Visualizing the Link Between Nanomorphology and Energetic Disorder in 3D Organic Solar Cells

arXiv:2507.20040 · physics.app-ph, physics.comp-ph · Submitted 2025-07-26 · Read on arXiv

Listen

Radio episode about this paper

Transcript

Introduction to the show: ident: Physics. The day in general physics: instruments, methods and results across the field.

Kai: I'm Kai.

Mira: And I'm Mira, here with Lev.

Kai: We're looking at the paper: "Visualizing the Link Between Nanomorphology and Energetic Disorder in 3D Organic Solar Cells".

Mira: Here's the gist of the paper. The paper presents results from a three-dimensional hybrid model capable of operating at both high carrier densities and incorporating the effects of energetic disorder in organic bulk heterojunction solar cells The results reveal that even when macroscopic percolation pathways remain intact, energetic disorder limits performance primarily through suppressed charge extraction in interconnected domains This suggests that optimizing molecular packing at the nanoscale is as critical as controlling phase separation at the mesoscale Modeling Morphology and Charge Generation The researchers first generate realistic morphologies using a phase-field approach that accounts for solvent evaporation during film formation This process involves simulating a three-component system consisting of a donor, an acceptor, and a solvent The morphology formation is governed by the Cahn–Hilliard equation which describes the temporal evolution of the domains representing the donor and acceptor materials within the blend Charge Transport Dynamics Charge transport is modeled using drift–diffusion equations solved in quasi-3D to understand configurational disorder The simulation incorporates an AM1.5G solar spectrum and models exciton generation, diffusion, recombination, and splitting Once the photogenerated charge distribution is generated, it is fed into a 3D drift–diffusion model with an exponential tail of trap states Incorporating Energetic Disorder Modeling charge carrier trapping is essential for accurately describing transport in disordered semiconductors such as those based on polymers and small molecules The model breaks up the distribution of trapped states into a series of independent trap levels each with its own independent quasi-Fermi-level Fermi–Dirac statistics are used for trapped carriers, and the charge distribution of trap states is described with an exponential density-of-states Numerical Challenges and Solutions Solving three-dimensional drift–diffusion problems in BHJ organic semiconductor devices presents significant computational challenges These challenges include the non-linearity of the problem, the physical system contains very small numbers and very large numbers, and memory/compute time grows by a factor of at least n cubed as the number of mesh points increases in each dimension The authors overcame these challenges by testing uncoupled 2D slices along one axis followed by a second simulation using orthogonal slices and then averaging the resulting current densities and carrier distributions Key Findings on Optimization The study demonstrates that both nanomorphology and energetic disorder play an important role in achieving a high efficiency The device with the highest photocurrent is the 50:50 blend, while the device with the lowest is the 95:5 blend ratio Achieving maximum efficiency requires a well-connected nanomorphology with low energetic disorder The paper concludes that full 3D modeling is essential for accurate predictions of Voc, jsc, and fill factor Electrical Parameters The simulation parameters include electron mobility in the acceptor at 1 × 10−5 m2V−1s−1 Hole mobility in the acceptor at 1 × 10−10 m2V−1s−1 The effective density of states for electrons is 5 × 10 25 m−3 The electron trap density is 1 × 10 24 m−3 The radiative decay rate kr is 4 × 10 9 s−1 Comparison of Models The comparison between the models shows that going from the 1D effective medium model to 2D to 3D models, the slope of the dark curve reduces, indicating less recombination and current flowing through the device at any given voltage The top panel of Figure 10 can be seen that going from the 1D effective medium model to 2D to 3D models, the slope of the dark curve reduces, indicating less recombination and current flowing through the device at any given voltage The bottom panel shows that Voc, FF and jsc increase as the dimensionality of the models increases This is for the same reasons as given above, mainly that the 3D model has more current paths through the medium and spatially separates the electron and hole distributions The 1D effective medium model assumes perfect spatial overlap of electrons and holes enabling them to easily recombine The 3D model captures both of these effects Visualization and Implementation The authors implemented a custom OpenGL-based visualization tool with a real-time insight into intermediate states such as potential distributions or carrier densities The simulation framework was implemented in C, using native libraries for both Linux and Windows The authors adopted a pragmatic decomposition strategy by performing simulations using uncoupled two-dimensional slices along one axis followed by a second simulation using orthogonal slices The final methodology settled on a hybrid strategy that balanced physical fidelity with computational tractability The benchmarking showed that for a full 3D system with a 10 × 10 × 10 mesh and four electron and four hole trap levels per mesh point, the solver completed a full device simulation in under two minutes on a standard desktop workstation The simulation framework was implemented in C, using native libraries for both Linux and Windows The visualization included dynamic controls such as sliders to traverse slices along each spatial axis enabling users to inspect 2D cross-sections of 3D fields in real time The final methodology settled on a hybrid strategy that balanced physical fidelity with computational tractability The simulation framework was implemented in C, using native libraries for both Linux and Windows The benchmarking showed that for a full 3D system with a 10 × 10 × 10 mesh and four electron and four hole trap levels per mesh point, the solver completed a full device simulation in under two minutes on a standard desktop workstation The simulation framework was implemented in C, using native libraries for both Linux and Windows The visualization included dynamic controls such as sliders to traverse slices along each spatial axis enabling users to inspect 2D cross-sections of 3D fields in real time The final methodology settled on a hybrid strategy that balanced physical fidelity with computational tractability

Paper summary: Mira: The paper is looking at generating realistic morphologies using a phase-field approach that accounts for solvent evaporation during film formation, which is governed by the Cahn–Hilliard equation >

Kai: They then feed those generated morphologies into a three-dimensional drift–diffusion model to understand how charge moves through this disordered landscape >

Lev: That sounds computationally heavy. What's the specific challenge they ran into with solving that kind of problem in three dimensions?

Kai: The challenge is solving these quasi-three-dimensional drift–diffusion problems because they have non-linearity, and the system involves both very small numbers and very large numbers >

Mira: And to handle that computational difficulty, how did the authors manage to get a result that is actually useful for predicting performance metrics like voltage or current?

Kai: They overcame it by testing uncoupled two dimensional slices along one axis first, and then running a second simulation using orthogonal slices before averaging the resulting current densities and carrier distributions >

Lev: So they used this hybrid approach to keep the physical fidelity high while staying computationally tractable for real hardware simulations >

Conclusion: Kai: Looking at this paper, "Visualizing the Link Between Nanomorphology and Energetic Disorder in three dee Organic Solar Cells," it really hammers home that you can't optimize the solar cell just by looking at one aspect > <ref:2507.20040#pg1,Visualizing the Link Between Nanomorphology and Energetic Disorder in 3D Organic Solar>

Mira: It’s about understanding that nanomorphology and energetic disorder both play a critical role in achieving high efficiency, and they show that you need both to be well-connected and low in disorder simultaneously >

Lev: In simple terms for someone who doesn't work in this area, it means you can't just build the best physical structure; you also have to control the quality of the energy states within that structure >

Kai: Right. So, if we only fix one thing—say, the way the donor and acceptor materials separate—we might still fail because of those traps or disorder that impede charge extraction in those domains >

Mira: This work suggests that full three-dimensional modeling is essential for getting accurate predictions for things like open circuit voltage or short circuit current density >

Lev: And they're quite pragmatic about their method, settling on a hybrid strategy because solving the full three dimensions with all the trap levels was just too much overhead on a standard workstation >

Pelin Cáiloglu, Carmen Tretmans, Carsten Deibel, Jan-F. Pietschmann, Martin Stoll, Roderick C. I. MacKenzie

Faculty of Mathematics, Technische Universit¤t Chemnitz · Institute of Mathematics, University of Augsburg · Institut für Physik, Technische Universit¤t Chemnitz · Institute of Mathematics and Centre for Advanced Analytics and Predictive Sciences (CAAPS), University of Augsburg · Department of Engineering, Durham University

physics.app-ph, physics.comp-ph

Submitted: 2025-07-26

Updated: 2025-07-26

Journal ref: Mater. Res. Express 13, 215001 (2026)

DOI: 10.1088/2053-1591/aeb515

Project page: https://www.oghma-nano.com

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Importance score: 83/100

The gist: The paper presents results from a three-dimensional hybrid model capable of operating at both high carrier densities and incorporating the effects of energetic disorder in organic bulk heterojunction…

Key concepts

Phase-Field Approach
This method simulates how materials like donor and acceptor blend together during film formation, accounting for solvent evaporation. It uses the Cahn–Hilliard equation to track the evolving domains of these materials over time, creating realistic nanoscale morphologies.
Drift–Diffusion Equations
These equations are used to model how charge carriers move through a material under an electric field. The quasi-3D version helps simulate configurational disorder, incorporating effects like exciton generation and recombination within the solar cell structure.
Energetic Disorder Modeling
This involves breaking down trapped states into independent levels, each with its own statistics. It uses Fermi–Dirac statistics for trapped carriers to accurately describe how disorder limits charge transport in disordered organic semiconductors.

Terminology

Summary

The paper presents results from a three-dimensional hybrid model capable of operating at both high carrier densities and incorporating the effects of energetic disorder in organic bulk heterojunction solar cells The results reveal that even when macroscopic percolation pathways remain intact, energetic disorder limits performance primarily through suppressed charge extraction in interconnected domains This suggests that optimizing molecular packing at the nanoscale is as critical as controlling phase separation at the mesoscale

Modeling Morphology and Charge Generation

The researchers first generate realistic morphologies using a phase-field approach that accounts for solvent evaporation during film formation This process involves simulating a three-component system consisting of a donor, an acceptor, and a solvent The morphology formation is governed by the Cahn–Hilliard equation which describes the temporal evolution of the domains representing the donor and acceptor materials within the blend

Charge Transport Dynamics

Charge transport is modeled using drift–diffusion equations solved in quasi-3D to understand configurational disorder The simulation incorporates an AM1.5G solar spectrum and models exciton generation, diffusion, recombination, and splitting Once the photogenerated charge distribution is generated, it is fed into a 3D drift–diffusion model with an exponential tail of trap states

Incorporating Energetic Disorder

Modeling charge carrier trapping is essential for accurately describing transport in disordered semiconductors such as those based on polymers and small molecules The model breaks up the distribution of trapped states into a series of independent trap levels each with its own independent quasi-Fermi-level Fermi–Dirac statistics are used for trapped carriers, and the charge distribution of trap states is described with an exponential density-of-states

Numerical Challenges and Solutions

Solving three-dimensional drift–diffusion problems in BHJ organic semiconductor devices presents significant computational challenges These challenges include the non-linearity of the problem, the physical system contains very small numbers and very large numbers, and memory/compute time grows by a factor of at least n cubed as the number of mesh points increases in each dimension The authors overcame these challenges by testing uncoupled 2D slices along one axis followed by a second simulation using orthogonal slices and then averaging the resulting current densities and carrier distributions

Key Findings on Optimization

The study demonstrates that both nanomorphology and energetic disorder play an important role in achieving a high efficiency The device with the highest photocurrent is the 50:50 blend, while the device with the lowest is the 95:5 blend ratio Achieving maximum efficiency requires a well-connected nanomorphology with low energetic disorder The paper concludes that full 3D modeling is essential for accurate predictions of Voc, jsc, and fill factor

Electrical Parameters

The simulation parameters include electron mobility in the acceptor at 1 × 10−5 m2V−1s−1 Hole mobility in the acceptor at 1 × 10−10 m2V−1s−1 The effective density of states for electrons is 5 × 10 25 m−3 The electron trap density is 1 × 10 24 m−3 The radiative decay rate kr is 4 × 10 9 s−1

Comparison of Models

The comparison between the models shows that going from the 1D effective medium model to 2D to 3D models, the slope of the dark curve reduces, indicating less recombination and current flowing through the device at any given voltage The top panel of Figure 10 can be seen that going from the 1D effective medium model to 2D to 3D models, the slope of the dark curve reduces, indicating less recombination and current flowing through the device at any given voltage The bottom panel shows that Voc, FF and jsc increase as the dimensionality of the models increases This is for the same reasons as given above, mainly that the 3D model has more current paths through the medium and spatially separates the electron and hole distributions The 1D effective medium model assumes perfect spatial overlap of electrons and holes enabling them to easily recombine The 3D model captures both of these effects

Visualization and Implementation

The authors implemented a custom OpenGL-based visualization tool with a real-time insight into intermediate states such as potential distributions or carrier densities The simulation framework was implemented in C, using native libraries for both Linux and Windows The authors adopted a pragmatic decomposition strategy by performing simulations using uncoupled two-dimensional slices along one axis followed by a second simulation using orthogonal slices The final methodology settled on a hybrid strategy that balanced physical fidelity with computational tractability The benchmarking showed that for a full 3D system with a 10 × 10 × 10 mesh and four electron and four hole trap levels per mesh point, the solver completed a full device simulation in under two minutes on a standard desktop workstation The simulation framework was implemented in C, using native libraries for both Linux and Windows The visualization included dynamic controls such as sliders to traverse slices along each spatial axis enabling users to inspect 2D cross-sections of 3D fields in real time The final methodology settled on a hybrid strategy that balanced physical fidelity with computational tractability The simulation framework was implemented in C, using native libraries for both Linux and Windows The benchmarking showed that for a full 3D system with a 10 × 10 × 10 mesh and four electron and four hole trap levels per mesh point, the solver completed a full device simulation in under two minutes on a standard desktop workstation The simulation framework was implemented in C, using native libraries for both Linux and Windows The visualization included dynamic controls such as sliders to traverse slices along each spatial axis enabling users to inspect 2D cross-sections of 3D fields in real time The final methodology settled on a hybrid strategy that balanced physical fidelity with computational tractability

Conclusion

Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.

Improvements for AI systems

  1. Incorporate a three-dimensional hybrid model that operates at both high carrier densities and incorporates energetic disorder to separate and visualize how nanomorphology and energetic disorder affect device performance. This improved system can accurately predict device parameters such as short-circuit current (jsc), open-circuit voltage (Voc), and fill factor (FF) by decoupling the influence of configurational and energetic disorder.

  2. Develop a phase-field model that accounts for solvent evaporation during film formation to generate realistic morphologies, which are then used as input into a quasi-3D drift-diffusion model with an exponential tail of trap states. This allows the system to capture both spatial charge separation and 3D percolation pathways, overcoming the limitations of one-dimensional effective medium approximations.

  3. Implement a numerical strategy that solves the 3D problem in uncoupled two-dimensional slices along one axis, followed by a second simulation using orthogonal slices, and then averages the resulting current densities and carrier distributions. This pragmatic decomposition allows for practical runtime on standard workstation hardware while capturing both directional current flows.

  4. Integrate a custom OpenGL-based visualization tool with a Python user interface to provide real-time insight into intermediate states like potential distributions and carrier densities. This enables users to inspect 2D cross-sections of 3D fields in real time, which is crucial for diagnosing convergence behavior and validating boundary conditions during solver development.

  5. Adopt a hybrid architecture that interfaces a single-threaded frontend with PETSc solvers running in MPI mode via shared memory communication using memory-mapped files and POSIX semaphores. This allows the system to leverage fully distributed PETSc solvers for large linear systems while preserving the flexibility and simplicity of a single-threaded interface for rapid experimentation.

  6. Implement a configurable reordering scheme that groups variables associated with a given mesh point—potential, carrier densities, and trap occupations—contiguously in memory before factorization. This aims to reduce matrix bandwidth and potentially lower fill-in during LU factorization, improving performance without sacrificing physical consistency through block-wise normalization.

Sources

Related papers