Free tool
dQ/dV analysis (incremental capacity)
Paste or upload voltage and capacity data from a slow charge or discharge. The tool resamples capacity onto an even voltage grid, smooths it, differentiates, and marks the peaks, for up to six curves at once so you can compare a fresh cell with an aged one.
Runs entirely in your browser. Nothing is uploaded. Sample curves are synthetic.
What dQ/dV peaks mean
Incremental capacity analysis plots the capacity gained per volt, dQ/dV, against voltage. Where the voltage curve is flat, a lot of capacity is stored over a small voltage change, so it appears as a peak. Those plateaus are associated with phase transitions or staging in the electrode materials, which makes each peak a fingerprint of a particular process in the cell.
Comparing curves over the life of a cell is the main use. Peaks that shrink or shift in voltage are commonly linked to loss of lithium inventory or loss of active material, although higher resistance and differences in current or temperature can look similar. Read the shifts as a pointer for further checks, not as a diagnosis on their own.
Smoothing and voltage step
Differentiation amplifies noise, so raw dQ/dV is rarely usable. This tool first interpolates capacity onto a uniform voltage grid with the step you choose, then smooths capacity with a Savitzky-Golay filter, a local polynomial fit that keeps peak position and height better than a plain moving average, and finally takes central differences. A very small step makes the curve noisier; a very large one blurs narrow peaks. A window that is too wide merges neighbouring peaks, so adjust it while watching the peak table.
Use low C-rate data, ideally C/20 or slower. At higher rates polarization smears and shifts the peaks, and the curve stops reflecting the electrode processes. For the impedance side of cell diagnostics, see the EIS Nyquist plotter.
FAQ
What data format does the tool need?
Two columns: voltage in V and capacity in Ah (or mAh if the header contains mAh), from a single charge or discharge. CSV, semicolon or tab separated files work, and comma decimals are accepted with semicolons or tabs. Headers are detected by name; otherwise the first two columns are used as voltage and capacity. If the voltage is not monotonic, the longest run in the chosen direction is used.
How should I choose the smoothing?
Start with Savitzky-Golay on capacity, a window of about 11 points on a 5 mV grid and polynomial order 2. Widen the window until noise peaks disappear, but stop before real peaks flatten or merge. A smaller voltage step needs a larger window in points for the same smoothing in volts. Compare the raw curve with no smoothing to check you are not removing real features.
Should I use the charge or the discharge segment?
Either works, and the choice mostly depends on which segment your test controls best. Use the same direction, temperature and current for every curve you compare. Charge and discharge peaks sit at slightly different voltages because of polarization, so do not mix them in one comparison.
What do peak shifts and shrinking peaks mean?
Peaks relate to phase transitions or staging in the electrodes. As a cell ages, peaks commonly shrink or move, which is often associated with loss of lithium inventory or loss of active material, but the same changes can also come from rising resistance or a different test current or temperature. Treat dQ/dV as a diagnostic hint and confirm it with other measurements.
Is my data uploaded?
No. Everything runs in your browser and nothing leaves your computer.