Periodic Table

The grid is a map of electron configuration

The periodic table arranges the chemical elements in order of atomic number, in rows called periods and columns called groups, and the shape of that grid follows their electron configurations. One step right adds a proton and an electron; across the s- and p-blocks that electron joins the same outer shell, while in the d- and f-blocks it fills an inner subshell. One step down starts a new shell. A group therefore collects elements with the same number of valence electrons, and since bonding is almost entirely a valence-shell affair, that shared count is why lithium, sodium and potassium behave alike, and why fluorine and chlorine both want exactly one more electron.

Block widths are subshell capacities: s holds 2 electrons, p holds 6, d holds 10 and f holds 14. Hence 2 columns on the left, 6 on the right, 10 across the transition metals and 14 for the f-block. Period lengths follow: 2, 8, 8, 18, 18, 32, 32.

The f-block sitting underneath is a printing decision, not chemistry. Those elements belong inside periods 6 and 7, between barium and hafnium and between radium and rutherfordium. Put them there and the table runs 32 columns wide, too wide to display legibly, so they are cut out and set below. The two strips are 15 wide, lanthanum to lutetium and actinium to lawrencium, one more than the f-block holds, because tables differ on whether lanthanum or lutetium belongs in group 3. This one leaves that open, so the group 3 cells under yttrium are empty. The ordering is untouched: hafnium still follows lutetium.

Why some cells carry a mass number instead

Almost every mass here is a standard atomic weight: an average over the isotopes as they occur naturally on Earth, weighted by abundance. Chlorine’s 35.45 is not the mass of any chlorine atom. It is roughly three parts chlorine-35 to one chlorine-37. Figures here are IUPAC CIAAW abridged values, to five significant figures.

That average exists only where an element has a settled natural composition. Technetium and promethium have no isotope long-lived enough to have survived since the Earth formed and no settled natural mixture to average, and the same holds from polonium upwards. Thorium, protactinium and uranium are the exceptions: they occur naturally with a characteristic composition and keep atomic weights of 232.04, 231.04 and 238.03. In all, 84 elements have a standard atomic weight, and 34 do not.

Those 34 show the mass number of the longest-lived known isotope instead, or of one of them where CIAAW lists several. It is flagged so it cannot be read as a weight, and the tell is the whole number: polonium 209, radon 222, plutonium 244. A mass number counts protons plus neutrons in one nuclide, so it is exact and involves no averaging. Use one as a molar mass and the answer holds only for that isotope.

Reading a trend off the grid

Across a period the nuclear charge rises one proton at a time while the added electrons all enter the same shell, and same-shell electrons shield one another poorly. The pull on the valence electrons therefore strengthens from left to right: atoms contract, ionisation energy climbs and electronegativity climbs with it. Fluorine ends up the most electronegative element for that reason.

Down a group the opposite dominates. Each period adds a shell, so the valence electrons sit further out behind more shielding, and the extra protons cannot compensate. Radius grows and ionisation energy falls. Caesium and lithium both have one valence electron, but caesium’s is the loosest held, and caesium is the far more violent reducing agent.

Mass does not track position cleanly:

• Argon 39.95 outweighs potassium 39.098, though potassium has the extra proton.
• Cobalt 58.933 outweighs nickel 58.693.
• Tellurium 127.6 outweighs iodine 126.9.

Each is an abundance effect. Argon is almost entirely argon-40 while potassium is mostly potassium-39; iodine is a single isotope at 127 while tellurium leans on its heavier ones. Mendeleev ordered by mass and broke his own rule at tellurium and iodine to keep the chemistry right, and ordering by atomic number removes the conflict.

Where the table becomes prediction

Past lawrencium at 103 the positions are certain and the chemistry largely is not. Everything from rutherfordium up is made an atom at a time by fusing lighter nuclei in an accelerator, and the heaviest decay in milliseconds. Oganesson has been detected a handful of times in total. There is no sample to weigh and, for most, no compound anyone has held.

Group extrapolation is least trustworthy where it is leaned on hardest. In very heavy atoms the inner electrons move fast enough for relativistic effects to matter: s orbitals contract and are stabilised, already visible in mercury being liquid at room temperature and in gold’s colour. At copernicium and beyond the corrections are large enough to reorder what a group trend predicts.

Oganesson is the clearest case. It sits under radon in group 18, but calculations put it as a solid at room temperature with a highly polarisable electron cloud, so reactive rather than inert. Treat the last two rows as positions with predicted properties attached.

Periodic trends, property by property

Each of these properties has a page of its own, drawn as a colour-coded periodic table from PubChem’s data, with the elements ranked and the exceptions to each trend picked out of the data:

• Why the trends run the way they do
• Electronegativity
• Atomic radius
• First ionisation energy
• Electron affinity
• Melting point
• Boiling point
• Density