© 2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim 1535 www.advmat.de www.MaterialsViews.com wileyonlinelibrary.com COMMUNICATION Adv. Mater. 2012, 24, 1535–1539 Alexander Mityashin,* Yoann Olivier, Tanguy Van Regemorter, Cedric Rolin, Stijn Verlaak, Nicolas G. Martinelli, David Beljonne, Jérôme Cornil, Jan Genoe, and Paul Heremans* Unraveling the Mechanism of Molecular Doping in Organic Semiconductors A. Mityashin, Dr. C. Rolin, Dr. S. Verlaak, Dr. J. Genoe, Prof. P. Heremans Imec, Kapeldreef 75, Leuven, 3001, Belgium E-mail: mityash@imec.be; heremans@imec.be A. Mityashin, Prof. P. Heremans ESAT, Katholieke Universiteit Leuven Kasteelpark Arenberg 10, Leuven, 3001, Belgium Dr. Y. Olivier, Dr. T. Van Regemorter, Dr. N. G. Martinelli, Dr. D. Beljonne, Dr. J. Cornil Laboratory for Chemistry of Novel Materials University of Mons Place du Parc 20, Mons, 7000, Belgium DOI: 10.1002/adma.201104269 Over the last decade molecular doping has been very success- fully used to control electrical conductivity in organic electronic devices, particularly in charge transport layers of organic light emitting diodes and solar cells. [1] It has been demonstrated that the conductivity of an n-(p-)type organic semiconductor can be reproducibly enhanced by several orders of magnitude when introducing a certain amount of strong electron-(hole-) donating molecules. [2] The doping mechanism is thereby usu- ally assumed to be a two-step process, similar to in inorganic semiconductors: [1] i) ionization of the dopant with donation of a charge carrier (electron or hole) to the host, and ii) dissociation of this charged pair, effectively creating a free charge carrier. The latter process is, however, not at all obvious in an organic semiconductor matrix, since the low electrical permittivity of organics (ε = 3–4) means that the donated charge carrier must be bound to the parental dopant by a strong Coulombic binding energy of 10 to 20 times kT/q. There is at present no detailed understanding of the dissociation process of strongly bound charge carriers from the ionized molecules. In this communication, we unravel microscopic details of the doping mechanism in organics, and in particular of the charge carrier dissociation event. To this end, we analyze the archetypal host–dopant system pentacene doped by tetrafluorotetracyano- quinodimethane (F 4 TCNQ). The key findings are, however, rather general and transpose to other host–dopant combina- tions. Considering the above-mentioned two-step process, we first show the impact of doping molecules on the charge-trans- port energy levels of the host and donation of charge carriers by the dopant to the host semiconductor. We then calculate the binding energy of the charge pair on neighboring dopant and host molecules, and estimate the dissociation probability at dif- ferent doping concentrations. We find that the doping efficiency is indeed strongly dependent on the doping concentration and that there is a threshold doping level below which doping simply has no effect on the electrical conductivity. These findings extend the understanding of doping of inorganic semiconductors to organic semiconductors, by adding to the consideration of thermal activation of dopants the concept of activation by the proximity of multiple doping sites. The first step is to determine the charge-transport energy levels of host semiconductor molecules in the vicinity of (one or several) dopant molecules by means of a molecular micro- electrostatics (ME) model, [3–4] which is parameterized against quantum-chemical calculations. Subsequently, we determine the fate of the charge carriers donated by the doping molecules using a master-equation formalism for charge migration, [5] in which the hopping rates are computed using the Marcus– Levich–Jortner theory. Technical details of these calculations are presented in the Supporting Information. An atomistic model of the doped material was constructed by introducing a certain amount of doping molecules into vacancies created in a three-dimensional pentacene crystal with the unit cell parameters of Siegrist et al. [6] We consider four samples, with 0.3, 1, 3, and 5% doping concentration. In these samples, the relative positions and orientations of the doping molecules are relaxed using a molecular mechanics model that allows for rotation and translation of the dopant molecule within the rigid pentacene crystal. [7] As shown in the inset of Figure 1, F 4 TCNQ molecules have a suitable size and shape to substitute pentacene molecules in the crystal structure; efficient substitution is also supported by a scanning tunneling micro- scopy study of this system. [8] This rather perfect structural match between dopant and host is not necessarily the case in other host–dopant combinations or at high doping concentrations. [9] Deviation from this match is expected to introduce structural disorder in the host and will be discussed later. Since the inter- molecular distances within a pentacene monolayer are much smaller than distances between monolayers, associated to the doping charge separation and transport are most likely to happen in the monolayer plane. To depict the energetic land- scape of this plane in a simple graph, two diagonal crystal direc- tions are defined within the monolayer (A and B in the inset of Figure 1), taking a dopant molecule as origin. By following the evolution of the electronic energy of a charge along these two directions, the full range of electrostatic effects induced by the dopant is explored. In the microelectrostatics (ME) approach, the overall effect of the molecular medium on the charge energy is decoupled in a