Chem. Lett. 2015, 6, 1773– 1779

[93] Wang, Q.; Shao, Y.; Xie, H.; Lyu, L.; Liu, X.; Gao, Y.; Huang, J. Qualifying composition dependent p and n self-doping in CH3NH3PbI3 Appl. Phys. Lett. 2014, 105, 163508

[94] Tao, C.; Neutzner, S.; Colella, L. 17.6% Stabilized Efficiency in Low-Temperature Processed Planar Perovskite Solar Cells Energy Environ. Sci. 2015, 8, 2365– 2370

[95] Chiang, C. H.; Tseng, Z. L.; Wu, C. G. Planar heterojunction perovskite/PC71 BM solar cells with enhanced open-circuit voltage via a (2/1)-step spin-coating process J. Mater. Chem. A 2014, 2, 15897−15903

[96] Yamamuro, O.; Oguni, M.; Matsuo, T.; Suga, H. J. Chem. Calorimetric and dilatometric studies on the phase transitions of crystalline CH3NH3I Thermodyn., 1986, 18, 939

[97] Huang, J.; Jiang, K.; Cui, X.; Zhang, Q.; Gao, M.; Su, M.; Yang, L.; Song, Y. Direct Conversion of CH3NH3PbI3 from Electrodeposited PbO for Highly Efficient Planar Perovskite Solar Cells Sci. Rep. 2015, 5, 15889.

[98] Kim, J. S.; Lagel, B.; Moons, E.; Johansson, N.; Baikie, I. D.; Salaneck, W. R.; Friend, R. H.; Cacialli, F. Kelvin probe and ultraviolet photoemission measurements of indium tin oxide work function: a comparison Synth. Met. 2000, 111, 311

[99] Vidyasagar, R.; Camargo, B.; Pelegova, E.; Romanyuk, K.; Kholkin, L. Controlling Surface Potential of Graphene Using dc Electric Field KnE Mater. Sci. 2016 183–189

[100] Ng, T.-W.; Thachoth Chandran, H.; Chan, C.-Y.; Lo, M.-F.; Lee, C.-S. Ionic Charge Transfer Complex Induced Visible Light Harvesting and Photocharge Generation in Perovskite ACS Appl. Mater. Interfaces 2015, 7, 20280−20284

１３１

초 록

유기물/무기물 혼합 페로브스카이트 태양전지는 최근 높은

전력변환 효율을 보이며 매우 많은 주목을 받고 있다. 현재

까지 대부분의 페로브스카이트 태양전지는 용액공정으로 제

작되어왔다. 용액공정을 이용한 페로브스카이트 태양전지의

중요한 주제 중 하나는, 박막의 균일도, 작은 구멍, 수분 및

대기에의 취약성 등을 제어하기 어려움에서 오는 재현성 문

제이다. 진공 증착은 박막 형성 인자들과 수분 및 용매가 없

는 깨끗한 환경에서 증착이 가능하기 때문에 재현성 문제를

해결할 수 있는 가능성이 큰 방법이다. 본 학위 논문에서는

모든 층을 진공으로 증착한 태양전지와, 페로브스카이트르

진공에서 고품질로 재현성 있게 만들 수 있는 방법론과 페

로브스카이트의 전구물질인 Methylammonium iodide (MAI) 의

진공공정에서의 거동을 이해 하기 위한 흡착 특성 및 기작

에 대해 논의하였다.

제 2장에서 전 층을 진공으로 제작한 페로브스카이트 태

양전지를 보고하였다. 기존 진공공정에서 안정적으로 사용되

던 Molybdenum oxide (MoO3)와 N,N′-Di(1-naphthyl)-N,N′-

diphenyl-(1,1′-biphenyl)-4,4′-diamine (NPB)를 이용하여 높은

１３２

개방 전압, 1.12 V, 과 함께 최고 13.7%의 효율을 갖는 태양전

지를 제작 보고하였다. 일반적인 기존 용액공정에 비해 높은

재현성을 보였으며 기존 진공공정에서 사용되던 전하수송층

을 페로브스카이트 소자에 성공적으로 적용한 의미가 있다.

제 3장에서는 진공공정에서 MAI를 효과적으로 제어하기

위한 방법을 제시하였다. 진공에서 MAI를 가열하게 되면 동

작압력이 증가하게 되고, 이것이 페로브스카이트의 조성을

제어 할 수 있는 지시계가 될 수 있었다. 다양한 동작 압력

에서 페로브스카이트의 물성은 크게 바뀌었으며, 압력에 따

른 물성 변화를 체계적으로 보고하였다. 최적화된 압력에서

페로브스카이트 태양전지는 최고 14.5%의 효율과 향상된 재

현성을 보여 진공도에 의한 제어가 성공적으로 이루어 졌음

을 알 수 있었다.

제 4장 에서는 진공에서 MAI의 성막 특성에 대해 논의하

였다. 페로브스카이트 진공공정에서 MAI의 거동을 제어하고

이해하는 것이 중요한데, 이는 성막 특성과 밀접한 관련이

있다. MAI는 표면에 의존적인 흡착 특성을 보였으며 PbI2 위

에서 특히 빠른 증착 속도를 보였다. PbI2와의 반응이 MAI

의 증착 속도를 증가시키며, PbI2안으로 확산을 하는 성질 때

문에 PbI2의 두께가 증가 할수록 이 효과가 오래 지속되었다.

１３３

또한 AFM 분석을 통해 MAI-MAI 성막이 일어나고 있는 구

간에서도 표면 포텐셜이 지속적으로 변하는 현상을 통해 초

기의 PbI2 층이 MAI-MAI의 증착 속도를 빠르게 한다는 현상

을 발견 하였다. 이러한 결과를 토대로 페로브스카이트의 강

유전성이 MAI의 배향을 바꾸며 표면포텐셜의 변화로 인한

쌍극자 인력이 흡착 특성을 강화 시키는 다이폴 유도 흡착

을 MAI의 진공에서의 성막 기작으로 제시하였다.

주요어: 유기물/무기물 혼합 페로브스카이트 태양전지, 전

진공공정, MAI 압력, MAI 흡착

학 번: 2015 – 30180

１３４

CURRICULUM VITAE

Beom-Soo Kim

Department of Materials Science and Engineering

Seoul National University, Seoul, 151-744, Korea

Education and Training

2015.3 ~ Current PH.D. in Materials Science and Engineering

Supervisor: Professor Jang-Joo Kim

Seoul National University, Seoul, Korea

2013.3 ~ 2015.2 M.S. in Materials Science and Engineering

Seoul National University, Seoul, Korea

2012.6 ~ 2013.3 Process Engineer in 3M company of Optical

Systems Division, Korea

2010.6 ~ 2011.6 Study abroad at University of Nottingham,

England

2005.3 ~ 2012.2 B.S. in Advanced Materials Science and

Engineering

Sungkyunkwan University, Seoul, Korea

１３５

Research Interests

- Metal-halide perovskite and optoelectronic devices physics

- Analysis, optimize and modification vacuum process system

- Kinetics and mechanisms for growth of perovskite films in

vacuum process

Professional Skills

- Design and fabrication of Perovskite and organic electronic

devices: thermal evaporator, sputtering system, glove-box, spin

coater

- Vacuum evaporator system set up, modification and maintenance

- The optoelectronic analysis on organic semiconductor: UV-vis-

NIR absorption spectroscopy, Photoluminescence spectroscopy,

Interpretation of current density-voltage characteristics and

incident photon-to-current efficiency, Impedance spectroscopy,

Photovoltaic parameters measurements, Transfer matrix methods

to calculate optoelectronic properties

- The structural characterization of organic films: Atomic forced

microscopy, Kelvin probe force microscopy, Alpha-step, X-ray

diffraction patterns, XPS, TOF-SIMS, SEM, EDX

１３６

Honors and Awards

- 2014. 10: Best poster Awards

Asian Conference on Organic Electronics, 2014

- 2009. 03: Honorable Merit-based scholarship

Department of Advanced Materials Science and

Engineering, SKKU

List of Publications

1. Beom-Soo Kim, Min-Hyung Choi, Min-Soo Choi and

Jang-Joo Kim “Composition-controlled organometal halide

perovskite via MAI pressure in a vacuum co-deposition

process” * J. Mater. Chem. A, (2016) ,4, 5663-5668

2. Beom-Soo Kim, Tae-Min Kim, Min-Soo Choi, Hyun-Sub

Shim and Jang-Joo Kim* “Fully vacuum–processed

perovskite solar cells with high open circuit voltage using

MoO3/NPB as hole extraction layers” Organic Electronics

(2015), 17, 102-106

3. Beom-Soo Kim and Jang-Joo Kim “Growth characteristics

of MAI on PbI2 in vacuum process” to be submitted

4. Min-Soo Choi, Tae-Min Kim, Hyun-Sub Shim, Beom-Soo

Kim, Hyo Jung Kim, and Jang-Joo Kim* “Enhancement of

１３７

the Fill Factor through an Increase of the Crystallinity in

Fullerene-Based Small-Molecule Organic Photovoltaic

Cells",

ACS Applied Materials & Interfaces, (2015) 7, 9134-9138

List of presentations

International Conference

1. Beom-Soo Kim, Min-Hyung Choi, Min-Soo Choi, Jang-Joo Kim,

"Composition-Controlled Organic/Inorganic Hybrid Perovskite via

Working Pressure of MAI in Vacuum Co-Deposition Process", The

8th International Workshop on Flexible & Printable Electronics

(2016 IWFPE), November 23-24 (November 23), 2016, Korea

(Poster)

2. Beom-Soo Kim, Min-Hyung Choi, Min-Soo Choi, Jang-Joo Kim,

"Composition-Controlled Perovskite Via MAI Pressure in Vacuum

Co-Deposition Process", International Conference on Hybrid and

Organic Photovoltaics (HOPV16), June 28-July 1 (June 29), 2016,

United Kingdom (Poster)

3. Beom-Soo Kim, Tae-Min Kim, Min-Soo Choi, Hyun-Sub Shim,

Jang-Joo Kim, "Fully vacuum-processed MAPbI3 perovskite solar

cells with high open circuit voltage using MoO3/NPB layer", The

１３８

International Symposium on Recent Advances and Future Issues in

Organic Electroluminescence (ISOEL2016), February 17-19

(February 18), 2016, Korea (Poster)

4. Min-Soo Choi, Tae-Min Kim, Hyun-Sub Shim, Beom-Soo Kim,

Hyo Jung Kim, Jang-Joo Kim, "Enhancement of the Fill Factor

through an Increment of the Crystallinity in Fullerene-Based

Small-Molecule Organic Photovoltaic Cells", The International

Symposium on Recent Advances and Future Issues in Organic

Electroluminescence (ISOEL2016), February 17-19 (February 18),

2016, Korea (Poster)

5. Beom-Soo Kim, Tae-Min Kim, Min-Soo Choi, Hyun-Sub Shim,

Jang-Joo Kim, "Fully vacuum co-deposited MAPbI3 perovskite

solar cell with 1.12 V of circuit voltage employing MoO3/NPB

layers as a hole transport layers", The 7th International Workshop

on Flexible & Printable Electronics 2015 (IWFPE 2015),

November 4-6 (November 5), 2015, Korea (Poster)

6. Min-Soo Choi, Tae-Min Kim, Hyun-Sub Shim, Beom-Soo Kim,

Hyo Jung Kim, and Jang-Joo Kim, "Enhancement of The Fill

Factor through An Increase of The Crystallinity in Fullerene-based

Small-Molecule Organic Photovoltaic Devices", KJF-ICOMEP

2015, September 6-9 (September 7), 2015, Korea (Poster)

１３９

7. Beom-Soo Kim, Tae-Min Kim, Min-Soo Choi, Hyun-Sub Shim,

and Jang-Joo Kim, "Fully Vacuum Deposited MAPbI3 Perovskite

Solar Cells with High Open Circuit Voltage Using MoO3/NPB

Layers", KJF-ICOMEP 2015, September 6-9 (September 7), 2015,

Korea (Poster)

8. Min-Soo Choi, Tae-Min Kim, Hyun-Sub Shim, Beom-Soo Kim,

Hyo Jung Kim, and Jang-Joo Kim, "Enhancement of fill factor

through increase of crystallinity in fullerene based small molecule

organic photovoltaic cells", 26th International Conference on

Molecular Electronics and Devices (IC ME&D 2015), May 21-22

(May 22), 2015, Korea (Poster)

9. Beom-Soo Kim, Tae-Min Kim, Min-Soo Choi, Hyun-Sub Shim

and Jang-Joo Kim, "Fully vacuum-processed MAPbI3 perovskite

solar cells with high open circuit voltage using MoO3/NPB layer",

International Conference on Hybrid and Organic Photovoltaics

2015 (HOPV15), May 10-13 (May 12), 2015, Italy (Poster)

10. Min-Soo Choi, Tae-Min Kim, Hyun-Sub Shim, Beom-Soo Kim

and Jang-Joo Kim, "Enhancement of Fill Factor via Increase of

Crystallinity in Fullerene Based Small Molecule Organic

Photovoltaic Cells Using Copper Iodide as a Templating Layer",

The 6th Asian Conference on Organic Electronics (A-COE 2014),

November 12-14 (November 13), 2014, Taiwan (Poster)

１４０

11. Beom-Soo Kim, Tae-Min Kim, Min-Soo Choi, Hyun-Sub Shim

and Jang-Joo Kim, "Fully-Vacuum Processed Perovskite Solar

Cells with High Open Circuit Voltage Using MoO3/NPB Layers",

The 6th Asian Conference on Organic Electronics (A-COE 2014),

November 12-14 (November 13), 2014, Taiwan (Poster)

Domestic Conference

1. 김범수, 최민형, 최민수, 김장주, "Composition-controlled

organic/inorganic hybrid perovskite via MAI pressure in vacuum

co-deposition process", 2016 한국고분자학회 춘계학술대회,

April 6-8 (April 7), 2016, Korea (Poster)

2. 김범수, 김태민, 최민수, 심현섭, 김장주, "Fully vacuum-

processed perovskite solar cells with VOC of 1.12 V employing

MoO3/NPB layer", 2015 한국고분자학회 춘계학술대회, April

8-10 (April 9), 2015, Korea (Poster)

3. 최민수, 김태민, 심현섭, 김범수, 김효정, 김장주,

"Enhancement of fill factor via increase of crystallinity in fullerene

based small molecule organic photovoltaic cells using copper

iodide as a templating layer", 2015 한국고분자학회 춘계학술대

회, April 8-10 (April 9), 2015, Korea

１４１

Invited talks

1. Beom-Soo Kim and Jang-Joo Kim "Vacuum Processed Perovskite

Solar Cells", 페로브스카이트 포토닉스 학술대회 (PPC 2017),

November 20-21 (November 20), 2017, Korea, Keynote

2. Hyun-Sub Shim, Beom-Soo Kim, Jang-Joo Kim, "Vacuum

deposited organic and perovskite solar cells", 2016년 한국유기태

양전지학회 동계 학술대회, February 23-24 (February 23), 2016,

Korea

Awards

- Beom-Soo Kim, Tae-Min Kim, Min-Soo Choi, Hyun-Sub Shim

and Jang-Joo Kim, “Best Poster Award”, The 6th Asian

Conference on Organic Electronics, Nov. 12-14, 2014