MAXA to speed up energy transition
Every battery that goes into an EV or a grid storage system starts with getting the cathode material right. If the crystal structure is off, the battery underperforms. Or worse, it degrades fast and ends up as waste. I work on NMC cathode compositions. The kind that go into the batteries powering electric vehicles and renewable energy storage. We were iterating on dozens of them. Every single one needed XRD (X-ray diffraction) to confirm the crystal structure before we could move forward with anything else. If the structure is wrong, nothing downstream matters. The problem was the XRD refinement itself. Open GSAS-II. Load the raw data. Set up the background. Run Le Bail. Refine scale, lattice, profile, atoms, cation mixing, preferred orientation. Converge. Check the fit. Maybe add impurity phases and do it again. That is roughly 10 steps, and every step has parameters you need to get right. A new person on the team? Couple of days just to learn the workflow. Someone experienced? Still two to three hours per sample to get a clean Rietveld refinement. Multiply that by dozens of compositions and you are burning weeks on something that should not be the bottleneck. Weeks spent on refinement is weeks not spent finding the next cathode composition that could make batteries cheaper, longer lasting, or faster to charge. That matters here because better batteries mean more affordable EVs, more reliable grid storage for solar and wind, and fewer reasons for anyone to stay on fossil fuels. So we built a tool at Maatria. We call it MAXA. The idea is simple: you should not need to babysit a 10-step refinement manually when the logic can be codified. Here is how it works: 1. Upload your raw XRD file and select your cathode composition. MAXA runs a 12-stage automated Rietveld refinement pipeline using GSAS-II under the hood. Background, Le Bail, scale, lattice, profile, atomic coordinates, cation mixing, preferred orientation, convergence, and final validation. 2. If the fit quality is poor, it detects residual peaks, suggests possible impurity phases, and lets you decide which to include for multi-phase refinement. 3. You get lattice parameters, c/a ratio, cation mixing percentage, bond distances, Rwp, and a full exportable report. Seconds, not hours.