Influence of Microstructure on Thermo- and Photo-stability in

Organic Bulk-heterojunction Solar Cell

Einfluss der Mikrostruktur auf die Thermo- und Lichtstabilität in

organischen Bulk-Heterojunction Solarzellen

Der Technischen Fakultät

der Friedrich-Alexander-Universität

Erlangen-Nürnberg

zur

Erlangung des Doktorgrades Dr.-Ing.

vorgelegt von

Chaohong Zhang

aus Guangdong, China

I

Als Dissertation genehmigt

von der Technischen Fakultät

der Friedrich-Alexander-Universität Erlangen-Nürnberg

Tag der mündlichen Prüfung: 19.12.2017

Vorsitzender des Promotionsorgans: Prof. Dr.-Ing. Reinhard Lerch

1. Gutachter: Prof. Dr. Christoph J. Brabec

2. Gutachter: Prof. Dr. Hin-Lap Yip

I

Acknowledgements

Back in bachelor time, I thought I already knew what I want; however, during PhD, I realized

that what I’ve wanted is not what I’ve thought. And now it occurs to me that human beings

normally want what they don’t have. The right attitude towards life is to be happy with what

one has AFTER FIGHTING. So, now, after fighting hard, although I am not satisfied with

myself within PhD, I do really appreciate the opportunity of knowing all of you in Germany,

especially my beloved boyfriend Liang Chen.

I would like to thank Christoph, my dear dear supervisor; thank you for all the mind-opening

and eye-lightening discussion and ideas, thank you for always being so nice and supportive. I

remember that I was all the time so panic and went to Christoph, asking what I should do,

what should I do. He was always so nice, giving instructions, and relieved my worries.

I would like to thank Ning Li; thank you for all the inspiring and enjoyable discussions and

supports. There were so many times that I had no idea what to do with my “chaos” data, but

you could always click, click, and told me: see, you have a very good story line! You are the

person I know best at telling stories as I said in my first year way before we have those

pleasant cooperation.

I would like to thank Stephan Langer, Thomas Huemüller, Andres Osvet, Corina Winkler

who has helped and supported me a lot in discussion and in the laboratory.

I would like to thank the collaborators in Russian and in Erlangen.

I would like to thank the Chinese Scholarship Council for the PhD scholarship.

I would like to thank all the colleagues in i-MEET, ZAE and Energy Campus.

I would like to thank my family and friends for cheering me up whenever I feel down.

Greatly thank the amazing and exceptional PhD journey!

I was trying my best to find out the true truth. Now it occurs to me that the truth is

uncertainty. We could do things precisely; however, the reality is a matter of probability.

With science, we could get (infinitely) close to truth, but we are never there and perhaps

never will; however, this doesn’t bother me anymore.

II

III

Abstract

Currently, one of the biggest challenges of OPV being commercially competitive is to

perform consistently throughout its lifetime. Generally, aging stresses, such as light source,

temperature, and atmosphere, can induce degradation in all layers including electrodes,

interfaces, and active layers. Heat and white light induced intrinsic degradation is deemed to

be investigated and addressed with the highest priority. The investigation of the PhD program

concentrates on microstructural evolution of bulk-heterojunction (BHJ) active layer under

thermal- and/or photo- stress.

The first attempt is to employ polymeric fullerenes. A polystyrene based side-chain

polymeric fullerene is successfully synthesized and applied in solar cell. However, side-chain

polymeric fullerenes suffer from poor solubility which limits its role of acting as the main

acceptor in the photovoltaic application. Two main-chain polymeric fullerenes, PPC4 and

PPCBMB, are employed as additives in PCE11:PCBM based solar cells. With up to 8 wt%

addition, the efficiency of the ternary solar cells stay as high as the binary solar cells;

however, with 20 wt% addition, the photovoltaic performance slightly decreases; moreover,

the addition of the main-chain polymeric fullerenes fails to address the photo instability issue

of polymer:fullerene solar cells.

As increasing insights concerning the BHJ morphology are gained, here comes the strategy of

combining materials with good miscibility to overcome the thermal instability of

polymer:fullerene solar cells. We demonstrate that the low miscibility between PCBM and

pDPPT5-2 or PTB7-Th is one of the fundamental origins of the low thermal stability. On the

contrary, two novel fullerenes, PyF5 and FAP1, with a significantly higher chemical

compatibility are introduced to overcome these limitations. Further, it is observed that the

IV

miscibility between the donor and acceptor dominates the optimized acceptor:donor ratios in

polymer-fullerene BHJ systems.

Owing to the extremely poor miscibility, the highly efficient PCE11:PCBM solar cells suffer

from strong burn-in losses even under room temperature. We demonstrate that crystalline

properties of PCE11 are highly influenced by molecular weight and polydispersity;

furthermore, polymer-fullerene BHJ film morphology is largely affected by the crystalline

nature of polymers and evolves differently under external stresses, like heat or light, which

eventually results in solar cells with different burn-in loss and lifetime. A detailed analysis of

the energy and intensity dependence of light-induced burn-in degradation suggests that

photo-excited carriers do affect amorphous polymer segments in a similar way as thermal

stress does.

V

Zusammenfassung

Gegenwärtig besteht eine der größten Herausforderungen von OPV darin, während ihrer

gesamten Lebensdauer beständig zu arbeiten, damit sie kommerziell wettbewerbsfähig ist. Im

Allgemeinen können Belastungen, wie z. B. Licht, Temperatur und Atmosphäre, eine

Verschlechterung in allen Schichten induzieren, einschließlich der Elektroden, Grenzflächen

und aktiven Schichten. Wärme- und weißlichtinduzierte intrinsische Degradation müssen mit

höchster Priorität untersucht und gelöst werden. Die Untersuchungen dieser Dissertation

konzentrierten sich auf die mikrostrukturelle Veränderung der Bulk-Heterojunction(BHJ)

aktiven Schicht unter thermischer und / oder Photo- Belastung.

Der erste Versuch ist die Verwendung von polymeren Fullerenen. Ein Polystyrol-basiertes

Seitenkettenpolymer-Fulleren wird erfolgreich synthetisiert und in Solarzellen eingesetzt.

Seitenketten-polymerische Fullerene leiden jedoch unter schlechter Löslichkeit, die die

Funktion des Hauptakzeptors in der Photovoltaikanwendung begrenzt. Zwei Hauptketten-

Polymer-Fullerene, PPC4 und PPCBMB, werden als Additive in PCE11: PCBM-basierten

Solarzellen eingesetzt. Mit Beigabe von bis zu 8 Gew .-% bleibt der Wirkungsgrad der

ternären Solarzellen so hoch wie bei den binären Solarzellen. Mit einer Zugabe von 20

Gew .-% nimmt jedoch die Leistung leicht ab. Darüber hinaus wird durch die Zugabe der

polymerischen Hauptketten-Fullerene das Problem der Photoinstabilität von Polymer:

Fulleren-Solarzellen nicht gelöst.

Aus den zunehmenden Erkenntnissen über die BHJ-Morphologie erfolgt die Strategie der

Kombination von Materialien mit guter Mischbarkeit zu kombinieren, um die thermische

Instabilität von Polymer: Fulleren-Solarzellen zu überwinden. Wir zeigen, dass die geringe

Mischbarkeit zwischen PCBM und pDPPT5-2 oder PTB7-Th der Grund für die geringe

thermische Stabilität ist. Im Gegensatz dazu werden zwei neue Fullerene, PyF5 und FAP1,

mit einer signifikant höheren chemischen Kompatibilität eingeführt, um diese

Einschränkungen zu überwinden. Des Weiteren wird beobachtet, dass die Mischbarkeit

VI

zwischen Donor und Akzeptor das optimierale Verhältnis in Polymer-Fulleren Solarzellen

dominiert.

Aufgrund der extrem schlechten Mischbarkeit leiden die hocheffizienten PCE11: PCBM-

Solarzellen selbst bei Raumtemperatur unter starken Burn-in Verlusten. Wir zeigen, dass die

kristallinen Eigenschaften von PCE11 stark vom Molekulargewicht und der Polydispersität

beeinflusst werden. Darüber hinaus wird die Polymer-Fulleren BHJ-Filmmorphologie stark

von den kristallinen Eigenschaften der Polymeren beeinflusst und entwickelt sich unter

äußeren Belastungen wie Wärme und Licht unterschiedlich, was letztendlich zu Solarzellen

mit unterschiedlichem Burn-in Verlust und Lebensdauer führt. Eine detaillierte Analyse der

Energie- und Intensitätsabhängigkeit der lichtinduzierten Burn-in-Degradation deutet darauf

hin, dass photoangeregte Ladungsträger in ähnlicher Weise wie thermische Spannungen

amorphe Polymersegmente beeinflussen.

VII

Abbreviations

AFM Atomic Force Microscopy

Ag Silver

AIBN 2,2′-Azobis(2-methylpropionitrile)

Al Aluminum

AM Air Mass

BHJ Bulk-heterojucntion

Ca Calcium

CB Chlorobenzene

CT Charge-transfer

C60 Fullerene

C70 Fullerene with 70 carbon atoms

CDCl3 Deuterated chloroform

CH2Cl2 Dichloromethane

D-A Donor - Acceptor

D:A Donor : Acceptor

DAB Debye-Anderson-Brumberger

DOS density of state

DSC Differential scanning calorimetry

EHOMO Energy level of Highest Occupied Molecular Orbital

ELUMO Energy level of Lowest Unoccupied Molecular Orbital

EQE External quantum efficiency

ETL Electron transporting layer

FF Fill factor

FTIR Fourier-transform infrared spectroscopy

FWHM full width at half maximum

GISAXS Grazing-incidence small-angle scattering

GIWAXS Grazing-Incidence Wide-Angle X-ray Scattering

HOMO Highest occupied molecular orbital

HPLC high-performance liquid chromatography

VIII

HSP Hansen Solubility Parameter

HTL Hole transporting layer

ICBA Indene-C60 bisadduct

IDT Indacenodithiophene

IML Intermediate layer

IPCE Incident photon to current efficiency

IQE Internal quantum efficiency

IR Infrared

ISOS International Summits on OPV Stability

ITIC

3,9-bis(2-methylene-(3-(1,1-dicyanomethylene)-indanone))-

5,5,11,11-tetrakis(4-hexylphenyl)-dithieno[2,3-d:2’,3’-d’]-s-

indaceno[1,2-b:5,6-b’]dithiophene

ITO Indium tin oxide

IV Current-voltage

J-V Current density-voltage

LBG Low band gap

LiF Lithium fluoride

LS light scattering

LUMO Lowest unoccupied molecular orbital

MEH-PPV poly[2-methoxy-5-(2’-ethylhexyloxy)-p-

phenylene vinylene]

MDSC Temperature-modulated Differential scanning calorimetry

MoOX Molybdenum trioxide

MPMC methyl prop-2-yn-1-yl malonate C61

MPMCPS methyl prop-2-yn-1-yl malonate C61 loaded polystyrene

derivative

N2 Nitrogen

NaOH Sodium hydroxide

NMR Nuclear Magnetic Resonance spectroscopy

oLED Organic light-emitting diodes

oDCB 1,2-Dichlorobenzene

OPV Organic photovoltaic

IX

OSC Organic solar cell

OSCs Organic solar cells

OXCBA o-xylenyl C60 bisadduct

OXCMA o-xylenyl C60 mono-adduct

OXCTA o-xylenyl C60 trisadduct

P3HT Poly (3-hexylthiophene-2,5-diyl)

PCBS [6,6]-phenyl-C61-butyric acid styryl ester

PCBSD [6,6]-phenyl-C61-butyric acid styryl dendron ester

PCBM/

PC61BM [6,6]-Phenyl-C61-Butyric-acid-Methyl ester

PC71BM Phenyl-C71-Butyric-acid-Methyl ester

PC61BPF [6,6]-phenyl-C61 butyric acid pentafluorophenyl ester

PCDTBT Polymer[N-9’-heptadecanyl-2,7-carbazole-alt-5,5-

(4,7)-di-2-thienyl-2’,1’,3’-benzothiadiazole]

PCPDTBT Poly[2,6-(4,4-bis-(2-ethylhexyl)-4H-cyclopenta[2,1-b;3,4-b’]

dithiophene)-alt-4,7(2,1,3-benzothiadiazole)]

PCE Power conversion efficiency

PCE10

Poly[4,8-bis(5-(2-ethylhexyl)thiophen-2-yl)benzo[1,2-b;4,5-

b']dithiophene-2,6-diyl-alt-(4-(2-ethylhexyl)-3-fluorothieno[3,4-

b]thiophene-)-2-carboxylate-2-6-diyl)]

PCE11 Poly[(5,6-difluoro-2,1,3-benzothiadiazol-4,7-diyl)-alt-(3,3’’’-di(2-

octyldodecyl)-2,2’;5’,2’’;5’’,2’’’-quaterthiophen-5,5’’’-diyl)]

pDDP5T diketopyrrolopyrrole–quinquethiophene alternating copolymer

PDI Polydipersity index

or Perylene Diimide

PEDOT:PSS Poly(ethylenedioxythiophene):poly(styrene sulfonic

acid)

PEG Poly(ethylene glycol)

PffBT4T-2OD Poly[(5,6-difluoro-2,1,3-benzothiadiazol-4,7-diyl)-alt-(3,3’’’-di(2-

octyldodecyl)-2,2’;5’,2’’;5’’,2’’’-quaterthiophen-5,5’’’-diyl)]

photo-CELIV Photogenerated charge carrier extraction by linearly increasing

voltage

PIA Steady-state photoinduced absorption

X

PL Photoluminescence

PTB7 Poly((3-fluoro-2-[(2-ethylhexyl)carbonyl]thieno[3,4-

b]thiophenediyl))

PTB7-Th

Poly[4,8-bis(5-(2-ethylhexyl)thiophen-2-yl)benzo[1,2-b;4,5-

b']dithiophene-2,6-diyl-alt-(4-(2-ethylhexyl)-3-fluorothieno[3,4-

b]thiophene-)-2-carboxylate-2-6-diyl)]

PV photovoltaics

RMS root-mean-square

SCLC Space-charge-limited current

TAS Transient absorption spectroscopy

TEM Transmission electron microscopy

TMS Tetramethylsilane

TLC thin layer chromatography

TQ1 Poly[[2,3-bis(3-octyloxyphenyl)-5,8-quinoxalinediyl]-2,5-

thiophenediyl]

ZnO Zinc oxide

XI

Symbols

Eg energy bandgap

FF fill factor

I light intensity

Jl current density under illumination at 100 mW cm-2

Jmpp current density at the point of maximum power output

Jo reverse saturation current

Jph photocurrent density

JSC short-circuit current

k Boltzmann constant

L active layer thickness

Mn number average molecular weight

Mw weight average molecular weight

n refractive index of the medium

𝑛1 The number of molecules of species 1

NA Avogadro’s number

Pin power of the solar radiation incident on the cell

Pout power of output from the solar cells

PCE power conversion efficiency

q electronic charge

R Resistance

or ideal gas constant

RRMS root mean square roughness

T Transmission

or temperature

Tg glass transition temperature

U applied voltage

Veff effective voltage

Vmpp voltage at the point of maximum power output

VOC open-circuit voltage

XII

V0 compensation voltage

x12 Flory-Huggins interaction parameter

𝜒𝑏 binodal

𝜒𝑐 critical interaction parameter

𝜒𝑠 spinodal

λmax absorption maxima

σ sigma

∅𝑐 critical composition

ε dielectric constant

εA molar extinction coefficient

μ charge carrier mobility

θ contact angle

Δ activation energy

∆𝐴𝑚𝑚𝑚 Helmholtz free energy of mixing

∆𝑆𝑚𝑚𝑚 entropy change upon mixing

∆𝑆𝑚𝑚𝑚 entropy change upon mixing per lattice site

τ time

π pi

Ω the natural logarithm of the number of ways to arrange molecules on

the lattice (the number of states)

Ω1 the number of states of each molecule of species 1

XIII

Contents

Acknowledgements ....................................................................................................................................... I Abstract ....................................................................................................................................................... III Zusammenfassung ....................................................................................................................................... V Abbreviations ............................................................................................................................................ VII Symbols ....................................................................................................................................................... XI Contents .................................................................................................................................................... XIII Chapter 1 General introduction .................................................................................................................. 1

1.1 Motivation ........................................................................................................................................ 2 1.2 From conducting to photovoltaic ................................................................................................... 3 1.3 Challenges of OPV toward commercialization ............................................................................. 5 1.4 Aim and outline of this thesis ....................................................................................................... 10

Chapter 2 Theory ........................................................................................................................................ 13 2.1 Working principles of organic solar cells .................................................................................... 14 2.2 The Flory interaction parameter ................................................................................................. 17 2.3 The solubility parameter .............................................................................................................. 20 2.4 Thermodynamics of mixing and demixing ................................................................................. 24

Chapter 3 State of the art ........................................................................................................................... 31 3.1 Degradation of organic solar cells ............................................................................................... 32 3.2 Progress in improving thermal stability of active layer ............................................................. 37 3.3 Understanding white light induced loss mechanisms ................................................................ 40

Chapter 4 Materials and methods ............................................................................................................. 45 4.1 Materials ........................................................................................................................................ 46

4.1.1 Active layer materials ........................................................................................................ 46 4.1.2 Interface and electrode materials ..................................................................................... 47

4.2 Device preparation ........................................................................................................................ 48 4.3 Methods of characterization ........................................................................................................ 49

4.3.1 J-V characteristics.............................................................................................................. 49 4.3.2 SCLC ................................................................................................................................... 49 4.3.3 Absorption, PL &FTIR ...................................................................................................... 50 4.3.4 DSC and MDSC ................................................................................................................. 50 4.3.5 GIWAXS and GISAXS ...................................................................................................... 50 4.3.6 Lifetime characterization .................................................................................................. 52

Chapter 5 Analysis of the application of polymeric fullerene derivatives in OPV ................................ 53 5.1 Introduction ................................................................................................................................... 54 5.2 Materials and reagents ................................................................................................................. 54 5.3 Characterization equipment and instrument ............................................................................. 55 5.4 Synthesis of side-chain polymeric fullerene derivatives (MPMCPS) ....................................... 55 5.5 Application of MPMCPS in solar cell devices ............................................................................ 58 5.6 Application of main-chain polymeric fullerene derivatives in solar cell devices ..................... 64 5.7 Conclusion ..................................................................................................................................... 68

XIV

Chapter 6 Correlation between miscibility, thermal-stability and optoelectronic properties .............. 69 6.1 Introduction ................................................................................................................................... 70 6.2 Materials and device fabrication ................................................................................................. 70 6.3 Optimization of D: A ratio of BHJ devices.................................................................................. 71 6.4 Thermal stability of fullerene-based BHJ films ......................................................................... 76 6.5 Evolution of surface morphology of BHJ films .......................................................................... 82 6.6 Evolution of bulk morphology of BHJ films ............................................................................... 84 6.7 Thermal behavior of materials in pristine and in blends .......................................................... 87 6.8 Correlation of miscibility and thermal stability of BHJ films................................................... 90 6.9 Correlation of miscibility and optoelectronic properties of solar cells ..................................... 94 6.10 Conclusion ................................................................................................................................. 103

Chapter 7 Correlation of JSC burn-in losses and microstructure metastabilities in organic solar cells .................................................................................................................................................................... 105

7.1 Introduction ................................................................................................................................. 106 7.2 Materials and device fabrication ............................................................................................... 106 7.3 Three batches of PCE11 ............................................................................................................. 107 7.4 Thermal- and photo- stability .................................................................................................... 108 7.5 crystalline properties and thermal behaviors ........................................................................... 110 7.6 Film morphology under illumination ........................................................................................ 115 7.7 Stability of PCE11:PC71BM ....................................................................................................... 116 7.8 Evolution of heat- and light- induced morphology .................................................................. 118 7.9 Equivalence of heat- and light- induced degradation .............................................................. 121 7.10 Conclusion ................................................................................................................................. 129

Chapter 8 Summary and outlook ............................................................................................................ 131 8.1 Summary...................................................................................................................................... 132 8.2 Outlook ........................................................................................................................................ 133

Appendix A Curriculum Vitae ................................................................................................................. 135 Appendix B Publications and Presentations ........................................................................................... 137 Appendix C Bibliography ........................................................................................................................ 139

1

Chapter 1 General introduction

Abstract

General motivation is briefly described in this chapter as the beginning. The history of

organic photovoltaic is introduced from the discovery of semiconducting polymer to the

latest achievements. The major challenges of OPV toward commercialization are

discussed and analyzed in aspects of efficiency, cost, and lifetime. Finally, the aim and

outline of the thesis are summarized.

2

1.1 Motivation

The development of human society relies on energy which functionalizes a role as gas to a

car. Green energy like solar energy is believed to play important part in sustainable

development of natural environment and human society. The sunlight reaching the Earth’s

surface every year is almost 10 thousand times more than the total power consumed by

human around the world in one year, which suggests a high potential of solar energy being

the world’s primary energy source in the future.

In 1839, the photovoltaic effect was experimentally demonstrated for the first time by French

physicist Edmond Becquerel. Nowadays, the confirmed efficiency of single-junction

terrestrial cell measured under AM1.5 spectrum for Si crystalline cell already reach 26.7%,

GaAs thin film cell 28.8%, perovskite 19.7%, and organic cell 13%. [1, 2]

Among the PV technology, OPV though not yet commercially competing with silicon solar

cells, draws exceptional interests due to the fact that organic semiconducting materials

possess comprehensive advantages, like light weight, non-toxic properties, flexibility, and

narrow absorption, which enable potable application and window-integrated semitransparent

devices, and remove the limitation of appearance of the devices.[3-5]

On the downside, the stability of organic solar cells has been one of the biggest obstacles in

entering the competitive photovoltaic market.[4] Increased research interest and effort have

been devoted to investigate and address the lifetime limitation.[6-9] Degradation induced by

extrinsic factors, like oxygen, humidity, and impurities in materials, can be tremendously

suppressed by appropriate encapsulation and purification. However, degradation caused by

intrinsic factors, like accumulated heat and illumination, must be understood and addressed in

order to enter the photovoltaic market.

3

1.2 From conducting to photovoltaic

The Nobel Prize in Chemistry 2000 was awarded to Alan Heeger, Alan MacDiarmid and

Hideki Shirakawa for showing how plastic can be made to conduct electric current. Since

then, a new object, semiconductive plastic, is drawn attention to the public. As a matter of

fact, their discovery has radically changed scientists’ view of plastic as a material dating back

to 1977 with their publication in The Journal of Chemical Society, Chemical Communications

about Synthesis of electrically conducting organic polymers: Halogen derivatives of

polyacetylene (CH)n.

The pre-story before 1977 began when Shirakawa had not known MacDiarmid and Heeger

and succeeded in synthesizing silvery trans-polyacetylene and copper-colored cis-

polyacetylene; in another part of the world, MacDiarmid and Heeger were investigating a

metallic-looking film of the inorganic polymer sulphur nitride, (SN)x. Accidentally or

fatefully, MacDiarmid met Shirakawa during a coffee break at a seminar, heard about the

silvery organic polymer, and invited Shirakawa to the University of Pennsylvania.

MacDiarmid and Shirakawa modified polyacetylene by oxidation with iodine vapor; Heeger

proposed the measurement of conductivity of the iodine-doped trans-polyacetylene. Eureka!

There the discovery of the electrically conducting organic polymers. Since then the field has

grown tremendously.

There are two distinguished bonds within polymer molecules: sigma (σ) bonds and pi (π)

bonds. The sigma bonds are fixed and immobile while the pi electrons, though are also

relatively localized, are not as strongly bound as the sigma bonds. After conjugated polymers

can be doping by oxidation (electrons being removed) or reduction (electrons being inserted),

the electrons constituting the pi bonds can move quickly along the molecule chain by

applying an electrical field.

One of the first and brilliant applications of organic semiconducting polymers is organic

light-emitting diodes (oLEDs) which in principle is electroluminescence. A typical oLED

consists of conductive and transparent polymer as an electrode on one side, semiconductive

polymer in the middle, and metal electrode at the other end. The semiconductive polymer will

emit light when a voltage is applied between the electrodes. Other applications, like antistatic

4

treatment of photographic film, computer screens, as corrosion inhibitor, soon follow,

boosting more and more interests into the organic semiconductors and eventually claiming

the Noble Prize in 2000.

After the introduction of bulk-heterojunction to the polymer photovoltaic, enormous progress

has been made in every aspect of OPV technology. [6,6]-phenyl-C61-butyric acid methyl ester

(PC61BM) and its corresponding C70 analogue PC71BM have been the most prominent

acceptor materials over the last two decades, and achieved the efficiency milestones in

combination with various benchmark polymer donors, such as Poly[2-methoxy-5-(3′,7′-

dimethyloctyloxy)-1,4-phenylenevinylene] (MDMO-PPV),[10, 11] Poly(3-hexylthiophene-

2,5-diyl) (P3HT),[12] Poly[2,6-(4,4-bis-(2-ethylhexyl)-4H-cyclopenta [2,1-b;3,4-

b′]dithiophene)-alt-4,7(2,1,3-benzothiadiazole)] (PCPDTBT), [13] Poly((3-fluoro-2-[(2-

ethylhexyl)carbonyl]thieno[3,4-b]thiophenediyl)) (PTB7), [14] Poly[4,8-bis(5-(2-

ethylhexyl)thiophen-2-yl)benzo[1,2-b;4,5-b']dithiophene-2,6-diyl-alt-(4-(2-ethylhexyl)-3-

fluorothieno[3,4-b]thiophene-)-2-carboxylate-2-6-diyl)] (PTB7-Th) [15, 16] and Poly[(5,6-

difluoro-2,1,3-benzothiadiazol-4,7-diyl)-alt-(3,3’’’-di(2-octyldodecyl)-2,2’;5’,2’’;5’’,2’’’-

quaterthiophen-5,5’’’-diyl)] (PffBT4T-2OD). [17] Currently, the power conversion

efficiencies of OPV has exceeded 13% in single-junction solar cells,[2] reached 13.8% in

tandem devices,[18] and attained 9.7% in modules.[1] The unique advantages of OPV are the

high absorption coefficient, flexibility, light-weight and non-toxic properties of organic

semiconducting materials, allowing the freedom of form and the applications in potable

appliances and residential area.[3-5]

5

1.3 Challenges of OPV toward commercialization

Bulk-heterojunction organic photovoltaics based on conjugated organic semiconductor

nanocomposites have achieved overwhelming breakthroughs in both scientific and

technological developments comprehensively in the past two decades. The ideal object of the

field is producing highly-efficient, cost-effective, and long-term stable photovoltaic modules

constructed on large-scale flexible substrates via high-volume roll-to-roll printing processing.

There are mainly three challenges of OPV toward the commercial market: efficiency, cost,

and lifetime (Figure 1-1).

Figure 1-1 Challenges of OPV toward commercial application

Efficiency

The spotlight of OPV has been on improving the power conversion efficiency throughout the

whole history of the field. There are three major aspects for enhancing photovoltaic

performance: material design, optimization of film morphology, and device engineering.

The photo-active layers of organic solar cells typically consist of two components, an

electron donor and an electron acceptor. PCBM has been a superstar acceptor for the two

CommercialApplication

Lifetime > 10 years

6

decades and witnessed benchmark efficiencies with various polymers like MDMO-PPV,

P3HT, PCPDTBT, PTB7, PCE10 and PCE11. To better match of absorption to the solar

spectrum for improving light harvesting, the polymers evolve from high bandgap to medium

bandgap and low bandgap by backbone engineering; to better match the LUMO level of

PCBM for effective charge generation and maximization of open circuit voltage, the LUMO

and HOMO levels are carefully tuned via side-chain modification. The development of

acceptors also follows the similar rules. The disadvantages of PCBM concerning efficiency

are the poor absorption and mismatch LUMO level. Fullerene based acceptors were

developed with better absorption as well as more optimized energy level; non-fullerene

acceptors, like PDI, ITIC and IDTBR, jumped out with tremendously improved absorption

and fabulous energy levels, advancing organic solar cells efficiency towards 14%.

Film morphology also plays vital role in realizing high efficient solar cells. The very first and

ever breakthrough is the introduction of donor:acceptor bulk-heterojunction which stimulate

pioneering researchers to recognize the important roles of the film morphology which is

induced by spontaneous nanoscale phase separation of the donor:acceptor composites.

Further, it is discovered that film nanomorphology can be adjusted by employing different

solvents owing to the solubility and compatibility of semiconductors with solvents, which can

be rationalized and guided by the Flory-Huggins solution theory. Moreover, small amount of

solvent additives can surprisingly reconstruct the microstructure of BHJ and heighten

photovoltaic performance, which results from the selective solubility of additive to donor or

acceptor component, paving bicontinuous pathway and facilitating charge carrier

transporting. Furthermore, thermal annealing and solvent vapor annealing of deposited films

can reshape mixed donor:acceptor, pure donor and acceptor phases, and consequently raise

solar cell efficiency. Additionally, a third photo-active component could also revise film

morphology, enhancing voltage or current density or fill factor and ultimately PCE.

Device engineering is another step acquiring high-performance organic solar cells. Lowering

metal electrode work function by inserting selectively hole/electron transporting layers

between active layer and electrodes is an innovate action, which tremendously strengthens

charge collecting efficiency and therefore boosts efficiency. By replacing conventional

device structure with inverted structure turn out work for certain donor-acceptor systems,

7

which reduce charge recombination and accordingly enhance photovoltaic performance.

Fabricating tandem device is also a helpful method to broaden absorption and deliver high-

performance solar cells.

Cost

Low cost of renewable PV devices is one of initial motivations into the OPV technology

which is compatible to solution processability as well as continuous roll-to-roll printing

techniques. However, to really realize cost-effective large-scale processing, there are still

quite some technical obstacles.

The highly resistive existing transparent electrodes would cause energy loss by Joule heating,

Ohmic loss.[19-22] The popular method to overcome the Ohmic loss in scaling up the sizes

of organic solar cell on an industrial scale is to fabricate monolithic modules comprising

serially connected stripe-patterned sub-cells with widths which are narrow enough to

regardless the sheet resistance of the transparent electrodes. Nevertheless, here comes a

derivative loss referred to as aperture loss which is attributed to the blank spaces between

stripe-patterned sub-cells and series connection regions of the sub-cells.[20, 23] Moreover,

additional undesired area losses in the module construction result from the poor patterning

resolution of conventional printing techniques. Millimeter-scale gaps between the stripe-

patterned sub-cells are required in the module to prevent intermixing between separately

ejected or transferred inks from the printing machines to the target substrates. Low geometric

fill factors, which is the ratio between the photoactive area and the total are, resulting from

the area loss cause big cut-down in module efficiency. Undesirable coating effects, like

pinholes, coffee rings and poor wettability would hugely lessen module efficiency when

printing the designed modules.[24-29] Therefore, for a successful transition from lab-scale

solar cells to commercial-scale printed modules, synergistic development of module

architectures and printing technologies require more research efforts.

The Ohmic and aperture losses can be significantly reduced by minimizing both stripe width

and the distance between stripes; Patterning definition of printing machines should be

considered when producing well-defined stripe patterns of interface layers, photoactive layers

and electrodes; The formation of homogeneous and pinhole-free thin films, which requires a

sophisticated printing system, is vital to fabricate the designed modules on a large scale; An

8

intensive cooperation with engineers who are familiar with roll-to-roll printing technologies

shall begin in order to enhance film quality and acquire good organic solar cell modules.

Lifetime

Current organic solar cells normally contain metal (e.g. Ag and Al) / metal oxide (e.g. indium

tin oxide, ITO) electrodes, semiconducting hole and electron transporting layers (e.g.

poly(3,4-ethylenedioxythiophene):poly(styrenesulfonate) (PEDOT:PSS), MoOx and ZnO),

and blends of organic semiconductors as the bulk-heterojunction (BHJ) active layer. Failures

in either layer translate into photovoltaic performance losses.[7, 9, 30] There are many factors

causing degradation of OPV devices, e.g. oxygen, humidity, impurities in semiconductors,

heat and light. Oxygen and water can induce chemical reactions of the active layer

component[31], and corrode certain interface materials[32, 33] (PEDOT:PSS, etc.) and

electrode materials (Ca, Al)[34, 35]. While long-time degradation induced by oxygen and

humidity is typically addressed by proper encapsulation, photo-oxidation in the presence of

residual traces of oxygen and light inducing photolytic and photo-chemical reactions of active

layer materials are more difficult to isolate due to accelerated kinetics.[36] Proper controlling

the residual amounts of oxygen and water in degradation experiments identified white light

illumination as the cause for a distinct degradation mechanism, which was later-on termed as

burn-in degradation originating from fullerene dimerization. [37], [38] We recently reported

burn-in like behavior in various high performance BHJ composites which could not be

explained by fullerene dimerization. [39, 40] Instead, microstructural instabilities arising

from a too large interaction parameter between the polymer and the fullerene were identified

as a further dominant degradation mechanism. In this PhD program we demonstrate the

equivalence of heat and white light in relaxing metastable bulk-heterojunction composites

towards their thermodynamic equilibrium states. Thermodynamic considerations suggest that

the equivalent impact of light and heat on metastable microstructures would allow

substituting long time photo-degradation trials by rather short time thermal degradation trials

without the loss of knowledge.

As the working temperature of solar cells is typically far below the decomposition

temperature of most photovoltaic materials, degradation owing to microscopic morphology

changes are dominantly induced by heat. Considerable methods have been promoted to

9

alleviate the thermal instability issue of polymer-fullerene solar cells. There are principally

two strategies in tackling the undesirable phase separation challenge: 1) introducing chemical

locking, which was stimulated by the strong phase separation of PCBM based solar

cells.[6],[41, 42],[43],[44], [45] 2) employing organic semiconductors with proper

miscibility.[46-48] As developing insights are gained about the microstructural phases of the

BHJ, more and more evidences imply that owing to the poor miscibility between the donors

and PCBM, the polymer:fullerene mixed regions turn out to be meta-stable, thus, causing the

exponential degradation (burn-in losses) of OPV devices upon heating.

Comparing to the straightforward thermal instability, the loss mechanisms induced upon

white light illumination are rather complicated and have not been fully understood.[49]

Research on light induced degradation is predominantly focused on open-circuit voltage

(VOC) losses and/or short circuit current density (JSC) losses. Adachi et al demonstrated that

accumulating charge carriers in the trap sites at the interfacial region to a P3HT:PCBM active

layer leads to deteriorated VOC.[50] Heumueller et al observed that a redistribution of charge

carriers in a broader density of states in amorphous polymers is the origin of the

photoinduced VOC losses, and that semi-crystalline materials are more stable against VOC

losses during white light illumination.[51, 52] Fullerene dimerization in the amorphous

region has been proven to be a reason for the JSC burn-in loss in C60-based fullerene solar

cells. [38, 53, 54]

The initial period of exponentially fast degradation, the burn-in degradation, is particularly

harmful as it is normally a 10-50% loss of initial performance in a short time.[49] The heat-

induced JSC burn-in loss is typically attributed to rearrangements of the composite

materials[41, 42, 44, 45] towards their thermodynamic equilibrium.[40] On the other hand,

the photo-induced JSC burn-in degradation so far was discussed in terms of fullerene

dimerization in certain C60-based fullerene solar cells.[38, 53, 54]. We complement that

insight by demonstrating that light-induced stress furthermore can transform metastable

microstructures towards their local thermodynamic equilibria, though on a much slower

timescale compared to thermal heating. Nevertheless, both ageing conditions do result in

identical microstructure changes.

10

1.4 Aim and outline of this thesis

It is widely observed that organic solar cells degrade over time. However, to be commercially

competitive, OPV is required to perform consistently throughout its lifetime. In principle,

aging stresses, such as light source, temperature, and atmosphere, can induce degradation in

all layers including electrodes, interfaces, and active layers. In order to understand and

address the instability issue, it is necessary to control aging conditions, limit loss

mechanisms, and focus on the dominant degrading layer at a time. Intrinsic degradation, such

as heat- and white light- induced degradation, should be investigated and addressed with the

highest priority. Therefore, the focusing investigation of the thesis is on thermal- and photo-

stability of bulk-heterojunction active layer concerning microstructural evolution. While

previous strategies to tackle the thermal instability issue concentrate on chemical locking in

this thesis we demonstrate that combination of materials with good miscibility is the essential

criteria for enhanced thermal stability, almost independent on the crystallinity of the

fullerenes. Another highlight of the thesis is the demonstration that the microstructural

changes induced by either thermal- or light-aging can be kinetically correlated to each other.

Our findings suggest that both, heat and light, are fundamentally equivalent in relaxing

metastable bulk-heterojunction composites towards their thermodynamic equilibrium states.

In Chapter 1, General motivation is briefly described. The history of organic photovoltaic is

introduced from the discovery of semiconducting polymer to the latest achievements. The

major challenges of OPV toward commercialization are discussed and analyzed in aspects of

efficiency, cost, and lifetime.

In Chapter 2, some basic theories employed in the thesis are briefly introduced. First, solar

cell structures, working principles, loss mechanisms and characterization are discussed.

Second, the Flory interaction parameters and Hansen Solubility Parameters are presented in

order to understand the thermodynamics of polymer:fullerene mixing and demixing behaviors.

In Chapter 3, the degradation stresses of organic solar cells are summarized and the general

loss mechanisms are elucidated in category of degradation of electrodes, hole/electron

11

transporting layers and active layer. Further, strategies and methods to overcome the intrinsic

degradation induced by heat are explicated and analyzed. In the end, progress in

understanding the photo-degradation by white light illumination is interpreted and discussed.

In Chapter 4, materials used in the thesis are described in terms of photoactive materials,

interface and electrode materials; further device structure and device preparation are

summarized; Finally, each characterization technique is briefly introduced.

Experimental results of the application of polymeric fullerene are described in Chapter 5. A

polystyrene based side-chain polymeric fullerenes was successfully synthesized and applied

in P3HT based solar cell. However, limiting by the strong aggregation and extremely low

solubility in common organic solvents, MPMCPS based binary and ternary solar cells fail to

achieve desirable performance. Two main-chain polymeric fullerenes, PPC4 and PPCBMB,

were employed as additives in PCE11:PCBM based solar cells. The photovoltaic

performance and photo-stability of the ternary solar cells were investigated.

In Chapter 6, we demonstrate binary organic solar cells based on PTB7-Th:fullerene and

pDPP5T-2:fullerene composites with decent photovoltaic performance and extraordinary

high thermal stability. We further in-depth investigate the carrier dynamics along with

structural evolution and analyze the acceptor loadings in optimized bulk-heterojunction

(BHJ) solar cells as a function of the polymer-fullerene miscibility. The polymer-fullerene

miscibility has more influential effects than crystallinity of single components on the

optimized acceptor:donor ratio in polymer-fullerene solar cells. The findings demonstrated in

this chapter suggest that the balance between the miscibility of the BHJ composites and their

optoelectronic properties has to be carefully considered for future development and

optimization of OPV solar cells based on the BHJ composites.

In Chapter 7, we demonstrate that the JSC burn-in loss can be either induced by thermal

annealing or white light illumination. Detailed microstructure studies confirm that demixing

of fullerenes from polymers in the amorphous regime is the primary mechanism for both

degradation conditions, although their kinetics is distinctly different. Both, light and heat,

12

provide enough energy to metastable bulk-heterojunction regimes and relax them into their

thermodynamic equilibrium, which typically is larger scale phase separated due to a positive

interaction energy. Notably, the microstructural changes induced by either thermal- or light-

aging can be kinetically correlated to each other. Similar to the phenomena that higher

temperature initiates faster degradation towards equilibrium, more intense light does as well

cause faster degrading.

13

Chapter 2 Theory

Abstract

In this chapter, some basic theories employed in the thesis are briefly introduced. First,

solar cell structures, working principles, loss mechanisms and characterization are

discussed. Second, the Flory interaction parameters and Hansen Solubility Parameters are

presented in order to understand the thermodynamics of polymer:fullerene mixing and

demixing behaviors.

14

2.1 Working principles of organic solar cells

Organic solar cells are devices that can convert light energy to electric energy through

organic photovoltaic materials. The structure of an organic solar cell contains: metal or metal

oxide act as electrodes; organic or inorganic materials functionalize as selectively

hole/electron transporting layers; organic semi-conductors work as the active layers. The

organic active layers normally consist of two semiconducting components, a low ionization-

potential material as donor and a high electron-affinity material as acceptor.

Semiconducting materials typically comprise of alternating single and double bonds of

Carbon atoms, forming the conjugated chemical structure. The configuration is the sp2-

hybridized orbitals. In sp2-hybridization, σ-bonds are formed in one plane by the overlap of

the sp2-orbitals between the carbon-carbon atoms, and the un-hybridized pz-orbitals form a π

bond between two carbon atoms, thus the produce of a double bond. The delocalized

electrons of the π orbital contribute to semiconducting properties. While the sp2-orbitals of

two carbon atoms overlap, bonding and anti-bonding states are formed. The filled π state with

highest energy is called the highest occupied molecular orbital (HOMO) while the empty π*

state with the lowest energy called the lowest unoccupied molecular orbital (LUMO). The

energy difference between the HOMO and LUMO level is the energy bandgap (Eg). In

modern organic solar cells, the donor and acceptor form bulk-heterojunction (BHJ) which can

maximize the interfacial area of the donor and acceptor domains.[55, 56]

The process of converting light energy to electric energy can be simplified as following

steps:[57]

a) Generation of excitons (electron-hole pairs) by absorbing photon;

b) Dissociation of excitons and formation of Coulomb-bound charge carriers (charge

transfer state) at the donor-acceptor interface;

15

c) Separation of Coulomb-bound charge carriers into free charges (hole and electron)

due to built-in electric field;

d) Transportation of holes and electrons through the active materials;

e) Collection of charge carriers to the electrodes.

The efficiency and recombination of every step are as following:[58-69]

The electron in the HOMO can be excited by photon with enough energy to LUMO, which is

the process called generation of excitons. Generally, step a) is limited firstly by the absorption

properties of semi-conductors. The absorption profile of the employed materials shall match

with the solar spectrum; further, the materials are supposed to possess high absorb

coefficient. Secondly, the film thickness of the active layers could also play important role on

the amount of excitons and the ultimate current density. This limitation can be characterized

by UV-vis absorption and External quantum efficiency (EQE). Thirdly, the excitons would

decay, meaning the electrons go back the HOMO without generating charges. This step

happens in the timescale of femtosecond region.

Typically, step b) and c) happen in femtosecond to picosecond timescale. The major losses

from these steps come from the geminate/monomolecular recombination. Owing to the low

timescale, this process is difficult to track and ultrafast techniques like transient absorption

spectroscopy (TAS) are necessary.

Step d) occurs at a relatively lower timescale of microsecond and has been understood

mostly. Loss mechanisms from this step include non-geminate/bimolecular recombination

and trap-assisted recombination. One of the effective methods to investigate these loss

mechanisms is the study of the current-voltage characteristics as a function of the light

intensity. Another method is to characterize the hole/electron mobility of the active layer by

means of space charge limited current (SCLC) or photo-CELIV.

16

Step e) was largely limited by the mismatch work function between the electrodes and the

photoactive layers in the early years, and has been greatly addressed by introducing hole- and

electron- selective transporting layers.

The final photovoltaic efficiency of the solar cell is determined by the above process. Each

step has its own limitations and loss mechanisms due to material shortcomings or device

imperfections.

The photovoltaic performance of a solar cell is characterized by:

𝑃𝑃𝑃 = 𝑃𝑜𝑜𝑜𝑃𝑖𝑖

= 𝐽𝑠𝑠× 𝑉𝑂𝑂 ×𝐹𝐹𝑃𝑖𝑖

(2-1)

where PCE is the power conversion efficiency, Pout and P in are the output and input powers,

respectively.

1) Short circuit current density (JSC)

JSC is defined as the current density running through the solar cells when the externally

applied voltage is 0.

2) Open circuit voltage (VOC)

VOC can be measured when the circuit current density is 0.

3) Fill factor (FF)

FF is defined as the ratio of the maximum output power to the value of 𝐽𝑠𝑐 × 𝑉𝑂𝑂:

𝐹𝐹 = 𝐽𝑚𝑚𝑚× 𝑉𝑚𝑚𝑚

𝐽𝑠𝑠× 𝑉𝑂𝑂 (2-2)

where Jmpp and Vmpp are the current density and voltage at the point of maximum power

output, respectively. FF represents dependence of current output on the internal field of

device, and is related to the series resistance and parallel resistance.

17

2.2 The Flory interaction parameter

Polymer solutions are mixtures of macromolecules with the low molar mass solvent (small

molecule). Interactions between different species can be either attractive or repulsive. [70]

In a polymer solution lattice, the lattice sites v0 are normally of the order of monomer sizes or

solvent sizes. One polymer molecule has molecular volume

v1 = N1 v0 (2-3)

and solvent/ small molecule has molecular volume

v2 = N2 v0 (2-4)

where N1 and N2 are the numbers of lattice sites occupied by a polymer molecule (N1=N>>1)

and a solvent molecule (N2=1).

In the mixtures, the volume fraction of polymer and solvent are

∅1 = 𝑉1𝑉1+𝑉2

= ∅ (2-5)

∅2 = 𝑉2𝑉1+𝑉2

= 1 − ∅ (2-6)

In regular solution theory, the energy of mixing is written in terms of three pairwise

interaction energies (u11, u12, and u22) between their first neighbors, with u11/ u22 meaning

interaction energy within two monomers of species 1/ species 2, u12 meaning interaction

energy between monomers of species 1 and species 2 (in the case, solvent or small

molecules).

In terms of species 1, the probability of a neighbor lattice occupying by species 1 is assumed

to be the volume fraction ∅1 = ∅; likewise, the probability of a neighbor lattice occupying

by species 2 is ∅2 = 1 − ∅; Each lattice site of a regular lattice has z nearest neighbors (z is

18

the coordination number of the lattice, for example, z = 4 for a square lattice and z = 6 a cubic

lattice); The number of lattice sizes occupied by species 1 is 𝑛∅ (n is the total number of

lattice sites in the polymer solution). In terms of species 2, it is similar. In addition, every

pairwise interaction is counted twice, once for the monomer in question and once for its

neighbor. Summing all the interaction gives the total interaction energy of the mixture:

𝑈 = 12

𝑛∅𝑧[𝑢11∅ + 𝑢12(1 − ∅)] + 𝑛(1 − ∅)𝑧[𝑢22 (1 − ∅) + 𝑢12∅]

= 𝑧𝑛2

[𝑢11∅2 + 2𝑢12(1 − ∅) + 𝑢22(1 − ∅)2] (2 − 7)

Before mixing, each species is only surrounded by itself. Ignoring the boundary effects (for

most macroscopic systems, the surface-to-volume ratio is very small), the total energy of both

species before mixing is the sum of the energies of the two pure components:

𝑈0 = 12

[𝑛∅𝑧𝑢11 + 𝑛(1 − ∅)𝑧𝑢22]

= 𝑛𝑛2

[𝑢11∅ + 𝑢22(1 − ∅)] (2-8)

The energy change upon mixing is:

∆𝐻 = ∆𝑈 = 𝑈 − 𝑈0 = 𝑛𝑛2

∅(1 − ∅)(2𝑢12 − 𝑢11 − 𝑢22) (2-9)

Because the volume is assumed to be constant, so ∆𝐻 = ∆𝑈.

However, it is more convenient to study the intensive property, which is the energy change

upon mixing per site:

∆𝐻𝑚𝑚𝑚 = 𝑈−𝑈0𝑛

= 𝑛2

∅(1 − ∅)(2𝑢12 − 𝑢11 − 𝑢22) = χ∅(1 − ∅)𝑘𝑘 (2-10)

A dimensionless measure of the differences in the strength of pairwise interaction energies

between species, the Flory interaction parameter, is defined:[70]

19

χ ≡ 𝑛2

(2𝑢12− 𝑢11−𝑢22)𝑘𝑘

= ∆𝐻𝑚𝑖𝑚∅(1−∅)𝑘𝑘

= ∆𝐻𝑚𝑖𝑚∅(1−∅)𝑅𝑘

(2-11)

As the Flory interaction parameter is defined in terms of energies per site, it is proportional to

the site volume v0. Therefore, the site volume must be specified whenever the interaction

parameter is discussed; the v0 are normally of the order of monomer sizes; to compare

various interaction parameters, the site volume of the compared interaction parameters have

to be the same.

If the interaction parameter of a binary mixture is small, it implies that the energy change

upon mixing is small; thus the pairwise interaction energy between the two species is similar

to the pairwise interaction energy within species. Therefore, the interaction parameter can act

as an index of miscibility between species. Systems with smaller interaction parameter have

better miscibility.

In the Flory-Huggins theory, one of the essential assumptions is that the volume stays

constant upon mixing. However, in most real polymer solution/blends, there is volume

change upon mixing. Some monomers may pack more condensed with other monomers.

Thus, the interaction parameter shows certain degree of composition-, chain length-, and

temperature- dependence. Till now, these effects are not fully understood yet. This

temperature- dependent of the Flory interaction parameter is empirically written as:

χ(𝑘) ≅ 𝐴 + 𝐵𝑘 (2-12)

where A is referred to the entropic part of the interaction parameter; B/T is the enthalpic part.

Furthermore, the parameters A and B are found to depend weakly on chain lengths and

composition. The interaction parameter usually contains the shortcomings of the Flory-

Huggins theory, where the reality does not match the ideality. With all the corrections

combined in interaction parameter, the Flory-Huggins equation includes all the

thermodynamic information needed to determine the equilibrium state of a mixture and

whether any metastable states are possible.

20

2.3 The solubility parameter

The Flory interaction parameter is an important parameter of a given mixture. One of the

methods to estimate the interaction parameter is developed by Hildebrand and Scott. It is

based on the solubility parameter related to the energy of vaporization of a molecule:[71]

δ1 ≡ ∆𝐸1𝑣1

(2-13)

where δ1, 𝑣1, and ∆𝑃1 are the solubility parameter or Hildebrand parameter, the volume,

and the energy of vaporization of species 1, respectively. The solubility parameter was

intended for nonpolar, non-associating systems. The dimension of solubility parameter is (cal

cm-3)0.5 = 2.046 × 103 (J m-3)0.5 = 2.046 MPa 0.5.

The energy of vaporization is the energy needed to remove a molecule from its pure state. ∆𝐸1𝑣1

is the cohesive energy density and the interaction energy per unit volume between the

molecules in the pure state. Thus, the interaction energy per site in the pure state of species 1

(zu11/2) is related to the solubility parameter:

−𝑛𝑢112

= 𝑣0∆𝐸1𝑣1

= 𝑣0𝛿12 (2-14)

Note that the minus sign is owing to the fact that the energy of vaporization is defined to be

positive, while the interaction energy is negative. Likewise, the interaction energy per site in

the pure state is:

−𝑛𝑢222

= 𝑣0∆𝐸2𝑣2

= 𝑣0𝛿22 (2-15)

The cohesive energy density of interaction between molecules of species 1 and 2 is estimated

from the geometric mean approximation:

−𝑛𝑢122

= 𝑣0∆𝐸1𝑣1

∆𝐸2𝑣2

= 𝑣0𝛿1𝛿2 (2-16)

21

By substituting Eqs (2-14) – (2-16) into the definition of the Flory interaction parameter,

comes the relation between interaction parameter and solubility parameter:[72]

χ12 ≈ 𝑣0𝑘𝑘

(𝛿1 − 𝛿2)2 = 𝑣0𝑅𝑘

(𝛿1 − 𝛿2)2 (2-17)

where v0 is mole volume in v0/RT.

In this equation, χ12 represents the enthalpic portion of polymer-solvent interaction

parameter. For nonpolar systems the entropic term χ𝑠 is usually a constant between 0.3 and

0.4, with χ𝑠 being 0.34 often used.[73, 74]

Complete miscibility can happen if the Hildebrand parameters are similar and the degree of

hydrogen bonding is similar between the species. This method works quite well for non-polar

interactions which only possess van der Waals forces between species, and does not work in

mixtures with strong polar or specific interactions, like hydrogen bindings. To address the

limitation of the Hildebrand parameter, Hansen and coworkers decomposed the Hildebrand

parameters into three terms (the Hansen Solubility Parameter)[75]:

δ𝑡2 = δ𝑑2 + δ𝑝2 + δℎ𝑏2 (2-18)

where 𝛿𝑡 is Hansen’s total solubility parameter, 𝛿𝑑 the dispersive term, 𝛿𝑝 the polar term, and

𝛿ℎ𝑏 the hydrogen bonding term. This was assumed that dispersion, polar, and hydrogen

bonding parameters are valid simultaneously. The values of the three terms can be determined

empirically. The Hansen’s total solubility parameter should be equal to the Hildebrand

parameter.

22

Figure 2-1. (a)Solubility of PTB7-Th in mix solvents using chlorobenzene as the good solvent; (b) Hansen Space and a sphere-fit matching the solubility limit of 10 mg mL-1 of PTB7-Th.

The solubility parameters of solvents or other small molecules can be determined directly.

For polymers, the solubility parameters cannot be determined from the heat of vaporization

due to their non-volatility. The Hansen solubility parameters (δd, δp, and δhb) of polymers can

be determined via the binary solvent gradient method (BGM), which was employed to probe

the surface of the Hansen sphere for a set of four different solvent mixtures.[76, 77] First the

solubility of each polymer was measured stepwise from good solvent to non-solvent.

Therefore, chlorobenzene was employed as good solvent, while acetone, propylene

carbonate, 2-propanol and cyclohexane were used as non-solvents (low solubility of the

polymers). Because of different weak forces of the non-solvents (propylene carbonate highly

polar or cyclohexane less polar), blends with altered interaction relative to the solute are

created. This results in a controlled change in solubility (Figure 2-1). Next, the Hansen

solubility parameters of each solvent blend were calculated by following equation:

𝐻𝑆𝑃𝑏𝑏𝑏𝑛𝑑 = 𝜙𝑆1 ∙ 𝐻𝑆𝑃𝑆1 + 𝜙𝑆2 ∙ 𝐻𝑆𝑃𝑆2 (2-17)

with ϕS1 and ϕS2 as the volume fraction of chlorobenzene and non-solvent, respectively. This

allows us to transfer the solubility data into HSP data, which are then plotted in the Hansen-

space. By using a solubility limit of 10 mg mL-1, a 0-1 scoring of the HSP data was made,

whereby blend with higher solubility were marked as 1, otherwise 0. Finally a sphere fit was

(a) (b)

23

performed by the software HSPiP. The program evaluates the input data using a quality-of-fit

function with the form:

DATAFIT = (𝐴1𝐴2 ⋯𝐴2)1 𝑛⁄ (2-18)

With n as the number of solvents and

𝐴𝑚 = 𝑒−(𝑏𝑒𝑒𝑒𝑒 𝑑𝑚𝑠𝑡𝑑𝑛𝑐𝑏)𝑖 (2-19)

where the error distance is the distance of the solvent in error to the sphere boundary.[76]

The center of the sphere represents then the Hansen solubility parameters of the polymers.

24

2.4 Thermodynamics of mixing and demixing

In the Flory-Huggins theory (the simplified lattice model), the assumption is the components

are mixed at constant volume, therefore, the energy change of mixing shall employ the

Helmholtz free energy of mixing.

∆𝐴𝑚𝑚𝑚 = ∆𝐻𝑚𝑚𝑚 − 𝑘∆𝑆𝑚𝑚𝑚 (2-20)

where ∆S𝑚𝑚𝑚 is the entropy change per site upon mixing.

The entropy S is defined as the product of the Boltzmann constant k and the natural logarithm

of the number of ways Ω to arrange molecules on the lattice (the number of states):

𝑆 = 𝑘 lnΩ (2-21)

The number of translational states of a given molecule is simply the number of independent

positions that a molecule can have on the lattice, which is equal to the number of lattice sites.

Before mixing, the number of states of each molecule of species 1 is equal to the number of

lattice sites occupied by species 1:

Ω1 = 𝑛∅1 (2-22)

The number of molecules of species 1 is:

𝑛1 = 𝑛∅1𝑁1

(2-23)

The entropy of species 1 in study is:

𝑆1 = 𝑛1𝑘 lnΩ1 = 𝑛∅1𝑁1

𝑘 ln 𝑛∅1 (2-24)

Likewise, the entropy of species 2 before mixing is:

25

𝑆2 = 𝑛2𝑘 lnΩ2 = 𝑛∅2𝑁2

𝑘 ln𝑛∅2 (2-25)

The entropy of species 1 and 2 upon mixing is:

𝑆12 = 𝑛𝑘 ln (𝑛∅1 + 𝑛∅2) = 𝑛𝑘 ln𝑛 (2-26)

The entropy change upon mixing is:

∆𝑆𝑚𝑚𝑚 = 𝑆12 − 𝑆1 − 𝑆2 = 𝑛𝑘 ln𝑛 − 𝑛∅1𝑁1

𝑘 ln𝑛∅1 − 𝑛∅2𝑁2

𝑘 ln𝑛∅2

= −𝑘𝑛(∅1𝑁1

ln∅1 + ∅2𝑁2

ln∅2) (2-27)

The entropy change upon mixing per lattice site is an intrinsic thermodynamic quantity:[72]

∆𝑆𝑚𝑚𝑚 = ∆𝑆𝑚𝑖𝑚𝑛

= −𝑘(∅1𝑁1

ln∅1 + ∅2𝑁2

ln∅2) (2-28)

In a polymer-solvent solution, the number of sites occupied by a polymer molecule N1=N>>1

and a solvent molecule N2=1:

∆𝑆𝑚𝑚𝑚 = ∆𝑆𝑚𝑖𝑚𝑛

= −𝑘 ∅1𝑁

ln∅1 + ∅2 ln∅2

= −𝑘[∅𝑁

ln∅ + (1 − ∅) ln(1 − ∅)] (2-29)

The Helmholtz free energy change per site upon mixing as a function of volume fraction of

polymer:

∆𝐴𝑚𝑚𝑚 = ∆𝐻𝑚𝑚𝑚 − 𝑘∆𝑆𝑚𝑚𝑚

= 𝑘𝑘 ∅𝑁1

ln∅ + 1−∅𝑁2

ln(1 − ∅) + χ∅(1 − ∅)

= 𝑘𝑘[∅𝑁

ln∅ + (1 − ∅) ln(1 − ∅) + χ∅(1 − ∅)] (2-30)

26

This is the Flory-Huggins equation for polymer solution, which can also use to describe

polymer-small molecule solid solution. The first two terms in the above equation have

entropic origin and always act to promote mixing (because ∅ is smaller than 1, the first two

terms are always negative.); however, the contribution is quite small with blends of long-

chain polymers. The last term has energetic origin, and can be positive (opposing mixing),

zero (ideal mixtures), or negative (promoting mixing) depending on the sign of the interaction

parameter.

If two species like each other better than they like themselves, net attraction, interaction

parameter will be negative and a single-phase mixture is favorable for all compositions. More

often in practice, there is a net repulsion between species, thus the interaction parameter is

positive. The equilibrium state of the mixture depends not on the sign of the free energy of

mixing at the particular composition of interest, but on the composition for the whole range

of compositions. This functional dependence ∆𝐴𝑚𝑚𝑚(∅) depends on the value of the

interaction parameter as well as on the degrees of polymerization of the polymers.

The definition of thermodynamic equilibrium is the state of the system with minimum free

energy. Stability of the mixture is determined by whether the free energy of the mixed state

A𝑚𝑚𝑚 is higher or lower than of a phase separated state A1+A2.

The local stability of the polymer-small molecule is determined by the sign of the second

derivative of the free energy with respect to composition:

∂2∆𝐴𝑚𝑖𝑚𝜕∅2

= ∂2∆𝐻𝑚𝑖𝑚𝜕∅2

− 𝑘 ∂2∆𝑆𝑚𝑖𝑚𝜕∅2

= 𝑘𝑘 1𝑁∅

+ 11−∅

− 2χ𝑘𝑘 (2-31)

If the temperature is lowered, the entropic term decreases, allowing the repulsive energetic

term to start to be important at intermediate compositions. Entropy always dominates the

extremes of composition, making those extremely stable. Below the critical temperature Tc,

compositions range with concave free energy appears; there is a range of compositions for

27

which there are phase separated states with lower free energy than the homogeneous state,

making the appearance of demixed states.

By considering the temperature dependence of the free energy of mixing, a phase diagram

can be built to summarize the phase behavior of the mixture, displaying regions of stability,

instability, and metastability. The phase boundary is determined by the common tangent of

the free energy at the compositions ∅′ and ∅′′ corresponding to the two equilibrium phases.

(∂∆𝐴𝑚𝑖𝑚𝜕∅

)∅=∅′ = (∂∆𝐴𝑚𝑖𝑚𝜕∅

)∅=∅′′ (2-32)

For the simple example of a symmetric polymer blend with N1=N2=N, the common tangent

line is horizontal.

∂∆𝐴𝑚𝑖𝑚𝜕∅

∅=∅′

= ∂∆𝐴𝑚𝑖𝑚𝜕∅

∅=∅′′

= 𝑘𝑘 ln ∅𝑁− ln(1−∅)

𝑁+ χ(1 − 2∅) = 0 (2-33)

The above equation can be solved for the interaction parameter corresponding to the phase

boundary — the binodal of a symmetric blend (as shown in Figure 2-2, solid line in the

bottom):

χ𝑏 = ln (∅/(1−∅))(2∅−1)𝑁

(2-34)

The binodal for binary mixtures coincides with the coexistence curve, since for a given

temperature (or Nχ) with overall composition in the two-phase region, the two compositions

that coexist at equilibrium can be read off the binodal. Any overall composition at

temperature T within the miscibility gap defined by the binodal has its minimum free energy

in a phase-separated state with the compositions given by the two coexistence curve

composition ∅′ and ∅′′.

28

Figure 2-2. Composition dependence of the free energy of mixing for a symmetric polymer blend with the product Nχ = 2.7 (top figure) and the corresponding phase diagram (bottom figure). Binodal (solid curve) and spinodal (dashed curve) are shown on the phase diagram. Figure is from the book Polymer Physics by Ruinstein M.[72] .

For an asymmetric blend like polymer-small molecule, the inflection points in ∆𝐴𝑚𝑚𝑚(∅) can

be found by equating the second derivative of the free energy to zero:

∂2∆𝐴𝑚𝑖𝑚𝜕∅2

= 𝑘𝑘 [ 1𝑁1∅

+ 1𝑁2(1−∅)

] = 𝑘𝑘 1𝑁∅

+ 11−∅

− 2χ𝑘𝑘 = 0 (2-35)

The curve corresponding to the inflection point is the boundary between unstable and

metastable regions and is called spinodal (the dashed line in the bottom part of Figure 2-2):

χ𝑠 = 12

( 1𝑁∅

+ 11−∅

) (2-36)

In a binary blend, the lowest point on the spinodal curve corresponds to the critical point:

N N c

29

𝜕χ𝑠𝜕∅

= 12

− 1𝑁∅2

+ 1(1−∅)2

= 0 (2-37)

The solution of this equation gives the critical composition:

∅𝑐 = 1√𝑁+1

≅ 1√𝑁

(2-38)

From this, we can see that the phase diagram of polymer-small molecule composite is

strongly asymmetric with low critical composition.

The critical interaction parameter can be determined by substituting this critical composition

back into the equation of spinodal:

χ𝑐 = 12

( 1√𝑁

+ 1)2 ≅ 12

+ 1√𝑁

(2-39)

The critical interaction parameter of polymer-small molecule composite is close to 0.5. The

spinodal and binodal for any mixture meet at the critical point (Figure 2-2). If the interaction

parameter below the critical interaction parameter (χ < χ𝑐 ), the homogeneous mixture is

stable at any composition. For χ > χ𝑐, there is a miscibility gap between the two branches of

the bimodal in Figure 2-2. For any composition in a miscibility gap, the equilibrium state

corresponds to two phases with compositions ∅′ and ∅′′ located on the two branches of the

coexistence curve at the same value of χ .

Consider a sudden temperature jump that brings a homogeneous mixture at the critical

composition ∅𝑐 into the two-phase region. The system will spontaneously phase separate into

two phases with compositions given by the values on the coexistence curve at that new

temperature. This spontaneous phase separation, called spinodal decomposition, occurs

because the mixture is locally unstable. Any small composition fluctuation is sufficient to

initiate the phase separation process.

The points of the phase diagram between the spinodal and the binodal curves (The two

regions that have positive second derivative of the free energy of mixing) correspond to

30

metastable mixtures. The metastable homogeneous state is locally stable against small

composition fluctuations and requires a larger nucleation event to initiate phase separation

into the equilibrium phases given by the coexistence curve. The nuclei of the more stable

phase must be larger than some critical size in order to grow in the metastable region because

of the surface tension between phases. The new phase can grow only when a sufficiently

large fluctuation creates a domain larger than the critical size.

31

Chapter 3 State of the art

Abstract

In this chapter, the degradation stresses of organic solar cells are summarized and the

general loss mechanisms are elucidated in category of degradation of electrodes,

hole/electron transporting layers and active layer. Further, strategies and methods to

overcome the intrinsic degradation induced by heat are explicated and analyzed. In the

end, progress in understanding the photo-degradation by white light illumination is

interpreted and discussed.

32

3.1 Degradation of organic solar cells

To be commercially competitive, one of the biggest challenges for OPV is to perform

consistently throughout its lifetime. It is widely observed that organic solar cells degrade over

time. There are generally various aging stresses, varying layer contributing losses, and

multiple loss mechanisms in the real world. In principle, aging stresses comes from light

source, temperature, atmosphere and etc (Figure 3-1). The International Summits on OPV

Stability (ISOS) outlined an experimental protocol to promote reproducibility across different

labs.[78] The aging conditions presented by ISOS are normally combined stresses that

include dark, laboratory weathering, thermal cycling and solar-thermal-humidity cycling.

However, to understand and address the instability issue, it is also vital to control aging

conditions, limiting loss mechanisms and focusing on the dominant degrading layer at a time.

Each aging factor can induce degradation in all layers, including electrodes, interfaces, and

active layers.

Figure 3-1. Degrading factors of OPV.

RH

33

Degradation of electrodes

There are three key functions on solar cells of electrodes: to set an electric field across the

active layer to drive electrons and holes to the appropriate contact; to provide a suitable

energy level to selectively extract the electrons or holes reaching the corresponding contact;

to offer a low resistance pathway to laterally transport charges out of the device, producing

electricity. The most commonly used electrode materials for OPV are indium tin oxide

(acting as transparent electrode), Calcium, aluminum and silver.

Low work function metals, like Ca, are well known to easily oxidize if exposing to the

atmosphere, particularly sensitive to humidity, even without exposure to photo

illumination.[79] Oxidation changes the conductivity and work function of the electrode

layer, which induce photovoltaic performance losses. It was first discovered in OLEDs that

metal electrode degraded in the presence of water, growing dark spots.[80-82] With

unpackaged solar cells, water can diffuse into electrode pinholes and cause dead zones of

solar cells to grow from the edges.[83] By employing inverted solar cell architecture, the low

work function Ca would be replaced, and thus the stability of the solar cells is improved

dramatically.[84, 85] Besides Ca, Al is reactive as well. Glatthaar presented in several papers

that Al2O3 would lead the IV curve changed from a standard exponential diode curve to a

curve with an inflection point, which might deteriorate the FF to some degree, or total fail the

solar cell creating an S-shape IV curve.[34, 86]. Silver is less reactive comparing to Ca and

Al.

Another loss mechanism from the electrodes is the diffusion of metal atoms into the interface

and active layer. It is evidenced that In from indium tin oxide and Al can diffuse into

poly(3,4-ethylenedioxythiophene):poly(styrenesulfonate) (PEDOT:PSS) and further the

active layer, which change the energy levels of interface layers and act as traps for charge

recombination, causing degradation of solar cells.[87, 88]

34

Degradation of interface layers

The hole or electron transporting layers are of importance to attain high efficient organic

solar cells.

The most commonly used hole transporting layers (HTL) are PEDOT:PSS and MoOx.

PEDOT:PSS is rather vulnerable to humidity and mechanical stress. Modification of

PEDOT:PSS with polymer additives, solvent additives, metal oxides or nanoparticles is

effective to improve stability of PEDOT:PSS. Kim[89] found out that with the addition of PS

nanoparticles in the HTL, the conductivity of PEDOT:PSS maintained, further, the solar cells

degraded fast less. Zhang[90] employed MoO3- PEDOT:PSS as HTL and increased solar cell

lifetime significantly, which is resulted from the less hydroscopicity of MoO3. Completely

replacing PEDOT:PSS with evaporated MoOx can greatly improve the air stability of solar

cells owing to the less hydroscopicity and better inoxidizability.[91] However, in the situation

under thermal stress, PEDOT:PSS based solar cells outperformance MoOx based solar cells

in terms of stability. Other metal oxides such as vanadium pentoxide[92] and tungsten

oxide[93] et al, are promising candidates acting as highly efficient and stable hole

transporting layers, however, more studies are required to spread the use of these materials.

The most popular electron transporting layers (ETL) are ZnO, Lithium fluoride (LiF). In

conventional geometry, LiF is an alternative to Ca acting as ETL. However, LiF is also

sensitive to air due to its reactivity with water and oxygen. Xu[94] and Yang[95] attempted to

employ metal oxide and polyelectrolyte respectively to improve the stability of ETL, which is

a good direction. ZnO used in inverted solar cells is unstable when exposing to air or light.

One helpful way to improve the stability issue of ZnO is to modify the fabrication method.

Zhang[96] treated the surface of ZnO film with ethanedithiol to passivate the surface defects

by removing reactive groups (-COO and -OH). Qiao[97] modified ZnO by inserting a poly-

ethylenimine ethoxylated layer and succeeded in suppress the oxidation of ZnO for long-term

stability in air. To overcome the photo-instability of ZnO, phosphonic acid was applied to

35

modify the ZnO layer, which was proved to be an effective method.[98] Research efforts on

achieving efficient and stable ETL are supposed to continue.

Degradation of active layers

Photoactive organic materials are well known to be unstable in the presence of light and

oxygen - photo-oxidation.[99, 100] Optical density of semiconductors would decrease

resulted from photo-oxidation, which can be easily and correctly tracked by measuring the

absorption of active layer films as a function of illumination time. Under one-sun intensity,

degradation of such kind is observed to happen on a time scale ranging from seconds to hours

depending on materials. A series of semiconducting materials, like P3HT, PPV, PTB7,

PCDTBT and et al, were investigated by means of infrared spectroscopy which demonstrated

losses of conjugated bonds and appearance of carbonyl, alkoxy and ester bonds, indicating

the chemical reactions.[31, 101-106]

Through electron spin resonance, the existence of free radicals is detected in aged polymer

semiconductors. Further, the concentration of free radicals is detected to increase with longer

illumination time.[107] The chemically loss mechanism of photo-oxidation is thought to

proceed by a typical free-radical reaction which normally consist of initiation, propagation,

and termination.[99] Generally, initiation happens when a bond is broken by light. Once a

free radical is formed, it could propagate throughout the whole film by reaction or diffusion.

During propagation, scission occurs to conjugated bonds, which is the direct contribution to

absorption loss. If two free radicals meet and recombine, termination of the reaction happens.

Concerning semiconducting organic materials, free radicals appear often due to abstraction of

hydrogen atoms from alpha carbons of side chains. With the presence of oxygen which can

readily flow through the film, the propagation is hugely accelerated. Through removing the

polymers’ side chains after deposition of the active layer, the photostability can be

improved.[108]

It is found out that not only chemical structures but also film morphology plays important

role on the photostability of active layers. For material systems that can deposited in both

crystalline and amorphous film, the films with better crystalline are normally happened to be

36

more stable.[104, 109, 110] Furthermore, film density is another affecting factor. Denser

films might suppress the diffusion of oxygen and free radicals to some extent, thus tend to be

more stable.[109] Moreover, the existence of PCBM also helps improve the photooxidative

stability.[31, 111] The stabilizing effects of PCBM are likely because: 1) the excited state of

polymer will be quenched by transferring electron to the fullerene which competes with

electron transferring from polymer to form O2− .[112] 2) fullerenes can act as free radical

scavengers in the film. It is found out that fullerene cages might be able to trap up to eight

free radicals.[113]

The above mentioned degradations of active layer are called the extrinsic degradation which

can be significantly or completely suppressed by proper encapsulation. However, there is

intrinsic kind of degradation which is unavoidable and must be addressed. Owing to the facts

that organic semiconductors normally have no absorption in the UV portion of the solar

spectrum, and that the photon energy of UV light can break chemical bonds, organic

electronic devices are normally in application with the UV light filtering out. Thus, white

light and accumulating heat from illumination are the two intrinsic stressors for organic solar

cells. Every layer, including electrodes and interfaces, would contribute to the photovoltaic

performance losses; however, the focus of this thesis is mainly the losses owing to the active

layers.

37

3.2 Progress in improving thermal stability of active layer

Intrinsic dark degradation of BHJ solar cells typically focus on the temperature dependence

of degradation which normally involves the movement of materials in the solar cells. When

assessing the effects of aging temperature on degradation, BHJ films or solar cells are aged

on hotplate at chosen temperature for a certain period of time in inert atmosphere. Previously,

most of the thermal stability tests run at accelerated temperature like 100 °C and even 150 °C.

As the working temperature of solar cells is typically far below the decomposition

temperature of most photovoltaic materials, heat induced degradation is dominantly owing to

microscopic morphology changes of the BHJ. There are principally two strategies in tackling

the undesirable phase separation challenge: 1) introducing chemical locking, which was

stimulated by the strong phase separation of PCBM based solar cells.[6],[41, 42],[43],[44],

[45] 2) employing organic semiconductors with intrinsic stability.[46-48]

Chemical locking

Previous research to tackle this challenge focused on chemically locking the fullerenes,

typically, by the following three strategies: 1) Adding cross-linkable groups to the donor

materials; 2) PCBM based locking strategy; 3) Addition of cross-linkable small molecule

additive.[6] Recently, Mynar and Yang designed new BDT based donors with fullerene-

reactive groups, and improved devices thermal stability to some extent.[41, 42] An alternative

even more universal strategy is to modify the fullerene-based acceptors. Poly(ethylene

glycol) (PEG) capped fullerene PCB-PEG and [6,6]-phenyl-C61 butyric acid

pentafluorophenyl ester (PC61BPF) were found to stabilize the P3HT:PCBM solar cells by

adding 8 wt%.[114, 115] [6,6]-phenyl-C61-butyric acid styryl ester (PCBS) and [6,6]-phenyl-

C61-butyric acid styryl dendron ester (PCBSD) were other successful examples, but were

only reported for P3HT:PCBM solar cells so far.[43] Durrant et al. gained enhanced

morphological stability and performance in several PCBM-based systems by light-induced

oligomerization.[116, 117] However, PCBM dimerization leads to performance losses and is

thus no viable strategy to enhance stability.[38, 53] More recently, Wantz et. al employed

cross-linkable small molecule 4,4'-Bis(azidomethyl)biphenyl in the active layer to enhance

38

the thermal stability by sacrificing some degree of efficiency.[44, 45] Xiaowei Zhan’s group

utilized 4,4'-Biphenol to stabilize solar cells with 80% efficiency preserved after annealing at

130 °C for 120 minutes.[45]

Unfortunately, there are certain shortcomings of this strategy: 1) the initial photovoltaic

performance of the cross-linked solar cells is usually lower than that of the non-cross-linked

analogs. 2) The thermal instability phenomena are not completely addressed. By crosslinking,

the macro scale phase separation of the donor and acceptor is indeed suppressed, such that the

photovoltaic performance preserve at high levels upon thermal annealing; nevertheless, there

are still burn-in type losses existing in the cross-linked solar cells, which is believed to result

from the demixing of amorphous donor:acceptor intermixed phase. 3) The chemically

functional groups introduced to realize the crosslinking are difficult to consumed 100% via

the intentional crosslinking and would later-on transform into the unstable elements during

the long-term stability especially under illumination.

Recombination of intrinsically stable materials

Developing insights are gained about the microstructural phases and the morphological

instability of the BHJ active layer. More and more evidences imply that owing to the poor

miscibility between donor and PCBM, the polymer: fullerene mix regions turn out to be

meta-stable, thus, leading to poor stability of OPV devices. The combination of more

miscible donor and acceptor materials is proposed to address the meta-stability of polymer:

fullerene solar cells. Xiaowei Zhan’s group introduce ICBA, which has better miscibility with

most polymers, to a number of polymer:PCBM composites and enhanced the thermal

stability of the devices.[48] The mechanism of this improved thermal stability lie in the fact

that the heat-induced crystallization of PCBM is suppressed due to the amorphous nature of

ICBA, thus, resulting in the better thermal stability of BHJ film. Simon Dowland employ

poly(fullerene) and PCBM as co-acceptor in combination with P3HT and attain ternary solar

cells with efficiency very close to the P3HT:PCBM binary devices.[47] Under thermal

stressing, the multi-acceptor based solar cells demonstrate significantly improved thermal

stability, outperforming the binary control devices by suppressing the PCBM migration and

39

aggregation. The influences of the poly(fullerene) on the devices’ performance and enhanced

stability are rationalized in terms of a thermodynamic model based on the Flory-Huggins

solution theory. In this PhD program, we replaced PCBM with two fullerene-based acceptors

which have better miscibility with PCE10 and pDDP5T and attained impressively

thermostable BHJ composite films.[46] We demonstrate that the low miscibility between

PCBM and pDPPT5-2 or PTB7-Th is the fundamental origin of the low thermal stability. On

the contrary, two novel fullerenes, PyF5 and FAP1, with a significantly higher chemical

compatibility are introduced to overcome these limitations. The benefit of chemical

miscibility as a novel design principle for improved stability is expected to pave alternative

guidelines towards designing and developing novel acceptors for efficient and thermally

stable OPV devices.

40

3.3 Understanding white light induced loss mechanisms

It is observed that for organic solar cells aged under white light illumination in nitrogen, there

exist three degrading time regimes: An initial period of exponential degradation; a period of

relatively linear and slow degradation; the final failure period. The second degradation period

can normally last for years thus it is rarely to see a completely failed solar cell. Intensive

attentions are paid to the steep degradation of the initial period, so-called burn-in loss. The

term burn-in is originally from the commercial practice of electronic device manufacturing

with a short-time thermal treatment before entering the final market. However, the burn-in

loss of organic solar cells tends to be more severe and protracted than that of other electronic

devices, in which during a time frame of tens or hundreds hours, initial PCE typically

dropped by 10-50%.

Some encapsulated organic solar cells are stable in the dark, nevertheless, once under

illumination, they degrade rapidly and show the exponential burn-in loss. Obviously, the

degradation is induced by the interaction white light. Still, multiple loss mechanisms exist

upon white light illumination and have not been fully understood. Researches concerning the

white light induced degradation are predominantly focused on the losses from the

deteriorated open-circuit voltage and/or short-circuit current density.

VOC burn-in Loss

Light-induced VOC loss is observed in a series of polymer-fullerene solar cells. Particularly,

the VOC loss in PCDTBT:PCBM solar cells is one of the most intensive systems, which draw

researchers’ interest. It has been reported that PCDTBT:PCBM solar cells can resist

temperature up to 80 degree and are rather stable at room temperature in the dark. McGehee’s

group re-apply new metal electrodes to of light-aged solar cells by peeling off the original

metal electrodes with scotch tape, and find out that FF recover to a high level but VOC does

not. This underpins that the VOC loss comes from the BHJ active layer.[51] With sensitive

absorption techniques, they have direct observation of defect states that form within the

bandgap in both aged solar cells and films.[118-121] By illuminating hole-only diodes of

PCDTBT:PCBM system, it is shown that the VOC loss correlates well with an increase in

41

energetic disorder on polymer. It is concluded that photo-induced defects on polymer are the

implications of the light-induced VOC loss in PCDTBT based solar cells.

Currently, there are two identified mechanisms that defect states in the active layer can cause

VOC loss. Firstly, the recombination rate constant would increase with more defect states,

which would shorten the carrier lifetime, and accordingly diminish the charge carrier density

at steady-state.[67, 122] Consequently, this lower the quasi-fermi level splitting as well as the

VOC. Secondly, the density of state (DOS) near the quasi-fermi level would broaden with

more defect states. The charge carrier lifetime and density are found to be the same for both

fresh and aged solar cell with combination of photo-voltage decay and charge extraction

measurements. The defect states are proved to rise in the DOS over aging time. Even though

the charge carrier density at open circuit is the same as fresh, the charges would prefer to fill

states with lower energy, hence, the ultimate DOS is filled up to a lower energy.[123, 124]

There are a couple of hypotheses of the chemical origin of the defect states causing the DOS

broadening. One hypothesis suggests that there might be a limited number of reactions sites

or reactive species in the active layer for the burn-in degradation gradually weakens and

eventually stops. These reactive species could be reactive capping groups at the polymer

chain ends, or impurities like palladium and bromine left from synthesis.[118] Another

assumption is hydrogen abstraction. Although the bond strength of a C-H bond is around 4-5

eV, the activation energy of hydrogen abstraction can be cut down to 2.2 eV due to atomic

relaxation within the polymer.[120] Further the possibility of burn-in VOC loss is the light

induced crosslinking between compounds because crosslinking can accounts for the self-

limiting nature of kinetics, which could stop the burn-in eventually by limiting the spatial

rearrangement.[125] It is also reported that solar cells possess lower initial VOC with higher

concentration of free radical species in polymer.[126] Further research efforts are needed in

the field.

JSC burn-in loss

A number of polymer:PCBM solar cells as well as small molecule:C60 solar cells are

observed to suffer from light induced JSC.[38, 53, 54, 127] There is external quantum

42

efficiency (EQE) loss in the absorption region of fullerenes, which indicates the involvement

of fullerenes in the degradation. What’s more, it is well known that in the absence of oxygen,

C60 based fullerenes are found to form oligomers under irradiation, which would change

their electronic properties.[128] PCBM dimers would be formed in similar conditions.[116,

129] PCBM dimers in degraded solar cells are demonstrated by an enhancement in the

absorption at 320 nm, and identified via high-performance liquid chromatography (HPLC).

By intentionally blending the PCBM dimers (from HPLC) with regular PCBM and making

solar cells, the J-V curves of degraded solar cells can be accurately reproduced, which implies

that PCBM dimers are the origin of the current density loss.

It is also observed that the relative degree of JSC burn-in loss varies depending on polymer

systems and film processing, which implies that film morphology affects the amount of

dimerization occurring. In fact, the ordering of the film, as well as the extent of polymer-

fullerene intermixing, impacts the dimerization reaction. In relatively amorphous polymers,

like PCPDTBT and PCDTBT, based BHJs where polymer and fullerene are intimately mixed,

fullerene dimers are suppressed. Furthermore, the JSC loss and extend of dimerization are also

influenced by the voltage condition during illumination. Solar cells aged at VOC condition

show more degradation than aged at JSC condition. According to the above observations, it is

proposed that dimerization reaction occurs via triplet excitons on the fullerene, which would

present in higher concentration at VOC condition and are quenched more effectively in highly

intermixed BHJs.[38]

It is thought that PCBM dimers acts as exciton trapping in the fullerene phase, such that

excitons generated in the PCBM phase are not effectively dissociated and collected, which is

evidenced by the EQE of aged and fresh solar cells. In bilayer solar cells, the fullerene

absorption wavelength is almost completely lost for aged solar cells. A device model

accounting for that decreased PCBM exciton diffusion length in degrading bilayer solar cells

can well-reproduce the J-V curves of aged solar cells. This device model is less fit with BHJ

solar cells, and a diminished exciton diffusion length in the PCBM domain cannot entirely

account for the JSC loss. Thus, it is proposed, in BHJ devices, that PCBM dimers also

influence exciton dissociation, which accounts for the EQE losses of the BHJ solar cells.

43

Nevertheless, the exact mechanism of dimerization in BHJ films still requires more

investigation.

In this PhD program, we find out that JSC burn-in loss is not purely induced by fullerene

dimerization. We investigate the degradation behavior of a semi-crystalline Poly[(5,6-

difluoro-2,1,3-benzothiadiazol-4,7-diyl)-alt-(3,3’’’-di(2-octyldodecyl)-2,2’;5’,2’’;5’’,2’’’-

quaterthiophen-5,5’’’-diyl)] (PffBT4T-2OD, or PCE11) based BHJ system through thermal

annealing and white light illumination. Even in the absence of fullerene dimerization, which

was guaranteed by using PC71BM, solar devices still suffer from JSC burn-in degradation

upon illumination or thermal stress. The degradation of PC71BM-based solar cells caused by

either thermal-annealing or illumination can be attributed to morphological changes,

particularly to the demixing of the donor and acceptor phases in the intimately mixed

amorphous region. Furthermore, by actively aging the solar cells under varying temperature

and various light intensities, we report for the first time that the microstructure changes

induced by either thermal- or light-aging can be correlated to each other as they are arising

from the same physical origin.

44

45

Chapter 4 Materials and methods

Abstract

In this chapter, materials used in the thesis are described in terms of photoactive

materials, interface and electrode materials; further device structure and device

preparation are summarized; Finally, each characterization technique is briefly introduced.

46

4.1 Materials

4.1.1 Active layer materials

The polymers and fullerene acceptors employed in the thesis are summarized in Figure 4-1

and Table 4-1. All materials were used as received without further purification.

Table 4-1 Materilas used in the thesis Materials Provider Mn (kg/mol) PDI Purity (%)

PTB7-Th 1 Material

pDPP5T-2 BASF

PffBT4T-2OD

/PCE11

1 Material 28k 1.8

1 Material 53k 1.6

Collaborator 43k 2.1

P3HT BASF

PyF5 Collaborator

FAP1 Collaborator

[60]PCBM Solenne 99

[70]PCBM Solenne 99

47

S

S

S

S

S

SF

O O

n

N

N

SS S

S S

O

O

n

PTB7-Th

pDPP5T-2

NSN

F F

SS

S

Sn

FBT-Th4 (1,4) or PCE11

S

C6H13

n

P3HT

OO

N

PyF5

O

OO

FAP1

[60]/[70]PCBM

O

OC10H21 C8H17

C10H21 C8H17

Figure 4-1. Chemical structures of polymers and fullerenes.

4.1.2 Interface and electrode materials

Doctor bladed PEDOT:PSS (from Nanograde, AI 4083) and evaporated MoOx are used as

hole transporting layer. Doctor bladed ZnO nanoparticles (from Nanograde, n-10) and

evaporated Ca are used as electron selective layer.

48

Evaporated Ag and Al as well as ITO are used as electrode.

4.2 Device preparation

The device structures are shown in Figure 4-2, which normally possesses electrodes and

interface layers and active layers in the middle. The sequence of film deposition is by coating

film from ITO layer. The active area of the final solar cells is 10.4 mm2.

Figure 4-2. Device structures employed in the thesis.

Besides evaporation, solution process is used to deposit interface layers and active layers.

Deposition techniques for solution processable layers are doctor blading and spin coating.

Both techniques can be employed to obtain homogeneous thin films on substrate. Spin

coating is very popular for lab scale film deposition for its simple and easy operation even

inside glovebox which is typically employed to facilitate inert atmosphere. The advantages of

doctor blading are the capabilities of depositing larger area films and fast transfer to slot-die

coating and roll-to-roll printing which enables large scale production. It is notable that the

drying kinetics and final film morphologies are different via spin coating or doctor blading.

During spin coating, the solvents evaporate at room temperature by reduced vacuum inducing

by fast spinning. On the contrary, during doctor blading, the solvent evaporate at elevated

temperature which provides by a hotplate that also heats up the solutes, thus leading to

different film morphology.

ITO

Electron transporting layer

Active Layer

Hole transporting layerAl/ Ag

ITO

Hole transporting layer

Active Layer

Electron transporting layerAl/ Ag

Normal structure Inverted structure

49

4.3 Methods of characterization

4.3.1 J-V characteristics

J-V curves of all solar cells are characterized with a source measurement unit form BoTest.

The light curves are measured under 100 mW/cm-2 illumination provided by a Newport SollA

solar simulator with AM 1.5G spectrum, while the dark curves are measured without

illumination.

In the dark, a working solar cell shows a typical diode behavior that only allows current to

flow in one direction, which is called rectifying behavior and is resulting from an asymmetric

junction in the solar cell that is necessary for charge separation.[130]

4.3.2 SCLC

Space charge limited current method is normally employed to characterized film mobility.

The single carrier devices are fabricated and measured dark current-voltage characteristics.

The space charge limited (SCL) regime is analyzed. The electron device structure for SCLC

characterization is ITO/Al/Active layer/Ca/Ag. For film making, pristine fullerene layer was

spin coated under ambient atmosphere. BHJ active layer was processed the same as solar

cells. 15 nm calcium and 100 nm silver were deposited subsequently under 6 × 10-6 Torr by

thermal evaporation through a shadow mask to form an active area of 10.4 mm2. The electron

mobility was estimated by fitting the current-voltage curves according to the SCLC modified

Mott-Gurney model: [131]

𝐽𝑆𝑂𝑆 = 98𝜀0𝜀𝑒𝜇

𝑉𝑖𝑖𝑖

𝑆3exp (0.89×𝛽

√𝑆√𝑉) (4-1)

Where 𝐽𝑆𝑂𝑆 is the current density in the space charge limited regime; 𝜀0 is the permittivity of

free-space (ε0=8.85*10-12 A.s/V.m = 8.85*10-11mA.s/V.cm); 𝜀𝑒 is the relative dielectric

constant of the active layer (3 for pristine, 2.7 for blends); 𝜇 is the charge carrier mobility; V in

is the built-in voltage; L is the thickness of the sample; n should be close to 2; 𝛽 should be

50

close to 0.

4.3.3 Absorption, PL &FTIR

Film absorption is characterized with a UV-vis-NIR spectrometer Lambda 950 from 300 nm

to 850 nm.

Photoluminescence spectra are recorded with a home-made LIBC which has a Si-CCD

attached to iHR320 monochromator (Horiba). Thin film PL measurements are conducted

under excitation from a 375 nm diode laser. The fluorescence spectra are corrected with the

optical density of the samples at the excitation wavelength.

The FTIR spectra were measured with a Vertex 70 from Bruker optics. A PCE11: PC71BM

blended solution was drop casted on Zinc sulfide based substrates and dried under vacuum

overnight. The light-aged sample was illuminated in the degradation chamber for 100 hour,

while the other samples were kept in the dark and in nitrogen atmosphere. The thermal aged

sample was aged at 85 °C for 2 hours on a hotplate inside a glovebox prior to FTIR

characterization. (data in chapter 7 were done by Xiaofeng Tang)

4.3.4 DSC and MDSC

DSC measurements were taken with a Q1000 from TA Instruments. The temperature can

range from -100 to 310 °C with a heating and cooling rate of 10 K/min.

The cooling and heating rate of MDSC measurements is set to 3 K/min.

To prepare the sample, solutions of pristine or blends are drop cast on clean glass substrates

and dried under inert atmosphere for 3 hours and under vacuum overnight.

4.3.5 GIWAXS and GISAXS

The GIWAXS/GISAXS patterns werce collected with the highly customized Versatile

Advanced X-ray Scattering instrumenT ERlangen (VAXSTER) at the chair for

Crystallography and Structural Physics, FAU, Germany. The system is equipped with a

51

MetalJet D2 70 kV X-ray source from EXCILLUM, Sweden. The beam was shaped by a 150

mm Montel optics (INCOATEC, Geesthacht) and two of the available four double-slit

systems with the last slit system equipped with low scattering blades (JJXray/SAXSLAB).

Aperture sizes were (0.7×0.7 mm2, 0.462 ×0.462 mm2) for GIWAXS and (0.15×0.06 mm2,

0.12×0.048 mm2) for GISAXS. The sample position was located within the fully evacuated

detector tube. The hybrid-pixel 2D Pilatus 300K detector (Dectris Ltd., Baden, Switzerland)

was used to collect the scattered radiation. The measurements were carried out at energy of

9.24 keV. The samples were mounted on a yz-theta goniometer allowing to adjust grazing

incidence angles which maximize the scattering volume and enhance the scattered intensity.

The incidence angle for GIWAXS measurements was 0.17 °. Grazing incidence geometry of

the incident X-ray with respect to the sample surface is used here to enhance the scattered

intensity, to maximize the scattering volume, and to access the three dimensional (3D)

structure of the studied thin films (lateral and normal direction). The sample-to-detector

distance (SDD) was calibrated with a silver behenate standard to 179 mm for GIWAXS and

1590 mm for GISAXS. An incidence angle around 0.25 ° was chosen to obtain a clear

separation between the Yoneda peaks of the involved materials and the specular peak in

GISAXS. Data were reduced with dpdak software.59 The structural model for the reduced

GISAXS data was fitted using the program SASfit.60 The PCE11: PC71BM blend films were

spin coated on silicon substrates. The light-aged film was aged for 140 hours and the thermal

aged film was thermal annealing for 1 hour under nitrogen atmosphere. Both GIWAXS and

GISAXS were performed on the same sample.

(done by Thaer Kassar and Wolfgang Gruber in the Institute of Crystallography and

Structural Physics, FAU, Germany)

GIWAXS patterns of the pristine fullerenes (PyF5, FAP1 and PCBM) were performed on the

ID-10 beamline at European Synchrotron Radiation Facilities (Grenoble, France). Diffraction

patterns were collected with a Pilatus 300k detector (172x172 μm pixel size). The wavelength

used was 1.24Å. The measurements were performed on thin films on Si substrate at an

incidence angle of 0.16º. The modulus of the scattering vector was calibrated using several

diffraction order of silver behenate. In-situ heating ramps were performed with Linkam

52

heating stage. The integration of 2D-WAXS patterns was performed in a home-made routine

written in Igor Pro software.

(done by Denis V. Anokhin and his collegues in Institute for Problems of Chemical Physics

of Russian Academy of Sciences, Semenov Prospect 1, Chernogolovka, 142432, Russia.)

4.3.6 Lifetime characterization

White light LEDs degradation characteristics: The solar cells were loaded into a sealed

chamber which can contain nine substrates of solar cells. This chamber was then

continuously purged with nitrogen. The water and oxygen level was kept below 0.5 ppm. For

white light illumination, the aging light sources are eight white light LEDs with an emission

spectrum between 400 nm to 800 nm. For thermal aging, the light sources for inline J-V

characteristics are eight white light LEDs and one UV-LED with wavelength of 365 nm. The

neutral-density filters are the FS-ND series from Newport.

High power illumination degradation setup: This is a homemade, highly-accelerated lifetime

setup which has precise temperature control and a variable light intensity ranging from 1 sun

to about 140 suns. The light source is a xenon short arc lamp (OSRAM XBO 1600W/XL

OFR). To increase illumination densities, the light is focused by a mirror and homogenalized

by a beam homogenizer onto the samples. To control sample temperatures and to keep the

solar cell at a constant temperature value, the devices under test were mounted on an actively

cooled copper heat sink which removes the heat impact of the light. The heat sink is also

equipped with electrical contacts to perform inline measurements of the cell parameters.

Furthermore, the samples ware mounted in a box and kept in nitrogen atmosphere while

illumination. For further details of this setup please refer to literature. The 500 nm and 650

nm cut-on long-wave pass filters are UV Grade Fused Silica form Edmund Optics. The 400

nm cut-on longpass filter is GG400 from SCHOTT AG.

53

Chapter 5 Analysis of the application of polymeric

fullerene derivatives in OPV

Abstract

A polystyrene based side-chain polymeric fullerenes was successfully synthesized and

applied in P3HT based solar cell. However, limiting by the strong aggregation and

extremely low solubility in common organic solvents, MPMCPS based binary and ternary

solar cells fail to achieve desirable performance. Two main-chain polymeric fullerenes,

PPC4 and PPCBMB, were employed as additives in PCE11:PCBM based solar cells. The

photovoltaic performance and photo-stability of the ternary solar cells were investigated.

Part of the results presented in the chapter was done by collaborators:

The synthesis of methyl prop-2-yn-1-yl malonate C61 (MPMC) were collaborated with

Institute of Organic Chemistry ӀӀ and done by Andreas Kratzer.

54

5.1 Introduction

Extensive work has been devoted to proposing a large number of approaches to improve the

stability of organic photovoltaics. One of the universally applicable approaches is to tether

fullerenes together by chemical bonding, which has great potentials to suppress the fullerene

diffusion and stabilize the microstructure of the solar cells. There are mainly two strategies of

this approach: one is by incorporating fullerenes as a pendent moiety in the side-chain of

conjugated or non-conjugated polymer, the so-called side-chain polymeric fullerenes[132,

133]; the other one is the poly(fullerene)s which contain the fullerenes in the backbone of the

polymers, so-called main-chain polymeric fullerenes[47, 134]. Each strategy has its own

advantages and disadvantages. The side-chain polymeric fullerenes can preserve the

maximized properties of the small fullerene molecules, however has tendency to aggregate.

The main-chain polymeric fullerenes have potential to improved solubility in organic

solvents however might suffer from reduced electron mobility. Both types of polymeric

fullerenes have not been fully investigated.

5.2 Materials and reagents

Chemicals were purchased from commercial sources and were used without further

purification (otherwise mentioned). C60 was purchased from IoLiTec Nanomaterials. For

reactions with fullerenes, HPLC grade solvents were used. 4-methylstyrene (Mme) and 4-

vinylbenzyl chloride (Mch) were distilled to remove inhibitor before use; 2,2′-Azobis(2-

methylpropionitrile) (AIBN) was recrystallized with 95% ethanol; toluene was stirred with

CaH2 and distilled under reduced pressure; CuBr was washed with acetic acid and ethanol.

(PPC4) and (PPCBMB) are synthesized according to literatures [135, 136] and supplied by

Belectric OPV GmbH. PC61BM/PCBM (99%) was purchased from Solenne BV. PCE11 was

synthesized according to literature [17] and supplied by South China University of

Technology.

55

5.3 Characterization equipment and instrument

Analytical TLC: Merck TLC silica gel 60 F254. Column chromatography was carried out on

silica gel (Macherey-Nagel, M-N Silica Gel 60A, 230−400 mesh). NMR spectroscopy:

Bruker Avance 400 or Avance 300 spectrometer and referenced to the residual solvent signal

(1H: CDCl3, δ=7.24 ppm). The chemical shifts are given in ppm relative to tetramethylsilane

(TMS). Abbreviations: s singlet, d doublet, m multiplet, br broad. Spectra were recorded at

room temperature.

5.4 Synthesis of side-chain polymeric fullerene derivatives

(MPMCPS)

Figure 5-1. Synthesis scheme of side-chain polymeric fullerene derivative (MPMCPS)

Step 1 Tradtional copolymerization of monomers

1.107 g Mch and 2.000 g Mme weighed with schlenk tube, and 0.0198 g AIBN with a 20 ml

bottle, used toluene to dissolve AIBN, added in the schlenk tube, total toluene 9.7 ml,

degassed with 4 freeze-pump-thaw circles, back filled with agron, light yellow transparent

solution. The mixture was stirred at 75 °C for 29 h, precipitated in 0 °C methanol, stirred for

56

1 h, filter under reduced pressure, washed with methanol 3 times (20 ml × 3), dried at 40 °C

under reduced pressure. 1.76 g white powder acquired, named P(Mme-Mch), yield 58%,

Mn=19600, PDI=1.79. 1H NMR (CDCl3): δ 1.11–2.40 (m, hydrogen in the backbone), 4.25–

4.75 (s, 2H, –CH2Cl), 6.10–7.10 (m, –CHPh) (Figure 5-2).

Step 2 Funcionalization with azide group

1.000 g P(Mme-Mch),0.456 g NaN3, 10 ml DMF added in 50 ml three-neck flask, stirred at

30°C under argon for 24 h. precipitated in 100 ml methanol/water (19/1) mix solvents,

filtered under reduced pressure, washed with methanol/water (19/1) mix solvents, dried at

45°C under reduced pressure. 860 mg white powder gained, named P(Mme-Ma) , yield 86%,

Mn=21400, PDI=1.77. 1H NMR (CDCl3): δ 1.11–2.40 (m, hydrogen in the backbone), 4.08–

4.23 (s, 2H, –CH2N3), 6.10–7.10 (m, –CHPh) (Figure 5-2).

Figure 5-2. 1H NMR spectral of P(Mme-Mch) and P(Mme-Ma)

From the blue line, the peak δ = 4.51 ppm represents hydrogen in –CH2Cl; from the black

line, the peak δ = 4.51 ppm completely disappears, meanwhile the peak δ = 4.21 ppm

representing hydrogen in − CH2N3 appears, which means chlorine group totally tuned into

azide group. By comparing the peak area ratios of peaks at δ 4.51 ppm (representing unit Ma)

and δ 6.1−7.1 ppm (standing for unit Ma and Mme), the monomer ratio of copolymers are

57

identified. Unit Ma to unit Mme is 34 to 66 (molar ratio), that is, of 100 monomer unit exists

34 azide group.

Step 3 and Step 4 were synthesized according to literatures.[137, 138] Product attained was

named methyl prop-2-yn-1-yl malonate C61 (MPMC).

Step 5 Attaching fullerene derivative to the polymer

150 mg (1.1eqa) fullerene derivative (MPMC), 60.5 mg (1eqa) P(Mme-Ma), 24.6 mg (1.1eqa)

CuBr, and 7.4 ml anhydrous toluene were added to 250 ml two-neck flask, degassed through

3 freeze-pump-thaw circle. 29.7 mg (1.1eqa) PMDETA mixed with 1 ml toluene was added

using syringe. The whole system degassed one more time. The mixture was stirred at 30 °C

for 24 h.

A thin layer chromatography (TLC) was performed with toluene/n-hexane (6/4) mix solvent.

By comparing the spots of raw material MPMC and the product, a new compound showed.

The majority of the product was the new compound which might be the fullerene grafted

polymer, methyl prop-2-yn-1-yl malonate C61 loaded polystyrene derivative (MPMCPS).

After column chromatography purification, from FTIR spectra (Figure 5-3) the absorption

around 2100 cm-2 representing the azide group disappears, which shows evidence of the

reaction between azide group and alkyne group.

58

Figure 5-3. FTIR spectra of P(Mme-Ma), MPMC and MPMCPS

5.5 Application of MPMCPS in solar cell devices

Solubility of MPMCPS and MPMC

The solubility of MPMCPS is very low in almost all the solvents, like toluene, CF, DMF, CB,

and oDCB. The best among the above solvents is oDCB. Although high temperature as

120 °C and ultrasonic bath are employed, the solubility of MCPS in oDCB is only 10mg/7.5

ml (1.3 mg/ml). The solubility of MPMC is quite high in CB (at least 10 mg/ml) and oDCB

(at least 20 mg/ml).

MPMCPS acted as acceptor in solar cell

Device structure is ITO/ PEDOT: PSS/ P3HT: MPMCPS/ Ca/ Ag. The blading temperature

of active layer is 92 °C. Due to the low total concentration of the active layer solution, all the

films are thin, which can be noticed by eyes. The total concentration of the solution matters

the most concerning the thickness of the films obtaining from Doctor Blade. While Run 1

3500 3000 2500 2000 1500 1000 500

Wavenumber (cm-1)

P(Mme-Ma) MPMCPS MPMC

alkyne group

N3

59

(Table 5-1) doctor-blade with the highest concentration, Film of Run 1 bears the darkest

color among the 6 films; Film of Run 3 is thicker than Film Run 2; with increasing number of

layer of active layer, the thickness of Film of Run 4 to Film of Run 6 increases.

Table 5-1 Experimental method Run D: A

(wt:wt)

Velocity (mm/s) Layers of active layer

1 1: 0.5 70 1

2 1: 1 50 1

3 1: 1 70 1

4 1: 2 70 1

5 1: 2 70 2

6 1: 2 70 3

From Figure 5-4 and Figure 5-5, the JSC of the solar cells is incredibly low, which might be

the main reason of the extremely low PCE; The series resistance is quite high from the J-V

curves; To increase the thickness of active layer by double or triple layers is not working.

60

Figure 5-4. Trend of photovoltaic parameters of all devices of Table 4-1

Figure 5-5. Current density-voltage characteristics (best PCE) under 100 mW cm-2

illumination (AM1.5G) (L) and in the dark (R)

Figure 5-6. Optical microscopic image (a) of MPMCPS and (b) of P3HT: MPMCPS films

61

The MPMCPS pristine and blended films are characterized under optical microscopy. As

depicted in Figure 5-6, big crystals, in the scale of micrometer, are observed in both pristine

MPMCPS and P3HT: MPMCPS blend films. Raman spectroscopy is employed to further

characterize the properties of the big crystals in the blend film.

Figure 5-7. Raman spectra (a) of P3HT and (b) of MPMCPS; (c) integrated optical

microscopic image of P3HT: MPMCPS blend films; (d), (e) and (f) are the Raman spectra of the position 1, 2 and 3 in (c).

The wavelength of the excitation laser of Raman spectroscopy is 532 nm. The laser spot is

around 100 nm2. The pristine MPMCPS and P3HT films are dropcasted on glass, while the

P3HT: MPMCPS blend film is bladed on PEDOT:PSS pre-coated ITO substrate. Through the

integratd optical microscpocy, an area from the center of a big crystal to outside of the crystal

is chosen to perform Raman spectroscopic mapping (Figure 5-7c). The scan area is

4.050×0.723 µm2; the scanning process is 8 lines per image and 40 points per line. The

spectra inside the aggregation/crystal circle are completely the same as pristine MPMCPS

spectrum; likewise, the homogenous part shows Raman spectra only exhibit the features from

pristine P3HT. From the edge of the aggregation circle to inside, position 1 to position 3,

different types of spectra are attained. All spectra from the area show both features from

62

pristine P3HT and MPMCPS, with more amount of P3HT in position 1 and gradually less

P3HT closing to the crystal.

MPMCPS acted as additive in solar cell

The devices’ structure is ITO/ PEDOT: PSS/ P3HT: PCBM: MPMCPS / Ca/ Ag. With o-

DCB as solvent, the concentration of P3HT, PCBM and MPMCPS is 20 mg/ml, 20 mg/ml

and 2 mg/ml, respectively. The active layer is doctor bladed in air under 92 °C and annealed

at 140 °C for 5 min inside the glovebox. The more detailed experimental parameters of the

active layer are summarized in Table 5-2.

Table 5-2 Experimental parameter of the active layer Run P3HT

(µl)

PCBM

(µl)

MPMCPS

(µl)

P3HT:C60:MPMCPS

(wt ratio)

Velocity

(mm/s)

1 120 120 0 1:1:0 25

2 200 200 20 1:1:0.01 25

3 200 200 40 1:1:0.02 30

4 200 200 80 1:1:0.04 30

5 200 200 120 1:1:0.06 35

6 100 100 100 1:1:0.1 35

63

Figure 5-8. Photovoltaic performance, Current density-voltage characteristics under 100 mW

cm-2 illumination (AM1.5G) and in the dark of solar cells

From Figure 5-8, small amount addition (0.5% - 3%) of MPMCPS in the P3HT: PCBM

blends leads to slight increase in current density while results in the more decrease of fill

factor; further the open circuit voltage of the solar cells keeps constant. From the dark J-V

64

curve, the addition of MPMCPS causes undesirable leakage.

5.6 Application of main-chain polymeric fullerene derivatives in

solar cell devices

Photovoltaic performance of solar cells with main-chain polymeric fullerenes as additive

The chemical structure of PPC4 and PPCBMB is presented in Figure 5-9. The devices’

structure is ITO/ ZnO/ PCE11: PCBM: poly-fullerenes / MoOx/ Ag. With CB:o-DCB:DPE

(6:4:0.03) as solvent, the concentration of PCE11 and PCBM is 15 mg/ml and 22.5 mg/ml,

respectively. The active layer is spin coated inside the glovebox. The detail experimental

parameters of the active layer are summarized in Table 5-3.

Table 5-3 Photovoltaic performance of solar cells Run Additive

(wt%)

Pre-ann.

JSC

[mA/cm2]

VOC

[V]

FF

[%]

PCE

[%]

1 - - 14.8 0.74 65 7.1

2 - 135 °C, 2 min 12.2 0.79 60 5.8

3 5% PPC4 - 15.4 0.75 66 7.6

4 135 °C, 2 min 13.6 0.79 59 6.4

5 20% PPC4 - 14.8 0.77 59 6.7

6 135 °C, 2 min 11.9 0.80 50 4.8

7 5% PPCBMB - 15.0 0.76 66 7.5

8 135 °C, 2 min 12.6 0.79 55 5.5

9 20% PPCBMB - 14.1 0.79 64 7.1

10 135 °C, 2 min 15.3 a) 0.80 a) 53 a) 6.4a)

a) Data was from degradation device measured under white LED at around 1 sun

From Table 5-3, 5 wt% of both poly fullerenes would not influence the photovoltaic

performance of the solar cells; however, the addition of 20 wt% of PPC4 would result in

slightly lower current density and fill factor, thus lower PCE. Pre-annealing of the active

layer films under inert atmosphere at 135 °C for 2 min led to impaired current density, fill

65

factor and PCE with or without poly fullerenes. The enhanced VOC of the ternary solar cells

was probably due to the higher LUMO of the poly fullerenes which are bis-adducted or tri-

adducted fullerenes[47].

Figure 5-9. Chemical structure of PPC4 and PPCBMB

Photo-stability of solar cells with main-chain polymeric fullerenes as additive

From literature[47], PPC4 can greatly stabilize the thermal stability of solar cells by adding

21 wt% of PPC4 into P3HT:PCBM solar device. The photo-stability of the PCE11: PCBM:

poly-fullerenes solar cells were investigated under white LED at inert atmosphere. From

Figure 5-10, the light-induced burn-in loss was serious in the first 30 minute; after around 20

hours, the solar cells became stable. Both PPC4 and PPCBMB could not stabilize the studied

system; the JSC burn-in loss of solar cells with 20 wt% polymeric fullerenes was slightly

larger than that 8 wt% addition of polymeric fullerenes.

66

Figure 5-10. Photo-stability of solar cells with PPC4 and PPCBMB as additive (without ann.)

As depicted in Figure 5-11, after pre-annealing at 135°C under inert atmosphere, JSC burn-in

loss of ternary solar cells reduced slightly. However, with 20 wt% addition of polymeric

fullerenes, the JSC burn-in loss became more severe comparing to that of solar cells with only

8 wt% addition. By comparing the two groups of photo degradation experiments, the

conclusion declared was, the addition of both PPC4 and PPCBMB would not enhance the

photo-stability of the PCE11:PCBM solar cells; on the contrary, the present of the polymeric

fullerenes might introduce other photo-instable factors into the solar cell, which causes more

serious JSC burn-in loss.

67

Figure 5-11. Photo-stability of solar cells with PPC4 and PPCBMB as additive (with 135°C pre-annealing)

68

5.7 Conclusion

A polystyrene backbone based side-chain polymeric fullerenes was successfully synthesized

and applied in solar cell. However, limiting by the strong aggregation and extremely low

solubility in common organic solvents, MPMCPS based binary and ternary solar cells fail to

achieve desirable performance, thus, the degradation test did not run on the MPMCPS based

solar cells.

Two main-chain polymeric fullerenes, PPC4 and PPCBMB, were employed as additives in

PCE11:PCBM based solar cells. With up to 8 wt% addition, the efficiency of the ternary

solar cells stay as high as the binary solar cells; however, with 20 wt% addition, the

photovoltaic performance already slightly decreased; pre-annealing at 135 °C for 2 min under

inert atmosphere would lead to reduced photovoltaic performance; the addition of the main-

chain polymer could not increase the photo-stability of the PCE11:PCBM based solar cells.

From these preliminary studies, certain knowledge about the limitation of the polymeric

fullerenes was gained: side-chain polymeric fullerenes normally suffer from the poor

solubility which limits the role of acting the main acceptor in the photovoltaic application;

polymeric fullerenes normally cannot attain comparable performance alone with the polymer

donor, and can only act as the secondary acceptor in the application of solar cells.

69

Chapter 6 Correlation between miscibility, thermal-stability and optoelectronic properties Abstract

In the chapter we demonstrate binary organic solar cells based on PTB7-Th:fullerene and

pDPP5T-2:fullerene composites with decent photovoltaic performance and extraordinary

high thermal stability. We further in-depth investigate the carrier dynamics along with

structural evolution and analyze the acceptor loadings in optimized bulk-heterojunction

(BHJ) solar cells as a function of the polymer-fullerene miscibility. The polymer-fullerene

miscibility has more influential effects than crystallinity of single components on the

optimized acceptor:donor ratio in polymer-fullerene solar cells.

Part of this chapter has been published (reproduced with permission from Copyright Wiley-

VCH Verlag GmbH & Co. KGaA):

[1] C. Zhang, S. Langner, A.V. Mumyatov, D.V. Anokhin, J. Min, J.D. Perea, K.L. Gerasimov, A. Osvet, D.A. Ivanov, P. Troshin, N. Li, C.J. Brabec, Understanding the correlation and balance between the miscibility and optoelectronic properties of polymer-fullerene solar cells, Journal of Materials Chemistry A, 5 (2017) 17570-17579.

[2] C. Zhang, A. Mumyatov, S. Langner, J.D. Perea, T. Kassar, J. Min, L.L. Ke, H.W. Chen, K.L. Gerasimov, D.V. Anokhin, D.A. Ivanov, T. Ameri, A. Osvet, D.K. Susarova, T. Unruh, N. Li, P. Troshin, C.J. Brabec, Overcoming the Thermal Instability of Efficient Polymer Solar Cells by Employing Novel Fullerene-Based Acceptors, Advanced Energy Materials, 7 (2017) 1601204.

Part of the results presented in the chapter was done by collaborators:

PyF5 and FAP1 were synthesized by Prof. Pavel Troshin’ group in Skolkovo Institute of

Science and Technology, Skolkovo Innovation Center, Nobel st. 3, Moscow, 143026,

Russian Federation. The GIWAXS/GISAXS patterns were done by Thaer Kassar in the

Institute of Crystallography and Structural Physics, FAU, Germany and by Denis V.

Anokhin and his collegues in Institute for Problems of Chemical Physics of Russian

Academy of Sciences, Semenov Prospect 1, Chernogolovka, 142432, Russia.

The Hansen solubility parameters of polymers and fullerenes were measured by Stephan

Langner and calculated by Jose Perea respectively.

70

6.1 Introduction

Organic photovoltaics is one of the most promising technologies for sustainable green energy

supply. Because of their high electron affinity and superior electron-transporting ability,

fullerene-based materials are deemed as very strong electron-accepting components in

organic solar cells. However, the most widely used fullerene-based acceptors, such as phenyl-

C61-butyric acid methyl ester, exhibit limited microstructure stability and unsatisfying

thermal stability in combination with many of the state-of-the-art organic donors due to

microstructural incompatibilities and subsequent demixing. Rational design rules for novel

fullerenes with enhanced compositional miscibility are employed to develop novel fullerenes

overcoming this limitation. Miscibility is proposed in addition to crystallinity as a further

design criterion for long lived and efficient solar cells.

6.2 Materials and device fabrication

ZnO-nanoparticle dispersion in isopropyl alcohol was received from Nanograde. pDPP5T-2

(batch No. : GKS1-001) was provided by BASF. PTB7-Th and PC61BM (99%) were

purchased for One Material and Solenne BV, respectively. PyF5 and FAP1 were synthesized

according to literatures.[139, 140] The structures of the active layer materials used in this

chapter are summarised in Figure 6-1.

All solar cells were fabricated in an inverted structure of ITO/ZnO/Active layer/MoOx/Ag, as

shown in Figure 6-1. The pre-structured ITO coated glass substrates were subsequently

cleaned in toluene, acetone and isopropyl alcohol for 10 min each. Then 30 nm ZnO

(Nanograde, N-10) were doctor bladed on the ITO substrate and annealed at 85 °C in ambient

air. The PTB7-Th based active layer from chlorobenzene: diphenylether (97:3) solution was

spin coated on top of ZnO in nitrogen atmosphere. The pDPP5T-2 based active layer from

chloroform: 1,2-dichlorobenzene (90:10) solution was bladed under ambient conditions. 10

nm MoOx and 100 nm silver were deposited subsequently under 6 × 10-6 Torr by thermal

evaporation through a shadow mask to form an active area of 10.4 mm2. For SCLC devices,

pristine fullerene layer was spin coated under ambient atmosphere. BHJ active layer was

71

processed the same as solar cells. 15 nm calcium and 100 nm silver were deposited

subsequently under 6 × 10-6 Torr by thermal evaporation through a shadow mask to form an

active area of 10.4 mm2.

Figure 6-1. a) Schematic device architecture of organic solar cells; b) Chemical structures of donor and acceptor materials.

6.3 Optimization of D: A ratio of BHJ devices

Figure 6-2 depicts the normalized absorption of pristine PTB7-Th, PC61BM, Py5 and FAP1.

The absorbing region of the three fullerenes mainly locates at around 320 nm while PTB7-Th

has a strongly absorbing area between 500 and 800 nm. The photoluminescence (PL) peak of

PTB7-Th appears at ~830 nm. When blended with PC61BM (donor (D): acceptor (A) =1.5:1),

PyF5 (D:A =1:1 or 1:3), or FAP1 (D:A =1:1 or 1:3), the PL emission of PTB7-Th is

significantly quenched by over 98% for all blends. These phenomena suggest that exciton

dissociation/charge separation process is highly efficient for PTB7-Th:PyF5 and PTB7-

Th:FAP1 with either high or low fullerene loadings.

72

Figure 6-2. Absorption of pristine materials and photoluminescence of polymer and blends

The photovoltaic characteristics of PTB7-Th:fullerene with different donor:acceptor ratios are

summarized in Figure 6-3 and Table 6-1. It is important to point out that PC61BM instead of

PC71BM was employed for reference devices to assure rational comparison with the two C60-

based fullerenes derivatives. The optimized photovoltaic performance of PyF5 and FAP1

based devices is slightly lower than that of PC61BM based devices, owing to the slightly

lower fill factor and JSC (Figure 6-4).

73

Figure 6-3. (a) Current density-voltage (J-V) characteristics of PTB7-Th:fullerenes organic solar cells; (b) PCE and (c) fill factor of the corresponding organic solar cells as a function of D:A ratios.

The solar cells based on PTB7-Th:PyF5 attained an average PCE of 6.5 % along with an

enhanced VOC of 0.84 V, which is significantly higher than that (0.80 V) of the PTB7-

Th:PCBM control devices. The higher VOC observed in the PyF5-based solar cells is

consistent with the higher LUMO of PyF5.[139] PTB7-Th:FAP1 solar cells display an

average PCE of 6.1% and again a significantly higher VOC of 0.87 V, indicating that the

bandgap to voltage loss is minimized by replacing PCBM with either of the two novel

fullerene derivatives.

(a)

(b) (c)

74

Table 6-1. Photovoltaic properties of PTB7-Th: fullerenes organic solar cells Active layer D:A

ratios

VOC

[V]

JSC

[mA cm-2]

FF

[%]

PCE

[%]

PTB7-Th : PC61BM 1:0.8 0.80±0.04 11.8±2.7 44±0.7 4.1±0.7

1:1 0.81±0.02 14.3±1.3 57±1.5 6.6±0.6

1:1.5 0.80±0.01 14.0±0.6 65±0.7 7.3±0.3

1:2 0.80±0.01 12.6±0.3 68±0.5 6.9±0.2

1:3 0.77±0.01 11.1±0.4 70±1.2 6.0±0.1

PTB7-Th : PyF5 1:0.8 0.84±0.01 8.9±0.6 36±0.4 2.7±0.1

1:1 0.84±0.01 14.6±0.6 45±1.1 5.6±0.3

1:1.5 0.84±0.01 13.8±1.1 53±1.0 6.2±0.5

1:2 0.84±0.01 13.7±0.6 56±0.6 6.5±0.5

1:3 0.83±0.01 11.2±0.3 62±0.8 5.8±0.1

PTB7-Th: FAP1 1:0.8 0.81±0.01 5.7±0.3 37±0.3 1.7±0.1

1:1 0.83±0.01 11.5±0.4 38±1.2 3.6±0.3

1:1.5 0.86±0.01 12.9±0.2 44±1.1 4.9±0.1

1:2 0.87±0.01 12.7±0.5 55±0.3 6.1±0.3

1:3 0.86±0.01 11.2±0.2 61±0.1 5.9±0.1

Figure 6-4. EQE spectra (a) of PTB7-Th:fullerene and (b) of pDPP5T-2:fullerne solar cells

As shown in Figure 6-5 and Table 6-2, pDPP5T-2:PyF5 and pDPP5T-2:FAP1 solar cells

exhibit comparable photovoltaic performance to the pDPP5T-2:PCBM control devices.

75

Again, given that both fullerenes are mono-substituted, impressively high VOC values of 0.70

V and 0.71 V were achieved for the low bandgap polymer donor in combination with PyF5

and FAP1, respectively.

As observed in Figure 6-3 and Figure 6-5, when the acceptor loading is low (D:A = 1:0.8),

JSC, fill factor as well as power conversion efficiency are much lower than the optimized

device performance for each corresponding system. With increasing the fullerene loadings,

the photovoltaic parameters of the three systems have a trend to increase. In PC61BM-based

BHJ systems, PCE reaches a summit at a D:A ratio of 1:1.5. Further increasing the ratio

beyond 1:1.5 only moderately increases the fill factor. In the case of PTB7-Th: PyF5 and

PTB7-Th:FAP1, PCE peaks at a D:A ratio of 1:2. Further increasing the fullerene ratio slowly

enhances the fill factor beyond 60% at a D:A ratio of 1:3. It is noticed that, at the D:A ratio of

1:3, when the fill factor of PyF5 and FAP1 based devices is above 60%, JSC and PCE are

almost the same for all three systems; Similar phenomena are observed for pDDP5T-2:PyF5

and pDDP5T-2:FAP1 solar cells.

Table 6-2. Photovoltaic properties of pDPP5T-2 : fullerenes organic solar cells Active layer Weight

ratios

VOC

[V]

JSC

[mA/cm2]

FF

[%]

PCE

[%]

pDPP5T-2 : PyF 5 1:0.8 0.67±0.01 4.9±0.1 38±0.5 1.2±0.1

1:1 0.66±0.01 6.1±0.1 43±0.5 1.7±0.1

1:1.5 0.66±0.01 8.5±0.5 48±2.0 2.7±0.2

1:2 0.67±0.01 9.5±0.4 52±1.4 3.3±0.1

1:3 0.65±0.01 7.8±0.3 60±0.6 3.0±0.1

pDPP5T-2 : FAP1 1:0.8 0.65±0.01 6.8±0.5 43±2.1 1.9±0.1

1:1 0.66±0.01 7.1±0.5 44±0.7 2.1±0.2

1:1.5 0.67±0.01 9.5±0.3 51±2.0 3.2±0.3

1:2 0.67±0.01 9.3±0.4 57±1.5 3.6±0.1

1:3 0.66±0.01 8.4±0.1 60±0.2 3.4±0.2

76

Figure 6-5. a) Current density-voltage (J-V) curves of pDPP5T-2:PyF5 and pDPP5T-2:FAP1; PCE (b) and FF (c) as a function of D:A ratios.

6.4 Thermal stability of fullerene-based BHJ films

As shown in Figure 6-6, Table 6-3 and Table 6-4, excellent thermal stability was found for

the PyF5- and FAP1-based solar cells. The PTB7-Th:PCBM control devices showed rapidly

decreased photovoltaic performance upon annealing and maintained only 19% of initial

performance after baking at 140 °C for 24 hours, while the photovoltaic performance of

PyF5- and FAP1-based devices remained almost unchanged under the same conditions. The

performance of pDPP5T-2:PyF5 and pDPP5T-2:FAP1 solar cells based on the two novel

fullerene acceptors remained ~100% after baking at 140°C for 24 hours, while the PCBM-

based solar cells maintained only 31% of initial PCE under the same conditions.

77

Figure 6-6. Normalized photovoltaic performance (a) of PTB7-Th:fullerene and (b) of

pDPP5T-2:fullerne organic solar cells as a function of annealing time at 140 °C under inert atmosphere (all the devices were annealed prior to the deposition of top electrode).

78

Table 6-3 Normalized photovoltaic performance of PTB7-Th:fullerene organic solar cells as

a function of annealing time at 140 °C under inert atmospherea)

Active layer Annealing time

[h]

VOC

[a.u.]

JSC

[a.u.]

FF

[a.u.]

PCE

[a.u.]

PTB7-Th:PCBM 0 1 1 1 1

0.5 1.04 0.94 0.87 0.85

2 1 0.39 0.58 0.23

24 0.94 0.31 0.61 0.19

PTB7-Th:PyF5 0 1 1 1 1

0.5 1.02 1.02 0.93 0.98

2 1.04 1.02 0.95 1

24 1.02 0.98 0.96 0.97

PTB7-Th:FAP1 0 1 1 1 1

0.5 1.01 0.93 1.06 0.98

2 1.02 0.96 1.04 1.02

24 1.01 0.93 0.98 0.93

a) Values are from Figure 6-6a.

79

Table 6-4 Normalized photovoltaic performance of pDPP5T-2:fullerene organic solar cells as

a function of annealing time at 140 °C under inert atmosphere Active layer Annealing time

[h]

VOC

[a.u.]

JSC

[a.u.]

FF

[a.u.]

PCE

[a.u.]

pDPP5T-2:PCBM 0 1 1 1 1

0.5 1.07 0.89 0.97 0.94

2 1.07 0.56 0.85 0.50

24 0.95 0.37 0.87 0.31

pDPP5T-2:PyF5 0 1 1 1 1

0.5 1.04 1.04 1.05 1.17

2 1.07 0.91 1.11 1.07

24 1.07 0.85 1.15 1.02

pDPP5T-2:FAP1 0 1 1 1 1

0.5 1.06 1 1.05 1.09

2 1.07 0.94 1.05 1.09

24 1.07 0.86 1.05 0.98

a) Values are taken from Figure 6-6b.

In order to exclude effects on D:A ratio, PTB7-Th:PCBM solar cells with D:A ratio at 1:2 are

made and tested the thermal stability under the same conditions (Figure 6-7 and Table 6-5).

After 2 hours’ annealing, the performance of the solar cells drops similarly as solar cells with

1:1.5 D:A ratio. PTB7-Th:[70]PCBM control devices are also fabricated to test the thermal

stability. The results are summarized in Figure 6-7 and Table 6-5. The photovoltaic

performance of this system is also decreasing with annealing time.

80

Figure 6-7. Current density-voltage (J-V) curves (a) and EQE spectra (b) of PTB7-Th:[60] / [70]PCBM solar cells under illumination of AM 1.5 (100 mW cm-2); Normalized

photovoltaic performance (c) of PTB7-Th:[60]/[70]PCBM organic solar cells as a function of annealing time at 140 °C under inert atmosphere (all the devices were annealed prior to the

deposition of top electrode).

Table 6-5 Photovoltaic characteristics of PTB7-Th:[60]/[70]PCBM organic solar cells Active layer D:A wt

ratios

VOC

[V]

JSC

[mA/cm2]

FF

[%]

PCE

[%]

PCEa)

[%]

PTB7-Th: PCBM 1:2 0.78 13.4 66 6.9 36

PTB7Th:[70]PCBM 1:2 0.78 15.8 64 7.9 69

a) Maintained PCE of OPV devices after annealing at 140 °C for 2 hours under inert

atmosphere. Values are from Figure 6-7c.

(a)

(b)

(c)

81

As shown in Table 6-6 and Table 6-7, PTB7-Th:PyF5 system is much more stable than

PTB7-Th:PCBM system under annealing at both low temperature and high temperature. At

lower temperature which has stronger effect on composite miscibility, PTB7-Th:PyF5

composite shows superior stability over PTB7-Th:PCBM blend; at such high temperature as

200 °C which is already destroying pristine material domain, PTB7-Th:PyF5 solar cells still

preserve higher initial photovoltaic performance than PTB7-Th:PCBM cells do.

Table 6-6 Normalized photovoltaic performance of PTB7-Th:fullerene organic solar cells

under longer time annealing test at 140 °C under inert atmosphere. Active layer D:A wt

ratio

Annealing time

[h]

VOC

[a.u.]

JSC

[a.u.]

FF

[a.u.]

PCE

[a.u.]

PTB7-Th:PCBM 1:2 0 1 1 1 1

120 0.94 0.31 0.53 0.16

PTB7-Th:PyF5 1:2 0 1 1 1 1

120 1 0.82 0.81 0.70

Table 6-7 Comparison of photovoltaic performance of PTB7-Th:fullerene organic solar cells

annealing at low temperature and high temperature under inert atmospherea)

Active layer D:A wt

ratio

Annealing T

[°C]

VOC

[a.u.]

JSC

[a.u.]

FF

[a.u.]

PCE

[a.u.]

PTB7-Th:PCBM 1:2 - 1 1 1 1

140 1.03 0.95 0.86 0.84

170 1.03 0.90 0.86 0.79

200 0.99 0.90 0.75 0.67

PTB7-Th:PyF5 1:2 - 1 1 1 1

140 1.02 1.02 0.93 0.98

170 1.05 0.93 0.91 0.91

200 1.02 0.82 0.91 0.77

a) annealing time is 0.5 hours.

82

6.5 Evolution of surface morphology of BHJ films

The surface morphology of BHJ thin films is studied by optical microscopy (Figure 6-8),

measurement of water contact angle (Figure 6-8, insert; Figure 6-9: pristine films) and

atomic force microscopy (AFM) (Figure 6-10).

From optical microscopy images, big crystals were formed on the surface of PTB7-Th:PCBM

film after annealing at 140 °C for 0.5 hour. Strong phase separation was observed for the 2

hours annealed film. The contact angle of water droplet on the films were determined to be

94°, 98° and 104° for PTB7-Th:PCBM annealed at 140°C for 0, 0.5 and 2 hours, respectively.

Figure 6-8. Optical microscopy images and contact angle of water droplets of PTB7-Th:PCBM films, PTB7-Th:PyF5 films and PTB7-Th:FAP1 films.

10 µm

PTB7-Th:PCBM 0.5 h 2 h0 h

PTB7-Th:PyF5 0.5 h 2 h0 h

PTB7-Th:FAP1 0.5 h 2 h0 h

83

Figure 6-9. Contact angle of water droplets of PTB7-Th, PCBM, PyF5 and FAP1 films.

The root mean square roughness (RRMS) measured by AFM increased dramatically for PTB7-

Th:PCBM BHJ films from 1.84 nm for no annealing film, to 23.3 nm for 0.5 hours annealing

and to 48.8 nm for 2 hours annealing. For both novel fullerene-based films, no significant

changes could be observed from the optical microscopy images. The contact angle of water

droplet on the films as well as the RRMS remain almost unchanged for all the films, indicating

that no phase separation or other severe morphology changes had happened to PyF5- and

FAP1-based BHJ films during annealing. This underpins the relevance of the extraordinary

high thermal stability observed for the corresponding photovoltaic devices.

PTB7-Th:PCBM PTB7-Th:PyF5 PTB7-Th:FAP1

Figure 6-10. Surface topographic and phase AFM images (size: 3 × 3 µm) of PTB7-

Th:fullerene films. A and B: no annealing; C and D: 0.5 hour annealing; E and F: 2 hours’ annealing.

PyF5~94︒

FAP1~95︒

PTB7-Th~103︒

PCBM~94︒

RMS 1.84 nm

RMS 23.3 nm

RMS 48.6 nm

1 µmA B

C D

E F

RMS 2.02 nm

RMS 1.10 nm

RMS 1.82 nm

1 µmA B

C D

E F

RMS 1.52 nm

RMS 1.10 nm

RMS 1.10 nm

1 µmA B

C D

E F

84

6.6 Evolution of bulk morphology of BHJ films

The crystalline properties of the three fullerenes are further confirmed by GIWAXS data as

shown in Figure 6-11. The room-temperature 2D GIWAXS patterns of PCBM reveal two

broad reflections of the crystalline phase in addition to the amorphous halo. The reflections

likely correspond to poorly organized PCBM crystals. The diffractograms of PyF5 exhibit

only amorphous halo. In contrast, the GIWAXS patterns of FAP1 reveal highly oriented

narrow peaks of a crystalline phase. Consequently, the degree of ordering in the studied

acceptors is different.

Figure 6-11. 2D GIWAXS patterns of PCBM (narrow reflections in small-angle region are

attributed to scattering from heating stage), PyF5 and FAP1

PCBM PyF5 FAP1

85

Figure 6-12. 2D GIWAXS patterns of (a1 and a2) PTB7-Th:PCBM, (b1 and b2) PTB7-

Th:PyF5 and (c1 and c2) PTB7-Th:FAP1. a1, b1 and c1: no annealing; a2, b2 and c2: 2 hours’

annealing. Regions in white corresponding to the position of the two inter-module gaps of the

Pilatus detector.

Figure 6-12, 6-13 and 6-14 depict the 2D GIWAXS patterns, the corresponding in-plane and

out-of-plane cuts, and the azimuthal distribution of PTB7-Th (100) peak of three BHJ blends.

All three non-annealed BHJ films show a bimodal texture with both edge-on and face-on

populations. The crystallinity of the PTB7-Th blended with FAP1 is slightly higher than that

of PTB7-Th:PyF5 and PTB7-Th:PCBM blends. After 2 hours’ annealing, the composites with

PyF5 or FAP1 have slightly enhanced intensity of (100) peak of PTB7-Th in respect to

amorphous halo, whereas PTB7-Th:PCBM blend develops a different crystal mosaicity.

86

Figure 6-13. Out-of-plane (Qz) and in-plane (Qy) cuts of PTB7-Th:PCBM, PTB7-Th:PyF5,

and PTB7-Th:FAP1 films without annealing or with 2 hours’ annealing.

Figure 6-14. Azimuthal distribution of the polymer (100) lamellar stacking peak intensity of PTB7-Th:PCBM, PTB7-Th:PyF5, and PTB7-Th:FAP1 films

87

As depicted in Figure 6-15, pristine PTB7-Th displays an intense photoluminescence (PL)

emission peak at 780 nm while the emission in the three non-annealed BHJ films is

significantly quenched. However, the PL emission of PTB7-Th:PCBM BHJ films

significantly increases after 2 hours’ annealing, and singlet emission of PTB7-Th can be

clearly distinguished, indicating a strong separation of the donor and acceptor phases. For

both PTB7-Th:PyF5 and PTB7-Th:FAP1 systems, the PL spectra remain almost unchanged,

which is again in excellent agreement with their extraordinary high thermal stability of BHJ

solar cells. We conclude this part by stating that PCBM and PTB7-Th have a strong tendency

to phase separate on a large scale. On the contrast, both PyF5 and FAP1 maintain their good

intermixing even under thermal stress of 140°C. This is remarkable, as both fullerenes have a

comparable glass transition temperature around 100°C. Further, FAP1 is crystalline while

PyF5 is dominantly amorphous.

Figure 6-15. UV-vis absorption (a) and PL spectra (b) of materials in pristine and in blends.

6.7 Thermal behavior of materials in pristine and in blends

The thermal behavior of pristine PTB7-Th and fullerene derivatives is summarized in Figure

6-16. No endothermic peak was observed for pristine PTB7-Th between 30 °C and 300 °C,

while the pristine PCBM exhibited melting peak in the 1st and 2nd heating scans. An

exothermic peak at 168 °C was observed during the 1st heating scan of PCBM, indicating

cold crystallization.[141] In contrast to PCBM, during the same temperature range, PyF5

(a)

(b)

88

exhibits a broad endothermic feature during the 1st heating run, implying it being a

dominantly amorphous material. This is most likely because PyF5 is a mixture of two

stereoisomers due to hindered nitrogen inversion.[139] A melting peak was observed in the 1st

heating scan of pristine FAP1, indicating the crystalline-feature of FAP1. However, no cold

crystallization or crystallization was observed in the 1st heating or 1st cooling scan for FAP1.

Moreover, the glass transition temperature (Tg) of pristine PCBM, PyF5 and FAP1 were

determined to be 129 °C, 142 °C and 110 °C, respectively, as shown in Figure 6-16e.

Figure 6-16. Thermal behavior of PTB7-Th, PCBM, PyF5 and FAP1 by DSC.

The mixing behavior of polymer:fullerene composites were further experimentally analyzed

by DSC. The DSC curves depicted in Figure 6-17 are taken from their 1st heating runs. In

(a)

(d)(c)

(b)

(e)

89

Figure 6-17a, the melting peaks of PCBM in pristine and in blends are more or less located

at the same temperature (282 °C). Cold crystallization of PCBM occurred in all the examined

blends at temperatures between 140 °C to 160 °C. The cold crystallization enthalpy and the

melting enthalpy for the pristine components as well as for the blends were calculated by

integrating the peak area. For pristine PCBM, the cold crystallization at ~160°C accounts for

only 21% of the melting enthalpy, however, for the blends, 90% of the melting enthalpy is

contributed from cold crystallization. We interpret this finding that the addition of PTB7-Th

to PCBM restricts the crystallization of PCBM during drying. However, upon heating, PCBM

gains enough energy to re-arrange and transform the aggregated but disordered regions into

crystallites, resulting in cold crystallization as observed in Figure 6-17a. Figure 6-17b

depicts the DSC heating curves for PyF5 and corresponding blends. Owing to the amorphous

nature of PyF5, DSC does not give specific insight into the thermal behavior of pristine as

well as blended PyF5. FAP1 behaves different to PCBM as well as to PyF5. The endothermic

peak of pristine FAP1 is located at 269 °C (Figure 6-17c). In contrast to PCBM, no cold

crystallization is observed for either pristine FAP1 or PTB7-Th:FAP1 blends, indicating the

absence of amorphous but aggregated pristine or mixed polymer:fullerene domains. Further,

FAP1 shows a melting point depression upon addition of PTB7-Th, with the melting

temperature being reduced to 249 °C at a mixing ratio of 50:50. The occurrence of melting

point depression suggests that FAP1 forms less-perfect crystals in the presence of PTB7-Th,

which is an obvious sign of chemical similarity and preferential mixing. Different from

PCBM, FAP1 can still crystallize in the presence of high polymer loadings probably by

forming polymer inclusions in the fullerene crystallites.

Figure 6-17. Thermal properties (a) of PTB7-Th:PCBM, (b) of PTB7-Th:FAP1 and (c) of PTB7-Th:PyF5 blends.

90

6.8 Correlation of miscibility and thermal stability of BHJ films

In order to rationalize the excellent thermal stability of PyF5- and FAP1-based solar cells, we

investigated the thermodynamic properties of single components and derived the miscibility

conditions in the solid-liquid equilibrium. This was done by calculating the interaction

parameters between the polymer donor and the fullerene-based acceptors. The difference of

the interaction energy in a mixture can be specified with the dimensionless Flory-Huggins

interaction parameter (x12). The interaction parameter is commonly used for evaluating the

miscibility of organic components or of diluted solutions.[72, 142] And it is calculated as

follows:

𝑥12 = 𝑣0𝑅𝑘

(𝛿1 − 𝛿2)2 (6-1)

Where v0 is the lattice site volume and is defined by the smallest unit; δ1 and δ2 are the

Hildebrand solubility parameters of the fullerenes and the polymer, respectively; R is the

ideal gas constant and T the temperature.

The Hansen solubility parameters (δd, δp, and δhb) of PTB7-Th was determined via the binary

solvent gradient method, which was employed to probe the surface of the Hansen sphere for a

set of four different solvent mixtures.[76, 77] First the solubility of each polymer was

measured stepwise from good solvent to non-solvent. Therefore, chlorobenzene was

employed as good solvent, while acetone, propylene carbonate, 2-propanol and cyclohexane

were used as non-solvents (low solubility of the polymers). Because of different weak forces

of the non-solvents (propylene carbonate highly polar or cyclohexane less polar), blends with

altered interaction relative to the solute are created. This results in a controlled change in

solubility (Figure 6-18). Next, the Hansen solubility parameters of each solvent blend were

calculated by following equation:

𝐻𝑆𝑃𝑏𝑏𝑏𝑛𝑑 = 𝜙𝑆1 ∙ 𝐻𝑆𝑃𝑆1 + 𝜙𝑆2 ∙ 𝐻𝑆𝑃𝑆2 (6-2)

with ϕS1 and ϕS2 as the volume fraction of chlorobenzene and non-solvent, respectively. This

allows us to transfer the solubility data into HSP data, which are then plotted in the Hansen-

space. By using a solubility limit of 10 mg mL-1, a 0-1 scoring of the HSP data was made,

91

whereby blend with higher solubility were marked as 1, otherwise 0. Finally a sphere fit was

performed by the software HSPiP. The program evaluates the input data using a quality-of-fit

function with the form:

DATAFIT = (𝐴1𝐴2 ⋯𝐴2)1 𝑛⁄ (6-3)

With n as the number of solvents and

𝐴𝑚 = 𝑒−(𝑏𝑒𝑒𝑒𝑒 𝑑𝑚𝑠𝑡𝑑𝑛𝑐𝑏)𝑖 (6-4)

where the error distance is the distance of the solvent in error to the sphere boundary. [76]

The center of the sphere represents then the Hansen solubility parameters of the polymers.

Figure 6-18. (a) Solubility of PTB7-Th in mix solvents using chlorobenzene as the good solvent; (b) Hansen Space and a sphere-fit matching the solubility limit of 10 mg mL-1 of PTB7-Th.

Table 6-8. Hansen solubility parameters (δd, δp, and δhb) and Hildebrand solubility parameter

(δΤ) of PTB7-Th, pDPP5T-2, PCBM, FAP1and PyF5

δd (MPa1/2) δp (MPa1/2) δhb (MPa1/2) δΤ (MPa1/2)d)

PTB7-Th a) 18.56 2.30 3.21 18.98

pDPP5T-2 b) 19.00 3.00 2.00 19.34

PCBMc) 20.60 4.93 4.23 21.60

[70]PCBMc) 20.95 2.80 1.64 21.20

PyF5d) 20.92 2.15 0.19 21.03

FAP1d) 20.75 2.60 0.33 20.91

(a) (b)

92

a) measured; b) data from reference;[143] by COMSO-RS;c) data from reference.[144]

e) 𝛿𝑘 = 𝛿𝑑2 + 𝛿𝑝2 + 𝛿ℎ𝑏22

The Hansen solubility parameters of polymers and fullerenes are summarized in Table 6-8.

The interaction parameters between the polymers and the fullerenes are summarized in

Figure 6-19 and Table 6-9. It is important to note that the values were estimated under the

same conditions and we concentrate our discussion on the relative trend given by the

interaction parameters rather than on the absolute values. Besides, since the Flory interaction

parameter is defined in terms of energies per site it is proportional to the site volume v0. The

site volume must be specified whenever interaction parameter is discussed.[72] Here we

decide to give the value of x12 in multiples of v0 for the purpose of general comparison. The

interaction parameter of PTB7-Th:PCBM is higher than that of PTB7-Th:PyF5 and PTB7-

Th:FAP1, which indicates that PyF5 and FAP1 are more miscible with PTB7-Th than PCBM

is. The same trend is observed for pDPP5T-2 mixtures, meaning better miscibility between

pDPP5T-2 and PyF5 or FAP1 than between pDPP5T-2 and PCBM. A clear trend is observed

between enhanced thermal stability and better polymer / fullerene miscibility.

Figure 6-19. Interaction parameters of of PTB7-Th:PCBM, PTB7-Th:FAP1 and PTB7-Th:PyF5 blends.

0.0

1.0E-3

2.0E-3

3.0E-3

PTB7-Th:FAP1PTB7-Th:PyF5

PTB7-Th:PCBM

X 12/v

0

93

Table 6-9 Interaction parameters between PTB7-Th, pDPP5T-2 and PCBM, PyF5, FAP1 Polymer:fullerene 𝑥12 𝑥12/𝑣0

PTB7-Th:[70]PCBM 1.99·10-3v0 1.99·10-3

PTB7-Th: PCBM 2.77·10-3v0 2.77·10-3

PTB7-Th: PyF5 1.69·10-3v0 1.69·10-3

PTB7-Th: FAP1 1.51·10-3v0 1.51·10-3

pDPP5T-2: PCBM 2.06·10-3v0 2.06·10-3

pDPP5T-2: PyF5 1.15·10-3v0 1.15·10-3

pDPP5T-2: FAP1 1.00·10-3v0 1.00·10-3

Combining the information on the interaction parameters together with the experimentally

measured DSC results, we conclude that the crystallization of PCBM is hindered in the

presence of PTB7-Th; a metastable solid state composite is formed under processing

condition, however, PCBM has rather low miscibility with the polymer. Although entropy

favors the formation of mixed aggregated regions, thermodynamics drives PCBM and PTB7-

Th into phase separation. As a consequence, when sufficient energy is achieved (e.g. by

heating, light excitation, etc.), PCBM is driven to phase separate from the polymer matrix,

causing large scale phase separation as evidenced by the deteriorated photovoltaic

performance and optical microscopy images. PyF5 is chemically more miscible with the

PTB7-Th due to the existence of the branched alky methoxy groups. PyF5 forms more

intimately mixed phases with PTB7-Th and pDPP5T-2 than PCBM does. Moreover, unlike

PCBM, PyF5 is rather amorphous and does not form ordered structure within/outside the

polymer matrix. To our surprise, the semi-crystalline FAP1 is also more chemically

compatible with PTB7-Th and pDPP5T-2 than PCBM. In contrast to PCBM, FAP1 does not

exhibit cold crystallization during heating, underpinning that the driving force for phase

separation is less than PCBM. The presence of PTB7-Th does not dramatically suppress

FAP1 crystallization, as the higher miscibility allows the formation of “impure” fullerene

crystals, again evidencing the enhanced compatibility between these two compounds.

94

6.9 Correlation of miscibility and optoelectronic properties of

solar cells

From our study, PyF5 and FAP1 are more miscible with PTB7-Th than PCBM, which

contributes to the excellent blend film thermal stability. However, we also notice that the

devices fabricated with the novel fullerene acceptors show a slightly low JSC and FF. As far

as we understand, the bi-continuous percolation pathways for transporting charge carriers are

not easy to form when the two components have good miscibility, which leads to slightly low

Jsc and FF for OPV devices based on functional fullerene acceptors. This finding shows a

clear direction to the future design of novel acceptors and donors with proper miscibility and

opto-electronic properties.

Carrier dynamics along with structural evolution

Space-charge-limited current (SCLC) measurements were performed to study the charge

carrier transport properties of fullerene:polymer blends for various fullerene loadings. The

electron mobility was estimated by fitting the current-voltage curves according to the SCLC

modified Mott-Gurney model (Figure 6-20).

Figure 6-20. The dark J-V characteristics of (a) PCBM, PyF5 and FAP1 pristine electron-only devices and (b-g) PTB7-Th: fullerenes electron-only devices. The solid line represent the best fitting using the space-charge-limited current (SCLC) modified Mott-Gurney model.

95

As shown in Figure 6-21 and Table 6-10, pristine PyF5 and FAP1, demonstrate high electron

mobility of 4.0×10-3 cm2 V-1 s-1 and 4.4×10-3 cm2 V-1 s-1, respectively, which are very similar

as determined for PC61BM (4.5×10-3 cm2 V-1 s-1), suggesting that electron transporting

properties of the functional fullerenes are very comparable to the ones of PC61BM. However,

the electron mobilities of PC61BM-based blends remain high (~10-4 cm2 V-1 s-1) in

composites with up to 50 wt.-% polymer loading. Such high electron mobility in blends is

essential to guarantee a high fill factor and a correspondingly high efficiency. In stark

contrast, the electron mobility of PTB7-Th:PyF5 (1:1) and PTB7-Th:FAP1 (1:1) BHJ systems

was significantly reduced to the level of 10-7 - 10-6 cm2 V-1 s-1. This is associated with the

rather low FF observed for such low fullerene loadings (Table 6-1). When more PyF5 or

FAP1 was added to the blends, the electron mobility was improved by two orders of

magnitude to 8.8×10-5 cm2 V-1 s-1 for PTB7-Th:PyF5 (1:3) and 9.3×10-5 cm2 V-1 s-1 for PTB7-

Th:FAP1 (1:3). This is in good agreement with the comparable JSC and PCE for all three

systems.

Figure 6-21. SCLC electron motilities of PTB7-Th:fullerene as a function of fullerene content. The dashed lines are a guide to the eye.

100 90 80 70 60 5010-7

10-6

10-5

10-4

10-3

10-2

PCBMPyF5FAP1

Elec

tron

mob

ility

(cm

2 V-1 s

-1)

Fullerene content (wt%)

96

Table 6-10. Electron mobilities of pristine fullerenes and PTB7-Th: fullerene composites

determined from SCLC measurements Materials Weight

ratios

Electron mobility

[cm2 V-1 s-1]

PTB7-Th:PCBM 0:1 4.5×10-3

1:1.5 9.0×10-5

1:1 9.8×10-5

PTB7-Th : PyF5 0:1 4.0×10-3

1:3 8.8×10-5

1:1 8.5×10-7

PTB7-Th : FAP1 0:1 4.4×10-3

1:3 9.3×10-5

1:1 4.4×10-7

We note that for PTB7-Th:PyF5 and PTB7-Th:FAP1 blends significantly higher fullerene

loadings are necessary to reach comparably high electron mobilities as for PTB7-Th:PC61BM

blends. Surprisingly, we interpret these findings that, the crystallinity of a fullerene acceptor

is not the sole material property dominating electron mobility of a BHJ composite. Our

studies demonstrate that the amorphous-dominated fullerene acceptor PyF5 has the same

mobility as the semi-crystalline fullerene acceptor FAP1. When blended with PTB7-Th, at

low fullerene loadings, electron mobility drops dramatically for both fullerene composites.

Upon increasing the fullerene loading, mobility increases for both, the amorphous as well as

the crystalline fullerene. This observation leads us to the conclusion that crystallinity per se is

an important material parameter but probably not the decisive parameter guaranteeing good

solar cell performance. This trend is best documented by Figure 6-21, where electron

motilities of PTB7-Th:fullerene as a function of fullerene content is presented. The much

distinct electron mobility delivered by the polymer:fullerene blends can be mainly attributed

to the different microscale BHJ morphology.

Morphology evolution of polymer-fullerene BHJ blends

97

In principle, there are three phases in OPV BHJ blends: pristine donor and acceptor domains,

and a mixed domain;[40, 145-147] the mixed domain should be as inter-mixed as possible to

facilitate charge generation and dissociation; the pristine domains should be as pure and

crystalline as possible in order to reduce recombination and lead to better charge

transportation. Previous studies discussed even more complete pictures of the BHJ

microstructure involving 5 or even more phases;[148-152] however, we will limit the

discussion here to one amorphous mixed phase which allows us to simplify the introduction

of the miscibility concept without limiting its generality. Previous research indicates that a

certain degree of miscibility between donor and acceptor is necessary to provide a sufficiently

mixed polymer-fullerene domain which is important for charge generation and dissociation.

Nevertheless, if the polymer donor and the fullerene acceptor have too fine inter-mixing, full

percolation of the fullerene phase gets hampered and electron mobility starts to break down,

resulting in reduced µτ products and worsened fill factors. Higher fullerene loadings favor a

more complete percolation of extended fullerene phases and domains and, within specific

tolerances, will typically lead to enhanced µτ products and improved photovoltaic

performance.[145, 153-156]

Figure 6-22. Morphology schematic of morphology evolution of polymer-fullerene with (a) poor miscibility and (b) good miscibility.

More fullerenes

(b)

FullerenesPolymer

Optimization

Energy

(a)

98

The rather low miscibility of PTB7-Th and PCBM typically does result in unfavorable thin

film morphology. Advanced processing recipes relying on additives provides a path to

microstructure optimization at rather low fullerene loading;[15] nevertheless, this optimized

microstructure turns out to be metastable.[46] The microstructure instability makes these

composites vulnerable against external stress (e.g. heating, light excitation, etc. summarized

in Figure 6-22a), resulting in the thermal instability of PTB7-Th:PCBM solar cells.[6, 44,

157] We previously demonstrated that PyF5 and FAP1 are better miscible with PTB7-Th than

PCBM, leading to superior thermal stability of the corresponding solar cells.[46] In our study,

we find that the PL signal of PTB7-Th is efficiently quenched by adding PCBM, PyF5 or

FAP1 (Figure 5-2), indicating equally efficient charge dissociation; however the electron

mobilities of PyF5- and FAP1-based blends at a D:A ratio of 1:1 are two orders of magnitude

lower than that of the PCBM-based blend, indicating deficient charge transport. At a D:A

ratio of 1:3, the electron mobilities as well as the photovoltaic performance of PyF5- and

FAP1-based composites advance to a performance level comparable to the PCBM-based

composite at a mixing ratio of 1:3. From the absorption spectra of thin films (Figure 6-23),

the low polymer content in PTB7-Th:PyF5 (1:3) and PTB7-Th:FAP1 (1:3) films results in

reduced absorption and JSC, compared to the optimized PTB7-Th:PCBM (1:1.5) blend with

the same layer thickness. This fullerene ratio dependence was reported multiple times in the

literatures and is schematically summarized in Figure 6-22b, visualizing that higher fullerene

loadings typically lead to an improved percolation.[145, 158-161]

Figure 6-23. Absorption spectra of ~80 nm thick PTB7-Th:PCBM (1:1.5), PTB7-Th:PyF5 (1:3) and PTB7-Th:FAP1(1:3) thin films.

400 500 600 700 8000.0

0.2

0.4

0.6

Abs

orba

nce

Wavelength (nm)

PTB7-Th:PCBM 1:1.5 PTB7-Th:PyF5 1:3 PTB7-Th:FAP1 1:3

99

Balance between miscibility and optoelectronic properties

The current state of the art does discuss the optimum fullerene ratios in relation to the

crystallinity of the donor.[51, 161, 162] Polymers with higher crystallinity are suggested to

require lower PCBM loadings to form effective percolation pathways. However, this

hypothesis cannot fully explain the optimized PCBM loadings reported for multiple organic

solar cells, such as MDMO-PPV:PCBM, PTB7-Th:PCBM or PffBT4T-2OD:PCBM. Both

MDMO-PPV and PTB7-Th are dominantly amorphous polymers, but PTB7-Th requires far

less PCBM than MDMO-PPV to attain optimized performance. The required fullerene

loading for PTB7-Th is very similar to that of the highly crystalline/aggregated PffBT4T-

2OD. Thus, we suggest considering that the D:A ratio in BHJ blends is not solely depending

on the crystalline nature of the polymer donor but may be governed by other factors. This

thought provoking impulse is further motivated by the data from this work where both, PyF5-

and FAP1-based, systems require more fullerene loadings than PCBM-based system

irrelevant of the crystalline nature of fullerene acceptors.

Figure 6-24. (a) Solubility of TQ1 in mixed solvents using chlorobenzene as the good solvent; (b) Hansen Space and a sphere-fit matching the solubility limit of 10 mg mL-1 of TQ1.

In addition to taking HSP of polymers from literatures, we further determined the HSP for

TQ1, as shown in Figure 6-24. The Hildebrand parameters as determined by Hansen

solubility parameters (HSP) are summarized in Table 6-11 for a selection of relevant of

polymer donors and fullerene acceptors. In addition to taking HSP of polymers from

100

literatures, we further determined the HSP for a number of polymers. The interaction

parameter for various polymer-fullerene combinations were calculated and summarized in

Table 6-12. The site volume v0 in equation (6-1) must be specified whenever discussing the

interaction parameter as it is defined in terms of energy per site.[72] When v0 is fixed in

equation (6-1), the value of (𝛿1 − 𝛿2)2 is proportional to the interaction parameter x12 which

is directly correlated to the polymer-fullerene miscibility. The optimized acceptor:donor

ratios of the corresponding polymer-fullerene solar cells were taken from literatures[10, 11,

16, 17, 163-168] (Table 6-12).

Figure 6-25 summarizes the acceptor: donor ratios for optimized devices as a function of the

relative polymer-fullerene miscibility. Most surprisingly we find a well expressed relation

between the acceptor:donor ratio and the miscibility. Two separate regions are observed in

Figure 6-25. Systems with a relatively low miscibility, below the threshold, and a significant

tendency to phase separate perform best at low fullerene ratios of 1:1 or 2:1 (acceptor:donor),

more or less independent of their crystallinity. On the other hand, composites with high

miscibility, above the threshold, show a more expressed dependency on the interaction

parameter and exhibit the best performance at high fullerene loadings of 3:1 to 5:1

(acceptor:donor).

Table 6-11. Hansen solubility parameters (δd, δp, and δhb) and Hildebrand solubility parameter (δΤ) of representative polymers and fullerenes

Materials δd (MPa1/2) δp (MPa1/2) δhb (MPa1/2) δΤ (MPa1/2)a) References

PCPDTBT 19.60 3.60 8.80 21.78 [169]

MDMO-PPV 19.06 5.62 5.28 20.56 [170]

MEH-PPV 19.06 5.38 5.44 20.53 [170]

TQ1 19.20 4.50 4.80 20.30 b)

pDPP5T-2 19.00 3.00 2.00 19.30 [143]

PffBT4T-2OD 18.56 4.07 2.31 19.14 [40, 171]

P3HT 18.50 4.60 1.40 19.11 [172]

PTB7-Th 18.56 2.30 3.21 18.98 this chapter

PCBM 20.6 4.93 4.23 21.60 [144]

PC71BM 20.95 2.80 1.64 21.20 [144]

PyF5 20.92 2.15 0.19 21.03 this chapter

101

FAP1 20.75 2.60 0.33 20.91 this chapter

a) 𝛿𝑘 = 𝛿𝑑2 + 𝛿𝑝2 + 𝛿ℎ𝑏22

b) Measured via the binary solvent gradient method

Table 6-12. Interaction parameters of polymer:fullerenes and the corresponding

Acceptor:Donor ratios from literatures Polymer:fullerene 𝒙𝟏𝟏/𝒗𝟎 (𝜹𝟏 − 𝜹𝟏)𝟏 Optimized devices

(10-3· cm-3 · mol) (MPa) A:D ratio References

PCPDTBT:PCBM 0.014 0.034 3.6:1 [164]

MDMO-PPV:PCBM 0.44 1.08 4:1 [10]

MEH-PPV:PCBM 0.45 1.13 5:1 [163]

TQ1:PCBM 0.69 1.70 3:1 [165]

pDPP5T-2:PCBM 2.06 5.11 2:1 [167]

PffBT4T-2OD:PCBM 2.44 6.05 1.2:1 [17]

P3HT:PCBM 2.49 6.18 1:1 [166]

PTB7-Th:PCBM 2.78 6.89 1.5:1 this chapter

PCPDTBT:PC71BM 0.14 0.34 3:1 [11]

MDMO-PPV:PC71BM 0.16 0.41 4:1 [10]

TQ1:PC71BM 0.33 0.82 3:1 [165]

pDPP5T-2:PC71BM 1.40 3.46 2:1 [167]

PffBT4T-2OD:PC71BM 1.71 4.24 1.2:1 [17]

P3HT:PC71BM 1.75 4.35 1:1 [166]

PTB7-Th:PC71BM 2.00 4.95 1.5:1 [16]

pDPP5T-2: FAP1 0.98 2.44 2:1 this chapter

PTB7-Th: FAP1 1.49 3.70 2:1 this chapter

pDPP5T-2: PyF5 1.11 2.76 2:1 this chapter

PTB7-Th: PyF5 1.65 4.10 2:1 this chapter

Although the miscibility of PC61BM and PC71BM was found to be different when blended

with the same polymer, the difference is still below the threshold of miscibility as shown in

Figure 6-25. In this case, the optimized acceptor: donor ratio is expected to be very similar

for a certain polymer. Figure 6-25 is in good agreement with the observation that polymers

102

having better miscibility with fullerenes require higher fullerene loadings to attain optimized

photovoltaic performance, quite independent of their crystallinity. The correlation between

miscibility and optimized acceptor:donor ratio is more clear from the Figure 6-26, where the

acceptor:donor ratio is plotted logarithmically.

Figure 6-25. Optimal fullerene acceptor: polymer donor ratios as a function of polymer-

fullerene miscibility

Figure 6-26. Fullerene acceptor: polymer donor ratios as a function of polymer-fullerene

miscibility (with the acceptor:donor ratio plotted logarithmically)

MDMO-PPV

PffBT4T-2ODPTB7-Th

MDMO-PPV

PffBT4T-2ODPTB7-Th

103

6.10 Conclusion

To summarize, we demonstrate that the low miscibility between PCBM and pDPPT5-2 or

PTB7-Th is the one of the fundamental origins of the low thermal stability. On the contrary,

two novel fullerenes, PyF5 and FAP1, with a significantly higher chemical compatibility are

introduced to overcome these limitations. PyF5 and FAP1 further exhibit more optimized

energy levels and a higher VOC than PCBM does, allowing for keeping the time-zero

performance losses at a minimum. Most importantly, the better chemical miscibility with

PTB7-Th and pDPP5T-2 allows overcoming short-time JSC burn-in losses and furthermore

results in OPV devices with superior thermal stability. The combination of promising

optoelectronic properties and thermal stability underlines the necessity for novel acceptor

design rules balancing performance and stability. The benefit of chemical miscibility as a

novel design principle for improved stability is expected to pave alternative guidelines

towards designing and developing novel acceptors for efficient and thermally stable OPV

devices.

We in-depth investigated the carrier transport loss mechanism for PTB7-Th:PCBM, PTB7-

Th:PyF5 and PTB7-Th:FAP1 BHJ composites as a function of their microstructure and

especially their miscibility. An obvious correlation between the microstructural compatibility

of the polymer-fullerene components on the electron transport properties and ultimate

photovoltaic performance of BHJ solar cells was observed. Surprisingly, this interaction is

independent of the crystalline nature of the fullerenes but rather depends on the miscibility of

the polymer donor with the fullerene acceptor. We further found that the miscibility between

the donor and acceptor dominates the optimized acceptor loadings in polymer-fullerene BHJ

systems. BHJ composites with good polymer-fullerene miscibility require higher fullerene

loadings than composites with a tendency to phase separation. It has to be carefully

considered for future design of the donor and acceptor BHJ composites that the miscibility

and the optoelectronic properties need to be well balanced in order to maximize the

photovoltaic performance of organic BHJ solar cells.

104

105

Chapter 7 Correlation of JSC burn-in losses and microstructure metastabilities in organic solar cells

Abstract

We demonstrate that the JSC burn-in loss can be either induced by thermal annealing or

white light illumination. Detailed microstructural studies confirm that demixing of

fullerenes from polymers in the amorphous regime is the primary mechanism for both

degradation conditions, although their kinetics is distinctly different. Both, light and heat,

provide enough energy to metastable bulk-heterojunction regimes and relax them into

their thermodynamic equilibrium, which typically is larger scale phase separated due to a

positive interaction energy. Notably, the microstructural changes induced by either

thermal- or light-aging can be kinetically correlated to each other. Similar to the

phenomena that higher temperature initiates faster degradation towards equilibrium, more

intense light does as well cause faster degrading.

Part of the results presented in the chapter was done by collaborators:

The GIWAXS/GISAXS patterns were done by Wolfgang Gruber in the Institute of

Crystallography and Structural Physics, FAU, Germany

PCE11 (P2) was synthesized and supplied by Prof. Lei Ying’s group by South China

University of Technology.

Part of the results has been submitted.

106

7.1 Introduction

Since the introduction of bulk-heterojunction into the field of organic photovoltaics (OPV),

astounding achievements have been attained in materials designs as well as device

optimizations. As the efficiency of organic solar cells is approaching 14%, more and more

research efforts have been dedicated to study and hopefully resolve the instability problems

of organic photovoltaics. Degradation induced by extrinsic factors, like oxygen, humidity,

and impurities in materials, can be tremendously suppressed by appropriate encapsulation and

purification. However, degradation caused by intrinsic factors, like accumulated heat and

illumination, must be understood and addressed in order to enter the photovoltaic market.

7.2 Materials and device fabrication

ZnO-nanoparticle dispersion in isopropyl alcohol was received from Nanograde. PCE11

(batch: YY10128CH) were purchased from One Material and from collaborators in South

China University of Technology. PC61BM and PC71BM (99%) were purchased from Solenne

BV, respectively.

Figure 7-1. Chemical structures of photo-active materials studied in the chapter

All solar cells were fabricated in an inverted structure of ITO/ZnO/Active layer/MoOx/Al

(Figure 7-2). The pre-structured ITO coated glass substrates were subsequently cleaned in

toluene, acetone and isopropyl alcohol for 10 min each. Then 30 nm ZnO (Nanograde, N10)

were doctor bladed at 30 °C on the ITO substrate and annealed at 85 °C in ambient air. The

107

PCE11 based active layer from chlorobenzene: 1,2-dichlorobenzene: diphenylether

(58.5:38.5:3) solution were spin coated on top of ZnO in nitrogen atmosphere. 15 nm MoOx

and 100 nm silver were deposited subsequently under 6 × 10-6 Torr by thermal evaporation

through a shadow mask to form an active area of 10.4 mm2.

Figure 7-2. Schematic device architecture of organic solar cells

7.3 Three batches of PCE11

As depicted in Figure 7-3, there are three batches of PCE11, namely P1, P2 and P3, which

have different molecular weight and polydispersity index (PDI). From the DSC data of the

pristine PCE11, P1, P2 and P3 exhibit melting temperature of 274 °C, 268 °C and 247 °C,

respectively, which is in good agreement with literature that polymer with higher molecular

weight presents higher melting point. Notably, the crystalline properties of the pristine

polymers are well demonstrated in the DSC curve. P1, with the highest molecular weight and

the lowest PDI, shows a narrow melting peak, indicating relatively bigger and more perfect

crystals; P3, with the lowest molecular weight and a slightly higher PDI, reveals a broader

melting peak, implying less perfect crystals. P2 with the broadest molecular weight

distribution displays two well-defined features, a narrow melting peak at a higher temperature

and a small bump at a relatively lower temperature, which suggests that P2 mostly forms

relatively more perfect crystals like P1 and a small fraction of smaller and less perfect crystal

like P3. The glass transition temperature of all three polymers is around 72 °C. The solar cell

108

architecture in this work employed an inverted structure as depicted in Figure 1c. With

[60]PCBM as acceptor, three batches of PCE11 based solar cells could attain current density

(JSC) of around 18 mA cm-2 and efficiency (PCE) at around 8.5%.

Figure 7-3.crystalline and optic-electronic properties of three batches of PCE11

7.4 Thermal- and photo- stability

The solar cells based on the three batches of PCE11 were aged with white LEDs and at 85 °C

in nitrogen atmosphere, respectively. As shown in Figure 7-4, the JSC of the solar cells

degraded overtime with either light aging or thermal aging, while the voltage and fill factor

remained at high levels (Figure 7-5 and Figure 7-6). As proved in our previous work, the

demxing of the intimately mixed donor:acceptor regions causes the JSC burn-in degradation.

Interestingly, the P2:PCBM solar cells suffered from less JSC burn-in losses than P1:PCBM

and P3:PCBM solar cells. After 1100 hours exposing to white light illumination (Figure 7-6),

P2:PCBM solar cells still preserved more than 80% initial JSC , while P1:PCBM and

P3:PCBM solar cells lost almost 40% initial JSC at round 600 hours. Similarly, after 2 hours’

annealing at 85 °C, P2:PCBM solar cells still preserved 80% initial JSC and researched

composite thermodynamics while P1:PCBM and P3:PCBM degraded nearly 40% initial JSC

109

and still have trend to further degradation. This is highly interesting because this is not purely

the effect of higher or lower molecular weight as the more stable system possesses medium

molecular weight; further, the three systems have similar photovoltaic performance.

Figure 7-4 Normalized JSC evolution of PCE11:PCBM solar cells (a) under white light illumination (~ one sun) and (b) at 85 °C annealing.

Figure 7-5 Normalized VOC, FF and PCE evolution of PCE11:PCBM solar cells at 85 °C annealing.

(a) (b) (c)

110

Figure 7-6 Normalized JSC, VOC, FF and PCE evolution of PCE11:PCBM solar cells under white light illumination.

7.5 crystalline properties and thermal behaviors

To acquire deeper insights into the composite film morphology, the three pristine and blended

PCE11films were characterized with Grazing-Incidence Wide-Angle X-ray Scattering

(GIWAXS) technique. The GIWAXS profiles collected from out-of-plane and in-plane cuts

are depicted in Figure 7-7 while the corresponding 2D patterns are shown in Supporting

Information Figure 7-8. The positions of the polymer lamella peak (100) and π−π stacking

(010) are well defined and in good agreement with literature. From Figure 3, both pristine

and blended PCE11 has strong lamella peak (100) in in-plane direction and π−π stacking

(010) in out-of-plane direction, which signifies preferential face-on orientation of pristine

polymers. The addition of PCBM did not change the preferential orientation of PCE11.

111

Figure 7-7. (a-d) GIWAXS linecuts of PCE11 in pristine and in blends; (e) FWHM of polymer (100) peak from in-plane cuts.

(a)

(b)

(c)

(d)

(e)

112

Figure 7-8. 2D GIWAXS patterns of polymers in pristine and in blends. The film thickness of the samples are P1: 880 nm, P2: 169 nm, P3: 299 nm, P1-PCBM: 1345 nm, P2-PCBM: 248 nm, P3-PCBM: 282 nm; the measured time of P3 and P2 is five minutes each spot while the

measured time of other samples is 10 minutes each spot. The linecuts of the samples are corrected with sample thickness and measured time.

The full width at half maximum (FWHM) of polymer (100) peak from in-plane cuts are

calculated and summarized in Figure 7-9a. According to the Scherrer equation,[173] the

crystal size and the FWHM are inversely proportional, thus, we can use the Scherrer equation

for a qualitative discussion of the crystallite sizes. For three pristine PCE11, the FWHM of P1

is slightly smaller than that of P2 while the FWHM of P3 is obviously larger than the FWHM

of P1 and P2, which demonstrates that the average size of the ordered domains of pristine P1

and P2 is similar while the average ordered domain size of pristine P3 is apparently smaller.

With the presence of PCBM, the FWHM of three PCE11 (100) became smaller, indicating

bigger ordered domain size. This might be due to the existence of dispersed PCBM in the

polymer ordered phase, which made the crystals less perfect but larger area. Noticeably, P3

P1 P3P2

P1:PCBM P2:PCBM P3:PCBM

113

crystals were affected the most with the addition of PCBM, which changed from forming the

smallest ordered domains in pristine to forming the biggest ordered domains in blend. Owing

to the imperfect nature of the P3 crystal, PCBM can easily intrude into the polymer packing.

P1 crystals were influenced the least by the addition of PCBM, which is mainly because P1

formed the most perfect crystals among the three PCE11, thus, the incursion of PCBM was

most difficult.

Figure 7-9 Thermal behaviors of polymers in pristine and in blends

The thermal behaviors of PCBM, polymers in pristine and in blends were further studied by

DSC. The 2nd scan heating curves are illustrated in Figure 7-9b. The melting point of PCBM

is in consistence with literature; the melting point of P1, P2 and P3 measured with DSC is the

same as the melting point measured with MDSC. P1:PCBM, P2:PCBM and P3:PCBM

composites display two melting peaks. The melting point of higher temperature, which is at

the exact position of pristine PCBM melting peak, is correspond to PCBM crystals; the

melting point at lower temperature is the melting peak of polymer crystals. The melting

temperature of polymer crystals in blends is lower than in pristine, indicating less perfect

crystals formed in blend than in pristine. Comparing the heating scan of P3 in pristine and in

blend together with GIWAXS data, it is quite obvious that the addition of PCBM strongly

interrupt the crystalline behavior of P3, which results in large amount of amorphous regions

(a) (b)

114

in BHJ film. Thus, when applying external energy, light or heat, the intimately mixed donor-

acceptor domain can easily demix into pure donor and acceptor domains, which results in

reduced donor-acceptor interface areas and fast JSC burn loss.

We further investigated the morphology of the BHJ films by measuring in-situ laser light

scattering (LS) of P1:PCBM and P2:PCBM composite films upon annealing at 85 °C in

nitrogen atmosphere. The LS was measured with 785 nm laser, which means it was only

measuring approximately 50-70 nm size of ordered aggregates/structures. From Figure 7-10,

the LS signal of P2:PCBM sample started at approximately 330 mV while the LS signal of

P1:PCBM started at approximately 140 mV. As the thickness of the two samples were similar,

this information suggests that there were more ordered structures around 50-70 nm in

P2:PCBM blend film than in P1:PCBM blend film. During annealing, the LS signal of

P2:PCBM sample slightly decreased, indicating the break-down of small amount of ordered

structures; on the contrast, the LS signal of P1:PCBM sample slightly increased, implying the

development of the ordered structures.

Figure 7-10. Light scattering of P1:PCBM and P2:PCBM composite films annealing at 85 °C

115

7.6 Morphology evolution under external stresses

Until now, we have better understanding about the correlation of the BHJ morphology and

solar cell stability. For donor:acceptor system with poor miscibility, like PCE11:PCBM, the

crystal phase is the key to the stability of the solar cells. As illustrated in Figure 7-11, there

are generally five phases within BHJ: crystalline pure polymer and fullerene phases,

amorphous pure polymer and fullerene phases, and amorphous polymer:fullerene mixed

phase. For P3:PCBM BHJ, the imperfect crystalline nature of polymer crystals resulted in

large area of amorphous phase; when external energy was applied, heat or light, the

amorphous pure domains failed to suppress the demxing of the donor:acceptor amorphous

regions. For P1:PCBM BHJ, owing to the strong self-aggregating nature of polymer chains,

the donor:acceptor mixed region has strong tendency to phase separate into ordered pure

polymer and fullerene domains upon the application of external energy (LS data). For

P2:PCBM BHJ, the polymer crystals comprised relatively more perfect crystals like in the

P1:PCBM BHJ and small amount of relatively less perfect crystals like in the P3:PCBM BHJ.

The big and more perfect crystals acted as anchors when there was external stress; the

imperfect crystals in the relatively more amorphous phase suppressed the movement of the

amorphous regions to some degree, thus, the JSC burn-in degradation was less profound in

P2:PCBM solar cells.

116

Figure 7-11. Film morphology evolution of polymers in pristine and in blends with or without aging

7.7 Stability of PCE11:PC71BM

PC71BM instead of PC61BM is used for this study to suppress dimerization. PCE11: PC71BM

solar cells were aged by thermal annealing at 85 °C or illumination provided by white Light-

emitting diodes (LEDs) with a spectral emission from 400 nm to 800 nm. Both thermal and

light aging were performed in electronically controlled degradation chambers in nitrogen

atmosphere. The J-V characteristics were probed periodically inline.

P1 P2 P3

Pristine

BHJ

AgedBHJ

117

Figure 7-12 (a) current density-voltage characteristics of fresh and aged samples; Normalized JSC, VOC, FF and PCE evolution of PCE11:PC71BM solar cells (b) at 85 °C annealing and (c) under white light illumination (~ one sun).

Figure 7-13 Normalized VOC and FF evolution of PCE11:PC71BM solar cells under 85 °C annealing. The first point was measured at room temperature (r.t.); the second point was measured at the 2nd minute after turning on heating (temperature stabilized at 85 °C); after 2 hours’ annealing, the solar cells were cooled down to room temperature (last point in the figure). The VOC and FF were temperature dependent and reversible. Thus, VOC and FF evolution in Figure 7-12 were normalized to the point measured at the second minute. Figure 7-12a shows the J-V characteristics of PCE11:PC71BM solar cells with and without

aging. The control device (kept in the dark) exhibited a JSC of 18.0 mA cm-2 and PCE of

9.2 %. After 2 hours’ annealing at 85 °C, the photovoltaic performances dropped by over 40%

compared to the reference, as depicted in Figure 7-12 and Figure 7-13. The loss in PCE is

mainly due to the strong JSC burn-in degradation, which decreased sharply in the first hour

118

and then remained almost unchanged. During degradation, both FF and VOC almost retained

their original values. Most interestingly, we observe a significantly slower but otherwise

identical degradation behavior under one-sun illumination with white light LEDs. Again, the

voltage and FF of the solar cell remained high, and the JSC burn-in loss dominated the total

performance loss. Interestingly, the degradation characteristics as well as the J-V curves after

degradation are identical for thermal aging and light aging.

7.8 Evolution of heat- and light- induced morphology

To understand the loss mechanisms, in particular the origin for these generic burn-in

phenomena, Fourier Transform Infrared (FTIR), Grazing-Incidence Wide-Angle X-ray

Scattering (GIWAXS) and Grazing-Incidence Small-Angle X-ray Scattering (GISAXS)

measurements on fresh and aged BHJ films were carried out. The FTIR spectra in Figure 7-

14a reveal that the features of both polymer and PC71BM in three samples are identical.

Neither heat nor light has chemically damaged the semiconductors in our experiments, or the

damage was negligible to be detected. The GIWAXS profiles collected from out-of-plane and

in-plane cuts are depicted in Figure 7-14b while the corresponding 2D patterns are shown in

Figure 7-15. The polymer lamella peak (100) and π−π stacking (010) are well defined and in

good agreement with literature. The perfectly overlapped GIWAXS profiles demonstrate that

the crystalline phases of PCE11 and PC71BM remained unchanged after thermal annealing or

light illumination. Thus, charge transport and non-geminate recombination in these samples

remained almost unaffected, which is in good agreement with the high FF retained for the

aged samples.

119

Figure 7-14. (a) FTIR spectra, (b) GIWAXS linecuts, (c) GISAXS Qy linecuts of PCE11:PCBM blend films with and without aging.

The morphological properties concerning the intimately mixed amorphous regions were

investigated by GISAXS measurement.[174, 175] The 2D GISAXS patterns and in-plane

GISAXS profiles of the films with and without aging are shown in Figure 7-16 and Figure

7-14c, respectively, and clearly indicate distinct differences for the degraded samples. The

form factor scattering of the pure PC71BM domains/clusters contributes mostly to the shape

of the shoulder in the linecuts. The in-plane profiles of three samples were fitted with Debye-

Anderson-Brumberger (DAB) model, and with a combined model of DAB model and poly-

dispersed spheres having a Schultz-Zimm size distribution. The poly-dispersed spheres are

supposed to describe the PCBM aggregates while the DAB model describes the large scale

network of the PCBM molecules distributed within the amorphous and around the crystalline

polymer molecular conformations.[176] However, the contribution of the poly-dispersed

spheres is so small that no reliable data can be extracted from the fits. Therefore, we restrict

ourselves to the DAB model. The correlation length of each sample derived from the DAB

model is summarized in Table 7-1. Both thermal and light-aged samples have increased

correlation lengths as compared to the reference, representing the diffusion of fullerene

molecules and the formation of domains with a stronger separation between the polymer

donor and the fullerene acceptor. In line with previous studies,[40] we suggest that the poor

miscibility between PCE11 and PC71BM is the driving force behind the donor / acceptor

demixing in the amorphous phase. Here, the combination of FTIR, GIWAXS and GISAXS

analysis allows us to uniquely conclude that light- and thermal- aging indeed induce the same

120

type of microstructural changes in the finely mixed, amorphous regions, which primarily

accounts for the burn-in degradation in these types of BHJ solar cells.

Figure 7-15. 2D GIWAXS patterns of PCE11:PC71BM blend films with and without aging. The color bars represent the intensity of the GIWAXS data. The import and export function of BoranAgain software[177] was used to depict the 2D patterns in Qy-Qz coordinate system.

Figure 7-16. 2D GISAXS patterns of PCE11:PCBM blend films with and without aging. The color bars represent the intensity of the GISAXS data. The import and export function of BoranAgain software[177]was used to depict the 2D patterns in Qy-Qz coordinate system.

(100)

(010)

PCBM

121

Table S1 Structural parameters determined by DAB model-fitting of GISAXS profiles of films with and without aging.

Sample Correlation lengtha) of the mixed region (nm)

Standard deviation of the correlation lengthb) (nm)

w/o aged 9.4 1.16 thermally aged 10.02 0.37

light aged 11.64 0.3 a) Each sample was measured at three spots to account for lateral inhomogeneity. Given are the averaged values measured at three spots; b) Given are the standard deviations of the three individual measurements.

7.9 Equivalence of heat- and light- induced degradation

To acquire deeper insights into the degradation behavior caused by thermal annealing and

light illumination, we aged the solar cells under different stress levels by varying the

degradation temperature, the light intensity as well as the photon energy. Figure 7-17a

reveals that the degree of burn-in degradation after finding equilibrium is directly related to

the applied temperature. Solar cells aged at 35 °C find equilibrium after about 100 hrs at

90 % of the initial performance. Increasing the aging temperature for the very same sample to

60 °C triggers another exponential-type JSC burn-in loss, reducing the performance to about

80 % in less than 100 hours. Finally, increasing the temperature to 85 °C caused a quick JSC

drop to about 70 % of the initial performance within few hours only. After that, JSC remained

constant. FF and VOC of the solar cells remained almost unchanged through the whole aging

process. PCE basically evolved following the JSC trend (Figure 7-18). As shown in

temperature-modulated differential scanning calorimetry (MDSC) in Figure 7-3, the glass

transition temperature (Tg) of the PCE11 batch used in this investigation was around 72 °C.

This fairly low Tg is the reason behind the sharp losses at 85 °C: with the segments of PCE11

moving more quickly, demixing of PCE11: PC71BM becomes accelerated, resulting in the

significant JSC burn-in loss.

122

Figure 7-17. Normalized JSC evolution of PCE11:PC71BM solar cells during aging under (a) varying temperature, (b) light of different photon energy (~ one sun), and (c) various light intensity. (d) JSC loss at 40th hour as a function of light intensity. While under illumination, the temperature of the chamber was kept at ~ 20 °C

123

Figure 7-18. Normalized VOC, FF and PCE evolution of PCE11:PC71BM solar cells aged at 85 °C in nitrogen atmosphere. The VOC decreased whenever the temperature increased because of the temperature-dependent effects, which is was found to be reversible as depicted in Figure 7-13.

Next, we analyzed the kinetics of the photo-induced burn-in as a function of light intensity.

Assuming the equivalence of photo- and thermal degradation in these composites, one would

further expect a direct correlation between the level of photo-induced burn-in degradation and

the photon density (light intensity). However, before studying the light intensity dependence,

it is of utmost importance to clarify the possible role of the photon´s excess energy on

microstructure induced burn-in degradation. The photon´s excess energy is the difference

between the material bandgap and the photon energy. Light with an energy of 3 eV (about

413 nm) will dissipate more than 1.5 eV of heat when absorbed by a semiconductor with a

bandgap of 1.5 eV (about 826 nm) That excess energy will dissipate heat into the lattice of

the polymer by exciting phonon vibrations. Several solar cells fabricated in the same run

were tested under different light conditions. First, six solar cells were aged under light with a

wavelength higher than ~ 650 nm by using two long-wave pass filters (650 nm cut-on and

500 nm cut-on filters); the light intensity was adjusted, so that the JSC of the solar cells was

identical to the that measured under simulated AM 1.5. To achieve one-sun equivalent light

intensities, a homemade, highly-accelerated lifetime setup based on concentrated metal halide

0 50 100 150 200 250 3000.0

0.2

0.4

0.6

0.8

1.0

Nor

m. P

erfo

rman

ce

Thermal aging time (h)

VOC

FF PCE

124

lamps with a variable light intensity ranging from 1 to about 140 suns was employed. Details

on the setup can be found in literature.[178] The wavelength window was chosen such that

the photon energy of the light above 650 nm (Ehν ≤ 1.9 eV) is slightly above resonance with

the bandgap of PCE11 (1.65 eV).[17] As illustrated in Figure 7-17b, the JSC lost

exponentially while the FF and VOC were preserved (Figure 7-19). After around 160 hours,

the filters were replaced by a 400 nm cut-on long pass filter transmitting light with

wavelengths higher than 400 nm (Ehν ≤ 3.1eV); the light intensity was adjusted, so that the

JSC of the solar cells was the same as before changing filters. To our surprise, Figure 7-17b

reveals that highly energetic light with a wavelength > 400 nm is resulting in JSC burn-in

losses of similar magnitude as low energetic light with a wavelength > 650 nm. Further

studies are being planned to better understand the insensitivity of microstructure related burn-

in degradation on the photon energy. For the moment we state a wavelength independence of

the JSC burn-in and turn our attention to the light intensity dependence.

Figure 7-19. Normalized VOC, FF and PCE evolution of PCE11:PC71BM solar cells aged under light with similar intensity but different photon energy in nitrogen atmosphere

0 100 200 3000.0

0.2

0.4

0.6

0.8

1.0

Nor

m. P

erfo

rman

ce

Light aging time (h)

< 1.9 eV < 3.1 eV

125

Different neutral-density filters were employed to investigate the light intensity dependence

of microstructure related burn-in degradation between 0.3 to 1.6 suns. The light intensity of

the highly accelerated lifetime setup was calibrated with a standard Si-diode by Newport.

Figure 7-17c summarizes the light intensity dependence of burn-in degradation. JSC is

diminished significantly faster when aging under stronger illumination, with a weakly,

sublinear slope (Figure 7-17d). FF and VOC again were not affected (Figure 7-20). Although

we are not able to model the weak light intensity dependence, such small scaling exponents

are more characteristic of indirect mechanisms like thermal dissipation. Most interesting, the

fact that burn-in is insensitive to photon energy but sensitive to light intensity suggests that

the mechanism is rather coupled to the density of free carriers than to the excess energy of

excitons. Recent transport studies reported that photo-generated carriers in disordered organic

semiconductors take nano- to microseconds to find energetic equilibrium.[179, 180] Further

investigations are necessary to understand whether these phenomena are related.

126

Figure 7-20. Normalized VOC, FF and PCE evolution of PCE11:PC71BM solar cells aged under different light intensity in nitrogen atmosphere

Finally, and to uniquely proof the equivalence of thermal and light aging for this bulk

heterojunction composite, solar cells fabricated in the same run were subsequently aged

under heat and light. As illustrated in Figure 7-21, there was no further JSC loss under

illumination when the solar cells were first thermally aged at 85 °C; similarly, no further JSC

loss was observed upon thermal aging when the solar cells were subject to photo-degradation

beforehand. The FF and VOC again were insensitive to this procedure and remained at high

level (Figure 7-22). We believe that Figure 7-21 is a strong proof that thermal-annealing and

photo-degradation seem to cause identical changes to the metastable microstructures of

PCE11:PC71BM. We believe that this equivalence is quite generic to bulk heterojunction

composites with metastable microstructures. We are aware that two observations for different

127

polymer systems are, strictly speaking, not a unique proof but a strong admissible hypothesis

for the generic equivalence of temperature / light-intensity burn-in degradation in metastable

microstructures. We believe that the data presented in this study will trigger further studies

challenging the universality of this hypothesis.

Figure 7-21. Normalized JSC evolution of PCE11:PC71BM solar cells during aging. Top: the solar cells were first under thermal aging at 85 °C then exposed to one-sun white-light illumination; Bottom: the solar cells were first under white light illumination then aged at 85 °C. While under illumination, the temperature of the chamber was about ~ 35 °C.

128

Figure 7-22. Normalized VOC, FF and PCE evolution of PCE11:PC71BM solar cells during aging. Top: the solar cells were first under thermal aging at 85 °C then exposed to white light LEDs; Bottom: the solar cells were first under white light illumination then aged at 85 °C.

129

7.10 Conclusion

In this chapter, we find out that crystalline properties of polymer, PCE11, are hugely

influenced by molecular weight and polydispersity; furthermore, polymer-fullerene BHJ

film morphology is largely affected by the crystalline nature of polymers and evolve

differently under external stresses, like heat and light, which eventually results in solar cells

with different burn-in loss and lifetime.

We investigated the heat- and light-induced degradation behavior of a semi-crystalline BHJ

system based on PCE11:PC71BM. Burn-in degradation exclusively affects JSC and is

equivalently induced by thermal-aging or photo-aging respectively. Both degradation

conditions induce the same polymer – fullerene demixing responses in the amorphous region.

We find out that the device stability, in particular the stability of the BHJ microstructure of

PCE11:PC71BM, is equally challenged by photo-illumination of about > 100 hours and

thermal-aging at 85°C of about ~1 h, respectively. The detailed ratios are certainly depending

on the light intensity, the temperature and the glass transition temperature of the composite. A

detailed analysis of the energy and intensity dependence of light-induced burn-in degradation

suggests that photo-excited carriers do stress amorphous polymer segments in a similar way

as thermal stress. This finding might be adaptive to many other BHJ systems, which suffer

from JSC burn-in degradation due to a metastable BHJ morphology, and provides a unique

way to extend the standard photo-stability test (typically >1000 hrs under one-sun

illumination) by accompanying with thermal-aging at elevated temperatures.

130

131

Chapter 8 Summary and outlook

Abstract

Main findings and observations are summarized in this chapter. The thermal- and photo-

instability of organic BHJ solar cells remains a challenge. Some interesting strategies are

suggested in the outlook.

132

8.1 Summary

Driving by the motivation of pushing OPV toward the PV market, main efforts of this PhD

program devote to investigate the heat and white light induced intrinsic degradation as it is

deemed to be investigated and addressed with the highest priority. The main findings of the

thesis are as following:

A polystyrene backbone based side-chain polymeric fullerenes was successfully synthesized

and applied in solar cell. Two main-chain polymeric fullerenes, PPC4 and PPCBMB, were

employed as additives in PCE11:PCBM based solar cells. However, side-chain polymeric

fullerenes suffer from the poor solubility which limits the role of acting the main acceptor in

the photovoltaic application; polymeric fullerenes (both side chain and main chain) normally

cannot attain comparable performance alone with the polymer donor, and can only act as the

secondary acceptor in the application of solar cells.

It is demonstrated that the low miscibility between PCBM and pDPPT5-2 or PTB7-Th is the

one of the fundamental origins of the low thermal stability. On the contrary, two novel

fullerenes, PyF5 and FAP1, with a significantly higher chemical compatibility are introduced

to overcome these limitations. Furthermore, It is observed that the miscibility between the

donor and acceptor dominates the optimized acceptor loadings in polymer-fullerene BHJ

systems. BHJ composites with good polymer-fullerene miscibility require higher fullerene

loadings than composites with a tendency to phase separation.

Crystalline properties of PCE11 are hugely influenced by molecular weight and

polydispersity; furthermore, polymer-fullerene BHJ film morphology is largely affected by

the crystalline nature of polymers and evolve differently under external stresses, like heat and

light, which eventually results in solar cells with different burn-in loss and lifetime. The

device stability, in particular the stability of the BHJ microstructure of PCE11:PC71BM, is

equally challenged by photo-illumination of about > 100 hours and thermal-aging at 85°C of

about ~1 h, respectively. This finding might be adaptive to many other BHJ systems, which

suffer from JSC burn-in degradation due to a metastable BHJ morphology, and provides a

unique way to extend the standard photo-stability test (typically >1000 hrs under one-sun

illumination) by accompanying with thermal-aging at elevated temperatures.

133

8.2 Outlook

Some general guidelines for future design of materials and device from the above findings

are:

The miscibility and the optoelectronic properties need to be well balanced in order to

maximize the photovoltaic performance of organic BHJ solar cells. In the aspect of materials

synthesis, it is advised to synthesize new semiconducting materials with high glass transition

temperature, no cold crystalline nature, and less strong self-packing properties. In the aspect

of device fabrication, it is suggested to create novel combinations of donor and acceptor

taking into account that the intrinsic interaction and miscibility between different materials

could result in completely different film morphologies and eventually influence the solar cell

stability.

The thermal- and photo- instability of organic BHJ solar cells remains a challenge although

more and more insights are gaining rapidly in the past few years. To understand the degrading

mechanisms, the lab scale stability studies of solar cells mainly employ constant conditions -

such as constant light intensity, constant temperature - or focus on single factor. As the

appearance of highly efficient and relatively more stable non-fullerene based solar cells, it is

of great interest to test the lifetime of these solar cells in conditions that are close to practical

application. Firstly, as the degradation might be load-dependent, the solar cells should be

aged at maximum power output (MPP) load. Secondly, the lab-scale solar cells should be

aged under controllable Solar-Thermal-Humidity Cycling. Finally, the lab-scale solar cells or

mini-modules should be degraded outdoor; and the power created can be collected for power

supply, for example, to support applied load at MPP. At this stage, real issues can be

identified. Degrading effects like heat- or light- induced degradation, degrading components

like active layer or interface, degrading mechanisms like electrochemical reactions or

physical rearrangements will be compared and studied under different conditions.

It is highly interesting to investigate how light induces physical re-arrangement of the BHJ

active layer.

Further, the instable effects resulting from interface layer should be investigated in parallel.

134

135

Appendix A Curriculum Vitae

Chaohong Zhang

Email: chaohong.zhang@fau.de, chaohong_zhang@126.com

Tel: (+49) 09131/8527730

Address: Martensstr. 7, 91058, Erlangen, Germany

Personal Information: Born on March 3rd 1987 in Enping Guangdong, China; Unmarried

Education Background: 2013−present Friedrich-Alexander-Universität Erlangen-Nürnberg

Doctor candidate of material science Prof. Christoph Brabec

2010 − 2013 Beijing University of Chemical Technology (BUCT)

Master of Chemistry (Polymer Chemistry and Physics) Prof. Jianping Deng

2006 − 2010 Beijing Institute of Technology (BIT)

Bachelor of Polymer Materials and Engineering

136

137

Appendix B Publications and Presentations

Publications: [1] C. Zhang, S. Langner, A.V. Mumyatov, D.V. Anokhin, J. Min, J.D. Perea, K.L.

Gerasimov, A. Osvet, D.A. Ivanov, P. Troshin, N. Li, C.J. Brabec, Understanding the correlation and balance between the miscibility and optoelectronic properties of polymer-fullerene solar cells, Journal of Materials Chemistry A, 5 (2017) 17570-17579.

[2] C. Zhang, A. Mumyatov, S. Langner, J.D. Perea, T. Kassar, J. Min, L.L. Ke, H.W. Chen, K.L. Gerasimov, D.V. Anokhin, D.A. Ivanov, T. Ameri, A. Osvet, D.K. Susarova, T. Unruh, N. Li, P. Troshin, C.J. Brabec, Overcoming the Thermal Instability of Efficient Polymer Solar Cells by Employing Novel Fullerene-Based Acceptors, Advanced Energy Materials, 7 (2017) 1601204.

[3] J.D. Perea, S. Langner, M. Salvador, B. Sanchez-Lengeling, N. Li, C. Zhang, G. Jarvas, J. Kontos, A. Dallos, A. Aspuru-Guzik, C.J. Brabec, Introducing a New Potential Figure of Merit for Evaluating Microstructure Stability in Photovoltaic Polymer-Fullerene Blends, The Journal of Physical Chemistry C, 121 (2017) 18153-18161.

[4] L. Ke, N. Gasparini, J. Min, H. Zhang, M. Adam, S. Rechberger, K. Forberich, C. Zhang, E. Spiecker, R.R. Tykwinski, C.J. Brabec, T. Ameri, Panchromatic ternary/quaternary polymer/fullerene BHJ solar cells based on novel silicon naphthalocyanine and silicon phthalocyanine dye sensitizers, J. Mater. Chem. A, 5 (2017) 2550-2562.

[5] C. Zhang, C. Song, W. Yang, J. Deng, Au@poly(N-propargylamide) Nanoparticles: Preparation and Chiral Recognition, Macromolecular Rapid Communications, 34 (2013) 1319-1324.

[6] F. Yao, D. Zhang, C. Zhang, W. Yang, J. Deng, Preparation and application of abietic acid-derived optically active helical polymers and their chiral hydrogels, Bioresour Technol, 129 (2013) 58-64.

[7] C. Song, C. Zhang, F. Wang, W. Yang, J. Deng, Chiral polymeric microspheres grafted with optically active helical polymer chains: a new class of materials for chiral recognition and chirally controlled release, Polymer Chemistry, 4 (2013) 645-652.

[8] C. Zhang, D. Liu, B. Zhou, J. Deng, W. Yang, Poly(N-propargylamide)s bearing cholesteryl moieties: Preparation and optical activity, Reactive and Functional Polymers, 72 (2012) 832-838.

138

Presentation at Conferences:

2017.7

Poster presentation in conference NEXT-GEN ш: PV MATERIALS, Groningen Netherlands

2016.11

Oral presentation in conference 2016 MRS fall Meeting & Exhibit, Boston USA

2016.11

Poster presentation in International Congress Next Generation Solar Energy, Erlangen

Germany

2015.10

Poster presentation in 3rd Erlangen Symposium on Synthetic Carbon Allotropes 953,

Erlangen Germany

139

Appendix C Bibliography

[1] M.A. Green, Y. Hishikawa, W. Warta, E.D. Dunlop, D.H. Levi, J. Hohl-Ebinger, A.W.H.

Ho-Baillie, Solar cell efficiency tables (version 50), Progress in Photovoltaics: Research and

Applications, 25 (2017) 668-676.

[2] W. Zhao, S. Li, H. Yao, S. Zhang, Y. Zhang, B. Yang, J. Hou, Molecular Optimization

Enables over 13% Efficiency in Organic Solar Cells, Journal of the American Chemical

Society, (2017).

[3] K. Zhang, Z. Hu, C. Sun, Z. Wu, F. Huang, Y. Cao, Toward Solution-Processed High-

Performance Polymer Solar Cells: from Material Design to Device Engineering, Chemistry of

Materials, 29 (2017) 141-148.

[4] H. Kang, G. Kim, J. Kim, S. Kwon, H. Kim, K. Lee, Bulk-Heterojunction Organic Solar

Cells: Five Core Technologies for Their Commercialization, Advanced Materials, 28 (2016)

7821-7861.

[5] C.J. Brabec, Organic photovoltaics: technology and market, Sol. Energy Mater. Sol. Cells,

83 (2004) 273-292.

[6] J.W. Rumer, I. McCulloch, Organic photovoltaics: Crosslinking for optimal morphology

and stability, Materials Today, 18 (2015) 425-435.

[7] M. Jørgensen, K. Norrman, S.A. Gevorgyan, T. Tromholt, B. Andreasen, F.C. Krebs,

Stability of Polymer Solar Cells, Advanced Materials, 24 (2012) 580-612.

[8] S.A. Gevorgyan, N. Espinosa, L. Ciammaruchi, B. Roth, F. Livi, S. Tsopanidis, S. Zufle,

S. Queiros, A. Gregori, G.A.D. Benatto, M. Corazza, M.V. Madsen, M. Hosel, M.J. Beliatis,

T.T. Larsen-Olsen, F. Pastorelli, A. Castro, A. Mingorance, V. Lenzi, D. Fluhr, R. Roesch,

M.M.D. Ramos, A. Savva, H. Hoppe, L.S.A. Marques, I. Burgues, E. Georgiou, L. Serrano-

Lujan, F.C. Krebs, Baselines for Lifetime of Organic Solar Cells, Advanced Energy

Materials, 6 (2016) 1600910-n/a.

[9] P. Cheng, X. Zhan, Stability of organic solar cells: challenges and strategies, Chemical

140

Society Reviews, 45 (2016) 2544-2582.

[10] S.E. Shaheen, C.J. Brabec, N.S. Sariciftci, F. Padinger, T. Fromherz, J.C. Hummelen,

2.5% efficient organic plastic solar cells, Applied Physics Letters, 78 (2001) 841-843.

[11] M.M. Wienk, J.M. Kroon, W.J. Verhees, J. Knol, J.C. Hummelen, P.A. van Hal, R.A.

Janssen, Efficient methano [70] fullerene/MDMO-PPV bulk heterojunction photovoltaic

cells, Angewandte Chemie, 115 (2003) 3493-3497.

[12] G. Li, V. Shrotriya, J. Huang, Y. Yao, T. Moriarty, K. Emery, Y. Yang, High-efficiency

solution processable polymer photovoltaic cells by self-organization of polymer blends,

Nature materials, 4 (2005) 864-868.

[13] J. Peet, J.Y. Kim, N.E. Coates, W.L. Ma, D. Moses, A.J. Heeger, G.C. Bazan, Efficiency

enhancement in low-bandgap polymer solar cells by processing with alkane dithiols, Nature

materials, 6 (2007) 497-500.

[14] Y. Liang, Z. Xu, J. Xia, S.T. Tsai, Y. Wu, G. Li, C. Ray, L. Yu, For the bright future-bulk

heterojunction polymer solar cells with power conversion efficiency of 7.4%, Adv Mater, 22

(2010) E135-138.

[15] S.H. Liao, H.J. Jhuo, Y.S. Cheng, S.A. Chen, Fullerene Derivative-Doped Zinc Oxide

Nanofilm as the Cathode of Inverted Polymer Solar Cells with Low-Bandgap Polymer

(PTB7-Th) for High Performance, Advanced Materials, 25 (2013) 4766-4771.

[16] Z.C. He, B. Xiao, F. Liu, H.B. Wu, Y.L. Yang, S. Xiao, C. Wang, T.P. Russell, Y. Cao,

Single-junction polymer solar cells with high efficiency and photovoltage, Nature Photonics,

9 (2015) 174-179.

[17] Y. Liu, J. Zhao, Z. Li, C. Mu, W. Ma, H. Hu, K. Jiang, H. Lin, H. Ade, H. Yan,

Aggregation and morphology control enables multiple cases of high-efficiency polymer solar

cells, Nat Commun, 5 (2014) 5293.

[18] Y. Cui, H. Yao, B. Gao, Y. Qin, S. Zhang, B. Yang, C. He, B. Xu, J. Hou, Fine-Tuned

Photoactive and Interconnection Layers for Achieving over 13% Efficiency in a Fullerene-

Free Tandem Organic Solar Cell, Journal of the American Chemical Society, 139 (2017)

7302-7309.

[19] G. Pirotte, J. Kesters, P. Verstappen, S. Govaerts, J. Manca, L. Lutsen, D. Vanderzande,

W. Maes, Continuous Flow Polymer Synthesis toward Reproducible Large-Scale Production

141

for Efficient Bulk Heterojunction Organic Solar Cells, ChemSusChem, 8 (2015) 3228-3233.

[20] H. Hoppe, M. Seeland, B. Muhsin, Optimal geometric design of monolithic thin-film

solar modules: Architecture of polymer solar cells, Solar Energy Materials and Solar Cells,

97 (2012) 119-126.

[21] B. Muhsin, J. Renz, K.-H. Drüe, G. Gobsch, H. Hoppe, Influence of polymer solar cell

geometry on series resistance and device efficiency, physica status solidi (a), 206 (2009)

2771-2774.

[22] F.C. Krebs, M. Jørgensen, K. Norrman, O. Hagemann, J. Alstrup, T.D. Nielsen, J.

Fyenbo, K. Larsen, J. Kristensen, A complete process for production of flexible large area

polymer solar cells entirely using screen printing—First public demonstration, Solar Energy

Materials and Solar Cells, 93 (2009) 422-441.

[23] J. Lee, H. Back, J. Kong, D.-W. Park, K. Lee, New series connection method for bulk-

heterojunction polymer solar cell modules, Solar Energy Materials and Solar Cells, 98 (2012)

208-211.

[24] R.R. Søndergaard, M. Hösel, F.C. Krebs, Roll-to-Roll fabrication of large area functional

organic materials, Journal of Polymer Science Part B: Polymer Physics, 51 (2013) 16-34.

[25] M. Helgesen, J.E. Carlé, F.C. Krebs, Slot-Die Coating of a High Performance

Copolymer in a Readily Scalable Roll Process for Polymer Solar Cells, Advanced Energy

Materials, 3 (2013) 1664-1669.

[26] R. Søndergaard, M. Hösel, D. Angmo, T.T. Larsen-Olsen, F.C. Krebs, Roll-to-roll

fabrication of polymer solar cells, Materials Today, 15 (2012) 36-49.

[27] F.C. Krebs, T. Tromholt, M. Jorgensen, Upscaling of polymer solar cell fabrication using

full roll-to-roll processing, Nanoscale, 2 (2010) 873-886.

[28] E. Bundgaard, O. Hagemann, M. Manceau, M. Jørgensen, F.C. Krebs, Low Band Gap

Polymers for Roll-to-Roll Coated Polymer Solar Cells, Macromolecules, 43 (2010) 8115-

8120.

[29] F.C. Krebs, All solution roll-to-roll processed polymer solar cells free from indium-tin-

oxide and vacuum coating steps, Organic Electronics, 10 (2009) 761-768.

[30] M. Jørgensen, K. Norrman, F.C. Krebs, Stability/degradation of polymer solar cells,

Solar Energy Materials and Solar Cells, 92 (2008) 686-714.

142

[31] M.O. Reese, A.M. Nardes, B.L. Rupert, R.E. Larsen, D.C. Olson, M.T. Lloyd, S.E.

Shaheen, D.S. Ginley, G. Rumbles, N. Kopidakis, Photoinduced Degradation of Polymer and

Polymer–Fullerene Active Layers: Experiment and Theory, Advanced Functional Materials,

20 (2010) 3476-3483.

[32] M. Hermenau, M. Riede, K. Leo, S.A. Gevorgyan, F.C. Krebs, K. Norrman, Water and

oxygen induced degradation of small molecule organic solar cells, Solar Energy Materials

and Solar Cells, 95 (2011) 1268-1277.

[33] J. Adams, M. Salvador, L. Lucera, S. Langner, G.D. Spyropoulos, F.W. Fecher, M.M.

Voigt, S.A. Dowland, A. Osvet, H.-J. Egelhaaf, C.J. Brabec, Water Ingress in Encapsulated

Inverted Organic Solar Cells: Correlating Infrared Imaging and Photovoltaic Performance,

Advanced Energy Materials, 5 (2015) 1501065-n/a.

[34] M. Glatthaar, M. Riede, N. Keegan, K. Sylvester-Hvid, B. Zimmermann, M.

Niggemann, A. Hinsch, A. Gombert, Efficiency limiting factors of organic bulk

heterojunction solar cells identified by electrical impedance spectroscopy, Solar Energy

Materials and Solar Cells, 91 (2007) 390-393.

[35] K. Norrman, F.C. Krebs, Lifetimes of organic photovoltaics: Using TOF-SIMS and

18O2 isotopic labelling to characterise chemical degradation mechanisms, Solar Energy

Materials and Solar Cells, 90 (2006) 213-227.

[36] N. Li, C.J. Brabec, Air-processed polymer tandem solar cells with power conversion

efficiency exceeding 10%, Energy & Environmental Science, 8 (2015) 2902-2909.

[37] I.T. Sachs-Quintana, T. Heumüller, W.R. Mateker, D.E. Orozco, R. Cheacharoen, S.

Sweetnam, C.J. Brabec, M.D. McGehee, Electron Barrier Formation at the Organic-Back

Contact Interface is the First Step in Thermal Degradation of Polymer Solar Cells, Advanced

Functional Materials, 24 (2014) 3978-3985.

[38] T. Heumueller, W.R. Mateker, A. Distler, U.F. Fritze, R. Cheacharoen, W.H. Nguyen, M.

Biele, M. Salvador, M. von Delius, H.-J. Egelhaaf, M.D. McGehee, C.J. Brabec,

Morphological and electrical control of fullerene dimerization determines organic

photovoltaic stability, Energy & Environmental Science, 9 (2016) 247-256.

[39] J.D. Perea, S. Langner, M. Salvador, B. Sanchez-Lengeling, N. Li, C. Zhang, G. Jarvas,

J. Kontos, A. Dallos, A. Aspuru-Guzik, C.J. Brabec, Introducing a New Potential Figure of

143

Merit for Evaluating Microstructure Stability in Photovoltaic Polymer-Fullerene Blends, The

Journal of Physical Chemistry C, 121 (2017) 18153-18161.

[40] N. Li, J.D. Perea, T. Kassar, M. Richter, T. Heumueller, G.J. Matt, Y. Hou, N.S. Güldal,

H. Chen, S. Chen, S. Langner, M. Berlinghof, T. Unruh, C.J. Brabec, Abnormal strong burn-

in degradation of highly efficient polymer solar cells caused by spinodal donor-acceptor

demixing, Nature Communications, 8 (2017) 14541.

[41] G. Griffini, J.D. Douglas, C. Piliego, T.W. Holcombe, S. Turri, J.M.J. Fréchet, J.L.

Mynar, Long-Term Thermal Stability of High-Efficiency Polymer Solar Cells Based on

Photocrosslinkable Donor-Acceptor Conjugated Polymers, Advanced Materials, 23 (2011)

1660-1664.

[42] Z. Li, F. Wu, H. Lv, D. Yang, Z. Chen, X. Zhao, X. Yang, Side-Chain Engineering for

Enhancing the Thermal Stability of Polymer Solar Cells, Advanced Materials, 27 (2015)

6999-7003.

[43] Y.J. Cheng, C.H. Hsieh, P.J. Li, C.S. Hsu, Morphological stabilization by in situ

polymerization of fullerene derivatives leading to efficient, thermally stable organic

photovoltaics, Advanced Functional Materials, 21 (2011) 1723-1732.

[44] L. Derue, O. Dautel, A. Tournebize, M. Drees, H. Pan, S. Berthumeyrie, B. Pavageau, E.

Cloutet, S. Chambon, L. Hirsch, Thermal Stabilisation of Polymer–Fullerene Bulk

Heterojunction Morphology for Efficient Photovoltaic Solar Cells, Advanced Materials, 26

(2014) 5831-5838.

[45] P. Cheng, C. Yan, T.K. Lau, J. Mai, X. Lu, X. Zhan, Molecular Lock: A Versatile Key to

Enhance Efficiency and Stability of Organic Solar Cells, Adv Mater, 28 (2016) 5822-5829.

[46] C. Zhang, A. Mumyatov, S. Langner, J.D. Perea, T. Kassar, J. Min, L.L. Ke, H.W. Chen,

K.L. Gerasimov, D.V. Anokhin, D.A. Ivanov, T. Ameri, A. Osvet, D.K. Susarova, T. Unruh,

N. Li, P. Troshin, C.J. Brabec, Overcoming the Thermal Instability of Efficient Polymer Solar

Cells by Employing Novel Fullerene-Based Acceptors, Advanced Energy Materials, 7 (2017)

1601204.

[47] S.A. Dowland, M. Salvador, J.D. Perea, N. Gasparini, S. Langner, S. Rajoelson, H.H.

Ramanitra, B.D. Lindner, A. Osvet, C.J. Brabec, R.C. Hiorns, H.-J. Egelhaaf, Suppression of

Thermally Induced Fullerene Aggregation in Polyfullerene-Based Multiacceptor Organic

144

Solar Cells, ACS Applied Materials & Interfaces, 9 (2017) 10971-10982.

[48] P. Cheng, C. Yan, Y. Wu, J. Wang, M. Qin, Q. An, J. Cao, L. Huo, F. Zhang, L. Ding, Y.

Sun, W. Ma, X. Zhan, Alloy Acceptor: Superior Alternative to PCBM toward Efficient and

Stable Organic Solar Cells, Advanced Materials, 28 (2016) 8021-8028.

[49] W.R. Mateker, M.D. McGehee, Progress in Understanding Degradation Mechanisms and

Improving Stability in Organic Photovoltaics, Advanced Materials, 29 (2017) 1603940-n/a.

[50] K. Kawano, C. Adachi, Evaluating Carrier Accumulation in Degraded Bulk

Heterojunction Organic Solar Cells by a Thermally Stimulated Current Technique, Advanced

Functional Materials, 19 (2009) 3934-3940.

[51] T. Heumueller, W.R. Mateker, I.T. Sachs-Quintana, K. Vandewal, J.A. Bartelt, T.M.

Burke, T. Ameri, C.J. Brabec, M.D. McGehee, Reducing burn-in voltage loss in polymer

solar cells by increasing the polymer crystallinity, Energy & Environmental Science, 7 (2014)

2974-2980.

[52] T. Heumueller, T.M. Burke, W.R. Mateker, I.T. Sachs-Quintana, K. Vandewal, C.J.

Brabec, M.D. McGehee, Disorder-Induced Open-Circuit Voltage Losses in Organic Solar

Cells During Photoinduced Burn-In, Advanced Energy Materials, 5 (2015) 1500111.

[53] A. Distler, T. Sauermann, H.-J. Egelhaaf, S. Rodman, D. Waller, K.-S. Cheon, M. Lee,

D.M. Guldi, The Effect of PCBM Dimerization on the Performance of Bulk Heterojunction

Solar Cells, Advanced Energy Materials, 4 (2014) 1300693.

[54] Q. Burlingame, X. Tong, J. Hankett, M. Slootsky, Z. Chen, S.R. Forrest, Photochemical

origins of burn-in degradation in small molecular weight organic photovoltaic cells, Energy

& Environmental Science, 8 (2015) 1005-1010.

[55] R. Schueppel, K. Schmidt, C. Uhrich, K. Schulze, D. Wynands, J.L. Brédas, E. Brier, E.

Reinold, H.B. Bu, P. Baeuerle, B. Maennig, M. Pfeiffer, K. Leo, Optimizing organic

photovoltaics using tailored heterojunctions: A photoinduced absorption study of

oligothiophenes with low band gaps, Physical Review B, 77 (2008) 085311.

[56] T.M. Halasinski, D.M. Hudgins, F. Salama, L.J. Allamandola, T. Bally, Electronic

Absorption Spectra of Neutral Pentacene (C22H14) and Its Positive and Negative Ions in Ne,

Ar, and Kr Matrices, The Journal of Physical Chemistry A, 104 (2000) 7484-7491.

[57] G. Dennler, M.C. Scharber, C.J. Brabec, Polymer-fullerene bulk-heterojunction solar

145

cells, Adv. Mater. (Weinheim, Ger.), 21 (2009) 1323-1338.

[58] P. Schilinsky, C. Waldauf, C.J. Brabec, Recombination and loss analysis in

polythiophene based bulk heterojunction photodetectors, Appl. Phys. Lett., 81 (2002) 3885-

3887.

[59] L.J.A. Koster, V.D. Mihailetchi, R. Ramaker, P.W.M. Blom, Light intensity dependence

of open-circuit voltage of polymer : fullerene solar cells, Applied Physics Letters, 86 (2005)

123509.

[60] J. Kurpiers, D. Neher, Dispersive Non-Geminate Recombination in an Amorphous

Polymer:Fullerene Blend, Scientific Reports, 6 (2016) 26832.

[61] G.-J.A.H. Wetzelaer, N.J. Van der Kaap, L.J.A. Koster, P.W.M. Blom, Quantifying

Bimolecular Recombination in Organic Solar Cells in Steady State, Advanced Energy

Materials, 3 (2013) 1130-1134.

[62] R.A. Street, Localized state distribution and its effect on recombination in organic solar

cells, Physical Review B, 84 (2011) 075208.

[63] J.D. Servaites, M.A. Ratner, T.J. Marks, Organic solar cells: A new look at traditional

models, Energy & Environmental Science, 4 (2011) 4410-4422.

[64] L.J.A. Koster, M. Kemerink, M.M. Wienk, K. Maturová, R.A.J. Janssen, Quantifying

Bimolecular Recombination Losses in Organic Bulk Heterojunction Solar Cells, Advanced

Materials, 23 (2011) 1670-1674.

[65] A. Baumann, T.J. Savenije, D.H.K. Murthy, M. Heeney, V. Dyakonov, C. Deibel,

Influence of Phase Segregation on Recombination Dynamics in Organic Bulk-Heterojunction

Solar Cells, Advanced Functional Materials, 21 (2011) 1687-1692.

[66] S.R. Cowan, A. Roy, A.J. Heeger, Recombination in polymer-fullerene bulk

heterojunction solar cells, Physical Review B, 82 (2010) 245207.

[67] M.M. Mandoc, F.B. Kooistra, J.C. Hummelen, B.d. Boer, P.W.M. Blom, Effect of traps

on the performance of bulk heterojunction organic solar cells, Applied Physics Letters, 91

(2007) 263505.

[68] L.J.A. Koster, V.D. Mihailetchi, P.W.M. Blom, Bimolecular recombination in

polymer/fullerene bulk heterojunction solar cells, Applied Physics Letters, 88 (2006) 052104.

[69] I. Riedel, J. Parisi, V. Dyakonov, L. Lutsen, D. Vanderzande, J.C. Hummelen, Effect of

146

Temperature and Illumination on the Electrical Characteristics of Polymer–Fullerene Bulk-

Heterojunction Solar Cells, Advanced Functional Materials, 14 (2004) 38-44.

[70] P.J. Flory, Principles of polymer chemistry, Cornell University Press, 1953.

[71] J.E. Mark, Physical properties of polymers handbook, Springer, 2007.

[72] M. Rubinstein, R.H. Colby, Polymer Physics, Oxford University Press, 2003.

[73] R.L. Scott, M. Magat, Thermodynamics of high-polymer solutions. III. Swelling of

cross-linked rubber, Journal of Polymer Science, 4 (1949) 555-571.

[74] R.F. Blanks, J.M. Prausnitz, Thermodynamics of Polymer Solubility in Polar and

Nonpolar Systems, Industrial & Engineering Chemistry Fundamentals, 3 (1964) 1-8.

[75] C.M. Hansen, Hansen solubility parameters: a user's handbook, CRC press, 2007.

[76] F. Machui, S. Langner, X. Zhu, S. Abbott, C.J. Brabec, Determination of the P3HT:

PCBM solubility parameters via a binary solvent gradient method: Impact of solubility on the

photovoltaic performance, Solar Energy Materials and Solar Cells, 100 (2012) 138-146.

[77] I. Burgués-Ceballos, F. Machui, J. Min, T. Ameri, M.M. Voigt, Y.N. Luponosov, S.A.

Ponomarenko, P.D. Lacharmoise, M. Campoy-Quiles, C.J. Brabec, Solubility Based

Identification of Green Solvents for Small Molecule Organic Solar Cells, Advanced

Functional Materials, 24 (2014) 1449-1457.

[78] R. Roesch, T. Faber, E. von Hauff, T.M. Brown, M. Lira-Cantu, H. Hoppe, Procedures

and Practices for Evaluating Thin-Film Solar Cell Stability, Advanced Energy Materials, 5

(2015) 1501407-n/a.

[79] S. Cros, M. Firon, S. Lenfant, P. Trouslard, L. Beck, Study of thin calcium electrode

degradation by ion beam analysis, Nuclear Instruments and Methods in Physics Research

Section B: Beam Interactions with Materials and Atoms, 251 (2006) 257-260.

[80] F. So, D. Kondakov, Degradation Mechanisms in Small-Molecule and Polymer Organic

Light-Emitting Diodes, Advanced Materials, 22 (2010) 3762-3777.

[81] P.E. Burrows, V. Bulovic, S.R. Forrest, L.S. Sapochak, D.M. McCarty, M.E. Thompson,

Reliability and degradation of organic light emitting devices, Applied Physics Letters, 65

(1994) 2922-2924.

[82] J. Buseman-Williams, K.D. Frischknecht, M.D. Hubert, A.K. Saafir, J.D. Tremel, Flat-

plate encapsulation solution for OLED displays using a printed getter, Journal of the Society

147

for Information Display, 15 (2007) 103-112.

[83] K. Feron, T.J. Nagle, L.J. Rozanski, B.B. Gong, C.J. Fell, Spatially resolved

photocurrent measurements of organic solar cells: Tracking water ingress at edges and

pinholes, Solar Energy Materials and Solar Cells, 109 (2013) 169-177.

[84] M.T. Lloyd, D.C. Olson, P. Lu, E. Fang, D.L. Moore, M.S. White, M.O. Reese, D.S.

Ginley, J.W.P. Hsu, Impact of contact evolution on the shelf life of organic solar cells, Journal

of Materials Chemistry, 19 (2009) 7638-7642.

[85] S.K. Hau, H.-L. Yip, N.S. Baek, J. Zou, K. O’Malley, A.K.-Y. Jen, Air-stable inverted

flexible polymer solar cells using zinc oxide nanoparticles as an electron selective layer,

Applied Physics Letters, 92 (2008) 253301.

[86] M. Glatthaar, N. Mingirulli, B. Zimmermann, T. Ziegler, R. Kern, M. Niggemann, A.

Hinsch, A. Gombert, Impedance spectroscopy on organic bulk-heterojunction solar cells,

physica status solidi (a), 202 (2005) R125-R127.

[87] S.T. Lee, Z.Q. Gao, L.S. Hung, Metal diffusion from electrodes in organic light-emitting

diodes, Applied Physics Letters, 75 (1999) 1404-1406.

[88] R. Franke, B. Maennig, A. Petrich, M. Pfeiffer, Long-term stability of tandem solar cells

containing small organic molecules, Solar Energy Materials and Solar Cells, 92 (2008) 732-

735.

[89] D.J. Kang, H. Kang, K.-H. Kim, B.J. Kim, Nanosphere Templated Continuous

PEDOT:PSS Films with Low Percolation Threshold for Application in Efficient Polymer

Solar Cells, ACS Nano, 6 (2012) 7902-7909.

[90] S. Shao, J. Liu, J. Bergqvist, S. Shi, C. Veit, U. Würfel, Z. Xie, F. Zhang, In Situ

Formation of MoO3 in PEDOT:PSS Matrix: A Facile Way to Produce a Smooth and Less

Hygroscopic Hole Transport Layer for Highly Stable Polymer Bulk Heterojunction Solar

Cells, Advanced Energy Materials, 3 (2013) 349-355.

[91] Y. Sun, C.J. Takacs, S.R. Cowan, J.H. Seo, X. Gong, A. Roy, A.J. Heeger, Efficient, Air-

Stable Bulk Heterojunction Polymer Solar Cells Using MoOx as the Anode Interfacial Layer,

Advanced Materials, 23 (2011) 2226-2230.

[92] S.-P. Cho, J.-S. Yeo, D.-Y. Kim, S.-i. Na, S.-S. Kim, Brush painted V2O5 hole transport

layer for efficient and air-stable polymer solar cells, Solar Energy Materials and Solar Cells,

148

132 (2015) 196-203.

[93] M. Qiu, D. Zhu, X. Bao, J. Wang, X. Wang, R. Yang, WO3 with surface oxygen

vacancies as an anode buffer layer for high performance polymer solar cells, Journal of

Materials Chemistry A, 4 (2016) 894-900.

[94] M. Wang, Q. Tang, J. An, F. Xie, J. Chen, S. Zheng, K.Y. Wong, Q. Miao, J. Xu,

Performance and Stability Improvement of P3HT:PCBM-Based Solar Cells by Thermally

Evaporated Chromium Oxide (CrOx) Interfacial Layer, ACS Applied Materials & Interfaces,

2 (2010) 2699-2702.

[95] Y.-M. Chang, R. Zhu, E. Richard, C.-C. Chen, G. Li, Y. Yang, Electrostatic Self-

Assembly Conjugated Polyelectrolyte-Surfactant Complex as an Interlayer for High

Performance Polymer Solar Cells, Advanced Functional Materials, 22 (2012) 3284-3289.

[96] S. Bai, Y. Jin, X. Liang, Z. Ye, Z. Wu, B. Sun, Z. Ma, Z. Tang, J. Wang, U. Würfel, F.

Gao, F. Zhang, Ethanedithiol Treatment of Solution-Processed ZnO Thin Films: Controlling

the Intragap States of Electron Transporting Interlayers for Efficient and Stable Inverted

Organic Photovoltaics, Advanced Energy Materials, 5 (2015) 1401606-n/a.

[97] S. Venkatesan, E. Ngo, D. Khatiwada, C. Zhang, Q. Qiao, Enhanced Lifetime of

Polymer Solar Cells by Surface Passivation of Metal Oxide Buffer Layers, ACS Applied

Materials & Interfaces, 7 (2015) 16093-16100.

[98] B.A. MacLeod, B.J. Tremolet de Villers, P. Schulz, P.F. Ndione, H. Kim, A.J. Giordano,

K. Zhu, S.R. Marder, S. Graham, J.J. Berry, A. Kahn, D.C. Olson, Stability of inverted

organic solar cells with ZnO contact layers deposited from precursor solutions, Energy &

Environmental Science, 8 (2015) 592-601.

[99] A. Rivaton, A. Tournebize, J. Gaume, P.-O. Bussière, J.-L. Gardette, S. Therias,

Photostability of organic materials used in polymer solar cells, Polymer International, 63

(2014) 1335-1345.

[100] A. Rivaton, J.-L. Gardette, B. Mailhot, S. Morlat-Therlas, Basic Aspects of Polymer

Degradation, Macromolecular Symposia, 225 (2005) 129-146.

[101] I. Fraga Domínguez, P.D. Topham, P.-O. Bussière, D. Bégué, A. Rivaton, Unravelling

the Photodegradation Mechanisms of a Low Bandgap Polymer by Combining Experimental

and Modeling Approaches, The Journal of Physical Chemistry C, 119 (2015) 2166-2176.

149

[102] J. Razzell-Hollis, J. Wade, W.C. Tsoi, Y. Soon, J. Durrant, J.-S. Kim, Photochemical

stability of high efficiency PTB7:PC70BM solar cell blends, Journal of Materials Chemistry

A, 2 (2014) 20189-20195.

[103] A. Tournebize, P.-O. Bussière, P. Wong-Wah-Chung, S. Thérias, A. Rivaton, J.-L.

Gardette, S. Beaupré, M. Leclerc, Impact of UV-Visible Light on the Morphological and

Photochemical Behavior of a Low-Bandgap Poly(2,7-Carbazole) Derivative for Use in High-

Performance Solar Cells, Advanced Energy Materials, 3 (2013) 478-487.

[104] A. Dupuis, P. Wong-Wah-Chung, A. Rivaton, J.-L. Gardette, Influence of the

microstructure on the photooxidative degradation of poly(3-hexylthiophene), Polymer

Degradation and Stability, 97 (2012) 366-374.

[105] M. Manceau, A. Rivaton, J.-L. Gardette, S. Guillerez, N. Lemaître, The mechanism of

photo- and thermooxidation of poly(3-hexylthiophene) (P3HT) reconsidered, Polymer

Degradation and Stability, 94 (2009) 898-907.

[106] S. Chambon, A. Rivaton, J.-L. Gardette, M. Firon, Reactive intermediates in the

initiation step of the photo-oxidation of MDMO-PPV, Journal of Polymer Science Part A:

Polymer Chemistry, 47 (2009) 6044-6052.

[107] L.A. Frolova, N.P. Piven, D.K. Susarova, A.V. Akkuratov, S.D. Babenko, P.A. Troshin,

ESR spectroscopy for monitoring the photochemical and thermal degradation of conjugated

polymers used as electron donor materials in organic bulk heterojunction solar cells,

Chemical Communications, 51 (2015) 2242-2244.

[108] M. Manceau, M. Helgesen, F.C. Krebs, Thermo-cleavable polymers: Materials with

enhanced photochemical stability, Polymer Degradation and Stability, 95 (2010) 2666-2669.

[109] W.R. Mateker, T. Heumueller, R. Cheacharoen, I.T. Sachs-Quintana, M.D. McGehee, J.

Warnan, P.M. Beaujuge, X. Liu, G.C. Bazan, Molecular Packing and Arrangement Govern

the Photo-Oxidative Stability of Organic Photovoltaic Materials, Chemistry of Materials, 27

(2015) 6345-6353.

[110] Y.W. Soon, S. Shoaee, R.S. Ashraf, H. Bronstein, B.C. Schroeder, W. Zhang, Z. Fei, M.

Heeney, I. McCulloch, J.R. Durrant, Material Crystallinity as a Determinant of Triplet

Dynamics and Oxygen Quenching in Donor Polymers for Organic Photovoltaic Devices,

Advanced Functional Materials, 24 (2014) 1474-1482.

150

[111] E.T. Hoke, I.T. Sachs-Quintana, M.T. Lloyd, I. Kauvar, W.R. Mateker, A.M. Nardes,

C.H. Peters, N. Kopidakis, M.D. McGehee, The Role of Electron Affinity in Determining

Whether Fullerenes Catalyze or Inhibit Photooxidation of Polymers for Solar Cells,

Advanced Energy Materials, 2 (2012) 1351-1357.

[112] A.J. Heeger, 25th anniversary article: Bulk heterojunction solar cells: understanding the

mechanism of operation, Adv Mater, 26 (2014) 10-27.

[113] I.C. Wang, L.A. Tai, D.D. Lee, P.P. Kanakamma, C.K.F. Shen, T.-Y. Luh, C.H. Cheng,

K.C. Hwang, C60 and Water-Soluble Fullerene Derivatives as Antioxidants Against Radical-

Initiated Lipid Peroxidation, Journal of Medicinal Chemistry, 42 (1999) 4614-4620.

[114] Q. Tai, J. Li, Z. Liu, Z. Sun, X. Zhao, F. Yan, Enhanced photovoltaic performance of

polymer solar cells by adding fullerene end-capped polyethylene glycol, Journal of Materials

Chemistry, 21 (2011) 6848-6853.

[115] M.-H. Liao, C.-E. Tsai, Y.-Y. Lai, F.-Y. Cao, J.-S. Wu, C.-L. Wang, C.-S. Hsu, I. Liau,

Y.-J. Cheng, Morphological Stabilization by Supramolecular Perfluorophenyl-C60

Interactions Leading to Efficient and Thermally Stable Organic Photovoltaics, Advanced

Functional Materials, 24 (2014) 1418-1429.

[116] Z. Li, H.C. Wong, Z. Huang, H. Zhong, C.H. Tan, W.C. Tsoi, J.S. Kim, J.R. Durrant,

J.T. Cabral, Performance enhancement of fullerene-based solar cells by light processing, Nat

Commun, 4 (2013) 2227.

[117] H.C. Wong, Z. Li, C.H. Tan, H. Zhong, Z. Huang, H. Bronstein, I. McCulloch, J.T.

Cabral, J.R. Durrant, Morphological Stability and Performance of Polymer–Fullerene Solar

Cells under Thermal Stress: The Impact of Photoinduced PC60BM Oligomerization, ACS

Nano, 8 (2014) 1297-1308.

[118] C.H. Peters, I.T. Sachs-Quintana, W.R. Mateker, T. Heumueller, J. Rivnay, R. Noriega,

Z.M. Beiley, E.T. Hoke, A. Salleo, M.D. McGehee, The Mechanism of Burn-in Loss in a

High Efficiency Polymer Solar Cell, Advanced Materials, 24 (2012) 663-668.

[119] R.A. Street, D.M. Davies, Kinetics of light induced defect creation in organic solar

cells, Applied Physics Letters, 102 (2013) 043305.

[120] R.A. Street, J.E. Northrup, B.S. Krusor, Radiation induced recombination centers in

organic solar cells, Physical Review B, 85 (2012) 205211.

151

[121] R.A. Street, A. Krakaris, S.R. Cowan, Recombination Through Different Types of

Localized States in Organic Solar Cells, Advanced Functional Materials, 22 (2012) 4608-

4619.

[122] S.R. Cowan, W.L. Leong, N. Banerji, G. Dennler, A.J. Heeger, Identifying a Threshold

Impurity Level for Organic Solar Cells: Enhanced First-Order Recombination Via Well-

Defined PC84BM Traps in Organic Bulk Heterojunction Solar Cells, Advanced Functional

Materials, 21 (2011) 3083-3092.

[123] G. Garcia-Belmonte, P.P. Boix, J. Bisquert, M. Lenes, H.J. Bolink, A. La Rosa, S.

Filippone, N. Martín, Influence of the intermediate density-of-states occupancy on open-

circuit voltage of bulk heterojunction solar cells with different fullerene acceptors, The

Journal of Physical Chemistry Letters, 1 (2010) 2566-2571.

[124] J.C. Blakesley, D. Neher, Relationship between energetic disorder and open-circuit

voltage in bulk heterojunction organic solar cells, Physical Review B, 84 (2011) 075210.

[125] G. Wantz, L. Derue, O. Dautel, A. Rivaton, P. Hudhomme, C. Dagron-Lartigau,

Stabilizing polymer-based bulk heterojunction solar cells via crosslinking, Polymer

International, 63 (2014) 1346-1361.

[126] D.K. Susarova, N.P. Piven, A.V. Akkuratov, L.A. Frolova, M.S. Polinskaya, S.A.

Ponomarenko, S.D. Babenko, P.A. Troshin, ESR spectroscopy as a powerful tool for probing

the quality of conjugated polymers designed for photovoltaic applications, Chemical

Communications, 51 (2015) 2239-2241.

[127] N. Wang, X. Tong, Q. Burlingame, J. Yu, S.R. Forrest, Photodegradation of small-

molecule organic photovoltaics, Solar Energy Materials and Solar Cells, 125 (2014) 170-175.

[128] A.M. Rao, P. Zhou, K.-A. Wang, G.T. Hager, J.M. Holden, Y. Wang, W.-T. Lee, X.-X.

Bi, P.C. Eklund, D.S. Cornett, M.A. Duncan, I.J. Amster, Photoinduced Polymerization of

Solid C₆₀ Films, Science, 259 (1993) 955-957.

[129] A. Dzwilewski, T. Wågberg, L. Edman, Photo-Induced and Resist-Free Imprint

Patterning of Fullerene Materials for Use in Functional Electronics, Journal of the American

Chemical Society, 131 (2009) 4006-4011.

[130] J. Nelson, The Physics of Solar Cells, Imperial College Press, London, 2003.

[131] P.N. Murgatroyd, Theory of space-charge-limited current enhanced by Frenkel effect,

152

Journal of Physics D: Applied Physics, 3 (1970) 151.

[132] F. Giacalone, N. Martín, Fullerene Polymers:  Synthesis and Properties, Chemical

Reviews, 106 (2006) 5136-5190.

[133] B. Gholamkhass, T.J. Peckham, S. Holdcroft, Poly(3-hexylthiophene) bearing pendant

fullerenes: aggregation vs. self-organization, Polymer Chemistry, 1 (2010) 708-719.

[134] H.H. Ramanitra, S.A. Dowland, B.A. Bregadiolli, M. Salvador, H. Santos Silva, D.

Begue, C.F.O. Graeff, H. Peisert, T. Chasse, S. Rajoelson, A. Osvet, C.J. Brabec, H.-J.

Egelhaaf, G.E. Morse, A. Distler, R.C. Hiorns, Increased thermal stabilization of polymer

photovoltaic cells with oligomeric PCBM, Journal of Materials Chemistry C, 4 (2016) 8121-

8129.

[135] H.H. Ramanitra, H. Santos Silva, B.A. Bregadiolli, A. Khoukh, C.M.S. Combe, S.A.

Dowland, D. Bégué, C.F.O. Graeff, C. Dagron-Lartigau, A. Distler, G. Morse, R.C. Hiorns,

Synthesis of Main-Chain Poly(fullerene)s from a Sterically Controlled Azomethine Ylide

Cycloaddition Polymerization, Macromolecules, 49 (2016) 1681-1691.

[136] M. Stephen, H.H. Ramanitra, H. Santos Silva, S. Dowland, D. Begue, K. Genevicius,

K. Arlauskas, G. Juska, G.E. Morse, A. Distler, R.C. Hiorns, Sterically controlled azomethine

ylide cycloaddition polymerization of phenyl-C61-butyric acid methyl ester, Chemical

Communications, 52 (2016) 6107-6110.

[137] L.K. Wasserthal, A. Kratzer, A. Hirsch, Sequential Fullerenylation of Bis-malonates –

Efficient Access to Oligoclusters with Different Fullerene Building Blocks, European Journal

of Organic Chemistry, 2013 (2013) 2355-2361.

[138] T. Wei, M.E. Perez-Ojeda, A. Hirsch, The first molecular dumbbell consisting of an

endohedral Sc3N@C80 and an empty C60-fullerene building block, Chemical

Communications, 53 (2017) 7886-7889.

[139] A.V. Mumyatov, F.A. Prudnov, L.N. Inasaridze, O.A. Mukhacheva, P.A. Troshin, High

LUMO energy pyrrolidinofullerenes as promising electron-acceptor materials for organic

solar cells, Journal of Materials Chemistry C, 3 (2015) 11612-11617.

[140] A.V. Mumyatov, Institute of Problems of Chemical Physics of the Russian Academy of

Sciences Chernogolovka, Moscow Region 142432 (RU), LANXESS Deutschland GmbH

50569 Köln (DE), in: E.P. Office (Ed.) EUROPEAN PATENT, Germany, 2014.

153

[141] C. Müller, T.A.M. Ferenczi, M. Campoy-Quiles, J.M. Frost, D.D.C. Bradley, P. Smith,

N. Stingelin-Stutzmann, J. Nelson, Binary Organic Photovoltaic Blends: A Simple Rationale

for Optimum Compositions, Advanced Materials, 20 (2008) 3510-3515.

[142] R.A. Orwoll, P.A. Arnold, Polymer–Solvent Interaction Parameter χ, in: J.E. Mark (Ed.)

Physical Properties of Polymers Handbook, Springer New York, New York, NY, 2007, pp.

233-257.

[143] J.G. Tait, T. Merckx, W. Li, C. Wong, R. Gehlhaar, D. Cheyns, M. Turbiez, P.

Heremans, Determination of solvent systems for blade coating thin film photovoltaics,

Advanced Functional Materials, 25 (2015) 3393-3398.

[144] J.D. Perea, S. Langner, M. Salvador, J. Kontos, G. Jarvas, F. Winkler, F. Machui, A.

Görling, A. Dallos, T. Ameri, C.J. Brabec, Combined Computational Approach Based on

Density Functional Theory and Artificial Neural Networks for Predicting The Solubility

Parameters of Fullerenes, The Journal of Physical Chemistry B, 120 (2016) 4431-4438.

[145] J.A. Bartelt, Z.M. Beiley, E.T. Hoke, W.R. Mateker, J.D. Douglas, B.A. Collins, J.R.

Tumbleston, K.R. Graham, A. Amassian, H. Ade, J.M.J. Fréchet, M.F. Toney, M.D.

McGehee, The Importance of Fullerene Percolation in the Mixed Regions of Polymer–

Fullerene Bulk Heterojunction Solar Cells, Advanced Energy Materials, 3 (2013) 364-374.

[146] B.A. Collins, Z. Li, J.R. Tumbleston, E. Gann, C.R. McNeill, H. Ade, Absolute

Measurement of Domain Composition and Nanoscale Size Distribution Explains

Performance in PTB7:PC71BM Solar Cells, Advanced Energy Materials, 3 (2013) 65-74.

[147] P. Westacott, J.R. Tumbleston, S. Shoaee, S. Fearn, J.H. Bannock, J.B. Gilchrist, S.

Heutz, J. deMello, M. Heeney, H. Ade, J. Durrant, D.S. McPhail, N. Stingelin, On the role of

intermixed phases in organic photovoltaic blends, Energy & Environmental Science, 6 (2013)

2756-2764.

[148] A.C. Mayer, M.F. Toney, S.R. Scully, J. Rivnay, C.J. Brabec, M. Scharber, M. Koppe,

M. Heeney, I. McCulloch, M.D. McGehee, Bimolecular crystals of fullerenes in conjugated

polymers and the implications of molecular mixing for solar cells, Adv. Funct. Mater., 19

(2009) 1173-1179.

[149] E. Buchaca-Domingo, A.J. Ferguson, F.C. Jamieson, T. McCarthy-Ward, S. Shoaee,

J.R. Tumbleston, O.G. Reid, L. Yu, M.B. Madec, M. Pfannmoller, F. Hermerschmidt, R.R.

154

Schroder, S.E. Watkins, N. Kopidakis, G. Portale, A. Amassian, M. Heeney, H. Ade, G.

Rumbles, J.R. Durrant, N. Stingelin, Additive-assisted supramolecular manipulation of

polymer:fullerene blend phase morphologies and its influence on photophysical processes,

Materials Horizons, 1 (2014) 270-279.

[150] E. Buchaca-Domingo, K. Vandewal, Z. Fei, S.E. Watkins, F.H. Scholes, J.H. Bannock,

J.C. de Mello, L.J. Richter, D.M. DeLongchamp, A. Amassian, M. Heeney, A. Salleo, N.

Stingelin, Direct Correlation of Charge Transfer Absorption with Molecular Donor:Acceptor

Interfacial Area via Photothermal Deflection Spectroscopy, Journal of the American

Chemical Society, 137 (2015) 5256-5259.

[151] M. Scarongella, J. De Jonghe-Risse, E. Buchaca-Domingo, M. Causa’, Z. Fei, M.

Heeney, J.-E. Moser, N. Stingelin, N. Banerji, A Close Look at Charge Generation in

Polymer:Fullerene Blends with Microstructure Control, Journal of the American Chemical

Society, 137 (2015) 2908-2918.

[152] M. Causa, J. De Jonghe-Risse, M. Scarongella, J.C. Brauer, E. Buchaca-Domingo, J.-E.

Moser, N. Stingelin, N. Banerji, The fate of electron–hole pairs in polymer:fullerene blends

for organic photovoltaics, Nature Communications, 7 (2016) 12556.

[153] A.M. Ballantyne, T.A. Ferenczi, M. Campoy-Quiles, T.M. Clarke, A. Maurano, K.H.

Wong, W. Zhang, N. Stingelin-Stutzmann, J.-S. Kim, D.D. Bradley, Understanding the

influence of morphology on poly (3-hexylselenothiophene): PCBM solar cells,

Macromolecules, 43 (2010) 1169.

[154] N.D. Treat, A. Varotto, C.J. Takacs, N. Batara, M. Al-Hashimi, M.J. Heeney, A.J.

Heeger, F. Wudl, C.J. Hawker, M.L. Chabinyc, Polymer-Fullerene Miscibility: A Metric for

Screening New Materials for High-Performance Organic Solar Cells, Journal of the American

Chemical Society, 134 (2012) 15869-15879.

[155] K. Vakhshouri, D.R. Kozub, C. Wang, A. Salleo, E.D. Gomez, Effect of miscibility and

percolation on electron transport in amorphous poly (3-hexylthiophene)/phenyl-C 61-butyric

acid methyl ester blends, Physical review letters, 108 (2012) 026601.

[156] S. Ulum, N. Holmes, M. Barr, A.D. Kilcoyne, B.B. Gong, X. Zhou, W. Belcher, P.

Dastoor, The role of miscibility in polymer: fullerene nanoparticulate organic photovoltaic

devices, Nano Energy, 2 (2013) 897-905.

155

[157] M. Jorgensen, K. Norrman, S.A. Gevorgyan, T. Tromholt, B. Andreasen, F.C. Krebs,

Stability of polymer solar cells, Adv Mater, 24 (2012) 580-612.

[158] D. Chirvase, J. Parisi, J.C. Hummelen, V. Dyakonov, Influence of nanomorphology on

the photovoltaic action of polymer–fullerene composites, Nanotechnology, 15 (2004) 1317.

[159] G. Dennler, A.J. Mozer, G. Juška, A. Pivrikas, R. Österbacka, A. Fuchsbauer, N.S.

Sariciftci, Charge carrier mobility and lifetime versus composition of conjugated

polymer/fullerene bulk-heterojunction solar cells, Organic Electronics, 7 (2006) 229-234.

[160] D. Veldman, Ö. İpek, S.C.J. Meskers, J. Sweelssen, M.M. Koetse, S.C. Veenstra, J.M.

Kroon, S.S. van Bavel, J. Loos, R.A.J. Janssen, Compositional and Electric Field Dependence

of the Dissociation of Charge Transfer Excitons in Alternating Polyfluorene

Copolymer/Fullerene Blends, Journal of the American Chemical Society, 130 (2008) 7721-

7735.

[161] S.H. Park, A. Roy, S. Beaupre, S. Cho, N. Coates, J.S. Moon, D. Moses, M. Leclerc, K.

Lee, A.J. Heeger, Bulk heterojunction solar cells with internal quantum efficiency

approaching 100%, Nat Photon, 3 (2009) 297-302.

[162] D.H. Wang, J.K. Kim, J.H. Seo, O.O. Park, J.H. Park, Stability comparison: A

PCDTBT/PC71BM bulk-heterojunction versus a P3HT/PC71BM bulk-heterojunction, Solar

Energy Materials and Solar Cells, 101 (2012) 249-255.

[163] S. Alem, R. de Bettignies, J.-M. Nunzi, M. Cariou, Efficient polymer-based

interpenetrated network photovoltaic cells, Applied physics letters, 84 (2004) 2178-2180.

[164] J.Y. Kim, K. Lee, N.E. Coates, D. Moses, T.-Q. Nguyen, M. Dante, A.J. Heeger,

Efficient tandem polymer solar cells fabricated by all-solution processing, Science, 317

(2007) 222-225.

[165] E. Wang, L. Hou, Z. Wang, S. Hellström, F. Zhang, O. Inganäs, M.R. Andersson, An

Easily Synthesized Blue Polymer for High-Performance Polymer Solar Cells, Advanced

Materials, 22 (2010) 5240-5244.

[166] M.T. Dang, L. Hirsch, G. Wantz, P3HT: PCBM, best seller in polymer photovoltaic

research, Advanced Materials, 23 (2011) 3597-3602.

[167] V.S. Gevaerts, A. Furlan, M.M. Wienk, M. Turbiez, R.A. Janssen, Solution processed

polymer tandem solar cell using efficient small and wide bandgap polymer: fullerene blends,

156

Advanced Materials, 24 (2012) 2130-2134.

[168] N. Li, D. Baran, K. Forberich, F. Machui, T. Ameri, M. Turbiez, M. Carrasco-Orozco,

M. Drees, A. Facchetti, F.C. Krebs, C.J. Brabec, Towards 15% energy conversion efficiency:

a systematic study of the solution-processed organic tandem solar cells based on

commercially available materials, Energy Environ. Sci., 6 (2013) 3407-3413.

[169] F. Machui, S. Abbott, D. Waller, M. Koppe, C.J. Brabec, Determination of solubility

parameters for organic semiconductor formulations, Macromolecular chemistry and Physics,

212 (2011) 2159-2165.

[170] D.T. Duong, B. Walker, J. Lin, C. Kim, J. Love, B. Purushothaman, J.E. Anthony, T.Q.

Nguyen, Molecular solubility and hansen solubility parameters for the analysis of phase

separation in bulk heterojunctions, Journal of Polymer Science Part B: Polymer Physics, 50

(2012) 1405-1413.

[171] J.D. Perea, S. Langner, M. Salvador, B. Sanchez-Lengeling, N. Li, C. Zhang, G. Jarvas,

J. Kontos, A. Dallos, A. Aspuru-Guzik, C.J. Brabec, Introducing a Figure of Merit for

Predicting Microstructure Formation in Organic Photovoltaic Blends, The Journal of Physical

Chemistry C, (2017).

[172] N. Li, F. Machui, D. Waller, M. Koppe, C.J. Brabec, Determination of phase diagrams

of binary and ternary organic semiconductor blends for organic photovoltaic devices, Sol.

Energy Mater. Sol. Cells, 95 (2011) 3465-3471.

[173] P. Scherrer, Bestimmung der Größe und der inneren Struktur von Kolloidteilchen

mittels Röntgenstrahlen, Göttinger Nachr. Math. Phys., 2 (1918).

[174] A.J. Parnell, A.J. Cadby, O.O. Mykhaylyk, A.D.F. Dunbar, P.E. Hopkinson, A.M.

Donald, R.A.L. Jones, Nanoscale Phase Separation of P3HT PCBM Thick Films As

Measured by Small-Angle X-ray Scattering, Macromolecules, 44 (2011) 6503-6508.

[175] F. Liu, W. Zhao, J.R. Tumbleston, C. Wang, Y. Gu, D. Wang, A.L. Briseno, H. Ade, T.P.

Russell, Understanding the Morphology of PTB7:PCBM Blends in Organic Photovoltaics,

Advanced Energy Materials, 4 (2014) 1301377-n/a.

[176] T.L. Lin, U. Jeng, C.S. Tsao, W.J. Liu, T. Canteenwala, L.Y. Chiang, Effect of Arm

Length on the Aggregation Structure of Fullerene-Based Star Ionomers, The Journal of

Physical Chemistry B, 108 (2004) 14884-14888.

157

[177] C. Durniak, M. Ganeva, G. Pospelov, W. Van Herck, J. Wuttke, BornAgain-Software

for simulating and fitting X-ray and neutron small-angle scattering at grazing incidence,

version 1.9, (2015) http://www.bornagainproject.org.

[178] C.J.B. K. Burlafinge, 2017 submitted

[179] N. Gasparini, M. Salvador, T. Heumueller, M. Richter, A. Classen, S. Shrestha, G.J.

Matt, S. Holliday, S. Strohm, H.-J. Egelhaaf, A. Wadsworth, D. Baran, I. McCulloch, C.J.

Brabec, Polymer:Nonfullerene Bulk Heterojunction Solar Cells with Exceptionally Low

Recombination Rates, Advanced Energy Materials, (2017) 1701561.

[180] N. Felekidis, A. Melianas, M. Kemerink, Nonequilibrium drift-diffusion model for

organic semiconductor devices, Physical Review B, 94 (2016) 035205.