· ! 5! abstract! overthepastthreedecades,tremendousamountofresearchhas...
TRANSCRIPT
Degradation Pathways in Organic Small Molecule and Hybrid Solar Cells
GOLNAZ SHERAFATIPOUR NanoSYD-Mads Clausen Institute SDU Sønderborg-University of Southern Denmark
2
3
Degradation Pathways in Organic Small Molecule and Hybrid Solar Cells
Golnaz Sherafatipour
Doctoral Thesis
2018
Nanoscience Centre NanoSYD Mads Clausen Institute Faculty of Engineering
SYDDANSK UNIVERSITET
Supervisor:
Morten Madsen
4
5
Abstract
Over the past three decades, tremendous amount of research has been devoted to the field of renewable energy in order to reduce our dependence on fossil fuels. Among them, organic (OSC) and perovskite (PSC) solar cells have gained enormous attention as they offer unique advantages compared to the traditional silicon solar cells, such as low fabrication cost, mechanical flexibility, light-‐weight modules and semi-‐transparency. These unique properties offer a wide range of applications and integration schemes, which make this field of research even more exciting. OSC and PSC have recently achieved power conversion efficiencies (PCE) of around 15% and 22%, respectively, emphasizing the great potential of the technologies. However, both of these devices suffer from low stability and short lifetime (degradation). Therefore, understanding the degradation mechanisms of these devices, paves the way for viable commercialization of this appealing technology in the market. This work is dedicated to investigate governing degradation mechanisms and pathways taking place inside organic and perovskite solar cell devices. First part of the work focuses on the performance and stability of DBP-‐C70 based organic solar cells in standard and inverted device configurations. We study their device stabilities by aging them under ISOS-‐D-‐3 (darkness, 85℃ and 85% RH-‐humidity) and ISOS-‐T-‐3 (darkness, -‐40℃ and room humidity) conditions. The results show that despite a change in the performance upon aging, there is a pronounced morphological stability at the DBP-‐C70 interface. Possible effects from the electron transport layer (ETL) on the device stability were investigated, demonstrating that this layer contributes significantly to the degradation of the inverted devices. The second part of this work focuses on understanding the degradation mechanisms taking place inside perovskite solar cells under real operational conditions. Results for indoor (ISOS-‐L-‐1, illumination, 60℃ and ambient humidity) and outdoor (ISOS-‐O-‐1, sunlight, ambient) degradation test conditions are conducted, showing reversible and irreversible degradation mechanisms under light-‐darkness cycles, which reveal interesting degradation pathways and emphasize on the importance of including these cycles in experimental protocols for the assessment of long-‐term stability of the PSCs.
6
7
Resume
I de seneste tre årtier har forskningen inden vedvarende energi været stærkt stigende, således at mængden og afhængigheden af fossile brændstoffer kan reduceres i fremtiden. Blandt teknologier indenfor vedvarende energi har organiske (OSC) og perovskite (PSC) solceller tiltrukket meget opmærksomhed, da de tilbyder unikke fordele sammenlignet med traditionelle silicium solceller, såsom lave fabrikationsomkostninger, mekanisk fleksibilitet, samt lette og gennemsigtige moduler. Disse unikke egenskaber tilbyder en lang række applikationer og integrationsmuligheder, som gør dette forskningsfelt yderligere interessant. OSC og PSC har for nyligt opnået ydeevner på hhv. 15% og 22%, hvilket understreger deres store potentiale. Begge disse typer solceller lider dog under relativt lav stabilitet og levetid, hvorfor en forståelse for degraderingsmekanismerne der finder sted i disse celler kan være med til at bane vejen for en kommercialisering af denne lovende teknologi.
Dette arbejde er fokuseret mod at undersøge de grundlæggende degraderingsmekanismer og retninger som finder sted i organiske og perovskite solceller. I første del af arbejdet undersøges ydeevnen og stabilitet af DBP-‐C70 organiske solceller i standard og inverteret konfigurationer. Vi studerer deres stabilitet ved at degradere dem under ISOS-‐D-‐3 (mørke, 85℃ og 85% luftfugtighed) og ISOS-‐T-‐3 (mørke, -‐40℃ og normal luftfugtighed) betingelser. Resultaterne viser at der, på trods af ændringer i ydeevnen under degradering, er en udtalt morfologisk stabilitet ved DBP-‐C70 grænsefladen. Mulige effekter fra elektron transport laget på solcelle stabiliteten blev undersøgt, hvilket demonstrerede at dette lag bidrager væsentligt til degradering af de inverterede solceller. Den anden del af dette arbejde fokuserer på forståelsen af de degraderingsmekanismer der finder sted I perovskite solceller under reelle arbejdsbetingelser. Resultater for indendørs (ISOS-‐L-‐1, belysning, 60℃ og rum luftfugtighed) og udendørs (ISOS-‐O-‐1, sollys) degraderingstest betingelser blev udført, og viste reversible og ikke-‐reversible degraderingsmekanismer under lys-‐mørke overgange, hvilket afslører interessante degraderingsmekanismer og understreger vigtigheden i at inkludere disse overgange i eksperimentelle stabilitetsprotokoller for PSC solceller.
8
9
Acknowledgments
I would like to express my gratitude to everyone who supported me throughout the course of my PhD studies. This work would not have been possible without their help and contribution in one way or another.
First and foremost, I am greatly thankful to my supervisor Dr. Morten Madsen for giving me the opportunity to join NanoSYD and be part of the THINFACE project. His aspiring guidance, patience and enthusiasm have always motivated me, and his friendliness and openness releases inevitable tensions of a PhD work.
Next, I would like to thank my supervisor at TU Dresden, Prof. Koen Vandewal, for sharing his incredible knowledge and skills in the field and all his support and encouragement during my stay at TUD and through this research work.
Thank you Prof. Horst-‐Günter Rubahn and Dr. Katharina Rubahn, for all your support and friendly advice during my PhD studies and THINFACE project.
My colleagues/friends, Mina, Peyman, Andre, Mehrad, Bhushan, Ela, Pawel, Michela, Elodie, Jani, Arkadiusz, Vida, Elahe and Fatemeh thank you for all the nice time and happy memories inside and outside of the office. Mina and Peyman, I was so lucky to have your valuable friendship during the long and dark days when things can challenge you to your very core. You turned all of them into great memories.
Thanks to beautiful souls around the department, Sabina, Charlotte, Lise, Ferran, Zora, Roana and Luciana for their kind and cheering attitude.
I would also like to thank our lab technicians and manager who are the backbone of the lab and the facilities, Mogens, Reiner, Jens and Arkadiusz.
10
I would especially like to thank my colleagues/friends at IAPP, TU-‐Dresden Johannes Benduhn for the sEQE measurements, and Dr. Donato Spoltore for his help with the device fabrication. It was a pleasure to work with you. I enjoyed the energetic and friendly research environment of IAPP. You made my stay memorable!
To my love Gerrit, I am deeply thankful for your unconditional support and love. You made it possible to come this far. You are truly one of a kind!
Last but not least, I am extremely grateful to my beloved family for their love and never ending support throughout all these years. In every moment of my life they have always stood by me and cheered me up.
11
List Of Papers I. Degradation pathways in standard and inverted DBP-‐C70 based organic solar cells Golnaz Sherafatipour, Johannes Benduhn, Bhushan Ramesh Patil, Mehrad Ahmadpour, Donato Spoltore, Horst-‐Günter Rubahn, Koen Vandewal, Morten Madsen Manuscript II. Dynamics of photoinduced degradation of perovskite photovoltaics: from reversible to irreversible processes Mark V. Khenkin, Anoop K. M., Iris Visoly-‐Fisher, Sofiya Kolusheva, Yulia Galagan, Francesco Di Giacomo, Olivera Vukovic, Bhushan Ramesh Patil, Golnaz Sherafatipour, Vida Turkovic, Horst-‐Günter Rubahn, Morten Madsen, Alexander V. Mazanik, and Eugene A. Katz ACS Appl. Energy Materials III. Reconsidering figures of merit for performance and stability of perovskite photovoltaics Mark V. Khenkin, Anoop K. M., Iris Visoly-‐Fisher, Yulia Galagan, Francesco Di Giacomo, Bhushan Ramesh Patil, Golnaz Sherafatipour, Vida Turkovic, Horst-‐Günter Rubahn, Morten Madsen, Tamara Merckx, Griet Uytterhoeven, João P. A. Bastos , Tom Aernouts, Francesca Brunetti, Monica Lira-‐Cantu and Eugene A. Katz RSC Energy & environmental science IV. Area-‐dependent behavior of bathocuproine (BCP) as cathode interfacial layers in organic photovoltaic cells Bhushan Ramesh Patil, Mehrad Ahmadpour, Golnaz Sherafatipour, Talha Qamar, Antón F. Fernández, Karin Zojer, Horst-‐Günter Rubahn, Morten Madsen Scientific Reports-‐Nature-‐Submitted, 2018
12
13
Contents
List of Figures ......................................................................................................... 15
1 Introduction ....................................................................................................... 19 1.1 A green horizon ........................................................................................................ 19 1.2 Photovoltaic devices .............................................................................................. 20 1.3 organic solar cells .................................................................................................... 21 1.3.1 Challenges toward commercialization of organic solar cells ...... 22
1.4 Perovskite solar cells ............................................................................................. 23 1.4.1 Challenges toward commercialization of perovskite solar cells 23
1.5 Aim of the thesis ...................................................................................................... 24 2 Fundamentals .................................................................................................... 27 2.1 Organic semiconductors ....................................................................................... 27 2.2 Operation principles of organic solar cells (OSCs) ................................... 29 2.2.1 exciton generation ......................................................................................... 29 2.2.2 exciton diffusion and dissociation .......................................................... 30 2.2.3 carrier transport ............................................................................................. 35 2.2.4 charge extraction at electrodes ................................................................ 36 2.2.5 Summery of the operation .......................................................................... 37
2.3 Solar cell architectures ......................................................................................... 38 2.3.1 Planar vs. bulk heterojunction .................................................................. 38 2.3.2 Standard vs. inverted structure ............................................................... 39
2.4 Materials ...................................................................................................................... 40 2.4.1 Donor and Acceptor ...................................................................................... 40 2.4.2 buffer layers ...................................................................................................... 42 2.4.3 Contacts .............................................................................................................. 44
2.5 Fabrication techniques ......................................................................................... 44 2.6 Stability of organic solar cells ............................................................................ 46 2.6.1 Intrinsic degradation .................................................................................... 46 2.6.2 Extrinsic degradation ................................................................................... 46
2.7 Perovskite solar cells ............................................................................................. 47 2.7.1 Fabrication ........................................................................................................ 49 2.7.2 Stability ............................................................................................................... 49
2.8 Characterization ...................................................................................................... 50 2.8.1 J-‐V characteristics .......................................................................................... 50 2.8.2 ISOS degradation tests ................................................................................. 56
2.9 Relation between VOC and CT states ................................................................ 56 3 Experimental Setup ........................................................................................ 61 3.1 Device fabrication ................................................................................................... 61 3.1.1 patterning of the substrates ...................................................................... 61 3.1.2 Pre-‐cleaning the substrates ....................................................................... 62
14
3.1.3 Vacuum deposition of organic/recombination layers and top electrode ........................................................................................................................ 63 3.1.4 Final structure ................................................................................................. 66
3.2 Characterization ...................................................................................................... 67 3.2.1 J-‐V measurements .......................................................................................... 67 3.2.2 Sensitive external quantum efficiency (sEQE) measurements .. 68 3.2.3 Degradation protocols ................................................................................. 69 3.2.4 Morphological characterization ............................................................... 70 3.2.5 Photoluminescence quenching measurements ................................. 70
4 Degradation pathways in Standard and Inverted DBP-‐C70 Based Organic Solar Cells ................................................................................. 73 4.1 Device Performance: .............................................................................................. 75 Standard vs. Inverted configuration ....................................................................... 75 4.2 Sensitive EQE measurements ............................................................................ 77 4.3 Morphology investigation at the D-‐A interface .......................................... 78 4.4 Degradation studies ............................................................................................... 80 4.5 ETL charge transport properties and its effect on the stability .......... 85 4.5.1 Electron-‐only devices ................................................................................... 86 4.5.2 Photoluminescence quenching measurements ................................. 88 4.5.3 Improved stability of the inverted devices ......................................... 90
5 Understanding the degradation mechanisms in perovskite solar cells ................................................................................................................... 93 5.1. Outdoor day/night degradation and recovery tests ............................... 94 5.2 Indoor degradation and recovery dynamics ............................................... 99 5.3 Correlation between indoor and outdoor stability measurements . 106
6 Summary and outlook ................................................................................ 109 Appendix A ............................................................................................................ 113 ISOS-‐T-‐3 degradation test for standard structure .......................................... 113
Appendix B ............................................................................................................ 115 TTF as donor material in organic solar cells ..................................................... 115
Bibliography ......................................................................................................... 133
15
List of Figures
Figure 1.1. Three generations of solar cells. Mono crystalline and poly crystalline Silicon based solar cells, thin film solar cells and organic solar cells .......................................................................................... 21
Figure 2.1. Illustration of HOMO and LUMO level of an organic material with bonding-‐antibonding interactions. ........................... 28
Figure 2.2. Three differnet types of excitons. a) Wannier-‐Mott excitons, with delocalized and weakily bound presented in inorganic materilas b) Strongly bound Frenklel excitons entirely located on one molecule in organic semiconductotors. ............... 30
Figure 2.3. Left) Charge-‐trasfer excitons, located on the adjacent molecules at the donor-‐acceptor hetetojunction; right a) Ground state at which the molecules are neutral and the HOMO of the molecules is filled; right b) Charge transfer state at which the donor is positively and the acceptor is negatively charges. Electron is transferred from donor to the acceptor molecule and it is still coulombically bounded to the hole in the donor molecule. ........................................................................................................... 31
Figure 2.4. Marcus theory. Electron transport can occur if the initial state overcomes the barrier of ΔG* to reach the crossing point of the potentials. Marcus derived the height of the barrier. ............ 34
Figure 2.5. The principle of charge separation in an organic solar cell operation. Summary of the 4 main steps. ........................................... 37
Figure 2.6. a) Planar heterojunction or bilayer solar cell b)Bulk heterojunction solar cell with a donor/acceptor blended active layer for more efficient charge generation. ....................................... 38
Figure 2.7. Two main geometries of PHJ organic solar cells: a) standard and b) inverted configuration. ............................................. 40
Figure 2.8. Skeletal formula of tetraphenyldibenzoperiflanthene (DBP). ................................................................................................................. 41
Figure 2.9. Schematic drawing (a) soccer-‐ball structure of buckyball C60 and (b) rugby-‐ball structure of common fullerene C70. (c) C70 Powder ............................................................................................................... 42
Figure 2.10. (left) Crystal structure of molybdenum trioxide (Hyung-‐Seok Kim/Nature Materials), (right) MoO3 powder ...................... 43
Figure 2.11. Left) skeletal formula of Bathocuproine, 2,9-‐dimethy-‐4, 7-‐diphenyl -‐1,10-‐phenathroline (BCP), right) BCP powder. ...... 44
Figure 2.12. Schematic drawing of spin coating procedure. Desired solution is applied over a substrate located on a spinner, then
16
the spinner is rotated and the material is spread over the surface by centrifugal force to form a film. ........................................ 44
Figure 2.13. Thermal evaporation chamber. Substrate is located upside down above the source. Material is thermally evaporated and is deposited on the surface of the rotating substrate. A Crystal monitors the deposition rate. ................................................... 45
Figure 2.14. Generic perovskite crystalline lattice arrangement of the form ABX3. Note that the lines represent crystal orientation and not the bonding patterns. The two structure are equivalent with left) atom B at the <0,0,0> position and right) atom A at the <0,0,0> position. ............................................................................................. 48
Figure 2. 15. Electric circuit equivalent to an OPV device. .................... 51 Figure 2. 16. Current-‐voltage (I-‐V) curve of an organic solar cell. a)
Without illumination, the solar cell behaves like a diode, b) after illumination, the curve is shifted downward as the cell start to generate power, c) increasing the light intensity results in further shift of the curve downward as the generated power increases. ........................................................................................................... 52
Figure 2.17. Operation of a solar cell under different applied biases; 1) large reverse bias; 2) small reverse bias; 3) positive bias, results in zero initial field, and corresponds to open-‐circuit condition; 4) positive bias and carrier injection .............................. 53
Figure 2.18. Solar cell J-‐V curve metrics. ....................................................... 54 Figure 2.19. a) free energy diagram for the ground state and lowest
excited state. b) Reduced EQE PV and EL spectrum with fits using formulas (10) and (11). Each parameter is indicated in the figure. .................................................................................................................. 59
Figure 3.1. Layout of the patterned ITO coated glass substrates. ...... 62 Figure 3.2. Cryofox deposition cluster system. Organic materials and
metals are deposited in two separated chambers in ultra high vacuum conditions. A robotic arm transfers the sample between the chambers without breaking the vacuum between the steps. Right. The system is connected to a glove box to avoid exposure of the samples to air. (http://www.polyteknik.com/index.php) ............................................................................................................................... 64
Figure 3.3. Organic sources located inside the ultra high vacuum chamber. Each source is placed inside crucible, and a shutter to cover the source while not in use. .......................................................... 65
Figure 3.4. Final outlook of the DBP-‐C70 based organic solar cells. Each sample consists of 4 cells. Overlap of the top electrode with
17
the bottom ITO electrode defines the cell area, which here is 10 mm2 ..................................................................................................................... 66
Figure 3.5. Final illustration of DBP-‐C70 based organic solar cells with a) standard and b) inverted configurations. C and d show energy band diagram for each structure. ........................................................... 67
Figure 3. 6. Sample holder with 4 sample positions ................................ 68 Figure 3.7. Sketch of the sensitive EQE measurement setup120. 69
Figure 4.1. J-‐V curves of DBP-‐C70 organic based PHJ devices having standard or inverted configurations. ........................................................ 76
Figure 4.2. sEQE measurements at 300 K and Marcus fits for standard and inverted structures. Dashed lines are fits to the EQE using Marcus theory. ..................................................................................................... 78
Figure 4.3. AFM images of interface layer (a) DBP on MoO3/ITO, b) C70 on DBP/MoO3/ITO, c)C70 on BCP/ITO, and d) DBP on C70/BCP/ITO .................................................................................................................................... 79
Figure 4.4. J-‐V curves for fresh (solid lines) and aged (dashed lines) devices at a) ISOS-‐D-‐3 b) and ISOS-‐T-‐3 degradation conditions. . 81
Figure 4.5. sEQE measurements and their corresponding Fits for fresh and aged devices at a) ISOS D-‐3 and b) ISOS-‐T-‐3 degradation conditions. ............................................................................................................. 83
Figure 4.6. sEQE measurements and Marcus Fits for fresh and annealed devices at 110 ℃ for 3 hours ......................................................................... 84
Figure 4.7. a) Electron only devices with different ETLs, and with deposited 10 nm BCP and 100 Ag on top b) JV measurements of the fresh EODs c) JV measurements after aging devices for 24 hours in ISOS-‐D3 and d) ISOS-‐T-‐3 degradation conditions. ............ 87
Figure 4.8. Photoluminescence (PL) measurements for five ETLs, fresh and degraded at ISOS-‐D-‐3 and ISOS-‐T-‐3. The stack has the structure: (0.5nm)/BCP (0.5nm) and BCP (2nm)/Ag (1nm)/BCP (2nm)/Ag (1nm)/BCP (2nm)). ........................................................................ 89
Figure 4.9. JV measurements for fresh and aged inverted device with 0.5 nm BCP and BCP/Ag stack. ..................................................................... 91
Figure 5.1. Normalized PCE evolution for indoor contentious simulated sunlight illumination of glass/ITO/SnO2/Cs0.05((CH3NH3)0.15(CH(NH2)2)0.85)0.95PbI2.55Br0.45/spiro-‐OMeTAD/Au cells (type I). .................................................................. 96
Figure 5.2. Normalized PCE evolution during two weeks of outdoor exposure to natural sunlight (a) type I (glass/ITO/SnO2/Cs0.05((CH3NH3)0.15(CH(NH2)2)0.85)0.95PbI2.55Br0.45/spiro-‐OMeTAD/Au), and (b) type II mini-‐modules
18
(glass/ITO/TiO2/CH3NH3PbI3/Spiro-‐OMeTAD/Au). All lines are guides for the eye. .............................................................................................. 96
Figure 5.3. Normalized evolution of daily energy output, Eday, of (a) cell type I, and (b) mini-‐modules type II. All lines are guides for the eye. ........................................................................................................................... 98
Figure 5.4. Evolution of PV parameters of PSCs under continuous 1-‐sun indoor illumination, interrupted at T80 (a) T60 (b), or T50 (c, d). Gray areas show their subsequent recovery in the dark. ............... 100
Figure 5.5. Evolution of PV parameters after turning on the light and continuous simulated 1-‐sun re-‐illumination of the PSCs after T50 and recovery in the dark (gray areas): (a) The cell whose PCE dark recovery reached saturation (as in Figure 5.4.c); (b) The cell whose PCE dark recovery did not reach saturation (as in Figure 5.4.d). ..................................................................................................................... 100
19
CHAPTER 1
Introduction
1.1 A green horizon The future generation is facing a broad and complex topic, which
is the foreseeable world’s energy crisis. The main supply of energy
today is Fossil fuels, however, these natural resources are in limited
supply and consumption of them is giving rise to further climate
change. By 2020, the global consumption of energy is expected to
increase by 50 percent, and the main debate is to find alternatives to
meet this demand. The best solution is to replace these finite
resources with renewable ones, known as green energy such as
wind, solar, biomass and geothermal. Green energy utilizes energy
resources that are readily available all over the world.
During the past three decades, there has been a tremendous
amount of research and development in the field of green energy
that has helped to reduce our dependence on fossil fuels. Among all
the renewable resources, solar energy has gained enormous
attention in past decades. Harnessing energy from the sun is the
20
most reliable and pragmatic approach to cater the global energy
needs. In fact, the energy from the sun that hits the earth’s surface in
just one hour can power the whole planet in energy for one year.
However, utilizing this great power to its maximum potential is
limited by many factors such as materials, manufacturing, etc.
1.2 Photovoltaic devices In order to harvest the sunlight, we use devices that can convert
sunlight into electricity. These devices are called solar cells, and the
process of converting sunlight into electricity is called photovoltaic;
therefore, solar cells are also known as photovoltaic devices.
Solar cells are categorized into three generations according to the
time sequence1 (see Figure 1.1). First generation solar cells are
crystalline and multi-‐crystalline Silicon (Si), with simple constitution,
but high manufacturing cost. The theoretical efficiency of this type of
solar cells is of about 30%2. Efficiency of a solar cell refers to the
portion of sunlight that can be converted into electrical power via
the photovoltaic effect.
Second generation solar cells or thin film solar cells made out of
amorphous Si with lower manufacturing cost, but poor stability and
inherent problems. The most successful materials of this generation
are Cu(In,Ga)Se2 (CIGS) and CdTe/CdS. The lab efficiency for this
generation is around 19%.
The third generation solar cells introduce new concepts in terms
of device architecture and materials; for example the idea of multi
junction solar cells to improve harvesting of photons and overcome
the 30% limit for efficiency. This 3rd generation includes dye-‐
21
sensitized (DSSCs)3,4 and Organic Solar Cells (OSCs), quantum dot
(QD) and Perovskite photovoltaic. DSSCs are fabricated using dyes,
metal oxides and electrolyte, and their efficiencies are in the range of
12% for small lab scale devices. However, the lifetime of the devices
are low compared to inorganic solar cells.
Figure 1.1. Three generations of solar cells. Mono crystalline and poly crystalline Silicon based solar cells, thin film solar cells and organic solar cells
1.3 organic solar cells An organic solar cell is a type of an organic optoelectronic device,
that deals with conductive low weight organic materials such as
small molecules and oligomers, or high weight molecules, i.e.,
polymers, which is vastly used5. These devices use photovoltaic
effect to directly convert sunlight into direct current (DC) electricity.
In 1906, Pochettino observed photoconductivity in an organic
compound named Anthracene6, and it was the beginning of a new era
for application of organic compounds in electronics.
In 1954, high conductivity in perylene-‐iodine complex was
discovered7, after which organic semiconductors attracted great
amount of attention. In recent decades, organic solar cells (OSCs)
have been under intense research due to their interesting
22
advantages compared to first and second generation. They exhibit
higher absorption coefficient, which makes it possible to fabricate
very thin films with low amount of materials. Moreover,
manufacturing them is easy and cost-‐effective7–10, and the final
product is flexible, semitransparent and lightweight. They also
provide a huge potential for large area applications and portable
solar panels. This means they can be printed by roll-‐to-‐roll (R2R)
machinery. These advantages connote a great potential for organic
solar cells to make a significant impact on the future of the PV
market. More details and technical information about organic solar
cells is provided in chapter 2.
1.3.1 Challenges toward commercialization of organic solar cells Despite all the advantages for organic solar cells, still achieving a
high efficiency and long-‐term stability11 is remaining the bottleneck
for potential commercialization of this appealing technology. To this
date, a world-‐record of 15% efficiency has been achieved for organic
solar cells12, however, this still needs to be improved to compete
with current photovoltaic technologies on the market.
Low stability and short lifetime of organic solar cells means that
over time, they loose their photovoltaic ability, and their efficiency is
decreased. These changes can be caused by extrinsic or intrinsic
factors such as oxygen and moisture from the air or dynamic and
active nature of organic materials11,13 (more details in chapter 2).
Despite improvements regarding these issues14, loss mechanisms
and degradation patterns yet need to be fully understood. In order to
push the limitations for application of organic solar cell beyond its
23
boundaries, a deep understanding of the degradation process that
limits and controls the stability of the devices is crucial. Our
comprehensions can serve as a guide for finding new solutions and
identify new materials for more advanced and stable next-‐generation
solar cells.
1.4 Perovskite solar cells Perovskite solar cells are another type of solar cells that have
attracted prominent attention over the past few years due to their
high efficiency of around 20%15, which is comparable to other types
of inorganic photovoltaic such as Cadmium Telluride. Their high
efficiency have made them rising star of the photovoltaic world since
their breakthrough paper of 201216, and a huge interest to the
academic community (more details in section 2.7).
Over the past two years, improvements in fabrication routines
and engineering of perovskite formulations has led to significant
increase in their power conversion efficiency which reached over
22%, as of April 201717. This dramatic rise is incredibly impressive
and the efficiency is comparable to Cadmium Telluride, which has
been used for near 40 years. Furthermore, their potential for much
lower processing cost makes them significantly interesting for its
future market.
1.4.1 Challenges toward commercialization of perovskite solar cells Despite their high efficiency, perovskite solar cells suffer from a
low stability and short operational lifetime. There are several
reaction pathways involving, oxygen, water and diffusion of metal
24
from the electrodes that leads to degradation (more details in section
2.7.2)18–25. Another problem is the use of lead in perovskite
compounds and products for commercial use that is a huge source of
toxic pollution. A lead alternative such as tin-‐based perovskites is
possible, however, the power conversion efficiency of such devices is
much behind the lead-‐based devices26.
Finally another issue is the lower optical density of these
materials, which means a higher thickness of the light-‐harvesting
layer is needed compared to organic solar cells, which again results
in some fabrication limitations in solution processed devices27.
Perovskite solar cells can also be fabricated based on vacuum
deposition techniques to give a better, more uniform film qualities.
However, evaporation of organic and inorganic materials requires
special evaporation chambers that are not available in many
research labs, and not easy to control in large area depositions.
Previously, vacuum deposition techniques offered the highest
efficiency devices, but recently, through the improvements in
solution-‐based deposition techniques, the record-‐breaking devices
has shifted to solution-‐based processing28.
Overall, to enable a reliable low cost-‐per-‐watt energy solution,
perovskite solar cells need to obtain longer device lifetime, and low
manufacturing costs. Although this has not yet been achieved,
perovskite-‐based solar cells still demonstrate enormous potential to
achieve this.
1.5 Aim of the thesis The broad title of this dissertation comprises degradation studies
on organic and perovskite solar cells. The presented work takes the
25
opportunity to investigate governing mechanisms taking place inside
these devices. For organic solar cells, we focus on factors affecting
the open-‐circuit value as one of the most important parameters
affecting both performance and stability of the devices. In this work
small molecule solar cells based on DBP and C70 are investigated. For
perovskite solar cells, their stabilities are tested under real
operational conditions and a correction between already existing
stability measurements for indoor and outdoor testing conditions is
suggested. The results points on possible and specific degradation
pathways that are important to understand in detail, in order to
facilitate the viable commercialization of organic and perovskite
solar cells in the future. The work is structured as follows.
Chapter 2 provides background information regarding organic
and perovskite solar cells, principles of operation, materials and
characterization techniques.
Chapter 3 describes the fabrication steps including the
measurement and characterization setups employed in this work.
Chapter 4 presents the results and discussion regarding the
performance and degradation mechanism in the DBP-‐C70 organic
solar cells investigated in this work.
Chapter 5 presents results and discussion related to degradation
mechanism in the perovskite solar cells tested under both indoor
cycling protocols and outdoor testing.
Chapter 6 provides conclusion remarks as well as an outlook for
the future studies.
26
27
CHAPTER 2
Fundamentals This chapter introduces basics of the organic and perovskite solar
cell (OSC) technologies, including the characterization of those.
Understanding these concepts is required to follow the studied
material subjected to this work.
2.1 Organic semiconductors Organic semiconductors are carbon-‐based materials in which
carbon atoms are covalently bonded to each other by alternating
single and double bonds (conjugated π-‐bonds), and a weak Van der
Waal’s force that holds two organic molecules within the solid29,30.
The bonding structure is the property that gives organic
semiconductors its unique advantages such as flexibility, lightweight,
and easy processing.
The band structure of an organic semiconductor can be viewed
similarly as inorganic semiconductor in which the valance band is
filled with electrons and conduction band is free of electrons. In
28
organic semiconductors, the lowest electronic transition (and the
most probable one) is between the π–band, which is the Highest
Occupied Molecular Orbital (HOMO), and the π*–band, which is the
Lowest Unoccupied Molecular Orbital (LUMO), (Figure 2.1). These
types of molecular orbitals are analogous to the valance and
conduction band of an inorganic semiconductor, respectively29.
There are techniques such as cyclic voltammetry and photoemission
yield spectroscopy to measure the HOMO and LUMO level of the
organic materials31.
The energy difference between the HOMO and LUMO in an
organic semiconductor is denoted as the band gap of the material,
which is generally between 1.1-‐3.5 eV. Hence, optical excitation can
happen in the range of visible light and near infra-‐red32.
When an electron is excited from HOMO to LUMO, it causes the
excitation of the whole molecule itself into a higher energy state.
This is different than the actual excitation of a free electron from the
valance band to the conduction band in inorganic semiconductor33 .
Figure 2.1. Illustration of HOMO and LUMO level of an organic material with bonding-‐antibonding interactions.
29
2.2 Operation principles of organic solar cells (OSCs) Operation of an organic solar cell can be simplified into 4 main
steps:
1. Photon absorption resulting in formation of electron-‐hole pairs
(exciton)
2. Exciton diffusion to heterojuction
3. Exciton dissociation at the heterojunction
4. Carrier transport and extraction at the electrodes
Each of these steps will be described here in detail.
2.2.1 exciton generation When light is shined over an organic solar cell, photons are
absorbed at the active organic materials. If the photon energy
exceeds the band gap of the material, it excites an electron to the
LUMO level, while leaving a hole in the HOMO. Due to low dielectric
constant (2-‐4)34 of the organic materials and low screening of
charges, upon transition of an electron from HOMO to LUMO, a
coulombically bounded electron-‐hole pair known as exciton is
formed with a binding energy of 0.1-‐1.4 eV. These types of excitons
are known as Frenkel excitons, which are entirely located on one
molecule, Figure 2.2.b. However, in inorganic semiconductors due to
a higher dielectric constant (12-‐16), the Coulomb attraction is
weakened and screening is increased; therefore, the average radius
between electron and holes is larger than the lattice spacing, leading
to a much lower binding energy of only 5-‐15 meV33. These types of
excitons are known as Wannier-‐Mott excitons and their binding
energies are sufficiently below thermal energy at room temperature
30
(KBT≈25 meV), which results in a higher probability of free charge
carrier generation after absorption of photons and dissociation of
the pairs by absorbing thermal energy, Figure 2.2.a.
Due to the fact that the absorption coefficient of organic materials
is higher than inorganic materials (~105 cm-‐1), a thin layer of a few
hundreds of nanometers of the active layer is sufficient to absorb
enough amount of light and generate carriers.
Figure 2.2. Three differnet types of excitons. a) Wannier-‐Mott excitons, with delocalized and weakily bound presented in inorganic materilas b) Strongly bound Frenklel excitons entirely located on one molecule in organic semiconductotors.
2.2.2 exciton diffusion and dissociation After excitons are formed, next step is to separate them and
generate free charges, which eventually leads to generation of
current (electricity). To achieve an efficient exciton pair separation,
Tang et al.10 provided a solution by employing two different kinds of
organic materials with properly aligned band energies. One material
acts as electron donor and the other is an electron acceptor, resulting
in the basic concept of the so-‐called donor-‐acceptor (D/A) in OSCs.
The interface between the two materials is called the D/A
heterojunction. The acceptor material is a strongly electronegative
31
and when placing the two materials adjacently, the electron can go to
a much lower energy state within the acceptor. This causes the
dissociation of the exciton at the heterojunction. When an exciton is
generated in the electron donor material, it migrates towards the
heterojunction, and charge transfer from donor to acceptor initiates
if the difference between the LUMO energy levels of the donor and
acceptor overcome the exciton binding energy. This makes the
electron to transfer from exciton to LUMO of the acceptor while a
hole remains in HOMO of the donor35, Figure 2.5.3.
2.2.2.1 Charge transfer states After the charge transfer at the heterojunction, the electron-‐hole
pair is still Coulombically bound and are located at the D/A interface
on adjacent but different molecules. This new type of exciton, which
is between the two other types in terms of binding energy and
binding distance, is called charge transfer (CT) exciton and its in a
charge transfer state, Figure 2.3.
Figure 2.3. Left) Charge-‐trasfer excitons, located on the adjacent molecules at the donor-‐acceptor hetetojunction; right a) Ground state at which the molecules are neutral and the HOMO of the molecules is filled; right b) Charge transfer state at which the donor is positively and the acceptor is negatively charges. Electron is transferred from donor to the acceptor molecule and it is still coulombically bounded to the hole in the donor molecule.
32
The charge separation at the CT state is about 1 nm, which is
roughly the distance between the donor and acceptor molecules. The
energy of the CT state is defined as the energy difference between
the ionization potential of the donor and electron affinity of the
acceptor (IPD-‐EAA) plus EB which is a term to account for the binding
energy of the CT which is typically estimated at approximately 0.1-‐
0.5 eV.
The energy of the CT state depends on different factors such as
the composition of the donor and acceptor molecules, processing
treatments, etc36. For example, Loi et al. noted that increasing the
concentration of the high dielectric constant PCBM in F8DTBT:PCBM
blend results in lowering the CT state energy due to an increased
effective dielectric constant of the blend37–39. In addition, it reduces
the effective Coulombic interaction between electron and hole, and
facilitates CT state dissociation into free carriers40. Moreover, it has
been reported that thermal annealing of the P3HT:PCBM decrease
the energy of its CT state38,41–43.
As mentioned before, CT pairs at the heterojunction are still
Coulombically bound and need to be separated by an internal field
before they recombine. In fact, CT states are intermediate states
between exciton recombination and dissociation; hence, they play a
crucial role in charge generation in organic photovoltaics44–48.
2.2.2.2 Marcus theory From a molecular perspective, charge transfer from donor
molecule to acceptor molecule is describing by a theory known as
Marcus theory49. In 1956 R. A. Marcus won a Nobel Prize for
introducing a method for calculating rates of electron transfer
33
reactions at which electrons can hop from electron donor molecule
to electron acceptor. This theory is now widely used in chemistry
and physics.
In Marcus theory, the potential of reactants and products are
sketched as parabolas, Figure 2.4, In this sketch, ΔG° is free energy
changes between the reactants and products (difference between
potential minima), ΔG* is activation energy and λ is reorganization
energy, which is the energy required to force the reactants to have
the same nuclear configuration as the products without electron
transfer. Electron transfer can only occur if the excited D/A pair
(D*/A) overcomes the barrier of ΔG* to reach the crossing point of
the potential wells. Marcus derived the height of the ΔG* from the
thermodynamic parameters of the system:
ΔG∗ =(𝜆 + 𝛥𝐺°)!
4𝜆 (2.1)
In organic solar cells, the theory has been successfully applied to
describe the absorption and emission shape of charge transfer states
in D-‐A systems.
In Figure 2.4, a parabola with the lowest energy shows the relaxed
ground state for donor and acceptor interface. After absorption of
the light, one of the molecules is excited (D* or A*) and the molecular
conformation changes at the D/A interface. If the excitation energy
can overcome the barrier ΔG*, the Frenkel exciton being localized at
one molecule separates at D/A interface and forms CT exciton (D*/A
D+/A-‐).
34
Figure 2.4. Marcus theory. Electron transport can occur if the initial state overcomes the barrier of ΔG* to reach the crossing point of the potentials. Marcus derived the height of the barrier.
2.2.2.3 Recombination As mentioned earlier, due to the low dielectric constant of the
organic materials, excited electron-‐hole pairs are still Coulombically
bound after dissociation and while they form a CT exciton. The pair
is called a geminate pair and before they fully dissociate into free
charge carriers, they can still recombine back to the ground state in a
process known as geminate recombination. Since charge transfer
process occurs faster than the charge recombination (~45 fm vs. ~1
ns)50,51, efficient exciton dissociation at the heterojunction is
possible.
Even after dissociation, recombination can happen again if the
dissociated free electrons and holes encounter each other once more.
This time, since the two carriers are separated, free carriers, this
process is called a nongeminate recombination.
In addition to these two situations, if after absorption of light,
generated excitons cannot reach the interface, they recombine back
35
to their ground state. The maximum length that an exciton can
diffuse before recombination is called the exciton diffusion length. In
organic materials this distance is a few tens of nanometer52,53. If the
distance where the excitons are generated is longer than their
diffusion length, they recombine before reaching the heterojunction.
Hence, in organic solar cells, the active layer is kept thin to make
sure that it is within the exciton dissociation length. However, a
thinner active layer can result in low absorption efficiency.
Therefore, as a solution to this tradeoff, bulk heterojunction solar
cells were invented to increase the efficient exciton dissociation by
increasing the amount of the interface between the acceptor and
donor materials.
In all cases being explained above, the energy of the photon is lost
in the form of radiation that results in fewer carriers that are
collected at the electrodes.
2.2.3 carrier transport After dissociation, generated electrons and holes have to travel to
electrodes for collection. The two main driving forces for transport
are drift and diffusion currents. Drift current causes the carriers to
move along the potential gradient inside the solar cell and it
determines by the choice of electrodes. Choosing a high work
function anode and low work function cathode creates a built-‐in
electric field, in which applying an external bias modifies the internal
electric field and the drift current. This current leads the carriers
toward the respective electrodes and from there they are collected.
Diffusion current is another mechanism, which transport the
carriers along the concentration gradient inside the solar cell.
36
Concentration of electrons and holes are higher around the
heterojunction due to generation of geminate pairs there; hence,
carriers diffuse away from the interface and cause diffusion current.
After this point, the ability of the material to transport the carriers
is critical. This ability is called mobility and characterizes the speed
of the carrier transport through the material. In general, electrons
have higher mobility54 than holes which means choosing donor and
acceptor materials with high mobility differences can lead to
accumulation of charges at the interface between active layer and
electrodes55. Therefore, balancing hole and electron mobilities is
critical to achieving efficient carrier transport in the organic solar
cells.
2.2.4 charge extraction at electrodes When the carriers reach at the active layer/electrode interface,
they can be extracted. In order to maximize the efficiency in charge
extraction, the potential barrier have to be minimized. Therefore
ideally the work function of the cathode should match the LUMO of
the acceptor and the work function of anode should match the HOMO
of the donor. This way an ohmic contact is formed and carrier
extraction condition is fulfilled. In organic solar cells, indium tin
oxide (ITO) with a work function of ~4.7 eV is commonly used as
anode. On the cathode side, a low work function material such as Al
(4.2 eV) can be used.
Another trick to align the work functions of the electrodes with
HOMO and LUMO of the organic materials is to use the interlayers
known as buffer layers at the active layer/electrode interface.
Molybdenum oxide (MoO3), Zinc oxide (ZnO) and Bathocuproine
37
(BCP) are a few examples of buffer layers, which have shown to
improve the carrier collection56. Apart from changing the electrode
materials and using buffer layers, increasing the roughness or
interface area of the active layer and electrodes can improve the
charge collection efficiency.
2.2.5 Summery of the operation Upon absorption of a photon in an organic solar cell, an exciton is
formed. This exciton travels toward the donor-‐acceptor interface,
and if it can reach the interface before decaying, the exciton can
dissociate, with the electron transferred to the adjacent acceptor
molecule and hole remained at the donor molecule. This state is
called a charge transfer (CT) state in which the charge carriers can
either dissociate or recombine back to the ground state. If they split
into free charge carriers, they can travel toward the electrodes and
be collected, or if they meet at an interface, they can recombine with
Figure 2.5. The principle of charge separation in an organic solar cell operation. Summary of the 4 main steps.
38
another charge carrier and loose their energy. The summery of the
operation is shown in Figure 2.5.
2.3 Solar cell architectures Solar cell architectures can significantly influence the
performance of the devices. Here, two main types with two common
architectures are explained.
2.3.1 Planar vs. bulk heterojunction Planar heterojunction or bilayer cell is the simplest interface
structure in which a layer of donor and a layer of acceptor on top of
each other (joined together) form the active layer. This layer is
sandwiched between charge collection layers (buffer layers) and
electrodes (anode and cathode), Figure 2.6.a. This simple structure is
based on the basic operating principles of the solar cell. However,
due to the low diffusion length of the excitons in organic material
(around 10 nm), there is a thickness limitation for the layers.
Figure 2.6. a) Planar heterojunction or bilayer solar cell, b)Bulk heterojunction solar cell with a donor/acceptor blended active layer for more efficient charge generation.
Furthermore, a thin active layer results in a weak absorption, and
a delicate thickness balance is thus present in these cells. Introducing
39
bulk heterojunctions (BHJ) solar cells in the mid 1990s7 provided the
solution to this tradeoff, Figure 2.6.b. In this structure, mixing or co-‐
evaporation of donor and acceptor materials forms the junction. The
resulting film is a network of donor/acceptor domains, and since the
scales of the domains are within the diffusion length of the excitons,
it provides a path for efficient carrier transport and dissociation.
Therefore, nearly all generated excitons are dissociated and free
charge carriers are transported towards electrodes and are collected.
This advantage makes it possible to form a thicker active layer
compared to bilayer solar cells. Nevertheless, controlling the film
morphology is harder in the films formed by spin coating compared
with vacuum deposition (more information about the techniques are
provided in section 2.5). This means that there are many parameters
that can affect the performance of the BHJ devices. There exist
various methods to improve the morphology of the BHJ solar cells,
such as thermal annealing, solvent annealing or modifying the
functionality of the organic materials57–59. These techniques have
improved the performance of the OSC devices.
2.3.2 Standard vs. inverted structure
There are two different solar cell geometries: standard (or
conventional) and inverted (see Figure 2.7). If the anode is directly
placed on the substrate it is a standard, and if the cathode is on the
substrate it is an inverted configuration. There are advantages and
drawbacks of each geometry. Standard configuration usually results
in higher efficiencies; however, inverted configuration usually has
higher stability60–65, although clear exceptions are present in both
cases.
40
Figure 2.7. Two main geometries of PHJ organic solar cells: a) standard and b) inverted configuration.
2.4 Materials 2.4.1 Donor and Acceptor Oligomers (small-‐molecules) and conjugated polymers are two
main categories of organic materials being commonly used in active
layer of the organic solar cells. Small molecules have low molecular
weights; while polymers are heavy molecules consist of long
molecular chains. Small molecules are usually deposited via ultra-‐
high vacuum thermal deposition whereas polymers are deposited
using solution-‐processed methods, such as spin-‐coating, or printing
techniques such as ink-‐jet printing and doctor blade.
Small molecules have several advantages compared to polymer
materials, which makes them attractive for their application in solar
cells. Some of the main advantages are:
• Ease of synthesis and reproducibility of the process.
• High purity of the material
• Better control of the morphology and structure of the film by
controlling the growth parameters, such as substrate
temperature, deposition pressure and rate.
41
• No need for solvents (some toxic)
• Fabrication of multilayer structures, such as tandem solar
cells is possible by controlling the thickness of the layer in
nanometer scale.
In the following, the chemical structure and key optoelectronic
properties of the materials used in this thesis are presented.
Tetraphenyldibenzoperiflanthene (DBP)
DBP is a p-‐type (electron donor) semiconductor with a
symmetrical molecular structure. This molecule is only composed of
carbon and hydrogen atoms (see Figure 2.8). Among other donor
molecules, DBP is a promising electron donor material, which has
been utilized in many laboratories since 200966–72.
High optical absorption and a deep HOMO level around 5.5 eV are
principle advantages of DBP in application of solar cells66,73. These
main advantages allow fabrication of a thinner active layer, which
facilitates reaching the exciton to the D/A interface. Moreover, its
HOMO level energy of around -‐5.5 eV is energetically compatible
with fullerene acceptors to be used in the active layer. Energy
difference between HOMO of the DBP and LUMO of the fullerene
results in a high open circuit voltage (VOC)4,66,71,74–78.
Figure 2.8. Skeletal formula of tetraphenyldibenzoperiflanthene (DBP).
42
Fullerene C70
Fullerene molecule is a hollow molecule cage consisting of sixty or
more carbon atoms. It can be in the shape of a sphere, ellipsoid, tube
or other shapes. Buckminsterfullerene or buckyball was the first
known example of the spherical shaped fullerenes. C70 is a fullerene
molecule with 70 carbon atoms, and resembles the shape of a rugby
ball. In 1985, Robert Curl, Harold Kroto and Richard Smalley
discovered Fullerenes (with the most common C60 and C70) (see
Figure 2.9), and they were awarded the 1996 Nobel Prize in
chemistry for their discovery79. Since then C70 has been used as
electron acceptor and electron transporting molecule. HOMO and
LUMO of C70 are -‐6.1 eV and -‐4.0 eV respectively80.
Figure 2.9. Schematic drawing (a) soccer-‐ball structure of buckyball C60 and (b) rugby-‐ball structure of common fullerene C70. (c) C70 Powder
2.4.2 buffer layers Molybdenum trioxide MoO3
Molybdenum oxide is known for its catalytic activity and semi-‐
conductive properties. MoO3 is an n-‐type material being used as hole
transport layer in organic solar cells81–84 (Figure 2.10).
Compared to other metal oxides such as WO3 and V2O5,
molybdenum oxide can be evaporated at very low temperatures
(~ 400 ℃) in vacuum from a crucible. Using MoO3 between the ITO
43
and donor layer improves the fill factor of the devices by reducing
the series resistance. This in general improves the power conversion
efficiencies of the solar cells83. Moreover, it has been shown that
using MoO3 instead of PEDOT:PSS in OPV devices improves their
stability85.
Figure 2.10. (left) Crystal structure of molybdenum trioxide (Hyung-‐Seok Kim/Nature Materials), (right) MoO3 powder
Bathocuproine BCP
Bathocuproine, 2,9-‐dimethy-‐4, 7-‐diphenyl -‐1,10-‐phenathroline or
BCP is a well-‐known electron transport layer being used in organic
solar cells86–88. BCP is a crystalline white or yellow powder, and is
insoluble in water (Figure 2.11). This molecule is a wide band gap
material, and importing an 8-‐10 nm thick layer BCP between
acceptor layer and top metal contact such as aluminum improves
electron transport properties by reducing geminate recombination
of the exciton at the acceptor-‐metal interface. Furthermore, it
protects the active layer from damages caused by the metal
deposition87,88.
44
Figure 2.11. Left) skeletal formula of Bathocuproine, 2,9-dimethy-4, 7-diphenyl -1,10-phenathroline (BCP), right) BCP powder.
2.4.3 Contacts
Active layer and buffer layers are sandwiched between indium tin
oxide (ITO) as anode and top Ag or Al as cathode. ITO is a
transparent conducting oxide with electrical conductivity and optical
transparency in the visible regime.
2.5 Fabrication techniques There are two common ways to deposit layers of a solar cell: spin
coating and vacuum deposition. In spin-‐coating processes, a small
amount of material in form of a solution is applied on the center of a
substrate being located on a holder of a spin-‐coater machine. Then
the substrate is rotated at a high speed and the coating material is
Figure 2.12. Schematic drawing of spin coating procedure. Desired solution is applied over a substrate located on a spinner, then the spinner is rotated and the material is spread over the surface by centrifugal force to form a film.
45
spread by centrifugal force (see Figure 2.12). Depending on the
viscosity and concentration of the solution, as well as the speed of
the spinner, a desired thickness of the film is achieved.
Vacuum thermal evaporation is another deposition technique, in
which an organic material is heated up in high vacuum. In this
technique, the substrate is place upside down in a distance above the
Figure 2.13. Thermal evaporation chamber. Substrate is located upside down above the source. Material is thermally evaporated and is deposited on the surface of the rotating substrate. A Crystal monitors the deposition rate.
source, where the evaporated material is directly deposited onto the
substrate, as shown in Figure 2.13. This allows for depositing many
layers of different materials without chemical reactions between the
layers. A crystal monitors the deposition rate to reach the desired
thickness. However, for large-‐area substrates, there may be
problems with film thickness and uniformity. In this work, we used
46
vacuum thermal evaporation technique to fabricate bulk and bilayer
DBP-‐C70 based organic solar cells.
2.6 Stability of organic solar cells One of the main drawbacks of organic solar cells is their low
stability and short lifetime due to device degradation. This means
that over time, they loose their photovoltaic ability to generate
electricity, which results in a decay in their PCE. Depending on the
importance of air exposure, it can be divided into two categories;
intrinsic and extrinsic degradation.
2.6.1 Intrinsic degradation Intrinsic degradation is caused by thermal intrusion of constituent
materials inside the solar cell. This incudes chemical degradation as
well as molecular segregation or rearrangement at the interfaces and
active materials13. This type of degradation can happen either in dark
conditions or under illumination (photo-‐induced)11,89,90.
Over a short period of time at dark conditions, molecular
segregation and rearrangement at material interfaces hinder charge
extraction. For longer periods, degradation causes phase separation
at BHJ solar cells and results in less efficient charge generation11,91–98.
Under illumination, photo-‐induced degradation causes JSC or VOC
losses especially at solution processed active layers due to increasing
the energy disorder of the polymer materials11. In fullerene
acceptors especially C60, photo-‐dimerization causes JSC loss99.
2.6.2 Extrinsic degradation Extrinsic degradation is penetration of air (oxygen and water)
into the active layer and interlayers, and causing chemical reaction.
47
This results in the loss of photocurrent from the active layer and
dead zones are grown in the device100,101. Reaction of oxygen with
the electrodes can increase the work function of metals by forming
surface dipoles and deteriorate the performance of the devices100,102.
Under light condition, photooxidative loss of absorption (bleaching)
of the semiconductor materials can cause degradation97,103,104.
There are solutions to prevent intrinsic and extrinsic
degradations11,13, such as encapsulation of the devices with low
permittivity materials105, or using pure, dense materials with highly
ordered film morphology to lower the oxygen and water diffusion
into the layers106. Using non-‐fullerene acceptor materials will also
reduce photo-‐induced JSC and VOC losses, which normally results from
photo-‐dimerization of fullerene acceptors107. Moreover, using high
work function metals is desirable to reduce extrinsic degradation102.
2.7 Perovskite solar cells Perovskite solar cells are made out of the perovskite mineral that
was named after Lev Perovskite who was the founder of the Russian
Geographical Society.
The perovskite structure can be simplified as a hybrid molecule
with organic atom A (positively-‐charged cation in the center of the
cube), inorganic atoms B (usually a metal cation) located on the
corners of the cube, and smaller halide atoms X (anion with negative
charge) occupying the faces of the cube108. In the world of solar cells,
perovskites and the perovskite structure are used interchangeably.
Any compound that has the same crystallographic structure as the
perovskite mineral is also called perovskite, which means the generic
48
form ABX3. For example, Perovskite mineral composed of Calcium,
Titanium and Oxygen forms the CaTiO3.
Depending on the atoms/molecules that are used in the structure,
perovskite can show interesting properties that make them very
exciting for physicists, chemists and material scientists. Their
intrinsic properties such as broad absorption spectrum, fast charge
separation and long carrier lifetime and long electron and hole
transport distance (over one micron109,110) have made them
promising materials for solid-‐state solar cells. The perovskite
crystalline lattice arrangement is demonstrated in Figure 2.14.
Figure 2.14. Generic perovskite crystalline lattice arrangement of the form ABX3. Note that the lines represent crystal orientation and not the bonding patterns. The two structure are equivalent with left) atom B at the <0,0,0> position and right) atom A at the <0,0,0> position.
Low costs of both raw materials and fabrication methods, high
efficiency, light weight and flexible perovskite solar cell have made
them commercially attractive. Perovskite conversion efficiency has
reached to about 22.7% in late 2017 (listed in an efficiency chart
provided by NREL). However, one of the main challenges that hold
them back from the market is their low stability and lifetime.
49
2.7.1 Fabrication Perovskite solar cells can be easily incorporated into standard or
thin film architecture. Vacuum deposition can be used to co-‐
evaporate the organic (methylammonium) and inorganic (lead
halide) components to give a uniform and high quality film.
However, co-‐evaporation of these materials requires special
chambers that may not be available in many labs. Moreover, this
may cause cross-‐contamination between organic and non-‐organic
sources in the deposition chambers, which is difficult to clean.
Development of solution deposition processes has provided
simpler methods to fabricate Perovskite solar cells28. A lead halide
and methylammonium halide can be dissolved in a solvent and spin
coated in one-‐step processing onto a substrate. However, other steps
and additives are required in order to achieve a homogenous and
high quality film.
For the interlayers, many of the standard electron and hole
transport layers from the world of organic solar cells such as
PEDOT:PSS, PCBM, ZnO and TiO2 can be used in perovskite solar
cells.
2.7.2 Stability
As mentioned before, one main challenge in the field of perovskite
solar cells is their poor stability and short lifetime. This is mainly
related to influence of the different factors such as oxygen and
moisture18,19, thermal intrinsic stability20, heating under applied
voltage21, UV photodegradation22, and mechanical fragility23. Water
solubility of the organic constituent of the perovskite absorber
material make these devices highly sensitive to moisture in the air
50
and causes rapid degradation24. This sensitivity can be reduced by
reducing the constitute material or by changing the cell’s
architecture. Moreover, encapsulation of the devices prevents the
immediate degradation of these devices under moist conditions;
however, long-‐term studies have yet to be performed24,25. It has been
shown that the UV Photodegradation of the perovskites with TiO2 is
linked to the interaction of the photogenerated holes inside the TiO2
with the oxygen radicals on the surface of the TiO2 111.
These two instability problems can be solved by using
multifunctional coatings to provide strong hydrophobic barrier and
block the UV light of the incident solar spectrum and converting it
into visible light112. In addition to these solutions, Insertion of a
mechanically reinforcing scaffold into the active layer of the
perovskite solar cells has provided a higher fracture resistance by
30-‐fold113. Nevertheless, despite all these efforts, long-‐term stability
of the perovskite solar cells needs to be further enhanced for
successful commercialization of this appealing technology in the
market.
2.8 Characterization
2.8.1 J-‐V characteristics Organic solar cells are typically characterized under 1000W/m2
light (1.5 AM solar spectrum). In the dark, the solar cell acts as a
simple diode. An ideal solar cell can be modeled by an equivalent
circuit in which the current source is in parallel with a diode.
However, since no solar cell is ideal, a series resistance (Rs) and a
shunt resistance (Rsh) component are added to the circuit. This
51
model is useful to understand the electronic behavior of a solar cell
(see Figure 2.15).
Figure 2. 15. Electric circuit equivalent to an OPV device.
Series resistance is resulted from all the resistances at the
interfaces, conductivity of the semiconductors and the electrodes.
Shunt resistance is related to the defects in the film and takes into
account the leakage of the current through these defects. Ideally a
low Rs and a high Rsh is desired. In this circuit, n is the ideality factor
of the diode, ID is the saturation current, which is the current in the
dark at the reverse bias, and Iph corresponds to photocurrent
generated during illumination.
Figure 2.16 shows the current-‐voltage (I-‐V curve) characteristic for a
typical solar cell. Without illumination, the solar cell electrical
characteristics are similar to a diode. After illumination, the IV curve
is shifted down to the 4th quadrant as the cell begins to generate
power. The greater the light intensity, the greater the generated
current and the amount of shift.
52
Figure 2. 16. Current-‐voltage (I-‐V) curve of an organic solar cell. a) Without illumination, the solar cell behaves like a diode, b) after illumination, the curve is shifted downward as the cell start to generate power, c) increasing the light intensity results in further shift of the curve downward as the generated power increases.
The operation of the solar cell under different applied biases can
be summarized into 4 steps (see Figure 2.17):
1. Large reverse bias: when a reverse bias is applied, it
reinforces the built-‐in electric field, and due to presence of a
strong electric field drift current is dominant. This enhances
the exciton dissociation and results in an efficient charge
transport and a large photocurrent.
2. Small reverse bias: when the applied bias is close to
zero, the built-‐in field is the only force exists in the device,
which drives the carriers to the corresponding electrodes for
collection. When the applied bias is increased in positive
direction, it opposes the built in field. Therefore, drift current
decreases and results in decrease of the current.
53
3. Positive bias and zero internal field: at the point where
the applied bias is equal to the built in field, the electric field
inside the device becomes very small and diffusion current
dominates the current. This condition is called open-‐circuit
condition and corresponds to the maximum voltage out of a
solar cell.
4. Positive bias: by further increasing the bias, the
applied bias becomes larger than the built-‐in field that
reverses the potential gradient in the device. Therefore,
carrier injection occurs through the tunneling and results in a
positive current.
Figure 2.17. Operation of a solar cell under different applied biases; 1) large reverse bias; 2) small reverse bias; 3) positive bias, results in zero initial field, and corresponds to open-‐circuit condition; 4) positive bias and carrier injection
It is more common to use current density (J) instead of current (I)
on the y-‐axis. Therefore, the key parameters of the solar cell that
define the performance of the cell can be extracted from its J-‐V curve.
These parameters are open circuit voltage (VOC), short circuit current
(JSC), fill factor (FF), the maximum power conversion efficiency
(Pmax), the voltage and the current at max power point (Vmax and
Imax). Figure 2.18 shows solar cell J-‐V curve metrics. Each of these
parameters is described here:
54
Figure 2.18. Solar cell J-‐V curve metrics.
Open circuit voltage (VOC)
Is defined as the voltage at which the current density output is
zero. The value can be extracted from the crossing point of the J-‐V
curve with V axis at 0 J.
Short circuit current (JSC)
Is defined as the current at which the externally applied voltage is
zero. The value can be extracted from the crossing point of the J-‐V
curve at 0 volt. The value presents the number of the generated
charge carriers that were collected at the electrodes at short circuit
condition. JSC value can be improved by using small bandgap
materials with high absorption coefficient and high carrier mobility.
Moreover, as discussed before, a bulk heterojunction with more
mixed phases (smaller domains) can improve the charge generation
efficiency.
Fill factor (FF)
The shape of the J-‐V curve is defined by the fill factor parameter. It
is defined as the ratio between the maximum obtainable output
55
power (product of Jmpp and Vmpp at maximum power point) and
product of the open-‐circuit voltage and short-‐circuit current.
FF =J!"" V!""J!" V!"
(2.2)
A low electron mobility causes increase in the recombination rate
before they reach the interface therefore a higher external bias is
needed to sweep the carriers to the heterojunction for dissociation.
This results in a strong dependency of the current on the applied
bias and leads to a lower FF.
Power conversion efficiency (PCE)
Efficiency of a solar cell is also known as the solar power
conversion efficiency, that is the ratio of produced electrical power
to the incident irradiation power.
𝑃𝐶𝐸 =𝑉!" 𝐽!"𝐹𝐹
𝑃!" (2.3)
Where Pin is the input power density.
External Quantum Efficiency (EQE)
The percentage of the photons that are converted to charge
carriers and collected at the electrodes is called the external
quantum efficiency (EQE). This parameter is defined as the product
of the efficiency of the four steps; absorption, diffusion, dissociation
and collection,
𝜂!"! = 𝜂! . 𝜂!"## . 𝜂!"## . 𝜂! (2.4)
A narrow absorption band or a thin active layer can lower the
absorption efficiency. Moreover, a high recombination rate and
photo generated loss due to quenching at metal electrodes or high
phase separation in the active layer can lower EQE.
56
2.8.2 ISOS degradation tests
In order to track the behavior of the organic and hybrid devices
over time and to understand the degradation mechanisms, solar cell
devices are tested under various stability test conditions known as
accelerated lifetime measurements. To increase reproducibility of
the conditions across different labs, International Summit on OPV
Stability (ISOS) protocol have been established. There are different
ISOS test conditions, such as dark, illumination, indoor, outdoor,
laboratory weathering, thermal and humidity cycling. These
conditions are simulated in systems known as climate chambers.
After putting the devices in the desired environment, the
performance of the devices is monitored over time and the result is a
plot that shows the efficiency versus time, known as decay curve.
There are three main regimes of degradation of solar cells known as
burn-‐in, long-‐term and failure. Analyzing the behavior of the solar
cells in these three zones can give much information about the
overall stability of the devices114–117.
2.9 Relation between VOC and CT states One of the prominent parameters that limits the efficiency of the
solar cells is open circuit voltage (VOC), that is the maximum possible
voltage out of a solar cell. Numerous studies have shown that VOC can
be predicted from the energy gap between the HOMO of the donor
and LUMO of the acceptor material in a BHJ. Therefore, VOC depends
on the choice of the materials; which means choosing light
harvesting materials with higher energy bandgaps can be beneficial.
However, analyzing VOC across a wide range of materials have shown
57
that although the optical bandgap of the light harvesting materials is
around 1.7 to 2.1 eV, VOC barely exceeds half of the incident photon
energy meaning around 1.0 V118. This difference between the
material bandgap and VOC has experimentally shown to be
approximately 0.6 eV, while it is only 0.3 to 0.45 eV for inorganic
materials such as Si, CIGS, and GaAs119. To date, there have been
many efforts to address the origins of the VOC losses and finding the
solutions to prevent these, however, it remains still an open
question.
There are many factors that can influence the open circuit voltage
in organic solar cells such as energetic disorders or density of states,
charge transfer states, donor-‐acceptor interface, carrier density and
interface morphology. Apart form these, various other factors such
as temperature, light intensity, recombination, etc. can influence the
VOC indirectly120. Among all, we are interested in charge transfer
states, which are relating the properties of the D-‐A interface to the
VOC value. Several studies have demonstrated that there is a strong
relation between VOC and CT state emission (ECT) being correlated
according to:
ECT−qVOC=0.6±0.1 eV (2.5)
Vandewal et al. obtained an analytical expression for VOC by
separating the losses from ECT into radiative and non-‐radiative
recombination:
𝑉!" =𝐸!"𝑞 − Δ𝑉!"# (𝑇)− Δ𝑉!"!#$% (𝑇) (2.6)
The two temperature dependent radiative and non-‐radiative
recombination losses are given by the following equations:
58
Δ𝑉!"# 𝑇 = − 𝑘𝑇𝑞 ln
𝐽!" ℎ!𝑐!
𝑓𝑞2𝜋 𝐸!" − 𝜆 (2.7)
Δ𝑉!"!#$% 𝑇 = −𝑘𝑇𝑞 ln 𝐸𝑄𝐸!" (2.8)
Here J!" represents the short circuit current, and EQE!" is the
external quantum efficiency of electroluminescence. Equation 2.6
shows a linier correlation between the photovoltage with the energy
of the CT states. A detailed deduction can be found in the paper by
Vandewal et al.121.
Since the energy of the CT state is one of the lowest energy states
in the SC, a highly sensitive absorption technique such as sEQE is
used, in which the CT state is determined at the shoulder of the
spectrum at very low energies38. The working principle of the sEQE
measurements is similar to the standard EQE measurements, and
will be explained in Chapter 3.
There are two methods both based on the principles of the Marcus
theory (see section 2.2.2.2) to extract the CT properties.
Based on this theory49,122, a formalism has been raised by
Vanderwal et al.121 to characterize the properties of the CTS, and has
been widely used to investigate the energy of the CT states123–125.
𝐸𝑄𝐸(𝐸) =𝑓!!
𝐸 4𝜋𝜆𝑘𝑇𝑒𝑥𝑝
−(𝐸!" + 𝜆 − 𝐸)!
4𝜆𝑘𝑇 (2.9)
Where k is Boltzmann’s constant and T is the absolute
temperature. ECT is the free energy difference between the CT
complex ground state and the CT excited state, 𝜆 is reorganization
energy, f is a parameter proportional to the density of CT states and
represents the strength of the donor/acceptor material interaction, h
is Planck’s constant, and c is the speed of light in vacuum. Through a
59
Gaussian fit over the shoulder of the sEQE curve at sub-‐bandgap
region, ECT, 𝜆, and f can be extracted (fitting parameter).
Another method is to measure both CT absorption via sEQE, and
CT emission via electroluminescence (EL). The crossing point of
these two spectra equals ECT, and 𝜆 and f can be extracted through a
fit over the sub-‐bandgap region and using Equations below. (see
Figure 2.19)126.
𝐸𝑄𝐸!" 𝐸 .𝐸 =𝑓!"!4𝜋𝜆𝑘𝑇
exp−(𝐸!" + 𝜆 − 𝐸)!
4𝜆𝑘𝑇 (2.10)
𝐸𝐿(𝐸)𝐸 =
𝑓!"4𝜋𝜆𝑘𝑇
exp−(𝐸!" − 𝜆 − 𝐸)!
4𝜆𝑘𝑇 (2.11)
The left hand-‐side of the above equations show a mirror image
relationship, and due to the multiplication and division of the spectra
by E, they are called reduced absorption and emission spectrum,
respectively.
Figure 2.19. a) free energy diagram for the ground state and lowest excited state. b) Reduced EQE PV and EL spectrum with fits using formulas (10) and (11). Each parameter is indicated in the figure.
60
61
CHAPTER 3
Experimental Setup
Organic solar cells used in this work are fabricated over pre-‐
patterned indium-‐tin-‐oxide (ITO) (≈ 100 nm) coated glass wafers
with a sheet resistance of 9-‐15 Ω/sq (purchased from University
wafer, Inc, USA). Patterning process is carried out in a class ISO7
cleanroom, and then prepared substrates are transferred to a
laboratory equipped with fabrication and measurement systems to
proceed with other steps. All organic active materials, metal oxides
and metal layers are formed using a vacuum deposition system.
In this chapter, each of these device fabrication steps and the
measurement setups are explained in detail.
3.1 Device fabrication
3.1.1 patterning of the substrates Pattering process is done using a positive lithography process,
through which a thin layer of adhesion promoter,
Bis(trimethylsilyl)amine (also known as hexamethyldisilazane, or
62
HMDS) is applied in 120°C, and a photoresist (AZ5214E,
Microchemicals GmbH, Germany) is spin-‐coated and baked at 90°C
for 1 min on a hot plate. Patterns are transferred using a UV Karl
Suss MA 150 Aligner. Then an ITO etching process is performed for 4
min in HCL:HNO3:H2O (1:0.08:1) solution at 40°C (etching rate≈ 0.5-‐
1 nm) to remove excess areas. Then the photoresist is stripped off in
Acetone placed in ultra-‐sonic bath for 10 min to reveal the final
patterns.
3.1.2 Pre-‐cleaning the substrates Patterned wafers are diced into 15 by 15 mm2 substrates, and are
cleaned using detergent, acetone and isopropanol in an ultra-‐sonic
water bath (10 min each), and blow-‐dried with a nitrogen (N2) gun.
Figure 3.1 shows the layout of the pattered ITO samples after being
diced into smaller substrates.
Figure 3.1. Layout of the patterned ITO coated glass substrates.
Prior to fabrication process, substrates are treated with air
plasma cleaning at 400 mTorr for 20 min. Then right after the
treatment, they are transferred to a glovebox connected to a high-‐
vacuum deposition system with a base pressure of about 10-‐8 mbar.
63
3.1.3 Vacuum deposition of organic/recombination layers and top electrode The whole fabrication process is done in a Cryofox cluster system
in UHV environment, and via thermal evaporation process (Figure
3.2). Cluster system is connected to a Nitrogen Glovebox to avoid air
exposure during the whole fabrication process. In order to avoid
material contamination in different layers, deposition of organic
materials, and metals are performed in separate chambers having
vacuum pressure of about 5 x 10-‐8 mbar. These chambers are
connected with a transfer chamber with a base pressure of around
10-‐9, where a robotic arm transfers the sample between the
chambers. This allows a sequential deposition in high vacuum
conditions without breaking the vacuum between the steps. Vacuum
is broken only when there is a need to use a different mask, during
which the samples are only exposed to the Nitrogen environment of
the glovebox. The whole process is controlled through a control unit
where sample recipes are loaded for deposition of each layer
separately.
To form the active layer, tetraphenyldibenzoperiflanthene (DBP)
(purchased from Luminescence Technology Corp., Taiwan) and
fullerene (C70) (purchased from Sigma-‐Aldrich, Germany) are used as
electron donor and acceptor, respectively. For exciton blocking layer
Bathocuproine (BCP), and for electron transport layer molybdenum
oxide (MoO3) (both purchased from Sigma-‐Aldrich, Germany) are
deposited.
64
Figure 3.2. Cryofox deposition cluster system. Organic materials and metals are deposited in two separated chambers in ultra high vacuum conditions. A robotic arm transfers the sample between the chambers without breaking the vacuum between the steps. Right. The system is connected to a glove box to avoid exposure of the samples to air127.
Figure 3.3 shows the organic sources for DBP, C70 and BCP inside
the organic chamber. Each crucible is covered with a shutter when it
is not in use, and a quartz crystal microbalances (QCM) are
controlling the deposition rate for each source with tenth of an
angstrom precision. Materials are thermally evaporated and are
deposited over the surface of rotating samples that are placed upside
down in a distance above the sources. This way thickness uniformity
of 10% across a 10cm sample holder is achieved. Moreover, if
needed, there is a possibility to heat up the substrates up to 250℃.
This allows annealing the samples in vacuum during evaporation or
between depositions of each layer.
65
Figure 3.3. Organic sources located inside the ultra high vacuum chamber. Each source is placed inside crucible, and a shutter to cover the source while not in use.
As the final step, 100nm silver (Ag) (purchased from AESpump
ApS, Denmark) is deposited as top electrode to finish the fabrication.
Deposition of MoO3 and Ag is performed in a separate chamber.
Deposition rate is kept at 0.3 Å/s for organic materials and 0.5-‐1 Å/s
for Ag top electrode using quartz crystals. In order to avoid a short
between the bottom ITO contact with the top Ag electrode, a shadow
mask with a wider opening is used to form the active layer, while a
mask with smaller opening is used during the metallization. Overlap
of Ag contact area with its underneath organic layers defines a cell
area of around 10 mm2. These fabrication steps takes around 3-‐4
hours. Figure 3.4 shows the final look of the fabricated sample. The
final sample consists of 4 cells with separated active areas sharing a
common ITO bottom electrode.
66
Figure 3.4. Final outlook of the DBP-‐C70 based organic solar cells. Each sample consists of 4 cells. Overlap of the top electrode with the bottom ITO electrode defines the cell area, which here is 10 mm2
3.1.4 Final structure In this work we fabricate organic solar cells based on the same
active and buffer materials but with two different standard and
inverted configurations, Figure 3.5.a and 3.5.b, respectively. The
order and thickness of each layer is as follow:
• Standard configuration:
ITO/MoO3 (10 nm)/DBP (20 nm)/C70 (30 nm)/BCP (10 nm)/
Ag (100 nm)
• Inverted configuration:
ITO/BCP (0.5 nm)/C70 (30 nm)/DBP (20 nm)/MoO3 (10
nm)/Ag (100 nm).
Note that the thickness of the DBP (donor) and C70 (acceptor) are
kept the same in both configurations. Energy diagram for each
configuration is illustrated in Figure 3.5.c and 3.5.d.
67
Figure 3.5. Final illustration of DBP-‐C70 based organic solar cells with a) standard and b) inverted configurations. C and d show energy band diagram for each structure.
3.2 Characterization
3.2.1 J-‐V measurements Fabricated devices are mounted in a homemade sample holder
(Figure 3.6), and are characterized in ambient condition using a
3000 class AAA solar simulator from Abet Technology Inc., USA with
a calibrated arc lamp. The system is connected to a Keithley 2400
source measure unit (Keithley instrument Inc., USA), and LAbView
software is controlling the measurements. The JV curve is recorded
applying a voltage sweep between +1 to −0.25 V under illumination
intensity of 100 mW/cm2.
68
Figure 3. 6. Sample holder with 4 sample positions
3.2.2 Sensitive external quantum efficiency (sEQE) measurements The working principle of the sEQE measurements is shown in
Figure 3.7, and is similar to the standard EQE measurements.
In this technique, the light from a quartz Halogen lamb (50 W, TS
Electric) is chopped at 140 Hz and is passed through a
monochromator (Cornerstone 260 1/4m, Newport). Then the
monochromatic light is focused onto a solar cell mounted in a holder,
and the produced current at the short circuit condition is sent to a
current amplifier (DHPCA-‐100, FEMTO). This current is then
analyzed by a LOCK-‐IN-‐amplifier (7280 DSP, Signal Recovery, Oak
Ridge, USA) to extract the EQE properties. EQE is defined as the ratio
of the photocurrent to flux of incoming photons, which is obtained
using a calibrated silicon (FDS100-‐CAL, Thorlabs) and/or indium-‐
gallium-‐arsenide (InGaAs) photodiode (FGA21-‐CAL, Thorlabs) 119. In
order to achieve a span spectrum over 5-‐7 decades, the time
constant of the amplifier was set to 1s, and its amplification was
69
increased to resolve low photocurrent. The key points to achieve a
higher sensitivity compared to the standard EQE are to use a halogen
lamp which has a broad and contentious spectrum down to infrared
(IR), avoid optical fibers for illumination, and using high optical
density filter.
Figure 3.7. Sketch of the sensitive EQE measurement setup128.
3.2.3 Degradation protocols Stability measurements are performed in four ISOS aging test
conditions:
1. ISOS-‐D-‐1 (darkness, room temperature and humidity) in a
dark shelf
2. ISOS-‐D-‐3 (darkness, 85℃ and 85% RH-‐humidity) in a
climate chamber with controlled humidity and temperature
3. ISOS-‐T-‐3 (darkness, -‐40℃ and room humidity) in a thermal
chamber with ambient humidity and controlled temperature
4. ISOS-‐L-‐1 (illumination, 60℃ and ambient humidity) in a
setup consisting a solar simulator with monitored ambient
conditions
70
The setup used for ISOS-‐L-‐1 condition is consisting of a wide-‐area
metal halide lamp solar simulator (metal halide display/optic lamp
HMI from Orsam). The lamp is calibrated in ambient air with a Cell
and Meter (91150V, Newport). Using intensity and temperature
sensors, the ambient conditions are monitored and recorded
together with the solar cell parameters. For ISOS-‐D-‐1, D-‐3 and T-‐3, J-‐
V characteristics of the fresh samples are first measured using the
solar simulator under 1 sun illumination, and then devices are placed
in each condition for 24 hours in darkness, after which, J-‐V
characteristics are recorded again for the aged samples.
3.2.4 Morphological characterization Morphological characteristics are recorded using atomic force
microscopy (AFM). For our morphology investigations, a Dimension
3100 Nanoman scanning probe microscope from Veeco is used to
record the AFM images of the samples in air. Between the two modes
of AFM scanning, which are contact mode and tapping mode, the
second one is used for our experiments to avoid adhesions or
frictions between the tip of the cantilever and the surface of the
organic thin-‐films.
3.2.5 Photoluminescence quenching measurements Photoluminescence (PL) intensity measurements are performed
using a florescence microscope (Nikon Eclipse ME600), equipped
with a mercury short arc lamp having a filtered excitation
wavelength centered between 330nm-‐380nm. The system is
connected to a Maya2000Pro Spectrometer from Ocean optics to
71
record the spectra, and PL is collected with a microscope objective
(Nikon E Plan 50X 0.75 EPL) and after 10 sec integration time.
Measurements were repeated for aged samples.
We use the techniques described in this chapter to fabricate and
characterize fresh and degraded DBP-‐C70 based organic solar cells
with standard and inverted configurations. The results are presented
in the next chapter.
72
73
CHAPTER 4
Degradation pathways in Standard and Inverted DBP-‐C70 Based Organic Solar Cells
Organic solar cells have become an emerging competitive
technology due to their unique advantages such as low fabrication
cost, semi-‐transparency, flexible and lightweight products. However,
as it was discussed previously in this literature, achieving long-‐term
stability in organic solar cells is a remaining bottleneck for the
commercialization of this otherwise highly appealing technology. For
organic solar cells to be commercially viable, at least 10 years of
long-‐term stability is required129,130, which has presented a
significant challenge to this field. Therefore, a deeper understanding
of the degradation mechanism of the devices is needed, which will
pave the way for strategies on device lifetime improvement.
74
Degradation mechanisms in organic solar cells can be divided into
two main categories caused by either intrinsic or extrinsic factors.
The former is caused by diffusion of oxygen and moisture from the
air into device layers, and the latter is due to dynamic and active
nature of organic materials, and includes chemical degradation and
molecular rearrangement in the active materials and interfaces13.
Encapsulation of the organic solar cells minimizes the availability of
air to the different layers, and can significantly slow down the
extrinsic degradation. However, intrinsic degradation happens over
time even for the best-‐encapsulated devices91–96. Degradation
mechanisms in organic solar cells are rather complex and include
degradation of interfaces and interlayers, electrode diffusion, and
morphological changes. This implies the strong impact of the
interface between every two adjacent layers on the stability of the
devices60–65,131. In organic solar cell, the interface between the donor
and acceptor materials plays a critical rule in degradation of the
devices, where chemical and morphological changes over time
directly affect charge photogeneration and recombination pathways.
As introduced in chapter 2, residing at the D-‐A interface, charge
transfer (CT) state represents an intermediate state between charge
generation and recombination and its properties directly affect the
performance of the devices, especially the open-‐circuit voltage
(VOC)38,47,121,132. Moreover, degradation paths are rather intricate and
differ between standard and inverted devices60–63. As an example,
Krebs et al. and Cros et al. reported a higher air stability for inverted
configuration compared to standard configuration, while both
devices have the same type of photovoltaic layer and undergo an
identical processing procedure64,65.
75
In the work presented in this chapter, we study the degradation
pathways in DBP-C70 based organic solar cells, having standard and
inverted device architectures. We utilize sensitive external quantum
efficiency (sEQE) measurements to detect differences in
morphological properties between standard and inverted device
configurations, as well as detecting the potential degradation at the
DBP/C70 interface after degrading the devices at ISOS-‐D-‐3 and ISOS-‐
T-‐3 aging conditions. Our investigations reveal that despite the
variations in the VOC values of the fresh and degraded devices, CT
state properties undergo only very minor changes, suggesting a
pronounced morphological stability at the interface of DBP and C70.
Instead, it is shown that presence of Bathocuproine (BCP) electron
transport layer (ETL) in inverted devices leads to higher VOC losses and
low stability of these devices133. To address this issue, as an alternative
ETL, we introduce BCP/Ag stack with improved exciton blocking
properties and carrier transport efficiency. Our final results show
enhanced device stabilities for inverted devices implementing BCP/Ag
stack as their ETL.
4.1 Device Performance:
Standard vs. Inverted configuration The experiments were performed with DBP-‐C70 based organic
solar cells having standard and inverted structure. More information
about fabrication processes and characterization techniques can be
found in chapter 3.
The performances of the devices were recorded via J-‐V
measurements by applying a voltage sweep between +1 to −0.25 V
76
under 1 sun illumination. Figure 4.1 and Table 4.1 show the J-‐V
characteristics for the two types of devices with identical active
layers and contacts. Solar cell parameters are averages over 7
devices of each type, and J-‐V curves show the devices with
characteristics closest to the average parameters.
Figure 4.1. J-‐V curves of DBP-‐C70 organic based PHJ devices having standard or inverted configurations.
Table 4.1. Photovoltaic performances of DBP-‐C70 organic based PHJ devices having standard or inverted configurations. The values extracted over seven best performing devices of each type.
Devices VOC (V)
Jsc
(mA/cm2) PCE %
FF
St-PHJ 0.85 ± 0.05 6.32 ± 0.36 3.66 ± 0.21 68.0 ± 1.8 In-PHJ 0.72 ± 0.07 5.78 ± 0.40 2.68 ± 0.29 62.0 ± 2.5
Although both devices have identical active layers, standard
structure show an overall better performance with 130 mV higher
77
VOC and higher PCE. However, a recent work has demonstrated that
this trend may depend on device area134. The reported efficiencies are
similar to the ones previously reported for PHJ DBP/C70 cells in the
literature72,135.
Previous studies have shown that morphological differences and
changes in molecular orientations at the D-‐A interface can result in
VOC changes by altering the material energy levels136 and CT states
properties. Based upon Marcus formalism (see section 2.9) and
Gaussian fitting, we investigate the VOC differences through the sEQE
measurements and CT state characteristics. Using this method121,
ECT, lambda and f are determined by fitting the low energy part of the
EQE spectrum with a Gaussian.
4.2 Sensitive EQE measurements Using Eq. 2.9 and fitting to the sEQE spectra, we obtained ECT
values as 1.44 eV for standard vs. 1.37 eV for inverted configuration.
The normalized sEQE spectra and their corresponding fits are shown
in Figure 4.2, and extracted CT parameters (ECT, 𝜆, and f) are listed in
Table 4.2.
In standard configuration, the 70 meV higher ECT is in agreement
with its higher obtained VOC value; however, it does not fully explain
the differences in VOC upon inverting the structure. A slightly higher f
value obtained for standard structure identifies as higher density of
CT states. This is correlated to higher amount of interface between
the donor and acceptor materials, and represent the strength of the
interaction between the donor/acceptor molecules137. Previous
finding show that larger interface area results in more VOC losses due
78
to a higher recombination current138. However, the difference in f
value obtained here is minor, and should not contribute to significant
VOC differences.
Figure 4.2. sEQE measurements at 300 K and Marcus fits for standard and inverted structures. Dashed lines are fits to the EQE using Marcus theory.
Table 4. 2. VOC and CT characteristics extracted from fit parameters.
4.3 Morphology investigation at the D-‐A interface In order to further study the morphological differences at the D-‐A
interface of the devices, AFM imaging was performed. For this
purpose, samples were prepared with following structures:
Devices VOC (V) ECT (eV) f (eV2) λ (eV) St-‐PHJ 0.85 ± 0.007 1.41 0.0015 0.17 In-‐PHJ 0.72 ± 0.07 1.37 0.0008 0.22
79
a) ITO/MoO3 (10nm)/DBP (20nm),
b) ITO/MoO3/DBP (20nm)/C70 (30nm),
c) ITO/BCP (0.5nm)/C70 (30nm),
d) ITO/BCP (0.5nm)/C70 (30nm)/DBP (20nm).
Figure 4.3. AFM images of interface layer (a) DBP on MoO3/ITO, b) C70 on DBP/MoO3/ITO, c)C70 on BCP/ITO, and d) DBP on C70/BCP/ITO
Samples a & b represent layers in standard configuration, and c &
d represent layers in inverted configuration. For each sample
vacuum deposition was performed without breaking the vacuum
between the layers, and AFM images were recorded in air right after
fabrication, Figure 4.3. The root-‐mean-‐square (RMS) and surface
area values were calculated by software techniques applied to image
analysis and are shown in the figure.
Figure 4.3.a and 4.3.c show the surface of the lower layer of the D-‐
A interface in each type of device, respectively. The DBP layer in
standard configuration (Fig. 4.3.a) has a slightly smoother surface
with lower surface area, although very close to the one for inverted
configuration. Non-‐conformal film coverage for DBP on the C70 in
inverted device configuration could lead to the slightly lower f value
80
seen for this type of devices. In general, variations in interfacial
molecular orientations and morphological differences alter the
interaction of the molecules by varying electronic coupling between the
two molecules, and influences the charge generation efficiency, CT
properties and VOC loss paths136,139–142. Therefore, we note that both
modified molecular orientations and interface area can result in slightly
different f values seen here. This can be further investigated using
grazing incident wide-‐angle X-‐ray scattering (GIWAXS), and high-‐
resolution transmission electron microscope (HR-‐TEM)140; however,
difficulties in controlling the molecular orientations and a thin active
layer make this investigations impossible or challenging.
4.4 Degradation studies So far our findings show that morphological changes affect the CT
properties, which can be detected using sEQE measurements.
In this section, we benefit from this finding to detect degradation
mechanisms caused by morphological changes or molecular
rearrangement at the D-‐A interface. In order to track dark
degradation, standard and inverted devices were aged under ISOS-‐D-‐
3 (85 ℃, 85% RH-‐darkness) and ISOS-‐T-‐3 (-‐40℃ and room humidity-‐
darkness) degradation protocols for 24 hours without encapsulation.
Figure 4.4 shows J-‐V characteristics measured for both fresh and
aged devices under 1 Sun illumination using a solar simulator.
Average values over 7 devices of each set are presented in Table 4.3.
81
Figure 4.4. J-‐V curves for fresh (solid lines) and aged (dashed lines) devices at a) ISOS-‐D-‐3 b) and ISOS-‐T-‐3 degradation conditions.
Table 4.3. Photovoltaic performance parameters of fresh and aged devices under ISOS-‐D-‐3 and ISOS-‐T-‐3 degradation conditions. Device parameters for standard configuration are marked in blue, and for inverted devices in red.
Devices VOC (V) Jsc (mA/cm2) PCE % FF
St-Fresh 0.85 ± 0.05 5.8 ± 0.26 3.50 ± 0.17 71.2 ± 1.1
In-Fresh 0.72 ± 0.07 5.88 ± 0.57 2.68 ± 0.29 62.0 ± 2.5
St-ISOS-D-3 0.36 ± 0.28 4.90 ± 1.44 0.68 ± 0.73 23.3 ± 16.7
In-ISOS-D-3 0.03 ± 0.05 5.60± 0.54 0.01 ± 0.02 7.3 ± 14.5
St-ISOS-T-3 0.87 ± 0.05 5.67 ± 0.34 2.94 ± 0.52 59.3 ± 6.7
In-ISOS-T-3 0.45 ± 0.07 5.86 ± 0.93 1.02 ± 0.55 37.9 ± 13.8
Results for ISOS-‐D-‐3 condition show a significant drop in the PCE
values of both devices. However, it is most significant for inverted
cells. This drop is mostly affected by reductions in VOC and FF values,
which may be resulted from an increase in the density of deeper
traps due to oxidation of the active layer143–145. These traps can act as
recombination sites and disturb the internal electric field
distribution144. Moreover, at high temperature condition, diffusion of
metal from the top contact into the buffer layer results in a
decreased FF90,146. Results for ISOS-‐T-‐3 condition show that the PCE
82
value of the standard devices is less affected by this aging condition
and the VOC value remains intact. This could be due to the used low
temperature resulting in a slowed down chemical reactions taking
place in the organic materials. This effect is demonstrated in
Appendix A, Figure A.1, where degradation of encapsulated standard
devices in ISOS-‐T-‐3 condition is compared with the ones in ISOS-‐D-‐1
(darkness, room temperature) condition. We observe that for ISOS-‐
T-‐3 condition, after 3500 hours, the VOC value of the devices remains
above 80% of its initial value, while for the devices kept at room
temperature (ISOS-‐D-‐1), T80 point is reached much earlier at around
500 hours121,147. However, results for inverted devices in ISOS-‐T-‐3
condition (Figure 4.4), show a notable change in the device stability
compared to standard devices. Since both devices have the same
material system, these results indicate a degradation process not
related to standard photo-‐oxidation of the active layer. Under ISOS-‐
D-‐3 conditions, when high temperature and humidity are present,
degradation is accelerated and is influence by both chemical
reactions in the active layer and instabilities of the interlayers
resulting in pronounced degradation.
In order to detect possible degradations due to morphological
changes at the D-‐A interface, sEQE measurements were performed
after aging the devices. Figure 4.5 shows normalized sEQE spectra
and their corresponding fits for fresh and degraded devices at each
degradation condition. Table 4.4 represents the extracted CT
parameters.
83
Figure 4.5. sEQE measurements and their corresponding Fits for fresh and aged devices at a) ISOS D-‐3 and b) ISOS-‐T-‐3 degradation conditions.
Table 4.4. Extracted Voc and CT parameters through fits on sEQE spectra for fresh and aged devices.
Devices VOC (V) f (eV2) λ (eV) ECT (eV)
St-PHJ-Fresh 0.85 ± 0.05 0.0017 0.17 1.44
In-PHJ-Fresh 0.72 ± 0.07 0.0008 0.22 1.37
St-PHJ-ISOS-D-3 0.36 ± 0.28 0.0017 0.19 1.44
In-PHJ-ISOS-D-3 0.03 ± 0.05 0.0011 0.23 1.38
St-PHJ-ISOS-T-3 0.87 ± 0.05 0.0014 0.18 1.44
In-PHJ-ISOS-T-3 0.45 ± 0.07 0.0006 0.20 1.40
The sEQE results indicate that regardless of a change in VOC value
of the devices after degradation, the CT properties remain almost the
same. Unchanged f value after degradation suggests a morphological
stability at the DBP/C70 interface that means morphological changes
at the D/A interface are not responsible for a drop in voltage upon
aging of the devices. We further supported this by investigations on
morphological stability of the DBP-‐C70 interface through annealing
standard PHJ and BHJ devices. We fabricated these devices and
84
annealed them at 110 ℃ for 3 hours in dark inside the glovebox.
Extracted J-‐V parameters and CT properties for fresh and annealed
devices are presented in Figure 4.6 and Table 4.5.
Figure 4.6. sEQE measurements and Marcus Fits for fresh and annealed devices at 110 ℃ for 3 hours
Table 4.5. Voc and CT parameters extracted through fits on sEQE spectra for fresh and aged devices.
The results show that although, as expected, a significant higher f
value (higher amount of interface) is observed for BHJ cells, no
changes in CT properties and f value of the devices is seen upon
annealing.
85
For the inverted configuration (Figure 4.5), during degradation,
the C70 absorption (shoulder at 1.8 eV) decreases slightly relative to
DBP absorption. We ascribe the reason for this instability of C70 to
changes in the underlying BCP layer. It has been reported that
insertion of a BCP as ETL in inverted devices, depending on the
device area134, can hamper the device performance with low FF and
PCE133,148, and lead to poor stabilities148. Moreover, BCP tendency for
crystallization can induce defects at the ETL-‐C70 interface that can
give rise to increased VOC losses149.
In standard configuration, due to diffusion of the top Ag electrode
into the BCP layer, a Ag-‐BCP complex is formed87. This complex
introduces a new LUMO level aligned with the LUMO of the C70 that
facilitates electron transfer150. Nevertheless, the inverted structure
suffers from an inefficient electron extraction due to lacking this
metal-‐BCP complex151. Small device area of 10 mm2 being studied
here show reasonable device efficiencies for inverted architectures,
but with limited device stability.
We further investigated the instabilities by assessing the ETL
properties of different thicknesses of BCP in inverted configuration.
Moreover, two BCP ETL stacks based on C70 and Ag were tested in
these devices.
4.5 ETL charge transport properties and its effect on the stability We investigated the possible effect of the ETL contact on stability
of the inverted devices by assessing the ETL properties of three
different thicknesses of BCP (0.5, 5 and 10nm) and two BCP ETL
86
stacks based on C70 and Ag. It has been shown that deposition of an
ultrathin layer of C70 between ITO and BCP can improve the ETL
properties and increase the device yield149. The BCP-‐Ag stack can
provide a BCP-‐Ag semi-‐complex without a need for directly doping
the BCP layer.
Electron transport and exciton blocking properties of each
implemented ETL were examined via Electron-‐Only Devices (EODs)
and photoluminescence (PL) spectroscopy, respectively. These
measurements were performed for both fresh and degraded devices.
4.5.1 Electron-‐only devices Electron transport properties of the implemented ETL films were
assessed by performing space-‐charge limited current (SCLC)
measurements on EODs, Figure 4.7.a. EODs were fabricated with the
same methods described in chapter 3 by sandwiching the C70 layer
between the respective bottom ITO/ETL and top BCP/Ag contact. In
these devices, electrons are injected from the Ag electrode into the
devices and extracted from the ITO side. Five following devices were
fabricated:
• ITO/No BCP (0nm)/C70/BCP/Ag
• ITO/BCP (0.5nm)/C70/BCP/Ag
• ITO/BCP (5nm)/C70/BCP/Ag
• ITO/C70 (0.5nm)/ BCP (0.5nm)/C70/BCP/Ag
• ITO/BCP (2nm)/Ag (1nm)/BCP (2nm)/Ag (1nm)/BCP (2nm)/
C70/BCP/Ag
J-‐V characteristics of the EODs were recorded by applying a
sweeping voltage between +1 and −1 V at reverse bias using a
87
Keithley 2400 source measure unit. Measurements were performed
directly after fabrication and repeated after degradation period.
Figure 4.7. a) Electron only devices with different ETLs, and with deposited 10 nm BCP and 100 Ag on top b) JV measurements of the fresh EODs c) JV measurements after aging devices for 24 hours in ISOS-‐D3 and d) ISOS-‐T-‐3 degradation conditions.
Figure 4.7.b and 4.7.c show EOD characteristics for fresh and
degraded devices in ISOS-‐D-‐3 and ISOS-‐T-‐3 conditions for 24 hour.
J-‐V measurements show a lower series resistance for EODs with
BCP/Ag stack, which indicates better electron transport properties
with ease of electron transportation through the implemented ETL
layer. Among the investigated BCP thicknesses, interestingly, devices
without BCP layer at the bottom (0nm BCP) show no or very minor
deterioration. Noting that these devices have a top BCP layer, these
results indicate that the bottom BCP layer is indeed responsible for
88
the degradation in inverted cells, while the top BCP layer has a high
stability and no major rule in degradation of the devices. Therefore, a
pure BCP layer hampers electron extraction in the inverted devices,
which is expected due to lack of a metal-‐BCP complex.
4.5.2 Photoluminescence quenching measurements The exciton blocking properties of the ETLs were characterized
using PL measurements (for more information see 3.2.5). Samples
are consisted of the ETL and 100nm C70 deposited on top:
• ITO/No BCP (0nm)/C70
• ITO/BCP (0.5nm)/C70
• ITO/BCP (5nm)/C70
• ITO/C70 (0.5nm)/ BCP (0.5nm)/C70
• ITO/BCP (2nm)/Ag (1nm)/BCP (2nm)/Ag (1nm)/BCP (2nm)/
C70
Figure 4.8.a demonstrates that upon increasing the BCP thickness,
PL intensity of the C70 layer increases, which means an increased
exciton blocking capability for the ETL and thus minimum quenching
at the cathode interface. Among all samples, BCP/Ag stack shows
stronger PL emission, and hence minimized exciton quenching at the
cathode interface.
After aging the samples in ISOS-‐D-‐3 and ISOS-‐T-‐3 conditions for
24 hours, stabilities of the ETLs were tested by repeating the PL
measurements. Figure 4.8.b and 4.8.c represent the results for each
aging condition, and show overall decreased PL intensity in both
conditions compared with the fresh samples. However, in each set,
89
BCP/Ag stack shows higher PL intensity meaning better exciton
blocking properties.
Figure 4.8. Photoluminescence (PL) measurements for five ETLs, fresh and degraded at ISOS-‐D-‐3 and ISOS-‐T-‐3. The stack has the structure: (0.5nm)/BCP (0.5nm) and BCP (2nm)/Ag (1nm)/BCP (2nm)/Ag (1nm)/BCP (2nm)).
90
From these results we conclude that the BCP/Ag stack ETL
provides both improved electron transport and exciton blocking
properties as well as less interface degradation. Therefore, we
implemented this ETL candidate as an alternative to pure BCP in full
inverted devices, and evaluated the performance and stability of the
devices.
4.5.3 Improved stability of the inverted devices
The stability of the inverted devices with the following structure
were tested: ITO/BCP (2nm)/Ag (1nm)/BCP (2nm)/Ag
(1nm)/BCP (2nm)/C70 (30nm)/DBP (20 nm)/MoO3 (10 nm)/Ag
(100 nm)
Devices were aged at ISOS-‐D-‐3 condition for 24 hours and J-‐V
characteristics were recorded under 1 sun illumination for fresh and
aged devices. Results are illustrated in Figure 4.9 together with the
previously obtained results for 0.5nm BCP as ETL. Table 4.6 shows
the extracted J-‐V parameters for both types of devices.
The obtained results show both improved VOC and stability for the
inverted devices with BCP/Ag stack ETL, demonstrating that for
inverted cells, the initial larger VOC drop along with the accelerated
degradation compared to standard devices is caused by the BCP ETL.
Since J-‐V characterizations were performed through a bottom
illumination, slightly lower JSC is achieved due to a reflection from the
thin layer of Ag at the lower layers. Interestingly, performance of the
degraded inverted devices based on BCP-‐Ag stack in ISOS-‐D-‐3
condition is similar to the performance of the standard configuration
cells after aging at the same condition. This indicates that now the
91
interlayers possess a similar stability as those implemented in the
devices with standard configuration.
Figure 4.9. JV measurements for fresh and aged inverted device with 0.5 nm BCP and BCP/Ag stack.
Table 4.6. Photovoltaic characteristics of fresh and aged inverted devices with 0.5 nm BCP or BCP/Ag stack as their HTL.
In summary the results show that despite the differences in VOC of
the standard and inverted devices upon degradation in ISOS-‐D-‐3 and
ISOS-‐T-‐3 conditions, no morphological or CT state changes are
observed. These results suggest that instabilities are mainly
originated from electrode or interlayer degradations, and in inverted
92
devices the BCP ETL has a significant effect on the performance and
degradation of these devices. Evaluating the performance of the
implemented ETLs via EODs and PL measurements backed up these
conclusions. We showed that implementing an alternative BCP/Ag
stack as ETL for inverted devices can enhance both performance and
stability of these devices.
93
CHAPTER 5
Understanding the Degradation Mechanisms in Perovskite Solar Cells
The efficiency of hybrid organic-‐inorganic metal halide perovskite
solar cells (PSCs) have reached PCE of around 22%15 and 12%152 for
laboratory scale and large area cells, respectively. However,
combining high efficiency with high stability in a device still remains
challenging. There are numerous reports addressing the instability
of the active and transport layers in PSCs under exposure to heat,
light, electric field and air153–160. In real operational outdoor
conditions, these factors can simultaneously take part in degradation
processes making it complicated to understand the precise origin of
the degradation mechanisms involved161,162.
Recently, reversibility of certain PSC degradation processes under
illumination-‐darkness cycling has drawn a lot of attention163–166,
94
since it provides a more complicated stability issue than typically
seen in photovoltaic cells. The loss in the device performance being
caused by the illumination can be recovered fully or partially by
resting at dark conditions. This means that in real operational
conditions with diurnal cycling, a dark period following natural
sunlight might help the cell to recover.
On the other hand, some groups reported recently an opposite
behavior in which recovery occurs under illumination in darkness-‐
light cycles, which is known as “fatigue” behavior. This behavior is
related to the well-‐known light soaking effect, and is mostly
attributed to trap filling upon illumination157,163–168. To observe the
aforementioned two types of behaviors for PSCs during day/night
cycle, we have tested long-‐term stability of the PSCs in outdoor and
indoor conditions, and compared the results with already existing
figures of merit in stability measurements. Then the degradation and
recovery dynamics observed under outdoor day-‐night cycling were
related to the ones from constant illumination condition. This
strategy helps us to understand the degradation mechanism that
may be responsible for both types of diurnal behaviors. This chapter
summarizes the results presented in paper II and III, and the work is
done in collaboration with Ilse Katz institute for Nanoscale Science
and GTechnolog (outdoor testing) and the Holst center (perovskite
cell fabrication). Indoor testing was done at MCI, SDU.
5.1. Outdoor day/night degradation and recovery tests Outdoor degradation experiments were performed with two types of
PSCs:
95
glass/ITO/SnO2/Cs0.05((CH3NH3)0.15(CH(NH2)2)0.85)0.95PbI2.55Br0.45/spiro-‐
OMeTAD/Au cells (type I) and glass/ITO/TiO2/CH3NH3PbI3/Spiro-‐OMeTAD/Au mini-‐modules (type II).
Devices were degraded indoor by continuous illumination of
simulated sunlight (ISOS-‐L-‐1 protocol) at 60°C, and outdoor natural
sunlight in the Negev desert (the spectrum was measured very close
to the AM1.5G, ISOS-‐O-‐1 protocol) with performance testing under
simulated sunlight three times a day. The initial PCE of the devices
were ~15% and ~10% for the devices of type I and type II,
respectively. Devices were degraded until their PCE dropped to 80%
of its initial value (T80). This point plays a crucial rule as it defines
the overall lifetime of the solar cell and total energy generated by the
cell during its lifetime169. For indoor measurements, the type I
devices were placed under continuous illumination. At this
condition, T80 was reached in about 1 hour (𝑇!"!"#$~1 ℎ), Figure 5.1,
while in outdoor test conditions for both types of devices the
degradation happened at much slower speeds, also due to the effect
of the day/night cycles, which appeared as fluctuations in PCE values
during day and night periods, Figure 5.2.a and 5.2.b. For type I
devices, in the first 11 days of two weeks exposure, the cell
performance degraded during the days and recovered during the
nights, which resulted in higher “morning” PCE values compared to
“evening” PCE values. Therefore, the effect of the night recovery
resulted in much slower long-‐time degradation dynamics compared
with continuous illumination. In this pattern since the cell PCE would
cross the 80% mark multiple times, T80 as a point for lifetime
determination is a misleading parameter. This result demonstrates
96
the significance of including a light/dark cycling in stability
measurements for defining the cell lifetime.
Figure 5.1. Normalized PCE evolution for indoor contentious simulated sunlight illumination of glass/ITO/SnO2/Cs0.05((CH3NH3)0.15(CH(NH2)2)0.85)0.95PbI2.55Br0.45/spiro-‐OMeTAD/Au cells (type I).
Figure 5.2. Normalized PCE evolution during two weeks of outdoor exposure to natural sunlight (a) type I (glass/ITO/SnO2/Cs0.05((CH3NH3)0.15(CH(NH2)2)0.85)0.95PbI2.55Br0.45/spiro-‐OMeTAD/Au), and (b) type II mini-‐modules (glass/ITO/TiO2/CH3NH3PbI3/Spiro-‐OMeTAD/Au). All lines are guides for the eye.
Nighttime recovery of the PSCs does not result in 100%
restoration of the initial PCE due to two superimposed factors:
presence of a permanent damage, i.e., irreversible degradation
97
mechanism and/or a recovery requiring longer time than one night.
Moreover, a number of different degradation mechanisms can occur
simultaneously and some can dominate at different degradation
stages and define the dynamics of the PCE changing during the day.
In the curve shown in Figure 5.2.a after a certain aging time, the
effect of these factors lead to much closer “morning” and “evening“
PCE values and even invert the pattern during days 12-‐14, at which
the cell demonstrates “fatigue-‐like” behavior, similar to type II
devices (Figure 5.2.b).
Cell lifetime can for example be estimated using the maximum
PCE values measured every day, so that 𝑇!"!"#~4 d (i.e., morning PCE
values up to day 10 in Figure 5.2.a). 𝑇!"!"# counts for the irreversible
losses and/or incomplete recovery during one night; however, since
the dynamics of the diurnal degradation are not taken into account,
the total energy generation by the cell during its lifetime can be
overestimated from this approach170,171.
𝐸!!"!"# = 𝑃𝐶𝐸 𝑡 ∗ 𝑃!" 𝑡 ∗ 𝑑𝑡
!!"!"#
! (5.1)
Where Pin is the power of incoming sunlight (for simplicity, Pin=1
sun=100 mW/cm2).
The evolution of the daily generated energy output by cell type I is
depicted in Figure 5.3.a.
𝐸!"# = 𝑃𝐶𝐸 𝑡 ∗ 𝑃!" 𝑡 ∗ 𝑑𝑡!!"#
! (5.2)
Where 𝑡!"# is the illumination time during one day. Now, we can
estimate 𝑇!"! ≈ 9 d at which Eday drops to 80% of its value. Therefore,
from this definition of T80, both the reversible and irreversible
98
degradation of the cell performance are taken into account, which
suggests a reliable figure of merit for PCE stability.
Figure 5.3. Normalized evolution of daily energy output, Eday, of (a) cell type I, and (b) mini-‐modules type II. All lines are guides for the eye.
Now, using 𝑇!"! ~9 d, we calculate the total energy generated by the
cell during its lifetime as:
𝐸!!"! = 𝑃𝐶𝐸 𝑡 ∗ 𝑃!" 𝑡 ∗ 𝑑𝑡!!"!
! (5.3)
This value connects the cell performance and its stability whose
overall improvement is the ultimate goal of the photovoltaic
technology. This relation does not depend on the certain PCE
changes day/night (illumination/darkness) cycle due to the reason
that in both device types, we observe a behavior resulted from
superposition of reversible and irreversible mechanism, which is
very different from that of continuous illumination experiment.
Overall these results suggest that a light-‐dark cycling is to be used
for precise stability measurements of PSC, and that a new set of
figures of merit are required to describe the performance, stability
and the interplay between them.
99
5.2 Indoor degradation and recovery dynamics To understand the dynamics of degradation and recovery process
under light/dark cycles further, we performed indoor degradation
under continuous light illumination and recovery at dark to different
extends. This allows us to suggest degradation mechanism that may
be responsible for both types of behaviors and relate them to the
results from continuous illumination tests.
For this purpose, we used:
glass/ITO/SnO2/Cs0.05((CH3NH3)0.15(CH(NH2)2)0.85)0.95PbI2.55Br0.45/
spiro-‐OMeTAD/Au cell (type I) (more information regarding the
materials and methods can be found in paper II172).
Indoor illumination was performed in ISOS-‐L-‐1 protocol in
ambient air and using a degradation setup equipped with a wide
area solar simulator (metal halide display/optic lamp HMI from
Osram). The ambient temperature (60 °C ) and intensity were
monitored by sensors, and recorded along with the solar cell
parameters. During the exposure test, solar cell performance was
recorded in situ every 2 minutes via J-‐V measurements in forward
direction, and the parameters were tracked. During the dark
recovery, the cells were kept in the dark at room temperature and
were exposed to light only when J-‐V measurements were conducted
every hour for the first 8 h of degradation.
Samples were divided into three batches and aged until their
efficiency value reached 80% (T80), 60% (T60) and 50% (T50) of the
initial value. Then cells were located in dark condition for recovery.
Figure 5.4 shows the evolution of cell performance parameters.
100
Figure 5.4. Evolution of PV parameters of PSCs under continuous 1-‐sun indoor illumination, interrupted at T80 (a) T60 (b), or T50 (c, d). Gray areas show their subsequent recovery in the dark.
Figure 5.5. Evolution of PV parameters after turning on the light and continuous simulated 1-‐sun re-‐illumination of the PSCs after T50 and recovery in the dark (gray areas): (a) The cell whose PCE dark recovery reached saturation (as in Figure 5.4.c); (b) The cell whose PCE dark recovery did not reach saturation (as in Figure 5.4.d).
For cells, which their PCE reached to 80% of its initial value, full
recovery was observed after a few hours of resting in dark (Figure
101
5.4.a). This result is in accordance with the previously reported
observation for indoor light-‐darkness cycling of state-‐of-‐the-‐art PSCs
(PCE~20%)164. At this stage, degradation is determined by the decay
of mainly open circuit voltage (VOC) and fill factor (FF) and minor
contribution of short circuit current (JSC). This result suggests that
this reversible mechanism with fast recovery should not cause a
permanent damage and a long-‐term deterioration of the PSC
performance under day/night cycle. For reversible degradation in
PSCs, three different mechanisms were previously suggested: (1)
migration of ion vacancies, upon which both halide and cation
vacancies are moving in the perovskite toward the interfaces with
charge transport layers that can affect the charge extraction from the
active layer164; (2) in inverted planer PSCs, light induced reversible
JSC degradation is attributed to the formation of metastable deep
traps in the perovskite layer163; (3) a fully reversible halide
segregation into Br-‐rich and I-‐rich domains, which can be observed
as two photoluminescence peaks that merge into one when allowed
to relax in dark166,173,174. These domains give rise to trap assisted
recombinations for photo-‐excited charge carriers175,176. However, in
our case, PL measurements reveal that phase segregation is not
likely the reason of our results, probably as the Br content is very
low (~15%) in our perovskite layer (more information is provided in
paper II172 and its supplementary information177). Moreover, we
observed that under illumination, both PL intensity and JSC decreased
and recovered in the dark. While the degradation caused by
interfacial effects would result in decreased JSC and increased PL
yield due to deterioration of charge transfer from the active layer.
Therefore, similarity in reversible degradation dynamics for these
102
parameters may be attributed to light induced formation of bulk
nonradiative recombination centers in the photoactive layer, which
may be related to the metastable deep traps (mechanism number 2)
or to lattice point defects178. Hence, formation of bulk traps is one of
the possible mechanisms that contribute to reversible degradation.
Reversible VOC and FF changes may have been caused by trap
formation/annihilation and/or light-‐induced migration of ionic
species in the perovskite layer, which lead to changes in the charge
extraction and electric field across the cell164,165.
To study the dynamics of later degradation stages (t>T80), the
second and third batches were illuminated further until the PCE of
the cell reached 60% and 50% of their initial value, and then the
devices where placed in dark condition for recovery (Figure 5.4.b
and Figure 5.4.c,d respectively). Dynamics of further aging the cells
to t~T50 were found to be dramatically different from early stage
degradation. Within the first 10 minutes of the dark storage, a rapid
drop in JSC and FF was observed for all cells resulted in further decay
of PCE from 50% to around 10% of its initial value (Figure 5.4.c,d).
Corresponding J-‐V curve showed an “s-‐shape” distortion (see Figure
S4 in supplementary information177 of paper II172), which has been
attributed to surface recombination or charge accumulation at the
interface between perovskite and transport layer179.
The PCE drop in the dark then followed by a slow recovery
process, which ended up in two cases: either reached saturation over
5 days of measurements or continued its evolution (Figure 5.4.c and
5.4.d respectively, after turning off the light). It is not clear if the
observed difference is qualitative or is due to a different recovery
time scale. Nevertheless, in both cases, due to occurrence of an
103
irreversible mechanism, a full recovery of the PCE was not observed,
and only reached a maximum saturation level of 70-‐80% of its initial
value for the behavior shown in Figure 5.4.c.
In order to investigate the dynamics of PCE drop after turning off
the light, the cells from both cases were re-‐illuminated (Figure 5.5).
For the cells whose PCE reached saturation in dark (as in Figure
5.4.c), no further light soaking improvement were observed (Figure
5.5.a). However, the cells for which the PCE did not saturate (as in
Figure 5.4.d), a rapid PCE improvement (mostly resulted from
increases in JSC and FF) was observed, which was then degraded
upon further illumination (Figure 5.5.b). Qualitatively, this type of
behavior is similar to type II behavior observed in outdoor
measurements in which the cells showed recovery under subsequent
illumination (Figure 5.2.b).
An intermediate case between degradation times of T80 and T50 is
when the cell is degraded to 60% of its initial PCE (T60) (Figure
5.4.b). After this point, a dark storage results in the following
dynamics: VOC was recovered rapidly (as in the T80 case) and drops in
JSC and FF were slowly recovered (as in the T50 case, but much
slower). The result was reflected in a slow PCE recovery that took
significantly longer than one night in time scale. Therefore, to our
view Figure 5.4.b and 5.4.d illustrate apparently irreversible
dynamics due to the fact that the recovery rate is such slow that the
full recovery is not practical under real operational day-‐night
stressing. At this stage J-‐V curve shows an s-‐shape when measured
without light soaking (see Figure S4 in supplementary information177
of paper II172).
Until now these dynamics have not been described in the
104
literature. Therefore, to explain this behavior we discuss a possible
scenario172: at later aging stages, shallow interfacial traps of a
different origin than the deep bulk traps are generated under
illumination. While the cell is under illumination, these traps are
being occupied by photogenerated carriers and are neutralized
which mitigate their effect on PSC performance. When the light is
turned off, detrapping of the charge carriers leads to charging of
these states that forms an interfacial charge extraction barrier.
Therefore, it may result in the J-‐V curve’s s-‐shape distortion and
correspondingly JSC and FF decrease as observed in our
experiment180,181.
These results reveal that, although the traps are formed during
illumination, their effect is appearance only after the light is turned
off, which is not the case for the experiments commonly performed
under continuous illumination. This emphasizes the importance of
considering a dark-‐light cycling as a viable strategy for stability
testing of the PSCs, and thus lifetime determination164.
As shown in Figure 5.5.b, by re-‐illuminating the cell whose PCE
did not reach saturation (as in Figure 5.4.d), both JSC and FF
increased significantly within the first 10 minutes due to trap filling
upon light soaking before further degradation takes place. According
to the literature, the effect of light soaking can take from minutes to
hours to reach saturation167,168. Notably, our findings are
considerably different than commonly observed light soaking effect
in the following way: (1) in our case, the cell developed following
~50% PCE degradation in contrast to a feature of fresh devices; (2)
the PCE improvement under illumination was mainly related to
increase in JSC and FF in our case rather than commonly observed
105
improvements in the VOC from light soaking182–186; and (3) in our case
the performance improvement was followed by degradation under
further illumination (Figure 5.5.b). As it has been previously
suggested, deep traps lead to VOC decrease, while shallow ones, being
located near the charge extraction interface, may contribute to JSC
variations by inducing interface charging and poor charge extraction.
Hence, the hypothesis of formation of shallow interfacial traps
suitably describes our observations. These shallow traps can be
originated from interstitial ions or ion vacancies at the
perovskite/transport layers interface178,187.
For the type of behavior shown in Figure 5.4.c, the PCE partially
recovered and saturated after a long time in the dark. Due to the fact
that upon re-‐illumination of the recovered cell no light soaking effect
was observed (Figure 5.5.a), this type of slow recovery can be related
to trap states disappearing, (e.g., due to back-‐migration of ionic
species). To look further into this matter, more investigations using
secondary ion mass spectroscopy (SIMS) are under way.
Figure 5.4.b and 5.4.c show irreversible long-‐term degradation in
which the PCE value reaches saturation level without full recovery.
This behavior was also noted in long-‐term outdoor degradation of
the cells172. Irreversible degradation mechanism may be due to
decay of the perovskite photoactive layer111,188–193, as well as
transport layers and contacts22,194,195, especially catalyzed by
interlayer ion diffusion161,195–197. We should also take into account
the effect of oxygen and water penetration through encapsulation as
our indoor and outdoor experiments were conducted in air.
There are factors that have been reported to contribute to the
irreversible degradation mechanism that can be considered, such as
106
chemical decomposition of the perovskite layer upon exposure to
light, effect of humidity, and temperature188–191, as well as
irreversible trap generation192 and ion migration to the perovskite
from contact/transport layers193,198.
The chemical decomposition is typically accompanied by PbI2
formation resulting in a color change to yellow in the samples. In our
case, we did not visually observe any color change in the cells during
the degradation. Moreover, Raman scattering measurements did not
reveal PbI2 peaks even after cells were degraded to 50% of the initial
PCE (more information can be found in section 7 in the supporting
information177 of paper II172). Although the sensitivity of the Raman
technique to small PbI2 quantities can be limited, we suggest that the
irreversible loss of the cell performance was not resulted from
significant decomposition of the perovskite layer itself.
5.3 Correlation between indoor and outdoor stability measurements Reversible and irreversible degradation mechanisms were
observed for both indoor and outdoor measurements.
In case of outdoor measurements, in the first stage of the aging
process when the cell is degraded to not more than 80% of its initial
PCE, the decrease in the VOC and FF was partially compensated by
nighttime recovery172 (type I dynamics). This effect slows down the
rate of the long-‐term degradation compared to a constant
illumination condition199, This mechanism is similar to the one
depicted in Figure 5.4.b (t<T50) in which the PCE decay is mainly
caused by the decrease in VOC and FF with minor contribution of JSC
107
deterioration, and then the performance is significantly recovered in
the dark. Full recovery was not reached in outdoor conditions due to
“apparently irreversible” processes occurring at t>T80 and a recovery
time longer than one night. However, we observed that increasing
the dark recovery time did not significantly improve the degree of
recovery. At later aging stages of outdoor degradation, the diurnal
behavior of the cells is switched from type I to type II in which PCE
decreased over night and increased under subsequent illumination.
This behaviors is roughly corresponding to the one depicted in
Figure 5.4.d in which after PCE decreased to less than 50% of its
initial value, a sharp PCE decrease is observed immediately after
turning off the light. At this stage both indoor and outdoor
experiments show a decrease in the PCE during the darkness and its
increase under subsequent illumination, which suggests a similar
degradation mechanism(s) for both cases.
Overall, summarizing the results that were shown for indoor and
outdoor degradation tests, we suggest the following scenario for cell
degradation: at the very early aging stages (t≤T80 in our case),
degradation mechanism is reversible. Then at t>T60, both reversible
and irreversible degradation mechanism take place. Finally, at t~T50,
a third type of unique dynamic kick in that changes the diurnal
dynamics of type I to that of type II, which result in further
degradation after turning of the light and subsequent improvements
in PCE under light soaking.
In summery, these results reveal that understanding the
degradation processes under real operational conditions is only
achievable by including the light-‐darkness cycles in experimental
protocols for the assessment of PSCs long-‐term stability.
108
109
CHAPTER 6
Summary and Outlook Organic and hybrid solar cells have become appealing
technologies due to their unique advantages. Efficiencies of organic
and perovskite solar cells have recently reached world records of
15% and 22%, respectively, however, in order for them to be
commercially viable they need to have at least 10 years of stability,
which has introduced challenges to the field.
In this work, we studied the stability of both OSC and PSC devices.
For organic solar cells, we fabricated and evaluated performance and
stability of DBP-‐C70 based organic solar cells with standard and
inverted device configurations. We focused on the morphological
changes at the DBP-‐C70 interface, which can lead to intrinsic device
degradations. We performed sEQE measurements to identify
molecular and morphological changes at D-‐A interface after aging the
devices at two different ISOS degradation test conditions (ISOS-‐D-‐3
and ISOS-‐T-‐3). The results revealed a pronounced stability at the D-‐A
interface for both devices. Instead, the results suggest that the drop
110
in performance and stability of inverted devices is resulted from
using a BCP bottom ETL in this configuration. This shows the
potential strong impact of all layers and interlayers on the stability of
the devices. To address this, we evaluated exciton blocking and
electron transport properties of different BCP film thicknesses as
ETL, and two alternatives based on C70/BCP and BCP/Ag stacks. The
ETL layers were evaluated using standard JV, SCLC and PL
measurements together with stability tests. The results showed
improved device performance and stability for inverted devices
implementing the BCP/Ag stack as ETL, which implies the significant
contribution of a pure BCP ETL to the hampered stability of the
inverted devices.
For perovskite solar cells, we tested long-‐term stability of the
devices under indoor and outdoor conditions during day/night
cycles, revealing both reversible and irreversible degradation
dynamics. We compared the results with already existing figures of
merit for stability measurements, and related the ones observed for
outdoor degradation tests to the results obtained from indoor
constant illumination condition. We concluded that under
illumination/dark cycles, at early stages of the aging process (t≤T80),
degradation mechanism is still reversible; however, as the aging
process continues (t>T60), irreversible degradation mechanism
appears. At a stage when t~T50, a third type of mechanism kicks in,
which changes the degradation pattern and degradation in dark is
seen, instead of typically observed degradation under light. These
results revealed the importance of including the light-‐darkness
cycles in the experimental stability protocols to achieve a deeper
111
understanding of the degradation processes under real operational
conditions.
This work thus touches upon some essential points in the stability
of both organic and hybrid solar cells, however, there are still lots of
investigations to be made and unknown to be explored. For the OPV
devices investigated here, there is still a need for a better ETL
candidate with improved properties to be used in inverted device
configuration. Moreover, as a useful method, Fourier-‐transform
infrared spectroscopy (FTIR) can be used to study the degradation
mechanisms related to pure chemical changes in the investigated
materials.
Further improvements in performance and stability of the organic
and hybrid solar cell devices can be achieved through systematic
investigations and proper characterization methods. Degradation
studies under real operational conditions can reveal more
degradation paths taking place inside the devices and link our
findings to the broader context in the literature. It is expected that
these results will be a path to further study the degradation
mechanisms, and will be built upon so that true implementation of
organic and hybrid solar cells in our everyday life will be realized
soon from all the joint efforts made on these topics.
112
113
Appendix A
ISOS-‐T-‐3 degradation test for standard structure DBP-‐C70 based organic solar cells with standard strictures were
encapsulated using a Dow Corning 732 silicon vacuum sealant and a
glass on top in order to decrease the effect of the air and extrinsic
degradation. Then half of the devices were placed inside a thermal
chamber under ISOS-‐T-‐3 test condition ((-‐40℃ and room humidity-‐
darkness), and the other half were stored in a shelf in darkness at
room temperature (ISOS-‐D-‐1). JV characteristics of the devices were
recorded regularly using a solar simulator until they reached their
T20 (where the efficiency reaches 20% of its initial value). Figure A.1
presents the decay curves of the devices.
As it can be seen, for the devices which were kept under ISOS-‐T-‐3
at low temperature, the VOC value remain above 80% of its initial
value after 3500 hours (almost 145 days). However, for devices kept
at room temperature, T80 point is reached much earlier and around
500 hours (20 days).
114
Figure A.1. Lifetime curves of the DBP-‐C70 based organic solar cells with Standard planar structure. Top” Under ISOS-‐D-‐1 (darkness, room temperature and room humidity) and bottom: under ISOS-‐T-‐3 (-‐40℃ and room humidity-‐darkness) aging condition
115
Appendix B
TTF as donor material in organic solar cells Tetrathiafulvalene (TTF)200–202 is recognized as a strong π-‐
electron organic donor and has been of great interest since the early
1970s due to the discovery of the first organic metal203,204. TTF is an
small organic molecule with good mobility and a wide spectrum of
applications205–207(Figure B.1). This molecule can reversibly undergo
a two-‐step oxidation process to form the radical cation and the
dication forms of the TTF and is in general thermodynamically stable
to many synthetic transformations. Moreover, it is relatively stable in
air, and its oxidation potentials can be finely tuned by the attachment
of electron-‐donating or electron-‐withdrawing substituents208.
Overall, interesting charge-‐transport properties, photoinduced
electron transfer leading to highly stabilized ion radical pairs,
together with possibility of tailoring the properties, makes TTF
derivatives appealing candidate for organic photovoltaic
applications.
116
Figure B.1. Various applications of TTF in super molecular chemistry and material chemistry202,209.
TTF is widely researched in charge transfer salts where it conducts
electrons200, and it is possibly the best organic electron-‐donor
material and used extensively in supramolecular chemistry210.
However, surprisingly their application in organic and dye
synthesized solar cells has been very limited211,212.
In this work we tested three generations of novel TTF derivatives
as electron donor with acceptor PC60BM in organic solar cells. We are
aiming to improve the performance and stability of the organic solar
cells by implementing and optimizing the properties of this molecule
in the active layer of the devices.
The novel TTF molecules have been synthesizes by our
collaborator Dr. Steffen Bähring at SDU FKF Odense, Denmark, and
solar cell device fabrication and optimization has been performed at
SDU NanoSYD.
117
B.1 First-‐generation TTF molecules We started the process by testing the received first generation of
the synthesized novel TTF derivatives named as TTF1 and TTF2 with
different electron-‐donor and absorption properties (Figure B.2).
Cyclic voltammetry (CV) and HUMO-‐LUMO levels of each molecule
are shown in Figure B.3. The molecules were tested in TTF:PC60BM
based solar cell devices with the structure of ITO/ZnO/TTF:PC60BM
/MoO3/Ag.
Figure B.2. Skeletal formula of TTF1 and TTF2 molecules.
Figure B.3. (left) Cyclic voltammetry and (right) HUMO LUMO levels for TTF1 and TTF2 molecules
118
To fabricate the devices, pre-‐patterned ITO coated substrates
were cleaned using acetone and isopropanol in ultrasonic bath and
blow-‐dried with nitrogen gun. In order to avoid exposure to air, the
whole fabrication process was done in a nitrogen glove-‐box. A
commercial ZnO semi-‐conductive ink (Genes’ Ink, France) was spin-‐
coated on top of ITO with 1000 rpm for 60 sec and annealed on a hot
plate at 130°C for 15 min. First, we started with TTF1 molecule from
which TTF1:PC60BM solutions were prepared by dissolving 20 mg of
TTF1 and 20 mg of PC60BM in Chlorobenzene (CB) solvent to result
in 40 mg/ml concentration with the ratio (1:1). To spin-‐coat the
active layer, three different rpms were used: 500, 1000 and 2000. In
parallel with device fabrication, the solutions were separately spin-‐
coated over cleaned glass substrates in order to investigate the
quality and thickness of the films. Finally, 10 nm MoO3 and 100 nm
Ag were vacuum deposited on top of the deposited films, and a mask
was used to define a cell area of 10 mm2 for each cell (design of the
devices are similar to the ones described in chapter 3).
J-‐V characteristics of the devices were recorded under 1 sun
illumination in a solar simulator (3000 class AAA solar simulator
from Abet Technology Inc., USA). Among three types of devices, only
the ones with deposited active layer at 2000 rpm showed
photovoltaic effect with 0.001% PCE and 0.08 mA/cm2 JSC. The
deposited films over the cleaned glasses were very thin and showed
absorption only in the UV range (Figure B.4).
119
Figure B.4. Absorption spectra of the three types of TTF1:PC60BM films deposited at three different rpms (500, 1000 and 2000).
B.1.1 increasing the concentration to increase
absorption In the next step, a solution with a 3 times higher concentration
(120 mg/ml) was made, and the whole experiment was repeated.
However, similar results were obtained without any improvement,
confirming that the devices were hampered by the poor visible
absorption from the material. Moreover, phase segregation of
PC60BM and TTF1 molecules were seen, which resulted in low
quality and uneven films.
B.1.2 Changing the solvent and using dynamic dispense To improve the solubility of the molecules, we used Chloroform
(CF) solvent instead and annealed the solution before and during
spin coating. In this experiment, 2 solutions with 40 mg/ml and 120
mg/ml concentrations were prepared over night by stirring on a hot
plate at 80 °C. This time 1000, 2000 and 3000 rpms were used to
120
deposit the films. Moreover, since CF is a high vapor pressure
solvent, each solution was applied while the substrate was spinning
(dynamic dispense) to avoid evaporation of the solvent and resulting
in a better film uniformity.
For each sample, the quality of the deposited films and their
absorption properties were evaluated (Figure B.5), and devices were
fabricated based on the highest absorber film. However, none of the
developed devices worked. It was noticed that right after deposition
of the films, crystalline zones were growing very fast on each
substrate, which may be the main result for the failure of this
experiment (Figure B.6).
Figure B.5. Absorption spectrum for deposited films from TTF1:PC60BM solutions with 40mg/ml and 120mg/ml concentration, deposited at 1000, 2000 and 3000 rpm.
Figure B.6. Crystallized zones formed right after spin coating.
121
Crystallization can occur either due to the reason that the two
materials reject each other (originating from their structure or
dipole dipole interaction) or due to a very strong attraction between
the TTF molecules. Unlike polymers, this often happens with small
molecules, as they are not connected in long chains, which makes
them more mobile. To solve this problem, we added a polymer
(polystyrene=PS) into the solution, which would act as a sort of
'spacer' that prevents them from aggregating.
B.1.2 Adding polystyrene (PS)
The experiments with both TTF1 and TTF2 molecules were
repeated, and TTF1:PC60BM and TTF2:PC60BM solutions with 40
mg/ml and 80 mg/ml concentrations containing either 10% or 5%
PS were made. SPS with 8 mg/ml concentration (10% PS in a 80
mg/ml concentration) was prepared by dissolving a 33mg
polystyrene bead in about 4.12 ml chloroform, and left for stirring
for a few days on a hot plate at 50 °C. Next, 8 different types of
TTF:PC60BM solutions were prepared by adding different ratios of
SPS (Table B.1).
Films were spin-‐coated over cleaned glass substrates with 1000,
2000 and 3000 rpms. Photographs and optical microscope images of
each substrate is shown in Figure B.7 and B.8. Absorption properties
of each film was recorded and is illustrated in Figure B.9
122
Table B.1. 8 different solutions with TTF1 and TTF2, with two different concentrations (80 and 40mg/ml) containing 5% and 10% PS.
TTF1:PC60BM
40mg/ml
5% PS 10% PS
20mg TTF1+20mg PCBM
0.25ml SPS+0.75ml CF
20mg TTF1+20mg PCBM
0.5ml SPS+0.5 ml CF
80mg/ml
5% PS 10% PS
40mg TTF1+40mg PCBM
0.5ml SPS+0.5ml CF
40mg TTF1+40mg PCBM
1ml SPS
TTF2:PC60BM
40mg/ml
5% PS 10% PS
20mg TTF2+20mg PCBM
0.25ml SPS+0.75ml CF
20mg TTF2+20mg PCBM
0.5ml SPS+0.5 ml CF
80mg/ml
5% PS 10% PS
40mg TTF2+40mg PCBM
0.5ml SPS+0.5ml CF
40mg TTF2+40mg PCBM
1ml SPS
Figure B.7. Deposited films from TTF1:PC60BM solutions with 40mg/ml and 80mg/ml concentration containing 5% and 10% PS.
123
Figure B.8. Deposited films from TTF2:PC60BM solutions with 40mg/ml and 80mg/ml concentration containing 5% and 10% PS.
Figure B.9. Absorption spectra of films deposited from 4 different solutions based on TTF2:PC60BM, each deposited with 3 different rpms.
124
Overall, TTF2:PC60BM films showed better film quality, and
stronger absorption properties in visible wavelengths. Therefore, the
TTF2 molecules were evaluate as the active layer material in the
solar cell devices.
B.1.3 Assessing TTF2 molecules in organic solar cells
As the next step, 4 different TTF2:PC60BM solutions were coated
over ITO/ZnO substrates to investigate the quality of the deposited
films. Figure B.10 shows optical and AFM images of the spin-‐coated
solution over ITO/ZnO substrates, and Figure B.11 illustrates
absorption properties for each sample.
Figure B.10. Optical microscope and AFM images of the films deposited from TTF2:PC60BM solutions with two different concentrations and two different PS concentrations
125
Figure B.11. Absorption properties of TTF2:PC60BM with 40mg/ml and 80mg/ml containing 5% and 10% PS.
As the results show, the quality and uniformity of the deposited
films has significantly improved. Based on these results, solar cell
devices based on TTF2:PC60BM active layer were fabricated. In this
process, the 4 types of solutions were deposited over ITO/ZnO
substrates with 1000 rpm, followed by vacuum deposition of MoO3
and Ag on top. Figure B.12 shows J-‐V curves and extracted
parameters for each type of device.
As it can be seen, although the efficiency is very low, this time the
cells show reproducible performance results. Low PCE is resulted
from a low current (JSC), which is expected as very little visible light
is absorbed. Therefore, the challenge were to incorporate modified
TTF molecules that absorb in the visible area.
126
Figure B.12. J-‐V curves and extracted parameters of 4 different types of devices based on TTF2:PC60BM active layer.
B.2 Second-‐generation TTF molecules For the second-‐generation TTF molecules, we introduced an
electron-‐accepting moiety (benzoyl groups) to lower the LUMO
energy level, and thereby increase the absorption band between 400
and 600 nm of varying intensity. We assessed two TTF editions in
our solar cells (Figure B.13).
127
Figure B.13. Skeletal formula of TTF2I and TTF2II molecules.
Devices were fabricated with the new TTF2 compounds in their
active layers. TTF2:PC60BM solutions with 40 mg/ml and 80 mg/ml
concentrations were prepared, and devices were fabricated as
described before. Final results showed only very low PCE around
0.004% for 80 mg/ml solutions, and not much improvement.
B.3 Third-‐generation TTF molecule This time our collaborator synthesized third-‐generation TTF
molecules with better absorption properties, which can assist with
improved cell characteristics. These Molecules are editions of the
previously incorporated TTF molecules in which an organic dye is
used as a platform (quinoidal porphyrin), which has higher visible
absorption with the TTF still being able to modulate the electronic
properties (ex-‐TTF porphyrin) (Figure B.14).
128
Figure B.14. Absorption spectra of exTTF-‐Porphyrin molecule
We received two derivatives of this ex-‐TTF porphyrin and for
convenience, we named them TTF56 and TTF80 (Figure B.15). CV
measurements of the two molecules are shown in Figure B.16. Fine-‐
tuning the electronic properties of the molecules to introduce more
electron withdrawing moieties such as COOMeTTF56 results in
0.227 V higher oxidation potential compared to SEt TTF80.
Figure B.15 Skeletal formula of TTF80 and and TTF56 molecules.
129
Figure B.16. CV recorded for TTF80 and TTF56.
With the same procedure as explained before, we made 4
solutions with these two molecules and PC60BM (1:1) having either
40 mg/ml or 80 mg/ml concentration, and containing 5% PS. This
time we slightly changed the recipe. We spin-‐coated ZnO over pre-‐
patterned ITO substrates at 2000 rpm for 60s and baked it at 130°C
for 15 min. Each substrate left in the vacuum for 30 min for complete
drying. Then, the TTF:PC60BM solution was dynamically spin-‐coated
at 1000 rpm for 60s followed by 1500 rpm for 5s resulted in 200-‐
260 nm thick film. Again, a vacuum drying procedure was carried out
for 30 min to give enough time for evaporation of the solvent. After
that, samples were transferred into a vacuum deposition system
where 10 nm MoO3 and 100 nm Ag were deposited as HTL and
electrode on top.
Final device characterization shows reproducible results with
very low PCE of 0.04% and 0.01% for TTF56 and TTF80,
respectively. The measured VOC is comparable to reference donor
130
materials (P3HT); however, the JSC and FF are very low (Figure B.17).
This may be due to a poor morphology, which may be improved
using a different donor and acceptor ratio in the active layer.
Figure B.17. J-‐V characteristics and extracted device parameters for solar cell devices based on TTF56:PC60BM and TTF80:PC60BM with (1:1) ratio.
B.3.1 changing the donor acceptor ratio In another effort, ratio of the PC60BM was increased three times
more and solutions were made with TTF (80 or 56):PC60BM (1:3)
with 40 mg/ml concentration containing 2.5% PS (2 solutions). Solar
cell devices were fabricated with these solutions and their
performances were characterized. Results show an improved PCE to
0.26% and 1.11 mA/cm2 JSC for devices based on TTF56:PC60BM
(Figure B.18).
131
Figure B.18. J-‐V characteristics and extracted device parameters for solar cell devices based on TTF56:PC60BM and TTF80:PC60BM with (1:3) ratio.
B.3.2 Degradation measurements As a final attempt, we tested stability of the TTF56:PC60BM based
solar cells by aging them in ISOS-‐D-‐3 condition for 24 h. Results
show a degradation path in which 50% drop in PCE is mostly caused
by a drop in VOC and JSC and less related to a change in the FF (Figure
B.19). Further investigations are needed to identify the origin of the
degradation mechanisms taking place in these devices.
132
Figure B.19. Stability characterization of TTF56:PC60BP based solar cells in ISOS-‐D-‐3 condition Overall, the results so far indicate the potential application of the
TTF molecules as donor material in organic solar cells, which
however requires much optimization of the material and device
properties. Ex-‐TTF derivatives show interesting electronic
properties to be used in organic photovoltaics. Although the results
for BHJ type OSC are poor, they show a potential to improve the
overall performance by a systematic approach to modify the
electronic properties of the molecules and device fabrication
techniques.
133
Bibliography 1. Bagnall, D. M. & Boreland, M. Photovoltaic technologies. Energy
Policy 36, 4390–4396 (2008). 2. Hanna, M. C. & Nozik, A. J. Solar conversion efficiency of
photovoltaic and photoelectrolysis cells with carrier multiplication absorbers. J. Appl. Phys. 100, (2006).
3. Gauthier, S. et al. Mild cognitive impairment. Lancet 367, 1262–1270 (2006).
4. Nazeeruddin, M. K. et al. Conversion of Light to Electricity by cis-‐X2Bis (2,2′-‐bipyridyl-‐4,4′-‐dicarboxylate) ruthenium (II) Charge-‐Transfer Sensitizers (X = Cl−, Br−, I−, CN−, and SCN−) on Nanocrystalline TiO2 Electrodes. J. Am. Chem. Soc. 115, 6382–6390 (1993).
5. Shauddin, S. Comparison among Various Emerging PV Cells with History, Current Status and Future Challenges. Energy and Power 3, 91–105 (2013).
6. A Pochettino, Acad.Lincei Rend, 1906, 15, 355. No Title. 7. Yu, G., Gao, J., Hummelen, J. C., Wudl, F. & Heeger, A. J. Polymer
Photovoltaic Cells: Enhanced Efficiencies via a Network of Internal Donor-‐Acceptor Heterojunctions. Science (80-‐. ). 270, 1789–1791 (1995).
8. Facchetti, A. π-‐Conjugated polymers for organic electronics and photovoltaic cell applications. Chem. Mater. 23, 733–758 (2011).
9. Zhokhavets, U., Erb, T., Gobsch, G., Al-‐Ibrahim, M. & Ambacher, O. Relation between absorption and crystallinity of poly(3-‐hexylthiophene)/ fullerene films for plastic solar cells. Chem. Phys. Lett. 418, 347–350 (2006).
10. Tseng, H.-‐R. & Lin, C.-‐L. Organic photovoltaic cell. 183, 1–19 (2011).
11. Mateker, W. R. & McGehee, M. D. Progress in Understanding Degradation Mechanisms and Improving Stability in Organic Photovoltaics. Adv. Mater. 29, (2017).
12. Che, X., Li, Y., Qu, Y. & Forrest, S. R. High fabrication yield organic tandem photovoltaics combining vacuum-‐ and solution-‐processed subcells with 15% efficiency. Nat. Energy 3, 8–13 (2018).
13. Cao, H. et al. Recent progress in degradation and stabilization
134
of organic solar cells. J. Power Sources 264, 168–183 (2014). 14. Mateker, W. R., Sachs-‐Quintana, I. T., Burkhard, G. F.,
Cheacharoen, R. & McGehee, M. D. Minimal long-‐term intrinsic degradation observed in a polymer solar cell illuminated in an oxygen-‐free environment. Chem. Mater. 27, 404–407 (2015).
15. Yang, W. S., Park, B.-‐W., Jung, E. H. & Jeon, N. J. Iodide management in formamidinium-‐lead-‐halide – based perovskite layers for efficient solar cells. Science (80-‐. ). 356, 1376–1379 (2017).
16. Simon, P. et al. References and Notes 1. 338, 1541–1545 (2011).
17. NREL. NREL. 18. Bryant, D. et al. Light and oxygen induced degradation limits
the operational stability of methylammonium lead triiodide perovskite solar cells. Energy Environ. Sci. 9, 1655–1660 (2016).
19. Chun-‐Ren Ke, J. et al. In situ investigation of degradation at organometal halide perovskite surfaces by X-‐ray photoelectron spectroscopy at realistic water vapour pressure. Chem. Commun. 53, 5231–5234 (2017).
20. Juarez-‐Perez, E. J., Hawash, Z., Raga, S. R., Ono, L. K. & Qi, Y. Thermal degradation of CH 3 NH 3 PbI 3 perovskite into NH 3 and CH 3 I gases observed by coupled thermogravimetry–mass spectrometry analysis. Energy Environ. Sci. 9, 3406–3410 (2016).
21. Yuan, Y. et al. Electric-‐field-‐driven reversible conversion between methylammonium lead triiodide perovskites and lead iodide at elevated temperatures. Adv. Energy Mater. 6, 1–7 (2016).
22. Guarnera, S. et al. Improving the long-‐term stability of perovskite solar cells with a porous Al2O3 buffer layer. J. Phys. Chem. Lett. 6, 432–437 (2015).
23. Rolston, N. et al. Mechanical integrity of solution-‐processed perovskite solar cells. Extrem. Mech. Lett. 9, 353–358 (2016).
24. Habisreutinger, S. N. et al. Carbon nanotube/polymer composites as a highly stable hole collection layer in perovskite solar cells. Nano Lett. 14, 5561–5568 (2014).
25. Van Noorden, R. Cheap solar cells tempt businesses. Nature 513, 470 (2014).
26. Noel, N. K. et al. Lead-‐free organic–inorganic tin halide perovskites for photovoltaic applications. Energy Environ. Sci.
135
7, 3061–3068 (2014). 27. Perovskites and Perovskite Solar Cells: An Introduction. at
<https://www.ossila.com/pages/perovskites-‐and-‐perovskite-‐solar-‐cells-‐an-‐introduction>
28. Nie, W. Y. et al. High-‐efficiency solution-‐processed perovskite solar cells with millimeter-‐scale grains. Science (80-‐. ). 347, 522–525 (2015).
29. Bredas, J. L., Calbert, J. P., da Silva Filho, D. A. & Cornil, J. Organic semiconductors: A theoretical characterization of the basic parameters governing charge transport. Proc. Natl. Acad. Sci. 99, 5804–5809 (2002).
30. Kymissis, I. Organic Field Effect Transistors. (2003). doi:10.1007/978-‐0-‐387-‐92134-‐1
31. Leonat, L., Sbârcea, G. & Bran̂zoi, I. V. Cyclic voltammetry for energy levels estimation of organic materials. UPB Sci. Bull. Ser. B Chem. Mater. Sci. 75, 111–118 (2013).
32. Brutting, W. & Rieß, W. Grundlagen der organischen Halbleiter. Phys. J. 7, 33–38 (2008).
33. Gregg, B. A. & Hanna, M. C. Comparing organic to inorganic photovoltaic cells: Theory, experiment, and simulation. J. Appl. Phys. 93, 3605–3614 (2003).
34. Torabi, S. et al. Strategy for enhancing the dielectric constant of organic semiconductors without sacrificing charge carrier mobility and solubility. Adv. Funct. Mater. 25, 150–157 (2015).
35. Deibel, C. & Dyakonov, V. Polymer-‐Fullerene Bulk Heterojunction Solar Cells. 096401, 68 (2010).
36. Vandewal, K. Interfacial Charge Transfer States in Condensed Phase Systems. Annu. Rev. Phys. Chem 67, 113–33 (2016).
37. Loi, M. A. et al. Charge transfer excitons in bulk heterojunctions of a polyfluorene copolymer and a fullerene derivative. Adv. Funct. Mater. 17, 2111–2116 (2007).
38. Vandewal, K. et al. The relation between open-‐circuit voltage and the onset of photocurrent generation by charge-‐transfer absorption in polymer: Fullerene bulk heterojunction solar cells. Adv. Funct. Mater. 18, 2064–2070 (2008).
39. Veldman, D. et al. Compositional and electric field dependence of the dissociation of charge transfer excitons in alternating polyfluorene copolymer/fullerene blends. J. Am. Chem. Soc. 130, 7721–7735 (2008).
40. Benson-‐Smith, J. J. et al. Formation of a ground-‐state charge-‐transfer complex in polyfluorene/[6,6]-‐ phenyl-‐C61 butyric
136
acid methyl ester (PCBM) blend films and its role in the function of polymer/PCBM solar cells. Adv. Funct. Mater. 17, 451–457 (2007).
41. Tvingstedt, K. et al. Electroluminescence from charge transfer states in polymer solar cells. J. Am. Chem. Soc. 131, 11819–11824 (2009).
42. Goris, L. et al. Observation of the subgap optical absorption in polymer-‐fullerene blend solar cells. Appl. Phys. Lett. 88, 1–3 (2006).
43. Clarke, T. M., Ballantyne, A. M., Nelson, J., Bradley, D. D. C. & Durrant, J. R. Free energy control of charge photogeneration in polythiophene/fullerene solar cells: The influence of thermal annealing on P3HT/PCBM blends. Adv. Funct. Mater. 18, 4029–4035 (2008).
44. Liu, Y., Zojer, K., Lassen, B., Kjelstrup-‐hansen, J. & Madsen, M. The Role of Charge Transfer State on the Reduced Langevin Recombination in Organic Solar Cells : A Theoretical Study. J. Phys. Chem. C in press, (2015).
45. Vandewal, K. et al. Efficient charge generation by relaxed charge-‐transfer states at organic interfaces. Nat. Mater. 13, 63–8 (2014).
46. Lee, J. et al. Charge transfer state versus hot exciton dissociation in polymer-‐fullerene blended solar cells. J. Am. Chem. Soc. 132, 11878–11880 (2010).
47. Vandewal, K., Tvingstedt, K., Gadisa, A., Inganäs, O. & Manca, J. V. On the origin of the open-‐circuit voltage of polymer–fullerene solar cells. Nat. Mater. 8, 904–909 (2009).
48. Alqurashi, R., Griffin, J., Alsulami, A. & Buckley, A. Open-‐Circuit Voltage in Inverted Polycarbazole:Fullerene Bulk Heterojunction Solar Cells. IEEE J. Photovoltaics 6, 918–923 (2016).
49. Marcus, R. A. Relation between charge transfer absorption and fluorescence spectra and the inverted region. J. Phys. Chem. 93, 3078–3086 (1989).
50. Brabec, C. J. et al. Tracing photoinduced electron transfer process in conjugated polymer/fullerene bulk heterojunctions in real time. Chem. Phys. Lett. 340, 232–236 (2001).
51. Gu, S., Neugebauer, H. & Sariciftci, N. S. Conjugated Polymer-‐Based Organic Solar Cells. 1324–1338 (2007).
52. Kietzke, T. Recent advances in organic solar cells. Adv. Optoelectron. 2007, (2007).
137
53. Nunzi, J.-‐M. Organic photovoltaic materials and devices. Comptes Rendus Phys. 3, 523–542 (2002).
54. Shrotriya, V., Yao, Y., Li, G. & Yang, Y. Effect of self-‐organization in polymer/fullerene bulk heterojunctions on solar cell performance. Appl. Phys. Lett. 89, 2–4 (2006).
55. Lee, J. H. et al. Enhanced solar-‐cell efficiency in bulk-‐heterojunction polymer systems obtained by nanoimprinting with commercially available AAO membrane filters. Small 5, 2139–2143 (2009).
56. Kyaw, A. K. K. et al. Improved Inverted Organic Solar Cells With a Sol–Gel Derived Indium-‐Doped Zinc Oxide Buffer Layer. IEEE J. Sel. Top. Quantum Electron. 16, 1700–1706 (2010).
57. Du, C. et al. Morphology and Performance of Polymer Solar Cell Characterized by DPD Simulation and Graph Theory. Sci. Rep. 5, 1–13 (2015).
58. Li, G. et al. ‘Solvent annealing’ effect in polymer solar cells based on poly(3-‐hexylthiophene) and methanofullerenes. Adv. Funct. Mater. 17, 1636–1644 (2007).
59. Chen, H.-‐Y. et al. Polymer solar cells with enhanced open-‐circuit voltage and efficiency. Nat. Photonics 3, 649–653 (2009).
60. Nam, S. et al. Inverted polymer fullerene solar cells exceeding 10% efficiency with poly(2-‐ethyl-‐2-‐oxazoline) nanodots on electron-‐collecting buffer layers. Nat. Commun. 6, 1–9 (2015).
61. Dey, S., Vivo, P., Efimov, A. & Lemmetyinen, H. Enhanced performance and stability of inverted organic solar cells by using novel zinc-‐benzothiazole complexes as anode buffer layers. J. Mater. Chem. 21, 15587 (2011).
62. Lee, S.-‐H., Seo, J.-‐W. & Lee, J.-‐Y. Stable inverted small molecular organic solar cells using a p-‐doped optical spacer. Nanoscale 7, 157–165 (2015).
63. Lan, J.-‐L. et al. The effects of Ta 2 O 5 –ZnO films as cathodic buffer layers in inverted polymer solar cells. J. Mater. Chem. A 2, 9361–9370 (2014).
64. Krebs, F. C., Gevorgyan, S. A. & Alstrup, J. A roll-‐to-‐roll process to flexible polymer solar cells: model studies, manufacture and operational stability studies. J. Mater. Chem. 19, 5442 (2009).
65. Cros, S. et al. Definition of encapsulation barrier requirements: A method applied to organic solar cells. Sol. Energy Mater. Sol. Cells 95, S65–S69 (2011).
66. Fujishima, D. et al. Organic thin-‐film solar cell employing a
138
novel electron-‐donor material. Sol. Energy Mater. Sol. Cells 93, 1029–1032 (2009).
67. Che, X., Xiao, X. & Forrest, S. R. Highly Efficient ( 11 . 1 %) Small Molecule Multi-‐junction Organic Photovoltaic Cells. 70, 140–142 (2014).
68. Hirade, M. & Adachi, C. Small molecular organic photovoltaic cells with exciton blocking layer at anode interface for improved device performance. Appl. Phys. Lett. 99, 2009–2012 (2011).
69. Yokoyama, D. et al. High-‐efficiency simple planar heterojunction organic thin-‐film photovoltaics with horizontally oriented amorphous donors. Sol. Energy Mater. Sol. Cells 98, 472–475 (2012).
70. Zheng, Y. et al. Highly efficient bulk heterojunction photovoltaic cells based on C 70 and tetraphenyldibenzoperiflanthene. Appl. Phys. Lett. 102, 143304 (2013).
71. Xiao, X., Zimmerman, J. D., Lassiter, B. E., Bergemann, K. J. & Forrest, S. R. A hybrid planar-‐mixed tetraphenyldibenzoperiflanthene/C70 photovoltaic cell. Appl. Phys. Lett. 102, (2013).
72. Wang, Z. et al. Highly efficient organic p-‐i-‐n photovoltaic cells based on tetraphenyldibenzoperiflanthene and fullerene C70. Energy Environ. Sci. 249–255 (2012). doi:10.1039/c2ee22952h
73. Hirade, M. & Adachi, C. Small molecular organic photovoltaic cells with exciton blocking layer at anode interface for improved device performance. Appl. Phys. Lett. 99, 4–7 (2011).
74. Wang, Z. et al. High fill factor and thermal stability of bilayer organic photovoltaic cells with an inverted structure. Appl. Phys. Lett. 106, 053305 (2015).
75. Peng, Y., Zhang, L. & Andrew, T. L. High open-‐circuit voltage, high fill factor single-‐junction organic solar cells. Appl. Phys. Lett. 105, 083304 (2014).
76. Horvath, H. Atmospheric light absorption -‐ a review. Atmos. Environ. Part A, Gen. Top. 27, 293–317 (1993).
77. Zhuang, T., Sano, T. & Kido, J. Efficient small molecule-‐based bulk heterojunction photovoltaic cells with reduced exciton quenching in fullerene. Org. Electron. physics, Mater. Appl. 26, 415–419 (2015).
78. Rand, B. P., Burk, D. P. & Forrest, S. R. Offset energies at organic
139
semiconductor heterojunctions and their influence on the open-‐circuit voltage of thin-‐film solar cells. Phys. Rev. B -‐ Condens. Matter Mater. Phys. 75, 1–11 (2007).
79. Kroto, H. W., Heath, J. R., O’Brien, S. C., Curl, R. F. & Smalley, R. E. C60: Buckminsterfullerene. Nature 318, 162–163 (1985).
80. Thirunavukkuarasu, K. et al. Rotational Dynamics in C 70 : Temperature-‐ and Pressure-‐Dependent Infrared Studies. J. Phys. Chem. C 115, 3646–3653 (2011).
81. Kröger, M. et al. P-‐type doping of organic wide band gap materials by transition metal oxides: A case-‐study on Molybdenum trioxide. Org. Electron. physics, Mater. Appl. 10, 932–938 (2009).
82. Meyer, J., Zilberberg, K., Riedl, T. & Kahn, A. Electronic structure of Vanadium pentoxide: An efficient hole injector for organic electronic materials. J. Appl. Phys. 110, (2011).
83. Kim, D. Y. et al. The effect of molybdenum oxide interlayer on organic photovoltaic cells. Appl. Phys. Lett. 95, 2007–2010 (2009).
84. Greiner, M. T. et al. Universal energy-‐level alignment of molecules on metal oxides. Nat. Mater. 11, 76–81 (2012).
85. Wang, F., Qiao, X., Xiong, T. & Ma, D. The role of molybdenum oxide as anode interfacial modification in the improvement of efficiency and stability in organic light-‐emitting diodes. Org. Electron. physics, Mater. Appl. 9, 985–993 (2008).
86. Peumans, P., Uchida, S. & Forrest, S. R. Efficient bulk heterojunction photovoltaic cells using small-‐molecular-‐weight organic thin films. Nature 425, 158–162 (2003).
87. Vogel, M., Doka, S., Breyer, C., Lux-‐Steiner, M. C. & Fostiropoulos, K. On the function of a bathocuproine buffer layer in organic photovoltaic cells. Appl. Phys. Lett. 89, 1–4 (2006).
88. Huang, J., Yu, J., Lin, H. & Jiang, Y. Detailed analysis of bathocuproine layer for organic solar cells based on copper phthalocyanine and C60. J. Appl. Phys. 105, 1–6 (2009).
89. Guerrero, A. et al. Shelf Life Degradation of Bulk Heterojunction Solar Cells: Intrinsic Evolution of Charge Transfer Complex. Adv. Energy Mater. 5, 1–8 (2015).
90. Guerrero, A. & Garcia-‐Belmonte, G. Recent advances to understand morphology stability of organic photovoltaics. Nano-‐Micro Lett. 9, (2017).
91. Yang, X., Duren, J. K. J. Van, Janssen, R. a J., Michels, M. a J. &
140
Loos, J. Morphology and Thermal Stability of the Active Layer in Poly (p -‐phenylenevinylene)/ Methanofullerene Plastic Photovoltaic Devices. Macromolecules 37, 2151–2158 (2004).
92. Zhao, J. et al. Phase Diagram of P3HT / PCBM Blends and Its Implication for the Stability of Morphology Phase Diagram of P3HT / PCBM Blends and Its Implication for the Stability of Morphology. J. Phys. Chem. B 113, 1587–1591 (2009).
93. Bertho, S. et al. Influence of thermal ageing on the stability of polymer bulk heterojunction solar cells. Sol. Energy Mater. Sol. Cells 91, 385–389 (2007).
94. Vandenbergh, J. et al. Thermal stability of poly[2-‐methoxy-‐5-‐(2′-‐phenylethoxy)-‐1,4-‐ phenylenevinylene] (MPE-‐PPV):Fullerene bulk heterojunction solar cells. Macromolecules 44, 8470–8478 (2011).
95. Ning, Y. et al. Investigation on Thermal Degradation Process of Polymer Solar Cells Based on Blend of PBDTTT-‐C and PC 70 BM. Int. J. Photoenergy 2014, 1–9 (2014).
96. Kesters, J. et al. Enhanced organic solar cell stability by polymer (PCPDTBT) side chain functionalization. Chem. Mater. 27, 1332–1341 (2015).
97. Turkovic, V. et al. Long-‐term stabilization of organic solar cells using UV absorbers. J. Phys. D. Appl. Phys. 49, (2016).
98. Turkovic, V. et al. Long-‐term stabilization of organic solar cells using hydroperoxide decomposers as additives. Appl. Phys. A Mater. Sci. Process. 122, 1–6 (2016).
99. Heumueller, T. et al. Morphological and electrical control of fullerene dimerization determines organic photovoltaic stability. Energy Environ. Sci. 9, 247–256 (2016).
100. Lloyd, M. T. et al. Impact of contact evolution on the shelf life of organic solar cells. J. Mater. Chem. 19, 7638 (2009).
101. Chand, P. K. and S. & Center. Recent progress and future aspects of organic solar cells. Prog. Photovoltaics Res. Appl. 20, 6–11 (2012).
102. Tavakkoli, M. et al. Progress in stability of organic solar cells exposed to air. Sol. Energy Mater. Sol. Cells 95, 1964–1969 (2011).
103. Clarke, T. M. et al. Photodegradation in encapsulated silole-‐based polymer: Pcbm solar cells investigated using transient absorption spectroscopy and charge extraction measurements. Adv. Energy Mater. 3, 1473–1483 (2013).
104. Jørgensen, M. et al. Stability of polymer solar cells. Adv. Mater.
141
24, 580–612 (2012). 105. Ahmad, J., Bazaka, K., Anderson, L. J., White, R. D. & Jacob, M. V.
Materials and methods for encapsulation of OPV: A review. Renew. Sustain. Energy Rev. 27, 104–117 (2013).
106. Mateker, W. R. et al. Molecular Packing and Arrangement Govern the Photo-‐Oxidative Stability of Organic Photovoltaic Materials. Chem. Mater. 27, 6345–6353 (2015).
107. Yan, C. et al. Non-‐fullerene acceptors for organic solar cells. Nat. Rev. Mater. 3, 1–19 (2018).
108. Cheng, Z. & Lin, J. Layered organic–inorganic hybrid perovskites: structure, optical properties, film preparation, patterning and templating engineering. CrystEngComm 12, 2646 (2010).
109. Liu, X., Ding, K., Panda, A. & Forrest, S. R. Charge Transfer States in Dilute Donor-‐Acceptor Blend Organic Heterojunctions. ACS Nano 10, 7619–7626 (2016).
110. Stranks, S. D. et al. Electron-‐Hole Diffusion Lengths Exceeding. Science 342, 341–344 (2014).
111. Leijtens, T. et al. Overcoming ultraviolet light instability of sensitized TiO2with meso-‐superstructured organometal tri-‐halide perovskite solar cells. Nat. Commun. 4, 1–8 (2013).
112. Bella, F. et al. Improving efficiency and stability of perovskite solar cells with photocurable fluoropolymers. Science (80-‐. ). 354, 203–206 (2016).
113. Watson, B. L., Rolston, N. J., Printz, A. D. & Dauskardt, R. H. Scaffold-‐Reinforced Perovskite Compound Solar Cells. Energy Environ. Sci. 10, 2500–2508 (2017).
114. Corazza, M., Krebs, F. C. & Gevorgyan, S. A. Lifetime of organic photovoltaics: Linking outdoor and indoor tests. Sol. Energy Mater. Sol. Cells 143, 467–472 (2015).
115. Kettle, J. et al. Using ISOS consensus test protocols for development of quantitative life test models in ageing of organic solar cells. Sol. Energy Mater. Sol. Cells 167, 53–59 (2017).
116. Gevorgyan, S. A., Jørgensen, M. & Krebs, F. C. A setup for studying stability and degradation of polymer solar cells. Sol. Energy Mater. Sol. Cells 92, 736–745 (2008).
117. Gevorgyan, S. A. et al. Improving, characterizing and predicting the lifetime of organic photovoltaics. J. Phys. D. Appl. Phys. 50, (2017).
118. Perez, M. D., Borek, C., Forrest, S. R. & Thompson, M. E.
142
Molecular and Morphological Influences on the Open Circuit. J. Am. Chem. Soc. 131, 9281–9286 (2009).
119. Graham, K. et al. Charge-‐transfer state energy determines open-‐circuit voltage in organic photovoltaics. SPIE Newsroom DOI: 10.1117/2.1201309.005006 (2013). doi:10.1117/2.1201309.005006
120. Elumalai, N. K. & Uddin, A. Open circuit voltage of organic solar cells: an in-‐depth review. Energy Environ. Sci. 9, 391–410 (2016).
121. Vandewal, K., Tvingstedt, K., Gadisa, A., Inganäs, O. & Manca, J. V. Relating the open-‐circuit voltage to interface molecular properties of donor:acceptor bulk heterojunction solar cells. Phys. Rev. B -‐ Condens. Matter Mater. Phys. 81, 1–8 (2010).
122. Gould, I. R. et al. Radiative and nonradiative electron transfer in contact radical-‐ion pairs. Chemical Physics 176, 439–456 (1993).
123. Sulas, D. B. et al. Open-‐Circuit Voltage Losses in Selenium-‐Substituted Organic Photovoltaic Devices from Increased Density of Charge-‐Transfer States. Chem. Mater. 27, 6583–6591 (2015).
124. Tang, Z. et al. A New Fullerene-‐Free Bulk-‐Heterojunction System for Efficient High-‐Voltage and High-‐Fill Factor Solution-‐Processed Organic Photovoltaics. Adv. Mater. 27, 1900–1907 (2015).
125. Graham, K. R. et al. Importance of the Donor:Fullerene Intermolecular Arrangement for High-‐E ffi ciency Organic Photovoltaics. J. Am. Chem. Soc. 136, 9608–9618 (2014).
126. Vandewal, K., Tvingstedt, K., Gadisa, A., Inganäs, O. & Manca, J. V. Relating the open-‐circuit voltage to interface molecular properties of donor:acceptor bulk heterojunction solar cells. Phys. Rev. B -‐ Condens. Matter Mater. Phys. 81, (2010).
127. Cluster. No Title. 128. Benduhn, J. Sensitive Spectroscopy of Charge-‐Transfer States
and Triplet Excited States in Organic Photovoltaics. (2015). 129. Mulligan, C. J. et al. A projection of commercial-‐scale organic
photovoltaic module costs. Sol. Energy Mater. Sol. Cells 120, 9–17 (2014).
130. Azzopardi, B. et al. Economic assessment of solar electricity production from organic-‐based photovoltaic modules in a domestic environment. Energy Environ. Sci. 4, 3741 (2011).
131. Ecker, B. et al. Degradation effects related to the hole transport
143
layer in organic solar cells. Adv. Funct. Mater. 21, 2705–2711 (2011).
132. Zou, Y. & Holmes, R. J. Correlation between the Open-‐Circuit Voltage and Charge Transfer State Energy in Organic Photovoltaic Cells. ACS Appl. Mater. Interfaces 7, 18306–18311 (2015).
133. Hao, X. et al. Novel cathode buffer layer of Ag-‐doped bathocuproine for small molecule organic solar cell with inverted structure. Org. Electron. physics, Mater. Appl. 15, 1773–1779 (2014).
134. Patil, B. R., Ahmadpour, M., Sherafatipour, G. & Qamar, T. Area dependent behavior of bathocuproine (BCP) as cathode interfacial layers in organic photovoltaic cells.
135. Galindo, S. et al. Influence of the density of states on the open-‐circuit voltage in small-‐molecule solar cells. Org. Electron. physics, Mater. Appl. 15, 2553–2560 (2014).
136. Hörmann, U. et al. Voc from a Morphology Point of View: the Influence of Molecular Orientation on the Open Circuit Voltage of Organic Planar Heterojunction Solar Cells. J. Phys. Chem. C 118, 26462–26470 (2014).
137. Vandewal, K., Tvingstedt, K., Manca, J. V. & Inganäs, O. Charge-‐Transfer States and Upper Limit of the Open-‐Circuit Voltage in Polymer:Fullerene Organic Solar Cells. IEEE J. Sel. Top. Quantum Electron. 16, 1676–1684 (2010).
138. Vandewal, K. et al. Increased open-‐circuit voltage of organic solar cells by reduced donor-‐acceptor interface area. Adv. Mater. 26, 3839–3843 (2014).
139. Zheng, Y. et al. Comparative investigation of molecular orientation and charge collection in highly efficient mixed heterojunctions based on three planar-‐shaped donors and C 70. J. Phys. D. Appl. Phys. 49, 465106 (2016).
140. Ran, N. A. et al. Impact of interfacial molecular orientation on radiative recombination and charge generation efficiency. Nat. Commun. 8, 1–9 (2017).
141. Fu, Y. T. et al. Structure and disorder in squaraine-‐C60 organic solar cells: A theoretical description of molecular packing and electronic coupling at the donor-‐acceptor interface. Adv. Funct. Mater. 24, 3790–3798 (2014).
142. Zimmerman, J. D. et al. Independent control of bulk and interfacial morphologies of small molecular weight organic heterojunction solar cells. Nano Lett. 12, 4366–4371 (2012).
144
143. Seemann, A. et al. Reversible and irreversible degradation of organic solar cell performance by oxygen. Sol. Energy 85, 1238–1249 (2011).
144. Schafferhans, J., Baumann, A., Wagenpfahl, A., Deibel, C. & Dyakonov, V. Oxygen doping of P3HT:PCBM blends: Influence on trap states, charge carrier mobility and solar cell performance. Org. Electron. physics, Mater. Appl. 11, 1693–1700 (2010).
145. Hauch, J. A., Schilinsky, P., Choulis, S. A., Rajoelson, S. & Brabec, C. J. The impact of water vapor transmission rate on the lifetime of flexible polymer solar cells. Appl. Phys. Lett. 93, (2008).
146. Kim, H. J., Lee, H. H. & Kim, J. J. Real time investigation of the interface between a P3HT:PCBM layer and an al electrode during thermal annealing. Macromol. Rapid Commun. 30, 1269–1273 (2009).
147. Cardinaletti, I. et al. Organic and perovskite solar cells for space applications.
148. Lee, C.-‐C. Degradation mechanism of organic photovoltaic devices with bathocuproine buffer layer. J. Photonics Energy 1, 011108 (2011).
149. Up-‐scaling inverted organic photovoltaic cells using bathocuproine (BCP) as cathode interfacial layers.
150. Yoshida, H. Electron Transport in Bathocuproine Interlayer in Organic Semiconductor Devices. J. Phys. Chem. C 119, 24459–24464 (2015).
151. Tong, X., Lassiter, B. E. & Forrest, S. R. Inverted organic photovoltaic cells with high open-‐circuit voltage. Org. Electron. physics, Mater. Appl. 11, 705–709 (2010).
152. Kim, J. et al. Overcoming the challenges of large-‐area high-‐efficiency perovskite solar cells. ACS Energy Lett. 2, 1978–1984 (2017).
153. Divitini, G. et al. In situ observation of heat-‐induced degradation of perovskite solar cells. Nat. Energy 1, 15012 (2016).
154. Zhao, X., Kim, H.-‐S., Seo, J.-‐Y. & Park, N.-‐G. Effect of Selective Contacts on the Thermal Stability of Perovskite Solar Cells. ACS Appl. Mater. Interfaces 9, 7148–7153 (2017).
155. Kim, N.-‐K. et al. Investigation of Thermally Induced Degradation in CH3NH3PbI3 Perovskite Solar Cells using In-‐situ Synchrotron Radiation Analysis. Sci. Rep. 7, 4645 (2017).
145
156. Gottesman, R. & Zaban, A. Perovskites for Photovoltaics in the Spotlight: Photoinduced Physical Changes and Their Implications. Acc. Chem. Res. 49, 320–329 (2016).
157. Lee, S. W. et al. UV Degradation and Recovery of Perovskite Solar Cells. Sci. Rep. 6, 1–10 (2016).
158. Bae, S. et al. Electric-‐Field-‐Induced Degradation of Methylammonium Lead Iodide Perovskite Solar Cells. J. Phys. Chem. Lett. 7, 3091–3096 (2016).
159. Yadav, P., Prochowicz, D., Alharbi, E. A., Zakeeruddin, S. M. & Grätzel, M. Intrinsic and interfacial kinetics of perovskite solar cells under photo and bias-‐induced degradation and recovery. J. Mater. Chem. C 5, 7799–7805 (2017).
160. Yang, J., Siempelkamp, B. D., Liu, D. & Kelly, T. L. Investigation of CH3NH3PbI3degradation rates and mechanisms in controlled humidity environments using in situ techniques. ACS Nano 9, 1955–1963 (2015).
161. Reyna, Y. et al. Performance and stability of mixed FAPbI3(0.85)MAPbBr3(0.15)halide perovskite solar cells under outdoor conditions and the effect of low light irradiation. Nano Energy 30, 570–579 (2016).
162. Li, X. et al. Outdoor Performance and Stability under Elevated Temperatures and Long-‐Term Light Soaking of Triple-‐Layer Mesoporous Perovskite Photovoltaics. Energy Technology 3, 551–555 (2015).
163. Nie, W. et al. Light-‐activated photocurrent degradation and self-‐healing in perovskite solar cells. Nat. Commun. 7, 1–9 (2016).
164. Domanski, K. et al. Migration of cations induces reversible performance losses over day/night cycling in perovskite solar cells. Energy Environ. Sci. 10, 604–613 (2017).
165. Bag, M. et al. Kinetics of Ion Transport in Perovskite Active Layers and Its Implications for Active Layer Stability. J. Am. Chem. Soc. 137, 13130–13137 (2015).
166. Hoke, E. T. et al. Reversible photo-‐induced trap formation in mixed-‐halide hybrid perovskites for photovoltaics. Chem. Sci. 6, 613–617 (2015).
167. Huang, F. et al. Fatigue behavior of planar CH3NH3PbI3perovskite solar cells revealed by light on/off diurnal cycling. Nano Energy 27, 509–514 (2016).
168. Liu, F. et al. Is Excess PbI2 Beneficial for Perovskite Solar Cell Performance? Adv. Energy Mater. 6, 1–9 (2016).
146
169. Reese, M. O. et al. Consensus stability testing protocols for organic photovoltaic materials and devices. Sol. Energy Mater. Sol. Cells 95, 1253–1267 (2011).
170. Turkovic, V. et al. Long-‐term stabilization of organic solar cells using hindered phenols as additives. ACS Appl. Mater. Interfaces 6, 18525–18537 (2014).
171. Reliability of Small Molecule Organic Photovoltaics with Electron-‐ Filtering Compound Buffer.
172. Khenkin, M. V. et al. Dynamics of Photoinduced Degradation of Perovskite Photovoltaics: From Reversible to Irreversible Processes. ACS Appl. Energy Mater. acsaem.7b00256 (2018). doi:10.1021/acsaem.7b00256
173. Park, B. W. et al. Chemical engineering of methylammonium lead iodide/bromide perovskites: Tuning of opto-‐electronic properties and photovoltaic performance. J. Mater. Chem. A 3, 21760–21771 (2015).
174. Yoon, S. J. et al. Tracking Iodide and Bromide Ion Segregation in Mixed Halide Lead Perovskites during Photoirradiation. ACS Energy Lett. 1, 290–296 (2016).
175. Eperon, G. E. et al. Formamidinium lead trihalide: A broadly tunable perovskite for efficient planar heterojunction solar cells. Energy Environ. Sci. 7, 982–988 (2014).
176. Jesper Jacobsson, T. et al. Exploration of the compositional space for mixed lead halogen perovskites for high efficiency solar cells. Energy Environ. Sci. 9, 1706–1724 (2016).
177. Khenkin, M. V. et al. Supporting information-‐Dynamics of Photoinduced Degradation of Perovskite Photovoltaics: From Reversible to Irreversible Processes. ACS Appl. Energy Mater. acsaem.7b00256 (2018). doi:10.1021/acsaem.7b00256
178. Yin, W. J., Shi, T. & Yan, Y. Unusual defect physics in CH<inf>3</inf>NH<inf>3</inf>PbI<inf>3</inf> perovskite solar cell absorber. Appl. Phys. Lett. 104, (2014).
179. Jacobs, D. A. et al. Hysteresis phenomena in perovskite solar cells: the many and varied effects of ionic accumulation. Phys. Chem. Chem. Phys. 19, 3094–3103 (2017).
180. Van Reenen, S., Kemerink, M. & Snaith, H. J. Modeling Anomalous Hysteresis in Perovskite Solar Cells. J. Phys. Chem. Lett. 6, 3808–3814 (2015).
181. Shao, S. et al. Efficient Perovskite Solar Cells over a Broad Temperature Window: The Role of the Charge Carrier Extraction. Advanced Energy Materials 7, (2017).
147
182. Zhao, C. et al. Revealing Underlying Processes Involved in Light Soaking Effects and Hysteresis Phenomena in Perovskite Solar Cells. Advanced Energy Materials 5, (2015).
183. Shao, S. et al. The Effect of the Microstructure on Trap-‐Assisted Recombination and Light Soaking Phenomenon in Hybrid Perovskite Solar Cells. Advanced Functional Materials 26, 8094–8102 (2016).
184. Peng, J., Sun, Y., Chen, Y., Yao, Y. & Liang, Z. Light and Thermally Induced Evolutional Charge Transport in CH 3 NH 3 PbI 3 Perovskite Solar Cells. ACS Energy Lett. 1, 1000–1006 (2016).
185. Lv, S. et al. One-‐step, solution-‐processed formamidinium lead trihalide (FAPbI (3−x) Cl x ) for mesoscopic perovskite–polymer solar cells. Phys. Chem. Chem. Phys. 16, 19206–19211 (2014).
186. Liu, C. et al. Hysteretic behavior upon light soaking in perovskite solar cells prepared via modified vapor-‐assisted solution process. ACS Appl. Mater. Interfaces 7, 9066–9071 (2015).
187. Du, M. H. Efficient carrier transport in halide perovskites: theoretical perspectives. J. Mater. Chem. A 2, 9091–9098 (2014).
188. Noh, J. H., Im, S. H., Heo, J. H., Mandal, T. N. & Seok, S. Il. Chemical management for colorful, efficient, and stable inorganic-‐organic hybrid nanostructured solar cells. Nano Lett. 13, 1764–1769 (2013).
189. Niu, G. et al. Study on the stability of CH 3 NH 3 PbI 3 films and the effect of post-‐modification by aluminum oxide in all-‐solid-‐state hybrid solar cells. J. Mater. Chem. A 2, 705–710 (2014).
190. Misra, R. K. et al. Temperature-‐ and component-‐dependent degradation of perovskite photovoltaic materials under concentrated sunlight. J. Phys. Chem. Lett. 6, 326–330 (2015).
191. Misra, R. K. et al. Effect of Halide Composition on the Photochemical Stability of Perovskite Photovoltaic Materials. ChemSusChem 9, 2572–2577 (2016).
192. Song, D. et al. Degradation of organometallic perovskite solar cells induced by trap states. Appl. Phys. Lett. 108, (2016).
193. Domanski, K. et al. Not All That Glitters Is Gold: Metal-‐Migration-‐Induced Degradation in Perovskite Solar Cells. ACS Nano 10, 6306–6314 (2016).
194. Abate, A. et al. Silolothiophene-‐linked triphenylamines as stable hole transporting materials for high efficiency perovskite solar cells. Energy Environ. Sci. 8, 2946–2953
148
(2015). 195. Zhao, Y. et al. Mobile-‐Ion-‐Induced Degradation of Organic
Hole-‐Selective Layers in Perovskite Solar Cells. J. Phys. Chem. C 121, 14517–14523 (2017).
196. Cacovich, S. et al. Gold and iodine diffusion in large area perovskite solar cells under illumination. Nanoscale 9, 4700–4706 (2017).
197. Akbulatov, A. F. et al. Effect of Electron-‐Transport Material on Light-‐Induced Degradation of Inverted Planar Junction Perovskite Solar Cells. Advanced Energy Materials 7, (2017).
198. Li, Z. et al. Extrinsic ion migration in perovskite solar cells. Energy Environ. Sci. 10, 1234–1242 (2017).
199. Khenkin, M. V. et al. Reconsidering Figures of Merit for the Performance and Stability of Perovskite Photovoltaics. Energy Environ. Sci. 739–743 (2018). doi:10.1039/C7EE02956J
200. Bryce, M. R. Current trends in tetrathiafulvalene chemistry: Towards increased dimensionality. J. Mater. Chem. 5, 1481–1496 (1995).
201. Jørgensen, T., Hansen, T. K. & Becher, J. Tetrathiafulvalenes as building-‐blocks in supramolecular chemistry. Chem. Soc. Rev. 23, 41–51 (1994).
202. Bryce, M. R. Tetrathiafulvalenes as π-‐electron donors for intramolecular charge-‐transfer materials. Adv. Mater. 11, 11–23 (1999).
203. Wudl, F., Wobschall, D. & Hufnagel, E. J. Electrical Conductivity by the Bis-‐1,3-‐dithiole-‐Bis-‐1,3-‐dithiolium System. J. Am. Chem. Soc. 94, 670–672 (1972).
204. Ferraris, J., Cowan, D. O., Walatka, V. & Perlstein, J. H. Electron Transfer in a New Highly Conducting Donor-‐Acceptor Complex. J. Am. Chem. Soc. 95, 948–949 (1973).
205. Lorcy, D. & Bellec, N. Dithiadiazafulvalenes: Promising precursors of molecular materials. Chem. Rev. 104, 5185–5202 (2004).
206. Rovira, C. Bis(ethylenethio)tetrathiafulvalene (BET-‐TTF) and related dissymmetrical electron donors: From the molecule to functional molecular materials and devices (OFETs). Chem. Rev. 104, 5289–5317 (2004).
207. Fabre, J. M. Synthesis strategies and chemistry of nonsymmetrically substituted tetrachalcogenafulvalenes. Chem. Rev. 104, 5133–5150 (2004).
208. Broggi, J. et al. Organic Electron Donors as Powerful Single-‐
149
Electron Transfer Reducing Agents in Organic Synthesis To cite this version : HAL Id : hal-‐01428063 Organic Electron Donors as Powerful Single-‐Electron Transfer Reducing Agents in Organic Synthesis. 0–30 (2017). doi:10.1002/anie.201209060/full>.<10.1002/anie.201209060>.
209. Jeppesen, J. O., Nielsen, M. B. & Becher, J. Tetrathiafulvalene cyclophanes and cage molecules. Chem. Rev. 104, 5115–5131 (2004).
210. Segura, J. L. & Martn, N. New concepts in tetrathiafulvalene chemistry. Angew. Chemie -‐ Int. Ed. 40, 1372–1409 (2001).
211. Liu, J. et al. A dopant-‐free hole-‐transporting material for efficient and stable perovskite solar cells. Energy Environ. Sci. 7, 2963 (2014).
212. Martín, N., Sánchez, L., Herranz, M. Á., Illescas, B. & Guldi, D. M. Electronic Communication in Tetrathiafulvalene (TTF)/C 60 Systems: Toward Molecular Solar Energy Conversion Materials? Acc. Chem. Res. 40, 1015–1024 (2007).
Front Cover Image: Echosolarhotel.com
150
151
Paper I
152
153
Paper II
154
155
Paper III
156
157
Paper IV