14/10/2025
In research and development, particularly in material development, challenges arise regarding the time and cost of experiments and securing human resources. In this context, simulation-based approaches using computational chemistry are receiving attention. Specifically, DFT (Density Functional Theory), which calculates electronic states based on quantum mechanics, offers a balance between computational cost and accuracy, and its application in research and development is expanding.
In this article, we will provide a clear explanation of representative software for performing DFT calculations, including selection criteria, compatible visualization tools, and the actual calculation process. Finally, we will briefly introduce a comparison with high-speed calculations using our product, “Matlantis”.
DFT Software Selection Guide
There are various types of software capable of performing DFT calculations, and it is essential to select the appropriate software to achieve the desired computational results. This guide provides an overview of the fundamental considerations for selecting DFT software, including the target materials, desired physical properties, computational environment, and budget.
Target material system
The choice of standard computational methods in DFT calculations depends heavily on the specific material system under investigation. Therefore, it is necessary to determine in advance whether your target material is a "solid system" or a "molecular system". These can be roughly classified as follows.
“Solid systems” include solids such as metals and semiconductors, as well as amorphous materials and solutions with aggregated structures. Assuming the input structure is periodically repeated (periodic boundary conditions) makes calculations for infinitely large systems possible. If you want to calculate the electronic states or physical properties of a collection of atoms or molecules, select DFT software designed for solid systems.
“Molecular systems” refer to molecules such as water (H₂O) or methane (CH₄), as well as molecular clusters formed by their interactions. Although these systems are typically treated in a vacuum, calculations considering solvent effects are also possible. For applications such as calculating the properties of individual molecules or the reactions of homogeneous systems with high precision, molecular system-oriented DFT software is more suitable.
Target physical properties and phenomena
Then, you need to figure out the specific properties you want to calculate. In general, any property related to electronic states can be evaluated through DFT calculations.
It is vital to compare calculated properties with experimental data to ensure their validity. Given that different software packages support varying sets of properties, it is crucial to verify beforehand if your chosen software is compatible with the experimental data you have available.
The table below lists common characteristics and phenomena that can be addressed by DFT calculations, along with representative examples of the results obtained.
Characteristic/Phenomenon for Calculation Example of calculation results
Structural Properties- Lattice constant, Surface structure, Equilibrium geometry, Volume, Density
Electronic Properties- Band structure, Density of states, Molecular orbitals, Partial Charge, Dipole moment
Thermodynamic Propertie- Specific heat, Heat capacity, Boiling point, Melting point, Formation energy, Surface energy, Free energy
Transport Properties- Electrical conductivity, Diffusion coefficient, Thermal conductivity, Viscosity
Response functions and optical properties- Elastic constants, Dielectric constant, Magnetic moment, Phonons, Molecular vibrations, UV-VIS absorption wavelength and intensity
Chemical Reactions- Reaction energy, Activation energy
While in theory, DFT can address many experimental properties linked to electronic states, it's important to recognize that practical application might be limited for certain time and spatial scales. Realistically, the upper bounds for time and spatial scales accessible via DFT calculations are typically on the order of nanoseconds and nanometers, respectively.
Viewer Options
Performing DFT calculations necessitates three-dimensional coordinates of the target system as input. While structures can sometimes be sourced from public databases or existing literature, structural modeling becomes essential if they are not readily available. Furthermore, DFT calculation outputs are typically presented in numerical, text-based formats. Depending on the complexity of the calculation, these outputs can be tens of thousands of lines long. This makes it hard, especially for beginners, to find the relevant information.
To address this problem, "viewers", which are visualization software, are widely employed. Many viewers also incorporate structure modeling functionalities, commonly used for constructing complex molecules or surface adsorption systems. Official viewers, when compatible with the DFT software, can directly retrieve and display information such as energy and charge from the output files. These tools are particularly effective for visualizing data like structures, molecular orbitals, and vibrations, which are hard to interpret from raw numerical data alone.
A variety of viewers exist, and it is crucial to select one that is compatible with your chosen DFT software. Viewers differ in their visualization capabilities, features, usability, support systems, and pricing (paid or free), making it important to review these aspects prior to implementation.
Available Environments and Resources
Before undertaking DFT calculations, it is critical to assess your available computational resources, including your computer hardware and overall computing environment. While calculations of smaller target systems can be performed on standard personal computers, dedicated computing machines are generally preferred for more extensive operations.
However, acquiring a new dedicated computing machine involves substantial upfront costs, the need for physical space, environmental considerations like power and cooling, and the requirement for specialized expertise and personnel for system setup and maintenance.
Moreover, calculations involving a large number of atoms (electrons) or those requiring high precision can significantly increase the computational load, with single calculations potentially taking several weeks. Consequently, limitations in computational resources often restrict the number of trials, posing a significant hurdle in research and development.
In recent years, alternative solutions have emerged to address computational resource constraints, including cloud services for computational environments and AI-driven technologies to accelerate calculations. For example, our cloud-based service, “Matlantis”, eliminates the need for infrastructure setup and utilizes machine learning for high-speed calculations, making it a valuable solution adopted in many research groups.