Two ways to walk the block
The previous guide walked you left to right across the first transition series, watching the 3d shell fill one electron at a time and the highest oxidation state rise and then fall. That is the horizontal story. But the periodic table has a second dimension, and the transition metals behave strangely along it. Below each first-row metal sit two heavier relatives in the same group, called its congeners: below titanium come zirconium and hafnium; below chromium, molybdenum and tungsten; below iron, ruthenium and osmium; below nickel, palladium and platinum. Reading straight down a single column is the vertical story, and it has its own surprises.
In a main-group column your intuition usually serves you well: go down and the atoms get steadily bigger, heavier, and more metallic, each member a clearly inflated version of the one above. You might expect the transition metals to do the same — three triplets of ever-growing size. They do not. The second and third members of a transition group are almost the same size as each other, behaving nearly like twins, while the first-row member is the odd one out, the smallest and often the least like the other two. Knowing why is the key to navigating the lower two-thirds of the d-block, where most of the precious, catalytic, and industrially vital chemistry actually lives.
The lanthanide contraction: a hidden squeeze
The culprit is an event that happens off to the side, between the second and third rows of the d-block, called the lanthanide contraction. Before the 5d metals can begin, the fourteen lanthanide elements step in and fill their 4f subshell. Here is the crucial fact: 4f orbitals are shaped so that they pull in close to the nucleus yet, paradoxically, do a poor job of screening it. You met this idea earlier as shielding and penetration — an electron only blocks the nuclear pull felt by others if it spends its time between them and the nucleus, and diffuse, lobed 4f electrons largely fail to.
So as we march across the fourteen lanthanides, fourteen protons pile into the nucleus while the fourteen new 4f electrons that arrive alongside them barely cancel that extra pull. The net effect on every outer electron is a steadily rising effective nuclear charge, which reels the outer shells inward. The atoms quietly shrink — about ten picometres of radius lost across the series, an unusually large drop for filling a single subshell. By the time the 5d metals finally start, this accumulated shrinkage has almost exactly cancelled the size increase you would normally get from adding a whole new principal shell. The expected growth and the contraction nearly tie, and the 5d atom ends up no bigger than the 4d atom above it.
Twins you cannot tell apart: zirconium and hafnium
The cleanest demonstration sits at the top of the block in group 4. Zirconium is a 4d metal; hafnium, directly beneath it, is a 5d metal one full row lower. Naively hafnium should be distinctly larger. Instead the lanthanide contraction has shrunk it back so hard that the two atoms are almost exactly the same size — their metallic radii both fall near 159 picometres, and the Zr4+ and Hf4+ ions are likewise twins. Same size, same charge, same outer-shell behavior: their chemistry is virtually identical. They form the same oxides, the same halides, the same complexes, in the same proportions.
That near-identity has a real and expensive consequence. Almost every property we use to separate two elements — solubility, the temperature a salt crystallizes at, how strongly an ion binds an extractant — turns on size and charge. When two metals match on both, the usual separation tricks barely bite. Zirconium and hafnium always occur mixed in the same ore, and prising hafnium out of zirconium is one of the genuinely hard jobs of industrial chemistry, requiring patient solvent extraction or ion exchange repeated many times. It matters because nuclear reactors need zirconium that is hafnium-free: zirconium is nearly transparent to neutrons and makes good fuel cladding, while hafnium greedily absorbs them — chemically twins, but nuclear opposites.
Heavier metals climb higher and pair tighter
Going down a group does more than fix the size — it changes the chemistry in a consistent direction. The heavier 4d and 5d congeners hold high oxidation states far more comfortably than the 3d metal above them. Compare a column directly. Chromium(VI), as the dichromate and chromate ions, is a fierce oxidizer always itching to drop to the placid Cr3+. But tungsten(VI) in plain WO3 is calm and ordinary, perfectly happy to stay put. Manganese(VII) in permanganate is explosive in the wrong company, yet rhenium(VII) oxide, Re2O7, is a stable, unremarkable solid. Iron is dominated by +2 and +3, but its congener osmium reaches a startling +8 in OsO4. The rule is clear: down a group, the top of the oxidation-state ladder becomes a comfortable home rather than a fierce oxidizer.
Why does the heavier metal hold tight to so many shared electrons? Two linked reasons. First, the 4d and especially 5d orbitals are more spatially spread out than the compact 3d set, so they reach farther toward the ligands and form stronger, more covalent bonds; that extra bonding energy pays for stripping the metal up to a high state. Second, those same diffuse orbitals lower the energy cost of approaching neighbors. A practical handle: the descriptive chemistry of a heavy metal centers on its high oxidation states and on neutral or anionic oxo and halo species, whereas its 3d cousin lives mostly in modest +2 and +3 cation chemistry.
The same diffuse, far-reaching orbitals also widen the d-orbital splitting. Recall the spin contest from the crystal-field rung: an electron pairs down low if the splitting delta beats the pairing energy P. Because 4d and 5d orbitals overlap ligands more strongly, their delta is roughly fifty percent larger per row than the 3d value, while the pairing energy actually drops a little (the bigger orbitals let paired electrons keep their distance). Delta climbs, P falls, and the contest stops being a contest. The consequence: heavier transition-metal complexes are almost always low-spin, with electrons paired into the lower set and few or no unpaired spins. The high-spin/low-spin drama that makes 3d chemistry so colorful and magnetic is largely settled in advance for the 4d and 5d rows.
Metals that bond to each other: clusters
Here is the most striking gift of those big, diffuse orbitals. In ordinary first-row chemistry a metal ion sits at the center of its ligands and ignores its metal neighbors. But 4d and 5d orbitals reach far enough that two metal atoms can overlap their d orbitals directly and form a genuine metal-metal bond — atoms of the same element holding hands. Because each metal offers several d orbitals, they can form not just a single bond but double, triple, even quadruple bonds. The famous example is the [Re2Cl8]2- ion, where two rhenium atoms are held by a quadruple bond, a bond order higher than anything carbon can manage. The 3d metals do this only rarely and weakly; their compact orbitals simply cannot reach.
Bonding down a group, at a glance
3d (e.g. Cr, Fe, Ni)
compact d orbitals -> weak overlap, isolated ions
high-spin common, M-M bonding rare
4d / 5d (e.g. Mo/W, Ru/Os, Pd/Pt; Zr ~ Hf in size)
diffuse d orbitals -> strong overlap
larger delta -> almost always LOW-spin
high oxidation states stable (WO3, Re2O7, OsO4)
direct M-M bonds & clusters:
[Re2Cl8]2- : Re=Re quadruple bond
[Mo6Cl8]4+ : octahedral 6-metal cluster corePush this further and several metal atoms can lock together into a metal cluster — a small cage of metals bonded to one another and wrapped in ligands. The [Mo6Cl8]4+ core, an octahedron of six molybdenum atoms with chlorides bridging its faces, is a classic; tungsten, rhenium, and the platinum metals build a rich zoo of such cages. Clusters are not a curiosity: they are how we model the surfaces of metal catalysts, where bonds form across several metal atoms at once, and they sit at the heart of compounds we lean on industrially. This is also the structural reason the platinum-group metals resist corrosion and make such durable catalysts — strong metal-metal and metal-ligand bonding from far-reaching 5d orbitals.
Tying the vertical trends together
Step back and notice that one root cause feeds nearly everything in this guide. The 4d and especially 5d orbitals are larger and more diffuse than the 3d set, and the lanthanide contraction keeps the 4d and 5d atoms the same size so that their orbitals overlap a neighbor's reach almost interchangeably. From that single fact flow the near-twin chemistry of congeners, the comfortable high oxidation states, the strong fields that force low-spin, and the metal-metal bonds and clusters. The horizontal trends of the last guide came from filling the d shell; the vertical trends here come from how far those d electrons reach.
A couple of honest caveats keep the picture truthful. These are trends, not laws — there are exceptions in every group, and the early heavy metals differ from the late ones just as they do in the first row. Remember too that oxidation state remains a bookkeeping device, not the literal charge sitting on osmium in OsO4; and that calling a complex low-spin describes its electron arrangement, which is a thermodynamic and magnetic fact independent of whether the complex swaps its ligands quickly or slowly. With those guardrails in place, the down-the-group lens is one of the most reliable predictive tools in the whole d-block: tell me where a metal sits in its column, and I can already guess a great deal about how it will behave.