Where the data comes from
One open dataset. Nothing invented.
Every number on this site — masses, melting points, electron shells, all of it — comes from one openly licensed dataset, fetched by an open script, shown as published apart from two disclosed corrections of known defects.
The pipeline
1 · Fetch
An open script downloads the published dataset directly from its source. If the source is down, the script fails — it never invents or fills in a value.
2 · Check & trim
The script verifies all 118 confirmed elements are present with the fields this site shows, keeps only those fields, and drops hypothetical elements beyond 118. Hypothetical elements beyond 118 are not shown: 119 Ununennium (predicted, never observed). It also fixes two known defects in the source, and says exactly which: Two known defects in the source are corrected, nothing else: atomic masses published with their uncertainty digit appended (helium as 4.0026022 for IUPAC 4.002602(2), iron as 55.8452 for 55.845(2)) have that digit removed, restoring the IUPAC standard atomic weight; and nihonium, filed as "probably transition metal", is labeled "probably post-transition metal" like the rest of group 13.
3 · Snapshot
The result is stored with its provenance — source URL, license, retrieval date (2026-08-21) — beside every number, and this page links straight back to the original.
The source
Periodic-Table-JSON (Bowserinator and contributors); element summaries derived from Wikipedia, published under CC BY-SA 3.0. Element summaries are drawn from Wikipedia articles (each element page links to its original). 118 elements in this snapshot.
Units, as published: Units as published by the dataset: temperatures in kelvin (K); density in g/cm³ for solids and liquids, g/L for gases; atomic mass in unified atomic mass units (u); ionization energies and electron affinity in kJ/mol.
The compounds source
The 52 compounds come from PubChem, National Center for Biotechnology Information, U.S. National Library of Medicine — Public domain (PubChem, U.S. National Institutes of Health), which permits any reuse. A second pipeline script fetches them by their PubChem identifiers; molecular weights and dissociation constants (pKa) are stored verbatim as published, and the only computation is parsing each formula into element counts (checked against the 118 elements).
Curation, disclosed: This is a curated selection of everyday compounds, not a database: which compounds appear, their everyday names, tags, and "commonly met as" acid/base roles are editorial choices. Every number and name beside them comes from PubChem verbatim. Snapshot retrieved 2026-08-21.
Is it okay that this comes from Wikipedia?
Yes — that's what the license is for. Wikipedia's text and this dataset are published under Creative Commons BY-SA, which explicitly allows anyone to reuse them, including in products, on two conditions: credit the source (done — this page, and every element page, links straight back) and share the text under the same license — so the element summaries you read here remain CC BY-SA, as they should.
Worth being honest about what that means: the summaries are encyclopedic writing — community-written and community-reviewed — while the numbers (masses, melting points, shells) are physical constants the dataset compiled from reference sources. For a future version, primary scientific databases (NIST, IUPAC, PubChem) are the upgrade path for the numbers; the pipeline was built so a new source slots in without changing the app.
What we won't do
- No invented values. Where the dataset has no number (some densities, melting points, and the chemistry of the newest elements), the page shows a dash and says so.
- Predicted ≠ confirmed. Elements whose chemistry has never been observed are labeled that way — and hypothetical elements beyond 118 aren't shown at all.
- Depth, not dumbing down. The Simple / Everyday / Student / Deep dive tabs on every element page tell the same facts at different depths — none of them a different truth.