Fullerene-Based Semiconductors

Dr. David Kronholm¹, Prof. Dr. Jan C.², Hummelen²

¹Solenne BV, The Netherlands, ²Molecular Electronics, Zernike Institute for Advanced Materials, and Stratingh Institute of Chemistry, University of Groningen, The Netherlands

Material Matters 2007, 2.3, 16.

Introduction

Since the first publication in 1995 describing a bulk heterojunction photodiode incorporating a methanofullerene,¹ significant progress has been made in improving device performance and the scope of device research has broadened widely. The most commonly used fullerene derivative in organic electronics is the methanofullerene Phenyl-C₆₁- Butyric-Acid-Methyl-Ester ([60]PCBM).² Various analogues (termed here PCBMs) have been made and tested as ntype semiconductors. The use in organic photovoltaics, photodetectors³ and organic field effect transistors (OFETs)⁴ among other applications, has been investigated and is under active development. After more than a decade of research, it appears that PCBMs have proven to be effective solution processable organic n-type semiconductors in a variety of thin film organic electronics applications.

Figure 1 is a schematic showing a typical bulk heterojunction organic photovoltaic device (OPV) architecture and performance characteristics. Fabrication of bulk heterojunction OPVs requires soluble fullerene derivatives in order to form blends with p-type polymer semiconductors. PCBMs preserve important electronic and optical properties of the parent fullerenes while providing a significant increase in solubility and processability. Some of these properties are fast electron transfer, an adequate dielectric constant, isotropic (in the case of C₆₀ derivatives) or relatively isotropic (in the case of C₇₀ and C₈₄ derivatives) electron accepting due to the symmetry of the fullerene acceptor, and good electron mobilities. Coupling these properties of the parent fullerenes with improved solubility in common organic solvents and the observed desirable precipitation kinetics of the PCBMs provides robust formation of uniform nanoparticulate n-type domains in the final film.

Figure 1. a) Present day standard architecture for bulk heterojunction OPV devices. The unnamed layers are a transparent conducting oxide (TCO, e.g. ITO; top) and an ultra-thin protective layer (e.g. LiF; bottom); b) I/V curves of optimized regioregular P3HT:methanofullerene (1:1) PV cells under AM1.5 illumination, 1 sun intensity. Current density values are corrected for spectral mismatch and real active area. (Presented at MRS spring meeting 2007; devices by Lacramioara Popescu, Univ. of Groningen). It can be seen that a relatively minor change in molecular structure of the n-type semiconductor influences the performance of the final device.

Figure 2. Library of the [60]PCBM analogues (PCBMs)

Analogues of [60]PCBM based on higher fullerenes (C₇₀ and C₈₄) have been synthesized and tested in devices, and [60]PCBM analogues based on alterations of the addend moiety to vary miscibility/solubility and electronic properties have also been developed and tested. Figure 2 shows the more commonly used and best performing PCBMs. Since the formation of thin film organic electronics devices is highly complex, especially due to morphology considerations, it is hard to generalize or extrapolate trends and predict which PCBM will give the best performance in a given device or architecture. These molecules do however represent a library available for the experimental researcher to explore optimization, and each has shown in different devices and architectures to provide advantages. Different purity grades have also been developed allowing for significantly lower prices and the availability of commercial scale volumes.

Overview of PCBMs as n-Type Semiconductors in Organic Electronics Applications

Organic Photovoltaics (OPV)

[60]PCBM is still by far the most commonly used n-type component in organic photovoltaics. Over at least the last 6 years, the published world record power conversion efficiency (ɳ) for an organic photovoltaic device to our knowledge has been continuously held by devices incorporating [60]PCBM, save for a period in which a device containing [70]PCBM held the record. MDMO-PPV:PCBM devices were thoroughly studied and characterized, eventually leading to ɳ=3.0%⁵ when [70]PCBM was substituted for [60]PCBM, which earlier had given ɳ=2.5%.⁶ The increase was due to the higher optical absorption of [70]PCBM in visible wavelengths compared to [60]PCBM. [84]PCBM has an even stronger absorption in the visible wavelengths, though the better electron accepting ability led to a diminished performance in OPV, because it was used in combination with a relatively strongly electron donating donor polymer.⁷ More recently, researchers and developers have transitioned to polythiophene/PCBM systems, and ɳ’s of 4.4%–6% have been published by several groups.^8,9 Careful control of morphology, either by annealing or slow evaporation, provides a significant improvement in performance.

The state of the art roadmap for research and development to achieve ɳ of 10% focuses on improving morphology and polymer characteristics¹⁰ leading to an inference that [60]PCBM is largely adequate as the n-type semiconductor for improved devices. However, improvements in morphology control with polythiophenes, where more extensive demixing of the PCBM and polymer phase has been observed, is also desired. This has led to the design and testing of a new molecule, [60]ThCBM, which in preliminary results does appear to give a slightly more advantageous morphology with P3HT.¹¹ [60]ThCBM also preserves the electronic properties (LUMO and mobility) of [60]PCBM. Increases in LUMO level of the n-type have also been long sought by OPV developers and a recently synthesized molecule, 2,3,4-OMe-PCBM, shows a modest though significant increase in LUMO. This molecule has been shown to give a higher open circuit voltage (V_OC) in combination with MDMO-PPV¹² but has not yet been fully characterized in OPV devices.

Organic Field Effect Transistors (OFETs)

Relatively high mobilities for an organic semiconductor have been demonstrated for [60]PCBM devices (1 x 10^–2–2 x 10^–1 cm²/Vs),⁴ as well as ambipolar transport which allowed for the construction of inverters.¹³ Stability has been an issue, though efficient passivation has been reported. [70]PCBM thus far has shown about an order of magnitude lower electron mobilities but allows for shorter annealing times and higher stability. [84]PCBM has shown very good stability, in combination with an electron mobility up to 3 x 10^–3 and a hole mobility of 10^–5–10^–4 cm²/Vs.¹⁴ Blends of conjugated polymers with PCBMs can also be used for ambipolar OFETs.¹⁵ Less work has been done with OFET devices using PCBMs compared to OPV, and it can be expected that mobility improvements can be obtained applying similar control of film morphology (optimal solvents and evaporation/annealing) as has been demonstrated with OPV.

Organic Photodetectors

Concurrent with the early development of OPV devices, bulk heterojunction organic photodetectors based on similar photodiodes were also developed.¹⁶ Performance adequate for commercial application was realized, with low dark currents, high external quantum efficiencies (80%), and fast transient behavior.³ Significantly, large area applications are envisioned due to the cost advantages of organic thin films over siliconbased devices.

Properties of the Individual Members of the PCBM Library

[60]PCBM (Prod. Nos. 684430, 684449, 684457)

[60]PCBM is present as a single isomer. An interesting feature of [60]PCBM which may correlate with its performance is that it preserves to a high degree the electronic and physical properties of C₆₀. Single crystal structure analysis shows that intermolecular spacing is essentially identical to C₆₀, with the shortest ball-to-ball spacing curiously being slightly smaller in PCBM than C₆₀.¹⁷ It has been consistently demonstrated that deviation to too great of a degree from the compact structure of [60]PCBM (and thus from the parent fullerene) leads to diminished performance.

The >99.5% grade of [60]PCBM (Prod. No. 684449) has been and is still most extensively used by researchers, but testing has shown that in some devices and architectures >99% grade (Prod No. 684430) is acceptable. However, it must be cautioned that since precipitation/crystallization behavior of the PCBM has a strong influence on morphology and small amounts of impurities may have a strong influence on precipitation/crystallization kinetics, the different grades must be examined to determine optimal performance.

[70]PCBM (Prod. No. 684465)

The motivation for the first synthesis of [70]PCBM was to take advantage of the increased optical absorption in the visible region compared to [60]PCBM. This can be especially advantageous in combination with relatively large bandgap donors like MDMO-PPV. It is prepared as a mixture of 1 major and 2 minor isomers and it is used as such. Since the LUMO energies of [70]PCBM and [60]PCBM are very close,⁵ it offers the opportunity for improvement in light harvesting where the film absorption is a function of the n-type as well as the ptype while preserving the electronic performance of [60]PCBM.

It should be noted that though the measured amount of [70]PCBM easily dissolved in common solvents is somewhat higher than [60]PCBM, the precipitation behavior in solution processed thin film devices is different (and consistent with behavior that would be expected for a less soluble molecule). This led to the necessity for the use of a stronger fullerene solvent ortho-dichlorobenzene (ODCB) for optimal performance, at least in combination with MDMO-PPV,⁵ as using chlorobenzene resulted in extensive de-mixing and large [70]PCBM domains. We speculate that the reduced symmetry of the C₇₀ molecule and mixed isomer form induces differences in precipitation kinetics compared to [60]PCBM.

[84]PCBM (Prod. No. 684473)

[84]PCBM comes as a mixture of mainly three isomers and it is used as such.⁷ [84]PCBM shows panchromatic absorption (extending into the NIR), in combination with a 350 mV lower LUMO level, compared to that of [60]PCBM. This lower LUMO led to a diminished performance in the OPV system in combination with MDMO-PPV, but it is most likely an important factor in the much better air stability of the single component OFET.¹⁸

[60]PCB-Cn esters (Prod. Nos. 685321 (C₄) and 684481 (C₈))

One strategy for morphology control has been to use stronger fullerene solvents, which reduces the precipitation driving force for the PCBM, thus leading to smaller PCBM domains and smoother films. Similarly, the solubility of [60]PCBM has been improved upon by replacing the methyl group by larger alkyl moieties. It has been demonstrated that for certain solvent systems (poorer solvents), higher solubility provides performance advantages, which improvements diminish with larger increases in solubility.¹⁹ Depending on solvent choice and device architecture, these higher solubility versions may provide advantages in forming the desired morphology. Alkyl chain lengths of n=4, 8, 12, 16, and 18 have been synthesized. [60]PCB-C₄ shows only a moderate increase in solubility, while n=8 and higher are significantly more soluble. It is likely that the crystal structure of the derivatives with longer chain lengths deviates from that of the compact structure of PCBM, and it is thought that this may lead to reduced mobilities. It may also influence recombination via lowering the dielectric constant.

[60]ThCBM (Prod. No. 688215)

Phase separation has been shown to be more extensive in polythiophene:PCBM systems, and so to increase miscibility a thienyl group was substituted for the phenyl of [60]PCBM. Though not yet fully characterized in devices, [60]ThCBM may offer more controllable morphology with polythiophenes due to improved miscibility. The LUMO and pure thin-film mobilities are very similar to those of [60]PCBM.¹¹

2,3,4-OMe-PCBM

The VOC in OPV has been shown to be a function of the HOMO of the donor and LUMO of the acceptor, which has led to efforts in increasing the LUMO relative to [60]PCBM. 2,3,4-OMe-PCBM has been synthesized for this purpose and does demonstrate a higher LUMO while moderately preserving the processability of [60]PCBM. Preliminary work shows an increased V_OC in combination with MDMO-PPV,¹² though the overall device characteristics have yet to be optimized.

d₅-PCBM (Prod. No. 684503)

As thin film organic electronics device performance is such a strong function of film morphology, information regarding the structure of donor–acceptor blends can be crucial. Dynamic secondary ion mass spectroscopy (SIMS) can be used with blends where PCBM is replaced by the deuterated version d₅-PCBM to provide a detailed elemental analysis of film morphology in three dimensions. This technique successfully elucidated the 3-D morphology of MDMO-PPV:PCBM OPV devices.²⁰

Practical Use

In solution processing of organic electronic device layers containing PCBMs, care should be taken to ensure complete dissolution of the PCBM. Incomplete dissolution can lead to seeding of the precipitation, leading to larger PCBM domain sizes and even in some cases micron-scale crystallite formation. Typically stirring for 8 hours or more at concentrations well below the solubility limits is adequate. Filtration should also be used. Sonication alone does not ensure adequate mixing and may leave sub-micron or nano-scale suspended particulates not visible to the naked eye.

PCBMs are relatively stable; though fullerenes typically form epoxides upon exposure to light and air, the process is slow. Storage in sealed, opaque containers is adequate, though for longer storage (> 6 months), purging with an inert gas (N₂ or Ar) may be desired. In practice we find that device performance in OPV applications does not degrade using PCBM that has been stored sealed in an opaque container (without inert gas purging or storage in a glovebox) for up to a year.

Conclusions and Outlook

PCBMs have shown to be broadly applicable as solution processable organic n-type semiconductors. The variations available represent opportunities for optimization in various devices and architectures. Volume and cost considerations have also been adequately addressed, allowing for commercial use of these materials. Further optimization of molecules is likely, though it appears that the compact structure of the PCBMs is a desirable property in preserving the inherently desirable properties of the parent fullerenes, while still providing adequate solution processability. The presently available library of PCBMs allows the researcher to vary a number of important parameters governing the action of the various molecular electronics devices. The present choice is in terms of miscibility, solubility, optical absorption, air stability, and LUMO energy. A likely experimental strategy to minimize R&D time and cost is the use of [60]PCBM for basic understanding and preliminary optimization, and the testing of various PCBM variations when undertaking fine tuning of the system.

References

1. Yu G, Gao J, Hummelen JC, Wudl F, Heeger AJ. 1995. Polymer Photovoltaic Cells: Enhanced Efficiencies via a Network of Internal Donor-Acceptor Heterojunctions. Science. 270(5243):1789-1791. https://doi.org/10.1126/science.270.5243.1789

2. Hummelen JC, Knight BW, LePeq F, Wudl F, Yao J, Wilkins CL. 1995. Preparation and Characterization of Fulleroid and Methanofullerene Derivatives. J. Org. Chem.. 60(3):532-538. https://doi.org/10.1021/jo00108a012

3. Rauch T, Henseler D, Schilinsky P, Waldauf C, Hauch J, Brabec CJ. 2004. Performance of bulk-heterojunction organic photodetectors. 4th IEEE Conf. on Nanotechnology; 15 Aug 2019; Munich, Germany IEEE. p. 632–634..

4. Anthopoulos TD, de Leeuw DM, Cantatore E, van ?t Hof P, Alma J, Hummelen JC. 2005. Solution processible organic transistors and circuits based on a C70 methanofullerene. Journal of Applied Physics. 98(5):054503. https://doi.org/10.1063/1.2034083

5. Wienk MM, Kroon JM, Verhees WJH, Knol J, Hummelen JC, van Hal PA, Janssen RAJ. 2003. Efficient Methano[70]fullerene/MDMO-PPV Bulk Heterojunction Photovoltaic Cells. Angew. Chem. Int. Ed.. 42(29):3371-3375. https://doi.org/10.1002/anie.200351647

6. Shaheen SE, Brabec CJ, Sariciftci NS, Padinger F, Fromherz T, Hummelen JC. 2001. 2.5% efficient organic plastic solar cells. Appl. Phys. Lett.. 78(6):841-843. https://doi.org/10.1063/1.1345834

7. Kooistra FB, Mihailetchi VD, Popescu LM, Kronholm D, Blom PWM, Hummelen JC. 2006. New C84Derivative and Its Application in a Bulk Heterojunction Solar Cell. Chem. Mater.. 18(13):3068-3073. https://doi.org/10.1021/cm052783z

8. Li G, Shrotriya V, Huang J, Yao Y, Moriarty T, Emery K, Yang Y. 2005. High-efficiency solution processable polymer photovoltaic cells by self-organization of polymer blends. Nature Mater. 4(11):864-868. https://doi.org/10.1038/nmat1500

9. Kim K, Liu J, Namboothiry MAG, Carroll DL. 2007. Roles of donor and acceptor nanodomains in 6% efficient thermally annealed polymer photovoltaics. Appl. Phys. Lett.. 90(16):163511. https://doi.org/10.1063/1.2730756

10. Scharber M, Mühlbacher D, Koppe M, Denk P, Waldauf C, Heeger A, Brabec C. 2006. Design Rules for Donors in Bulk-Heterojunction Solar Cells?Towards 10?% Energy-Conversion Efficiency. Adv. Mater.. 18(6):789-794. https://doi.org/10.1002/adma.200501717

11. Popescu M, van’t Hof P, Sieval AB, Jonkman H, Hummelen JC. 2006. App. Phys. Lett.. 89, 213507.

12. Kooistra FB, Knol J, Kastenberg F, Popescu LM, Verhees WJH, Kroon JM, Hummelen JC. 2007. Increasing the Open Circuit Voltage of Bulk-Heterojunction Solar Cells by Raising the LUMO Level of the Acceptor. Org. Lett.. 9(4):551-554. https://doi.org/10.1021/ol062666p

13. Anthopoulos TD, de Leeuw DM, Cantatore E, Setayesh S, Meijer EJ, Tanase C, Hummelen JC, Blom PWM. 2004. Organic complementary-like inverters employing methanofullerene-based ambipolar field-effect transistors. Appl. Phys. Lett.. 85(18):4205-4207. https://doi.org/10.1063/1.1812577

14. Anthopoulos TD, de Leeuw DM, Cantatore E, van ?t Hof P, Alma J, Hummelen JC. 2005. Solution processible organic transistors and circuits based on a C70 methanofullerene. Journal of Applied Physics. 98(5):054503. https://doi.org/10.1063/1.2034083

15. Meijer EJ, de Leeuw DM, Setayesh S, van Veenendaal E, Huisman B-, Blom PWM, Hummelen JC, Scherf U, Klapwijk TM. 2003. Solution-processed ambipolar organic field-effect transistors and inverters. Nature Mater. 2(10):678-682. https://doi.org/10.1038/nmat978

16. Yu G, Yong C. 2002. U.S. Patent App.. [dissertation]. 20020017612:

17. Rispens MT, Meetsma A, Rittberger R, Brabec CJ, Sariciftci NS, Hummelen JC. Influence of the solvent on the crystal structure of PCBM and the efficiency of MDMO-PPV:PCBM ?plastic? solar cells. Chem. Commun..(17):2116-2118. https://doi.org/10.1039/b305988j

18. Anthopoulos T, Kooistra F, Wondergem H, Kronholm D, Hummelen J, de?Leeuw D. 2006. Air-Stable n-Channel Organic Transistors Based on a Soluble C84 Fullerene Derivative. Adv. Mater.. 18(13):1679-1684. https://doi.org/10.1002/adma.200600068

19. Zheng L, Zhou Q, Deng X, Yuan M, Yu G, Cao Y. 2004. Methanofullerenes Used as Electron Acceptors in Polymer Photovoltaic Devices. J. Phys. Chem. B. 108(32):11921-11926. https://doi.org/10.1021/jp048890i

20. Bulle-Lieuwma C, van Duren J, Yang X, Loos J, Sieval A, Hummelen J, Janssen R. 2004. Characterization of poly(p-phenylene vinylene)/methanofullerene blends of polymer solar cells by time-of-flight secondary ion mass spectrometry. Applied Surface Science. 231-232274-277. https://doi.org/10.1016/j.apsusc.2004.03.251