Materials Square
Run DFT, molecular dynamics, and CALPHAD from a browser — no Linux, no compute environment to set up. Bring simulation to your own research from day one.
- Run calculations in the cloud with no install or build step
- Use a GUI for every step — no Linux or scripting needed
- DFT · MD · CALPHAD in one platform
- 10+ years of use in research, education, and industry
See Materials Square in motion
A short demo walk-through of setting up, running, and analyzing a calculation in the browser.
Proven in the field
- 30,000+
- Registered users
- 100+
- Countries reached
- 10+
- Years in service
- 40+
- Universities & labs
Core capabilities
Configure, run, and analyze calculations from a single web UI — no separate Linux environment required.
DFT simulation
Automated workflows on Quantum ESPRESSO. Configure calculations from a GUI and templates — no hand-written input decks.
Molecular dynamics
Run LAMMPS-based MD on the same platform and chain results into downstream analysis automatically.
CALPHAD
OpenCALPHAD-based phase-equilibrium calculations from the same interface — alloy design and thermodynamic assessment, closer at hand.
No Linux required
Environment setup, terminal commands, and scheduler tuning all disappear. Run everything from clicks and forms.
Elastic cloud compute
Spin up hundreds of cores when you need them. Skip the cluster install and scheduling work, and focus on results.
Visualization & analysis
Built-in viewers for structures, electronic structure, DOS, and bands — interpret results without leaving the platform.
What gets calculated, and how
Three families of calculation share one interface. Which one you reach for depends on the length and time scale the property you want lives at.
DFT — properties from the electronic structure
Density functional theory solves for electron behaviour directly, giving lattice constants, formation energies, band structures, densities of states and surface adsorption energies. Unlike methods that lean on measured data, it applies to compositions with no experimental values and to structures nobody has synthesised yet. The backend is Quantum ESPRESSO, and instead of writing input cards you adjust parameters in a template built for the calculation you want. Cutoff energy, k-point mesh and pseudopotential choice drive the result, so they carry defaults but stay on screen — hidden, they leave you unable to explain later where a number came from.
Molecular dynamics — systems that move in time
MD integrates the equations of motion under a force field, reaching the tens to hundreds of thousands of atoms and the nanosecond timescales DFT cannot. Diffusion coefficients, thermal conductivity, elastic and mechanical properties, and polymer chain behaviour all need that statistical average to mean anything. The backend is LAMMPS, and finished trajectories flow straight into the analysis tools on the same platform.
CALPHAD — phase equilibria and alloy design
CALPHAD computes phase equilibria across composition and temperature from thermodynamic databases. In alloy design it narrows down which phases appear where, and which composition windows are single-phase, before any experiment is run. It is built on OpenCALPHAD, and formation energies from DFT can feed the thermodynamic assessment inside the same account.
Choosing between them
A few hundred atoms and an electronic-structure answer means DFT; tens of thousands of atoms and time-dependent behaviour means MD; a map over composition and temperature means CALPHAD. Real projects move between all three, which is why keeping them on one platform is faster than installing and learning three separate toolchains.
How far to trust a calculated number
A simulated value means nothing as a number unless the method and settings come with it. Convergence testing — raising the cutoff and k-point density until the property of interest stops moving — a check against a similar system with known measurements, and a record of the functional and pseudopotentials used, are what let somebody else reproduce and cite the result. That is why the platform keeps the settings of every calculation alongside its output in your account.
How it works
Three steps from sign-up to results — no infrastructure setup needed.
-
01
Sign up and load a structure
Register by email, then upload a crystal structure or pull one from the built-in database.
-
02
Configure and run
Pick a calculation template, adjust the parameters, and submit to the cloud with one click.
-
03
Analyze and continue
Visualize results in the built-in viewer, download data, or chain into the next calculation.
Where teams use it
Battery & energy materials
Screen electrode materials, map ion diffusion paths, and analyze electrolyte reactions.
Catalysis & surface science
Surface adsorption energies, NEB reaction paths, and identification of active sites.
Semiconductors & devices
Defect formation energies, electronic structure analysis, and quantitative doping effects.
Undergraduate & graduate teaching
Computational chemistry and materials courses — instantly usable by a full class, no install burden.
Materials Square — frequently asked questions
What calculations can I run on Materials Square?
DFT with Quantum ESPRESSO, molecular dynamics with LAMMPS, and CALPHAD phase equilibria with OpenCALPHAD — all from the same web interface, from structure preparation through electronic structure, DOS and band visualisation.
Do I need to know Linux or how to set up a compute environment?
No. Environment setup, terminal commands and scheduler configuration all sit on the platform side. You load a structure in the browser and adjust the parameters of a calculation template.
Is there anything to install, or a server to prepare?
No. A web browser is enough. Calculations run in the cloud, so there is no workstation or cluster to prepare, and hundreds of cores can be brought in the moment you need them.
How is it priced?
Signing up is free, and you pay only for the cloud computing resources your simulations actually consume — no idle hardware to carry. Team and institutional terms are arranged through the contact page.
Can it be used for university teaching?
Yes. It is used in computational chemistry and materials engineering lab courses. Students install nothing, so there is no class setup burden, and seats can be provisioned per head.
Which research fields use it most?
Battery and energy materials (electrode screening, ion diffusion), catalysis and surface science (adsorption energies, NEB reaction paths), and semiconductors and devices (defect formation energies, doping effects).
The fastest way to start is to run one
You can run a demo calculation on a free account today. Team and institutional onboarding conversations are always open.