Metal-Organic Complexes for Doping Organic Semiconductors and Surface Doping

Stephen Barlow, Seth R. Marder

Center for Organic Photonics and Electronics and School of Chemistry and Biochemistry, Georgia Institute of Technology

Introduction

The conductivity of organic semiconductors can be increased, and the barriers to charge-carrier injection from other materials can be reduced, by the use of highly reducing or oxidizing species to n- or p-dope, respectively, the semiconductor.¹ Molecular species with well-defined redox chemistry which form stable cations or anions provide one such approach. Dopant molecules can also potentially undergo redox reactions at the surfaces of inorganic electrodes such as gold² and indium tin oxide (ITO),³ and of 2D materials such as graphene,⁴ enabling tuning of the work function (WF) and in some cases, conductivity. This article surveys two classes of metal-organic compound that have recently been introduced as effective dopants for both organic semiconductors and for electrode surfaces.

Although alkali metals are the ultimate n-dopants in terms of doping strength, they are highly reactive and not easily handled; moreover, they are not directly applicable to controlled solution doping. The corresponding ions are rather small and can potentially diffuse within doped films and/or act as deep coulombic traps for electrons on adjacent semiconductor molecules. Larger molecular dopant ions will tend to minimize diffusion and trapping, although they may lead to more structural disruption. Development of strong molecular dopants is challenging; in particular, strong n-dopants that react by a simple one-electron transfer are inevitably highly air-sensitive due to their low ionization energies, complicating their handling.¹ The coupling of chemical reactions to the electron-transfer reaction offers the possibility of achieving more stable dopants. Although several such approaches have been developed,^5,6 we considered that an approach in which a symmetrical dimer is cleaved to release a highly reducing monomer might be a particularly clean way to achieve this goal without the inevitable formation of side products. In particular, we considered that the “2 × 18-electron” dimers formed by certain 19-electron sandwich compounds including rhodocene,⁷ 1,2,3,4,5-pentamethylrhodocene,⁸ 1,2,3,4,5-pentamethyliridocene,⁹ and pentamethylcyclopentadienyl mesitylene ruthenium¹⁰ (Figure 1) might be suitable candidates for dopants.

Figure 1. Top) Structures of some organometallic dimers suitable for use as n-dopants and the monomeric cations from which they are synthesized and to which they are converted on reaction with acceptors; Lower left) synthesis of the dimers; and Lower right) mechanisms by which they react with acceptors such as organic semiconductors or electrode surfaces, where D2 denotes the neutral 2 × 18-electron dimer, D+ the corresponding 18-electron cationic monomer, and D the 19-electron neutral monomer.

Indeed, the dimers shown do react with a variety of acceptors to form the corresponding monomer cations and the acceptor anion (or dianion, if the potentials and stoichiometries allow). Whether the C—C bond cleavage precedes or follows electron transfer depends on the dimer and acceptor combination in question.^11,12 The reaction of one example with 6,13-bis(tri(isopropyl)silylethynyl)pentacene (TIPS-pentacene, Product No. 716006) is shown in Figure 2.

Figure 2. Left) UV-vis. spectra showing the reduction of TIPS-pentacene (Product No. 716006) to its radical anion by ruthenium pentamethylcyclopentadienyl 1,3,5-triethylbenzene dimer (a close analogue of (RuCp*mes)2,) in chlorobenzene solution. Right) View of the crystal structure of the doping product showing organometallic monomeric cations in red, two types of crystallographically distinct TIPS-pentacene radical anions in green and blue, and benzene molecules in yellow, turquoise, and magenta. Reproduced with permission from reference 11 (©2012 Wiley-VCH).

They can be used for the n-doping of materials including fullerenes, perylene diimides, copper phthalocyanine (702854, 546674, and 546682) (Figure 3), and TIPS-pentacene, i.e. materials with solid-state electron affinities (EAs) as low as ca. 3 eV.^13,14 The estimated potential, determine their thermodynamic ability to act as reductants, for the reactions:

0.5D₂ = D⁺ + e^–

are ca. –1.7 V vs. ferrocene in the case of (RhCp₂)₂ and ca. –2.0 V for the other examples.¹² Since this reaction involves both electron transfer and bond cleavage, these compounds are much more tolerant of air and moisture, and are therefore more easily handled than simple one-electron reductants with similar potentials, such as decamethylcobaltocene (401781, CoCp*­₂, –1.9 V). The solid dimers can be handled in air (for weighing or for transfer to a vacuum chamber), but are best stored in inert atmosphere. Solutions are also more sensitive to air than the solids, but are considerably more stable than solutions of CoCp*­₂. Suitable solvents include THF, toluene, and chlorobenzene (all of which should be dried); however, the dimers react with dichloromethane and chloroform. It should be emphasized that while the moderate air stability of the dimeric dopants may be useful for handling the pure dopants, the precautions necessary for handling the resultant doped films (and, assuming rapid reaction with semiconductors, during processing) will generally be dictated by the sensitivity of the reduced semiconductors.

Figure 3. Left) Ultraviolet photoelectron spectroscopy (UPS) spectra of 10 nm evaporated films of undoped CuPc (solid line, Product Nos. 702854, 546674 and 546682) and of CuPc doped with 3.5 wt% (RhCp2)2; the significant shift of the filled levels away from the Fermi level is indicative of n-doping. Right) J-V characteristics of simple sandwich structures based on undoped CuPc and CuPc doped with 5 wt% (RhCp2)2 showing the dramatic effects of doping on conductivity. Reproduced with permission from reference 14 (©2012 Wiley-VCH).

Organic semiconductors can be n-doped with the dimers through co-sublimation or through reaction in solution prior to spin-coating. Controlled vacuum deposition of monomeric CoCp₂ and CoCp*₂ requires specialized equipment,¹⁵ the dimers have lower vapor pressures, enabling their use in conventional evaporation crucibles. As well as demonstrating doping through the current-voltage characteristics of simple sandwich devices and ultraviolet photoelectron spectroscopy (UPS) measurements, the dimeric species have also been incorporated into devices. Ultralow doping of OFETs based on unpurified C₆₀ has been shown to largely compensate for the charge-trapping effects of defects, resulting in a reduction in the threshold voltage to a value comparable to that obtained using highly purified C₆₀ (572500), as shown in Figure 4, while retaining a high current on-off ratio (>10⁵).¹⁶ In another study, heavy doping of a C₆₀ transistor in the immediate vicinity of the source and drain electrodes was shown to reduce the contact resistance and, therefore, increase the effective field-effect mobility, while retaining useful on-off ratios.¹⁷

Figure 4. Left) Effects of different levels of doping on the Fermi level of C60, plotted both as work function (WF) (i.e., vs. vacuum) and as the position relative to the onset of ionization from the C60 HOMO. The regimes of strong dependence of the Fermi level position on doping level are attributed to trap filling, while the shallower regions are attributed to contribution of free carriers. Reproduced with permission from reference 18 (©2012 American Physical Society). Right) Effect of ultralow doping with (RuCp*mes)2 on the threshold voltage of OFETs fabricated with two batches of C60 with differing purities. The on/off ratios for all but the most heavily doped device are all 105 or higher. Reproduced with permission from reference 16 (©2012 AIP Publishing LLC).

In addition to their use as dopants for semiconductors and surfaces (see below) the reducing strength and stability of these dimers, coupled with the stability of the resulting monomer cations, may make them convenient reagents for generating reduced species for other applications. For spectroscopic studies in solution the cations are colorless and diamagnetic.

Molybdenum Tris(dithiolene)s as p-Dopants

2,3,5,6-Tetrafluoro-7,7,8,8-tetracyanoquinodimethane (376779, F₄-TCNQ, E_1/2(red) = +0.15 V vs. ferrocenium/ferrocene in CH₂Cl₂;¹⁹ solid state EA = 5.2 eV²⁰) is one of the most widely used p-dopants for organic semiconductors. It is highly volatile (making controlled sublimation challenging), poorly soluble, and prone to diffusion in some host materials. With some semiconductors, significant donor-acceptor orbital overlap precludes an integral electron-transfer reaction.²¹

The recent introduction of the more three-dimensional molybdenum tris(1,2-bis(trifluoromethyl)ethane-1,2-dithiolene, Mo(tfd)₃ (795712, Figure 5)^22,23 – a long-established one-electron oxidant that has been used in synthetic organometallic chemistry²⁴ – as a p-dopant helps address some of these drawbacks. It can be readily evaporated in standard vacuum deposition equipment and has been co-sublimed with organic semiconductors including a-NPD (765015),^19,25 pentacene (698423 and 684848),²⁶ and CuPc (702854, 546674 and 546682),¹⁴ to give n-doped films. The Mo(tfd)₃^– anion has been found to show stability with respect to diffusion in a-NPD at temperatures of up to 110 °C.²⁵ Mo(tfd)₃ has been applied in pentacene OFETs; heavy p-doping of a pentacene OFET in the vicinity of source and drain electrodes was found to reduce contact resistance in similar way to the contact n-doping of C₆₀ described above.²⁶ The “remote doping” of a pentacene OFET, in which the dopant is located within an adjacent a-NPD layer, has also been demonstrated.²⁷

Figure 5. Synthesis of Mo(tfd)3 and more soluble analogues.

Although Mo(tfd)₃ has reasonable solubility in organic solvents, its doping products are not always very soluble in solvents suitable for spin-coating of semiconductor films. For example, attempts to dope the archetypal OPV polymer P3HT at all but ultralow doping levels led to the very rapid formation of precipitates. Accordingly more soluble analogues molybdenum tris-(1-(methoxycarbonyl)-2-(trifluoromethyl)ethane-1,2-dithiolene), Mo(tfd-CO₂Me)₃, and molybdenum tris-(1-(trifluoroacetyl)-2-(trifluoromethyl)ethane-1,2-dithiolene), Mo(tfd-COCF₃)₃, were developed for solution p-doping of semiconductors (Figure 5, Table 1). P3HT doped with 4 wt% Mo(tfd-CO₂Me)₃ has been found to be an effective hole-collecting contact for organic photovoltaics that is processible from organic solution (Figure 6),²⁸ while Mo(tfd-COCF₃)₃ has been used in solution-processed field-effect transistors based on p-doped TIPS-pentacene / polymer blends.²⁹

Figure 6. Left) J-V characteristics for bulk heterojunction OPV devices under AM1.5 illumination and Right) in the dark where the interlayer is either P3HT (B), P3HT p-doped with Mo(tfd-CO2Me)3 (C and D), or PEDOT:PSS (E). Note the similarity of the data for D and E. Reproduced with permission from reference 28 (©2013 Wiley-VCH).

All three of these p-dopants are air stable in the solid state. They are soluble in solvents such as dichloromethane, chlorobenzene, and toluene; these solutions, however, slowly decompose on exposure to air, presumably through reduction by water. The compounds are incompatible with more Lewis basic solvents such as THF, acetonitrile, and acetone, in which they are reduced or otherwise decomposed.

Applications to Surface Doping

Some of the dimeric n-dopants discussed here have been used for surface doping, both using immersion in toluene solutions and by sublimation onto substrates. The work functions (WFs) of indium tin oxide (ITO), ZnO, and Au can all be significantly lowered.³⁰ Thus, ITO can be rendered an effective electron-injecting electrode. The WF reduction is particularly dramatic when dopant is sublimed onto a single-crystal of ZnO.³¹ In all cases, the major origin of the WF change is the interface dipole formed by the partial monolayer of dopant-derived monomeric cations on the negatively charged electrode material.

When applied to monolayer graphene³² the WF change has significant contributions from both population of the graphene conduction band and the interface dipole. Moreover, the conductivity of the graphene is significantly increased, with only minimal effects on its transparency. Graphene has also been p-doped using Mo(tfd-COCF₃)₃, leading to increased conductivity, but to significantly increased WF.³² Thus, using n- and p-dopants, the WF of graphene can be tuned over a range of ca. 1.8 eV as shown schematically in Figure 7. In principle, similar effects should be possible on other 2D materials; indeed, a dopant based on the dimer of an organic radical, but with similar reactivity to the organometallic dimers described here, has recently been used in the n-doping of large-area trilayer MoS₂.³³

Figure 7. Schematic showing effects of n- and p-doping of graphene with metal-organic dopants based on data from reference 32.

Summary

The dimers of group-9 metallocenes and of RuCp*mes are versatile and powerful n-dopants that are relatively easily handled and can be applied to doping both organic semiconductors and surfaces using either solution or vacuum processing. Molybdenum tris(dithiolene) complexes are a similarly versatile family of p-dopants.

References

1. Walzer K, Maennig B, Pfeiffer M, Leo K. 2007. Highly Efficient Organic Devices Based on Electrically Doped Transport Layers. Chem. Rev.. 107(4):1233-1271. https://doi.org/10.1021/cr050156n

2. Bröker B, Blum R, Frisch J, Vollmer A, Hofmann OT, Rieger R, Müllen K, Rabe JP, Zojer E, Koch N. 2008. Gold work function reduction by 2.2eV with an air-stable molecular donor layer. Appl. Phys. Lett.. 93(24):243303. https://doi.org/10.1063/1.3049616

3. Osikowicz W, Crispin X, Tengstedt C, Lindell L, Kugler T, Salaneck WR. 2004. Transparent low-work-function indium tin oxide electrode obtainedby molecular scale interface engineering. Appl. Phys. Lett.. 85(9):1616-1618. https://doi.org/10.1063/1.1785873

4. Chen W, Chen S, Qi DC, Gao XY, Wee ATS. 2007. Surface Transfer p-Type Doping of Epitaxial Graphene. J. Am. Chem. Soc.. 129(34):10418-10422. https://doi.org/10.1021/ja071658g

5. Werner AG, Li F, Harada K, Pfeiffer M, Fritz T, Leo K. 2003. Pyronin B as a donor for n-type doping of organic thin films. Appl. Phys. Lett.. 82(25):4495-4497. https://doi.org/10.1063/1.1583872

6. Wei P, Oh JH, Dong G, Bao Z. 2010. Use of a 1H-Benzoimidazole Derivative as ann-Type Dopant and To Enable Air-Stable Solution-Processedn-Channel Organic Thin-Film Transistors. J. Am. Chem. Soc.. 132(26):8852-8853. https://doi.org/10.1021/ja103173m

7. Fischer EO, Wawersik H. 1966. Über aromatenkomplexe von metallen. Journal of Organometallic Chemistry. 5(6):559-567. https://doi.org/10.1016/s0022-328x(00)85160-8

8. Gusev OV, Denisovich LI, Peterleitner MG, Rubezhov AZ, Ustynyuk NA, Maitlis PM. 1993. Electrochemical generation of 19- and 20-electron rhodocenium complexes and their properties. Journal of Organometallic Chemistry. 452(1-2):219-222. https://doi.org/10.1016/0022-328x(93)83193-y

9. Gusev OV, Peterleitner MG, Ievlev MA, Kal'sin AM, Petrovskii PV, Denisovich LI, Ustynyuk NA. 1997. Reduction of iridocenium salts [Ir(?5-C5Me5)(?5-L)]+ (L = C5H5, C5Me5, C9H7); ligand-to-ligand dimerisation induced by electron transfer. Journal of Organometallic Chemistry. 531(1-2):95-100. https://doi.org/10.1016/s0022-328x(96)06675-2

10. Gusev OV, Ievlev MA, Peterleitner MG, Peregudova SM, Denisovich LI, Petrovskii PV, Ustynyuk NA. 1997. Reduction of ruthenium arenecyclopentadienyl complexes reactions induced by electron transfer. Journal of Organometallic Chemistry. 534(1-2):57-66. https://doi.org/10.1016/s0022-328x(96)06888-x

11. Guo S, Mohapatra SK, Romanov A, Timofeeva TV, Hardcastle KI, Yesudas K, Risko C, Brédas J, Marder SR, Barlow S. 2012. n-Doping of Organic Electronic Materials Using Air-Stable Organometallics: A Mechanistic Study of Reduction by Dimeric Sandwich Compounds. Chem. Eur. J.. 18(46):14760-14772. https://doi.org/10.1002/chem.201202591

12. Mohapatra SK, Fonari A, Risko C, Yesudas K, Moudgil K, Delcamp JH, Timofeeva TV, Brédas J, Marder SR, Barlow S. 2014. Dimers of Nineteen-Electron Sandwich Compounds: Crystal and Electronic Structures, and Comparison of Reducing Strengths. Chem. Eur. J.. 20(47):15385-15394. https://doi.org/10.1002/chem.201404007

13. Qi Y, Mohapatra SK, Bok Kim S, Barlow S, Marder SR, Kahn A. 2012. Solution doping of organic semiconductors using air-stable n-dopants. Appl. Phys. Lett.. 100(8):083305. https://doi.org/10.1063/1.3689760

14. Guo S, Kim SB, Mohapatra SK, Qi Y, Sajoto T, Kahn A, Marder SR, Barlow S. 2012. n-Doping of Organic Electronic Materials using Air-Stable Organometallics. Adv. Mater.. 24(5):699-703. https://doi.org/10.1002/adma.201103238

15. Chan CK, Amy F, Zhang Q, Barlow S, Marder S, Kahn A. 2006. N-type doping of an electron-transport material by controlled gas-phase incorporation of cobaltocene. Chemical Physics Letters. 431(1-3):67-71. https://doi.org/10.1016/j.cplett.2006.09.034

16. Olthof S, Singh S, Mohapatra SK, Barlow S, Marder SR, Kippelen B, Kahn A. 2012. Passivation of trap states in unpurified and purified C60 and the influence on organic field-effect transistor performance. Appl. Phys. Lett.. 101(25):253303. https://doi.org/10.1063/1.4772551

17. Singh S, Mohapatra SK, Sharma A, Fuentes-Hernandez C, Barlow S, Marder SR, Kippelen B. 2013. Reduction of contact resistance by selective contact doping in fullerene n-channel organic field-effect transistors. Appl. Phys. Lett.. 102(15):153303. https://doi.org/10.1063/1.4802237

18. Olthof S, Mehraeen S, Mohapatra SK, Barlow S, Coropceanu V, Brédas J, Marder SR, Kahn A. Ultralow Doping in Organic Semiconductors: Evidence of Trap Filling. Phys. Rev. Lett.. 109(17): https://doi.org/10.1103/physrevlett.109.176601

19. Qi Y, Sajoto T, Barlow S, Kim E, Brédas J, Marder SR, Kahn A. 2009. Use of a High Electron-Affinity Molybdenum Dithiolene Complex to p-Dope Hole-Transport Layers. J. Am. Chem. Soc.. 131(35):12530-12531. https://doi.org/10.1021/ja904939g

20. Gao W, Kahn A. 2001. Controlled p-doping of zinc phthalocyanine by coevaporation with tetrafluorotetracyanoquinodimethane: A direct and inverse photoemission study. Appl. Phys. Lett.. 79(24):4040-4042. https://doi.org/10.1063/1.1424067

21. Salzmann I, Heimel G, Duhm S, Oehzelt M, Pingel P, George BM, Schnegg A, Lips K, Blum R, Vollmer A, et al. Intermolecular Hybridization Governs Molecular Electrical Doping. Phys. Rev. Lett.. 108(3): https://doi.org/10.1103/physrevlett.108.035502

22. King RB. 1963. Organosulfur Derivatives of the Metal Carbonyls. III. The Reaction between Molybdenum Hexacarbonyl and Bis-(trifluoromethyl)-dithietene. Inorg. Chem.. 2(3):641-642. https://doi.org/10.1021/ic50007a056

23. Davison A, Edelstein N, Holm RH, Maki AH. 1964. Synthetic and Electron Spin Resonance Studies of Six-Coordinate Complexes Related by Electron-Transfer Reactions. J. Am. Chem. Soc.. 86(14):2799-2805. https://doi.org/10.1021/ja01068a010

24. Connelly NG, Geiger WE. 1996. Chemical Redox Agents for Organometallic Chemistry. Chem. Rev.. 96(2):877-910. https://doi.org/10.1021/cr940053x

25. Qi Y, Sajoto T, Kro?ger M, Kandabarow AM, Park W, Barlow S, Kim E, Wielunski L, Feldman LC, Bartynski RA, et al. 2010. A Molybdenum Dithiolene Complex asp-Dopant for Hole-Transport Materials: A Multitechnique Experimental and Theoretical Investigation. Chem. Mater.. 22(2):524-531. https://doi.org/10.1021/cm9031623

26. Tiwari SP, Potscavage WJ, Sajoto T, Barlow S, Marder SR, Kippelen B. 2010. Pentacene organic field-effect transistors with doped electrode-semiconductor contacts. Organic Electronics. 11(5):860-863. https://doi.org/10.1016/j.orgel.2010.01.029

27. Zhao W, Qi Y, Sajoto T, Barlow S, Marder SR, Kahn A. 2010. Remote doping of a pentacene transistor: Control of charge transfer by molecular-level engineering. Appl. Phys. Lett.. 97(12):123305. https://doi.org/10.1063/1.3491429

28. Dai A, Zhou Y, Shu AL, Mohapatra SK, Wang H, Fuentes-Hernandez C, Zhang Y, Barlow S, Loo Y, Marder SR, et al. 2014. Enhanced Charge-Carrier Injection and Collection Via Lamination of Doped Polymer Layers p-Doped with a Solution-Processible Molybdenum Complex. Adv. Funct. Mater.. 24(15):2197-2204. https://doi.org/10.1002/adfm.201303232

29. Belasco J, Mohapatra SK, Zhang Y, Barlow S, Marder SR, Kahn A. 2014. Molecular doping and tuning threshold voltage in 6,13-bis(triisopropylsilylethynyl)pentacene/polymer blend transistors. Appl. Phys. Lett.. 105(6):063301. https://doi.org/10.1063/1.4892809

30. Giordano AJ, Pulvirenti F, Khan TM, Fuentes-Hernandez C, Moudgil K, Delcamp JH, Kippelen B, Barlow S, Marder SR. 2015. Organometallic Dimers: Application to Work-Function Reduction of Conducting Oxides. ACS Appl. Mater. Interfaces. 7(7):4320-4326. https://doi.org/10.1021/am5087648

31. Schlesinger R, Bianchi F, Blumstengel S, Christodoulou C, Ovsyannikov R, Kobin B, Moudgil K, Barlow S, Hecht S, Marder S, et al. 2015. Efficient light emission from inorganic and organic semiconductor hybrid structures by energy-level tuning. Nat Commun. 6(1): https://doi.org/10.1038/ncomms7754

32. Paniagua SA, Baltazar J, Sojoudi H, Mohapatra SK, Zhang S, Henderson CL, Graham S, Barlow S, Marder SR. Production of heavily n- and p-doped CVD graphene with solution-processed redox-active metal?organic species. Mater. Horiz.. 1(1):111-115. https://doi.org/10.1039/c3mh00035d

33. Tarasov A, Zhang S, Tsai M, Campbell PM, Graham S, Barlow S. 2015. Controlled Doping of Large-Area Trilayer MoS2with Molecular Reductants and Oxidants. Adv. Mater.. 27(7):1175-1181. https://doi.org/10.1002/adma.201404578