Classification of Elements: From Dobereiner to the Modern Periodic Table

In 1800, chemists knew of just 31 elements. By 1865, that number had more than doubled to 63, and today 114 are recognised, several of them entirely artificial, synthesised rather than found in nature. Studying every one of these elements and their countless compounds individually, with no organising structure at all, would be an almost impossible task. Scientists needed a systematic way to organise their growing knowledge, one that would not just tidy up known facts but genuinely predict new ones, pointing chemists toward elements and compounds that hadn't even been discovered yet. That need is what eventually produced the periodic table, arguably the single most useful organising idea in the whole of chemistry.

The German chemist Johann Dobereiner made the earliest documented attempt, noting by 1829 that several groups of three elements, triads, shared similar physical and chemical properties, with the middle element's atomic weight sitting almost exactly halfway between the other two, lithium, sodium and potassium being one such triad. Since it only worked for a handful of elements, it was dismissed largely as coincidence. English chemist John Newlands went further in 1865, arranging elements by increasing atomic weight and noticing that every eighth element shared properties with the first, exactly like the eighth note in a musical octave repeating the first, an idea he called the Law of Octaves. It held up neatly only as far as calcium, and was not widely accepted at the time, though Newlands was eventually recognised for it, awarded the Davy Medal in 1887, twenty-two years after he first proposed it.

Working independently, Dmitri Mendeleev and Lothar Meyer both proposed in 1869 that arranging elements by increasing atomic weight reveals properties repeating at regular intervals. Mendeleev is generally credited with the breakthrough, since his classification went further than Meyer's, using a broader range of chemical and physical properties, and where atomic weight ordering conflicted with observed similarities, such as iodine and tellurium, he trusted the pattern of properties over the raw weight measurement, reasoning the weight data itself might simply be wrong. Where his table had genuine gaps, he didn't force elements into ill-fitting boxes; he left them empty, naming the missing elements Eka-Aluminium and Eka-Silicon, and confidently predicted their density, melting point, and oxide and chloride formulas. When gallium and germanium were discovered years later, filling exactly those gaps, their measured properties matched Mendeleev's predictions astonishingly closely, cementing his Periodic Law: the properties of elements are a periodic function of their atomic weights.

Mendeleev built his entire table without knowing anything about atomic structure, since protons and electrons hadn't been discovered yet. In 1913, physicist Henry Moseley studied the characteristic X-ray spectra of elements and found that plotting the square root of X-ray frequency against atomic number gave a perfectly straight line, while plotting it against atomic weight did not. This showed decisively that atomic number, the count of protons, is the more fundamental organising property, not atomic weight. The Periodic Law was updated accordingly into its modern form: the physical and chemical properties of elements are periodic functions of their atomic numbers, not their atomic weights. This single correction also resolved every earlier anomaly Mendeleev had to argue around by instinct, like placing iodine before the heavier tellurium, since ordering by atomic number puts every element exactly where its actual properties say it belongs.

The modern long-form periodic table arranges elements into seven horizontal periods and eighteen vertical groups, replacing Mendeleev's older, more confusing IA through VIIA, VIII, IB through VIIB and 0 labelling with a clean, sequential 1 to 18 numbering, per IUPAC's recommendation. Each period's number corresponds directly to the highest principal quantum number occupied by that period's elements: period 1 holds just 2 elements (filling only 1s), period 2 and 3 hold 8 each, periods 4 and 5 hold 18 each, and periods 6 and 7 hold 32 each, in principle, though period 7 remains incomplete. To keep the table's structure manageable, the 14-element-wide lanthanoid and actinoid series from periods 6 and 7 are conventionally pulled out and displayed as two separate rows beneath the main table, rather than being squeezed into it directly.

The right to name a new element traditionally belonged to its discoverer, subject to IUPAC ratification, but for extremely heavy, unstable, artificially synthesised elements, produced sometimes only a few atoms at a time, this occasionally led to genuine international disputes, most famously between American and Soviet scientists both claiming credit for element 104, one side naming it Rutherfordium, the other Kurchatovium. To sidestep this, IUPAC adopted a systematic temporary naming scheme built directly from an element's atomic number, using fixed numerical roots (0=nil, 1=un, 2=bi, 3=tri, and so on) strung together in digit order with the suffix '-ium' added at the end. Element 104, for instance, becomes 'unnilquadium' (un-nil-quad-ium) under this scheme until its permanent name is officially settled by an IUPAC vote, at which point it can honour a country, place, or scientist, exactly how element 106 eventually became seaborgium, honouring Glenn Seaborg's work discovering the transuranium elements.

Each period's length follows directly and predictably from the Aufbau filling order established for electron configurations. Period 1 fills only the 1s orbital, giving exactly 2 elements, hydrogen and helium. Period 2 fills 2s then 2p, 2 plus 6 orbitals' worth of electrons, giving 8 elements from lithium to neon. Period 4 is where things get more interesting: after filling 4s, filling the 3d orbitals becomes energetically favourable before 4p, kicking off the 3d transition series starting at scandium (3d1 4s2) and completing at zinc (3d10 4s2), before 4p finally fills and the period closes at krypton, 18 elements total. Period 6 layers in a further complication, since after 6s fills, the 4f orbitals fill next, starting at cerium and ending at lutetium, the lanthanoid series, before 5d and 6p complete the period at 32 elements.

Elements sharing a vertical group share something specific and consequential: an identical valence-shell electron configuration pattern, differing only in principal quantum number. Every alkali metal, lithium, sodium, potassium, rubidium, caesium, francium, has exactly the ns1 configuration in its outermost shell ([He]2s1, [Ne]3s1, [Ar]4s1, and so on), and this shared arrangement is precisely why they all react similarly, readily losing that single outer electron. This is the deep reason group membership matters more than raw atomic number alone: chemical behaviour tracks valence electron arrangement, not simply how many total electrons an atom carries, which is also why the Periodic Law depends fundamentally on atomic number and its resulting electron configuration, not on atomic mass, resolving the last of Mendeleev's own occasional inconsistencies.

The Aufbau principle divides the entire periodic table naturally into four blocks, named for whichever subshell receives an element's last electron. The s-block (groups 1 and 2) covers reactive metals with ns1 or ns2 configurations. The p-block (groups 13 to 18) spans ns2np1 through ns2np6, ending each period with a noble gas. The d-block (groups 3 to 12), the transition elements, fills (n-1)d orbitals and sits between the highly reactive s-block and the less reactive p-block, forming a genuine chemical bridge between them. The f-block, the lanthanoids and actinoids, fills (n-2)f orbitals and is displayed separately below the main table. Two elements resist tidy categorisation: helium, though structurally s-block by its 1s2 configuration, is placed in the p-block's noble gas column since it behaves exactly like one, a completely filled valence shell; and hydrogen, with just one lone s-electron, could arguably sit with the alkali metals or, since it can also gain one electron to reach a noble gas configuration, with the halogens, so it's conventionally placed on its own at the very top of the table.

Beyond the four electronic blocks, the periodic table also splits along a broader, purely observable line: metals versus non-metals. Metals, over 78% of all known elements, occupy the left and centre of the table, are typically solid at room temperature (mercury is a famous exception), have high melting and boiling points, conduct heat and electricity well, and are malleable and ductile. Non-metals cluster at the upper right, are usually solids or gases with low melting points (boron and carbon are notable exceptions), and are poor conductors, often brittle where solid. The transition from metallic to non-metallic character isn't a clean, sudden cutoff; it happens gradually along a thick, zig-zagging diagonal line running through the table, and elements bordering that line directly, such as silicon, germanium, arsenic, antimony, and tellurium, show genuinely mixed properties of both categories, earning the name metalloids, or semi-metals.