A5034992311- Organic Electronics and Photovoltaics
- Conducting polymers and applications
- Computational Drug Discovery Methods
- Machine Learning in Materials Science
- Organic Light-Emitting Diodes Research
- Protein Structure and Dynamics
- Perovskite Materials and Applications
- Molecular Junctions and Nanostructures
- Electron and X-Ray Spectroscopy Techniques
- Liquid Crystal Research Advancements
- Thin-Film Transistor Technologies
- Organic and Molecular Conductors Research
- Advanced Electron Microscopy Techniques and Applications
- Bioinformatics and Genomic Networks
- Microbial Natural Products and Biosynthesis
- Photochemistry and Electron Transfer Studies
- Metabolomics and Mass Spectrometry Studies
- X-ray Diffraction in Crystallography
- Force Microscopy Techniques and Applications
- Thermal Radiation and Cooling Technologies
- Semiconductor Quantum Structures and Devices
- Enzyme Structure and Function
- Gene expression and cancer classification
- Machine Learning in Bioinformatics
- Semiconductor materials and interfaces
Max Planck Institute for Polymer Research
2012-2022
University of Cambridge
2017-2022
University of Applied Sciences Mainz
2017
Max Planck Society
2012-2015
Imperial College London
2011
Determining the stability of molecules and condensed phases is cornerstone atomistic modelling, underpinning our understanding chemical materials properties transformations. Here we show that a machine learning model, based on local description environments Bayesian statistical learning, provides unified framework to predict atomic-scale properties. It captures quantum mechanical effects governing complex surface reconstructions silicon, predicts different classes with accuracy,...
A key breakthrough in modern electronics was the introduction of band structure engineering, design almost arbitrary electronic potential structures by alloying different semiconductors to continuously tune gap and band-edge energies. Implementation this approach organic has been hindered strong localization states these materials. We show that influence so far largely ignored long-range Coulomb interactions provides a workaround. Photoelectron spectroscopy confirms ionization energies...
This review summarizes the current understanding of electrostatic phenomena in ordered and disordered organic semiconductors, outlines numerical schemes developed for quantitative evaluation induction contributions to ionization potentials electron affinities molecules a solid state, illustrates two applications these techniques: interpretation photoelectron spectroscopy thin films energetics heterointerfaces solar cells.
We investigate the relationship between molecular order and charge-transport parameters of crystalline conjugated polymer poly(3-hexylthiophene) (P3HT), with a particular emphasis on its different polymorphic structures regioregularity. To this end, atomistic dynamics is employed to study an irreversible transition metastable (form I′) stable I) P3HT polymorph, caused by side-chain melting at around 350 K. The predicted backbone–backbone arrangements in unit cells these polymorphs are...
Abstract The functionality of organic semiconductor devices crucially depends on molecular energies, namely the ionisation energy and electron affinity. Ionisation affinity values thin films are, however, sensitive to film morphology composition, making their prediction challenging. In a combined experimental simulation study zinc-phthalocyanine its fluorinated derivatives, we show that changes in as function orientation neat or mixing ratio blends are proportional quadrupole component along...
Organic solar cells rely on the conversion of a Frenkel exciton into free charges via charge-transfer state formed molecular donor–acceptor pair. These states are strongly bound by Coulomb interactions and yet efficiently converted charge-separated states. A microscopic understanding this process, though crucial to functionality any cell, has not been achieved. Here we show how long-range order interfacial mixing generate homogeneous electrostatic forces that can drive charge separation...
Efficiencies of organic solar cells have practically doubled since the development non-fullerene acceptors (NFAs). However, generic chemical design rules for donor-NFA combinations are still needed. Such proposed by analyzing inhomogeneous electrostatic fields at donor-acceptor interface. It is shown that an acceptor-donor-acceptor molecular architecture, and alignment parallel to interface, results in energy level bending destabilizes charge transfer state, thus promoting its dissociation...
We establish a link between the microscopic ordering and charge-transport parameters for highly crystalline polymeric organic semiconductor, poly(2,5-bis(3-tetradecylthiophen-2-yl)thieno[3,2-b]thiophene) (PBTTT). find that nematic dynamic order of conjugated backbones, as well their separation, evolve linearly with temperature, while side-chain parameter backbone paracrystallinity change abruptly upon (also experimentally observed) melting side chains around 400 K. The distribution site...
We develop a generic coarse-grained model for describing liquid crystalline ordering of polymeric semiconductors on mesoscopic scales, using poly(3-hexylthiophene) (P3HT) as test system. The bonded interactions are obtained by Boltzmann-inverting the distributions degrees freedom resulting from canonical sampling an atomistic chain in Θ-solvent conditions. nonbonded given soft anisotropic potentials, representing combined effects π–π and entropic repulsion side chains. demonstrate that can...
We present a method for evaluating electrostatic and polarization energies of localized charge, charge transfer state, or exciton embedded in neutral molecular environment. The approach extends the Ewald summation technique to effects, rigorously accounts long-range nature charge-quadrupole interactions, addresses aperiodic embedding charged cluster its cloud periodic illustrate by density states ionization thin films heterostructures organic semiconductors. By accounting mesoscale fields,...
Cryo-Electron Microscopy (cryo-EM) is a pivotal tool for determining the 3D structures of biological macromolecules. Current cryo-EM workflows, while effective, are computationally demanding and require manual intervention, creating bottlenecks use in high-throughput scenarios such as structure-based drug discovery. Often discovery, one can assume that all instances protein equivalent at resolutions needed alignment it therefore should be possible to harness information about particle poses...
We develop a stochastic network model for charge transport simulations in amorphous organic semiconductors, which generalizes the correlated Gaussian disorder to realistic morphologies, transfer rates, and site energies. The includes an iterative dominance-competition positioning vertices (hopping sites) space, distance-dependent distributions vertex connectivity electronic coupling elements, moving-average procedure assigning spatially field dependence of hole mobility semiconductor,...
A multiscale simulation scheme, which incorporates both long‐range conformational disorder and local molecular ordering, is proposed for predicting large‐scale morphologies charge transport properties of polymeric semiconductors. Using poly(3‐hexylthiophene) as an example, it illustrated how the energy landscape its spatial correlations evolve with increasing degree structural order in mesophases amorphous, uniaxial, biaxial nematic ordering. It shown that formation low‐lying states more...
We present a multi-scale model for charge transport across grain boundaries in molecular electronic materials that incorporates packing disorder, electrostatic and polarisation effects.
Abstract Organic photovoltaics (PV) is an energy-harvesting technology that offers many advantages, such as flexibility, low weight and cost, well environmentally benign materials manufacturing techniques. Despite growth of power conversion efficiencies to around 19 % in the last years, organic PVs still lag behind inorganic PV technologies, mainly due high losses open-circuit voltage. Understanding improving open circuit voltage solar cells challenging, it controlled by properties a...
In developed countries 60% of the electricity consumed is attributable to commercial and public buildings. Even in UK, solar energy incident on buildings more than 7× electrical they consume. This represents a problem (the management heat gain glare) but also an opportunity that may be taken advantage using complementary concentrator technologies. We are investigating conventional geometric luminescent concentrators combined optimally harvest direct diffuse components sunlight within double...
Machine learning (ML) is widely used in drug discovery to train models that predict protein-ligand binding. These are of great value medicinal chemists, particular if they provide case-specific insight into the physical interactions drive binding process. In this study we derive ML from over 50 fragment-screening campaigns introduce two important elements believe absent most -- not all studies type reported date: First, alongside observed hits use our models, incorporate true misses and show...
Predicting bioactivity and physical properties of molecules is a longstanding challenge in drug design. Most approaches use molecular descriptors based on 2D representation as graph atoms bonds, abstracting away the shape. A difficulty accounting for 3D shape designing can precisely capture while remaining invariant to rotations/translations. We describe novel alignment-free QSAR method using Smooth Overlap Atomic Positions (SOAP), well-established formalism developed interpolating potential...
Abstract We introduce a machine-learning (ML) framework for high-throughput benchmarking of diverse representations chemical systems against datasets materials and molecules. The guiding principle underlying the approach is to evaluate raw descriptor performance by limiting model complexity simple regression schemes while enforcing best ML practices, allowing unbiased hyperparameter optimization, assessing learning progress through curves along series synchronized train-test splits....
Generative models for structure-based molecular design hold significant promise drug discovery, with the potential to speed up hit-to-lead development cycle, while improving quality of candidates and reducing costs. Data sparsity bias are, however, two main roadblocks 3D-aware models. Here we propose a first-in-kind training protocol based on multi-level contrastive learning improved control data efficiency. The framework leverages large resources available 2D generative modelling datasets...
Cryo-EM is a powerful tool for understanding macromolecular structures, yet current methods structure reconstruction are slow and computationally demanding. To accelerate research on pose estimation, we present CESPED, data set specifically designed supervised estimation in cryo-EM. Alongside provide package to simplify cryo-EM handling model evaluation. We evaluate the performance of baseline model, Image2Sphere, which shows promising results but also highlights need further improvements....
Abstract Binding sites are the key interfaces that determine a protein’s biological activity, and therefore common targets for therapeutic intervention. Techniques help us detect, compare contextualise binding hence of immense interest to drug discovery. Here we present an approach integrates protein language models with 3D tesselation technique derive rich versatile representations combine functional, structural evolutionary information unprecedented detail. We demonstrate associated...