Case Study 2: Transition Metals and Rare Earths: Where Simple Rules Break Down

The Trouble with Rules

The Aufbau principle, the Madelung rule, and Hund's rules are powerful. They correctly predict the ground state electron configuration and term symbol for the vast majority of elements. But they are approximations to a complex many-body problem, and they break down in precisely the regions of the periodic table where physics becomes most interesting: the transition metals ($3d$, $4d$, $5d$ series), the lanthanides ($4f$), and the actinides ($5f$).

These breakdowns are not failures of quantum mechanics --- they are failures of simple rules applied to quantum mechanics. The full Hartree--Fock calculation (or more advanced methods like configuration interaction or density functional theory) handles these cases correctly. The exceptions teach us where our approximations reach their limits and what physics we have been neglecting.

The $3d$ Transition Metals: Chromium and Copper

The most famous exceptions to the Madelung rule occur in the first transition series:

Why Chromium Prefers $3d^5\,4s^1$

The explanation involves the exchange energy --- the energy lowering that occurs when electrons with parallel spins occupy different orbitals. The exchange energy is proportional to the number of pairs of parallel-spin electrons:

$$E_{\text{exchange}} \propto -K\binom{n_{\uparrow}}{2}$$

where $K$ is a positive exchange integral and $n_{\uparrow}$ is the number of spin-up electrons in the subshell.

For chromium, compare the two candidate configurations:

Configuration A: $3d^4\,4s^2$ --- four $3d$ electrons (assume all spin-up for maximum $S$) plus two paired $4s$ electrons. - $3d$ exchange pairs: $\binom{4}{2} = 6$

Configuration B: $3d^5\,4s^1$ --- five $3d$ electrons (all spin-up) plus one $4s$ electron. - $3d$ exchange pairs: $\binom{5}{2} = 10$

Configuration B has $10 - 6 = 4$ extra exchange pairs. If the exchange integral $K$ is large enough, this gain compensates for the energy cost of promoting a $4s$ electron to $3d$.

The energy balance is delicate: the $4s$ and $3d$ orbitals are nearly degenerate in the mid-transition series, and the exchange energy tips the balance. This is why the exception occurs at Cr (half-filled $d$) but not at V or Mn (where the exchange gain is insufficient to overcome the orbital energy difference).

The Copper Case

For copper, the comparison is:

Expected: $3d^9\,4s^2$ --- nine $3d$ electrons, exchange pairs $\binom{5}{2} + \binom{4}{2} = 10 + 6 = 16$.

Actual: $3d^{10}\,4s^1$ --- ten $3d$ electrons (completely filled), exchange pairs $\binom{5}{2} + \binom{5}{2} = 10 + 10 = 20$.

The gain of 4 exchange pairs, plus the additional stabilization energy of a completely filled $3d$ subshell (which has zero orbital angular momentum and is spherically symmetric, reducing electron--electron repulsion), makes the $3d^{10}\,4s^1$ configuration lower in energy.

The $4d$ and $5d$ Series: More Exceptions

The $4d$ transition metals have even more exceptions:

Palladium ($Z = 46$) is the most extreme case: both $5s$ electrons are "stolen" by the $4d$ subshell. This occurs because the $4d$ and $5s$ orbital energies are even more closely spaced than $3d$ and $4s$, making the energy balance more sensitive to correlation and exchange effects.

In the $5d$ series, relativistic effects become important. The $6s$ orbital contracts relativistically (its electron spends appreciable time near the heavy nucleus), lowering its energy. Meanwhile, the $5d$ orbitals expand because the contracted $6s$ electrons provide additional screening. The result is that $5d$ and $6s$ orbital energies are remarkably close, producing numerous exceptions and the rich chemistry of the platinum group metals.

The Lanthanides: Where the $4f$ Subshell Fills

The lanthanide series (La through Lu, $Z = 57$--$71$) presents the most confusing filling pattern in the periodic table. The $4f$ subshell begins to fill here, but the $4f$, $5d$, and $6s$ orbitals are so close in energy that configurations are difficult to predict from simple rules.

Why the Lanthanides Are Chemically Similar

The $4f$ orbitals are deeply buried inside the atom --- they have their maximum probability density inside the $5s$ and $5p$ shells. This means:

1. $4f$ electrons are poor screeners. They contribute little to the effective potential seen by the outermost ($6s$) electrons.
2. $4f$ electrons are chemically inaccessible. They do not participate directly in bonding. The valence electrons are the $5d$ and $6s$ electrons.
3. All lanthanides have essentially the same outer electron configuration: $5d^{0\text{--}1}\,6s^2$, with the $4f$ shell filling silently in the interior.

This is why all 15 lanthanides have nearly identical chemical properties, typically forming $+3$ ions (by losing the $5d$ and two $6s$ electrons, or three electrons from the $6s$/$5d$/$4f$ combination). Separating lanthanides from each other is notoriously difficult --- historically, it required hundreds of fractional crystallizations.

The Lanthanide Contraction

As the $4f$ shell fills from La to Lu, the nuclear charge increases by 14 units, but the $4f$ electrons are poor screeners of the $5d$ and $6s$ orbitals. The result is a steady increase in $Z_{\text{eff}}$ experienced by the outer electrons, producing a monotonic decrease in atomic radius across the series:

This cumulative contraction of 17 pm (about 17%) across the series has profound consequences for the $5d$ elements that follow. Hafnium ($Z = 72$, $5d^2\,6s^2$) has nearly the same atomic radius as zirconium ($Z = 40$, $4d^2\,5s^2$), despite having 32 more electrons. This makes Hf and Zr almost indistinguishable chemically, and the two elements were not separated until 1923.

The Actinides: Where Things Get Really Complicated

The actinide series ($Z = 89$--$103$) involves the filling of the $5f$ subshell, and the situation is even more complex than the lanthanides. The $5f$, $6d$, and $7s$ orbitals are nearly degenerate, and relativistic effects are substantial for these heavy elements. The configurations are irregular:

Unlike the lanthanides (which uniformly form $+3$ ions), the early actinides (Th through Am) exhibit a wide range of oxidation states: U can be $+3$ through $+6$, Pu shows oxidation states from $+3$ to $+7$. This is because the $5f$ orbitals are less deeply buried than $4f$ --- they extend farther from the nucleus and can participate in bonding. The $5f$ electrons are on the knife-edge between being localized (as in lanthanides) and being itinerant (as in transition metals).

This variability in oxidation states makes the actinides chemically richer than the lanthanides but also much more difficult to separate. The Manhattan Project required the separation of gram quantities of plutonium from uranium and fission products --- a chemical feat that was accomplished by Glenn Seaborg and colleagues using the newly discovered chemistry of the actinides.

Variable Oxidation States of Transition Metals

One of the most distinctive features of transition metals is their ability to form ions with multiple oxidation states. Consider manganese:

The colors arise from $d$--$d$ transitions --- electronic transitions within the partially filled $d$-subshell. In a free atom, all five $d$-orbitals have the same energy (for a given $n$). But in a crystal or molecular environment, the surrounding ligands create an electric field (crystal field) that splits the $d$-orbital energies. Photons with energies matching these splittings are absorbed, and the complementary color is observed.

The quantum mechanical foundation is clear: the variable oxidation states reflect the closely spaced orbital energies ($3d$ vs. $4s$), and the colors reflect the $d$-orbital splittings that occur in non-spherical environments. Neither phenomenon is remotely explainable without quantum mechanics.

What the Exceptions Teach Us

The exceptions to simple filling rules are not embarrassing failures --- they are informative windows into the physics of electron--electron interactions. Each exception reveals a competition between:

1. Orbital energy ordering (favoring lower $n + l$)
2. Exchange stabilization (favoring parallel spins and half-/fully-filled subshells)
3. Electron--electron repulsion (favoring configurations that minimize pairing)
4. Relativistic effects (contracting $s$/$p$ orbitals, expanding $d$/$f$ orbitals)

In light atoms, (1) dominates and the Madelung rule works. In transition metals, (2) competes with (1) and produces exceptions at half-filled and fully-filled $d$-subshells. In lanthanides and actinides, (3) and (4) become important and configurations become erratic.

The lesson is this: the Schrodinger equation (or the Dirac equation for relativistic atoms) is always correct. It is our approximation schemes that sometimes fail. When simple rules break down, it is not a crisis --- it is a signal that we need more sophisticated methods. The Hartree--Fock method, configuration interaction, and density functional theory are the tools that handle these cases correctly.

Questions for Reflection

1. Palladium ($Z = 46$) has the configuration $[\text{Kr}]\,4d^{10}$ with no $5s$ electrons. Does this mean the Madelung rule has completely failed? Or is this better understood as an extreme case of the exchange/correlation effect that produces the Cr and Cu exceptions?

2. Why are the lanthanides difficult to separate chemically? Connect your answer to the radial distribution functions of $4f$ orbitals.

3. The colors of transition metal compounds arise from $d$--$d$ transitions. Why are $s$-block and $p$-block compounds typically colorless? (Hint: consider the magnitude of orbital energy splittings.)

4. Uranium can exist in oxidation states from $+3$ to $+6$, while the analogous lanthanide neodymium (same position in $4f$ series) is almost exclusively $+3$. Explain the difference in terms of the spatial extent of $5f$ vs. $4f$ orbitals.

5. If all electrons were bosons (no Pauli exclusion, no exchange interaction), would there be any exceptions to the Madelung rule? Would the concept of "filling order" even make sense?