Thin films of [6,6]-phenyl-C 61

(1)

Thin films of [6,6]-phenyl-C 61

-butyric acid methyl ester for

application in organic solar cells

preparation and effects of exposure to light and heat Thomas van Pelt

Faculty of Health, Science And Technology Chemistry, Bachelor Thesis

15 ECTS Ellen Moons Maria Rova

(2)

(3)

methyl ester for application in organic solar cells

—

preparation and effects of exposure to light and heat

Thomas van Pelt

Thesis submitted in fulfilment of the requirements for the degree of

’Professional Bachelor in Chemistry, Chemistry Major’

Leuven, 2013

(4)

methyl ester for application in organic solar cells

—

preparation and effects of exposure to light and heat

Thomas van Pelt

Promotor: dr Ellen Moons Mentor: dr Herman Faes

Department of Physics and Electrical Engineering Karlstad University

651 88 Karlstad Sweden

(5)

(6)

Acknowledgements

‘The acknowledgements’, the place to acknowledge, maybe even confess a little who helped, in- spired and supported me, the place to thank some people that helped me, directly or indirectly, to get this bachelor thesis to the point as it is now.

Maybe a bit traditional, but I would really like to start by thanking my promoter, Ellen Moons.

She gave me this opportunity, took me into her lab, far passed all deadlines.

Griet t’Servranckx, helped me — and near the end even pushing me — to get everything organ- ised and realise my ‘Swedish adventure’, I am so glad she did.

I would like to thank the people from the department physics and electrical engineering at Karl- stad (now I am sure I have not forgotten anyone), but two people in particular: Rickard Hansson and Ana Sofia Anselmo. You helped me out whenever I had any questions, problems or was in need of advice, thank you so much for your help (and company).

There is also one person who contributed directly to this thesis, but has no scientific background.

I would like to thank Thomas Gaumont for being ‘my private photographer’, helping me image some very reflective samples I was unable to capture myself.

Maya, hoewel deze stage op het vlak van werk voorbij gevlogen is, was een half jaar toch lang om gescheiden te zijn, dankje dat je me altijd in mijn keuze bent blijven steunen.

Papa en mama, heel mijn leven hebben jullie me aangemoedigd om nieuwe ervaringen op te doen, zelfstandig te zijn(/worden). . . Tegelijkertijd waren jullie (zoals altijd) er voor mij, steun- den jullie me altijd.

Lastly I would like to thank you as a reader. I am aware that this is ‘only’ my bachelor thesis and will most likely not change the world, but it still took a lot of effort. . .

My thanks to all of you, and all the other people I forgot to mention here.

(7)

Solution-processed organic solar cells, typically consist of an active layer of two components: an electron-accepting material and a electron-donating material. A commonly used electon-acceptor is the fullerene derivative[6, 6]-phenyl-C61-butyric acid methyl ester (PCBM). In this thesis, the effects of light and heat on thin films of pristine PCBM are examined.

Various PCBM films were prepared by spin-coating using different solution recipes, with either chloroform and chlorobenzene as solvent. These films were subsequently assessed visually and using optical microscopy. Numerous micrographs were evaluated. A rating system was formulated to order these samples according to their smoothness — based on the most common defects — to find the preparation conditions that resulted in a smooth, homogeneous film that covered the entire substrate. Prolonged agitation and/or heating had a negative effect on the smoothness of the resulting films. The most optimal film was obtained from a chlorobenzene solution with short agitation. Subsequently, the preparation recipe uses a shortly-stirred PCBM solution in chlorobenzene, without further agitation or heating.

Exposing PCBM thin films to light, showed no significant effect on the film surface when examined by atomic force microscopy (AFM). However, exposure of the PCBM films led to a transition from a soluble to an insoluble state, observed when immersing the PCBM films in a toluene or chloroform:acetone mixture. Analysis using UV/Vis-spectroscopy shows a clear spectral change in the visible region, confirming a chemical change to the molecules of the film.

The smooth and homogeneous surface of a PCBM film underwent a significant change, when exposed to heat (200^◦C) both in air and vacuum. When the annealed film was prepared from an unheated solution, AFM images of the surface showed the formation of randomly orientated crystallites of a sub micrometre size. Films prepared from a heated solution showed the formation of crystallites that are aligned. Additionally, in optical micrographs, the latter also showed the formation of star-like crystals, several micrometres in diameter, that are absent in the case of films spin-coated from an unheated solution. This led to the conclusion that heating as well as the way in which PCBM solutions are prepared, has a significant impact on PCBM thin films.

(8)

Uit oplossing bereide organische zonnecellen bevatten meestal een actieve laag die is opgebouwd uit twee componenten: een elektronenaccepterend materiaal en een elektronendonerend materiaal. Een veel gebruikte elektronenacceptor is de fullereenderivaat[6, 6]-phenyl-C61-butyric acid methyl ester (PCBM). In deze thesis is het effect van licht en warmte op gespincoatte dunnen lagen van zuivere PCBM onderzocht.

Verscheidene PCBM-films werden bereid via spincoating, gebruikmakend van oplossingen met verschillende bereidingswijzen, zowel met chloroform als chlorobenzeen als solvent. Deze films werden achtereenvolgens visueel en met behulp van optische microscopie geanalyseerd.

Vele optische microscopie beelden werden geanalyseerd. Een scoringssysteem werd geformu- leerd om deze stalen te ordenen volgens hun gladheid — gebaseerd op de meest voorkomende defecten — om de bereidingscondities te bepalen, die resulteerde in een effen, gladde en homogene film die het hele substraat dekte. Hieruit bleek dat langdurig mengen en/of verwarmen, een negatieve impact heeft op de effen- en gladheid van de film. De meest optimale film werd bekomen van een kort geagiteerde chlorobenzeen-oplossing. En dus is voorgestelde berei- dingswijze: een kort geroerde PCBM in chlorobenzeen oplossing, zonder verdere agitatie of ver- warming.

Dunne films van PCBM blootstellen aan licht, vertoonde geen significant effect op het oppervlak van de film wanneer onderzocht met behulp van atomairekrachtmicroscopie (AFM). Maar blootstelling van PCBM-films leidde wel tot een overgang van een oplosbare naar een onoplos- bare staat, wanneer ondergedompeld in toluene of een chloroform:aceton-mengsel. Analyse met behulp van UV/Vis-spectroscopie toonde een duidelijke spectrale verandering in het zichtbare gebied. Dit bevestigde een chemische verandering van de moleculen van de film.

Het gladde en homogene oppervlak van een PCBM-film onderging een significante verandering wanneer deze werd blootgesteld aan warmte (200^◦C), zowel in lucht als vacuüm. Wanneer de geannealde film bereid werd van een onverwarmde oplossing, vertoonde de AFM-beelden van het oppervlak een vorming van willekeurig georiënteerde kristallieten met een submicrom- eter grootte. Films bereid van een verwarmde oplossing vertoonde kristallieten die zich parallel oriënteerden. Optische microscoopbeelden van deze laatsten vertoonden eveneens gevormde stervormige kristallen, met een diameter van enkele micrometers. Deze kristallen werden niet waargenomen bij films van een onverwarmde oplossing. Dit leidde tot de conclusie dat zowel verwarmen als de manier waarop de PCBM oplossing bereid werd, een significante impact had op de resulterende PCBM film.

(9)

Organiska solceller, tillverkade från lösning, har i regel ett aktivt lager med två komponenter:

ett elektronaccepterande material och ett elektrondonerande material. En ofta använd elektron- donator är fullerenderivatet PCBM. I detta arbete undersöks hur ljus och värme påverkar spincoatade skikt av PCBM.

Olika PCBM-skikt tillverkades genom spincoating från olika lösningar och med både klo- roform och klorbensen som lösningsmedel. Dessa skikt bedömdes med blotta ögat och med optisk mikroskopi. Flertalet optiska bilder analyserades och inspekterades, Ett bedömningssys- tem skapades för att klassificera och ordna dessa prover efter släthet — baserat på de vanligaste defekterna — för att hitta en provpreparering som ger släta och homogena skikt som täcker hela substratet. Det bästa skiktet erhölls från lösning i klorbensen med kortvarig omröring. Långvarig omröring och/eller värmning hade negativ effekt på det resulterande skiktets släthet. Därför in- nefattar provprepareringen kortvarigt omrörd PCBM-lösning i klorbensen utan vidare omröring eller värmning.

Exponering av tunna PCBM-skikt för ljus uppvisade ingen större effekt vid undersökning i AFM. Däremot ledde exponeringen till en övergång för PCBM från ett lösligt till ett icke-lösligt tillstånd vilket observerades när PCBM-skikten utsattes för toluen eller en blandning av kloro- form och aceton. UV/Vis-spektroskopi visar en tydlig förändring i det synliga området vilket bekräftar en kemisk förändring av molekylerna i skiktet.

Den jämna och homogena ytan genomgick en väsentlig förändring när den utsattes för värme (200^◦C) i såväl luft som i vakuum. När det värmda skiktet hade tillverkats från lösning som inte värmts visade AFM-bilder av ytan slumpmässigt orienterade kristalliter av storlek under en mikrometer. Skikt tillverkade från värmd lösning däremot visade kristalliter som var ord- nade. Vid undersökning med optisk mikroskopi syntes också stjärnformade kristaller av åtskil- liga mikrometer i diameter, vilka inte syntes i skikt spincoatade från lösning som inte värmts.

(10)

List of Figures

2.1 Schematic representation of the mechanism of the photon-to-electron conversion

process . . . . 5

2.2 Annual number of articles in scientific journals on PCBM and PCBM and solar cell 5 2.3 Synthesis of PCBM . . . . 6

2.4 The different steps of the spin-coating process . . . . 7

2.5 Air Mass coefficients, simulation of certain positions of the sun . . . . 8

2.6 Diagram comparing the main components of a STM and an AFM . . . . 9

2.10 The different dynamic AFM modes . . . . 10

2.7 Schematic representation of measuring a sample using an AFM and possible scanning directions . . . . 11

2.8 Schematic representation of the AFM head . . . . 11

2.9 Schematic representation of a scanner tube of an AFM . . . . 11

2.11 Schematic representation of a reflected light microscope . . . . 13

4.1 General spin-coating diagram . . . . 19

4.2 An example of a spin coated PCBM film, comparing the size to a two euro coin and a Swedish one crown coin . . . . 19

5.1 Three dimensional representation of the height image of a scarred PCBM thin film, using TM-AFM . . . . 23

6.1 Schematic representation of a TM-AFM chip . . . . 25

8.1 Relation between the spin speed and the film thickness of pristine PCBM thin films 31 8.2 Relation between the number of particles of a specific size and the speed of spin- coating . . . . 31

8.3 The rating of smoothness of pristine PCBM thin films in function of the speed at which they were spin-coated . . . . 31

8.4 Photograph of a PCBM film, prepared from a filtered solution . . . . 32

9.1 TM-AFM images of pristine PCBM thin films, before and after exposure to light . . 33

9.2 Photographs of pristine PCBM thin films, exposed and unexposed to light . . . . . 34

9.3 Graph comparing the UV/Vis-spectrum of pristine PCBM thin films: exposed to light for 15 h, not exposed to light, exposed to light for 15 h and washed with toluene, not exposed to light and washed with toluene . . . . 35

10.1 AFM images of PCBM films exposed and unexposed to heat . . . . 37

B.1 Optical microscope image of a used TM-AFM cantilever . . . . 52

B.2 Image of a TM-AFM cantilever above a scarred PCBM thin film ready for film thickness analysis . . . . 53

B.3 Optical microscope image of a PCBM thin film from an unheated solution . . . . . 53

B.4 Optical microscope image of a PCBM thin film from an unheated solution, annealed in air . . . . 54

B.5 Optical microscope image of a PCBM thin film from an unheated solution, annealed in vacuum . . . . 54

B.6 Optical microscope image of a PCBM thin film from a heated solution . . . . 55

B.7 Optical microscope image of a PCBM thin film from a heated solution, annealed in air . . . . 55

(13)

C.1 Four TM-AFM images with different scan sizes, of a crystal formed after exposure of the PCBM film to heat . . . . 57 C.2 Computer interface of the computer controlling the used AFM . . . . 58 C.3 Image showing the three steps used to determine the thickness of thin films . . . . 59 D.1 Tuning graph of a silicon cantilever . . . . 60 D.2 UV/Vis spectrum of a glass substrate . . . . 60 D.3 UV/Vis-spectra of pristine PCBM, before and after illumination . . . . 61 D.4 Solar emission spectrum of: the sun and of the standardised classes AM 1,5 G and

AM 1,5 . . . . 62

List of Tables

2.1 Solubility of PCBM in several organic solvents . . . . 7 4.1 Standard spin-coating recipe . . . . 19 A.1 Classes used for the determination of the partial scores for the assessment of pris-

tine PCBM thin films . . . . 45 A.2 Sample list of the films prepared to determine the most optimal pristine PCBM

thin film . . . . 46 A.3 Power densities of published standards of solar conditions . . . . 51

(14)

List of abbreviations

AFM atomic force microscope/microscopy

CB chlorobenzene

CF chloroform

CM-AFM contact mode atomic force microscope/microscopy D.I. deionised water

FM-AFM force modulation mode atomic force HOMO highest occupied molecular orbital IPA isopropyl alcohol, isopropanol LUMO lowest unoccupied molecular orbital NC-AFM non-contact mode atomic force NOC number of different colours NOP number of particles

ODCB orthodichlorobenzene OPV organic photovoltaic

PCBM [6, 6]-phenyl-C61-butyric acid methyl ester PCE power conversion efficiency

< rating of film smoothness, see equation 5.1 on page 22 rpm rotations per minute

ß spin speed (unit: rpm)

SPM scanning probe microscope/microscopy STM scanning tunnelling microscope/microscopy TM-AFM tapping mode atomic force microscope/microscopy

(15)

Introduction

(16)

1 | Introduction and outline

Without doubt the early 21^stcentury will go into the history books as, the period man realised its traditional resources — as used extensively since the industrial revolution — are not limitless available. Combined with a climate change threatening to dramatically alter the world, the search for alternative energy sources is ever more pressing.

The most abundant source of energy known to man is of course the sun. And although solar cells and other devices to convert the sun’s light into usable forms of energy are not new, they are still far from perfect.

One of the points of improvement is the efficiency with which they convert the energy i.e.

their PCE, power conversion efficiency. In addition, the fabrication of solar cells, in a durable and usable way should also be kept in mind. A device can prove to have a very high PCE, but at the same time the production can rely on resources that are scarce, hard to implement because of manufacturing limitations, or the device can have a short lifetime. There are multiple ways to try and find a solution to these problems, and continued research into converting solar energy is needed.

The traditional photovoltaics — devices that convert sunlight into electrical power — are made from semiconductors, often silicon. At the moment these traditional devices achieve the best PCE. And although this specific area of research is still far from fully explored, conversion efficiency did not improve so much any more over the recent years.

Currently many research groups are looking into the prospects of organic photovoltaics (OPVs).

OPVs promise to be more cost effective and have different mechanical properties such as flexi- bility. OPVs could ultimately be produced in a continuous process, making them far more inter- esting for large area production than the traditional solar cells.[4, 12, 33]

The work presented in this thesis explores some of the properties of one component often used in solution processed organic photo voltaics:[6, 6]-phenyl-C₆₁-butyric acid methyl ester.

Light has an influence on many chemical processes. And because this study has the use of PCBM in OPVs in mind, it is even more important to know what influence light has — more particularly sunlight — on pristine PCBM. In an article by Eklund et al., a change from a toluene- soluble to an insoluble-state is described for C₆₀, after exposure to light. This was also examined for PCBM.

Heating of samples (annealing) is a technique commonly used in the preparation of polymer based OPVs, to improve the properties. Assessing whether a material that is blended with the polymer — in this case PCBM — has changed properties as well, it is important to know how the pristine material behaves. For this reason, the effect heat treatment has on pristine PCBM is examined in this work. The formation of crystals in films of PCBM-polymer blends is described in literature e.g in PCBM:P3HT blends, in the article by Swinnen et al. This will be assessed as well.

(17)

In this first part, ‘Introduction’, some contextual information and theoretical background for the used techniques is given, as well as a statement of the aims of this thesis.

In the second part, ‘Methodology’, the used materials and methods are described for the preparation of samples and for the three main experiments: ‘Determination of an optimal pristine PCBM thin film recipe’, ‘Effects of light exposure’ and ‘Effects of heat exposure’.

The results for the above mentioned three main experiments, are presented and discussed in the third part, ‘Results and discussion’.

The final part, ‘General discussion’, combines all findings and discusses further research possi- bilities.

‘Appendices’ can be found, starting at page 44, assembling additional tables, images and graphs.

(18)

2 | Background

2.1 Organic photovoltaics

In its simplest form an OPV device consists of a donor material and an acceptor material, both connected to an electrode. As the name OPV implies, the two materials are made up of organic molecules. In polymer based solar cells, the electron donating material is a polymer and the electron accepting material a fullerene derivative.

To create an OPV, multiple methods are possible, for example: solution processed (e.g. spin- coated or spray-coated) or vacuum depositioned (e.g. sputtering, thermal evaporation). This work only investigates solution processed, spin-coated thin films. Figure 2.1 on page 5, explains schematically the mechanism of conversion of light into an electrical energy in five steps for a simple OPV device (a bulk heterojunction OPV).[1, 2, 11, 19, 33]

(I) Exciton formation

The energy of the absorbed light (hυ) allows an electron (e⁻, represented by a white circle) in the donor material to get excited from the highest occupied molecular orbital (HOMO) to the lowest unoccupied molecular orbital (LUMO) and form an exciton, i.e. an electron-hole pair coupled by coulombic forces (e⁻and h⁺), that is able to move along the polymer chain.

(II) Diffusion to the D-A interface

The formed exciton diffuses to the donor-acceptor (D-A) interface. Important is to note that the maximal diffusion length in an organic material is typically in the range of 10 nm.

(III) CT-exciton formation

At the D-A interface the exciton undergoes a charge-transfer (CT) and forms a CT-exciton. The electron transfers from the LUMO of the donor material to the LUMO of the acceptor material while the electron hole stays in the donor material, but are still held together coulombically.

(IV) Dissociation of the CT-exciton

The CT-exciton dissociates and the electron and electron hole are transported to their respective electrodes.

(V) Flow of charges

The potential difference between anode and cathode results in an external electric current when connected via a conductor.

(19)

LUMO

HOMO

LUMO

HOMO

E (eV)

anode HTL donor acceptor cathode

e^-

h⁺ h⁺

e^- e^-

e^-

(I)

(II)

(III)

(IV)

(IV) (IV)

(V) e^- (II)

hυ

(I)

Figure 2.1: Schematic representation of the mechanism of the photon-to-electron conversion process in a bulk heterojunction OPV. (The dimensions are not proportionate.) Adapted from [33]

2000 2002 2004 2006 2008 2010 2012

0 20 40 60 80 100 120 140 160 180 200

date

numberofarticles PCBM

PCBM and solar cell

Figure 2.2:Annual number of articles in scientific journals on PCBM and PCBM and solar cell. (Data acquired on 2013-04-11, from searches via the scientific database Science Direct (http://www.sciencedirect.com) with boolean search queries: TITLE-ABSTR-KEY(PCBM) AND LIMIT-TO(contenttype, "1,2","Journal") and TITLE-ABSTR-KEY(PCBM) and TITLE-ABSTR-KEY("solar cell") AND LIMIT-TO(contenttype, "1,2","Journal").)

(20)

2.2 PCBM

The studied molecule [6, 6]-phenyl-C61-butyric acid methyl ester or commonly abbreviated as PCBM or PC60BM, is a fullerene derivative of buckminsterfullerene. The molecule is seen as a good candidate to be used as an acceptor material in OPVs. Research into PCBM has risen significantly in the last years especially its use in solar cells as shown in figure 2.2.[3]

2.2.1 Fullerenes

Fullerenes are a family of molecules consisting only of carbon (Cn) of which 12 pentagons and m hexagons (other than one), conforming to the relation m = (n−20) ·2⁻¹. The best known fullerene is C₆₀and is called buckminsterfullerene, named after the American architect Buckmin- ster Fuller. For this fullerene its 60 carbon atoms are arranged in 12 pentagons and 20 hexagons, according to the above-mentioned formula.

Compared to other structures composed solely of carbon e.g. diamond and graphite, C₆₀ displays a considerable chemical reactivity. This is caused by the strain release that is associated with the bent geometry of the C−C double bonds. Two advantages of using a fullerene derivative are that they tend to aggregate less and be more soluble in commonly used solvents than their pristine counterparts. Both are very desirable properties with electronic purposes in mind.

Another property of C₆₀is its ability to reversibly accept up to six electrons. This is one of the main reasons to develop C₆₀-derivatives, especially PCBM, as an electron acceptor, to be used in photovoltaic devices.[5, 9, 26]

2.2.2 PCBM synthesis

PCBM was first synthesised by the research group of J.C. Hummelen as described in their research article in 1994 (see [14]). This article describes the synthesis route as shown in figure 2.3.

First a stable diazo compound is formed in situ starting from 4-benzoylbutyric acid which react with C₆₀ in orthodichlorobenzene (ODCB) to form[5, 6]-phenyl-C61-butyric-acid-methyl- ester. This product is heated for a long period to fully convert to[6, 6]-phenyl-C₆₁-butyric-acid- methyl-ester, PCBM.

C60

ODCB

heating

[6,6]-Phenyl C61 butyric acid methyl ester (PCBM)

[5,6]-Phenyl C61 butyric acid methyl ester

Figure 2.3:Synthesis of PCBM as described by J.C. Hummelen et al.[14, 32]

(21)

2.2.3 PCBM properties

Because PCBM is still a relatively new molecule, most properties have not yet been thoroughly studied, and are therefore not readily available. As a solid, PCBM occurs as black to light brown crystals with a molecular mass of 910,9 Da. It is relatively good soluble in organic solvents, as shown in table 2.1, but its solubility in water is negligible.[28]

Table 2.1:Solubility of PCBM (s_PCBM) in several organic solvents[24]

solvent sPCBM

(mg ml⁻¹)

toluene 9

xylene 15

mesitylene 29

chloroform 26

chlorobenzene 35 2-chlorotoluene 47

2.3 Spin-coating

Spin-coating is a technique, used to produce uniform films from solutions, with a thickness in the order of micro- and nanometres. The process is traditionally divided into four stages: deposition, spin-up, stable fluid outflow and evaporation. These stages of the spin-coating process are visualised in figure 2.4

In the first stage, deposition (figure 2.4a), solution is applied to the substrate. This stage ends when the deposition is terminated and the desired spin speed is reached.

The second stage, spin-up (figure 2.4b), is characterized by aggressive fluid expulsion from the substrate surface by the rotational motion. For this reason the applied solution during the deposition did not have to cover the entire substrate, any superfluous solution will be removed.

During the third stage, stable fluid outflow (figure 2.4c), the substrate is spinning at a constant rate and fluid viscous forces dominate fluid thinning behaviour.

In the final stage, evaporation (figure 2.4d), the centrifugal out flow ends and further thinning of the film is due to the evaporation of solvent, resulting in a thin film on the substrate.[27]

(a) (b) (c) (d)

Figure 2.4:The different steps of the spin-coating process: (a) deposition, (b) spin-up, (c) stable fluid outflow and (d) evaporation. Adapted from [27]

(22)

2.4 Solar simulators

To create reproducible and comparable data, when studying the effects of sunlight, it is essential to use a light source with a known, constant spectrum and intensity. For this purpose, solar simulators have been developed, that emit light conform a specific Air Mass coefficient (AM).

Several standard classes have been introduced, based on their power density, simulating the light from the sun which is dependant from the Zenith angle — the angle between the normal of the earth and the path between the sun and the point of measuring — as displayed in figure 2.5.

A table showing the different power densities for some standard classes is given in figure A.3 on page 51. The spectrum of sunlight and of class AM 1,5 light, is shown in figure D.4 on page 62.[23]

Figure 2.5:Air Mass coefficients, simulation of certain positions of the sun[23]

2.5 Scanning probe microscopy

The main technique that has been used for the analysis of the PCBM thin films was atomic force microscopy (AFM), more specifically tapping mode atomic force microscopy (TM-AFM). AFM together with its predecessor; scanning tunnelling microscopy (STM), are called scanning probe microscopy techniques (SPM), techniques that use a probe (tip) to scan a surface. The resulting data is used to create a three-dimensional map of that surface, with a nanometre resolution. The following sections will cover the different SPM-techniques with focus on the used technique, TM-AFM.

The main parts of a SPM are: its probe, a laser, a piezoceramic scanner — to which either the tip or the sample is attached — and a detector. (The different parts are explained more thoroughly in section 2.5.5 on page 12). The detector signal is used in a feedback loop to adjust the cantilever- sample distance, by changing the voltage applied to the scanner. A schematic representation is shown in figure 2.6 on page 9.[20]

2.5.1 Scanning tunnelling microscopy

Scanning tunnelling microscopy was developed by researchers at IBM’s research lab in Zurich in 1981. Its invention was revolutionising, it gave way to image individual atoms of a conducting surface. The technique is based on the quantum-tunnelling of electrons between a tip and sample.

When an bias voltage is applied between the tip and sample, there are three regimes depending on the distance between both, in order of decreasing separation: tunnelling, electronic contact and mechanical contact. Therefore the possible samples are limited to conductors or semi-conductors.[20, 21]

(23)

Figure 2.6:Diagram comparing the main components of a STM and an AFM[20]

2.5.2 AFM, operating principle and modes

Even though STM has proved itself to be an excellent technique at imaging at a nanometre scale, it is limited to image conducting surfaces. Also to image a surface with a high resolution (up to atomic scale), working under vacuum (or even ultra-high vacuum) is necessary.

To image non-conducting surfaces, atomic force microscopy was developed and introduced in 1985/86. Although STM and AFM have many components in common (see figure 2.6), the principles on which they are based are far from identical. Whereas STM is based on interactions with the valence electrons of the sample, AFM is based on the interaction with all electrons, in- cluding the core electrons.[20, 21]

A probe or tip is mounted on a flexible cantilever (at the tip of a so called chip, see figure 6.1 on page 25) which is placed in a cantilever holder. The cantilever holder also houses a small piezoelectric element. A laser-beam is precisely focused on the tip of the cantilever, and the reflected beam directed onto the middle of a split photo detector by means of a mirror. If the cantilever undergoes a buckling or a torsion (see figure 2.7 on page 11), the laser-beam’s spot on the split photo detector will change — some of its quadrants will be illuminated more than others

— thus register the cantilever’s deflection.

For cantilevers with a length of between 100 µm and 200 µm, the reflected laser-beam will register the changes detected by the tip by a factor of 750 – 1500. This makes it possible to use an AFM to perform measurements in the sub-ångström range.

In contrast to STM, the sample holder of a typical AFM will be mounted on the scanner. To create an image, the sample is moved with respect to the tip, using a piezoceramic scanner. A typical AFM scanner, as shown in figure 2.9 on page 11, will consist of multiple segments. These segments, piezoelectric elements, will expand or contract when a current is applied and are con- trolled individually. Controlling the current applied to segments X, ¯X and Y, ¯Y will move the

(24)

sample laterally. By changing the current applied to segment Z, the scanner tube will contract or expand, changing the sample’s height, move the sample vertically.[10, 20, 30]

There are several different ways in which an AFM tip can scan a sample’s surface, both static and dynamic (visualised schematic in figure 2.10 on page 10). These different modes are listed below, and will be discussed in the following sections. [20]

• static

– contact mode AFM (CM-AFM)

• dynamic

– non-contact mode AFM (NC-AFM) – tapping mode AFM (TM-AFM) – force modulated AFM (FM-AFM)

2.5.3 Contact mode AFM

When working in contact mode, the tip is kept in permanent contact with the surface. This is realised by keeping the cantilever deflection constant by altering the z-position of the sample via the scanner. The deflection results in a force exercised by the tip on the sample which lies in the region of 0 – 100 nN.

The AFM results can be presented either in the form of height or deflection (force) images.

The latter is based on the experienced lateral force during scanning, which causes buckling or torsion of the cantilever.

This mode has a real disadvantage, the permanent contact of the tip can damage a sample and sample material can be picked up by the tip. This makes it more difficult to image soft samples, without altering the surface structure.

Partly to overcome these difficulties, several dynamic scanning modes have been developed to: use the attractive forces of a sample’s surface, NC-AFM; avoid damage to sample surfaces, TM-AFM and examine surface mechanical properties, FM-AFM.[20]

(a) (b) (c)

Figure 2.10:The different dynamic AFM modes: (a) non-contact mode AFM (NC-AFM), (b) tapping mode AFM (TM-AFM) and (c) force modulation mode (FM-AFM). (Amplitudes are disproportionate.)[20]

2.5.4 Non-contact mode AFM

In non-contact mode the cantilever is vertically oscillated by the small piezoelectric element in the cantilever holder at or slightly above its resonance frequency with a small amplitude of typically less than 10 nm. The constant changing deflection of the cantilever is magnified by the laser-beam reflection onto the photo-detector, generating an AC signal.

(25)

(a) (b)

Figure 2.7: (a) Schematic representation of measuring a sample using an AFM. (b) Possible scanning directions[20]

Figure 2.8: Schematic representation of the AFM head: 1 laser, 2 mirror, 3 cantilever, 4 tilt mirror, 5 photodetector[30]

Figure 2.9:Schematic representation of a scanner tube of an AFM[30]

(26)

During a measurement a feedback signal is generated from this AC signal to keep either the amplitude or frequency of oscillation constant.

This feedback signal also ensures that the tip never comes into contact with the surface itself, but interacts with the long range forces of the sample’s surface ( approx. 1 nm – 10 nm).

Unfortunately the advantage of not exerting a force on the sample’s surface, is counteracted by its limited field of application. Usually NC-AFM only works on extremely hydrophobic samples with a very thin absorbed fluid layer. A ticker layer would result in the tip getting trapped, causing damage to tip and/or sample.

Also the scan speed and lateral resolution are lower compared to other dynamic mode AFM techniques.[20, 30]

2.5.5 Tapping mode AFM

In tapping mode AFM, the tip is oscillated at its resonance frequency with a high amplitude (10 – 100 nm). The cantilever needs to be much stiffer than a CM-AFM cantilever in order to achieve this. An optical microscope image of an TM-AFM cantilever is shown in figure B.1 on page 52.

The tip just touches the surface of the sample intermittently. To maintain this constant light tapping the amplitude is use as feedback. The moment the tip touches the surface, it limits the amplitude of oscillation. This lowers the observed amplitude, which is used to adjust the tip- sample distance via the scanner. This tip-sample distance is used to generate a height image of the sample’s surface.

Simultaneously the phase shift — the difference betweenthe phase of the set oscillation and the phase of the registered oscillation — is mapped as well. This phase image can be used to examine specific viscoelastic properties of the sample surface, e.g. stiffness.

This method minimizes lateral force, and the short tip-sample contact prevents inelastic mod- ification of the sample’s surface, preventing damage to both tip and sample.[20, 30]

2.5.6 Force-modulation mode AFM

In force-modulation mode, the cantilever is oscillated with a much smaller amplitude and frequency. This way there is a constant contact between tip and sample. When analysing the changes in amplitude and phase of the cantilever it is possible to determine several physical properties of the sample’s surface, e.g. spatial variation in elasticity and viscoelasticity of a sample surface.[20]

2.6 Optical microscopy

Optical microscopy, the study of the micrometre sized samples, using light from the visible spectrum to create an enlarged image of the studied sample via a series of lenses.

In the case of a transparent sample, it is possible to use a traditional transmission set-up, whereby: light source, sample and lenses are all aligned. However, when a sample is opaque, an incident light set-up might be necessary, e.g. reflected light microscopy.

Reflected light microscopy uses a light source that is mounted right-angled to the microscope (figure 2.11 on page 13). The light beam is directed onto the sample using a half-mirror (reflector).

The light that is reflected off of the surface, is send back into the microscope, through the half- mirror and further enlarged via a series of lenses. The enlarged image can than be observed either directly by means of an eyepiece and/or by use of a (digital) camera.[6, 22, 31]

(27)

Figure 2.11:Schematic representation of a reflected light microscope[31]

2.7 UV/Vis-spectroscopy

Ultraviolet-visual light (UV/Vis) spectroscopy, studies the absorption characteristics of a sample, using light with a wavelength of approx. 100 nm – 400 nm, ultraviolet and 400 nm – 750 nm, visible. Light absorbed in this spectral region causes electronic and vibrational excitations.

As a technique it is routinely used in quantitative analytical chemistry. Using the well known

‘Lambert-Beer relation (equation 2.1: A, absorbance; I₀, intensity of the incident light; I, intensity of the transmitted light; ε, extinction coefficient; c, concentration and l, path length) it is fairly simple to determine the concentration of a solution.

A=logI0

I =ε·c·l (2.1)

Although this equation, in this form, can only be used for liquids; UV/Vis-spectroscopy is not limited to liquids. The absorption characteristics of gases and solids, if sufficiently transparent, can also be examined.

An UV/Vis-spectrum on its own is not enough to identify a compound. However an absorption spectrum is specific for a compound, relating to its molecular structure, and therefore usable to gain qualitative information.[8, 13, 15]

(28)

3 | Aims

This thesis attempts to determine whether light or heat have an influence on the chemical properties of[6, 6]-phenyl-C₆₁-butyric acid methyl ester, using spin-coated thin films of pristine PCBM.

Determining the optimal condition to prepare thin films is the first major target; choosing the best parameters and solvent to prepare the optimal solution for spin-coating thin films.

Next these films were exposed to either; light, in the form of standardised light using a solar simulator or heat, using an oven. These treated films were compared with unexposed films, in order to determine whether such treatment has an influence on PCBM using atomic force microscopy, optical microscopy and UV/Vis-spectroscopy, to detect changes to the surface of a PCBM thin film and its chemical properties (solubility and absorption characteristics).

(29)

Methodology

(30)

4 | Sample preparation

In this chapter, the materials and methods used in the formation of PCBM thin films — both for the study into the optimal preparation of films and exposure test of PCBM thin films — are described. During every manipulation involving PCBM, in any form, exposure to light was kept at a minimum. This by working in a darkened room and covering any container (vials, sample boxes, etc.) in aluminium foil.

4.1 Preparation of PCBM solutions

Solutions of [6, 6]-phenyl-C61-butyric acid methyl ester were made, starting from PCBM pow- der (PC60BM, 99,5 % by Solenne b.v.). The required amount was weighed (Precisa XR205A by Precisa) directly into a brown glass vial after blowing it out with nitrogen to remove possible contaminations, e.g. dust particles.

The required volume of solvent, either chlorobenzene or chloroform, was added to the vial using a 1000 µl micropipet (Finnpipette Digital Pipette by Thermo Scientific).

The choice for chloroform and chlorobenzene as a solvent, was based on two factors. First of all the solubility of PCBM is quite good in both solvents, see table 2.1 on page 7. Secondly these solvents are commonly used for solution processed OPVs, see for example the article of Björström et al. or Casalegno et al.[2, 3]

Hereafter the solution was homogenised by one of the following manipulations:

a agitated by vortexing (MS2 Minishaker by IKA)

b first agitated by vortexing, than sonicated for 60 min (Ultrasonic bath, 5510 by Branson) c first agitated by vortexing, than filtered by one of the following filter discs together with a

plastic syringe (BD Plastipak 2 ml by Becton Dickinson S.A.):

c.i 25 mm PTFE filter with pores of 1 µm by PALL c.ii 13 mm PTFE filter with pores of 0,2 µm by VWR d sonicating for 60 min after vortexing and filtering

e vortexing, heating for 60 min at 50^◦C in a hot water bath, than sonicated for 60 min f vortexing, heating for 60 min at 50^◦C in a hot water bath, filtered and sonicated for 60 min g stirred for 30 min using a magnetic stirrer

h stirred for 30 min using a magnetic stirrer and filtered

i stirred for 30 min using a magnetic stirrer, filtered and sonicated

(31)

j stirred for 16 h at 50^◦C using a magnetic stirrer and a hot water bath k nitrogen bubbling for 30 min

l nitrogen bubbling for 30 min at 50^◦C using a hot water bath

The vials of the homogenised solutions were labelled, wrapped in paraffin film and aluminium foil.

4.2 Substrate treatment

The substrates on which the PCBM films were spin coated were pieces of silicon wafer (SI-MAT N-100, diameter: 3±0,02 inches, thickness: 356 – 406 µm) or for a few samples glass in the form of microscopy cover glass (L4094-3 coverglasses No.2, 20 mm², thickness: 196 – 250 µm by Van Loenen Instruments). The substrates, were cut by hand to the desired size, using a diamond tipped pen. In most cases a size of roughly 1 cm²was used, as this is close to the maximal size, that fitted the AFM.

The used substrate treatments were:

• none, used directly after cutting to size

• RCA-cleaning (chapter 4.2.1)

• toluene-cleaning, dried with a wipe (chapter 4.2.2)

• toluene-cleaning, dried with nitrogen (chapter 4.2.2)

• isopropanol-cleaning (chapter 4.2.3)

• UV/ozone-treatment (chapter 4.2.4)

Although the treatments are often called ‘a cleaning’, the main reason is not to simply remove any contamination, but to alter the surface properties, to alter the affinity for polar or apolar solutions. Doing so will influence the ‘wetting’ of the substrate, thus having an impact on how well a droplet of a solution will cover the substrate.

4.2.1 RCA-cleaning

For the bonding of many solutions an hydrophilic wafer surface is necessary. A commonly used treatment is RCA-cleaning (named after the facilities it was developed at; RCA Laboratories, Princeton, and RCA Solid-State Division, Somerville, NJ Laboratories[17]). It was developed to remove both organic and inorganic contaminants. And consists typically of two or three steps depending on the desired surface properties. The technique used for the preformed experiments consisted of the following two steps.

The first step, often called SC-1 (standard cleaning 1), consists of putting the wafer or wafer pieces into a solution of H₂O : H₂O₂(30 %) : NH₄OH(25 %) in a 5:1:1 ratio. This solution is heated for 10 – 15 min at a temperature of 70 – 80^◦C on a hot plate. After this period, the wafer pieces are thoroughly rinsed with deionised water (D.I.).

In the second step, often called SC-2 (standard cleaning 2), the rinsed wafer pieces are put into a next chemical bath. The solution consists of H₂O : H₂O₂(30 %) : HCl(25 %) in a 5:1:1 ratio.

This solution is likewise heated for 10 – 15 min at a temperature of 70 – 80^◦C on a hot plate and

(32)

After this step the wafer pieces were stored in D.I. till they were used. This to prevent contamination and preserve the realised hydrophilicity. The wafer pieces were used within 24 hours of cleaning to preserve the realised properties.[18]

The wafer pieces were dried before usage, using nitrogen gas.

4.2.2 Toluene-cleaning

The toluene-cleaning consisted of placing the cut wafer pieces into a receptacle with toluene in an ultrasonic bath. The solution was sonicated for 60 min and the cleaned wafer pieces were kept under toluene until they were used. The wafer pieces were dried, using either a laboratory wipe (Precision wipes by Kimtech Science) or nitrogen gas.

4.2.3 Isopropanol-cleaning

Isopropanol or isopropyl alcohol (IPA), was used in the same way as toluene (chapter 4.2.2) to clean the wafer pieces.

The cleaned wafer pieces were kept in the IPA-solution until they were used. To dry the wafer pieces, nitrogen gas was used.

4.2.4 UV/ozone-treatment

UV/ozone cleaning is a technique commonly used in nanotechnology to treat substrates, cantilevers, etc. It is an ideal technique to remove organic contaminations, without the use of a solvent (which could leave traces). An UV-lamp, usually a mercury discharge lamp, is used to generate ultra violet (UV) light. This UV-light generates atomic oxygen and ozone (and absorbed by organic compounds themselves), braking down organic contaminations.[16]

Wafer pieces were placed in the UV/ozone machine (UV probe and surface decontaminator, PSD-UV by Novascan) for 30 min.

4.3 Spin coating of PCBM films

The technique used to make PCBM thin films from the prepared solutions (chapter 4.1), on the treated substrates (chapter 4.2), was spin-coating. Like the solution preparation, spin-coating films was always preformed with as little ambient light as possible, to avoid chemical changes during the preparation process.

After a precautionary clean with nitrogen gas a substrate was placed on the spin-coater (Model P- 6708D, Desk-Top Precission Spin-Coating system by Specialty Coating Systems) using tweezers. The solution was applied to the substrate, normally using a 100 µl micropipet or a plastic or glass syringe. The volume needed to be enough to cover the entire surface. Slightly more — normally 80 µl for a 1 cm²sample — was applied to ensure complete coverage; any superfluous solution was removed in the first phase of spin-coating. Hereafter the spin-coating cycle was initiated.

(33)

R1 P1 R2 P2 R3 P3 R4 t (s)

ß(rpm)

Figure 4.1:General spin coating diagram; R1, R2, R3 and R4 are ramps and P1, P2 and P3 are plateaus where the spinspeed (ß) is kept constant

Table 4.1:Standard spin-coating recipe

t ß

(s) (rpm)

R1 1 100

P1 1 100

R2 1 100

P2 1 100

R3 3 2000

P3 90 2000

R4 5 0

The spin-coating recipe was programmed in advanced according to the required spin speed.

On page 19, the general spin-coating diagram is shown in figure 4.1 and the recipe used for the preparation of thin film (unless stated differently) is shown in table 4.1. An example of a spin- coated film is shown in figure 4.2.

After spin-coating the samples were placed in a labelled plastic sample box and wrapped in aluminium foil.

(a) (b)

Figure 4.2: An example of a spin coated PCBM film, comparing the size to: (4.2a) a two euro coin and (4.2b) a Swedish one crown coin. The film was spin-coated from a PCBM in chlorobenzene solution (c = 15,0 mg ml⁻¹).

4.4 Treatment of the samples

For the experiments on the effects of light or heat exposure, spin coated films were prepared as described in the previous sections and further treated as described in the following sections.

4.4.1 Exposing to light

To test the effects of light on PCBM thin films, a solar simulator with an AM 1,5 G spectrum, with power density of 100 mW cm⁻² (Oriel Sol2A, 94022A, Class ABA Solar Simulator by Newport) was used. The samples were placed in the machine without any packaging, an open sample box, and illuminated for the desired period.

(34)

To detect alterations to a film, it was either analysed before illumination or an identical sample was kept unexposed and analysed at the same time as the exposed sample.

4.4.2 Exposing to heat

For the exposure to heat, both in vacuum and air, a laboratory oven was used (APT.Line VD, vacuum drying oven with RD3 microprocessor program controller by Binder). The oven was set to the desired temperature and preheated. The samples were placed on a piece of aluminium foil and put in the oven for the desired time.

To heat the samples in (low) vacuum, an external vacuum pump was used and left on running for the entire duration of the treatment. In these cases, in order to open the oven, nitrogen gas was used to bring the oven chamber to atmospheric pressure.

(35)

tine PCBM thin film recipe

5.1 Testing and analysing different recipes

In order to find the most optimal recipe for pristine PCBM thin films, it was important to de- fine what a homogeneous and defect-free film should look like: it should completely cover the substrate; it would need to be as uniformly flat as possible, not showing any secondary structures (e.g. stripes or comet-shaped structures) resulting from the spin coating and contain as few particles or impurities as possible.

Different recipes where tested, changing solution preparation — both with chlorobenzene and chloroform as a solvent — and substrate treatment. These samples are listed in table A.2 on page 46.

5.1.1 Visual assessment of the samples

The first step in assessing the samples was with the naked eye. The two points judged were:

substrate coverage and the presence of secondary structures and large impurities. The substrate coverage, resulting from the wetting properties of the substrate–solution-pair, could be: complete, nearly complete or incomplete. The presence of secondary structures — e.g. comet shaped structures, resulting from the presence of a particle on the film during spin coating — was divided into three classes: many, some or few.

5.1.2 Imaging and analysing samples using optical microscopy

In the second step to assess the samples, an optical microscope (LV-IM with LV-UEPI, reflected light microscope by Nikon) equipped with a digital camera (DS-Fi1 and controller DS-U2 and NIS-Elements F3.0 for windows microscopy software, by Nikon) was used to imaging the samples. An image of the center of each sample was taken at a magnification of 20 times.

The first assessment involved counting the number of colours — resulting from height differ- ences — in each image. Secondly the particles present on these captured images, were counted using the program ImageJ (ImageJ 1.46r 64-bit, using Java 1.6.0_24 64-bit on Ubuntu 12.10). The resulting outputs, data series, showed the determined particles and their surface area.

Using a spreadsheet program (LibreOffice Calc, version 3.6.2.2 on Ubuntu 12.10) the data were sorted into four categories according to their surface area. The chosen surface areas, were these corresponding to the area of a circular particle with a radius of:

• tiny particles: less than 1 µm (π µm²)

Thin films of [6,6]-phenyl-C 61