Chemistry 1.13 - 1.17 - Building the periodic table

Mendeleev: patterns and predictions

In 1869, Dmitri Mendeleev faced an incomplete set of elements and no knowledge of protons, neutrons or electrons. He put the known elements mainly in order of increasing relative atomic mass, but he did not let mass alone decide their positions. He compared observable properties of the elements and the properties of their compounds, placing chemically similar elements in the same vertical column.

This made Mendeleev's table more than a list:

• If strict mass order put an element beside others with very different properties, he changed the order so that the property pattern was preserved. Tellurium and iodine were a famous example: iodine belonged with other halogens because of its chemical properties.
• If a property pattern required an element that was not known, he left a gap instead of forcing the wrong element into it.
• He used the elements around a gap, especially those in the same column, to predict the missing element's existence and its physical and chemical properties. The later discovery of gallium, with properties close to those he had predicted for "eka-aluminium", provided evidence that the pattern was useful.

Mendeleev thought his table increased in relative atomic mass, but some pairs did not follow that order. We can now explain why. Relative atomic mass is a weighted mean: it depends on both the masses of an element's isotopes and their relative abundances. For example, argon has atomic number 18 and relative atomic mass about 39.95, while potassium has atomic number 19 and relative atomic mass about 39.10. The mass value decreases even though the proton number increases because the two elements have different isotope mixtures.

Mendeleev did not know about isotopes, so he could not give this explanation. His decision to trust repeating chemical properties over a perfectly smooth mass order was nevertheless scientifically productive: later evidence explained the apparent exceptions.