The field of electronics has been a veritable powerhouse of the economy, driving technological breakthroughs that affect all aspects of everyday life. Aside from silicon, there has been growing interest in developing a novel generation of electronic devices based on pi-conjugated polymers and oligomers. While the goal of these efforts is not to exceed the performance of silicon technologies, these new species may result in greatly reduced fabrication costs as well as completely new functionalities (e.g. mechanical flexibility, transparency, impact resistance). The performance of such organic devices (e.g. field-effect transistors, light-emitting diodes, photovoltaic cells) is heavily dependent on the organization and electronic structures of pi-conjugated molecules or polymer chains at the molecular level. To achieve the full potential of such materials, technological developments require fine-tuning of the relative orientation/position of the pi-conjugated moieties, which provide a practical means to enhance electronic properties. The discovery pace of novel materials can be accelerated considerably by the application and development of efficient computational schemes, where the structure-property relationships could be evaluated in silico. These computational approaches should enable the description of the structural organization, electronic properties as well as the electron dynamics of pi-conjugated frameworks.
Within this context, our proposed workshop will offer the opportunity to (i) gather the key players in the field involving both theoreticians and experimentalists; (ii) assess the validity and range of applicability of current available computational approaches; (iii) identify future major challenges in both the experimental and theoretical areas.

The attractive properties of pi-conjugated molecules make them ideal candidates for functional units in molecular wires, organic solar cells, organic light-emitting diodes and organic field-effect transistors.[1] Similarly, molecular switches, motors and artificial muscles typically rely on pi-conjugation converting an optical or electrochemical signal into a mechanical response.[2]
To achieve these functionalities, the identification of materials, which exhibit enhanced charge carrier (electrons and/or holes) mobility, constitutes a primary objective. Indeed, carrier mobility is one of the key characteristics of organic semiconductors that strongly influence their performance in these electronic applications. The ionization potential (electron affinity) that relates to the hole (electron) injection from HOMO (LUMO) is also determinant to the nature and magnitude of the semiconducting properties. The optical bandgap represents another critical parameter that controls the nature of the signal in light-emitting diodes and the efficiency of light absorption in solar cells. From the computational perspective, the mandatory steps are to (i) describe the relative orientation, interactions, charge carrier mobility (i.e. how fast the carrier travels through the pi-systems) between pi-conjugated moieties as well as their excited states; and (ii) to identify relevant structure-property relationships. As described hereafter, the apparent simplicity of the chemical composition of pi-conjugated systems sharply contrast with the complexity of their electronic structures.
Kohn-Sham density functional theory[3] (DFT) is the most commonly used electronic structure method to describe the properties of relatively large-scale systems, such as pi-conjugated frameworks. For this reason, DFT has often been used to describe the intermolecular interactions between model systems mimicking neutral (i.e. resting state) and charged radical (i.e. charge carrier) oligomers.
The most obvious failures are the overly repulsive DFT binding energies of neutral oligomer assemblies, which are dominated by weak dispersive interactions. Long-range correlation effects are neglected in most popular semi-local (hybrid) approximation, which are unable to stabilize neutral assemblies.[4] Fortunately, these interactions can be conveniently supplemented by a posteriori atom-pair wise dispersion corrections such as those recently introduced.[5] Standard density functionals also fail to properly describe systems carrying fractional charges characteristic of dimers of radical cations. At large intermolecular distances the delocalization (or self-interaction) error artificially stabilizes the delocalization of one positive charge over two molecules as compared to one positive charge and a neutral molecule.[6] Around equilibrium, the errors are smaller, but the description of mixed valence states remains subtle. Highly doped (e.g. doubly charged) π-dimers such as tetracyanoethylene, (TCNE)2_2-[7] present yet another issue arising from static correlation. As a result, the dissociation of the singlet (TCNE)2_2- is impossible without breaking spin symmetry. In addition, the description of such systems is even difficult around equilibrium, due to important degree of multi-reference character.

The recognized DFT failures, such as spurious self-interaction error, are also commonly invoked to explain the time-dependent-DFT deficiency including the poor description of charge transfer excitations,[8] which are relevant to the characterization of the optical properties of these organic materials. Improved schemes involving finely tuned range-separated hybrid functionals[9] or standard functionals using constrained variational DFT[10] have been recently proposed but are awaiting in depth validations.

Taken together, the failures of standard density functionals for dispersion interactions, mixed-valence states and multi-reference character, the prospects for investigating π-functional molecules with common DFT approximations are rather discouraging. In the meantime, the size of the materials of practical interest precludes the application of generally robust, highly accurate ab initio methods to compute binding energies (e.g., coupled cluster, CASPT2 or multi-reference coupled cluster). Over the last decades, advancement in promising alternatives such as the density matrix renormalization group[11] and the reduced-density matrix theory[12] has fostered the development of new paradigm in theoretical and computational chemistry that promises to promote unprecedented growth in our ability to address the most relevant chemical questions such as those challenging the field of organic electronics.
Ab initio methods tend to focus on the description of electronic structure. However, the charge migration that leads to conductivity in conducting polymers relies on charge dynamics. This is also true, for example, in photovoltaic systems where excited state energy transfer determines the efficiency. The computation of the carrier charge dynamics is a rather tedious and time-consuming task, but the rapid advance of the field becomes clear when we consider how realistic computations and simulations are today in comparison to 10 or 20 years ago.[13] Progress is still ongoing and promises to continue in the future. The choice of systems that are as complete and realistic as possible yet constitutes another challenge. For polymers, this should require consideration of the short-range disorder effects using multiscale methods and mesoscopic simulations.[14]
In addition to the various methodological developments and benchmarking of electronic structure approaches achieved over the last decades, numerous experimental and computational research groups have excelled in identifying key structure-property relationships associated with pi-conjugated molecules.[15] In particular, they have addressed underlying issues regarding the stability of alternative topologies to acenes,[16] the physical origin of pi-interactions,[17] the nature of the charge carriers,[18] as well as the introduction of innovative schemes to probe the effect of electron delocalization on molecular properties.[19]

All these approaches and realizations are ideally designed to tackle and resolve some of today’s relevant aspects associated with the properties of molecular precursors to organic electronics. Nevertheless, despite much progress in the last decade, there remain considerable challenges in the computational prediction of properties of organic electronic materials and the rational design of such materials based on first principles. Overall, a lack of understanding of the structure–property relationships of organic semiconductors still hampers their rational design. Below we outline three open challenges central to the field, which will constitute the core of the proposed workshop.