variable oxidation states

A sodium atom can only really give up one electron, so sodium is almost always +1; you never argue about it. The transition metals are completely different. Manganese turns up as +2 in pink salts, as +4 in the brown solid in old batteries, and as +7 in the deep purple of permanganate used to disinfect water. The same element wears several different oxidation states depending on what it is reacting with, and switching between them is the heart of its chemistry.

The reason lies in their electron arrangement. A first-row transition metal has its outer electrons split between the 4s and 3d subshells, and these two are very close in energy. After the 4s electrons (there are usually two) are removed, peeling off the 3d electrons one at a time costs only a little more each step, with no big jump. So a metal can lose two electrons, or three, or more, reaching whichever oxidation state a given reaction makes worthwhile — there is no single 'natural' charge it is forced into. As a rough guide, the highest possible state climbs across the early part of the row (matching the total of 4s plus 3d electrons available, peaking around manganese with +7) and then falls back, because pulling out that many electrons from a heavier nucleus becomes too costly. Remember that the oxidation state is a bookkeeping number — a formal way of counting which atom 'owns' the shared electrons — not the literal charge sitting on the metal.

Variable oxidation states are why transition metals are everywhere in batteries, catalysts, pigments and biology. A catalyst often works precisely because the metal can flip up and down between two states, accepting electrons in one step and handing them back in the next — iron does this in countless reactions, and so do the metals in the enzymes that run your body. The diagrams that organize all this (Latimer, Frost and Pourbaix) are tools for predicting which state will be stable under given conditions.