MXene Atlas — an evidence-linked map of MXene propertiesv2
What has been built so far— each round: what it measured and how it ended · «undecidable» is not failure but «to be filled when data arrives»
Values extracted from open-access experimental papers · one dot = one sample under one condition · click a dot to see the source sentence · measurement conditions differ, so do not compare values directly
⚠️ Not “verified” yet — values were checked against the source by two AI judges only (no human check yet) · Help us check — click a point and answer «did we copy the number from the sentence correctly?» with ✅ / ❌ (no need to read the paper). The “strict” tab filters on machine-extracted fact fields, and the accuracy of that filter has not been measured (the test stopped as undecidable for lack of samples · Oct 8). Etchant question — cations from fluoride-salt etching (NH₄HF₂·NH₄F·NaF·KF) are treated as part of etching, like LiF/HCl samples, and stay in the strict tab · TMAOH is a deliberate intercalant and is excluded · basis = Gogotsi-lab reviews and protocol papers (to be revisited when outside experts reply)
Structure view— ① click an element in the periodic table to filter → ② in the network (3D · drag to rotate · wheel to zoom) click a paper dot to see its values and source → ③ click a property dot to jump to “Find by target value” below · follows the sample tab and etchant-question answer below
Try these
Loading the network… (needs internet — the 3D library comes from a public CDN)
● Blue = which MXene (composition · e.g. Ti₃C₂ · V₂C)● Orange = which etchant (etching · e.g. HF · LiF/HCl)● Pink = how layers were separated (delamination · capacitance data only · skipped when not stated)● Grey = sample form (e.g. free-standing film · film on substrate · electrode · powder)● Green = what was measured (property, by unit · e.g. conductivity S/cm · capacitance F/g · shielding dB)● Small purple dot = one source paper (hover for its ID · click for source sentences)Thicker lines = more values connecting the two · How to read: paths run composition → etchant → delamination → form → property · a thick “Ti₃C₂ — LiF/HCl — free-standing film — S/cm” path means many conductivity values for LiF/HCl-etched Ti₃C₂ made into free-standing films
🎚 How the filters work
Sample — Strict = pure MXene only · All = including mixed materials
Verification — “Passed only” keeps just the checked values
Unit — choose the property to see (“being verified” = for reference only)
Cation question — change how samples etched with fluoride salts are counted
● supported · ○ not judged · ✕ judged wrong (kept visible) · ▲ off the chart
Sample
Verification
Mixed withauto-classified · 35 of 40 correct in a fresh sample (13 Oct)
UnitConductivityEMI shielding (few samples · precision gate not passed)Battery (being verified)Interlayer spacing (being verified)Surface area (being verified)Photothermal (being verified)Work function (being verified)Thermal conductivity (being verified)
Treat etchant cations as intercalants?
● Supported (two judges agree or source checked)○ Not judged✕ Judged wrong (kept visible)▲ Off-axis (click to check · e.g. single-flake local value converted per mass)● Faint dot = not yet verified (from a round that did not pass · small sample · flagged for re-check)
Click a point to see the source sentence · conditions · paper here · then you can answer «did we copy the number correctly?» with ✅ / ❌ below it.
Ask — if you need this property, which MXene was made, and how?
How to use: ① click an example below or ② type a question (e.g. «high conductivity MXene» · «large surface area» · «low work function») → it recognises the property and «high/low», gathers samples in the top 10% (bottom 10% for «low») of that property, and counts which etchant and which form they were made with. Version 2 (10/9): also understands conditions — etchant (LiF/HCl · HF · fluoride salt · molten salt) · «non-Ti» · a composition (e.g. Nb2C) · «film». v3 (10/9 afternoon): conditions can be combined — element names (vanadium · niobium · molybdenum …) · delamination (TMAOH · sonication · shaking) · «pure» (no mixed material) · form (powder · aerogel · fiber · free-standing film) · add «compare» to get a table of values, papers and median per etching route.
Examples
The answer appears here — it is a list of source candidates («samples made this way reached this value»), not a guarantee that «making it this way gives this value».
Find by target value — how were the samples that reached it made?
How to use: ① pick a property in the «Unit» row above → ② type the value you need → ③ «Find». Samples that reached it are grouped by etchant; click a line to see the source sentence.
at leastFollows the sample tab, unit and etchant-question answer chosen above · conditions differ, so this does not promise “this method gives this value” — it is a list of candidates to check in the source
Enter a target value and press “Find” to list the samples that reached it, grouped by etchant.
Same recipe, different values — how much do values differ between papers for samples with the same composition, etchant, delamination and form?
Uses the unit and sample tab above · one vote per paper (its median) · values judged wrong ✕ excluded · groups with ≥2 papers only · largest spread first · click a row to compare the highest and lowest source sentences side by side — the cause may be conditions not in these fields (thickness, method, post-treatment, oxidation) · this shows where to look, not the cause
Show as table
⚠️ The default sample tab (strict · with etch ions · all) is not decided yet (a domain expert’s call) · not “verified”: machine judging only · no human check yet Basis (all in intelligence/mxene-atlas/) = RESULT-P2-pilot (capacitance) · RESULT-P3-1 (volumetric) · RESULT-P3-2e (conductivity) · RESULT-P3-3 (shielding) · REVISIT (assumption ledger)