Skip to main content

Lund University Publications

LUND UNIVERSITY LIBRARIES

Computational Predictions of Conjugated Polymer Properties for Photovoltaic Applications

Hedström, Svante LU (2015)
Abstract
Organic solar cells employing fullerenes blended with conjugated polymers as the main light-absorbing material have achieved power conversion efficiencies exceeding 10%. They hold promise as an alternative energy source with many advantages in terms of long-term sustainability and reduced greenhouse gas emissions. Detailed information on the electronic and geometric structure of the molecules involved is generally not accessible through experimental means, as the typically amorphous polymer films are not readily studied with e.g. X-ray crystallography. Computational chemistry, and in particular quantum chemistry as used for the research presented in this thesis, can however provide molecular level insight into the properties of these... (More)
Organic solar cells employing fullerenes blended with conjugated polymers as the main light-absorbing material have achieved power conversion efficiencies exceeding 10%. They hold promise as an alternative energy source with many advantages in terms of long-term sustainability and reduced greenhouse gas emissions. Detailed information on the electronic and geometric structure of the molecules involved is generally not accessible through experimental means, as the typically amorphous polymer films are not readily studied with e.g. X-ray crystallography. Computational chemistry, and in particular quantum chemistry as used for the research presented in this thesis, can however provide molecular level insight into the properties of these conjugated polymers. (Time-dependent) density functional theory calculations are here employed on various polymers, mainly of donor–acceptor (D–A) and D–A1–D–A2 types. Systematic studies demonstrate how the energy levels and optical properties relate to each other, as well as to the chemical composition of the polymers. In particular focus are the traits that are important for efficient solar cells: strong absorption, suitably narrow band gap, appropriate LUMO energy vs. the fullerene LUMO, and extended conjugation promoting high charge carrier mobilities. Several polymers with high-performance solar cells are studied, including TQ1 where a computationally revealed unique helical geometry is used to partially rationalize its 7.08% efficiency, and the D–A1–D–A2 polymer P3TQTIF whose two distinct acceptor units allow two strong low-energy electronic transitions, greatly enhancing its spectral coverage. Size-converged optical properties are obtained through a scheme based on extrapolations from oligomer calculations, and a detailed comparison to experiments has facilitated the development of an empirical correction for absorption energies and strengths. These corrections are subsequently used for a priori predictions of polymer absorption spectra with good agreement to experiments. Finally, a strategy is presented that includes the effect of temperature, in form of thermally populated conformations with reduced conjugation and weaker and blue-shifted absorption, yielding trends in excellent agreement with experimental optical properties. Calculations are in summary able to provide deeper insights into the fundamental properties of conjugated polymers, constituting a valuable tool for the ongoing development of materials for application in high-performance organic solar cells. (Less)
Abstract (Swedish)
Popular Abstract in English

The energy demands of human society are huge and will keep increasing for the foreseeable future. Fossil fuels have largely been able to meet those demands for the last centuries, but issues of sustainability and global warming have brought about a need for alternative energy sources. The emission of the sun supplies earth with an energy that is many thousand times larger than our consumption. It is therefore very promising as a major contributor to our electricity production, through the use of solar cells, also known as photovoltaics. The currently dominating solar cell technology is based on inorganic silicon, which converts light into electricity in a so called p–n junction. The fabrication... (More)
Popular Abstract in English

The energy demands of human society are huge and will keep increasing for the foreseeable future. Fossil fuels have largely been able to meet those demands for the last centuries, but issues of sustainability and global warming have brought about a need for alternative energy sources. The emission of the sun supplies earth with an energy that is many thousand times larger than our consumption. It is therefore very promising as a major contributor to our electricity production, through the use of solar cells, also known as photovoltaics. The currently dominating solar cell technology is based on inorganic silicon, which converts light into electricity in a so called p–n junction. The fabrication of silicon p–n junction solar cells is however complex and expensive. Another type of solar cells is made from organic, carbon-based materials and were invented over fifty years ago. Thanks to intense research, they have demonstrated sharply increasing efficiencies over the last two decades. They have many advantages over inorganic cells, such as being cheaper and easier to manufacture, very thin and mechanically flexible, and producible from renewable materials. Although their power conversion efficiency as of yet is roughly half of the silicon cells, their light-absorption properties are tunable, depending on what organic molecules are used, unlike the static absorption profile of silicon.

The research presented in this thesis concerns the properties of these light-absorbing molecules, in which the energy of the absorbed light is converted to electric power. The molecules of interest are special plastics or polymers: very large molecules made up of many identical repeating units connected along a chain. Although larger than most other synthetic molecules, they are still micro¬scopic, and many trillions of them make up a solar cell of typical lab-scale size: 1 cm2 in area and 100 nm thick. Despite consisting of so many polymer molecules, many of the solar cell properties can be deduced from studies of only single polymers, which is the main approach of the research herein. The most relevant properties are those that concern the absorption of light and the generation and transport of electric charge. The polymers of interest are conjugated, an intrinsic molecular property that promotes strong light-absorption and good conductivity, crucial for solar cell performance. Hundreds of new conjugated polymers are currently developed each year thanks to intense research efforts, each with different properties and varying efficiency in the solar cells where they are used.

Unlike the traditional view that chemistry is a very practical, hands-on discipline, the research in this thesis is purely computational. That means that advanced computer software is used to predict the properties of a molecule without ever coming in physical contact with it. This may sound a bit like magic, but consider that we in principle can calculate the time it takes for a ball to fall to the ground if we know the height, the ball’s weight, and the size of the gravitational force. In the same way, we know the forces that act on the atoms in a molecule, and can thus predict many of its properties. The effect of light absorption is also in principle known, and can be treated in a similar, yet somewhat different way. However, these forces that act on this microscopic level are quantum mechanical in nature, so the resulting equations are immensely more complex than the example of the falling ball, and that is why they are done with quantum chemistry software in a supercomputer rather than by hand.

Since the forces and resulting equations are so complicated, many approximations and simplifications are required, introducing uncertainties regarding the accuracy of the computational methods. This is an eternal issue for theoretical scientists, who have to rely on comparisons to experiments to validate that the calculations are reasonable representations of reality. The smaller the molecule, the less computationally demanding are the calculations. So for the large polymers studied herein, the most accurate quantum chemical methods are out of reach with today’s computers. Nevertheless, the results presented in this thesis show that methods based on density functional theory are approaching the capability of quantitatively predicting some of the most important properties of conjugated polymers for photovoltaic applications, including absorption spectra and energy levels. This is very valuable, as it can significantly decrease the workload associated with the design, synthesis, and characterization of new polymers with improved properties. Quantum chemical calculations are also useful for the further interpretation and understanding of experimental observations, since experimental methods generally are restricted to the study of many molecules at a time, lacking the ability to study the details of a single molecule, for example with regards to its geometry or the movements of single electrons.

A large number of polymers have been investigated during the PhD project presented in this thesis, often in collaboration with experimental groups at Chalmers and Linköping Universities. One very successful collaborative project resulted in the P3TQTIF polymer of so called D–A1–D–A2 type, showing 7.0% efficiency in solar cells. Several other D–A1–D–A2 polymers are computationally predicted to possess even better light-harvesting traits. One of the most important achievements is the development of a computational strategy to describe how different temperatures affect the optical and electronic properties of the polymers. With the ongoing development of computer hardware and computational methods, theoretical calculations are likely to play an increasing part in the development of new polymeric materials for use in solar cells and elsewhere. If the increase in efficiency of organic solar cells keep increasing as it has in the last decades, we are likely to see more of them sold commercially, for example for use on clothes, in windows, etc. where mechanical flexibility and absorption-tunability is vital.



Popular Abstract in Swedish

Mänsklighetens energibehov är enormt, och kommer att fortsätta öka. Fossila bränslen har använts för att tillgodose dessa behov under de senaste århundradena, men krav på långsiktig hållbarhet och drastiskt minskad klimatpåverkan har gjort att även andra energikällor behövs. Flera tusen gånger mer solenergi träffar hela tiden jorden än den energi vi använder och därför är solceller lovande för att till stor del bidra till vår framtida elektricitets¬försörjning. Den solcells¬teknologi som idag dominerar marknaden är så kallade p–n-övergångar av oorganiskt kisel som omvandlar ljus till elektricitet. Tillverkningen av dessa är dock svår och dyr. Solceller som istället tillverkas av organiska, kolbaserade material uppfanns för över femtio år sedan, och har tack vare intensiv forskning visat snabbt ökande verkningsgrader de senaste tjugo åren. De har många fördelar jämfört med oorganiska celler, bl.a. är de billigare och enklare att tillverka, de är tunna och böjbara, och kan tillverkas av återvunnet material. Deras verkningsgrad, uppåt 10%, är än så länge bara ungefär hälften så hög som i kiselceller, men deras ljusabsorptionsegenskaper kan å andra sidan justeras genom att man använder olika organiska molekyler, till skillnad från kiselcellers statiska absorptionsprofil.

De forskningsresultat som presenteras i denna avhandling handlar om dessa ljusabsorberande molekyler som är det material som omvandlar solens ljusenergi till elektrisk energi. Dessa molekyler är speciella plaster, polymerer – väldigt stora molekyler som består av ungefär fem till hundra identiska mindre repeterande enheter som alla är sammanlänkade längs en kedja. Dessa polymerer är större än de flesta syntetiska molekyler men är ändå mikroskopiska; flera biljoner av dem får plats i en liten solcell i labbskala, dvs 1 cm2 i yta och bara 100 nm tjock. Trots att en solcell innehåller så många molekyler kan många av dess egenskaper härledas från studier av enskilda polymermolekyler och det är utgångspunkten för forskningen i denna avhandling. De mest relevanta egenskaperna är de som relaterar till ljusabsorption samt laddningsgenerering och -transport. Polymererna som studerats är konjugerade, en inneboende molekylär egenskap som främjar stark ljusabsorption och god ledningsförmåga, något som är helt avgörande för solcellers effektivitet. Hundratals nya konjugerade polymerer tas fram varje år tack vare intensiv forskning. Alla ger dock inte förbättrade solceller.

Till skillnad från den traditionella föreställningen av kemi som en genomgående praktisk disciplin, är utgångspunkten för forskningen i denna avhandling ren teoretisk beräknings¬kemi. Det innebär att avancerad datormjukvara används för att förutspå molekylers egenskaper, utan att någonsin komma i fysisk kontakt med dem. Det kan låta som magi, men tänk på att vi utan att göra experimentet t.ex. kan beräkna hur lång tid det tar för en boll som släpps att nå marken om vi vet höjden, bollens vikt, och gravitationskraften. På samma sätt vet vi vilka krafter som påverkar atomerna i en molekyl, och vi kan därför noggrant förutsäga många av dess egenskaper. Ljusabsorptionen, vilken är central i detta arbete, påverkar också molekylerna och dess effekt kan också beräknas. Dock är de krafter som verkar på denna mikroskopiska nivå av kvantmekanisk natur, vilket resulterar i matematiska ekvationer som blir enormt mycket mer komplicerade än i exemplet med den fallande bollen. Därför utförs beräkningarna med kvantkemisk mjukvara i en superdator. Ett antal olika superdatorer på LUNARC i Lund och NSC i Linköping har använts i detta doktorandprojekt.

Eftersom krafterna och de därav följande ekvationerna är så komplicerade krävs många approximationer och förenklingar, vilket leder till osäkerheter i hur noggranna beräkningsmetoderna och resultaten blir. Detta är en evig fråga för teoretiska forskare, som måste jämföra sina resultat med experiment i den mån det är möjligt, för att säkerställa att beräkningarna väl representerar verkligheten. Ju större molekyl, desto mer krävande blir beräkningarna. Därför är de stora polymermolekylerna som studeras här utom räckhåll för de mest exakta beräkningsmetoderna, med dagens datorkapacitet. Icke desto mindre visar de resultat som presenteras i denna avhandling att metoder baserade på så kallad täthetsfunktionalteori närmar sig förmågan att kvantitativt förutsäga några av de viktigaste egenskaperna hos konjugerade polymerer för solcellsapplikationer, t.ex. absorptionsspektra och energinivåer. Sådana förutsägelser är mycket värdefulla eftersom de kan medföra en minskning av arbetet med att designa, syntetisera och karaktärisera nya polymer med nya egenskaper. Kvantkemiska beräkningar är också värdefulla för att tolka och förstå experimentella observationer. Experiment behandlar som regel många molekyler i taget, och kan därför inte användas för att studera enskilda molekyler på detaljnivå, t.ex. deras geometri eller enskilda elektroners rörelser.

Ett stort antal polymerer har studerats i detta doktorandprojekt, ofta i samarbete med experimentella grupper på Chalmers och Linköpings Universitet. Ett sådant samarbetesprojekt har resulterat i polymeren P3TQTIF av så kallad D–A1–D–A2-typ som uppvisat solceller av 7.0% verkningsgrad. Ett antal andra D–A1–D–A2-polymerer har från beräkningar förutspåtts ge ännu bättre absorptions¬egenskaper, vilket är mycket lovande. Ett av de viktigaste resultaten är utvecklandet av en beräkningsbaserad strategi för att beskriva hur olika temperaturer påverkar polymerernas elektroniska och optiska egenskaper. Tack vare den snabba utvecklingen av datorhårdvara och beräkningsmetoder kommer teoretiska beräkningar sannolikt att spela en allt större roll i utvecklandet av nya polymera material för solceller och inom andra områden. Om verkningsgraden för polymera solceller fortsätter att öka som den gjort de senaste årtiondena kommer vi sannolikt se mer av dem på marknaden, t.ex. för användning i kläder, på fönster, etc. där tunnhet, böjbarhet eller justerbar absorption är viktiga egenskaper. (Less)
Please use this url to cite or link to this publication:
author
supervisor
opponent
  • Eriksson, Leif A., University of Gothenburg
organization
publishing date
type
Thesis
publication status
published
subject
keywords
Conjugated polymers, density functional theory, organic photovoltaics, light-harvesting capabilities, quantum chemistry, absorption spectra, electronic structure, computational predictions
pages
216 pages
publisher
Department of Chemistry, Lund University
defense location
Hall B, Chemistry Center, Getingevägen 60, Lund
defense date
2015-09-11 13:15:00
ISBN
978-91-7422-405-4
language
English
LU publication?
yes
additional info
The information about affiliations in this record was updated in December 2015. The record was previously connected to the following departments: Theoretical Chemistry (S) (011001039), Department of Chemistry (011001220)
id
0fa33c84-5a49-42c6-86b3-38a684856865 (old id 5470126)
date added to LUP
2016-04-04 10:38:00
date last changed
2018-11-21 20:59:54
@phdthesis{0fa33c84-5a49-42c6-86b3-38a684856865,
  abstract     = {{Organic solar cells employing fullerenes blended with conjugated polymers as the main light-absorbing material have achieved power conversion efficiencies exceeding 10%. They hold promise as an alternative energy source with many advantages in terms of long-term sustainability and reduced greenhouse gas emissions. Detailed information on the electronic and geometric structure of the molecules involved is generally not accessible through experimental means, as the typically amorphous polymer films are not readily studied with e.g. X-ray crystallography. Computational chemistry, and in particular quantum chemistry as used for the research presented in this thesis, can however provide molecular level insight into the properties of these conjugated polymers. (Time-dependent) density functional theory calculations are here employed on various polymers, mainly of donor–acceptor (D–A) and D–A1–D–A2 types. Systematic studies demonstrate how the energy levels and optical properties relate to each other, as well as to the chemical composition of the polymers. In particular focus are the traits that are important for efficient solar cells: strong absorption, suitably narrow band gap, appropriate LUMO energy vs. the fullerene LUMO, and extended conjugation promoting high charge carrier mobilities. Several polymers with high-performance solar cells are studied, including TQ1 where a computationally revealed unique helical geometry is used to partially rationalize its 7.08% efficiency, and the D–A1–D–A2 polymer P3TQTIF whose two distinct acceptor units allow two strong low-energy electronic transitions, greatly enhancing its spectral coverage. Size-converged optical properties are obtained through a scheme based on extrapolations from oligomer calculations, and a detailed comparison to experiments has facilitated the development of an empirical correction for absorption energies and strengths. These corrections are subsequently used for a priori predictions of polymer absorption spectra with good agreement to experiments. Finally, a strategy is presented that includes the effect of temperature, in form of thermally populated conformations with reduced conjugation and weaker and blue-shifted absorption, yielding trends in excellent agreement with experimental optical properties. Calculations are in summary able to provide deeper insights into the fundamental properties of conjugated polymers, constituting a valuable tool for the ongoing development of materials for application in high-performance organic solar cells.}},
  author       = {{Hedström, Svante}},
  isbn         = {{978-91-7422-405-4}},
  keywords     = {{Conjugated polymers; density functional theory; organic photovoltaics; light-harvesting capabilities; quantum chemistry; absorption spectra; electronic structure; computational predictions}},
  language     = {{eng}},
  publisher    = {{Department of Chemistry, Lund University}},
  school       = {{Lund University}},
  title        = {{Computational Predictions of Conjugated Polymer Properties for Photovoltaic Applications}},
  url          = {{https://lup.lub.lu.se/search/files/5584783/5470242.pdf}},
  year         = {{2015}},
}