Atomic and Ionic Radii: Trends

Descending any single group of the periodic table, atomic radius increases steadily because a new principal shell is added at each step. Moving across a row of the d-block, however, the trend is quite different, because the electron being added at each step goes into an inner ($3d$) orbital rather than into the outermost shell.

The trend, in three phases. From scandium to chromium/manganese, the atomic (metallic) radius decreases fairly steadily, because each successive $3d$ electron only partially shields the correspondingly increased nuclear charge from the outer $4s$ electrons -- so the effective nuclear charge felt by the outer electrons rises, and the atom contracts. Through the middle of the series (roughly chromium to nickel), the radius becomes nearly constant, because from this point on, each additional $3d$ electron is added to an orbital that is already partly occupied, and the resulting increase in electron-electron repulsion within the $3d$ subshell very nearly cancels the pull of the increasing nuclear charge -- the two effects roughly balance. Finally, from nickel through copper and zinc, the radius rises slightly again, because the $3d$ subshell is now completely (or almost completely) filled, and the increased electron-electron repulsion among the ten paired $3d$ electrons begins to outweigh the (by now, only slowly increasing) nuclear attraction, pushing the outer electron cloud outward again.

Approximate metallic radii (in picometres) illustrating this fall-then-nearly-flat-then-slight-rise pattern are given for reference: $\text{Sc}\ 144 > \text{Ti}\ 132 > \text{V}\ 122 > \text{Cr}\ 118 \approx \text{Mn}\ 117 \approx \text{Fe}\ 116 \approx \text{Co}\ 116 \approx \text{Ni}\ 115 < \text{Cu}\ 117 < \text{Zn}\ 125$. Zinc, with its completely filled and comparatively poorly-shielding $3d^{10}$ subshell sitting beneath a filled $4s^2$ outer shell, is noticeably larger than its immediate neighbours. …